488th Advisory Committee on Reactor Safeguards - December 5, 2001 - Morning Session
Official Transcript of Proceedings
NUCLEAR REGULATORY COMMISSION
Title: Advisory Committee on Reactor Safeguards
488th Meeting - Morning Session
Docket Number: (not applicable)
Location: Rockville, Maryland
Date: Wednesday, December 5, 2001
Work Order No.: NRC-132 Pages 1-74
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
+ + + + +
488th MEETING
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
(ACRS)
MORNING SESSION
+ + + + +
WEDNESDAY
DECEMBER 5, 2001
+ + + + +
ROCKVILLE, MARYLAND
+ + + + +
The Advisory Committee met at the Nuclear
Regulatory Commission, Two White Flint North, Room
T2B3, 11545 Rockville Pike, at 8:30 a.m., Dr. George
E. Apostolakis, Chairman, presiding.
COMMITTEE MEMBERS:
GEORGE E. APOSTOLAKIS Chairman
MARIO V. BONACA Vice Chairman
F. PETER FORD Member
THOMAS S. KRESS Member-at-Large
DANA A. POWERS Member
STEPHEN L. ROSEN Member
WILLIAM J. SHACK Member. COMMITTEE MEMBERS:
JOHN D. SIEBER Member
ROBERT E. UHRIG Member
GRAHAM B. WALLIS Member
ACRS STAFF PRESENT:
JOHN T. LARKINS
PAUL A. BOEHNERT
SAM DURAISWAMY
CAROL A. HARRIS
HOWARD J. LARSON
. I-N-D-E-X
AGENDA PAGE
Opening Remarks by ACRS Chairman . . . . . . . . . 4
Dresden and Quad Cities Core Power Uprate. . . . . 5
. P-R-O-C-E-E-D-I-N-G-S
(8:30 a.m.)
DR. APOSTOLAKIS: The meeting will now
come to order. This is the first day of the 488th
Meeting of the Advisory Committee on Reactor
Safeguards. During today's meeting, the Committee
will consider the following: Dresden and Quad Cities
core power uprate, discussion of topics for meeting
with the NRC commissioners, risk-informed 10 CFR Part
50 Pilot Program, Option 2, and proposed ACRS reports.
A portion of this meeting will be closed
to discuss General Electric nuclear energy proprietary
information applicable to Dresden and Quad Cities core
power uprate.
In addition, the Committee will meet with
the NRC commissioners to discuss matters of mutual
interest. This meeting is being conducted in
accordance with the provisions of the Federal Advisory
Committee Act. Dr. John T. Larkins is a designated
federal official for the initial portion of the
meeting.
We have received no written comments or
requests for time to make oral statements from members
of the public regarding today's sessions. A
transcript of portions of the meeting is being kept,
and it is requested that the speakers use one of the
microphones, identify themselves, and speak with
sufficient clarity and volume so that they can be
readily heard.
I bring the members' attention to the
items of interest handout, which contains four
speeches by the commissioners, two on nuclear security
issues by Chairman Meserve, one on the future of
radiation protection by Commissioner Dicus and one by
Commissioner Diaz on predictability and balance.
Our first item is the Dresden and Quad
Cities core power uprate, and Dr. Wallis will lead us
through this.
DR. BERG: Mr. Chairman, I think Dr. Ford
would like to make a comment first.
DR. APOSTOLAKIS: Dr. Ford.
DR. FORD: Since I have a --
DR. APOSTOLAKIS: Microphone.
DR. FORD: Since I'm a GE retiree, I have
a conflict of interest on this topic.
DR. APOSTOLAKIS: So noted.
DR. WALLIS: Well, the Committee will
remember that we discussed this matter at our last
meeting, and the Committee had a few questions that
remained unanswered. And we also wanted to hear the
resolution of the outstanding issue regarding large-
scale tests. And we hope to be satisfied with what we
hear today regarding those matters. So I think we're
going to start with the staff, then we'll move on to
Exelon.
MR. MARSH: Good morning. My name is Tad
Marsh, and I'm the Deputy Director of the Division of
Licensing Project Management at NRR. We're here today
to brief you on the open item on the draft safety
evaluation, the extended power uprate for Dresden and
Quad Cities. The issue specifically before us is the
large transient tests and the recirc runback test.
