Protecting People and the EnvironmentUNITED STATES NUCLEAR REGULATORY COMMISSION
Official Transcript of Proceedings
NUCLEAR REGULATORY COMMISSION
Title: Advisory Committee on Reactor Safeguards
Joint Meeting of the Subcommittees on
Materials and Metallurgy, Thermal-Hydraulic
Phenomena, and Reliability and Probabilistic
Risk Assessment
Docket Number: (not applicable)
Location: Rockville, Maryland
Date: Friday, March 16, 2001
Work Order No.: NRC-110 Pages 1-239
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)
+ + + + +
JOINT MEETING OF THE ACRS SUBCOMMITTEES ON
MATERIALS AND METALLURGY, THERMAL-HYDRAULIC
PHENOMENA, AND RELIABILITY AND PROBABILISTIC RISK
ASSESSMENT
+ + + + +
FRIDAY,
MARCH 16, 2001
+ + + + +
ROCKVILLE, MARYLAND
+ + + + +
The Subcommittee met at the Nuclear
Regulatory Commission, Two White Flint North, Room
T-2B3, 11545 Rockville Pike, at 8:30 a.m., William J.
Shack, Chairman of the Materials and Metallurgy
Subcommittee, presiding.
COMMITTEE MEMBERS:
WILLIAM J. SHACK, Chairman, Materials and
Metallurgy Subcommittee
COMMITTEE MEMBERS: (cont'd)
GRAHAM B. WALLIS, Chairman, Thermal-Hydraulic
Phenomena Subcommittee
GEORGE APOSTOLAKIS, Chairman, Reliability and
Probability Risk Assessment Subcommittee
MARIO V. BONACA, Member
THOMAS S. KRESS, Member
JOHN D. SIEBER, Member
I-N-D-E-X
AGENDA ITEM PAGE
Introduction . . . . . . . . . . . . . . . . . . . 4
NRC Safety Research Program. . . . . . . . . . . . 5
Industry Presentation. . . . . . . . . . . . . . .69
NRC Staff Presentation . . . . . . . . . . . . . 155
ACRS General Discussion and Adjournment. . . . . 220
P-R-O-C-E-E-D-I-N-G-S
(8:29 a.m.)
CHAIRMAN SHACK: The meeting will now come
to order.
This is a joint meeting of the Advisory
Committee on Reactor Safeguards, Subcommittees on
Materials and Metallurgy, Thermal-Hydraulic Phenomena,
and Reliability and Probabilistic Risk Assessment.
I am William Shack, Chairman of the
Subcommittee on Materials and Metallurgy. Graham
Wallis is Chairman of the Subcommittee on Thermal-
Hydraulic Phenomena. And George Apostolakis is
Chairman of the Subcommittee on Reliability and PRA.
Subcommittee members in attendance are
Mario Bonaca, Thomas Kress, and Jack Sieber.
The purpose of this meeting is to discuss
the status of risk-informed revisions to the technical
requirements of 10 CFR 50.46 for emergency core
cooling systems. The subcommittees will also discuss
the proposed final report on the NRC Safety Research
Program. The subcommittees will gather information,
analyze relevant issues and facts, and formulate
proposed positions and actions, as appropriate, for
deliberation by the full committee.
Michael T. Markley is the cognizant ACRS
staff engineer for this meeting.
The rules for participation in today's
meeting have been announced as part of the notice of
this meeting previously published in the Federal
Register on March 1, 2001, and later amended to
provide for discussion of the report on the NRC Safety
Research Program.
A transcript of the meeting is being kept
and will be made available as stated in the Federal
Register notice. It is requested that speakers first
identify themselves and speak with sufficient clarity
and volume so they can be readily heard.
We have received no written comments or
requests for time to make oral statements from members
of the public regarding today's meeting.
For those who came to attend the 50.46, we
will be taking about an hour. That will -- we'll
start the discussion of 50.46 at about 9:35.
At the moment, I'll turn it over to George
Apostolakis to discuss the Safety Research Report.
MEMBER APOSTOLAKIS: Thank you, Bill.
As the members know, we have a new version
of the report to the NRC on the Reactor Safety
Research Program, and we would like to discuss some of
the outstanding issues today and possibly vote on it.
So, Dr. Kress, would you lead us, please,
through this?
MEMBER KRESS: Yes. Members should have
a handout, the new draft version, which fortunately
now has page numbers on it. And there are a number of
items of disagreement or contention, and what I'm
proposing we do is you have a sheet -- I think you
should have a sheet that looks like this. These are
the page numbers where those items are.
There's only really a few of them, but
some of them are more contentious than others. And my
proposal is that we take up the two most contentious
ones first, and see if we can discuss it and come to
some sort of agreement. That would be pages 11, 12,
13, and 15.
MEMBER APOSTOLAKIS: Okay. So you are
skipping the others?
MEMBER KRESS: Well, not yet. But we'll
come back to them.
MEMBER APOSTOLAKIS: Oh, okay.
MEMBER KRESS: I just want to start with
these --
MEMBER APOSTOLAKIS: All right.
MEMBER KRESS: -- because I think the
others are probably relatively easy.
MEMBER APOSTOLAKIS: Okay. So you're
saying 11?
MEMBER KRESS: Pages 11, 12, 13, and shut
down --
MEMBER APOSTOLAKIS: Okay.
CHAIRMAN SHACK: This whole thing? We
have two versions.
MEMBER APOSTOLAKIS: It is an alternate to
what. To the previous paragraph? The one on page 10?
MR. EL-ZEFTAWY: The one on 12 was a
recommendation to scratch the whole section of the
standard for PRA.
MEMBER APOSTOLAKIS: Oh, oh, wait a
minute. Wait a minute. So up to page 10, line 198,
there is nothing, there is no change.
MR. EL-ZEFTAWY: Right.
MEMBER APOSTOLAKIS: So then you have one
paragraph versus the standard for the PRA.
MR. EL-ZEFTAWY: Right. The one that's
starting in -- on page 11, that's a new paragraph.
MEMBER APOSTOLAKIS: So line 199, line
220, those would be the alternatives.
MR. EL-ZEFTAWY: Right. And then you have
to decide if you're going to scratch from 220 all the
way to 236.
MEMBER APOSTOLAKIS: Yes, and adopt 199
through 218.
MR. EL-ZEFTAWY: Correct.
MEMBER KRESS: That's the issue.
MEMBER APOSTOLAKIS: Okay. Okay.
MEMBER KRESS: Now, should we let members
have time to read both of these first and then discuss
them?
MEMBER WALLIS: Well, what's that page?
It seems to be the one from --
MEMBER KRESS: That's the problem. The
alternate has a much different thought process and
much different than the other one.
CHAIRMAN SHACK: Well, actually, what I
proposed for the alternate was it would replace the
lines from 184, starting on the risk management tools
used by industry, through 198, so it's --
MEMBER APOSTOLAKIS: See, that's what I
thought.
CHAIRMAN SHACK: It's an alternate that
replaces a fair chunk of that paragraph, rather than
an addition.
MEMBER KRESS: So that replaces 184
through 198.
MEMBER WALLIS: Well, that makes some
sense but that really changed the subject.
MEMBER APOSTOLAKIS: Yes. I thought the
issue of standards for PRAs is an entirely different
--
CHAIRMAN SHACK: That's a different --
yes, it just happens to come together here, but --
MEMBER APOSTOLAKIS: Oh, okay. So the
alternate, then, in our report would replace the risk
management tools. Because I thought when I saw the e-
mail that this was really a rephrasing of that --
CHAIRMAN SHACK: Right. It's a rephrasing
of the 184 to 198 section.
MEMBER APOSTOLAKIS: Well, I'm for the
alternate.
CHAIRMAN SHACK: I would also just suggest
that from reading last night the 177/179, I would take
out the "It can be argued that licensees are
adequately managing risk during planned outages." And
just go directly to "The nuclear industry has made
substantial efforts."
MEMBER APOSTOLAKIS: Wait a minute.
MEMBER KRESS: I think I would support
that.
MEMBER APOSTOLAKIS: Where is that?
CHAIRMAN SHACK: It's on 177.
MEMBER KRESS: Now, let me tell you --
MEMBER APOSTOLAKIS: And then you would go
directly to where?
CHAIRMAN SHACK: I would take that
sentence out and then take off the introductory
phrase, "It is certainly true that," and just say,
"The nuclear industry has made substantial" --
MEMBER KRESS: And start there. But let
me -- now, I was against this change.
CHAIRMAN SHACK: Completely.
MEMBER KRESS: Completely. That was the
word. And let me tell you why. The major message in
these lines 184 down can be found in lines 184 and 185
and 186 and part of 187. And that major message has
been done away with, and that's a message I think
needs to be said because it -- to me, it is the -- it
highlights the major difference between what the
industry does and needs and what the regulatory agency
does and needs.
So I didn't want to lose that message,
which the alternate -- the alternate proposal loses
that.
CHAIRMAN SHACK: Deliberately.
MEMBER KRESS: Deliberately, of course.
MEMBER WALLIS: Why did you want to lose
it?
CHAIRMAN SHACK: Because I personally
think that, you know, we've made that argument in our
'99 letter, that there was this distinction between
what the NRC was interested in and what the licensee
was interested in. I think that's a false
distinction. You know, if we're out here to just --
to computer this number, I don't think that's
important.
MEMBER KRESS: It's not a false
distinction, because there are two reasons for having
shutdown risk assessments. One of them is to manage
that shutdown risk, as it is ongoing, and that -- the
tools for that are what the industry uses and they're
good tools. NRC needs to know about those tools and
needs to be able to do that also.
But that doesn't help them at all when
they go to risk-inform the regulations. Those tools
tell them nothing. It just gives them a little -- a
few insights. This tells them nothing about the risk
contribution of shutdown. That's the problem.
And if you want the risk contribution due
to shutdown to factor in to your risk-informing the
regulations, you have to have this difference. And
it's a different animal. It's not easy to do. And
that's the reason it hasn't been done; it's not easy.
And that's why some research is needed.
But that's my whole problem. If you want
to risk-inform the regulations, you have to have this
component in there. And that's my problem.
CHAIRMAN SHACK: I mean, I will argue that
the reason I want to do shutdown risk studies is I
want to be able to make the statement that we made
that there are unlikely to be any major contributors
to risk that have not been identified that we can make
about normal operations. I don't think we can make
that statement about shutdown risk. I don't think we
can make it about fire risk.
MEMBER KRESS: I don't mind making that
statement also. I just don't want to lose this
distinction, though, because to me it's the major
distinction. I wouldn't mind adding --
CHAIRMAN SHACK: Being the major one is,
in fact, that you want the assurance that you've
identified to measure the contributors to this.
MEMBER KRESS: Well, I think this is
equally important. I think they have equal
importance.
CHAIRMAN SHACK: That's why we completely
disagree.
MEMBER WALLIS: I don't understand really
why there's so much disagreement. Can't you retain
the sentence -- two sentences that Tom would like to
retain? Is there something offensive about them?
CHAIRMAN SHACK: What's offensive about
it, of course, is it was stuffed the last time we sent
it up.
MEMBER KRESS: Well, you know, that
shouldn't be any reason why we -- we shouldn't give
good advice.
CHAIRMAN SHACK: I think it puts the wrong
emphasis on why you're doing it.
MEMBER APOSTOLAKIS: I don't see the two
points of view being very different, and I don't see
why you would have to say which one is more important
than the other, although I tend to agree with Bill.
I think, you know, if there is a question of not
knowing of some vulnerability, that is really the most
important thing you would like to know about it.
But why do we have to say what's more
important?
MEMBER KRESS: Well, I don't think you
have to. I would be in favor of keeping both
sentences. I just don't want to lose this thought.
MEMBER WALLIS: I think you could keep one
of these, you can just take it on the end of that
section. It's not incompatible, is it?
MEMBER KRESS: That would suit me.
MEMBER APOSTOLAKIS: Okay. So somebody
will do that?
MEMBER WALLIS: Tom will do that. Tom has
dictatorial authority.
MEMBER APOSTOLAKIS: Editorial or
dictatorial?
MEMBER WALLIS: It's one of those Latin
terms that you would have to -- the concept is
probably unknown in --
MEMBER KRESS: Okay. Now, that sort of
gets us on the road for maybe resolving that one. The
next one is this section on -- starting on line 219.
And I think the proposal there was to just zap that
section all together. That's the question. Do we
want to zap that section?
MEMBER APOSTOLAKIS: Well, since I
proposed that, it's not that I'm against the agency
spending resources to support the development of
standards for PRA. It's just that I thought that this
is something that the agency has committed to do.
It's something that they will do. And I don't view
that as research. So I didn't think it belonged
there, but I didn't really --
CHAIRMAN SHACK: Well, there is research
in lines 231, 232, 233, where you essentially define
necessary features of PRAs.
MEMBER APOSTOLAKIS: Well, but that's part
of this activity. This is not where we talk about
necessary and sufficient, is it?
MEMBER KRESS: No. We'll get to that.
MEMBER WALLIS: I thought we ought to
retain it, because this is a major issue with the --
it's on the Commission's radar screen. They are
worried about PRA quality and how to respond to
critics. The quality is so lousy you can't use it,
and --
MEMBER BONACA: We can put in a statement,
George, that affects your point of view. This has
already been dealt with in large part. We recognize
that. But it's as important --
MEMBER KRESS: Well, let me tell you what
my take is on this. My take is is the industry and
the agency are on divergent courses. Industry is
going out to -- through a certification process to
certify the plant-specific PRAs. And the agency is
going forth with this development of standards.
And I see the two as somewhat similar but
not completely compatible. And what I think will
happen is the agency will have this set of standards
to look at, but the industry will come in for some
request for an exemption or a change or rule --
rulemaking or whatever, and the plant that comes in
will bring his certified PRA certified by the industry
process.
And the staff will be sitting there with
another whole set of standards. And they will have to
somehow reconcile the two, and that's what I'm asking
for here is to give some thought to how they're going
to reconcile the two and see if there is some
relationship between the standards and the
certification, and maybe even adopt one or the other
or both of them, show how they're related to each
other.
So I thought we needed a section on
standards to deal with what I see as an upcoming
issue. And that's why I didn't want to zap this
section.
CHAIRMAN SHACK: But I thought they
already had a program in place to --
MEMBER KRESS: Yes, they have --
CHAIRMAN SHACK: -- at how good -- you
know, whether the peer review was a grade 3 -- is
adequate for Option --
MEMBER KRESS: Well, I'm sorry, I didn't
see --
MEMBER APOSTOLAKIS: And also, didn't the
staff also report some time ago where they had the
necessary features? That was really nice, where they
also had Appendix B and we recommended that they
expand the --
MEMBER KRESS: They had necessary but not
sufficient.
MEMBER APOSTOLAKIS: But it's not
sufficient here either. In fact, it says sufficiency
is very difficulty. So all of these things either
have been done or are in the process of being
completed. That's all.
MEMBER KRESS: Well, you know, everything
we talk about is -- I don't know why this one should
be different.
MEMBER WALLIS: They don't want to --
MEMBER APOSTOLAKIS: I don't see it as
research. But, anyway, I'm not going to --
MEMBER KRESS: Well, research is a good --
has got a broad envelope in this agency.
MEMBER APOSTOLAKIS: All right. Well,
that's fine with me. We can keep it. I didn't feel
strongly about it. It's just that I thought it was
something that was being done anyway.
But this also is nice because it says
clearly that you cannot define "sufficiency"
requirements, which we will remember a little bit
later.
MR. EL-ZEFTAWY: That word has been taken
out, the 242 -- line 242. When it comes to
sufficient, we took the word "sufficiency" --
MEMBER APOSTOLAKIS: Where is 242? That's
on -- what do you mean?
MR. EL-ZEFTAWY: But, I mean, originally
we had "sufficient and necessary," and we took the
word "sufficient" out.
CHAIRMAN SHACK: On 242, it was "necessary
and sufficient features of probabilistic risk
assessment" once upon a time. That -- we nailed that
--
MEMBER APOSTOLAKIS: Okay. So that's an
old letter.
CHAIRMAN SHACK: Yes.
MEMBER WALLIS: So as long as the PRA
methods are insufficient, that would be okay?
CHAIRMAN SHACK: Yes.
MEMBER APOSTOLAKIS: Okay. So we could --
so we keep that.
MEMBER KRESS: So we keep this, and that
resolves that issue. Well, we're making headway.
These other things I think --
CHAIRMAN SHACK: Well, I think there's a
question why George wants SPAR out.
MEMBER KRESS: Yes, that's on 13.
MEMBER APOSTOLAKIS: Because it's --
really, the whole thing addresses the issue of codes.
SPAR is a model. I mean, they are taking the IPs and
putting them on SAPPHIRE, right? Is that what SPAR
is, essentially? No?
MR. KING: Well, it's not taking the IPs.
It's taking our own models, which in many cases are
better than what the IPs had.
MEMBER APOSTOLAKIS: Right. Right.
MEMBER KRESS: But they are in a sense
PRAs. They're very --
MR. KING: No, plant-specific now. We've
got -- we're developing SPAR models for each plant.
MEMBER KRESS: But they could not --
MEMBER APOSTOLAKIS: But the whole point
of this paragraph was to address the need to peer
review the fundamental tool of SAPPHIRE. Now, whether
you use SAPPHIRE to do other things, I mean, I would
--
CHAIRMAN SHACK: Well, I thought it was
more risk assessment tools. I was going to suggest
changing lines 241 and 242 to read, "An agency effort
to define the kinds of risk assessment tools needed to
support regulatory processes might well provide the
agency" --
MEMBER APOSTOLAKIS: In my mind, the
important recommendation is 244. The SAPPHIRE code
has reached a stage of development that the public
deserves to see a comprehensive peer review of this
code. This is the message here.
MEMBER KRESS: Yes.
CHAIRMAN SHACK: But I think Dana has been
adamant that they need better risk assessment tools,
and so I think he would look at both the --
MEMBER APOSTOLAKIS: But we say that
somewhere else. This is not the place. This is -- he
doesn't discuss SPAR. He just mentions them in
passing.
MEMBER KRESS: That's the only place in
here we say anything about SPAR.
CHAIRMAN SHACK: Well, I mean, you might
say we ought to say more about SPAR, but I'm not sure
we should argue we should say less.
MEMBER BONACA: I agree with that.
MEMBER APOSTOLAKIS: Look at the first
sentence. "The NRC risk assessment codes and models
continue to undergo development, and the vision of
this code" --
CHAIRMAN SHACK: These codes and models.
MEMBER KRESS: I think it's --
MEMBER APOSTOLAKIS: I think it dilutes
it. I think it really is the SAPPHIRE thing that
needs the review.
MEMBER KRESS: Do you want to take us --
MEMBER APOSTOLAKIS: I mean, we're asking
them to do a peer review of a SPAR model?
CHAIRMAN SHACK: No. No, no. All we're
saying is they should continue the development, and
we'd just like a little better, more organized picture
of what they really intend to get to with the SPAR
models. How good do they need to be? How good are
they? How good --
MEMBER APOSTOLAKIS: Where does it say
that? It doesn't say that?
CHAIRMAN SHACK: Well, it's -- if we say
an agency -- you have to define the kinds of risk
assessment tools needed to support regulatory process
-- might well provide the agency with a more scrutable
strategy for the development of these models and
codes.
MEMBER KRESS: Yes, I wouldn't be against
that.
CHAIRMAN SHACK: You know, changing the
words a little bit.
MEMBER APOSTOLAKIS: Sometimes asking for
too much means you are getting nothing back. I think
if you have a specific recommendation, take this code
and peer review it. It's very hard to say, "There are
ways around it." Developing better models, yes, we
are development better models. What do you want? I
think it dilutes the message. The message is breaks
of power are so important to have some sort of peer
review to -- I don't care about the --
MEMBER KRESS: Let's take the -- the
proposal is to remove the word "SPAR" from here and
just have this paragraph focus specifically on
SAPPHIRE. Those in favor of that, please raise your
right hand. Those opposed? So it passed three to two
to -- so we're going to remove that "and SPAR," and
this paragraph is just going to --
MEMBER APOSTOLAKIS: Now, what are we
going to do about line 241? The necessary features of
probabilistic risk assessment support -- are you still
leaving that there? I think it's okay to leave it.
I mean --
MEMBER WALLIS: Let's leave it.
MEMBER APOSTOLAKIS: -- it's sort of a
model for --
MEMBER KRESS: Let's leave it. Let's
leave it.
MEMBER APOSTOLAKIS: Okay.
MEMBER BONACA: It's mentioned as part of
the -- from your report?
CHAIRMAN SHACK: I believe so.
MR. EL-ZEFTAWY: So it's even from line
237.
MEMBER APOSTOLAKIS: Yes, the SAPPHIRE
code.
MR. EL-ZEFTAWY: Okay.
MEMBER APOSTOLAKIS: Right.
MEMBER KRESS: Okay. That takes care of
that problem. Now, where should we go? Let's see
what's on page 25. Let's look at 15. That's the next
one. That's the next contentious issue. That's the
quantification of uncertainties.
MEMBER APOSTOLAKIS: I was the one getting
kind of --
MEMBER WALLIS: No, it's very important
that we do that.
MEMBER APOSTOLAKIS: Well, my point was
that this is also too dilute. Let's --
MEMBER WALLIS: Be stronger about it?
MEMBER APOSTOLAKIS: We worry about
uncertainty, and so on. I think our message is that
in the context of these new thermal-hydraulic codes,
we'd like to see statement of model uncertainty.
MEMBER KRESS: This had to do with PRA.
MEMBER WALLIS: We'd like uncertainty
evaluated everywhere, including PRA, as a separate --
MEMBER APOSTOLAKIS: Well, I mean, if you
do that there, that's PRA.
MEMBER WALLIS: Yes, well, this is a PRA.
We're talking about PRA.
MEMBER APOSTOLAKIS: And what they say in
the PRA context, "Please address the issue of
uncertainties and quantify them," again, is a
motherhood statement. Because PRA is supposed to do
that. If you don't do an uncertainty calculation, you
are not doing a PRA. So I thought the message was
clearer in the other sections.
MEMBER KRESS: Yes. But what this makes
a point is that the uncertainty development in PRAs is
mostly epistemic and they don't deal with the
aleatory --
MEMBER APOSTOLAKIS: And they don't know
that?
MEMBER KRESS: Well, we're saying here the
only place you have both of those combined is in the
NUREG 1150. And you need to somehow separate the two
out and use some generic measure of the aleatory and
let the codes go ahead and develop the epistemic. But
at the end you add the two together some way.
You have to deal with both uncertainties
some way in your development. Either -- either you do
it in your -- how you make your decision using the
results, or you make an assessment of them some way.
But it says you need to deal better with
uncertainties.
And it also makes an interesting point
that -- to be careful somewhat with the Bayesian
process because it knocks off details. And that's an
important message, too.
MEMBER APOSTOLAKIS: I wrote that.
MEMBER KRESS: You wrote that?
MEMBER APOSTOLAKIS: But I'm proposing to
eliminate it, because I think it's --
MEMBER KRESS: No, no, it's an important
message.
MEMBER WALLIS: George, it's not as if
they don't know it. If they know it, we're simply
reinforcing it. And the fact that we wrote it in our
report will help them. So --
MEMBER KRESS: The staff probably knows
99 percent of everything that --
MEMBER APOSTOLAKIS: Can we get -- well,
it's not a matter of that. Well, one thing we don't
do right now is send the reader to other sections
where similar things are discussed. Like here I think
it would be very appropriate on line 301 if we
actually send them to the thermal-hydraulic section.
MEMBER WALLIS: That would be fine. That
would be fine.
MEMBER APOSTOLAKIS: That would make me
happy.
MEMBER WALLIS: That would be fine.
MEMBER KRESS: I wouldn't be opposed to
that.
MEMBER APOSTOLAKIS: Yes. I mean, this
thing about -- as you know, we sent many e-mails to
Dana back and forth. When I discussed this issue of
updating the distributions it was in a very different
context.
MEMBER KRESS: Yes.
MEMBER APOSTOLAKIS: So I thought it was
a little bit out of the blue. But it's okay. I mean
--
MEMBER KRESS: Well, it fits you. I guess
the word fits. So we'll retain this, and maybe add a
sentence at the end that refers to the thermal-
hydraulic section that deals with this same issue.
MEMBER APOSTOLAKIS: Now, if we retain it,
look at 294. "Uncertainness in the models used for
the analysis are seldom discussed." Is that correct?
I don't think it's correct.
MEMBER KRESS: That one we might want to
change.
MEMBER APOSTOLAKIS: They may not be
quantified.
MEMBER WALLIS: We should take that out.
MEMBER APOSTOLAKIS: But they are
certainly discussed. So let's take that out.
MEMBER KRESS: I thought we could deal
with those kinds of things later. Well --
MEMBER APOSTOLAKIS: Also, on 293, of
propagating parameter uncertainties, we don't need the
word "epistemic" there.
MEMBER KRESS: 293? Yes, you're right.
It's redundant.
MEMBER APOSTOLAKIS: Now --
MEMBER KRESS: Besides, anywhere I can
mark out the words "epistemic" and "aleatory" I am
willing to do --
(Laughter.)
MEMBER APOSTOLAKIS: 291. Yet careful
quantification of the -- it's not -- there's a typo
there.
MEMBER KRESS: Yes, there are a couple of
typos.
MEMBER APOSTOLAKIS: Yes. Of
uncertainties seldom appears in risk-informed
regulatory discussions. Do we all agree with that?
MEMBER KRESS: Well --
MEMBER APOSTOLAKIS: Quantification. Yes,
probably right. Quantification is correct. The
discussion was incorrect.
Okay. So just make sure that at the end
we put "see also Section 6." Go to the end of the
paragraph, and --
MEMBER WALLIS: Is it only in Section 6
that we want to --
MEMBER APOSTOLAKIS: Well, that's where
the thermal-hydraulic is. Is there another place? I
think that's the main place.
MEMBER KRESS: I think that's the main
place it's in here.
MEMBER APOSTOLAKIS: See Section Roman
II.6. Thank you. Okay. Let's keep it.
MEMBER KRESS: Good. Let's go to page 25.
That's another simple one. Oh. The suggestion was to
delete that -- what should be the -- at the end of
that. I oppose that suggestion because --
MEMBER APOSTOLAKIS: Well, isn't the
second bullet -- sub-bullet -- where is it? Oh.
Asking the same thing? When is human performance at
the nuclear plant good enough? In fact, I prefer that
than what it should be, and that was what dictating --
the human error contribution should be 30 percent. So
that's why I proposed to take it out. Personally --
MEMBER KRESS: I think you're right,
George.
MEMBER BONACA: I think so.
MEMBER KRESS: I think you're right.
Let's zap that.
MEMBER APOSTOLAKIS: You win a few, you
lose a few.
MEMBER KRESS: Yes.
MEMBER APOSTOLAKIS: You lose a lot, you
win a few.
(Laughter.)
No, no. No. Delete, "What should they
be?" The red.
MEMBER KRESS: Okay.
MEMBER APOSTOLAKIS: Then, you want us to
go to 33?
MEMBER KRESS: Yes, we might as well go
right down the line here. Then we'll get back to six,
eight, and nine.
MEMBER APOSTOLAKIS: So this is alternate
now to what again?
CHAIRMAN SHACK: Hold on a second.
MEMBER APOSTOLAKIS: To all the bullets?
CHAIRMAN SHACK: Yes. Yes.
MR. EL-ZEFTAWY: Yes. It's all the --
MEMBER KRESS: The suggestion was to zap
out all of --
MR. EL-ZEFTAWY: Actually, it's line 631.