These items were discussed with you previously at the
last full Committee meeting, and we have reached
resolution on them.
Specifically, the issue was a proposal by
the Licensee to not conduct the generator load reject
test and the main steam isolation valve closure test
as described in ELTR1. In addition, per your request,
we are also here to discuss with the recirc runback
test.
Now, if you recall, during the last
meeting, we had indicated that we were not able to
discuss these open items with you, because it was
still being evaluated by the NRR staff and by senior
management. This decision that we were pondering at
that point we have now made, and it was a significant
decision with both pros and cons. And we wanted to
make sure that we had a solid basis for our
conclusion.
We've completed our evaluation of the
Licensee's proposal to not conduct the large transient
tests, and we've concluded that it is an acceptable
proposal. Our conclusion is based on the scope and
the extent of the modifications that are being made to
support the power uprate, the numerous tests that the
Licensee is proposing to perform at the system and
component level, evaluations of the systems and
components that are important to the tests under
consideration and the safety benefit, or lack thereof,
from these large transient tests.
I'd like to emphasize that the decision
that we're making is associated only with Dresden and
Quad Cities. It's not a generic decision that we're
making. We will be considering other power uprate
applications on a case-by-case basis with respect to
these tests.
DR. POWERS: Tad, are you laying out
something so that an applicant can reasonable
anticipate what he should have to do in order to avoid
having to do the integral tests or is it really a case
by case sort of thing?
MR. MARSH: Dr. Powers, I think the safety
evaluation lays out the staff's considerations for
both the pros and for the cons, and I believe the
Licensee, in reading the safety evaluation, will see
the elements that we used in coming to that conclusion
so that they would know how to apply or what to do.
DR. POWERS: I guess maybe licensees are
more prescient than I am, because I think I would have
a hard time saying, "Okay, if I do this, this and
this, I have high confidence that I will not have to
do this test." I think I'd have a hard time doing
that.
MR. MARSH: Okay. Let's go through the
presentation. You hear the considerations, and I
believe you'll hear a lot of the elements that we did
that we evaluated that I think would guide licensees.
You're going to hear agreement, to a large extent,
with the specific proposals that were made. And
you're going to hear what these tests do and don't do,
okay? So to that extent, the plant-specific
evaluation may be broader than just this particular
Plant; it may be. But we wanted to leave ourselves
some room in case systems, structures, components,
trips were added to the Plant, which would guide us in
another way.
We've also evaluated the Licensee's
proposal to not conduct the recirc runback test, and
we'll be discussing that with you today.
Let me now introduce the presenters.
Mohammed Shuaibi on my far right will be -- he was the
Lead Project Manager for Power Uprate -- will be
talking about the tests, the large transient tests.
Also, Stew Bailey, who is to my immediate right here,
who is the Project Manager for Quad Cities, will be
presenting the staff's evaluation for the recirc
runback tests.
If no further questions, we'll start with
the presentation. Mohammed?
MR. SHUAIBI: Good morning. My name is
Mohammed Shuaibi, and I'm going to cover the large
transient testing, as Tad said. The tests under
consideration are those included in ELTR1. These are
the load rejection tests and the MSIV closure test.
As you know, the load rejection test is included in
ELTR1 for power uprates greater than 15 percent. The
MSIV closure test is included for power uprates
greater than ten percent.
As I'm sure you're aware, the Licensee
referenced ELTR1 in their application; however, they
proposed to not do these tests, which are included in
ELTR1. So, in effect, what we have is a proposed
deviation from an approved topical report.
The Licensee provided a justification for
this deviation. Since the Licensee covered its
justification in detail during the last few meetings
with the ACRS, I'm only going to summarize what they
said. First, the power uprate is being achieved while
maintaining a constant steam dome pressure. This is
called the constant pressure power uprate -- I'm sure
you've heard that term before. Because the steam dome
pressure is not changing, the effect of the power
uprate on the Plant is significantly reduced. This
also led GE to conclude that these tests are not
necessary for constant pressure power uprates.
DR. POWERS: When GE put together ELT
whatever 1, they thought they were necessary.
MR. SHUAIBI: That's true.
DR. POWERS: What caused them to change
their mind?
MR. SHUAIBI: Basically, ELTR1 covers
different ways that you could achieve the power
uprate.
DR. POWERS: That's right.
MR. SHUAIBI: It allows plants to increase
pressure. It's not limited to constant pressure power
uprates. Also, at the time -- this was back in 1995
and 1996. At that time, they had no experience with
these major power uprates. And I'm not here to tell
you that they have significant experience now --
DR. POWERS: They don't have much
experience now either.
MR. SHUAIBI: Right. But they do have a
couple of data points.
DR. POWERS: Which --
MR. SHUAIBI: Which I'll be covering a
little bit later in terms of how we evaluated that.
That's not the basis for our conclusion.
DR. POWERS: You mean GE just said, "Gee,
we were wrong, we don't need these things anymore"?
MR. SHUAIBI: I think they looked at the
effects of the power uprate, of a constant pressure
power uprates on the Plant. And, basically, they were
convinced that they were able to model the Plant
without having to do these tests for confirmation.
They were convinced that they can do that.
MR. MARSH: I want to point out there's
also a proposal before the staff, a modification to
the licensing topical report to not do these tests,
which is under staff consideration too. So I believe
their thinking has evolved --
DR. POWERS: What I'm trying to understand
is why their thinking evolved. I mean somebody said,
"Gee, GE, this costs us a fortune to do these tests.
Change your mind"?
MR. SHUAIBI: I guess, maybe if I could go
through our presentation, we could tell you why we
came to our conclusion.
DR. POWERS: Right now I'm trying to
understand -- I mean if we don't understand why GE
changed its mind, just say so.
MR. SHUAIBI: We have asked several times
GE for why it was proposed and what changed that would
-- you know, what changed since then. And, basically,
it is that it is a constant pressure power uprate and
that the effect of the constant pressure power uprate
is not that significant on the Plant. I believe GE is
here, and they want to talk to this.
MR. KLAPPROTH: Yes. This is Jim
Klappworth from GE. Let me just reiterate, I think,
Mohammed's statements. I think that's exactly the
situation we're in. When we first came in with the
ELTR, that approach allowed a pressure increase or no
pressure increase, and at the time we did not have a
lot of experience with the extended power uprates, so
we proposed the large transient testing at that point
in time. Since that time, we have evolved in our
power uprate process and evolved to the CPPU, the
constant pressure power uprate process.
In addition, there's been several plants
that have done testing -- KKM, KKL -- and we have some
transient test results from Hatch. So we have some
actual plant data that we've been able to actually
model after the fact and also predict before the tests
are run. So we have high confidence that there is no
need to do the large transient testing. We don't
really gain that much from these testings under a
constant pressure power uprate scenario.
DR. POWERS: What was the rationale for
having the proposal -- I mean when -- I can never
remember the acronym -- ELTR1 came in, it considered
both pressure increases and no pressure increases, but
the testing was for both. And there must have been
some reason for lacking the confidence.
MR. KLAPPROTH: Again, at that time, we
had no direct experimental test results or predictions
at that point in time. We had just started the
extended power uprates. And we did not differentiate
in the ELTR1 between a no pressure and a pressure
increase power uprate.
DR. POWERS: Trust me, I know that's true.
A mystery.
MR. SHUAIBI: Well, maybe -- again, maybe
we could cover what we evaluated when we through the
Licensee's proposal, and maybe that will explain, at
least, how we came up with our conclusion for Quad
Cities and Dresden again. We'll be evaluating this on
a case-by-case basis for the other plants.
Okay. Moving on, the Licensee also stated
that these tests will not provide new information
about transient model --
DR. WALLIS: That's a strange conclusion.
I mean every time you run a test you might find
something happens that you were surprised by. So you
can never say that a test will never give you any new
information.
MR. SHUAIBI: That is true, Dr. Wallis.
I think what I did here is I'm trying to present --
summarize what the Licensee presented, but I totally
agree with you. Any time you run a test, you will
gain some information, even if it's confirmation that
you did things correctly. So there's additional data
that gets collected, and that is some new information.
And if surprises occur, then, yes, you're right.