MEMBER KRESS: Yes, and replace it with
this tiny little sentence. Yes. And I think Bill
Shack could -- could discuss why he thinks this is a
good idea maybe.
CHAIRMAN SHACK: Okay. I just felt that
basically you had better -- there were better points
made in the discussion of the specific topics than
there were here. I mean, these bullets didn't really
-- couldn't even make the case. You know, whether the
model is -- doesn't have the technical sophistication
that you have at NIST is not really the question. The
question is, is it good enough?
I don't know that it's, you know,
specialized activities that can't be done by the
regional staff and require -- you know, it does --
this didn't strike me as very forceful arguments for
why I needed research. I had much more forceful
statements I thought in the discussion of the specific
tools. And so I thought it actually strengthened the
argument to get on with it.
MEMBER KRESS: And my feeling was that
these are relatively true statements, all of them --
the bullets. So it didn't hurt much to leave them in
to set in -- give a context for the -- it didn't hurt
to leave them in. And Dana -- they were close to
Dana's heart, and so my feeling was it -- if it didn't
hurt to leave them in, why not just leave them in?
CHAIRMAN SHACK: Well, for example, in
631, this his what -- you know, we argue about when
the staff should be doing things and when the industry
should be doing things. It's not at all clear to me
that this is a -- you know, that it's something that
shouldn't be done by the industry, for example.
MEMBER KRESS: Well, if you're going to
risk-inform it -- the process, then it's something
that the agency ought to do.
CHAIRMAN SHACK: I think the industry
should be -- you know, required to analyze
consequences of the accidents.
MEMBER KRESS: Maybe what you -- how you
risk-perform it is --
CHAIRMAN SHACK: Well, to facilitate the
circuit analysis sounds to me like a licensee --
MEMBER BONACA: You know, I don't think,
though, the bullets here are created equal. I think
we should look one by one, because some of them, for
example, I agree to retain. That first one -- it's an
important observation. I think the fact that, you
know, the NRC to have the technical sophistication of
-- developed by -- it's an observation of --
MEMBER KRESS: I think we've used up our
hour. Should we defer this to --
CHAIRMAN SHACK: No, we've got until 9:30.
We've got half an hour.
MEMBER KRESS: Oh, do we?
MEMBER BONACA: Yes.
MEMBER KRESS: Okay. Thank you.
CHAIRMAN SHACK: We're doing good.
MEMBER BONACA: Yes. The second bullet,
for example, I could do without. I mean, so what's
the problem? If you need to obtain a specialist for
the important stuff, I mean --
MEMBER KRESS: I wouldn't want to be
without that second bullet.
MEMBER BONACA: What?
MEMBER KRESS: I wouldn't want to be
without that second bullet. Neither would Dana.
MEMBER SIEBER: Neither would I.
MEMBER KRESS: I've got two votes. If I
raise my left hand, it's Dana. If I raise my right
hand, it's --
MEMBER BONACA: No. I mean, I think we
should walk through the bullets now and discuss --
MEMBER KRESS: Yes.
MEMBER BONACA: -- them all. I think some
of them I agree with and some of them I don't agree
with.
MEMBER KRESS: Well, I certainly wouldn't
want to get rid of bullet number three. I might be
willing to get rid of bullet number four.
CHAIRMAN SHACK: Okay.
MEMBER KRESS: And five. I would like to
retain the first three bullets and get rid of --
MEMBER BONACA: I can go with that.
MEMBER APOSTOLAKIS: How many?
MEMBER KRESS: The first three.
MEMBER APOSTOLAKIS: The first two?
MEMBER KRESS: The first three.
MEMBER APOSTOLAKIS: And replace them by
the alternate or just --
MEMBER KRESS: Well, the alternate might
be something we'd want to replace them with. No, no,
the alternate would -- we said that to some extent in
the --
MEMBER APOSTOLAKIS: Right. So eliminate
the first three bullets.
MEMBER KRESS: Yes. No, the last two.
(Laughter.)
MEMBER APOSTOLAKIS: Wait a minute. Wait
a minute. You are eliminating --
CHAIRMAN SHACK: The last two.
MEMBER APOSTOLAKIS: The significance of
the termination process?
MEMBER KRESS: Yes. But read the whole
thing, George. I mean, what -- I mean, I don't think
that's very helpful to the --
MEMBER APOSTOLAKIS: Right, right, right.
MEMBER KRESS: You know, the important --
MEMBER APOSTOLAKIS: I would say it is
based on evaluations that are not at all transparent
to the public.
MEMBER KRESS: Well, I would probably
agree with leaving it in if you --
MEMBER APOSTOLAKIS: Yes.
CHAIRMAN SHACK: Transparency to the
public is --
MEMBER APOSTOLAKIS: Or transparent,
period.
CHAIRMAN SHACK: Until we've gone through
the STP, I'm -- you know, that's sort of my thing, is
that we --
MEMBER KRESS: We haven't really reviewed
the --
CHAIRMAN SHACK: -- we haven't reviewed
this.
MEMBER APOSTOLAKIS: Well, but in that
spirit, have you really reviewed the NIST code? And
do you know that it's much better than --
CHAIRMAN SHACK: Hey, I voted for
eliminating all of the bullets myself.
(Laughter.)
MEMBER APOSTOLAKIS: I think something
about the STP is important. I mean, you can say they
are not at all transparent, but, you know --
CHAIRMAN SHACK: It would be helpful if I
had read it and I knew what it was.
MEMBER APOSTOLAKIS: It would have been
helpful, yes.
MEMBER WALLIS: I think Dana has read it.
CHAIRMAN SHACK: Yes. But my comment was
that I think Dana may be the only one that has read
it. And, you know, this is a committee position.
MEMBER BONACA: This could be a true
statement and transparent -- and much of the
regulations are transparent to the public.
CHAIRMAN SHACK: right.
MEMBER APOSTOLAKIS: So why would we say
"not at all transparent to the public"? But it
doesn't matter? Because the other regulations are the
same way.
MEMBER KRESS: Yes, that's my point. I
don't think it has to be transparent.
MEMBER WALLIS: Well, I'd be happy to
remove it. We don't really seem to be certain that we
want to say --
MEMBER KRESS: Let's zap those two.
MEMBER APOSTOLAKIS: Yes.
MEMBER KRESS: Yes.
MEMBER APOSTOLAKIS: I'm not sure about
the first bullet.
MEMBER KRESS: Okay. Let's go back to --
MEMBER APOSTOLAKIS: Is it the case of the
grass being greener on the other side?
MEMBER KRESS: Well, I think it's a case
of --
CHAIRMAN SHACK: But even if it's not, it
doesn't -- you know, I'm sure there are lots more
technical sophisticated ways to do lots of things.
The question is, you know, is it good enough?
MEMBER APOSTOLAKIS: And I'll submit the
overall fire risk assessment methodology that is used
by the fire community is not as sophisticated as ours.
Maybe individual tools are a little better.
MEMBER BONACA: But given the significance
of fire risk, okay, given the significance of fire
risk, I think that, you know, that's a statement that
says we have expectations that the NRC had the better
-- had these available, acknowledge that it's not
being used right now.
MEMBER APOSTOLAKIS: Well, really, the
message should be that the technical sophistication of
our tools is behind that of the state of the art.
That is more accurate I think.
MEMBER KRESS: Yes, why don't we say that.
MEMBER APOSTOLAKIS: But to compare now
with NIST and --
MEMBER KRESS: Yes, I think you're right,
George. Let's do it that way.
MEMBER APOSTOLAKIS: Okay.
MEMBER KRESS: Give her the --
MEMBER APOSTOLAKIS: Huh?
MEMBER KRESS: Give her what that --
CHAIRMAN SHACK: We'll work on it later,
and we'll -- just move on.
MEMBER APOSTOLAKIS: The technical
sophistication of models that appear in the
literature --
MEMBER KRESS: I think we're going to zap
the red part of it.
MEMBER APOSTOLAKIS: We're zapping what?
MEMBER KRESS: This.
MEMBER BONACA: The alternate?
MEMBER APOSTOLAKIS: Oh, the alternate.
Yes, that goes.
MEMBER KRESS: Okay. We will work on that
bullet.
MEMBER APOSTOLAKIS: Okay. And we're
keeping everything else? What?
CHAIRMAN SHACK: Did you take out -- the
last two bullets went.
MEMBER APOSTOLAKIS: Oh, the
significance --
CHAIRMAN SHACK: The significance of
termination and --
MEMBER SIEBER: And the first one gets
rewritten.
MR. DURAISWAMY: Hey, Tom, excuse me. You
just took out the last two bullets? The last one,
too?
MEMBER KRESS: Yes.
MR. DURAISWAMY: But the last one I think,
you know, they've got some problems between the
industry and the staff.
MEMBER KRESS: But that's why --
MR. DURAISWAMY: I don't think you should
take it out.
MEMBER APOSTOLAKIS: No. But I agree with
Bill here. It's not obvious to me that there are
computational methods that could be developed to
facilitate it. Have we ever investigated that? Is it
obvious to everybody else?
MR. DURAISWAMY: Well, that's why I think
we've got to set up a subcommittee to --
MEMBER APOSTOLAKIS: Yes.
MR. DURAISWAMY: -- talk about that and --
MEMBER APOSTOLAKIS: I cannot say right
now that obvious computational methods could be --
CHAIRMAN SHACK: The "obvious" has to go,
if nothing else goes.
MEMBER WALLIS: If you're going to say
this, I would have a period after "fires," and cut out
this and simply start, "Computational methods should
be developed to" --
MEMBER APOSTOLAKIS: That makes me much
happier.
MEMBER KRESS: Well, do we want to put --
MEMBER APOSTOLAKIS: Take out "when
obvious." And capitalize "computational."
MEMBER WALLIS: But a period here, too.
MEMBER APOSTOLAKIS: Where are you now?
MEMBER WALLIS: No, no, no.
MEMBER APOSTOLAKIS: We zapped it.
MEMBER KRESS: Let's put it back in.
MR. EL-ZEFTAWY: So you're just taking out
one "obvious"? Is that the only thing you're going to
take out?
MEMBER WALLIS: "Computational methods
should be developed." Should instead of could.
MR. EL-ZEFTAWY: And there's a period
after "developed."
MEMBER WALLIS: To facilitate the --
analysis, risk-inform the--
MR. EL-ZEFTAWY: Okay.
MEMBER KRESS: I would put a statement
when --
MEMBER APOSTOLAKIS: Is it really a
controversy with the licensees?
MEMBER WALLIS: Yes, I think you might
want to change that, too. I would say something --
"Staff finds itself disagreeing with licensees" or
something like that.
MEMBER APOSTOLAKIS: Yes.
MEMBER WALLIS: Or in disagreement with --
MEMBER APOSTOLAKIS: Yes, finds itself in
disagreement.
MEMBER WALLIS: Or it simply disagrees.
Do we need to "find itself in disagreement"?
MEMBER APOSTOLAKIS: Disagrees. That's --
MEMBER KRESS: Get rid of "finds" in that
sentence.
MEMBER APOSTOLAKIS: So you go to fires,
put a period, on the second line?
MEMBER KRESS: Yes.
MEMBER APOSTOLAKIS: Delete "when obvious"
and capitalize C.
MEMBER WALLIS: And then have "should"
instead of "could."
CHAIRMAN SHACK: And we'll leave it to the
highly paid arbitrator to straighten out the
constructions of the other bullets.
MEMBER KRESS: Yes, Sam can do that.
MR. DURAISWAMY: Not highly paid, but --
(Laughter.)
MEMBER KRESS: Let's go to 36. This is
a --
MEMBER APOSTOLAKIS: Oh, you're going to
another page? Oh, I had a question on line 640.
MEMBER KRESS: 640?
MEMBER APOSTOLAKIS: Yes. The ACRS has
reviewed the plan and the concurs with the research
program -- that the plan sets forth. When did --
MEMBER KRESS: Oh, we did that. Yes, that
was -- we did that a couple of weeks -- months ago.
That was --
MEMBER APOSTOLAKIS: A couple of months
ago we wrote a letter?
MEMBER KRESS: Steve Arntz.
MEMBER APOSTOLAKIS: The fire protection
--
MEMBER KRESS: No, it wasn't.
MEMBER APOSTOLAKIS: Yes, that was --
MEMBER WALLIS: Wasn't this a Jack Sieber
thing or --
MEMBER KRESS: Yes, that was --
MEMBER SIEBER: We all got copies of
the --
MEMBER WALLIS: Did we actually agree on
it. There's a huge fat thing that came out and --
MEMBER SIEBER: Yes, three-eighths of an
inch thick.
MR. DURAISWAMY: Did you write a letter,
Jack?
MEMBER SIEBER: Pardon?
MR. DURAISWAMY: Did you write a letter?
MEMBER SIEBER: No.
MR. DURAISWAMY: So then you can't say
ACRS completed the --
MEMBER SIEBER: We never got it
officially.
MEMBER APOSTOLAKIS: Okay. Why don't we
delete the sentence?
MEMBER KRESS: Just zap the last sentence.
Thank you, Sam.
MEMBER APOSTOLAKIS: Keep going. Keep
going. Is it longer or shorter now?
MEMBER KRESS: Keep going to the -- whoa.
MEMBER APOSTOLAKIS: Wait. No, no, no,
no. We are in the next section now. Go back. Well,
it's not 640 anymore. Go back more.
MEMBER KRESS: Okay. There it is. 641.
No, it's --
MEMBER APOSTOLAKIS: Go back some more.
MEMBER KRESS: Okay. Thank you.
Now, let's look at page 136. George, I
think this is a debate between you and Mario.
MEMBER BONACA: Yes. I would say that the
statement we had before that says "would surely" is
too strong. I agree with that. The words I had
originally was "is likely to."
MEMBER APOSTOLAKIS: Sorry? What were the
words?
MEMBER BONACA: "Is likely to." It might.
So I would change "would surely" to "is likely to" and
Bill is proposing "could." I just wanted to make sure
it wouldn't be too --
MEMBER APOSTOLAKIS: I would say "could"
is more neutral, isn't it?
MEMBER KRESS: Yes. "Is likely to" means
it is likely to. "Could" means there's --
MEMBER BONACA: That's my judgment. At
this stage it's a judgment. Certainly, if it is just
a might --
MEMBER APOSTOLAKIS: No, the "might" I
think is too weak.
CHAIRMAN SHACK: "Could" is stronger than
"might" and weaker than "is likely to."
MEMBER KRESS: "Could" is so weak that it
always applies. "Could" is about as weak as you can
get.
MEMBER APOSTOLAKIS: Replace "might" by
"could."
MEMBER BONACA: I don't know.
MEMBER APOSTOLAKIS: No?
MEMBER BONACA: Are you sure? You don't
think about "is likely to"? I understand that we're
implementing a regulation and that affects --
therefore, you know, PRA, you may have --
MEMBER WALLIS: Well, I sort of support
Mario that this is old, and surely something has
changed.
MEMBER BONACA: If you make it too weak,
the whole section becomes into question. Why have a
full section proposing something if you're really
making a statement that is so weak that says "so
what?" I mean, if really aging is not an issue, and
you can have this plant at 600 years of age, I mean,
why propose --
MEMBER KRESS: Well, we could say results
could show increases in risk metrics. That's almost
certain. I don't think they're going to go down.
MEMBER WALLIS: It implies that we think
it's likely.
MEMBER KRESS: Yes. But even if it's
likely, this doesn't address the question of whether
George's statement that it's already so low that an
increase doesn't make much difference gets lost in the
noise. I think that's a significant statement.
MEMBER BONACA: I think George was
focusing mostly on main components for the vessel
rather -- I'm thinking about total -- I'm thinking
about those nozzles that we saw cracked, and the UT
failed to detect it.
MEMBER WALLIS: Well, I suggest no word at
all. PRA that could account for aging of structures,
systems, and components. Oh, I see. I'm sorry. I
screwed up on that.
MEMBER BONACA: I agree that, you know,
the statement "would surely" was too strong. But I
think that -- I think I believe it's likely to show
some -- I think it surely would --
MEMBER APOSTOLAKIS: The end part program
of several years ago did not find any significant
failures due to aging. It found partial degradation,
things like that, but, hence, to see the impact of
aging on failure rates have reached no conclusions.
There is no evidence that the failure rates increase.
When we did the small study at MIT, again,
we were hard pressed to really find a significant
change in probabilities of core damage, and so on.
That's why I'm reluctant to be very positive that,
yes, we will find an impact. On the other hand, I'm
not ready to say, no, there will be no impact. So, I
mean, what --
MEMBER BONACA: So this is -- I've been
thinking about the experience we had from industrial
facilities other than LOCA, is that when it reaches
end of life, even with the proper maintenance it
becomes so troublesome that they get shut down just
for economic reasons.
Now, here the implication is that those
kinds of, you know, compounding failures you see more
leaks here and there, some problems, are not going to
create a problem from a safety standpoint. It will
only create a problem from an economic standpoint, so
the plants would be retired. That's really the
conclusions we are reaching.
Well, I think it's a stretch to reach just
a conclusion. There is complexity that says, you
know, common sense is telling me that it could affect
those there. If I use "might" -- a section of this
size, it would be --
MEMBER APOSTOLAKIS: I am with you on the
"might," but do you disagree on using "could"? Is
that too weak?
MEMBER BONACA: "Could" is a little weak.
I mean, I -- you know, I thought that -- I really
believe it is likely to show, but I have never made it
-- we are assuming that --
CHAIRMAN SHACK: But you have Tom's
problem -- is that surely there are increases. The
question is whether they're significant or not.
MEMBER BONACA: Yes, I understand. Well,
whatever it says, we believe in fact that they are not
insignificant because of the --
MEMBER WALLIS: I have a different way of
putting it. A PRA that could account for aging of
structures would provide measures of increases in --
MEMBER BONACA: That's not the meaning of
this. What we intended to say is that we would see
some increase due to the fact that you have increased
failure rates of some type, and then the discussion of
how you would contribute to those increases.
MEMBER WALLIS: But I think your
colleagues object to the distinction that there are
going to be increases.
CHAIRMAN SHACK: Graham gets us around
that problem. I think his says --
MEMBER WALLIS: Yes, I had a time where
there's going to be increases or not.
MEMBER BONACA: Could you repeat your --
MEMBER WALLIS: I said -- well, take what
we've got. A PRA that could account for aging
structures, systems, and components, would provide
measures of increases in risk metrics, such as core
damage frequency. We don't even need to say
"increase."
MEMBER APOSTOLAKIS: It's a neutral
statement.
MEMBER BONACA: I agree.
MEMBER APOSTOLAKIS: It's more neutral.
MEMBER BONACA: I agree.
MEMBER APOSTOLAKIS: It says we want to
know.
MEMBER BONACA: Right. I agree. And
that's important for the --
MEMBER WALLIS: Without trying to guess if
it's going to be --
MEMBER BONACA: I agree with that.
MEMBER APOSTOLAKIS: All right. So will
you tell Sheri what to do there?
MEMBER WALLIS: Where are we? Take out
all the red. Take out the "might." Take out the red
stuff. Okay. And then take out "show." Would
provide measures of --
MEMBER APOSTOLAKIS: No, "would surely"
goes.
MEMBER WALLIS: Would provide -- instead
of "increases," which occur -- okay. Take that out.
Okay. Would provide measures of -- a long sentence.
Increases in -- has aged.
MEMBER APOSTOLAKIS: And I would put a
period there and say, "These increases are due to" --
CHAIRMAN SHACK: Yes, the sentence is too
long.
MEMBER APOSTOLAKIS: Go back. Not right
there. You want to put it after "operations" comma.
As aged from 40 years to 60 years of operation. And
then this increase would be expected due to increases
-- there are too many "expecteds."
MEMBER WALLIS: Just take out the second
"expected."
MEMBER APOSTOLAKIS: Yes.
MEMBER WALLIS: The second "expected."
Well, due to a higher failure probability.
CHAIRMAN SHACK: An increase in failure
probability along with components. Good enough.
That's okay. Is that --
MEMBER WALLIS: Should we leave up to the
"highly" --
CHAIRMAN SHACK: Yes.
MEMBER APOSTOLAKIS: So, essentially, what
we're saying is that even though the plants meet the
requirements of the license renewal rule, the risk
will increase as they're allowed to operate from 40 to
60 years.
MEMBER WALLIS: And we want to know. And
we want to -- they should find out.
CHAIRMAN SHACK: But George has a comment
on page 37 about some -- you know, the -- we didn't
say anything that, you know, people are trying to
manage this.
MEMBER BONACA: But the statement before
says the risk increase is found to be small because of
the implementation of requirements which are --
MEMBER APOSTOLAKIS: Where is that?
MEMBER BONACA: The statement right before
that. It gives credit to -- like 137. I mean -- has
extensive modeling programs in place. That's implicit
in that statement that --
MEMBER APOSTOLAKIS: Where is that
sentence that it says --
CHAIRMAN SHACK: 137.
MEMBER APOSTOLAKIS: I know the page, the
line.
CHAIRMAN SHACK: The line is 2536 on your
printed copy.
MEMBER KRESS: 538.
MEMBER APOSTOLAKIS: The risk increase is
small because implementation reserves regulatory
margins. Where does it say about --
CHAIRMAN SHACK: Well, I was going to
change that line to read, "The risk increase may well
be found to be -- may well be found small because the
license renewal process is intended to provide
insurance that aging management programs preserve
regulatory margins," duh, duh, duh, duh, duh, duh,
duh.
MEMBER APOSTOLAKIS: Good. That's what --
sounds good. That's good.
MEMBER BONACA: So give it to Sheri?
CHAIRMAN SHACK: I'll just give it --
while we're on that page, too, I thought Tom -- we
focus on the other metrics for power uprates. That
sentence on 137 that starts at 2544, I was going to
suggest changing to, "Assessments of the increases in
risk associated with license and renewal and power
uprates may need to consider risk metrics other than
CDF and LERF."
MEMBER KRESS: Oh, great. I love that.
MEMBER BONACA: What is the line?
CHAIRMAN SHACK: It's this line. We just
focused it purely on power uprates, and I think Tom
would say that you really need to look at that in a
broader sense.
MR. EL-ZEFTAWY: If we were to leave the
last one -- if you look at page 139, line 2579, can
you tell me if you agree on that change?
MEMBER APOSTOLAKIS: Where?
MR. EL-ZEFTAWY: Page 139.
MEMBER APOSTOLAKIS: Yes, it was approved,
I thought.
MEMBER KRESS: Yes.
MR. EL-ZEFTAWY: So that's approved?
MEMBER APOSTOLAKIS: Yes.
MR. EL-ZEFTAWY: The e-mail exchange. All
right.
MEMBER APOSTOLAKIS: Yes, I have no
problem with that.
MEMBER KRESS: Okay. Now, I suggest we
look at --
MEMBER APOSTOLAKIS: Any corrections?
MEMBER KRESS: No, not yet. We'll save
that to last. But look on pages 8 and 9, and let's
deal with those. I think Bill's suggestion was to
move some of this around and delete some of it. He
wanted to delete the lines highlighted in 140 and 141
and move lines 152, 153, and 154 up to replace them --
I think was the suggestion, wasn't it, Bill?
MEMBER WALLIS: What we have is actually
the new 140/141. That's an addition to -- the
highlighted 140/141 is an addition. It comes from
152, which has been --
MEMBER KRESS: That's just -- you're
right. And my suggestion was to not do that.
MEMBER WALLIS: Okay.
MEMBER KRESS: Is to -- is to delete that
line 140/141 and retain 152 and --
CHAIRMAN SHACK: Right. Yes. I wanted to
insert that line up there, but I didn't know where to
put it is all.
MEMBER KRESS: Yes. My feeling was that
that sentence doesn't belong up there, because it's an
entirely different subject matter than the rest of the
paragraph.
CHAIRMAN SHACK: Yes, but read the
sentence before that, starting on 137, and then look
at the sentence that starts on 152. "One would expect
that the Commission would -- would have available
comprehensive state-of-the-art assessment tools. One
would expect that site-specific risk information would
be readily available to line organizations
implementing risk assessment process." To me they're
saying exactly the same thing.
MEMBER WALLIS: Yes, it is. I think
you're right. I thought this was okay, sort of a
compression of the idea.
MEMBER KRESS: Yes, but I would want to --
I'd still want to get rid of that statement that risk
assessment remains an activity that --
MEMBER WALLIS: Do you want to remove it
all together?
MEMBER KRESS: Yes, that part of it I'd
like to get rid of.
CHAIRMAN SHACK: But would you then retain
any part of 152 and 154, or that just all goes?
MEMBER KRESS: I would get rid of all of
it, yes.
CHAIRMAN SHACK: Less is more.
MEMBER KRESS: Yes. Now, on page 9, Bill
had added in the shaded part, and I suggested that we
not add it in. And his reasoning I think was that
this is something we're leaving out of the statement
that's of use, but I -- my feeling was that these
limited numbers of shutdowns are not useful at all to
shutdown risk. I don't even want to refer to them,
because I think they're useless.
CHAIRMAN SHACK: For your purposes,
they're useless. You know, if you're looking for
insight, I think they are useful. You know, they
don't allow you to compute the average lifetime risk.
MEMBER KRESS: I think they're -- I think
the statement that they're -- I mean, I think it's
overly strong to say that you've got scoping
assessments of shutdown risk at two representative
plants. Period.
MEMBER WALLIS: I think they need more
than that. They used more information than that.
CHAIRMAN SHACK: That, to me, is just, you
know, a -- we didn't do it; therefore, we ain't going
to look at it.
MEMBER KRESS: Well, I'm not strong about
-- you're right. I can go along with leaving it in.
It doesn't hurt to say this.
MEMBER WALLIS: It's not a lie, is it?
MEMBER KRESS: No. Well, it's a very
limited --
CHAIRMAN SHACK: Put in a very --
MEMBER APOSTOLAKIS: Let's leave it in.
MEMBER WALLIS: They did use this
information. I mean, it --
MEMBER KRESS: Yes, let's leave it in.
Okay. Now, I didn't have a chance to send
everybody the suggestions that Graham Wallis had on
what to do with the --
MEMBER WALLIS: I sent them to all of the
ACRS --
MEMBER KRESS: So you got some suggestions
from Graham on what to do with the -- with some things
that may be wrong with the introduction. And what I
did was you have before you something that looks like
this.
MEMBER APOSTOLAKIS: Oh, this is Graham's?
MEMBER KRESS: No. This is my response to
Graham's suggestions.
MEMBER APOSTOLAKIS: Oh.
MEMBER KRESS: Before you look at it, I
want to make one correction. Under two on page 2,
there's supposed to be an introductory sentence that
says the examinations of the research programs by the
ACRS then did not focus on the initial need for the
research results. Instead, the exemptions focused on
the questions. And then --
MEMBER APOSTOLAKIS: Right. That's fine.
MEMBER KRESS: That's supposed to be --
MEMBER APOSTOLAKIS: That's fine.
MEMBER KRESS: But this is my -- what I've
done is just rearranged things.
CHAIRMAN SHACK: Well, you've wiped out
the whole user needs stuff.
MEMBER KRESS: No, it's supposed to be in
there.
CHAIRMAN SHACK: Did it move somewhere?
MEMBER WALLIS: It wasn't supposed to.
MEMBER KRESS: It just moved.
MEMBER WALLIS: It should be removed
because --
MEMBER KRESS: Well --
MEMBER WALLIS: -- as it was.