DR. WALLIS: Then you'd have to use a
different argument, saying that the expected small
amount of information to be gained is not worth the
risk or some task --
MR. MARSH: We're going to get there.
We're going to get there.
MR. SHUAIBI: And that is in our
evaluation. I'm just summarizing what was presented
to you before.
DR. WALLIS: But if they made statements
like that, they're very suspicious statements.
MR. SHUAIBI: Again, once I get to our
evaluation, you'll see those statements a little
differently.
DR. WALLIS: Okay.
MR. SHUAIBI: Okay. Also, they stated
that experience with power uprate shows that GE is
able to predict Plant response following power uprate
in an acceptable manner, and they provided some
information related to that.
DR. WALLIS: That means acceptable
prediction of previous similar tests; is that what it
means?
MR. SHUAIBI: In some cases, it's previous
similar tests; in other cases, although it was not GE
that analyzed the transients, they were plant
transients at Plant Hatch in which no significant
surprises or no surprises related to power uprates
were identified.
In addition, predictions were made for the
Dresden and Quad Cities Plants, and the predictions
showed that no significant change will result from the
power uprate.
And, lastly, the Licensee evaluated
components important to the transient tests under
consideration, again, the load rejection test and the
MSIV closure test, and they concluded that testing is
not necessary based on the effect of the power uprate
on those components.
And to summarize the Licensee's
conclusion, based on its evaluation, the Licensee
concluded that no significant information is expected
to be gained by performing these tests, and that from
a risk and safety perspective, unnecessary Plant
transients should not be induced without a
commensurate benefit.
DR. WALLIS: They didn't actually make
some evaluation of the risk and benefit, did they?
They made a qualitative argument.
MR. SHUAIBI: They did provide some risk
information related to performing the task; however,
it's, I believe, hard to quantify the benefits of the
test. But they did provide risk information -- you
know, the risk associated with performing the test.
And we looked at that.
Okay. Moving on to our evaluation of what
was presented. The staff evaluated the Licensee's
request and justification for not doing these tests.
First, we considered the scope and extent of
modification made to the Plant. We found that the
major modifications are basically on the secondary
systems, not on safety systems.
Next, we examined the effect of the power
uprate on the Plant. Basically, the power uprate will
result in an increase in power level, obviously, and
decay heat; increase in steam flow and feed flow, I
think those numbers were presented as about 20
percent; decrease in pressure and temperature at the
turbine inlet, that's a small decrease, I think it's
less than five percent; and an increase in loading on
some of the electrical equipment.
We then considered the testing that the
Licensee is planning to perform. The Licensee's power
ascension and test plan includes hold points at 50
percent of the pre-EPU power level, at 75 percent, at
90 percent and at 100 percent. The Licensee will
conduct testing and collect data related to steady-
state Plant response.
In addition, after reaching 100 percent of
the pre-EPU power level, the Licensee will increase
power in increments of less than or equal to five
percent, as recommended in ELTR. I believe they
stated at the last meeting that they'll be doing that
in three percent increments.
They will perform additional testing and
collect more data at each increment, at each increase
in power to ensure that the Plant is behaving the way
that they expect it to. What they'll be doing, too,
is they'll be projecting new values for some of these
parameters for the next increment in power level and
making sure that they're able to predict that.
The kind of testing and data collection
includes vibration, system equipment performance,
feedwater pump run-out, fuel delta P, and these are
just examples of the kinds of things that they're
going to be doing.
In addition, the Licensee's performing a
significant amount of system and component level
testing. In addition to post-modification testing,
the Licensee is testing systems and equipment whose
performance requirements have changed as a result of
the power uprate.
DR. BONACA: I need to ask a question.
Are these tests part of start-up testing when the
Plants were being built, so they were standard start-
up tests?
MR. SHUAIBI: Yes. So are the load
rejection tests and MSIV closure tests, though.
DR. BONACA: I mean the question seems to
be there should be, on the part of GE, some objectives
that they had regionally that they wanted to achieve.
And whether or not the power uprate is so significant
that it is almost like a new plant that you have to
do, because I think if it is almost as a new plant,
then you would say you want to have one. If it isn't,
understand the reasons why. But I just was wondering
again if we could get more insights from GE regarding
the original objectives of those tests as part of this
sort of testing.
MR. MARSH: Care to comment?
MR. KLAPPROTH: I guess at this point --
this is Jim Klapproth again from GE -- just to
reiterate the original thinking at the ELTR stage,
and, again, that was back eight, ten years ago, our
thinking has evolved, and I guess I really can't
comment any further as to this issue, other than the
fact that we have learned, we've been able to model
these events, we've been able to predict these events
very well.
In fact, back when we were here with the,
I guess the Subcommittee in June of this year, we
presented the results of the KKM test. I'm not sure
if it was KKM or KKL. It was a KKL test, and we
showed that we were able to model these events very
well. So based on that predictive capability and the
additional Hatch test and the fact, again, that the
Dresden/Quad Cities power uprate is a zero pressure
increase power uprate, we feel there's no need for
testing.
DR. BONACA: I understand.
MR. MARSH: Well, from the standpoint of
the original purposes of the test, and that is why the
staff wants these tests to be done originally, it is
as you say. This is part of the start-up testing
program. When there is a new plant that's been
constructed with new systems, new structures, new
components, new electronics, new trips, new integrated
operations that you want to verify, it is important to
run these tests up to a certain point, which is
normally the full power point.
The question is when you increase power by
a certain point, does it raise questions about those
system structures and components, trip set points to
the extent that you need to run an integrated system
test? In our judgement, it does not.
And that's the -- we looked at the safety
benefit that you would achieve from doing this test,
whether there's enough questions that would be raised
and answered by this power uprate, and we've come to
the conclusion that there is not. There's enough
testing that's done reaching going up to the new full
power level to flush any types of weaknesses that
would occur.
DR. BONACA: I understand. I was asking
the question because if in fact we had some memory of
what the original objectives were, regarding the very
points that you made -- instrumentation --
MR. MARSH: Exactly.
DR. BONACA: -- then one could, you know,
pretty well evaluate why a power uprate of 20 percent
or 30 percent, or whatever it's going to be, is really
necessary, because -- but, I guess, we don't have now
a detailed --
MR. MARSH: If the Plant were modifying
itself in some fundamental way, that is adding new
trip features, if it were challenging limits in some
fundamental way or if there were code prediction
questions, such that we were unsure that ODYN's
capability to predict how the Plant would perform, I
think we'd be at a different place.
DR. BONACA: Okay.
MR. ROSEN: Well, using that exact logic,
Tad, I would conclude that the recirc runback test
fits that, that you ought to run them, because it's
new systems and new equipment. I mean it fits exactly
to use that logic. So help me through why even if
that logic applies to the large transient test, but if
applied to the recirc test, it would seem to lead you
to a different conclusion than I anticipate you'll
tell us in a minute.
MR. MARSH: Can I leave us to that point
to lead us through that logic, rather than diverting
at this point?
MR. ROSEN: But just park that thought.
MR. MARSH: Okay.
MR. ROSEN: I mean I think you just made
an argument for the recirc runback test.
MR. MARSH: Okay.
MR. SHUAIBI: Okay. We also evaluated the
effect of the power uprate on mechanical and
electrical components important to the tests under
consideration. We looked at the effect on MSIVs, we
looked at the effect on control rod insertion, relief
and safety relief valve performance, turbine stop
valves, turbine control valves, scram signal timing,
turbine bypass systems, main generator, an on-class 1-
E switch gear, unit aux transformer, reserve aux
transformer. We looked at all that to convince
ourselves that in fact these tests are not needed and
that the Licensee can justify and that we could accept
this proposal.
Based on this evaluation, we concluded
that the effects of the power uprate on these
components are small. Components are covered by other
tests. In cases, for example, of the MSIV, we have
tech spec requirements that say that the valve must
close no faster than three seconds, no later than five
seconds, things like that. Components are covered by
other tests, and/or the effects are adequately covered
in models used for the analyses. In certain areas
where there's no testing, we look to see, well, what
do we normally expect in that area? And we found that
it's adequately covered in the modeling.
We also considered the need for performing
these tests for code validation. We do not need these
tests for code validation. As you know, analytical
codes are validated using data from numerous test
facilities and operational experience at other BWRs or
at BWRs, in general, that bound proposed operation of
Dresden and Quad Cities.