MEMBER SIEBER: Well, I just moved it
because I wanted you to look at this and see how it
read and then make a decision whether to black out the
--
MEMBER WALLIS: Well, you put it in now
with your added sentence under two. You said we
didn't refer to the user needs. We used these other
criteria, and that clarifies it. I don't think we
need to talk about user needs again, do we? Because
we sort of take a swipe at them which isn't justified.
MEMBER KRESS: Where is the user needs
part in here?
MEMBER WALLIS: On page 5. Nevertheless,
motivation -- did not -- we could remove that because
we've already said that.
MEMBER KRESS: Yes, that's the question.
Do we want to leave that in or --
MEMBER WALLIS: Let's take that out
because it sort of takes a swipe at something we never
expand upon in any way.
MR. EL-ZEFTAWY: You're talking page 5,
line -- the paragraph at the top?
MEMBER KRESS: Paragraph in the middle on
page 5.
MR. EL-ZEFTAWY: Yes.
MEMBER WALLIS: Actually, we're talking
about concerns about using the process at the top of
page 5.
CHAIRMAN SHACK: Right. It's the first
paragraph 5.
MEMBER WALLIS: I would like to remove
both of those paragraphs.
CHAIRMAN SHACK: Yes, I don't think it
helps.
MEMBER WALLIS: It says -- it sort of says
the user needs process is in place, but we've ignored
it completely because we think it's pretty lousy. I
don't think that's what we want to say.
CHAIRMAN SHACK: I'm sure that's what Dana
wants to say.
MEMBER WALLIS: Is it? But I'm not sure
that's what we want to say, is it?
MR. LARKINS: I think Dana's point was
that, you know, all of the research shouldn't be
driven by the user need requests, that it should be
some portion or percentage of the work that's done
outside of this process, research on its own.
MEMBER KRESS: I suspect everybody would
agree with this statement. So I don't know why we
don't --
CHAIRMAN SHACK: But I thought it was
cryptic enough as it was written.
MEMBER KRESS: Yes, it is pretty cryptic,
isn't it?
CHAIRMAN SHACK: I mean, I had a rewrite
of the paragraph that says, "As now constituted, the
user needs process may lead to an overemphasis on
short-term work to support immediate needs and not
result in adequate support for research needed to
improve line organization capabilities."
MEMBER KRESS: Is that a rewrite of the
first paragraph?
CHAIRMAN SHACK: Yes, that's a rewrite of
that first paragraph.
MEMBER WALLIS: You need something like
that.
MEMBER KRESS: Yes, that sounds pretty
good.
MEMBER WALLIS: I would support that and
keep the first paragraph. I'd support that. And then
the second paragraph I'm not sure we need that because
they've already said that in the sentence that you
added earlier.
MEMBER KRESS: I would go along with both
of those suggestions, Bill's rewrite of the first
paragraph and --
MEMBER WALLIS: And removing the second
one on page 5?
MEMBER KRESS: Yes.
MEMBER WALLIS: Okay. So, Bill, are you
going to --
CHAIRMAN SHACK: I'll give you some words,
and you can figure out how to work them in.
MEMBER APOSTOLAKIS: So you all agree with
the third bullet here on this page, too?
MEMBER KRESS: Oh, yes, that's another
question.
MEMBER APOSTOLAKIS: That was the ACRS
approach?
CHAIRMAN SHACK: Well, I think we did try
to consider that.
MEMBER APOSTOLAKIS: Some place.
CHAIRMAN SHACK: Some place.
MEMBER APOSTOLAKIS: Not really -- it was
not a uniform --
CHAIRMAN SHACK: Nothing is ever uniform.
We have no criteria, George, no process. We --
MEMBER KRESS: No privatization.
CHAIRMAN SHACK: In fact, I had a
paragraph if you'll look on the back of that, that
says we -- can you give me that thing back?
MEMBER KRESS: No.
(Laughter.)
CHAIRMAN SHACK: I was going to put --
following the three bullets, I was going to add
something like, "In previous reports, we have argued
that processes and criteria need to be developed to
address such questions. These have not been
developed, and our current assessment is based on our
own intuitive judgments."
(Laughter.)
Get into the Jocelyn-Graham job there.
MEMBER WALLIS: I'd like to remove the
"intuitive."
MEMBER APOSTOLAKIS: Isn't it true that
the whole risk-informed revision of the regulations
really should be done by the licensees? Who is
benefitting from all of that? Why do we have to do
it?
MEMBER KRESS: What did you say?
MEMBER APOSTOLAKIS: Yes. Why do we have
to do it? Why do we have to develop performance
indicators?
MEMBER KRESS: That's the job of the
agency.
MEMBER APOSTOLAKIS: Why? Who is
benefitting from it? I would argue that the licensees
are benefitting and we --
CHAIRMAN SHACK: We all benefit, George.
Benefit is not -- benefit is not the criteria. That
may be one of the elements of the thing.
MEMBER APOSTOLAKIS: I don't know. Is
this work that needs to be done independently by the
NRC? Why is the revised oversight process something
that we need to do? They should propose all -- we
review and approve.
CHAIRMAN SHACK: That's why we need
criteria and judgment for doing that.
MEMBER APOSTOLAKIS: But that's not what
it says. It says that we actually use this.
CHAIRMAN SHACK: We tried to think our way
through that based on our own judgment of when things
needed to be done independently, and when we could
just review the license --
MEMBER APOSTOLAKIS: I think it was done
in a very awkward way.
CHAIRMAN SHACK: Of course. Of course it
was.
MEMBER APOSTOLAKIS: We could put it front
up here that -- I mean, I understand the first two
bullets. I mean, that we really did.
MEMBER KRESS: I think we've exceeded our
one hour now. So we can put it on the agenda --
MEMBER WALLIS: Well, I do want to make
some resolution about this question of by what
criteria is this decided.
MEMBER APOSTOLAKIS: Yes. And that is
something that bothers me, too. I mean, the only
thing we didn't put are those little angels around the
--
MEMBER KRESS: Well, I think we would just
indent it and not bold it.
MEMBER APOSTOLAKIS: Huh?
MEMBER KRESS: I think we would just
indent it and not bold it and --
CHAIRMAN SHACK: We already have it as one
of the three bullets. Why repeat it again? Is sort
of my theory. You know, it's the same as this -- this
work that needs to be done independently by the NRC
rather than depending on information supplied by the
licensee. We've said it once. We don't have to say
it again.
MEMBER WALLIS: Well, it's said again in
response to this tension and competition. I want to
--
CHAIRMAN SHACK: Well, it's mentioned in
--
MEMBER WALLIS: Why don't we black that
whole thing about tension and competition?
MEMBER APOSTOLAKIS: Although the
identification? This paragraph? This paragraph?
MEMBER KRESS: I might be in favor of
that, particularly because there is something in the
wind and we may want to add some more to this in --
MEMBER APOSTOLAKIS: You're talking about
the paragraph that starts, "Although there are" --
MEMBER KRESS: Yes.
MEMBER APOSTOLAKIS: Oh, that would make
me very happy to take that out.
MEMBER WALLIS: That until the end or the
whole thing?
MEMBER APOSTOLAKIS: Until the end, the
present end.
MEMBER WALLIS: The present.
MEMBER APOSTOLAKIS: There would be
another end.
MEMBER WALLIS: There would be another
end?
MEMBER APOSTOLAKIS: Yes.
MEMBER WALLIS: You're rewriting the end
of the story?
MEMBER APOSTOLAKIS: Shall we move on now
and --
CHAIRMAN SHACK: We have to move on now.
We have --
MEMBER APOSTOLAKIS: I think we have to
take action here and vote. But the thought occurred
to us earlier that perhaps we should add a few
paragraphs to this as to where the agency -- what are
the challenges in the future, in particular with new
reactors. We have some of that in Roman 3, but we
should move it up maybe from last year.
So what I would propose is that Tom and
Dana -- Dana will be back on Monday -- add a few
paragraphs and circulate them by e-mail, but we take
a vote today, subject to that condition.
Tom, is that correct?
MEMBER KRESS: Yes. I was under --
MEMBER WALLIS: I think it would be very
appropriate if we can do it well. I think it should
be in this report. We're not going to --
MEMBER APOSTOLAKIS: Okay. So that's why
we are assigning Tom to do it.
MEMBER KRESS: It says I have to do it
well.
MEMBER APOSTOLAKIS: Yes.
MEMBER KRESS: I didn't agree to that.
(Laughter.)
MEMBER APOSTOLAKIS: I move that this
report be accepted by the committee, subject to this
condition that Dr. Kress will supply a few paragraphs
to be added to the introduction regarding future
challenges.
MEMBER BONACA: And the last paragraph?
The introduction is scratched?
MEMBER APOSTOLAKIS: That has been a --
MEMBER KRESS: And I also met something
that -- if you'll let me raise this with --
MEMBER APOSTOLAKIS: That's part of the
motion.
CHAIRMAN SHACK: Yes. Dr. Kress has full
power to do the cleanup work.
MEMBER KRESS: Okay. Good.
CHAIRMAN SHACK: He's the cleanup man.
MEMBER APOSTOLAKIS: Yes.
CHAIRMAN SHACK: Second.
MEMBER APOSTOLAKIS: So there is a motion
on the table and it has been seconded. Any
discussion? Hearing none, those in favor of the
motion raise your hand, please. The research report
is approved.
CHAIRMAN SHACK: We're running a little
bit late. I think I still want to take a five-minute
break.
MEMBER APOSTOLAKIS: Yes. Maybe we can
shorten the lunch break. Okay? So we will be back
when, at quarter of?
CHAIRMAN SHACK: Quarter of.
(Whereupon, the proceedings in the
foregoing matter went off the record at
9:36 a.m. and went back on the record at
9:45 a.m.)
CHAIRMAN SHACK: We'd like to come to
order now and begin our discussion of the 50.46. And
I believe we'll start with the industry presentation
by Mr. Heymer, assorted support from a wide variety of
people.
MR. HEYMER: Good morning. My name is
Adrian Heymer. I'm a project manager at NEI dealing
with risk-informed regulation under Tony Petrangelo,
who's our director, and I've been following the option
three as well as some of the option two activities.
This morning what we're going to talk to
you about is 50.46, and specifically what we believe
the most important element to look at in 50.46. And
the one with the highest priority from a safety
enhancement as well as a resource benefit is
redefining the large break LOCA.
And I have with me here Lewis Ward from
Southern Nuclear; Bob Osterrieder, Westinghouse Owners
Group; Dave Bajumpaa, Millstone and the CEOG; and
Terry Rieck from Excelon and representing the BWR
Owners Groups.
We did have another representative from
the B&W Owners Group, but something happened and he
couldn't make it this morning. Otherwise, we would
have had the complete spectrum of the owners groups
here.
And I think one of the messages we want to
provide today, that this is an industry-wide activity.
The owners groups are on board, and I'm going to go
over some of those issues and the industry structure
and background as we move forward with this
presentation.
So what we are really focusing on today is
redefining the large break LOCA. I will go over some
background information how we got here and the general
approach that the industry sees to improving this
aspect of the regulation. Then, the Westinghouse
Owners Group will talk about a specific approach and
some of their activities.
And then we'll have an example from the
CEOG of what we call an application, what flows from
redefining the large break LOCA to give you an idea of
where the benefits are, and then Terry Rieck will say
a few words on behalf of the BWR Owners Group.
I guess where we started off on risk-
inform in the regulations was several years ago with
SECY98-300, and at that time we were using this slide
of what we believe is the important aspect of risk-
informing NRC technical requirements, which was the
improved efficiency and effectiveness of the NRC
regulatory regime, to provide an increased focus on
those issues that are safety significant while
reducing unnecessary burden.
And that's the sort of fundamental element
that we've looked at as we've moved forward. And,
obviously, to do that, you've got to look just not at
the regulations but also at the guidance documents and
at the industry codes and standards activities.
MEMBER WALLIS: Well, there's another one
of their objectives which is maintain safety.
MR. HEYMER: Well, yes, but I --
MEMBER WALLIS: Remember, that's one of
the constraints.
MR. HEYMER: Yes. Yes. I mean,
increasing the focus on safety-significant issues
should -- we believe should enhance safety.
And I guess as we went through this,
following the Commission's SRM on 98-300, we went
through and we looked at the technical requirements
and we came up with a list, and there was an NRC
workshop and we discussed some of those things with
the NRC. And we went out to the industry with a
survey, and we included the list of regulations but we
said, "What do you, the industry, think that we should
focus on to improve our focus, focus on the safety-
significant issues, and provide some benefit?"
And we got a list back from the industry,
and we provided that list to -- the results of that
survey to the Commission in January of 2000. And in
that, there were three specific areas. One was -- the
first priority was focus your activities on finishing
what you've already started, which was things like the
oversight process, fire protection, and the technical
specifications, and then look at 50.46 and 50.44.
And the reason why they put 50.44 up there
was that they felt with the amount of work that had
gone on on 50.44 that was something that we could move
forward fairly expeditiously. But on 50.46, there was
a number of issues why we came up with 50.46, and it
just wasn't really associated with financial issues.
The initial 50.46, it has a large number
of tentacles that go out. Throughout the regulations,
there's a number of issues that are linked to 50.46,
and it was felt that if we could identify some of the
items that perhaps don't have the same degree -- high
degree of safety significance and where we perhaps
could better focus our activities, we could make
improvements both in safety and in -- and in the
financial profile of the plant by looking at 50.46.
MEMBER WALLIS: Are you going to enhance
safety as well as try to reduce burden by removing the
focus on low-risk significant events, but you can
actually also look at the other side of the coin, that
there are other things that are more important where
you can enhance safety?
MR. HEYMER: Other things that are more
important that we should place greater emphasis and
resources --
MEMBER WALLIS: There's going to be a
tradeoff. It's not going to be all just reducing
burden. It's going to be actually --
MR. HEYMER: Well, as has been said on
many occasions, there's two sides to this equation,
and we accept that. And if there's things that come
up that are required for safety --
MEMBER WALLIS: I think for public
reassurance, there's sort of -- there's a big drama
associated with the large break LOCA. And if you sort
of want to -- not that the agency should back off on
that. There's going to be some good arguments that
your -- because now your attention is focused on
something else, you are actually improving safety.
Otherwise, it looks as if you're -- the agency is just
backing off. I don't think that's very good for
public confidence.
MR. HEYMER: No. And I think that's the
reason why we're couching it the way we do. And I
think if you look at the -- the large break LOCA, if
you just look at some of the studies that have been
done out there, it is a relatively low probability
event.
And if you take that on as, we believe,
negligible public risk, and -- and but we're not
intending to sort of just throw everything away
associated with that. I think we've developed,
through risk-informed ISI, a much better process of
looking at what inspections we need to do on those
activities. We still have detection. We're still
going to carry out inspections in that regard.
Now, as I said, the effect from a safety
perspective associated with redefining the large break
LOCA is -- we think is very significant. As I said,
it's an essential element in the regulatory structure.
And if you can redefine what the break size is, then
the follow-on activities and your resources can be
adjusted to focus on the more probable activities and
those matters that are of safety significance.
And you'll see as we go through the
presentations here today some of the activities that
we get involved in link specifically to the large
break LOCA, which if you take a more realistic
approach to it we wouldn't have to be expending
resources in that area. So we hope to --
MEMBER WALLIS: The fact that it is a
central and controlling element, the way you've
identified it here, means that in the past it was
assessed as being important enough to have this role
of being a central and important controlling element.
MR. HEYMER: That is true. But I think as
-- and as we started out the regulations, there was a
very conservative approach to say it's the double-
ended guillotine break of the largest pipe. And
that's why we're emphasizing it's the redefining.
It's not the elimination of the large break LOCA.
We're redefining it.
MEMBER WALLIS: So it's not, though, as if
you're asking to simply redefine something of low-risk
significance. You're asking to redefine something
which is a central and controlling element. It's a
major step.
MR. HEYMER: It is a significant step.
But on the other hand, it has, we believe, significant
benefit, both in terms of safety and finance. Having
determined it was 50.46 and perhaps it should be
redefining large break LOCA, Westinghouse Owners Group
already had an activity underway and have already done
some extensive evaluations of redefining the large
break.
And through those activities, we pulled
the other owners groups together, and I think we've
developed over the past 18 months sort of an industry
approach, which I'll go into and which we've described
in several meetings and workshops with the NRC staff
as we've moved through and discussed the options of
what to look at first, because 50.46 is a very large
and complex regulation. And we think if you're going
to look at 50.46 you need to focus on what is really
going to provide the benefit, and to us that is
redefining the large break LOCA.
We've listed some of the -- some of the
safety enhancements that we see from here. I think on
a number of unnecessary plant transients it gets back
into how many times you can sort of begin to run up
against the limits in the technical specifications
that are linked back to the large break LOCA. And so
if you don't have to have a power train, you shouldn't
be imposing one, not only from a financial
perspective, but also from a plant safety perspective.
MEMBER WALLIS: Can you explain that a bit
more? You say "unnecessary plant transients." You
mean deliberate transients. You have to run the plant
through some transient under the regulations.
MR. HEYMER: Because of the regulation.
MEMBER WALLIS: It means you have to test
things or something?
MR. HEYMER: On a testings or come down in
power or shut down or come to a halt -- shut down
while you fix something, or even go to a cold
shutdown. So the number of times you exercise -- move
through those -- those plant states --
MEMBER WALLIS: Is all motivated because
of the large break LOCA?
MR. HEYMER: Or linked to requirements and
technical specifications or other -- that are linked
into the large break LOCA.
MEMBER WALLIS: It would be nice if you
could have some numbers associated with that, and let
us know what's the number and the cost or something.
MR. HEYMER: Yes, we can get back in
subsequent presentations on that.
MR. OSTERRIEDER: We do have a couple of
examples later in the presentation.
MR. HEYMER: Yes. We'll speak to that.
We speak here about improved worker safety
profile, and I guess the potential to rebalance the
ECCS system so that we -- we focus on more probable
events, such as the small break LOCA or the breaks of
a smaller size, intermediate and small, I think that
once you come up with whatever the new break size
would be and you start running that through the PRA,
your safety assessments and your PRA assessments
become even more meaningful and just improves the
general process. So --
MEMBER WALLIS: So you're claiming that
because of the focus on large break LOCA we have some
requirements for ECCS that may actually be detrimental
in the case of other kinds of LOCA?
MR. WARD: Yes, sir.
MEMBER WALLIS: Yes?
MR. WARD: The particular example that
we've discussed several times is the balancing of the
ECCS system so that when you have a large break LOCA
all of the water does not go to the broken loop, that
there is a certain amount of it that goes into the
intact loops and then goes to the core. And those --
we put orifices in typically to do that and balance
them to prevent runout on the pumps in that condition.
But what that does is if we had, for
example, a smaller loss of coolant accident or a small
leak, that -- those orifices are still in place and it
throttles back the amount of water that would be
delivered to the loops in that condition.
If we were designed to some intermediate-
sized break, then you could decide -- you could open
up the orifices, provide more flow to the core for the
small break than we do now, because you would not have
the pump runout concerns on the high end. That's one
example that we thought about.
MR. HEYMER: Our overall approach for
redefining the large break LOCA takes into account
that we have varying designs out there and varying
designs -- we have boilers, we have pressurized water
reactors, we have CEABB plants, we have Westinghouse
plants, we have B&W plants.
And, therefore, what our approach is is a
relatively straightforward rule change. At the
moment, the regulations say that you will analyze with
the double-ended guillotine break of the largest pipe.
And we think the add-on would be you would just add a
phrase "or alternative break sites as approved by the
Commission."
And to the extent of the rule change,
there would be some other conforming changes where in
other parts of the regulations, perhaps the general
design criteria, you'd define what a loss of coolant
accident is, so that there would be some conforming
changes there.
And having done that, and having started
to progress with the rulemaking, each owners group
would develop and submit what they believe would be
the justification for redefining the break size for
their particular designs. And that would be an owners
group specific activity.
But just redefining the break size alone
doesn't really get you there, and so you start looking
at applications. And so once there is a good
understanding on what the break size would be as we
begin to focus down and reach an understanding with
the staff on what the break size would be, you can
then start looking at the specific applications, be it
diesel generator start times, balancing ECCS that we
spoke about.
And they would be done, again, on an
owners group basis, generic as much as we could for
each owners group, so that when the licensee came
along they could just submit a license amendment based
on the topical reports that have already been approved
by the staff. And we think that would be the most
efficient use of resources of moving through this.
So it's really simple that the initial
step is -- is to move forward with a rule change that
would say -- as I've said, allow an alternative break
site as approved by the Commission, but don't define
that break size and leave that from the -- for the
technical interactions between the owners group and
the staff to come to some conclusion on what that
break size is.
MEMBER WALLIS: Wouldn't you still have to
analyze the large break LOCA anyway in order to show
that it's not significant? You have to do something
with it. You can't just ignore it.
MR. HEYMER: Well, no, you just don't
ignore it. That's some of the issues that we're going
to be --
MEMBER WALLIS: But you'd still have to do
an analysis and convince the Commission that this
break is not important or something.
MR. HEYMER: Well, as we --
MEMBER WALLIS: Wouldn't that just be a
risk analysis? Or would that be a technical analysis?
MR. HEYMER: Well, we will continue --
MEMBER APOSTOLAKIS: Mechanistic, you
mean.
MR. OSTERRIEDER: We would continue with
large break LOCA in the risk models for the plants.
But, you know, so that -- it wouldn't be taken out, so
--
MEMBER WALLIS: So it would still be in
the risk models, but it wouldn't be in the sort of
technical requirements.
MR. OSTERRIEDER: Right. It's like other
things in the risk models that aren't necessarily part
of your design basis.
MEMBER WALLIS: Yes. But in order to do
the risk model, you have to do a thermal-hydraulic
type analysis and everything. You have to look at
consequences and all that.
MR. OSTERRIEDER: Right. You need to do
appropriate success criteria analysis.
MEMBER WALLIS: So it wouldn't go away.
You'd have --
MR. OSTERRIEDER: That's correct.
MEMBER SIEBER: Well, it's an interesting
thing. If you enlarge the size of the orifice, which
you said was going to be an advantage to having a
smaller break size, then if you actually did have the
large break the pumps would run out, and the outcome
would be different than you currently have now. So
the risk numbers and consequences would go up. Is
that not true?
MR. OSTERRIEDER: Right. Yes, they
would --
MEMBER SIEBER: In other words, you could
not handle the --
MR. OSTERRIEDER: Right. The risk of that
occurring would be assessed into the plant's risk
model, since it --
MEMBER SIEBER: That's right. But if it
did occur, whatever the probability, the ECCS couldn't
handle it.
MR. OSTERRIEDER: That's right.
MEMBER SIEBER: Under those circumstances.
MR. OSTERRIEDER: Well, we'd assess with
the success criteria -- certainly more likely less
success probability, certainly.
MR. HEYMER: And as we get into the
discussions here, I think you'll also hear that our
emphasis isn't necessarily doing extensive
modifications based on this, but allowing for
operational margin to -- for us to operate with that
margin so you don't have to get involved in some of
these evaluations and activities that Dave here will
talk about from the CE perspective -- so perhaps
ultimately the sink calculations, containment, heat
removal.
So we're not talking about ripping out
pumps and replacing pumps. What we're talking about
is, okay, we don't have to -- perhaps the engineering
specification is going to be the same, but the actual
licensing and technical specification may be a little
bit different.
MEMBER KRESS: Why shouldn't I view this
from the perspective of Reg. Guide 1.174 and say
here's a suggested change in the licensing basis for
lots of plants, not just one. That will result in
these changes to specific plants, the listed changes.
I suspect you have those. And then that will change
the risk status of each of these plants by this much,
and looking at the guides in 1.174 say whether that's
acceptable or not. Why isn't -- I don't see a
perspective -- well, it seems like we're viewing this
strictly from the design basis accident space and not
--
MR. HEYMER: Well, I mean, the initial
step coming out of this isn't to say, well, we want to
redefine the large break LOCA because we want to have
a different pump size there. But on the other hand,
if somebody then wanted to go and implement a
modification, they would then use the 1.174 as the
guideline. And the guideline and the baseline for the
plant would be adjusted based on whatever the new
break size came out to be.
So, I mean, I think what you're saying is
that, yes, okay, once you've redefined it, you may
some stage down the road want to perform a
modification, and, yes, you would use 1.174.
MEMBER KRESS: I see. So that would come
in at the point where the plant decided -- a specific
plant decided that --
MR. HEYMER: Yes.
MEMBER KRESS: -- take advantage of the
new definition and make some changes.
MR. HEYMER: That's right. Within the
confines of 1.174 and the new rule and the guidelines
and the technical documents, but not as a direct --
MEMBER KRESS: Why couldn't they do that
already?
MR. OSTERRIEDER: I think it's more simple
than that. We are planning to use the 1.174 framework
and assess the risk significance of large break LOCA.
So I think the answer is simply, yes, we are intending
to do what you're suggesting that we should be doing.
MEMBER KRESS: Yes.
MR. OSTERRIEDER: As you'll see it in my
summary of what we've been able to do. I think we're
doing that.
MEMBER KRESS: Maybe I should wait until
we hear that.
MR. OSTERRIEDER: Or tell me if we're not
answering the question, certainly.
MEMBER WALLIS: Adrian, you talk about
alternative break size. This is really in the old
deterministic world where you have sort of specified
things you have to consider. But in a risk-informed
world, you really ought to look at all break sizes,
including large break LOCA.
MR. OSTERRIEDER: We do. And we will --
MEMBER WALLIS: Make an assessment --
MR. OSTERRIEDER: Yes, we do.
MEMBER WALLIS: -- and if you change your
orificing, then your consequences change and your risk
assessment changes for all of them.
MR. OSTERRIEDER: That's correct.
MEMBER WALLIS: And you don't -- there's
no real change. You have to look at the complete
spectrum of breaks.
MEMBER KRESS: There's only a change if
they make modifications to the plans and procedures.
MEMBER WALLIS: Right.
MEMBER KRESS: I think that's what you
have to look at.
MEMBER BONACA: Probably to do that I
think they only leave -- they can assign a very low
probability to that event. So, no, that -- so they
are going to consider that.
MEMBER KRESS: Well, that's almost
irrelevant. I think what the relevant thing is is
what modifications and what changes will result from
the change in the definition.
MEMBER BONACA: Yes.
MEMBER KRESS: It doesn't matter how low
the probability is. It's what --
MEMBER BONACA: What I'm saying is that
they're not going to eliminate the possibility.
They're going to consider it still. They're only
saying that the likelihood of the large break LOCA,
the way he's --
MEMBER KRESS: It may still not contribute
much to risk --
MEMBER BONACA: That's right. They're
saying --
MEMBER KRESS: -- for that sequence, but
the changes to the plant that result from the change
-- you see, the tentacles of design basis accident go
beyond a specific sequence you look at or --
MEMBER BONACA: That's obvious. But I'm
saying that -- that they are not neglecting that.
That's all I'm saying. I'm only saying that since it
is assuming low probability most likely, then it
should not be the design basis event you are designing
it for. And then there are tentacles we have to look
at. I agree with that.
MR. HEYMER: And, in fact, what we're
getting into here is some of the more detailed
discussions, and I think it's a good point at which we
can hand over to the Westinghouse group to get into
some of the more specifics of the technical approach.
And I'll ask Lewis Ward to lead off on that.
MR. WARD: Yes. I'm Lewis Ward with the
Southern Nuclear Operating Company of Bogle, Farley
and Hatch. I'm the Chairman of the Westinghouse Large
Break LOCA working group.
This project started off within the
Westinghouse Owners Group a little over two years ago.
We had an invitation from Commissioner Diaz to each
identify the most single important program that we
could work on that would help our fleet of plants.
And the WOG identified large break LOCA elimination,
I believe is the way it was phrased at that time.
Shortly after that we put together a
steering committee to start through the process of
following up with that letter to Commissioner Diaz,
and really deciding what approach we would take on
going about a rule change. We looked at the rules
themselves as very simple. There's about three places
in Appendix K and Appendix -- in 50.46 and Appendix A
that, you know, we need one sentence basically.
But as we realized right off, there is
much, much more to it than that. There are numerous
Reg. Guides and other documents below that that spin
off from that. So we internally worked for about a
year to try to put together a framework within our
owners group on how we would get our owners to fund
such a program, realizing it was going to take a big
commitment of resources on our part to even launch off
into that.
About a year ago we also got all three of
the other owners group involved with us, and NEI
started doing a coordination role with us. And right
now, all four of the owners groups representing all of
the plants in this country are on board with us on
this project.
What we have done is tried to think
through many of the questions that you're asking and
-- and put some thought behind how we would go about
addressing those issues.
One of the program approaches that we've
put together in our framework is that we really need
an implementation plan that would be exactly what
you're asking about, so that, you know, once we get a
rule change, what do we -- what do I as a licensee --
how do I go about implementing a particular change on
a particular system or component or design basis with
my plant?
And the approach that we are looking at is
to come up with an implementation guideline that is an
industry-wide guideline. Each of our owners groups
feels like, you know, we would put together a
guideline, have it reviewed and agreed to by the
staff, and possibly endorsed by a Reg. Guide, and
probably have a predefined list of the things that we
could go about doing once we got the rule change.
Now, there may be many, many other things
that we did not think about as we went through that
we'd then follow up with the normal licensing process,
either under 1.174 tech spec submittals and that kind
of thing.
That's the general approach we've been
working on. Over the last year, we have had numerous
internal meetings with all of the owners groups to get
this plan more or less laid out. We've started
gathering data to support the specific analysis for
the Westinghouse fleet. And we have proceeded
forward, and we -- we've kept the staff fully
informed.
I believe we've met six times with the
staff over the period of the last year or two, to lay
out our game plan and to get staff's feedback.
The safety benefits -- I'll go over part
of this and Bob will go over part of it. The safety
benefit -- I think all of us recognize that safety has
to be our first priority. We, as owners, investors,
and operators, and citizens who live next door to
nuclear power plants, do not -- did not want to go
down any path that we did not feel like was right from
safety. And so that has been our first focus.
We believe that doing this process will
allow us to put our limited resources on other
activities that have greater risk significance. Right
now, we spend a lot of engineering time, we spend our
highest level of engineering expertise on areas such
as large break LOCA. Our training staff spent a lot
of time on large break LOCA. If you're a licensed
operator, which I had an SRO license at Farley, you
can expect a large break LOCA on one of your requal
drills on the simulator.
As we've seen an event within the last
year not having to do with LOCA, there are more subtle
accidents that are more realistic that the operators
need to learn to deal with more than the "here's the
big one," "I know how to deal with this one," and we
go on.
Surveillance testing -- we do a tremendous
amount of surveillance testing for tech spec
surveillance requirements that are directly hinged to
large break LOCA, such things as accumulator level
transmitters in containment, very, very tight
tolerance bands, extremely time-consuming dose
activities that would go away or be extensively
broadened.
There's a considerable amount of
maintenance. We do fast starts on diesel generators
every month or more often. It puts wear and tear on
diesel generators; we have to tear them down, overhaul
them, and keep them in shape so they will pass the
surveillance over and over again in the event that
they are needed for the LOCA with a station blackout.
There are design issues that we have to
resolve. We work on design issues every day. There's
a current issue on containment sumps that all of you
are aware of. There are other issues that we have to
deal with on a design basis every day in the plants to
try to get resolved for this very unlikely event.
MEMBER KRESS: Now, if I had two
categories and one of them was safety benefits and
burden reduction, it seems to me like most of those
things you just talked about would fall under burden
reduction.
MR. WARD: They are burden reductions
which, recognizing we have limited resources in terms
of technical expertise --
MEMBER KRESS: Is this a zero-sum
activity? So those resources -- the money and
activities actually go into other things?
MR. WARD: Yes.
MEMBER KRESS: You save money that way.
MR. WARD: No. We don't -- I don't
envision any of us laying any people off. I think it
would allow the people that are doing these activities
to focus on something else.
MEMBER WALLIS: I think he's saying that
the operators would be better trained if they were
trained to face up to real events and not have so much
emphasis on LOCAs. Actually, the plant would be
better. It's not just reducing burden. It's better
use of people and resources.
MEMBER KRESS: I'm not so sure that
redefining the large break LOCA has anything to do
with the training process.
MEMBER WALLIS: Well, I think that's what
you're saying is you have a lot of people spending
time on something which is just very unlikely to
happen.
MR. WARD: Yes, that's correct. And that
was my point.
MEMBER KRESS: You shouldn't do that. I
mean, I don't see that that has anything to do with
this definition.
MEMBER BONACA: But isn't the bigger issue
that -- from what I've seen is that the requirements
of the large break LOCA on equipment are forcing a
very tight margin on equipment. I mean, simply there
isn't time on diesels to wait. You have to start them
and you have to test them cold. And so that's because
you have such a strict requirement coming from the
largest demand, which is the LOCA.
That's true of HVAC systems. They would
have to be reconfigured in -- with them in, like the
clock or -- isn't that issue of marginality of the
equipment that is really the bigger driving issue?
You're bumping limits, you're bumping the tech specs,
you have to find out because you are so marginal in
that your -- the demand is maximum for this, isn't it?
MR. WARD: Yes, that's correct. So when
we hit one of those limits, the -- our expertise works
on that problem, not something else. And that's the
-- it's a zero net sum, I think, but it's -- where do
you want to put the focus? On something that's most
likely never going to happen or something that is
likely to happen.
MEMBER BONACA: Yes. From my experience,
I mean, the problem is always that the plant is just
barely making those requirements of the LOCA.
Therefore, it's easy to bump into, you know -- the
diesel start is 10 and a half seconds, and the diesel
is not starting in -- it starts in 10.6 seconds. And
that one-tenth of a second is just killing you if you
have to do all kinds of jumping around to show that
you can come out. That's really where I see a
significant impact. I mean, from the requirement you
are making. Okay? Just the equipment barely makes it
today.
MEMBER KRESS: So that's one change. And
there must be a list of these changes. And the
question I have is: do those particular changes have
any significance in maintaining the risk profile of
plants to an acceptable level? Somehow I haven't seen
that case made yet, but I --
MEMBER BONACA: I agree totally with you.
That's my thought process, too. I would like to see
at some point in all these presentations the list of
the benefits and what they mean.
MEMBER KRESS: Yes. You know, I --
MEMBER BONACA: And then I could decide --
MEMBER KRESS: Well, it's clear that there
would be some benefits with this. And it's not
exactly as clear that the -- that this thing doesn't
have safety significance. But, you know, it looks --
I mean, the implications is that it doesn't, but I
haven't really seen the case made yet. I've seen, you
know, assertions to it.
MR. WARD: There is a list of benefits on
toward the back of the page that we'll get to in a
little while, and there are many that are -- that's
just a partial list. But I think what we wrestled
with for the qual was the -- it's going to take quite
a bit of effort on all of our parts to work through
this program to develop the list of benefits and the
approach on how we could benefit or how we can achieve
those benefits with the rule change.
And that's why we've been working real
closely with the staff on a -- you know, before we go
commit all of our resources and put together a 100
percent complete package, and then no assurance that
-- that anybody is going to listen to it, you know,
that's why we've been having a continuing dialogue
with the hopes of moving forward on that basis.
But we do have a list, and Bob will go
over them a little while later.
Another point is the consistency within
the regulations. Right now, leak before break is an
approved methodology for certain actions --
elimination of whip restraints, baffle -- reactor
vessel internal -- it's a baffle bolt issue. So it's
been recognized by the Commission for 15 years that
this is approved technology for certain uses.
And it presents us a problem by having one
set of requirements for one application and a
different set of requirements for different
applications. And what we are proposing is something
that would clean up that inconsistency within the
regulations.
Okay. Bob, I'm going to turn it over at
this point to Bob Osterrieder with Westinghouse, who
is our lead manager.
MR. OSTERRIEDER: Okay. What I'm going to
do is try to briefly summarize what our approach is in
redefining the large break LOCA, and then we'll move
on, after I talk about that a little bit, to some of
the example applications that you're asking about.
Essentially, you know, we view this as a
risk-informed initiative based on SECY98-300, Option
3, and as part of that we're envisioning this to be an
optional implementation where you could return your
current licensing basis in regard to large break LOCA.
Adrian already mentioned that there is
essentially three key places where the definition of
LOCA is in the regulations, that it has to be a
double-ended -- if they analyze up to a double-ended
rupture of the largest primary piping, and we're
envisioning changing that in these three places --
Appendix K, Appendix A, and 50.46.
What we intend to do is redefine the
maximum size and the attendant consequences while
maintaining an acceptable margin of safety.
MEMBER APOSTOLAKIS: But let me understand
here -- the issue -- the question was asked earlier
regarding 1.174, and now we have Option 3. I mean,
how do these things play against each other?
Tom, you raised the question of 1.174. I
mean -- yes, go ahead. I'm sorry.
MR. OSTERRIEDER: I was going to say,
we're going to be -- you'll see on the next slide
we're actually looking at 1.174. That's just part of
the Option 3 approach. I think they're all --
MEMBER APOSTOLAKIS: Okay. Well, if you
come to it later, we'll wait until then.
MEMBER KRESS: One of the places where the
double-ended rupture shows up is in the general -- in
the design basis accident, the containment. You know,
that doesn't affect -- hey, guys, you're going to go
in and weaken your containment just because of this
change in rule.
But it might affect future plants if we
did something like this. You know, a future plant
could have a new -- are we going to exempt
containment, or are we going to keep that -- give the
-- make a new design basis for containment? Is that
--
MR. OSTERRIEDER: Well, what we've
discussed to date is, as you indicated, not changing
the actual physical containment but allowing some
operational flexibility. You know, you may be able to
change some --
MEMBER KRESS: You can change your leak
range measurement --
MR. OSTERRIEDER: Yes. And --
MEMBER KRESS: -- for one thing. I see
where that --
MR. OSTERRIEDER: Right.
MR. HEYMER: The CE Owners Group is going
to -- the specific application where they talk about
containment -- and I think that they will be able to
really get into that situation. With regard to new
plants, we think that Option 3 should be kept separate
from new plants. And if you're going to go forward
with new plant regulations and thinking about a
framework for the regulatory regime --
MEMBER KRESS: Try to do --
MR. HEYMER: -- you need --
MEMBER KRESS: -- another -- Option 4
maybe or --
MR. HEYMER: You'll have a totally
framework and you're going to start off with a clean
sheet of paper and really -- really go through it.
MEMBER KRESS: I agree with that. An
Option 4 type thing.
MR. HEYMER: Yes.
MEMBER KRESS: Okay.
MR. OSTERRIEDER: Again, the approach
we're looking at is on the technical justification
slide. We will be using risk-informed technology to
show the low-risk significance of the large break
LOCA. Utilizing Reg. Guide 1.174, we're going to be
assessing the likelihood and the consequences of large
break LOCAs to demonstrate that they're of low-risk
significance and that these --
MEMBER APOSTOLAKIS: And, again, 1.174
utilizes the current CDF and LERF, right? I mean, and
the delta CDF and delta LERF. Are you going to do
this in a generic way? And if you do, what kind of
CDF are you going to use?
MR. OSTERRIEDER: We're doing some generic
work. Each of the owners groups is looking at the
risk significance of large LOCA. We, in fact, have
calculated some new initiating event frequencies for
large LOCA, but even if we hadn't what we are
intending to do in assessing the risk significance is
look at all of the plants, the importance of large
break LOCA, and --
MEMBER APOSTOLAKIS: For each plant.
MR. OSTERRIEDER: For each plant.
MEMBER APOSTOLAKIS: Oh.
MR. OSTERRIEDER: And we're containing
that in a -- at least for the Westinghouse Owners
Group, we're going to put that into one risk
significance document that's going to explain how risk
significant is large break LOCA in terms of core
damage frequency and large early release frequency,
and then addressing the -- the five principles of
1.174. That's what our plan is, and we intend to
submit that for review.
MEMBER KRESS: Well, let me ask you a
question about that. I could envision that -- that
this change would allow you to go to a higher leak
rate from the containment, possibly, because, you
know, you're holding the pressure down lower. And so
for you applying this -- the source term that you have
to apply, you could end up with a lower leak rate.
That implies to me that for other
accidents, other sequences, that it's possible, then,
that in the -- if you shift now to the PRA, that the
frequency which you exceed certain releases of
activity in the low level for -- you don't break
containment. You don't have a LERF. But you still
have core damage of some sort.
Those frequencies are going to increase,
although you wouldn't see that at all in CDF, and you
wouldn't see it in LERF. And it seems to me that
those things are of interest at least. They are to
me. And how would that be dealt with in a 1.174
sense?
MR. OSTERRIEDER: Well, I'm not sure about
in the 1.174 sense, but we will be looking at the
effects of any plant changes. If we do do a change to
the leak rate parameters, then we need to assess does
that affect the calculated dose rates for other events
that use those parameters in the analysis.
MEMBER KRESS: Yes. But that would be in
the deterministic space, in the Chapter 15 space.
MR. OSTERRIEDER: Right. It would --
MEMBER KRESS: Which doesn't, you know --
you have to meet those surely, but -- but it doesn't
show up in 1.174 anywhere, which wouldn't -- you know,
the only place it shows up in 1.174 is a little
sentence that says, "You will also meet the rest of
the requirements, the rest of the regulation."
But I'm worried that CDF and LERF doesn't
capture small releases and doesn't capture late
releases and doesn't deal with things like injuries to
workers and injuries to the population, that there are
less deaths. I worry about those things that it seems
to me like 1.174 doesn't properly capture.
MEMBER SIEBER: Well, there's other
requirements besides 1.174. You have 50.2 that has a
dose-limited defense line, and 50.35, and other
general design criteria which you have to meet anyway.
And that's in a deterministic and absolute sense.
MR. HEYMER: Yes. I mean, at the moment,
we're just focusing on 50.46 and redefining the large
break LOCA. And I agree that those -- those
requirements are still in place, and we would still
have to show that we meet those requirements.
MEMBER SIEBER: That's right.
MR. HEYMER: And the same for OSHA and
other worker safety requirements that -- that are
there. We have to meet those. So we're not looking
at changing those, and so that's what we would still
be governed by, regardless of what you might be --
MEMBER KRESS: Yes. The problem I have is
those other requirements are not necessarily risk-
informed. And we're trying now to go to a risk-
informed process.
MR. OSTERRIEDER: Okay.
MEMBER KRESS: But, anyway, that's a
personal problem I have. I don't want to --
MR. OSTERRIEDER: One other --
MEMBER KRESS: -- dwell on it.
MR. OSTERRIEDER: One other point.
There's no guarantee that you can reduce your leak
rate testing because a lot of plants are governed by
steam line break pressures for --
MEMBER KRESS: Yes, that's right.
MR. OSTERRIEDER: You know, so if you have
to look at the entire picture for your plant and
decide --
MEMBER KRESS: So it'll be plant-specific.
MR. OSTERRIEDER: Sure. You need to
holistically evaluate any potential plant changes.
Okay. The second part of -- after we
evaluate the risk significance, and demonstrate how
risk significant the event is, then we'll also be
relying on a deterministic piece of looking at leak
before break analysis to justify the break size that
we'll be submitting to the staff for approval as the
new maximum break size.
And then we'd have to do further analysis
to evaluate real plant changes as a result of the rule
change.
MEMBER KRESS: That's what I want to see.
MR. OSTERRIEDER: And just to kind of
follow up on what we mentioned before, we're already
allowed to use leak before break and not analyze
certain aspects of the plant for the full double-ended
guillotine break. And that is, GDC4 allows for
application of leak before break to high energy piping
to -- involved with the evaluation of the dynamic
effects.
And a number of plants have applied this
for main coolant piping, pressurizer surge line, and
a few other examples here.
MEMBER KRESS: Has leak before break been
approved for the big-sized pipes that we're talking
about?
MR. OSTERRIEDER: We've had leak before
break approved for certain applications down to and
including I believe we even have a six-inch approved
for the plants.
MEMBER SIEBER: But that's just for
Westinghouse, the combustion plants, right, and B&W?
But not for --
MR. OSTERRIEDER: Well, it's not a generic
approval. Each plant may have different --
MEMBER SIEBER: Right. Gets it on its
own, right.
MR. OSTERRIEDER: Right. And submitted
their own, you know, work and gotten it approved for
that particular plant.
Okay. What we intend to do is take the
existing leak before break work that's already been
accepted for certain applications and extend that to
other applications beyond the dynamic effects. And
what we have envisioned was using the same methodology
that was used in those cases.
And then what we would do, depending on
the maximum size that the -- that you're looking to
put as the largest LOCA size you must analyze, we'd
perform -- potentially perform leak before break on
additional lines, if you hadn't already covered those
lines with your existing leak before break work.
And we're intending to consolidate and
make more efficient the review by justifying one
maximum size, at least for the owners -- Westinghouse
Owners Group, and each owners group is going to
decide, you know, how do they want to approach that.
MEMBER KRESS: I guess implied in that
approach, then, is if you can just -- if you can
invoke leak before break, that that renders the
frequency of failure of those pipes that you invoke it
for to a low enough value that they wouldn't show up
significantly in the risk profile? Is that the
MR. OSTERRIEDER: Not -- I believe it's
slightly different than that. I believe it's more the
-- the leak before break is a -- a demonstration that
you will detect this before you get large LOCAs, and
the large LOCA may never --
MEMBER KRESS: Plus, you reduce the
frequency way down, because you're going to detect it
in the --
MR. OSTERRIEDER: Well, because it's based
on frequencies and propagation of cracks and to
potential leaks, and then the leak detection
capabilities of the plant. But the frequency itself
that you would use in your PRA, we have recalculated
frequencies using fracture mechanics. But you
wouldn't have to do that.
I mean, the main purpose of the leak
before break analysis is to support the idea that we
show that large LOCAs are not risk significant, and
then we show that we have some mechanism to evaluate
the actual plant and that we will not have a large
break before you would detect it.
It sounds like I'm not answering your
question.
MEMBER KRESS: Well, it sounds like you
answered it -- yes, to what I said, but I guess --
CHAIRMAN SHACK: But I think you will end
up essentially calculating a frequency of rupture as
a function of pipe -- that will -- that will go into
your PRA evaluations of delta CDF and delta LERF.
MEMBER KRESS: My question is if that
number is below, say, 10-6, then you say, "Well, we
won't worry about that pipe." If it gets above --
around 10-6, we'll say, "Okay. That may be the size
we're dealing with for large break LOCA." I was
wondering if that was the rationale.
MR. OSTERRIEDER: Well, that's part of the
approach that we've done in the Westinghouse Owners
Group and we're going to be submitting is looking at
the probability of all these different pipes in the
plant leaking above a certain amount, which, you know,
that will define the size. And we're looking at all
the pipes and, yes, that's exactly what we're doing.
MEMBER KRESS: Okay. That was the nature
of my question.
MR. OSTERRIEDER: Okay. Again, one aspect
that, you know, we've talked about, if -- if you
change your maximum LOCA break size, the question has
come up, "Do we need to identify other events that we
possibly don't analyze now?" And we believe that you
should currently be analyzing all of the significant
events for different plant systems, but we do need to
make sure if we lower the maximum size that we haven't
in the past said we don't need to analyze a certain
event because it's bounded by this.
We need to make sure that we don't now
have that resurface, and then we -- you know, so we
may end up having to do additional analysis. That
we're looking to do that as part of our comprehensive
program.
Adrian already mentioned that following
the rule change plant-specific changes to the maximum
size would require a submittal and approval of the
NRC. So you'd start with the rule change, allowing
you to change the maximum, and then you would have to
get approval from the NRC to --
MEMBER KRESS: So the rule wouldn't
specify the size.
MR. OSTERRIEDER: That's correct. We're
envisioning a broad rule, and each owners group would
need to decide on what size is appropriate based on
their designs.
MEMBER KRESS: So this could be a plant-
specific size, depending on --
MR. OSTERRIEDER: Yes. We'd envision each
plant submitting with a certain size, and we would
envision a lot of plants may submit with the same
size. But it depends on the plant design and the --
MR. WARD: Really, we're looking at fleet
size, the Westinghouse fleet. We would -- we are
doing a scoping study now, just have one size for the
whole Westinghouse fleet.
MEMBER KRESS: One size fits all
Westinghouse.
MR. WARD: Yes. And that is --
MEMBER KRESS: One size for --
MEMBER WALLIS: Can we talk about what you
mean by "size" now? If you say -- say, a six-inch
break, do you mean a six-inch pipe break, or do you
mean a break of an area in a bigger pipe, or what kind
of thing are you thinking of?
MR. OSTERRIEDER: We are envisioning a --
we pick a certain break flow rate, and then we would
analyze any leaks in any piping sizes to begin --
MEMBER WALLIS: It ought to say that
you've sort of got a six-inch pipe breaking and
snapping off or something. It's the same thing as the
equivalent area somehow opening up mysteriously in an
18-inch pipe or whatever, 30-something-inch pipe or --
MEMBER KRESS: It seems --
MEMBER WALLIS: They seem to be completely
different beasts.
MEMBER KRESS: Well, it seems like the --
the concept ought to be leak -- how fast you leak.
MEMBER SIEBER: Yes.
MR. OSTERRIEDER: Right. And I think I
was trying to say that. We're looking at how much
your leakage rate is. In fact, if you look at double-
ended guillotine ruptures, the risk significance is
really low, really low, for the double-ended ruptures
of any -- the initiating event frequencies are orders
of magnitude lower.
When you look at the leakage rates, at a
certain leakage rate for all of the different sizes,
then it does raise the frequency up. So that's what
we're looking at -- that, not just the double-ended
rupture of all of this piping.
MEMBER KRESS: Yes. It seems to me like
you ought to get away from that concept, and the rule
ought to specify a limiting leak rate, or something
like that.
MR. OSTERRIEDER: Right. Although we
think the rule should specify that you're allowed to
justify and determine what the leak rate is based on
your design.
MEMBER KRESS: Based on the design. Yes,
that's --
MR. OSTERRIEDER: I mean, we really
wouldn't want to have the leak rate in the rule
directly.
MEMBER KRESS: Yes, that's what I had in
mind.
MEMBER WALLIS: It matters where the leak
is.
MEMBER KRESS: Well, it certainly would,
yes. But that would be part of the rule, too.
Somehow that would be captured as --
CHAIRMAN SHACK: It's part of their
justification for the size that they pick, yes.
MR. OSTERRIEDER: That's correct. Where
we think the breaks are, and so forth. That's
correct.
MEMBER KRESS: Your justification would
have to be in terms of risk some way.
MR. OSTERRIEDER: Well, we're doing a
justification based on the risk, and then we're
supporting it with the leak before break deterministic
work.
Okay. Once you would have a specific size
approved for your plant, the licensee then -- any
additional plant changes or benefits they would go to,
and we're going to give a few examples here in a
minute, would follow the appropriate plant change
control processes, because we've had a discussion, you
know, do the plants -- do the licensees need to submit
future changes? And we believe that the processes in
place should dictate that.
And if you're falling -- say you want to
change something in the technical specification,
certainly you need to submit that or a current
guideline. So you may be able to allow the change in
technical specification, and I'll show -- well, we
might as well just go to the examples on the following
page.
Many of these are technical specification
numbers. So approving the rule change and even
accepting the new break size for that plant does not
mean the plant can go in and just make these changes.
They still have to follow all of the rules.
And we've talked about increased diesel
generator start time. You know, that's typically in
the tech specs, and this would allow you --
MEMBER KRESS: Explain to me, once again,
why they can't already do that.
MR. OSTERRIEDER: Because there's a limit.
They have to show that they're meeting the design
basis requirements, which includes large break LOCA,
which that's the event driving the quick diesel start
time.
MEMBER KRESS: I mean, could it -- when
they come in for a change to the licensing basis,
couldn't they -- couldn't that be part of the change
request?
MR. OSTERRIEDER: It could if they had a
basis. But if you have to analyze large break LOCA,
depending on which methodology you're using, you may
not be able to justify much longer start times.
MR. WARD: Right. And I have to get flow
to the core in 40 seconds, or whatever, to meet a
large break LOCA, which means diesel has got to start
at 10, the pump has got to sequence on in the next
five, come up --
MEMBER KRESS: And what I was saying, you
come in for -- request to increase that start time,
and your justification is not that it meets the
requirements, the justification is I don't need it
because of these risk considerations. And 1.174 plus
the other --
MR. OSTERRIEDER: Well --
MEMBER KRESS: -- it seems like that's a
perfectly legitimate thing to do under the --
MR. OSTERRIEDER: Well, we felt in
assessing the different options that the rule change
was a more holistic approach that would also
consolidate review times, etcetera. We could come in
with a bunch of exemption requests.
MEMBER KRESS: Okay. Now that, to me, is
a different reason and probably a valid one. It gives
everybody a start on the same page.
MR. OSTERRIEDER: Right.
MEMBER KRESS: So that's --
MR. OSTERRIEDER: Right.
MEMBER KRESS: Okay.
MR. OSTERRIEDER: Again, I'll just --
MEMBER APOSTOLAKIS: Didn't we just say,
though, that they would still get to come back and
request changes on individual units?
MEMBER KRESS: Yes. But at least it gives
them all sort of a systematic and consistent approach
to it, I think.
MR. WARD: I would have to come in and get
an application for my unit to apply the new rule as my
design basis. Now, as part of that, if -- if I could
change my diesel start time, and if it was in the tech
spec, I would have to have that in as the tech spec
submittal.
If I had already taken that specific
number out because I had adopted the approved tech
spec, and the specific number is not in the tech spec
but it's in the bases, then I would not have to come
in for that specific approval after I got the design
basis approval. That's how we envision it.
MEMBER APOSTOLAKIS: And I guess a lot of
the generic technical work will be done by the owners
group rather than individual --
MR. WARD: Yes.
MEMBER APOSTOLAKIS: -- licensees. That's
really a great benefit.
MR. WARD: Right.
MEMBER APOSTOLAKIS: I mean, in principle,
one could use 1.174 to come and request all of these
changes. But then each application would have to go
through a reevaluation of the large break LOCA
essentially. That's what you're saying.
MR. OSTERRIEDER: Right. That's right.
And there is justification, and I'm not sure if you'd
need an exemption request each time or not. You
would. You would, I guess, versus allowing you in the
rule.
MEMBER SIEBER: You would need a very
extensive exemption request because of all these other
deterministic rules that are out there as part of your
license conditions or, you know, if you have a --
well, a 104-type license, you have that, and so that
would not be a simple thing.
MR. OSTERRIEDER: I really -- I'll just
point to a couple of examples here. I guess in the
interest of time I won't go through them all unless
you have specific questions.
The third item on here -- we've talked
about the second item, which is flow balancing. We've
talked about -- you know, this list of some of the
things that we were looking at that plants may do or
desire to do after this rule changes.