We then examined the applicability of the
tests to safety analyses. Obviously, these are real
Plant transients that would happen. Non-safety
related equipment would be there to mitigate these
events. Anticipatory trips, which were discussed
before, will be there to trip the Plant. And all
those reasons together, you know, these tests are
going to be much more benign than the actual safety
analyses. So although, as Dr. Wallis said before, you
would get some information, that information doesn't
necessarily confirm that the safety analyses are done
correctly. We do those by other means.
We also considered power uprate experience
presented by the Licensee. The Licensee presented
information related to Hatch, which was uprated to 113
percent of the original rated thermal power, and the
KKL Plant, which uprated to 117 percent of original
rated thermal power. I'd like to note, and I think we
said this before, that we do not consider the
experience at Hatch and KKL to be directly applicable
to Dresden and Quad Cities. Obviously, these Plants
are not identical and the tests may have been done for
different reasons. For example, KKL, I believe, was
a scram avoidance test, because they have different
runback capabilities.
In addition, this experience alone would
not be sufficient to approve the Licensee's request if
all that they presented was the Hatch and KKL data.
We do not believe that we would be here today telling
you that we accept this. However, because of the
implementation of similar modifications at those
Plants to those that were done at Quad Cities and
Dresden, we do believe that to some extent GE's
ability to predict the effect of the power uprate on
component performance is validated. Again, it
wouldn't by itself be sufficient, but we do believe
that it's another data point that we could rely on.
To summarize our evaluation, we believe
that the combination of system and component
evaluation proposed and proposed testing at Quad
Cities and Dresden is sufficient to satisfactorily
demonstrate successful Plant modification without
performing these large transient tests. We believe
that the benefits of large transient tests are not
sufficient to justify the challenges to the Plant and
its equipment, the potential risk associated with
these tests and the burden.
We did not identify any safety concerns
that would warrant the performance of these tests. In
addition, the Licensee's application meets all
regulatory requirements without performing these
tests.
So based on our evaluation, we concluded
that the value added of these tests at Dresden and
Quad Cities is minimal, and therefore we accept the
Licensee's proposal to not perform these tests. And,
again, I'll reiterate what Tad said earlier: This is
for Dresden and Quad Cities. We'll be looking at
other plants when they come in, or as we're reviewing
them.
And that concludes my presentation on
large transient testing. I'll take any questions now.
MR. MARSH: Dr. Powers, you asked a
question whether licensees would know what the staff
would find acceptable or not in terms --
DR. POWERS: I think I'm still kind of at
sea on this, because there's not a number, there's not
a quantification. This is a sense sort of thing.
MR. MARSH: Yes. Our safety evaluation
has in it a table, and the table talks about the
components, the features of the Plant that are
affected by or that would affect the test itself. And
it talks about the tests that are already on those
structures and components, technical specifications or
surveillances or how they're covered by the existing
requirements.
That really forms, I think, the guidance
to licensees about how we judged structures and
equipment and its relevance to these tests. And if
plants had changed those structures and systems or if
they were not tests, as we described in that table,
then they would have to justify deviations from -- or
justify not doing a test in that context.
DR. BONACA: Why did you have to develop
a table? Why didn't the Licensee have to develop a
table referring to the original tests explaining the
basis so that you could support it or not? I don't
understand. I don't understand. The Licensee comes
and says, "We don't want to do the test," and gives
some words. And now you are here developing a basis
for it in your SER.
MR. MARSH: Right. And is your question
why are we doing that?
DR. BONACA: No, I'm saying why shouldn't
it be part of the application? I mean I don't have
anything against not running the tests, but there is
a lot of information licensees -- that GE developed
originally to justify the tests, what the objectives
were.
MR. MARSH: Right.
DR. BONACA: And if you take those
objectives, you could probably, easily demonstrate
that you don't need these tests by doing certain
things. Now, we haven't seen that. All I hear is
that the SER has a table -- and we haven't seen that
either -- that explains these details that Dr. Powers
is asking for and other licensees would want to know.
It seems like a reversal of roles.
MR. MARSH: Well, the Licensee did provide
information that -- you want to answer?