And in the area of accumulators, for
instance, we're looking at potentially some relaxation
in the tech specs where now if you're outside of a
spec on boron concentration or -- or water volume, you
may have to shut the plant down and do a plant
transient, whereas it's not a risk-significant event.
These are essentially relied on for the large break
LOCA.
So we're looking at some, again,
relaxation of operating parameters to avoid
potentially plant shutdowns, which -- obvious economic
benefit, but we believe there's a safety benefit of --
of the thermal cycling on the plant when -- when it's
not a risk-significant reason to be shutting down.
MEMBER SIEBER: Is that a real phenomenon?
I can't ever remember a plant shutting down because
accumulators were out of spec.
MR. WARD: The shut down is not such a big
risk as the startup.
MEMBER SIEBER: Well, you have to fiddle
-- you have to fiddle on startup to get it right.
MR. WARD: On one of my units we had a
two-day delay on startup last year with one
transmitter, trying to get it within a quarter of an
inch.
MEMBER SIEBER: Yes, it's pretty tough.
On the other hand, I can't remember a plant being shut
down because of that. Of course, I don't -- I don't
know every event that has happened in the last 30 or
40 years either.
MR. OSTERRIEDER: But even if they don't
shut down, there may be a lot of work spent at the
plant if they're on the edge as far as, you know, the
volume spec. You know, you certainly need to meet
their specs, but they may be doing -- spending effort
in dealing with the idea that they're close.
MEMBER SIEBER: My point is your slide is
sort of misleading to me.
MR. OSTERRIEDER: Well, the intent of the
slide really is just to give you some idea of the
things we're looking at. These -- you know, we need
to assess these down the road, and I agree with you.
I don't want to mislead you, but I don't want to
mislead you and not put something like an accumulator
spec change and then have you come back later and say,
"Geez, if he was thinking about that, we should have
put it on the list."
So we just -- we're trying to get it on
here to give you some examples. And, really, I guess
I wasn't planning on talking any specifics on this
anymore unless you have a specific example you'd like
us to talk about because I think we're going to get
into some more examples.
MEMBER WALLIS: But if the large break
LOCA went away, would the fan cooler water hammer
problem go away?
MR. OSTERRIEDER: We would have to assess
the fan cooler water hammer problem. We would have to
assess what's driving some of these issues and whether
or not they're prudent --
MEMBER WALLIS: You haven't gotten that
far yet to reach a conclusion?
MR. WARD: I think there's a high
likelihood that one may go away or get -- get better
anyway. We've added a number of relief valves on
containment penetrations for that reason. That could
have been avoided.
MEMBER KRESS: When the agency was
redefining the source term for use with the design
basis accidents, what they did was speculated on what
possible changes might result if a plant opted for the
new source survey. And then they took those changes
and calculated the risk impact of those, and then made
the decision whether or not that risk impact was
significant enough to do or allow a new source survey.
This sounds like it's very similar to that.
MR. OSTERRIEDER: It's the same thing.
What you don't see here is part of our activities
throughout this program is to identify and assess,
from a risk perspective and from a deterministic
perspective, some of these changes that we have on
this list. We fully intend to do that as part of our
demonstration analysis, but we --
MEMBER KRESS: That's down the road some.
MR. OSTERRIEDER: Well, we need to get
endorsement that we think that the -- that the
philosophy of the rule change makes sense. We believe
it fully does, and we're hoping to get endorsement so
that we could go ahead and proceed with that work.
But, you know, we have to evaluate the risk, but
that's what is currently planned in our activities.
MR. HEYMER: I'd also like to point out,
as Bob said before, that there are specific control
requirements imposed on licensees for making changes
today, and that if you move forward with the large
break LOCA, then want to go and do a change, you have
to meet those control requirements which may or may
not require you to go to the NRC staff to seek prior
review and approval.
And under the current process, you've got
1.174 from the plant-specific basis that would govern
that, so --
MEMBER KRESS: And we've got 50.59, of
course, so --
MR. HEYMER: So I think you're covered as
regards, can you overstep demand?
MR. OSTERRIEDER: Right. And just to kind
of try to bring this to closure here, our part, the
safety margin area, we've talked about most of these
items. As part of the risk-informed approach, 1.174
Reg. Guide, I mean, you're assessing defense in depth.
So that's going to maintain defense in depth.
And what we're doing, we've already
mentioned that, you know, we're going to be looking at
the CDF, the LERF, the effects on the health and
safety of the public. From that perspective, we're
not eliminating LOCA from these designs; we're looking
at, you know, retaining LOCA, just defining what the
maximum size is allowable.
And we believe that other design basis
accidents continue to maintain adequate margin. You
know, the idea of, do we need to look for additional
accidents? I mean, these accidents should already be
on the table, and that's what we're going back to look
at and make sure that we're not increasing the
importance of some event that we said was bounded
before.
And, again, we feel this focuses our
resources on greater risk-significant activities.
MR. WARD: So just to quickly wrap up our
part of it, we believe that this approach will
maintain the typical margins of safety. We do believe
this will bring consistency within the regulations.
It will help -- if we can follow this approach, it
will help reduce the amount of resources that the
staff and us have to spend on Option 3. And that
helps our efficiency, our manpower efficiency, and
effectiveness of the regulatory process.
We believe large break LOCA redefinition
is the preferred industry approach on Option 3. We
have looked at the other options that have been
floated around, and large break LOCA is the only one
that really makes sense to us to approve. And we do
have industry consensus on this one, on this one.
This is one we would like to move forward with.
But like Bob said, we need some assurance
that before we do another two years of work and invest
a tremendous amount of money in it that we're going to
have a success path to get there.
MEMBER WALLIS: Well, it seems to me you
read some -- I mean, this sounds reasonable, but then
I haven't seen the numbers. And it may be that some
of the gains may turn out to be small, other ones may
be bigger than you thought, and so on. Until we
really get an evaluation of them, it's hard to make
the decision.
MR. WARD: They put so many tentacles into
everything that I don't think any of us can envision
what all it may impact down the road. We'd like to
lay out a framework on how to use it.
MR. OSTERRIEDER: But we are doing the
quantitative work regarding risk significance. This,
in the near term -- we're currently scheduled for a
July submittal of the risk-significant arguments, at
least from the Westinghouse Owners Group.
MR. HEYMER: Okay. With that, we get on
to a presentation from the Combustion Engineering
Owners Group. Dave Bajumpaa from Millstone will go
over some of the --
MEMBER APOSTOLAKIS: Adrian, just a
second. Are we going to go until 12:00 with the
meeting with industry?
CHAIRMAN SHACK: Yes.
MEMBER APOSTOLAKIS: I thought maybe we
should take a break, then.
CHAIRMAN SHACK: Yes. I think there seems
to be a groundswell for a five-minute break here -- a
10-minute break.
(Whereupon, the proceedings in the
foregoing matter went off the record at
10:56 a.m. and went back on the record at
11:06 a.m.)
MR. BAJUMPAA: Good morning. I'm Dave
Bajumpaa. I'm a Senior Engineer in the Nuclear Fuel
and Safety Analysis Group at Northeast Nuclear Energy
Company, Millstone Nuclear Power Station.
I'm here this morning to -- actually, I
work in the deterministic thermal-hydraulic analysis
area, which includes the FSAR Chapters 14 and 15
accident analyses. I'm here this morning to present
the CEOG position on large break LOCA definition.
And as we talked about earlier, large
break LOCA -- by "large break LOCA" redefinition we
mean to -- we mean the use of leak before break
technology to really define a maximum mechanistic
break size that we need to analyze in a design basis
space, and then continue to analyze the spectrum of
LOCAs up to and including that maximum break size.
As we've talked about already this
morning, and as you well know, GDC4 currently allows
for the application of leak before break analyses to
eliminate dynamic effects associated with the large
break LOCA. And as part of risk-informing 10 CFR Part
50, we would -- we endorse extending this science to
the remainder of the 10 CFR Part 50 regulations.
As we talked, GDC4 doesn't specify a
specific break size. It generically -- essentially
allows a -- it's a generic statement to the effect of
essentially having the individual licensees justify
a maximum break size that needs to be analyzed.
We at the CEOG support the consistent
application of this large break LOCA redefinition
through both the Appendix K and the containment-
related analyses.
Adrian, Lewis, and Bob talked this morning
about -- identified I think some of the safety
benefits and the programmatic approach for the large
break LOCA redefinition. The CEOG approach that we
would take would be very similar to what -- the WOG
approach as has been discussed previously already. We
would continue to use risk-informed technology to show
the low risk of large break LOCA, use leak before
break analysis to justify a maximum break size, and
then continue to analyze the spectrum of LOCAs up to
this maximum break size.
The next slide, please.
There's, again, two major areas where we
see extending the application in this redefinition of
large break LOCA to 10 CFR 50.46, the ECCS performance
analysis, and then to the containment-related analyses
that offshoot from the 10 CFR Part 50 regulations.
And this morning I'd like to just present
-- for the remainder of my presentation, I'd like to
just discuss -- focus a little more on the containment
area, as those areas tend at times to be a little more
subtle.
Next slide, please.
I think the first two bullets on this
slide are pretty obvious to -- the most obvious here
is that we look at containment-related LOCA design
limits. We're looking from the perspective of peak
containment pressure and inside containment EQ
temperature profiles.
The next few bullets there are dealing
with the -- a more subtle analysis that we have to do.
We analyze -- as for peak containment pressure, we'll
analyze to maximize containment pressure and
temperature. But we also have to perform an analysis,
a separate LOCA analysis, that looks to maximize the
close cooling water system temperatures.
It's a similar kind of containment LOCA
analysis, except that we'll -- instead of using a
fouled set of CAR coolers to maximize the containment
temperature and pressure consequences, we'll actually
use clean CAR coolers to put as much energy into our
closed cooling water systems as possible. And you
continue to use a fouled heat exchanger on our closed
cooling water to service water; that's our ultimate
heat sink.
So we do two distinctly different large
break LOCA containment-related analyses. The analyses
that we do to maximize the RBCCW -- I'm sorry, I use
RBCCW because that's reactor-building, closed cooling
water system. That's what I call it in my plant, so
I apologize if I stumble through that.
Some of the key parameters we analyzed,
design limits we analyzed for peak RBCCW temperature-
related effects are the -- our containment -- our
safeguards rooms, our ECCS and containment spray pump
temperature profiles. They're for the room
temperature profiles that are in a building, and
that's going to affect the EQ of our HPSI/LPSI
containment spray pumps -- safety injection and
containment spray pumps.
Other design limits that we have to
analyze with this peak -- related to this peak RBCCW
analysis are closed cooling water inlet and outlet
temperatures at the different components in the closed
cooling water system at the containment air recirc
coolers, the shutdown cooling heat exchangers, and the
spent fuel pool cooling heat exchangers.
Again, it was brought up a little earlier,
the generic -- other issues related to containment
design limits, issues related to the Generic Letter
96-06, potential water hammer loads associated with
the LOCA with the concurrent LMP, and the potential
for voiding in the car coolers. And then, once you
resequence your closed cooling water pumps on, you
will get some significant hydrodynamic loads in the
CAR coolers.
Some of the other components -- the
subcompartment pressurization analysis. That's a
traditional design basis analysis where we looked at
the double-ended guillotine, and we looked to apply
this large break LOCA redefinition effort in order to
-- to limit the maximum break size we analyze.
Another issue -- the last bullet on that
slide is I've got a relatively significant issue
that's still out in the Generic -- I guess it's
Generic Safety Issue 191 dealing with debris
generation in the transport over to the -- with
regards to the containment sump screens and our
containment sump design verification.
We believe that the application of large
break LOCA redefinition is appropriate in this arena
as well as the rest of the containment design areas
and 50.46.
Looking at the next slide, the -- if we're
looking at the containment-related design limits, we
typically have very little analytical margins of these
design limits. What I will calculate for a peak RBCCW
temperature, say at the outlet of my CAR cooler, is
the actual limit that our CAR cooler piping can
handle.
I don't have any margin in many of the --
related to many of these design limits. Changing
these design limits is costly. For example, the CAR
cooler outlet temperature, if I have to increase that
containment temperature, I have many, many, many
calculations of structural supports and many stress-
related calculations that have to be redundant.
It's very expensive to us, so it's --
analytic margin is a very good thing to have. So if
we have increase analytic margin, we can accommodate
some unforseen plant problems that we run into on a
day-to-day basis at our plants.
And looking at -- you know, trying to
quantify some of the margin, you know, if we look at
the containment pressure design limit, the containment
pressure, we'd expect a reduction if we're allowed to
limit our -- apply -- redefine our large break LOCA.
We'd expect about a 10 percent increase or a 10
percent reduction in containment pressure, which would
increase our analytic margin by that 10 percent.
We wouldn't look to change any of the
design -- containment design, you know, thicknesses or
any kind of structural integrity of the containment.
But we'd use it for -- use it to get that and
establish the analytic margin.
Next slide.
I'd like to look at a little more detail
on containment design pressure here, just to show you
what typically is out there. This is actually related
to my plant at Millstone. These specific numbers are
related to my plant at Millstone but are very similar
to the rest of the CEOG fleet.
The containment design pressure I have is
a 54-pound containment design. My peak calculated
containment pressure falling in my LOCA is 52.9 psi.
Peak calculated pressure falling in the main steam
line break is 53.8. So I'm actually steam line break
limited at Millstone, at this point -- Millstone II at
this point.
So redefinition of large break LOCA is not
going to directly gain the --
MEMBER KRESS: Do you consider those two
numbers significantly different?
MR. BAJUMPAA: No, I do not.
MEMBER KRESS: Okay. So both of them
limit the --
MR. BAJUMPAA: Right. Right. But we
would certainly advocate the redefinition of large
break LOCA to get more margin for the LOCA. If we can
get that extra 10 percent or so margin, it would
certainly simplify our design change process to -- to
perhaps allow us to do -- right now, if I had proposed
a design change, because I'm so close to the limits,
I have to look at both accidents. If I had more
analytic -- and I have to look at both accidents
quantitatively.
If I was able to get a little more margin,
analytic margin for my LOCA, perhaps I could look at
that one qualitatively. But I would still have to
look quantitatively at the steam line break.
So there's not a different benefit on
containment pressure here, but it is certainly -- it
does simplify my life in the -- maintaining my plant's
configuration and our design change process.
Some of the other -- next slide, please.
Some of the other areas -- the inside
containment EQ profile. Again, increased analytic
margin between the post-accident calculated
temperature profile and the EQ temperature envelope of
equipment inside containment, that's -- again, we're
going to get some increased analytic margin here if we
are allowed to redefine the large break LOCA.
Similarly, the ECCS room temperature
profiles, if I have a -- if I am able to limit my
break size, for example, to something as large as
branch line break, up to something like that, that
would get me some additional temperature margin, so
that I could -- so it's just an increase in margin
there as well.
And the same thing between the CCW
temperature when it might have increased analytic
margin between what I calculate post-LOCA with clean
CAR coolers versus my coolant CCW temperature when
it's in my design limits right now.
MEMBER SIEBER: With respect to the EQ
limits on electrical equipment, how does that benefit
you since you already have the equipment qualified to
the original profile? Is it in replacement parts or
aging life, or how does that come up in -- in some
cases, you may be able to recall.
MR. BAJUMPAA: Yes, there is some
equipment that actually has -- I have in my plant
right now that -- it's definitely an aging issue.
MEMBER SIEBER: Okay.
MR. BAJUMPAA: It also comes from the
perspective of showing the long-term LOCA profile
compared to the test profile and actually analytically
proving that the test profile that the equipment is
tested to bounds the actual predicted LOCA profiles.
I actually have some equipment that is
very marginal that we have to do a pretty
sophisticated analysis internally to demonstrate that
with the double-ended guillotine LOCA that our
equipment would survive that for the 30-day time of
the accident and including them in a four-year life.
So there would be some equipment potentially that we
could avoid having to replace.
MEMBER SIEBER: I think I'll ask no more
questions about that.
MR. BAJUMPAA: Okay.
MEMBER SIEBER: Thank you.
MR. BAJUMPAA: Sure. So I've sort of
established to this point, hopefully, that what we're
trying to do here is increase the analytic margins to
our design limits. Now, what would we do with that
analytic margin? And this next slide sort of leads
into some of the areas in the containment-related
design benefits that we would get here.
One area, we would look to relax perhaps
the CCW flow limits through our CAR coolers. Right
now at my plant I have a very, very small window in
which I can set my CAR cooler flow outlet valves to
get the proper flows. I have to make sure that I have
enough minimum flow through the CAR coolers, so that
I am assured to pull off enough heat removal so that
I don't exceed my containment design pressure when I'm
looking to maximize my containment design pressure.
But I also to make sure that I don't have
too much CCW flow going through my CAR coolers. If I
start having too much flow, I might -- with a clean
CAR cooler, then I might actually increase my peak
calculated RBCCW temperature following the LOCA. And
so I actually have a very tight constraint right now
that I have at my plant to maintain a minimum flow
that ensures the containment pressure is met and a
maximum flow that ensures that I don't exceed the peak
CCW temperature limits.
MEMBER SIEBER: Do you have service water
temperature limits?
MR. BAJUMPAA: Yes. We do have service
water temperature limits. Primarily, with the
containment-related analysis, that comes in from the
ultimate heat sink, the --
MEMBER SIEBER: Okay.
MR. BAJUMPAA: -- sound temperatures, yes.
MEMBER SIEBER: And do you also have RWST
temperature limits?
MR. BAJUMPAA: That is correct.
MEMBER SIEBER: Which often is hard to
meet in the summertime?
MR. BAJUMPAA: That is -- yes, that is
correct.
MEMBER SIEBER: Okay. And those would be
relaxed if you had a smaller break size?
MR. BAJUMPAA: That would -- it would
certainly -- it would help.
MEMBER SIEBER: Okay.
MR. BAJUMPAA: I'm not sure on the RB we
have to look, you know, in an integrated fashion
through all of the -- but, yes, that would certainly
help the containment.
MEMBER SIEBER: I know that Farley had
those problems, correct?
MR. BAJUMPAA: Right.
MEMBER SIEBER: Okay.
MR. BAJUMPAA: Right. So I should have a
bigger window as far as CCW temperature to flow -- to
set a bigger window in my plant right now. I think
I've only got a 25 gpm target window to set through a
CAR cooler, so that I don't have a minimum flow. When
I take the minimum flow, I also have to drop off to
account for pump degradation and instrument
uncertainty.
So I have a very tight window. It's only
like 25, 30 gpm that I can set my CAR cooler outlet
valves within, which is a bit of a challenge and we do
have to fiddle with the valves from time to time to
make those -- to maintain our configuration
management.
MEMBER WALLIS: Fiddling with the valves,
is that done remotely?
MR. BAJUMPAA: No. These are actually
manual.
MEMBER WALLIS: Someone has to go in there
and turn them?
MR. BAJUMPAA: Yes. Somebody would go in
there and turn them, not during an accident, but
during the refueling outages and stuff when -- or
whenever we do anything that might change the
configuration of our CCW system we'd have to go back
and do a flow balance.
MEMBER SIEBER: Well, you end up having a
range of adjustments because as you operate the plant
the CAR coolers become fouled and flow goes down.
MR. BAJUMPAA: Yes. We typically --
MEMBER SIEBER: To some extent.
MR. BAJUMPAA: To some extent, that is
true I guess perhaps for some plants. But on our
plant we currently do cool our CAR coolers with a
closed cooling water system and maintain pretty decent
chemistry there.
MEMBER SIEBER: So you're better off.
MR. BAJUMPAA: We typically don't have
issues with car cooling, fouling.
MEMBER SIEBER: Okay.
MR. BAJUMPAA: Yes. Other potential
applications that we use for this increased analytic
margin we'd get with our potential large break LOCA
redefinition here, I might be able to increase my tube
plugging limits that I have to maintain in my
configuration right now on my CCW to service water
heat exchangers, my shutdown cooling heat exchangers,
and the CAR coolers.
The service water areas, being a raw water
system, that's probably the area where I'd be most
sensitive about tube plugging. Other areas -- as
alluded to a little earlier, I have actually made
submittals on our docket, and we just received a
license amendment regarding the increase in our
ultimate heat sink temperature limits. Actually, it's
more of an issue there that I just went through on my
plant.
Another potential area of use for this
increased margin would be to accommodate any future
potential power upgrades. Because I'm so close right
now with the LOCA on containment pressure, increasing
the power, that event turns out to -- that may
actually put the LOCA containment calculated pressures
higher than the design pressure containment.
If I were to do a power upgrade in the
future with a double-ended guillotine steam line break
there, even though it's up there right up and close,
that's actually limited by my zero power case by the
additional water inventory and the steam generators.
I guess the next slide -- this next slide
really is a summary of actually when Millstone was
going through its 54(F) configuration management
related outages, this is a listing of the areas where
we had analytic problems that we had to straighten out
our configuration. And that resulted in us having to
reanalyze our containment pressurization analysis
associated with the large break LOCA.
We had increased safety injection. We had
to increase our safety injection pump flows. We had
to increase the spent fuel pool cooling heat loads
that were assumed in the analysis. We had increased
containment spray header fill times when we looked at
in more detail.
Small -- very, very small issue here
obviously, the containment paint thickness, but that
does play into -- I have to -- I have a maximum paint
thickness that I have to assume on my passive heat
structures inside containment. And during the 54(F)
outage we identified that the paint micrometer was
thicker than what we had assumed in the analysis.
So that was another contributor to making
us reanalyze that event and in the increased refueling
water storage tank temperature. That's -- those are
things that we had to deal with during our 54(F)
outage, and now we have a solid configuration in our
running well.
Another area where we just wanted to make
one point with this slide, that we really feel -- the
CEOG feels that we should consistently redefine the
large break LOCA across the entire spectrum of the 10
CFR Part 50 requirements. We think that is the way to
go. It's currently acceptable to use LOCA
redefinition for GDC4, and to extend that application
through both the 50.46 and the containment-related
areas, we think that will allow us to take advantage
of some of the potential safety benefits.
I throw up as an example on this slide the
emergency diesel generator start times to -- to try to
enhance the diesel reliability. But that diesel
generator start time is driven by a lot of things.
It's driven by the 10 CFR 50.46 ECCS performance
analysis, where we are looking at getting our high
head and low head safety injection pumps up to speed
in the adequate time to combat the double-ended
guillotine LOCA.
But I also have to have my CAR fans start
my closed cooling water pumps for the containment-
related areas. They need to sequence on the diesel
generator and start to accommodate these double-ended
guillotine LOCAs.
If I am going to increase this diesel
start time, it's a major effort on behalf of my plant
from a design change perspective. I have to look to
make sure that my steam line break containment
performance does not now become more limiting than the
LOCA with regards to maximizing the water hammer loads
on the CAR coolers, the RB temperatures, and things
like that.
So it's -- to -- to -- right now, on
Millstone, my diesel start is tied in right now to --
in order to prevent too many -- too large hydrodynamic
loads on my CAR coolers, I need to start my diesel in
the same -- I can't relax that directly right now
unless I get -- unless I'm able to redefine the LOCA.
Then I will not be able to increase my start time, if
I just apply it to the -- if I apply LOCA redefinition
to the 50.46 area only.
So it's got to be an integrated overlook,
and I've got to look at not only the containment-
related LOCA analysis, I've got to look at the ECCS
analysis, and I've got to look at all the rest of the
Chapter 14/Chapter 15 accident analyses to make sure
I'm not affecting aux feedwater start times for my
loss of normal feed analysis in Chapter 14. So it's
got to be a big integrated look to get this increased
diesel start time and yet still maintain the proper
configuration at my plant.
And I guess the last slide is just a
summary that the CEOG -- as I have indicated and
alluded to in earlier slides, the CEOG does support
the consistent application of the large break LOCA
redefinition throughout the entire 10 CFR Part 50
regulation.
CHAIRMAN SHACK: Let me catch up one thing
back to Mr. Ward. Something slipped through that I'd
sort of like to understand. Although we're doing this
under Option 3, is your selection of the maximum pipe
size going to be basically a deterministic argument
based on the 1061 kind of arguments, that you're going
to have some margin to leak and then some margin on
crack size? Or are you going to go through and look
at the risk significance of various leak sizes?
MR. OSTERRIEDER: It's a combination.
We're considering the risk significance and the
deterministic piece.
CHAIRMAN SHACK: Okay.
MR. WARD: Yes. And, historically, we had
started this before Option 3 became a program. We
wanted to get into redefining the LOCA. So I guess we
have been doing this in conjunction or in support of
the staff's effort to come up with some options under
Option 3.
MR. HEYMER: Terry Rieck from Excelon will
talk to you about the BWR perspective. He's
representing the BWR Owners Group.
Terry?
MR. RIECK: Good morning. For those of
you who don't know who Excelon is, we're now in the
merged organization of Con Ed and PECO, so -- I know
a lot of you don't know how much we're changing out
there in the industry.
But I'm here representing the BWR Owners
Group. Excelon does own 10 BWR units, and so we are
a big part of that owners group.
We formed a committee recently, a
technical committee, to look into Option 3 risk-
informing Part 50, the technical requirements.
Frankly, we were behind on where the PWRs were, PWRs
having started a year and a half ago or so to look
into this. We are behind because we saw we had some
margin in large break LOCA, and we saw where the PWRs
were going and -- and didn't jump in right away.
But within the last few months, we felt we
needed to get more involved to see where the benefits
might be for us. And we formed this technical
committee. We've now met a couple of times, and we
started talking about the same things that the PWRs
were doing.
Very quickly, we got into determining that
redefining the large break LOCA should be our highest
priority. As I said, we kind of dismissed that a year
ago or so, but once we started talking about it we saw
that it met the criteria, the framework, that the
Commission was talking about in terms of Option 3.
That is, there was some significant safety
enhancements that could be made on our plants, but
also have a burden reduction for us, and there was
some cost that we had to incur to get there.
So when you looked at it, when we looked
at it in terms of the PWR approach, we saw the same
safety enhancements that the PWR saw. We talked about
diesel generators starting in 30 seconds instead of 15
seconds, and what would that mean to the reliability
of the diesels to the reliability of the electrical
system.
And we saw that our CDF for our large
break LOCA was way down in the weeds, very low, and
our small break LOCA was higher and other -- other
actions that relied on loss of -- relied on offsite
power were higher, and, thus, the diesels are very
important.
So we ended up with the same safety
benefits and felt we didn't meet the safety of the
plant significantly. And, you know, as we talked more
and more about it, the members of our committee from
the various utilities said, "Well, of course, you
know, the setting up of valves in certain ways and the
ECCS pumps would all enhance the small break LOCA."
And the large break LOCA was so low in CDF, you know,
it might, as mentioned here, increase its CDF. But it
was so low it was an insignificant increase.
We very quickly said, "Yes, we can gain
the safety benefit, and the burden reduction could
also be significant." The same thing is talked about
on the PWR side.
As I mentioned, we do have more margin in
the large break LOCA, so it kept us from getting in as
quickly. And we realized also that we haven't delved
in as much into fracture mechanics probability and
leak before break like the PWRs, so our cost might be
more to get into this than the PWRs.
But what that meant is we would incur a
larger cost for the burden reduction, but it might
also mean an alternate break size, as we may not be
able to justify as small of a break size as the PWRs.
And as each of the owners groups talked about here
today, that's part of our industry proposal is that
each vendor would have to propose their own break size
based on what they could justify. So we realize we
may not be able to get the same break size.
So, again, just to summarize, we felt the
-- that large break LOCA redefinition was our most
beneficial path to go down. But we did make this a
committee that looked at other options, because the
other owners groups had already done that. But we
hadn't had the chance to look and see what might be
our highest priority, and, thus, we did look at other
options and came up with half a dozen other options
that we ought to pursue. So we are looking other
places.
So we see large break LOCA as a high
priority. We see a benefit for that as well as
working with the Commission and the owners groups in
terms of -- of the framework and how it might be
further development through this process in the large
break LOCA.
So our follow-on activities on the second
slide here are to continue in the large break LOCA
arena, better define the safety benefits as well as
our burden reduction, essentially do the cost-benefit
analysis, but also assess the various approaches on
fracture mechanics and leak before break.
So we've got a lot of work ahead of us to
catch up to where the PWR is. But, again, I want to
emphasize that is our top priority in the working
group, but we are also looking at other options. And
we're playing in -- or we're working in this arena to
get more experience in that and know where our cost
benefit might be for other options down the road.
CHAIRMAN SHACK: Would you be looking at
changes in GDC4 on the requirements that you can't
apply leak before break, for example, to piping within
a granular stress corrosion cracking?
MR. RIECK: We have -- the BWRs have had
IGSCC problems, and we have also done some mitigation
to those problems over the years. And so, in our
discussions, we felt that we'd have to do a lot of
justification to show that we have improved on the BWR
pipes. And the smaller pipe size may not be
justified, but the larger pipe size, where we may not
have seen that, we might be able to justify that. So
that's the path that we have taken.
That's all I have to present. I'd be
repeating a lot of what the PWR said, but I made the
points. We are moving down this path.
MR. HEYMER: The B&W Owners Group couldn't
be here today, but they have a similar program
underway. They are supporting this activity. They
support this approach, and they believe, looking at
50.46, that redefining the large break LOCA is the
priority that we should be looking at in this
activity.
And so I guess, in summary, what we've
tried to do today is explain to you -- give you some
background of how we reached our conclusions on 50.46
and redefining the large break LOCA. All owners
groups have evaluated this and have continued to look
at it as we've interacted with the staff over the last
15 months or so and explained our positions and
thoughts to the staff on where we think the emphasis
needs to be placed in improving 50.46.
You heard today some discussion on the
benefits as regards to margin, the operational margin,
scheduling of work activities, the priority of work
activities, and consistency in the regulation. And we
believe that by going down this path of redefining the
large break LOCA we can more effectively focus our
resources on those issues that are safety significant
which will ultimately enhance the overall safety
profile of the plants.
So that's what we came with today. We're
looking to go to the next step, and you heard
Westinghouse talk about evaluations that are underway.
As I said, owners groups have provided funding because
we saw recognition that if we're not going to go down
the large break LOCA path that we need to hear about
that, because we do really believe that it is the
priority and the one that we should be looking at to
get the maximum benefit, both from a resource and a
safety perspective.
And before these folks go off and spend a
lot more money, I think we've got to come to a
resolution of the direction in which we're going.
MEMBER WALLIS: So what are you hoping for
from the ACRS at this point?
MR. HEYMER: Well, we understand that this
was really a familiarization discussion with you to
explain our position, and we look forward to hearing
what the staff is going to say on the matter. And we
understand that there are some papers going up to the
Commission, because as you rightly said this is --
this is a significant step and they want Commission
involvement.
And we hope that it's -- it's the decision
that supports what we've explained to you this
morning, because we think that is the right thing to
do.
MEMBER WALLIS: We have to listen to the
staff and then -- I don't know what the staff will
say. If the staff agrees with you, then it would seem
we don't have much to do. If the staff disagrees with
you, maybe we need to get involved and --
MR. HEYMER: Well, I mean --
MEMBER WALLIS: -- try to sort things out.
MEMBER KRESS: I think it's a question of
whether we agree or not, not whether the staff does.
MEMBER WALLIS: Yes, but we haven't seen
that much. And we've seen sort of the perspective of
you're about to go down this path, and there may be
some nice country to discover if you take the path.
But we haven't really seen much in the way of numbers
or technical analysis or -- I mean, there's hopes that
if you did this you might be able to get something
here. I think we need something a little more
substantial in order to give a solid yes or no or
evaluation.
MR. HEYMER: And I think that what we're
looking for from an industry side is the fact that,
yes, okay, it's worth progressing, and, yes, we can
continue to go down this path, and we will do the
analysis. And we can come back once we've done those
detailed analyses and provide you a further -- a much
more detailed description of what we're finding and
where we're going.
But we feel we should be confident that we
can do that, and we can achieve --
MEMBER WALLIS: Before you plan the
expedition to climb this mountain, you want to know if
there's going to be some taboo or something that says
you can't go there anyway.
MR. HEYMER: Well, yes. And if someone
says that whether you -- you can do what you want but
you're wasting your time, and, you know --
MEMBER WALLIS: Right.
MR. HEYMER: -- then we're going to have
to take a look at doing risk-informed, and are we
actually risk-informing the regulation, because we
think this is a central element and it -- it does send
a very clear signal that this is what we want to move
towards, improving the process.
MR. OSTERRIEDER: Right. In all the
discussions we've had to date, we haven't heard of any
issues brought up that we didn't feel were -- that
were insurmountable. We thought that all of the
issues that were identified to us thus far are issues
we can deal with. And this continues to make sense
and the most sense, but, you know, we just need to
understand if there are issues there and get them on
the table and move forward.
CHAIRMAN SHACK: Well, I think perhaps
this -- we really do need to hear from the staff.
That's the next step. And we'll be doing that at
12:30. We'd like to take a break for lunch now. Come
back at 12:30.
(Whereupon, at 11:42 a.m., the
proceedings in the foregoing matter went
off the record for a lunch break.)
. A-F-T-E-R-N-O-O-N S-E-S-S-I-O-N
(12:32 p.m.)
CHAIRMAN SHACK: We're ready to restart
the meeting, and we'll have the staff presentation.
I guess we're going to have an overview from Tom King,
Mary Drouin, and Mike Mayfield, among others -- the
usual suspects.
MR. KING: My name is Tom King. I'm with
NRC's Office of Research. With me at the table is
Mike Mayfield, Division Director, Division of
Engineering Technology and Research; Mary Drouin from
the PRA Branch; and Alan Kuritzky from the PRA Branch.
If I could have the second slide.
What we want to talk about today is give
you a status report on where we stand on looking at
50.46. We haven't sent you anything in advance.
We're not asking for a letter. This is work in
progress, but I think we've reached a point where it's
worthwhile having some interactions, and at least
informally getting your reactions, your feedback. And
we'll get to a schedule slide later on, and we will be
back several times before this is all done.
This is sort of an initial overview of
what we've done, sort of some general conclusions
we've reached, and some additional work things,
follow-up things we're still working on. And like I
said, we'll be back later to talk about further
progress in this area.
I'm going to give a little introduction
and overview of where we stand. Mike is going to talk
about in more detail redefining the large break LOCA.
And then Alan Kuritzky is going to talk about where we
stand in terms of some near-term conclusions or
recommendations that we're thinking of proceeding with
and the longer term aspects of 50.46. If I could have
the third slide, please.
As you recall, this is work being done
under what we call Option 3, and Option 3 was
basically an effort to go in and look at the
regulations, technical aspects of the regulations, try
and identify what the -- from a technical standpoint
what things were candidates to be changed based upon
risk insights. And those changes could go either way.
They could fill gaps in the current regulations that
risk insights say need to be filled, or they could
remove some unnecessary burden that risk insights say
are not very risk significant.
Our plan in Option 3 is to do what we call
a feasibility study. That is, go to the Commission,
do enough work to go to the Commission and say, "Hey,
here are some things that are feasible to change."
Not give them a rulemaking, but get them to buy in up
front that, yes, we've done enough work to establish
certain changes are feasible, and then the rulemaking
would actually get into the details and the exact
wording of the changes, and go through the normal
process which is usually about a two-year process.
The rulemaking process involves a lot of
public interaction, but we've also tried to have with
-- to work under Option 3 public interaction as well.
And you heard this morning we've had a number of
meetings with the owners groups, and they've all been
public meetings. We had a workshop to try and get as
broad input as we could on this activity, and we'll
continue to do that as we proceed.
MEMBER WALLIS: Are you getting what I
would call real public input?
MR. KING: We have not gotten much from
outside the industry, no. We've invited people, and
some of them attended, but we haven't gotten much in
terms of feedback.
MEMBER WALLIS: I think this is an issue
where, you know, I would really be interested in what
the sensible, informed public would think of about
this sort of change, and not people who have something
to gain by it one way or another but those who -- you
know, whose interests you are trying to protect.
MR. KING: I'd like that feedback as well.
I'm not sure exactly how to get it other than you keep
asking for it and hope it comes in.
MS. DROUIN: We have had at the public
meetings real public. They have come to some of the
meetings.
MR. KING: Yes, but they haven't said
much.
MS. DROUIN: They haven't said much.
MEMBER APOSTOLAKIS: I don't understand
why they are part of -- they use different names,
means --
(Laughter.)
MR. KING: Non-industry.
MEMBER APOSTOLAKIS: Okay. Non-industry.
Much more factual. In fact, to protect the sort of
public that's sitting in front of us. Let's not
forget that. The public has safety.
MR. KING: We're following in all of these
Option 3 activities what we call this framework
document, which we've discussed with the committee
before. It sort of lays out the ground rules as how
we go through and look at regulations and what the
guidelines are for making judgments on whether we'd
make a change or not make a change, and it involves
defense in depth, safety margins considerations, as
well as some risk guidelines.
And the question came up this morning,
well, why not use Reg. Guide 1.174? In theory, any
licensee today could come in with 1.174 and propose a
change to the large break LOCA for their plant, go
through all the analysis it would require as to staff
review and approval.
Option 3 is really trying to make some
generic changes, not require plant-specific stuff in
every case. And that's why the framework in Option 3
is a little different than what's in 1.174. 1.174 is
delta CDFs, delta LERFs. You're making changes from
the current plant risk profile.
MEMBER APOSTOLAKIS: Let me understand
that a little better, Tom, because it's something I
have on my mind. When you change a rule, you can
still use the idea of delta CDF and delta LERF to
evaluate the change you're about to make. In other
words, you are using the idea behind the 1.174, even
though you are not changing the licensing basis of a
specific unit.
MR. KING: When you change a rule,
particularly when it's mandatory on everybody, and you
go through the backfit analysis, you have to show that
the incremental improvement in safety is substantial.
And there are guidelines in the reg. analysis
guidelines, delta CDF, for example, that's used to do
that.
MEMBER APOSTOLAKIS: But in addition to
those, though, shouldn't delta CDF be small? Or do
you think that's covered already by the regulatory
analysis?
MR. KING: If we're imposing a new
requirement, we want to make sure that the improvement
in safety is not minuscule, that it has some
substantial improvement in safety. So there's a delta
CDF guideline in the reg. analysis guidelines.
MEMBER APOSTOLAKIS: I guess now I'm a bit
more -- when you are imposing a requirement, that
means that the delta CDF is negative, is it not? And
that's why the regulatory analysis applies.
MR. KING: But the change is an
improvement in CDF, and CDF goes down.
MEMBER APOSTOLAKIS: Okay.
MR. KING: Yes.
MEMBER APOSTOLAKIS: If, on the other
hand, you change a rule in a way that, really, there
is only relief, then I would expect a positive delta
CDF. So there shouldn't be a question of regulatory
analysis. But there should be a question of whether
the delta CDF, which is not positive, is acceptable.
MR. KING: Yes.
MEMBER APOSTOLAKIS: Which brings into
this 1.174 what I feel the idea is behind 1.174.
MR. KING: Right. Right.
MEMBER APOSTOLAKIS: So that would be one
where you're doing it. But then, this morning --
again, I'm trying to understand how this works -- we
said that, yes, even if you change the rule, then each
licensee will have to come to you with a request to
actually be allowed to change it for their facility.
And that's not clear to me.
MR. KING: Well --
MEMBER APOSTOLAKIS: Why do they have to
do that?
MR. KING: It depends on the rule change
that's made. I guess in an ideal situation, you'd
make a generic rule change, specify a new break size,
and maybe people now would have to submit on their
reload analysis to conform to that new break size.
But they wouldn't have to come in and justify that
break size.
I think what we've kicked around is there
is -- you know, it's not clear we can define a generic
break size for everybody -- a new one.
MEMBER APOSTOLAKIS: Okay.
MR. KING: Perhaps its class of plant-
specific or vendor-specific -- or perhaps it's plant-
specific, in which case you -- I think you heard this
morning that if it's plant-specific, maybe each
licensee -- what the rule change would do, it would
open the door for each licensee to come in and propose
for his plant the new break size that makes sense
based upon his risk assessment, his plant design.
That's one option.
And what this rule change would do,
basically, then, would eliminate having to have an
exemption to the current regulation to do that.
MEMBER APOSTOLAKIS: So they wouldn't need
to concern themselves with Regulatory Guide 1.174 at
all, because that's in the rule.
MR. KING: What we would have to do in
that case is put out a new Reg. Guide that says, okay,
if you're going to come in under this new rule that
opens the door, here's the analysis we'd want to see
and here's the acceptance criteria.
MEMBER APOSTOLAKIS: So you might repeat
some of the requirements of 1.174 --
MR. KING: Yes.
MEMBER APOSTOLAKIS: -- if it is
appropriate.
MR. KING: If people -- right, exactly.
MEMBER APOSTOLAKIS: In other words, you
may request a risk assessment, an evaluation of delta
CDF.
MR. KING: But today, if somebody wants to
come in under 1.174, in this particular case, they're
going to have to ask for an exemption to 50.46. And,
you know, the Commission has told us in pretty clear
terms that if -- if you're starting to make a lot of
exemptions in a certain area, you ought to start
thinking about a rule change. And that's one reason
for not just saying, "Go use 1.174" on something
generically.
MEMBER KRESS: That's the best reason --
explanation that I've heard so far. That's helpful.
MR. KING: Okay. Next, side 4.
Where we stand -- within the past four or
five months, we've pretty much been in an information-
gathering mode, gathering information from the
industry, from the owners groups you heard this
morning, in terms of what their thoughts are, what
work they've done looking at this issue, and we've
been doing our own work in-house looking at various
options and the technical basis for those options,
using our framework and looking at the risk
information associated with those options and how it
stacks up against the framework.
And we're now at the point where we've
gathered enough information where I think we're
settling in on some at least preliminary conclusions
and preliminary approaches, and that's what we're
going to talk about today.
We're going to talk about what we consider
is feasible to go forward with in the short term. By
the "short term" I mean with a paper to the Commission
in June, and then some things that we think are
potential longer term improvements, but we need some
more time to work on those. So you'll hear about both
of those.
If I can have the next slide we'll talk
schedule a minute.
This is sort of the first time we've come
forward in a public setting and talked about what
these preliminary conclusions are and what we want to
pursue. So we think, clearly, we want to have at
least another public meeting before June to talk about
these in more detail and get some feedback from others
besides this committee. So that is factored into our
schedule.
I think we want to come back to the
committee after that and, you know, tell you where we
stand, having factored in input from that public
meeting and additional work we're doing. We would
intend to get you a draft Commission paper to look at
in mid-May, and then we would want to come back to the
committee at your June full committee and ask for a
letter on this before we actually send the paper to
the Commission the end of June.
MEMBER APOSTOLAKIS: So the subcommittee
meeting will follow the SECY paper. We'll have an
opportunity to read the SECY paper before the meeting,
right? Or is that after?
MR. KING: We are proposing another
subcommittee before you actually get the paper, and
then give you the draft paper before the June full
committee.
MEMBER APOSTOLAKIS: The reason being?
MR. KING: The reason being that I thought
it might be useful before we actually take the time to
write this stuff down in a SECY paper to have some
more discussion. If, you know, the committee has
different views and you want to adjust that schedule,
we're certainly open to discussing it. So we can come
back to that maybe at the end of the day today.
MEMBER APOSTOLAKIS: But isn't it
typically too short a time? If the subcommittee makes
comments, I mean, May -- mid-May sounds like too short
a time for any real changes in the document. Should
the subcommittee meeting be perhaps late April or --
my feeling is with you guys that two weeks means
nothing.
(Laughter.)
And I'm learning from --
MR. KING: Is that a compliment or is that
a --
(Laughter.)
MEMBER APOSTOLAKIS: I'm not saying that,
you know, blaming you for anything. But I realize
that, you know, there are reviews to be made and
various offices to concur, and all that. And two
weeks is not enough.
MR. KING: I don't think it's worth
getting back together until we have our next public
meeting. I think after that would be the right time
to get together.
MEMBER APOSTOLAKIS: So, in essence, what
you're saying is that this June letter might not be
real. It could be a July letter.
MR. KING: That's always an option. We
send our paper forward and you follow up in July with
a letter. That's an option also. But what we --
ideally, we'd like to have a letter before our paper
goes up so we can reference it in our paper. And the
steps to get to that point are negotiable, but I think
we'd like to have our public meeting first and then
get back and have some further interaction.
MEMBER APOSTOLAKIS: Well, anyway, if you
don't think it's a problem, that's fine. But I
thought it would be a problem. You know, May -- mid-
May, early June, I mean, it sounds like things are
happening with lightning speed there.
Now, you are committed to give something
to the Commission by June 29th?
MR. KING: Yes.
MEMBER APOSTOLAKIS: Okay.
MR. KING: That's our current commitment
date, due date, to the Commission.
MEMBER APOSTOLAKIS: And what is it that
says that the public meeting cannot be in early April?
MR. KING: There's nothing that says that,
other than getting everybody's calendars coordinated
and picking a time and scheduling it.
MEMBER APOSTOLAKIS: Well, that would make
everything else much easier.
MR. KING: Yes. And Mary Drouin will be
setting that up.
(Laughter.)
MEMBER APOSTOLAKIS: So if it's not early
April, we know whose fault it is.
(Laughter.)
MS. DROUIN: Alan's.
(Laughter.)
MR. KING: Okay. Slide 6. As I said,
we're going to talk basically about two things. We're
going to talk about the work we've done looking at
risk-informing the definition of the large break LOCA,
and then we're going to talk about other things that
are on -- that we've considered beyond just redefining
the large break LOCA. And, again, these are going to
break down into some near term and some longer term
changes.
And with that, I'm going to turn it over
to Mike, who is going to talk about the large break
LOCA redefinition area.
MR. KURITZKY: Actually, before Mike takes
over the large break LOCA redefinition, I just wanted
to mention something up here which I hope after Tom
has clarified things for you this isn't going to
further confuse them. But we had -- under this Option
3 we had come up with a number of various options for
risk-informing the technical requirements of 50.46 and
the large break LOCA and associated GDCs, etcetera.
And we have about three of those options
for various takes on redefining large break LOCA, and
then another seven or so involve actual changes to
just the technical -- the various technical
requirements.
These three right here are the three that
we have that all fall into the category of redefining
large break LOCA. And when I explain the difference
between the three, hopefully that will further clarify
what Tom was saying and not further muddy it up.
But in the first -- when we were going to
redefine the large break LOCA by plant, you know,
except what we do there is a plant would then be
permitted to define its own maximum design basis LOCA
using leak before break and probabilistic fracture
mechanics analyses. It's very similar to what
Westinghouse or NEI industry was discussing earlier
today.
MEMBER KRESS: Tell me, how would they do
that? That sounds like a task that I see has no
relevance to this issue.
MR. KURITZKY: Well, this one is focused
on -- would be focused on the frequency of the break.
MEMBER KRESS: Yes, but that's just a PRA
issue.
MR. KURITZKY: No, it's actually even
before that. It's -- well, actually, the next thing
I'm going to tell you about, actually bring more of
the -- the carrying the risk. This one would be on
the frequency, and this --
MEMBER KRESS: Well, what you're doing is
using leak before break technology to -- to refine the
frequencies of certain leak sizes, to the PRA issue.
I don't see its relevance to this issue of defining a
design basis large break LOCA. I mean, unless you're
going to say something which I think is the wrong
thing to do, that -- if that frequency turns out to be
below 10-6 or some magic number, then we -- then we've
screened all pipes or leakages bigger than that, which
is -- which sounds like what it sounds like.
But I think that's the wrong way to go,
because the issue is really if I change this design
basis LOCA definition, what changes will I expect to
see in the plants? And are those changes acceptable
from a risk standpoint?
That really has nothing -- very little to
do with the actual frequency of it.
MR. MAYFIELD: But don't changes in the
plant hinge on what size break you go --
MEMBER KRESS: Oh, absolutely. But that
-- what I'm saying is you -- you make your rules say
you will -- your large break LOCA will be a leak size
specification that doesn't allow your plant to exceed
these risk acceptance criteria, something like that
1.174 but not exactly.
And then, they say -- all right. And then
use the leak before break technology to translate that
leak size into a pipe size, but that's -- you know,
that's sort of a detail. And it doesn't have anything
to do with the rule.
MR. MAYFIELD: But it's a sticky detail to
get to.
MEMBER KRESS: Yes, it's not easy. It's
not easy. I didn't mean to say that. But it doesn't
have anything to do with the rule, I don't think.
That's my point. Other than if that helps define the
frequency that goes into a PRA to do your calculations
for risk. It's useful there, too, but, you know,
presumably we've already got that in the PRA to some
extent.
But that's my point. I don't see how leak
before break really plays much of a role here except
in translating your leak size into pipe size.
MR. MAYFIELD: I think that's -- why don't
we go and then -- why don't we go ahead and walk
through this, and then come back to it, Tom, if we
don't address your issue.
We spent some time -- slide 8. We spent
some time thinking about what would go into the
analysis, the probabilistic fracture mechanics
analysis that might support this kind of change,
whether you go at it to determine what would be an
acceptable frequency or the analysis that you'd have
to have to work backwards from at leak size to pipe
sizes.
We spent some time thinking about what
should go into that, what kinds of things would have
to be considered. First of all, we think service
experience is the right place to start. We'll talk a
little bit about that. Some analysis to predict
piping failure frequency -- and, again, it's the same
basic analysis, whichever direction you're going.
And the other one is to look at what kinds
of failures -- other failures where the large break is
currently a surrogate to cover those, and we'll talk
a little bit about that.
We've had some prior analyses that have
addressed these same kinds of subjects; it's
probabilistic fracture mechanics kind of stuff. Those
have been presented to the committee going back to
1985, '86, the changes to general design criterion 4;
more recently, the risk-informed, in-service
inspection programs.
Neither of those approaches were
sufficiently rigorous to support the kind of rule
change we're talking about here. And I'll talk a
little bit more about why we believe that. We think
that the analysis to support this kind of rule change
-- and, again, whether it's going to the frequency or
backing from leak size to pipe size -- that level of
analysis ought to be at least as rigorous as what
we're going through on the PPS rule change.
We think that's beginning to set a
standard for the level of rigor that we think should
be in the analyses to support these fundamental type
of rule changes.
Can we have slide 9?
The prior staff study that, interestingly,
was referenced this morning was NUREG CR-5750,
estimated large break LOCA frequencies. For the PWRs,
that was a five times 10-6 kind of frequency, with a
90 percent confidence interval running from 10-7 to 10-
5 per year.
The recent experience at V.C. Summer,
while it certainly wasn't a large break LOCA, did
contribute to the same kinds of cracking that was
looked at. That would increase the best estimate that
the staff has looked at to something on the order of
seven times 10-6, which is certainly within the band
that came out of the earlier study.
We think that overall these numbers are
conservative, and then it gets to be a challenge of,
well, how conservative are they? And there are a lot
of assumptions. If the committee wants to go there,
we've got the staff here that can talk to them. But
I think the point that I wanted to make is that the
numbers aren't low enough that just out of hand you
say the large break LOCA can be dismissed. So it
requires a little more looking than that.
MEMBER KRESS: Where is that?
MR. MAYFIELD: Sir?
MEMBER KRESS: Where is that level?
MR. MAYFIELD: I'm not going to try and
pick it. But when you're in the mid 10-6s --
MEMBER KRESS: You know you're not there.
MR. MAYFIELD: -- we think we're not
there. It may be one of these things where,
collectively, the Commission would decide, well this
is -- you know, some number is low enough; I don't
think we have that target today.
CHAIRMAN SHACK: It's okay for vessels to
fail at that rate but not pipes?
MR. MAYFIELD: Absolutely.
MR. KING: Remember, the framework
document had laid out initiating events, sort of in a
frequency -- by frequency category. And when it got
down to rare events, then it had, you know, basically
collectively we want rare events not to add up to be
more frequent than 10-6 per year. And any individual
one ought to be substantially less than that.
We sort of proposed a rule of thumb of,
you know, at least a factor of 10 lower than the 10-5,
which would say, really, to exclude something you've
got to be below 10-6 per year frequency. It's sort of
the guideline we've been using in the framework
document.
MR. MAYFIELD: Slide 10, Alan.
We wanted to look at some of the other
changes that we've made, places where we have accepted
what's -- analyses that are put in a general class of
leak before break. These are basically probabilistic
fracture mechanics analyses or determinations that
derive from those kinds of analyses.
I mentioned the general design criterion
4 change from 1986. There we -- the underlying notion
was that pipes will leak before they will break, and
that was accepted for eliminating the dynamic effects
associated with the big pipe fracture.
What that led to was the removal or relief
from having to install pipe whip restraints and jet
impingement barriers. There were a few other spinoffs
to that, but it -- those were the big issues at the
time the rule was promulgated.
Subsequently, we've used these same kinds
of arguments to get some relaxation on break opening
time for the baffle bolt analyses, the risk-informed
changes to the in-service inspection programs, and we
used the same basic approach looking at the resolution
to GSI-190 on environmental effects on fatigue a year
or so ago.
Slide 11. The traits from those prior
applications -- they all seem to have both
probabilistic and deterministic aspects. Defense in
depth, however, hasn't been challenged. We were
always still covered by the large break LOCA, the
50.46 requirement. So we weren't getting to
fundamental changes in the plants. Rather, we were
dealing with some ancillary things.
The general design criterion 4 change,
back when the staff made the presentation to the ACRS,
they described the pipe whip restraints as the evil
pipe whip restraints. The notion was that, by and
large, these devices were being contrary to safety.
They were contributing to containment heat
load, they were impeding access for in-service
inspection, and if they didn't get shimmed up right
you could actually introduce loadings that hadn't been
previously analyzed, if the pipe managed to hang up on
the whip restraint. So there was an underlying notion
that these were good changes.
Similarly, with the risk-informed ISI,
there was the underlying notion that inspection
resources could be better shifted to other piping
systems, and there was also a -- both a quantitative
and qualitative approach, and they gave similar
results. So in each of these cases there was
something else that -- that supplemented the change
rather than just the analysis.
And as I mentioned, we don't think those
prior analyses, the probabilistic fracture, were as
rigorous as what we would apply today in supporting a
rule change, at least a rule change like this.
MEMBER APOSTOLAKIS: But let me, again,
say --
MR. MAYFIELD: Sure.
MEMBER APOSTOLAKIS: -- what would be the
change of the rule under this option? What exactly
would be the change? I mean, you are arguing that
they don't have sufficient analytical basis.
MR. MAYFIELD: But the notion here is that
if you -- if, for example, you wanted to use this type
of analysis to redefine the large break LOCA
frequency, or -- or subsequently use this approach to
determine what is an appropriate break size.
MEMBER APOSTOLAKIS: But large break LOCA
frequency -- do you mean the frequency per year?
MR. MAYFIELD: Yes, sir.
MEMBER APOSTOLAKIS: Wouldn't that be a
fundamental change in the rule there? Because the
rule deals with the conditional stuff. Given that
there is a break, it shows this and this and that.
MR. MAYFIELD: And if you -- yes. Now,
the notion is here -- I guess the notion I've been
working to is that the approach would go along the
lines of showing that the large break -- the frequency
of the large break LOCA is so low that it shouldn't be
the governing consideration. And so now, what is the
break size that is more appropriate to include in the
analysis?
MEMBER APOSTOLAKIS: But does the rule, as
written now, it does not take into account the
frequency of the LOCA at all?
MR. MAYFIELD: Then, it assumes it's one.
MEMBER APOSTOLAKIS: Does it?
MR. MAYFIELD: By --
MEMBER APOSTOLAKIS: It assumes it?
MR. MAYFIELD: It just assumes it.
MEMBER APOSTOLAKIS: But if you go this
way, then you are really changing the argument here in
a fundamental way.
MR. MAYFIELD: Well, either that or you're
saying that the frequency -- well, that's right. This
is --
MEMBER APOSTOLAKIS: But that's what I
don't understand. That's a fundamental change in the
rule. The rest is the mechanics of doing it. Now,
what there is fundamental? I mean, I read the rule.
I'm trying to --
MR. MAYFIELD: Well, the idea is you're
going to redefine the break size, the design break
size, as something other than the largest pipe.
MEMBER APOSTOLAKIS: So then you would go
back to this idea of conditional that --
MR. MAYFIELD: You're going to ultimately
go back to this idea of some conditional break and use
that as the design criterion. And you would use this
type of analysis to support defining that alternate
break.
MEMBER APOSTOLAKIS: Again, you know, it's
not obvious to me why you have to switch back to that
and not deal with the frequencies of the accident
sequences of various LOCAs, and deal with the whole
sequence.
MR. MAYFIELD: Again, I think the same
analysis schemes would come to play.
MEMBER APOSTOLAKIS: That's correct.
MR. MAYFIELD: Whether you did it with --
just as an alternative to the existing conditional
approach or deal with a spectrum. The underlying
analysis would be the same.
MEMBER APOSTOLAKIS: But you would still
deal with the frequency of the initiator that way.
MR. MAYFIELD: That's correct. But
don't --
MEMBER APOSTOLAKIS: If we could get a
listing of these changes, that would be very helpful
to me.
MR. KING: There's two parts you're going
to hear. Mike is talking about changing the
definition of a large break LOCA, which is really what
size pipe you assume. There's another part we've
looked at; it's all the assumptions that are made.
Regardless of what pipe size you assume,
there's a whole bunch of other assumptions that are in
the way we do business today and we're looking at
those as well, because of them don't make sense when
you look at it from a risk perspective. And Alan
Kuritzky is going to talk about those, so there's two
parts to this.
MEMBER BONACA: So what you're talking
about now is really addressing what has been presented
this morning by the industry.
MR. MAYFIELD: In part, yes.
MEMBER APOSTOLAKIS: In part.
MEMBER BONACA: And then, one thing that
would be interesting, of course, in that respect would
be if you define large break LOCA as a smaller break
than the current one. What kind of criteria, I mean,
would -- is there a possibility of accepting some
higher consequences for a large break LOCA that we
have today? Because this was to get done in the FSAR.
It would probably lead to more than 2,200
degrees Fahrenheit temperature. It would lead to some
more percent oxidation.
All I'd like to say is that it seems to me
that the only conflicts from the presentations we had
this morning and now is abandoning large break LOCA,
and that becomes almost like a severe accident
consideration where you can have, you know, core
damage to whatever extent.
Isn't there some possibility right now you
are setting your target to a smaller break size, but
you can also look at large break LOCA and see what --
given the -- all the changes that a plant would make,
you have some results that you might find as
acceptable now. See, that would allow you to deal
with the fact that you're not abandoning completely
some expectation for large break LOCA.
MR. KING: Well, I think what was said
this morning was the risk assessments will still have
the large break LOCA in them.
MEMBER BONACA: Yes.
MR. KING: And the risk assessments are
still going to need to show that the large break LOCA
is a very small contributor to risk for this whole
scheme to work. I mean, if we make changes and now
all of a sudden the large break LOCA is a high
contributor to risk, then we haven't done our job.
MEMBER KRESS: I think that's true, that
if it came out to be a high contributor to risk for
specific plants, then you need to consider it in your
design basis. But I don't think the inverse is true,
that if it comes out to be a low contributor to risk,
that doesn't necessarily mean that makes it a bad
design basis.
And the reason for that is when you invoke
and impose a design basis, you end up with certain
design features and certain procedures and certain
things that help control the total risk, not just for
that set of sequences but for all of the other
sequences involving LOCAs, and so forth. And you want
that total risk to be within acceptable limits.
So, you know, my perspective is your
Option 3, if you look at -- if I go from -- to a
specific leak rate or a specific size different than
this, what changes will the plants make? And are
those changes acceptable from a risk standpoint? And
that, to me, seems to be the only rational way to view
this.
MR. KING: I agree.
MEMBER KRESS: Not deal with, is the large
break LOCA a risk contributor? Maybe it is, maybe it
isn't. You know, if it is, you have to deal with.
But if it's not, I still don't think it's the right
perspective.
MR. KING: No, I agree with everything you
said. The large break LOCA probably is a surrogate
for some other things that you don't have to worry
about because you take care of it through dealing with
the large break LOCA, although that has to be
considered.
MEMBER KRESS: Okay.
MR. MAYFIELD: Which I think is at least,
in part, the message here is that this is perhaps a
tougher analysis than has been thought about.
And that -- on Slide 12, the notion here
is, again, we think that we've started setting some
guidelines on what is an appropriate level of rigor in
these analyses, based on what we're doing for the PTS
rule. And the presumption here is that we're going to
actually be successful with the PTS rule.
You can look, of course, at transients, at
thermal-hydraulic response, and so on. There are a
couple of these that get to be more interesting if you
were actually going to try and do this for piping.
The flaw distribution is one that gets significantly
more challenging.
We've actually done more work on flaw
distributions for vessels than we have for piping.
The other one that we haven't gotten into for these
kinds of analyses are the uncertainty analyses, and to
do that rigorously dealing with both epistemic and
aleatory uncertainties.
We think that would become a significantly
more challenging approach for this piping analysis
than it is for the vessel. There are more degradation
mechanisms, there are more unknowns, there are more
model uncertainties, and on and on. It gets to be a
much tougher challenge than it is for the vessel.
If we can have Slide 13, Alan.
The next several slides I don't propose to
go through in any detail. They are more for your
information. They're the kinds of things that we
think would have to be treated, some examples of
issues that we think haven't been dealt with very
well. The analyses themselves would have to deal with
both the initiation of subcritical cracking, the
subcritical crack growth, leak rate detection under
normal operating loads, and the fracture during upset
loads. Those are just the kinds of analyses that have
to be dealt with.
The main inputs to the probabilistic
fracture, the material property input, the crack
morphology -- what the crack surfaces look like -- the
loads under normal and transient conditions, the flaw
distributions.
On Slide 15, we need to deal with the
current and as-yet unknown mechanisms. This is
something, how do you know what hasn't happened yet,
and that's part of the speculation that gets to be
challenging in doing this.
There are some kind of notable examples of
things that we didn't think were likely.
MEMBER WALLIS: It's not just what hasn't
happened. It's what you --
MR. MAYFIELD: It's what we don't know.
MEMBER WALLIS: -- are not yet aware of.
MR. MAYFIELD: Yes. Exactly. So there
are some examples here of the kinds of things that
have caught us a bit by surprise in the past, the most
recent one perhaps being the primary water stress
corrosion cracking in the Inconel 182 welds. It's not
that we didn't know that material would crack. We
just didn't think that the loading and conditions were
right for it to crack in that particular application.
There are a host of things on Slide 16
that go into the leak rate analyses. So if we really
wanted to follow Dr. Kress' suggestion and take a leak
size and work backwards to a pipe size, that gets to
be a -- first of all, you can get a range of pipe
sizes depending on the crack size and loading. So it
-- there would be a fair bit of uncertainty to deal
with in that analysis.
But things like residual stresses come
into it, the pipe to system boundary conditions, are
you getting some restraint of free thermal expansion,
some restraint of bending in the pipe, crack face
pressures, some things that haven't been included.
And it -- as you work backwards from a leak size to a
pipe size, those things get to be important and you
change from a conservative approach to a non-
conservative approach, just depending on the direction
you're going.
MEMBER KRESS: Would this analysis also be
time-dependent, then?
MR. MAYFIELD: Absolutely.
MEMBER KRESS: So that's a factor we never
really --
MR. MAYFIELD: Yes.
MEMBER KRESS: -- had in this design basis
before.
MR. MAYFIELD: That's correct. That's
correct.
On Slide 17, the fracture mechanics
analyses, this would go to an elastic plastic fracture
rather than the net section collapse or limit load
kinds of approaches. We have historically assumed
some idealized through wall circumferential cracks,
nice regularly-shaped things. But if the subcritical
cracking caused large surface cracks, long on the
inside surface and relatively short on the outside
surface, it introduces another complication in either
direction in the analysis.
You'd need to look at things like the
earthquake rate -- loading rate effects on material
properties. These are not dynamic loading rates in
the classic sense, but there is a phenomena known as
dynamic strain aging that at these loading rates does
get to be an issue.
One of the interesting things that we've
discovered from some of our experimental work at
Battelle is that if the bending plane happens to be
different than the normal -- the primary plane for the
crack, you can get some significantly different
fracture behavior, and that is physically possible.
We've seen it. So that gets to be an interesting
twist to the analyses.
On Slide 18, the material property input,
these are normal kinds of things that you would hear
about, and I think you've heard about them before.
The difficulty now comes in quantifying them and
quantifying the distributions on them into the
analysis and to feed into the uncertainty analysis.
This is something that we met with the
committee and talked about just on the fracture
toughness for the vessel steels. And that, in and of
itself, is fairly complicated. This expands
significantly.
On Slide 19, the loads, looking at normal
operating versus upset loadings, and looking at the
actual stress levels versus the design basis levels --
and that got to be an important consideration when we
were looking at the environmental effects on fatigue,
to not just roll in the design basis stresses but to
actually come up -- try to come up with the true
operating stresses. And in some cases they were
substantially lower.
There are some thermal gradient effects
that I guess I hadn't been aware of until a
conversation that was being held on the steam
generator tube ruptures, some helical thermal gradient
in the hot leg. How much that changes with time, is
it a consideration, it's not something we have rolled
into the previous analysis; it's something that ought
to at least be considered here. Whether it has to be
incorporated rigorously is something you'd have to
decide as you went along.
The initial flaw distributions -- again,
this is something that will be even more complicated
for the piping evaluation than it was for the vessel
because of the nature of the loadings, the potential
for the environment to expand the number of flaws, the
way you would have to treat the potential for pre-
service or fabrication-induced flaws, to pick up and
grow during operation, which is not something that we
have seen or are able to predict in the vessel
analysis.
So, again, the piping is a more difficult
analysis to perform.
We've talked about other sources of large
breaks where the large break LOCA or the -- at least
the failure of the big pipe is a surrogate for these
things. Losing the steam generator manways, the
potential for those few plants that have loop stop
valves, the potential for losing the bolts that hold
the bonnet on. We've seen in some smaller valves some
valve body cracking; would that be an issue here?
Then, there is the subject of indirect
sources of pipe failure, and the notion here is moving
heavy loads inside the containment during power
operations.
In the work that was done by -- at
Lawrence Livermore in support of the GDC4 changes,
these indirect sources of pipe failure -- actually,
you got up to where that was the dominating
contributor to pipe failure frequency. So it's
something that needs to at least be considered here.
MEMBER BONACA: Moving heavy loads, the
power operations, is that containment?
MR. MAYFIELD: Well, at any rate, in
summary on Slide 22, the service data alone don't
appear to us to support eliminating a large break LOCA
without some further evaluation. The data, such as
they are, with some -- potentially with some limited
additional analysis would appear to be sufficient to
support some other types of changes.
The probabilistic analyses to support
eliminating the large break LOCA we believe would have
to be significantly more rigorous than what we've done
in the past, both from the staff as well as the
industry. And we think that active degradation of the
piping is not the only consideration.
The analyses would have to consider other
breaks, and they'd have to consider things like weld
repair history to make sure we capture potential for
other degradation mechanisms.
MR. KING: Okay. Now Alan Kuritzky is
going to talk about putting the redefinition of the
break size aside, what other things that we looked at
in terms of potential changes in risk-informing 50.46.
MR. KURITZKY: Yes. As I mentioned
earlier right before Mike began speaking, there was
about 10 different options we looked at during this
program. The first three all involved various -- you
know, different variations of redefining the large
break LOCA, and now here we have seven more options.
And these look more specifically --
MEMBER APOSTOLAKIS: Again -- I'm sorry,
maybe it's not a good day for me. Why are they risk
informed?
MR. KURITZKY: Why are they risk informed?
Well, because as I go through these items that are on
this list I'll explain the risk implications and what
risk aspects led us to come up with these as options.
MEMBER APOSTOLAKIS: And you have criteria
for deciding that these aspects are good or bad?
MR. KURITZKY: Well, what we're using
right now, as Tom mentioned before, is based on our
framework document. We have about a 10-6 roughly for
something that may be low enough that it's, you know,
considered to be a rare event and may not need to be
considered. And so that's kind of like a threshold.
MS. DROUIN: That's only one threshold.
Also remember that one of our ground rules under
Option 3 is that when you look at the analyses, if the
analyses are conservative, we see that in a risk arena
you want to be realistic. So some of the options
might -- it may have no relationship to the
quantitative guideline. It's because there is excess
conservatism, and our ground rules -- one of the
ground rules in the framework was to be realistic.
MEMBER APOSTOLAKIS: But, again, we are
talking about a fleet of 103 units. So, I mean if I
want to delete loss of offsite power consideration,
that may have different impact on CDF on different
units. On what basis do I decide to delete? By the
largest impact or when we --
MEMBER KRESS: The average or --
MEMBER APOSTOLAKIS: Yes.
MEMBER KRESS: On a specific basis?
MEMBER APOSTOLAKIS: Yes, I don't know.
You see, that's where I get confused.
MEMBER KRESS: Yes, I had that same
problem.
MR. KING: Or you set some criteria.
MEMBER APOSTOLAKIS: So we don't have them
yet.
MR. KING: We have not reached the point
where we've nailed down every detail of what we're
going to propose. What we're talking about are
certain areas that we think make sense for -- for
looking at it a little further and possibly going
forward with recommendations to the Commission.
Just take the first one -- simultaneous
loss of offsite power at the same time the large break
LOCA occurs. You know, you can use frequency
information and risk information to say that that
assumption is pretty -- pretty low in terms of
frequency and pretty low in terms of risk --
MEMBER KRESS: Well, let me --
MR. KING: -- for the large break LOCA.
MEMBER KRESS: Yes, let me tell you what
bothers me about that. Let's go back to the framework
document where you have these categories and ranges of
frequencies for basically specific sequence sets.
Now, if I look at, say, a large break LOCA
of a given size, and it fits into one of those
categories, that's the wrong way to use that framework
document. That's not what it was intended for. What
you need to say is, "We've defined this as my large
break LOCA." What does it do to the frequencies of
all the sequences that go into the PRA, and how does
this shift them all with respect to that framework
document? And is that total shift --
MEMBER APOSTOLAKIS: Which means look at
some global method.
MEMBER KRESS: Some global method, yes.
MEMBER APOSTOLAKIS: Like the CDF.
MEMBER KRESS: Well, like the CDF. Well,
not exactly global because I think defense in depth
gets involved in terms of this allocation that you
have. I think that's a defense in depth concept. So
that's one way you bring that in.
But I think just to say large break LOCA,
if it's this size, that particular sequence provides
this contribution that we have in this box, is the
wrong way to use that framework.
MR. KING: I think you have to do both.
I think you have to start somewhere, and you have to
say if large break LOCAs are very low frequency, then
maybe we ought to consider not calling them design
basis accidents anymore. Then you have to go and
look, well, what does that do to my risk for the
plant? And if it turns out that that drives other
things up, then maybe you don't want to do that.
If it turns out it doesn't drive other
things up, that it really is a very low risk type of
item --
MEMBER KRESS: I think that's the key
element right there.
MR. KING: Yes, I agree with you.
MEMBER KRESS: I think that part of it --
MR. KING: But you've got to start
somewhere.
MEMBER KRESS: Yes. Yes, because you
can't really do that other part yet, the second part,
the important part, the risk -- and it's awfully hard
to do. You probably don't have the capabilities of
doing that just yet.
MR. KING: Well, I think you have the
capabilities. I think, you know, you have to take
some time and some, you know, effort to do that.
MEMBER KRESS: Each plant-specific PRA
would have to be -- would have to decide what changes
in a plant are made and see how that impacts the PRA,
and do it on a plant-specific basis, and average it
out or add it up or something. It's not easy.
MR. KING: Everything we're talking about
today, not just the break size, there is the question
of how much can you do generically, and how much do
you have to do plant-specific?
MEMBER KRESS: And that's --
MR. KING: And the question applies to
everything we're talking about. And we haven't nailed
down where you draw that line yet, but we recognize
that that line has to be drawn somewhere.
MEMBER APOSTOLAKIS: Now, what is the
Commission's attitude towards Option 3? Have they
approved it or what? I don't remember.
MR. KING: Option 3 was approved, yes, a
couple of years ago. Remember, 50.44 was the first
one that came out of that, and this is the second one.
MS. DROUIN: It's in the SRM in response
to SECY98-300.
MR. KING: Yes.
MEMBER KRESS: Yes, 50.44 was a breeze.
That's one of the ones that --
MEMBER APOSTOLAKIS: Mary, you live with
those numbers. We come here once a month. You know,
SRM does this and that. It's not as alive to me as it
is to you.
(Laughter.)
MR. KING: Why don't you let Alan talk
about these, and then we can come back and maybe talk
about some of these generic questions.
MEMBER APOSTOLAKIS: Okay.
MR. KURITZKY: Okay. One of the first
things that we just talked about -- just momentarily
about the simultaneous loss of offsite power
assumption. One of the ways that we can envision
something like that playing out is -- is if the NRC
were to come up with some acceptable frequency
distribution for large break LOCAs, Tom already
mentioned that in NUREG 5750 we have -- or I guess
Mike may have mentioned that we had a range from 1E-7
to 1E-5 for PWRs for large break LOCA.
If 1E-5 is your upper bound, you could say
that's the upper bound for large break LOCA and you
can show that your conditional loss of offsite power
-- you know, loss of offsite power frequency is on the
order of 10-2, 10-3, and that could put you below some
threshold. And you may decide that that doesn't need
to be one of your design basis considerations, you
know, assuming loss of offsite power with the large
break LOCA.
Numbers like that have already been -- as
part of the resolution I think of generic issue 171,
there was some look into the conditional loss of
offsite power probabilities and --
MEMBER KRESS: As a function of time after
the LOCA?
MR. KURITZKY: No, it was just a -- it was
just given that you had --
MEMBER KRESS: You may have one eventually
within --
MR. KURITZKY: Right. In fact, what was
driving the initial failure probability was the
starting of the large ECCS pumps.
MEMBER KRESS: Okay.
MR. KURITZKY: And so that was in the
range I think for PWRs about 1.1-something, 10-2, and
for BWRs I think it was 6-2. So we can see that the
numbers -- given the large break LOCA frequencies, and
those numbers put you somewhere in that -- in that
threshold range. It was something that may be
feasible, but --
MEMBER KRESS: Once again, I fail to see
the relevance of the, say, 10-7, because what I was
saying is if I want to drop the loop, simultaneous
loop LOCA rule, why does that change? Why do people
want to change? And does that change the risk status?
You know, that's irrespective of the frequency.
MR. KURITZKY: And I would have to -- and
if someone were to come through and supply the
justification why they feel their initial loss of
power is -- why they would make a change, then
whatever plant change they would then want to --
MEMBER KRESS: You know --
MR. KURITZKY: -- make because of it they
would have to justify on a risk basis, and it could
impact many things besides just the large break LOCA.
They have to show that the delta risk is acceptable,
you know, maybe per Reg. Guide 1.174 or --
MEMBER KRESS: But, see, my point is I
don't see that you can do ahead of time a look at the
conditional loop, for example, and say you come up
with a frequency of 10-8 or 10-7 or 10-9 even. I don't
see that it tells you anything that's useful in
saying, "Okay. We'll do away with that part of it."
I don't see that that's helpful to you.
MR. KING: I don't see why you don't see
it's helpful. I mean, if you're making assumptions
that are rather remote in likelihood, does it really
make sense? And particularly if it's causing the
plant --
MEMBER KRESS: There may be some very
specific things just related to that. I don't want to
be, you know, completely on one side of that. There
may be some very specific things relating to that that
has no impact on anything else, and, in fact, may have
a negative impact that you could obviously change.
But I think just to use it as an ultimate
reason is --
MR. KING: But if it's causing the diesel
generators to have to start up very quickly --
MEMBER KRESS: Well, there may be other
reasons that you want the diesels to start up quickly.
MR. KING: There may be. There may be
other things that would catch -- I agree with that.
MEMBER KRESS: You can't just throw it
out.
MS. DROUIN: No. And if you go back to
the framework, nothing is ever thrown away or added in
just because of a number. And that's where we bring
in this -- we bring in the defense in depth and we
have six things that we had identified there.
Also, another part of the framework is
that before -- one of the other ground rules before we
delete something, we have to go in and go back and use
50.44 as an example, look at all of the tentacles and
where it would have impact. And that has to be
brought in, and that's all part of the ground rules
under which we make the decisions.
So it's not just, oh, well, there's a
number and it's below it, so we can throw it away.
It's not -- that's not what we do.
MR. KURITZKY: And just to follow -- just
in direct response to what you said, Dr. Kress, in the
case of if we were going to relax -- you know, the
start time had to be relaxed, then we would let -- one
of the things we would need to consider is, do we need
some other type of design-based accident that -- that
governs diesel start time? Because there may be some
other parameters that need to be looked at. So,
you're right. That would have to be considered.
Okay. The second one on there is the
excluding highly unlikely combinations of large break
LOCA initiators and single -- it's going to fall along
the same lines of what we just talked about -- loss of
-- initial loss of offsite power. It would run in the
same -- have that same --
MEMBER KRESS: Yes, it's the same issue.
MR. KURITZKY: So whenever fair, the same
thing holds.
Again, so that is something we would
consider. There would be a threshold, and we'd have
to consider other aspects, too, what would be the
overall risk impact of changes associated with that.
The next bullet is to look at the
conservatisms in Appendix K and decide whether or not
there are certain conservative models in Appendix K
that we feel can be replaced with more realistic
models or assumptions.
MEMBER KRESS: I see that as sort of a
separate issue. I don't know -- I don't see that's
related to this particular issue.
MR. KURITZKY: Only in the sense that
we're looking to make things more realistic.
MR. KING: Yes. Isn't risk-informing also
being more realistic? I mean, the risk analysis is
supposed to be a realistic analysis.
MEMBER KRESS: Well, the risk analysis is
supposed to be realistic. But Appendix K doesn't
necessarily have to be realistic. It can be risk-
informed and be very --
MR. KING: If they're going to be risk-
informed, that to me also implies we want to be
realistic.
MEMBER WALLIS: Is the idea that something
like the peak clad temperature might be a function of
the risk? That you might allow a 2300 if it's less
likely or --
MR. KING: Well, it depends. Is the peak
clad temperature 2200? Is that a very conservative
number? Is that a --
MEMBER WALLIS: Even if it's conservative.
Does it -- if it's a very likely event, you might want
some more conservatism.
MR. KING: Yes, I agree with that. That's
the margin issue.
MEMBER WALLIS: So you'd reexamine on the
basis of risk.
MR. KING: Yes.
MR. KURITZKY: In fact, the last bullet up
there is -- that's pretty low on the screen for you
guys, but it's modify the ECCS acceptance criteria.
It's another one that --
MEMBER WALLIS: We got that.
MR. KURITZKY: Okay. And then, actually,
the next three bullets all are things that are
associated with the best estimate calculations. And
just various ways of -- well, in the first case, they
would be using -- we would be considering, you know,
would it be possible to use a distribution of break
size allocation as input to the best estimate file and
propagate it just like we do with other parameters.
Therefore --
MEMBER WALLIS: It's just one of the other
uncertainties.
MR. KURITZKY: Right. And, in fact, you
have the very tail end of the LOCA, the very large
break -- you know, it's driving what you're getting
out because it's --
MEMBER WALLIS: It's not --
MEMBER KRESS: Sorry. Once again, I'm a
little bothered by that one because to me the break
size or, alternatively, the leak rate was the
independent variable. And you don't normally attach
uncertainty distributions to your independent
variable. You're looking at something that results
from that independent variable, which is a specific
number.
And you put the uncertainties on the
things that give you the result, and that one -- I
just don't quite understand what that one is saying to
me.
MEMBER WALLIS: I don't have any problem
at all admitting I'm uncertain about the break size
and trying to put it into the analysis.
MEMBER KRESS: Well, sure you're uncertain
about the break size if you're going to go in and
decide on a frequency of a given break size to feed
into a PRA, as part of PRA. But --
MEMBER WALLIS: You don't think it's part
of the thermal-hydraulics?
MEMBER KRESS: Oh, I'm sorry. I think
it's part of this element of fracture mechanics that
Mike said when you go from a given leak rate to decide
what kind of a break size that relates to, it's
certainly a part of that. I didn't mean that. But to
me it's not part of risk-informing the 10 CFR 50.46.
It's a -- it's how you implement that risk-informed
version. Once you go back to -- if you're going to go
so -- say, now what break size can I live with?
MEMBER WALLIS: I'm just sitting here
looking at all of these things and thinking of how
much work it would take to do them.
MEMBER KRESS: Oh, this is -- this looks
like a lot of work.
MR. KING: This is a list of things we've
considered. What you're going to hear is how we split
those up as to what we think is reasonable to go
forward with in the short term and which ones need
more work. This is one of the ones that falls on the
"more work" list. It's not anything we're to the
point we feel comfortable to go forward with now.
MR. KURITZKY: These are all the things we
-- actually, we've had these on the table for probably
close to a year now, and we've presented them at
various public meetings, so -- and in the last --
well, the next two there as far as the best estimate
calculation, the fifth bullet, meet improved
efficiency of the calculations using improved
statistical sampling methods like Latin Hypercube
sampling to try and make it a little more efficient.
And the sixth one was one we had
considered the possibility of having the NRC approve
certain uncertainty increments that would then be
input, and licensees could then just do their best
estimate calculation without having to necessarily do
all of the costs of the uncertainty analysis. There
would be some fixed offer or safety margin or
something that would be associated with a
predetermined uncertainty increment.
Again, these are just things that we had
put out on the table for possible consideration.
And the last one was modifying the ECCS
acceptance criteria for looking at peak cladding
temperature, the oxidation, and determining whether or
not there is any better way, you know, to -- in other
words, of those acceptance criteria, is there a better
way to handle it?
MEMBER WALLIS: About like rewriting the
works of Shakespeare.
MR. KURITZKY: In our spare time.
MEMBER WALLIS: Spare time.
(Laughter.)
MR. KURITZKY: Okay. Based on the --
those are some of the options we had thought about.
Of those options, there was a few that we felt in the
-- in the short term we felt we could establish the
feasibility of. And when we say "short term" we mean
in order to get something in the Commission paper by
June.
And the ones that we thought we had a fair
shot at establishing the feasibility or determining
the feasibility of are the large break LOCA, the
simultaneous loss of offsite power, large break LOCA
assumption, and that effects -- these, in fact, don't
affect -- I think none of these actually affect 50.46
-- the actual 50.46 itself, but rather Appendix K or
some of the GDCs.
Also, we have the single failure
assumptions just like we talked about in the previous
slide.
MEMBER WALLIS: Associated with the large
break LOCA.
MR. KURITZKY: Like we have unlikely
combinations. And an additional thing with the single
failure assumption is -- is to pursue maybe the use of
some type of risk-informed approach consistent with
the framework that would us to determine how and where
we might want to change the single failure criteria.
But that's a little broader than just looking at the
single failure associated with that large break LOCA.
Reducing decay heat conservatism --
Appendix K is also one we feel we might --
MEMBER WALLIS: That might be one of the
easiest ones.
MR. KURITZKY: That's what we like to
hear.
MEMBER WALLIS: Well, is that your
statement? Isn't this one of the easier ones?
MR. KURITZKY: That depends who you ask.
MEMBER WALLIS: Oh, okay.
MR. KING: We can have Norm Laubin explain
that to you, but we've done some work on it, and it's
not as easy as you may think.
MR. KURITZKY: A lot of these are -- some
of them appear easier, and then if you ask certain
people there are reasons why they don't think they're
that easy, so it --
MS. DROUIN: A lot of it -- why it didn't
get so easy is that, as you see, just the effects of
it and how far you have to dig down and the things
that it affects is sometimes not as straightforward.
And as you start uncovering all these layers it just
becomes a little bit more complicated than you thought
at the onset.
MR. KURITZKY: And another one, actually,
is not on this slide, but we are also kind of tossing
about is the possibility of some relaxation in the
break opening time. Right now, I think it's specified
in the standard review plan, Section 3.6.2, as a one
millisecond break opening time. And we're considering
--
MEMBER WALLIS: If it goes to two
milliseconds, it won't make much difference.
MR. KURITZKY: Or one and a half.
(Laughter.)
Or maybe something else along the lines of
20 milliseconds or something, and one of the areas
that they may have significant benefit is for the
barrel -- the baffle form of bolt.
MEMBER WALLIS: On the loads, is it when
you get sort of --
MR. KURITZKY: Yes.
MEMBER WALLIS: -- propagations and things
that --
MR. KURITZKY: Yes. Lastly, we have up
there consider enhancements based upon risk insights,
and then we want to look through the risk insights and
see if there -- if it indicates that there may be any,
you know, related safety concerns.
MEMBER WALLIS: What do you mean by
"enhancement" here?
MR. KURITZKY: Enhancement would be if
there is a safety concern that is indicated by the
risk insights, is there something we feel needs to be
beefed up?
MEMBER WALLIS: Oh, beefed up. Okay. It
means toughening up the regulation.
MR. KING: An example is maybe automatic
switchover to ECCS recirculation for those plants that
don't have it. Is that something the risk insights
are telling us ought to be in place? You know, that's
an example of what we're looking at.
MEMBER WALLIS: It makes sense that you
should look at both directions, if you can improve
safety based on risk insights.
MR. KURITZKY: Right. And then our
Option 3 framework. That's part of the Option 3
program.
Okay. Some of the benefits that we've
identified with these near term -- these possible
near-term changes, and as they relate to the agency's
performance goals -- to maintain safety, we feel that
these changes would help to maintain safety. In fact,
there may even be some improvement in safety.
Diesel generator liability, as was
discussed earlier today, is one area where there may
be some improvement in safety. We would maintain the
elements of defense in depth so there would be no
degradation there.
And, again, safety system reliability
would either -- we would -- there would be no
significant decrease based on these changes, and, in
fact, in some cases there may be an increase,
depending on how that -- how we would address the
single failure criteria. There's actually some places
where there may be enhanced system reliability.
Also, these changes -- these near-term
changes would make the PRSA activities more effective,
efficient, and realistic. Particularly in the
realistic arena we would have more realistic
assumptions in the DBAs. We'd have, you know, maybe
more realistic assumptions for Appendix K, and so that
would -- in keeping with that second goal.
MEMBER APOSTOLAKIS: How would you decide
whether key elements of defense in depth are
maintained? Is that a judgment call, really?
MR. KURITZKY: Well, I guess the actual
elements themselves we're going to take from doing the
framework document, which is fairly similar to the one
that is in Reg. Guide 1.174. And you would -- I guess
it is -- I mean, it's necessarily a qualitative
analysis. I mean, there may be some qualitative
pieces to it, but it's going to be some kind of
judgment qualitative decision that you're not
degrading anything.
MR. KING: Do you still have prevention
and mitigation? Do you still have sufficient, you
know, redundancy based upon the reliability of the
system?
MEMBER APOSTOLAKIS: You can look at the
single failure criterion and you'll decide to abolish
it. You will rely on some quantitative analysis to
show that you have not really degraded the --
MR. KING: Some, for example,
probabilistic definition of the failure criteria.
Maybe it results in not having to assume a single
failure. Maybe it results in having to assume a
multiple failure, depending upon the system
reliability and the break you're looking at. So it
can work both ways again.
MEMBER WALLIS: Some people might argue
that certainly large break LOCA is the biggest thing
that could ever happen. Is that kind of element -- it
has an element of defense in depth to it. I mean,
that you -- because you're uncertain, you look at the
extreme case and defend against that?
MR. KING: Sounds like pre-TMI
discussions.
MEMBER WALLIS: Isn't that the sort of
argument that was originally used in support of the
regulation?
MR. KING: Yes, I think it probably was.
MEMBER WALLIS: Was that defense in depth
or is that something else?
MR. KING: I'm not sure I'd call it
defense in depth. You may argue, well, that gives you
a margin because you know you can handle this big
break. Therefore, you can handle anything smaller.
MEMBER WALLIS: Defense in depth in depth
is what you went to when you were uncertain, because
you sort of have a nagging feeling that if something
were to happen you're not quite certain about, then
you --
MR. KING: Defense in depth to me is
different ways to accomplish the same function.
Again, you know, just assuming the large break covers
everything that --
MEMBER WALLIS: The containment is there
for the same sort of reason. But there are different
ways to perform the function that containment
performs. But because you're uncertain you put it
there anyway.
MR. KING: Okay.
MR. KURITZKY: And the last thing here we
have the performance goal reduce the unnecessary
regulatory burden. And the near-term changes that we
listed previously should help in a number of these
areas that are listed up there. Extension of the
diesel generator start and loading time, which was
discussed already. That's one of the big issues.
In fact, we have some cost information
that was provided to us from industry which shows that
extending the diesel generator start and loading time
could, in fact, save upwards of $400,000 to as much as
$1.2 million per plant per year. The big swing there
is whether or not it's on the critical path outage.
Also, relaxation of some AOTs or, in fact,
even removing equipment from the tech specs, like the
accumulator, which, again, I think we had some data
from industry that shows that could save upwards of
around $17,000 per plant per year. And additional
analytic margin for plants that are limited by
Appendix K right now.
It's only going to be some plants, and
exactly how they're limited and what -- unnecessary
regulatory burden, we feel, again, is going to be very
plant-specific. But it's certainly one area where
there's some potential.
Okay. We recognize that, obviously, as we
move forward with these short-term potential changes
that there's a number of implementation issues that
will have to be addressed. And we have them listed I
think on the next two slides. We don't need to go
into a lot of detail on them because there is stuff
that we can still -- the staff needs to still think
about the nuances of them, but basically they're many
of the similar things that we've discussed already.
If we're going to use something like a
combined reliability threshold or a frequency
threshold for the San Mateas LOOP assumption or the
single failure criterion, then we're going to need to
have some kind of frequency of the large break LOCA
distribution. And since we can't make a convenient
cut necessarily at the six inches or wherever a PRA
says a large break LOCA is, we would actually need to
have that distribution across all pipe sizes or all
affected break sizes. That's one thing we'd have to
-- we have to consider.
When going to the San Mateas LOOP
assumption there are a number of concerns, at least
from doing something generically because there are a
lot of plant-specific aspects. Plants have different
types of procedures and designs to handle loss of
offsite power, particularly delayed loss of offsite
power concerns, and so there may be, you know, some
significant work that may have to be done to make sure
that that's feasible.
Again, as we mentioned before for the
single failure criterion, we may consider whether
there's some kind of risk-informed replacement for
what will be combined initiator frequency and failure
probabilities.
One of the issues we want to keep in mind
as we move forward is that we want to try and utilize
a performance-based approach wherever possible.
MEMBER WALLIS: How do you apply that to
LOCA?
MR. KURITZKY: Well, I think that
performance-based to my mind -- maybe in the example
we've talked before about the conditional loss of
offsite power where if you have some kind of curve for
the -- you know, for the frequency of the large break
or breaks, then the utility may come in and try and
demonstrate that there conditional loss of offsite
power probability is of a certain value.
MEMBER WALLIS: It also would involve
perhaps utilities showing that their ultrasound method
for detecting cracks really could detect cracks. Is
that performance-based, too?
MR. KURITZKY: I leave that one to Mike.
MEMBER WALLIS: They would just go through
the ritual, but it actually works.
MR. MAYFIELD: There is clearly a
performance-based element in that, yes.
MR. KURITZKY: Delayed LOOP we just talked
about. Just a couple of the issues there -- the
double sequencing and the degraded voltage issues are
the ones that we have to kind of do some more thinking
on.
And also, the impact other modes of
operation and potential accidents. Low power
shutdown, for example -- whatever we do for --
anything for 50.46 or large break LOCA, we have to
make sure that while we're thinking primarily of this
operation, we've got to make sure that for low power
shutdown modes we're not going to do anything dumb,
not giving anything away, we need to be conscious of
it.
MEMBER APOSTOLAKIS: Are you going to be
able to do that without very good risk assessments for
shutdown?
MR. KURITZKY: There are limitations that
we have to deal with. I mean, that's obvious. And
even for operation for --
MEMBER APOSTOLAKIS: What does it mean
"limitations"? So how would you do it? Would you be
more concerned with it?
MR. KURITZKY: That's one possibility.
You have to be -- I mean, where there's uncertainty,
you go to conservatism if you can't address certain
things.
Okay. Some things aren't --
MS. DROUIN: We don't have an answer to
that, George, at this point of how we're going to deal
with it.
MEMBER APOSTOLAKIS: All right.
MR. KURITZKY: Okay. As Tom mentioned
before, there is also some things that we're thinking
about more for the longer term, other changes to
10 CFR 50.46 or associated GDCs.
One is to redefine large break LOCA. As
was discussed previously, that goes back to the first
three options that we talked about right before Mike
spoke and also what industry spoke of this morning.
We would want to continue working with industry on the
scope and depth of what work would be entailed.
MEMBER WALLIS: I've been waiting to ask
you the question, and it seems to me your presentation
and the industry presentation don't have too much
overlap.
MR. KURITZKY: Where the overlap would
have been is those first three options or parts of
this first step which we've kind of -- we skipped over
pretty quickly.
MEMBER WALLIS: Maybe we need to discover
what this extent -- the extent of this work with
industry is. I mean, they have their point of view,
and you have your point of view. They seem to be
rather different. Are you working together or --
MR. KURITZKY: We are now having exchanges
of information, but our first three options that we
discussed in the beginning where we -- particularly
Option 1 is very closely related to what industry is
doing.
MR. KING: We've had a number of meetings
with the industry as was mentioned, and I think what
you heard Mike say was we're not closing the door on
the path that they're pursuing. What Mike's
presentation tried to do was lay out the issues that
we felt needed to be addressed if we're going to go
down that path.
It doesn't mean ultimately we won't go
down that path, but it means between now and June not
enough work has been done for us to go to the
Commission and say, "Let's proceed down that path."
MEMBER WALLIS: What they seem to be
looking for, though, was for you to open the door and
say, "We're going to encourage you to do more work
because we think it's fruitful," rather than to hang
a plaque on the door which discourages them from doing
a lot more work.
MR. KING: I think one of the things we
have to talk about is, how wide is the door? You
know, how wide do they see it, and how wide do we see
it?
MEMBER WALLIS: We don't seem to agree yet
on that.
MR. KING: We probably don't agree on
that, but I think from our perspective the door is
open, and we've got to talk about is how wide is it
open.
MEMBER WALLIS: I'm not sure that at this
stage the ACRS has enough evidence from both sides to
give much advice about how open the door is. If
that's what we're being asked to do.
MEMBER BONACA: One question I have is,
have you given second thoughts about the generic
implications of this step? Really, this is the LOCA.
It was presented as the first DBA that can be changed
where there is implications to load rejection.
It's equally unlikely, I think, as an
event that you will have a sheer load rejection the
way that you have in the FSAR. It's true of the steam
line break.
So if you make a change to the LOCA and
then you go back into the risk evaluation and look at
the comparisons between, you know, LOCA, you are
reducing now the expectation resulting from the LOCA,
but you are looking at the one imposed by the steam
line break. Well, tomorrow you will be changing the
steam line break. It will change further.
I think you have to look globally at here
you are really fundamentally changing the way you're
looking at your design of the plant, I mean, and
you'll do it for a LOCA. That's a point that also Dr.
Kress has made, and, you know, without the
implications for the other accidents. I mean, it's --
MR. KURITZKY: I think your point is very
good. Even if --
MEMBER BONACA: I don't think you can wait
until you have gone through this gate, and then decide
how you're going to treat the other accident. I think
you have to think about and have some position on
that, because it will have implications about, you
know, anything that -- I mean, the example I made of
doing your risk assessment based on the consequences
of the LOCA and comparing it to other restrictions
being imposed by the steam line break, and then coming
in and having changes to the steam line break that
will affect the changes you have made now with the
LOCA. I mean --
MR. KING: What you are suggesting is
let's take the whole set of DBAs, take a look at what
risk information says about them.
MEMBER BONACA: Yes. I mean, if you're
making --
MR. KING: Make whatever changes you think
make sense and then assess that.
MEMBER BONACA: Or even if you progress
aggressively just with the LOCA, I think you ought to
have in mind some thoughts about how do you agree with
the fact that this is a new approach that most likely
is going to be applied with time to the other
accidents in the way you design your plant. I think
you have to have an understanding; at least I think we
have to have an understanding of how that will come
together.
MR. KING: I agree there are other things
on the plate to be looked at, like the rod ejection
accident.
MEMBER BONACA: Yes. Steam line break,
too, most likely.
MR. KING: Steam line break, probably some
others as well.
MEMBER BONACA: Yes.
MR. KING: And I agree you need to think
about the implications of this for those other events,
particularly if -- if those other events are
preventing some of the -- the benefit or some of the
improvements you're really expecting when you go to
make this change, does it make sense to do that
individually, or would you be better to go in and look
at them as a group?
MEMBER BONACA: Right.
MR. KING: One of the things we need to
spend more time on is looking at, if we would make
such a change here, what are the other things that are
going to catch you? And how do we deal with those
things, so that we take more of a collective or
holistic look at this whole LOCA situation?
CHAIRMAN SHACK: Of course, you know, one
of the problems that Tom keeps pointing out, when you
tackle the design basis approach here, you have a
touch time evaluating the risk, because you don't --
you're not really dealing with a specific change in a
plant.
You know, it's easier to estimate the
change in CDF if the guy comes in and says, "I want to
change my tech spec for diesel startup." I can sit
down and compute a delta CDF score. Here you're going
to change a design basis accident, and you really
don't know exactly what he's going to do in response
to that.
MR. KING: Right.
CHAIRMAN SHACK: So it becomes very
difficult to go back and try to estimate the risk
impact of this change because you don't know what the
changes are, whereas if you're coming the other way
it's a good deal easier.
MR. KING: You have to think through, what
are those changes that would ensue? And what is the
risk impact? And maybe some licensees will implement
them all and some won't.
CHAIRMAN SHACK: But that's the fortune-
telling aspect of this.
MR. KING: Right. But you've got to look
at that. I mean, there's no way around it.
MEMBER BONACA: Although, I mean, for
those changes which are being driven by LOCA -- for
example, the diesel start times -- and you can
determine that there are no other accidents driving
that, okay, then you can, you know -- some assessment
of it can be done.
There are others which are more
complicated because they are driven by other events,
and so I agree with --
CHAIRMAN SHACK: Yes. But could you do a
-- you know, a standard tech spec change for diesel
start times? You know, I mean, we -- I was just going
to make a wise remark that, you know, it was a good
thing you got the jet pipe whip restraints out there
before we decided to risk-inform the regulations.
MR. MAYFIELD: They'd still be there.
CHAIRMAN SHACK: They'd still be there.
MR. MAYFIELD: You were talk about the
tentacles this thing has. There was mention made this
morning that one of the reliefs you'd get is some
improvement. Without making a change in containment
design, you'd get some improvement in the calculated
margin against failure.
If you had that improved calculated margin
and then discovered that your containment was being
degraded, and it was going to be a major deal to go
get to it to repair it, would you then be inclined to
use up that margin, or a part of it, to avoid a repair
for a degraded containment?
I'm not going to speculate, but it seems
to me that that's a question that if you were going to
look at how the -- what changes might be made to a
plant, you get captured in those kinds of things. And
it -- I think that's a very tough thing to do to -- to
get out the crystal ball and guess at all the things
people might do.
MEMBER KRESS: It's crystal-balling,
that's right, yes. That's why it's hard.
MR. WARD: That's one reason that we had
-- are proposing an implementation guide with the
large break LOCA redefinition, so that we go to you
with a complete package of standard changes that had
already been reviewed, so that if I get the rule
change for my plant, then what we have already looked
at on a generic basis for the diesel start times,
accumulator number reduction, ECCS flow balancing, on
and on and on, and have a complete package already
reviewed and looked at as a compiled list, instead of
trying to go at it like this, you know, piecemeal one
at a time, and then figure out the impact, and then to
piecemeal another one and figure out the impact of it.
There are other things, you know, that
would go beyond that, of course, but if we -- if we
try to piecemeal it one piece at a time like this, I
don't think we'll get through it in my lifetime or
most of our careers. I think we will still be working
on that list -- the near-term list, you know, when all
of us are retired.
I think that's the reason we felt like we
had to go all the way to the beginning and start from
there and then work down and come up with a reasonable
list of things to work on at the same time and not try
to piecemeal this thing to death, because I think
that's what will happen if we try to take that
approach.
MR. KURITZKY: Okay. In any case, so on
the rest of these things, I think most of these items
we've already discussed at various times. Some of the
things we're looking at in the longer term, maybe
changing the ECCS acceptance criteria, the propagation
of break size frequency, looking at ECCS availability
for other modes of operation, and things like multiple
steam generator tube rupture, you know, whether
there's a need for multiple steam generator tube
rupture DBA.
Going back to what we discussed before,
even -- or maybe even in the short term looking at the
need for maybe a different diesel generator DBA if
we're going to relax that loss of offsite power
assumption.
So these are some of the things that in
the longer term we still -- we think we can come to
some kind of grips on their feasibility.
MR. KING: Okay. The last viewgraph -- I
mean, the purpose of the meeting today was to sort of
put our cards on the table to show you where we stand.
We don't have all the answers. We still have some
work to do. But at least we wanted to give you the
opportunity here, from all the work that's been done
so far, you know, where we've -- where we are today on
some of these issues and how we see perceiving in the
future both the near term and the long term.
Again, we're subject to scheduling. You
know, we'd like to have some future interactions with
the committee leading up to our June paper. That's
it.
CHAIRMAN SHACK: Just, you know, if you
take the industry's suggestion with, you know, this --
you're not defining -- you're not redefining the large
break LOCA in the rule. You're simply saying the
large break LOCA can be redefined, and they will have
to submit a package showing all the implications of
that in terms of a risk argument.
Doesn't that still leave you in control
and let you evaluate the things the way they -- rather
than trying to crystal ball it --
MR. KING: That's certainly one way to do
it, just open the door for any plant to come in and
say, "Here's my proposal. These are the things I want
to change."
CHAIRMAN SHACK: But I'd have to do a
fully risk-informed version of that.
MR. KING: Risk study, fracture mechanics
study, everything that comes along with it. I think
the question -- maybe Mike is better to answer this
than me -- but are we in the position to lay out the
Regulatory Guide that would have to be met, and the
acceptance criteria that would have to be met?
There's a lot of things on Mike's list of
things that should be addressed in such an analysis,
and, you know, it's not clear to me that at this point
we're ready to say, "This is the list, and this is how
it ought to be approached."
MR. MAYFIELD: I think that that really
was the point I was trying to get to. It's not that
it can't be done. It's that it's, we believe, a much
more significant undertaking than has been suggested
by some of the other discussions we've had.
As we've looked at what would -- what
kinds of things would have to be addressed, it's a
tougher analysis than has been suggested, where we
don't believe what was done before for the leak before
break rule change, in and of itself, was sufficient.
Is it a starting point? Sure. Of course.
But it's a much tougher analysis to do.
The hurdle is higher than it has been for these other
kinds of changes. I mean, it's a tough job. It's not
that it can't be done, although I personally am
skeptical that it's practical, but it's not that it
can't be done. But it's going to be a major resource
investment.
MEMBER BONACA: You said that that was
because for leak before break defense in depth was not
challenged. You said that it was covered by large
break LOCA in 50.46 requirements. So in case we
really missed it, and you have a large break LOCA, you
still have the ECCS capable of dealing with its own
certain criteria.
MR. MAYFIELD: Right.
MEMBER BONACA: Well, have you thought
about the possibility of the criteria being -- I'm
talking about, you know, there are a number of
restrictions to the plant that you could relax based
on a more likely break -- maximum break size. And you
could still impose some fundamental requirements, for
example, coolability of the core, for, you know, that
will end -- you can break.
I'm just throwing out a thought because --
because you still will have some results by the ECCS
that will not deliver exactly for a full -- give you
what you are supposed to, but it will, you know --
MR. KING: You say relax the acceptance
criteria for the large break, is that --
MEMBER BONACA: I'm thinking about the
fact that if you take out the DBA, not only the LOCA,
and you reevaluate those with these new concepts in
mind, you may think about the criteria as something
with -- with expectations for the current DBAs to
still be fulfilled, and probably systems are capable
of delivering that. And, you know, just --
MR. KING: Yes. I think I'm still not
sure exactly what you're proposing. Maybe we need
some discussion.
MEMBER BONACA: Well, once you relax your
diesel starting time, and all of the other things we
have seen here, okay, you still deliver a flow, and
you will have some expectations for coolability of the
core in case -- in case you have the truly double-
ended guillotine break.
MR. KING: I don't know. Given the fact
that the plants already have the right size pumps and
pipes and everything in there --
MEMBER BONACA: That's exactly right.
MR. KING: For existing plants, that
probably is --
MEMBER BONACA: As was said this morning,
that for new plants there will be a different kind
of --
MR. KING: Yes, new plants will be a
different story.
I may want to go back and mention one
other fundamental assumption. We're talking about,
you know, we have a short-term and a long-term list.
One of our fundamental assumptions was, I mean, we
could continue to work on the issues Mike has raised
and do the -- do all of the analysis and develop the
Reg. Guide and then have a complete package type
change.
We're not ready to do that by June, but
one of our fundamental assumptions was we thought it
would be important to get some short-term successes
under our belt on this issue and some short-term, you
know, changes that would get rid of some of these
unnecessary burdens and maybe -- maybe plug some
safety enhancement holes.
That's why we're proposing to go forward
in June with some -- some things that we think are
reasonable, some things that will provide some
advantages, some safety improvements, and continue to
work on the long-term.
I think if we decide not to do anything
until all of the work is done, it's going to take some
more time, and it's going to have negative
implications for all of Option 3. I think getting
some successes under our belt is important, and that's
one of the reasons we're proposing to have this two-
tiered approach. So, you know, I don't know what the
committee's views on that are, or the industry's
views, but that's our view.
MEMBER WALLIS: Well, my impression, if
you want a view, is that you have thought enough about
these issues that you will probably come up with a
good document in April, whenever this -- you know,
April/May, that timeframe. I mean, I've got a feeling
that you know enough about it that --
MR. KING: For the short-term things.
MEMBER WALLIS: -- that you will come up
with something by then, yes.
MR. KING: Right.
MEMBER WALLIS: So I'm sort of encouraged
by the progress I've seen so far.
MEMBER KRESS: I would like to throw out
one of my favorite comments, and that is when you
actually get around to doing this, the risk
implications of the changes -- which may be way down
the road on this thing -- I urge you to reconsider the
guidance in 1.174 a little more, because the overall
objectives of this agency are not just CDF and LERF.
They're releases of all magnitudes of
fission products, including late failures and
contamination, and even worker exposure. So if those
things are impacted and are going to change, you need
to know it also. And when you do the risk impact
evaluation, don't just focus on CDF and LERF. That's
my plain issue.
MR. KING: Okay. We're finished with our
presentation.
CHAIRMAN SHACK: Okay. I believe Mr.
Heymer wanted to make some comments.
MR. HEYMER: Adrian Heymer, NEI. I've
listened to the staff. We've been discussing the
issues with the staff for 15 months now. I think what
you've heard today is very much along the lines that
the discussions have gone for the last 15 months.
We've looked into this. We've evaluated it.
The owners groups have got funding
authorized to move forward on redefining the large
break LOCA, and we believe that's where the priority
should be.
So that's what we believe at the moment,
and we are under no illusions that it's going to be a
tough task. But it's a tough task that's got, we
believe, rewards commensurate with what they're going
to have to expend to achieve our aim, in the order of
probably quite substantial funding from the owners
groups. And by substantial I'm talking in the
millions, not in the thousands.
I hear the staff talk about short term and
near term, and I don't want to ask a question of the
staff, but at some stage I would like a definition of
what near term and short term is in their mind,
because I've got a nasty feeling that what I think is
near term is what they're thinking is something else.
So I'd be interested in hearing from the staff at that
point in time.
We need to think about what's been said
here today again and go back and discuss it amongst
ourselves as an industry, and then see where we go
from here. So that's what I wanted to say.
MR. KING: I mean, I think near term is
propose some changes this June. That's near term. We
want to be able to --
MR. HEYMER: And do rulemaking by the end
of the year?
MR. KING: No, of course not. You know,
we go to the Commission in June and propose specific
changes. They say yes; the rulemaking starts. Now,
rulemaking doesn't take six months. Generally, it
takes two years. Maybe 18 months if --
MR. HEYMER: But, I mean, the -- it would
be out by, say, six to nine months from the time the
Commission said move ahead.
MR. KING: Do you mean the proposed -- the
notice of proposed rulemaking?
MR. HEYMER: Yes.
MR. KING: With the proposed rule?
MR. HEYMER: Yes.
MR. KING: That's probably a reasonable
timeframe.
CHAIRMAN SHACK: Are there any more
comments or questions from the committee members? Any
more questions or comments from the members of the
audience?
Okay. Mike tells me our next subcommittee
meeting on this subject will be May 8th. Whether
that's --
MEMBER WALLIS: It will be a joint
subcommittee like these three subcommittees together.
MEMBER APOSTOLAKIS: Why was -- why
May 8th for the May meeting?
MR. MARKLEY: You decided at the last full
committee meeting on that date.
MEMBER APOSTOLAKIS: Well, that's a good
reason.
(Laughter.)
So May 8th is -- okay.
CHAIRMAN SHACK: That basically means that
the full committee can attend the meeting because it's
sort of banged up against the full committee meeting,
which is probably a good thing.
If there are no more questions or
comments, then I think we can adjourn.
(Whereupon, at 2:07 p.m., the proceedings
in the foregoing matter were adjourned.)