MR. HAEGER: Well, in some sense -- my
name's Al Haeger, I'm with Exelon. We did provide a
table in our submittal of the analysis of the
components and how they're tested and why we believe
that that covers the points that were made by the NRC.
So we did provide a table to them. Now, we did not
specifically talk about the original purpose of these
tests. We hadn't thought of that at the time. Since
then we have seen some other licensees start to do
that. But our submittal did provide a table of the
components and how they're tested and why we believe
the effects of EPU were minimal.
MR. SHUAIBI: That was actually provided
in response to an RAI, but, yes, they did provide a
table. I don't know if it's the same table that
you're talking about, but they did provide a table of
these components and what effect the power uprate has
on these components and their evaluation of the
effects and why it's not necessary to do these tests
in order to confirm that.
DR. POWERS: And the components in that
table do get tested as we make these incremental three
--
MR. SHUAIBI: Several ways that things get
tested. For example, the MSIVs know they're not going
to be --
DR. POWERS: Well, yes, but --
MR. SHUAIBI: Yes. There are several ways
that components are getting tested. Some have tech
spec requirements, some are -- some data is being
taken as they're going up in power. And if you have
one in mind, I could probably go through that with
you.
MR. MARSH: The table has in it, for
example, safety relief valves. Those aren't tested
during this power ascension. The point that's made in
the table is that there's no change in set point to
the safety relief valves, but there is a change in
capacity. We rely on more safety relief valves to
open as a result of the increased power. Turbine stop
valves, the control valves are in that table. There
is some testing of that, but I don't believe it's done
in increments on the way up. There's some testing
that's described in the table that you would do
normally, and it's still contained in the
requirements. Can you remember the features that's in
the table?
DR. POWERS: I can assure you I do not
remember the table.
MR. MARSH: Okay. The point we're trying
to convey is staff looked at the safety benefit from
these tests and found it was not sufficient safety
benefit. I don't want you leave with the impression
that there's no benefit to tests. There would be
benefit to tests. You'd get information, you could
confirm. There are reasons why these tests may be
good to do. And what the staff was faced with was
trying to make a judgment about the merit, the
usefulness of the tests compared to the safety benefit
of the tests. And the difficulty in the decision was
there are some good arguments for doing these tests.
We had to look carefully at the safety benefit and the
impact that you would be -- what you use these tests
fundamentally for and how can you come to a
conclusion.
So we -- we're conveying to you the basis
for making the decision, but I don't want to leave you
with the impression that it was all that clear, that
we ran to the decision because it was overwhelmingly
so. We had to very carefully consider the pros and
the cons. And the staff -- there are still members of
the staff who think we should be doing these tests.
So I don't want to leave you with the impression that
--
DR. POWERS: What is their argument?
MR. MARSH: Well, let me give you some of
the arguments. First, it's consistent with the
topical report. Staff approved the topical report,
and GE proposed doing these tests, so that formed an
Agency opinion. Okay? Now, we can argue why it was
proposed, the bases for it, why we're deviating, but
it still formed an Agency opinion. So that's the
reason why.
It is a demonstration of integral Plant
performance, which from a safety perspective you'd
think that would be a proper direction to go. But
from the standpoint of what do you fundamentally
challenge in this test, are you fundamentally
challenging safety set points, are you challenging
safety equipment? You're not.
But you can see your conservative thinking
leads you towards we should be doing a test. And then
we had to weigh carefully the impact, the benefit, the
purpose, what you gain from it. And we came to the
conclusion that it was not necessary. But we want to
leave ourselves the thought that if something does
change, if there's a system or a structure or if
there's a code issue, we may have to.
DR. KRESS: In making this judgment, did
you run the analysis through the standard back fit
regulatory analysis? Is that the kind of judgment you
made?
MR. MARSH: Probably not, though some of
the thinking may be there, but in the rigorous way.
It's a topical report, so you don't have to do that in
that regulatory process.
DR. KRESS: No, but it's a place where
there are real criteria and real limits and
cost/benefit levels that you can make judgments on.
And I just wondered if you actually went to a
quantification of those things.
MR. MARSH: We did not do that. We did
not.
DR. KRESS: It's a judgement call.
MR. MARSH: It's a judgment. We did ask
ourselves in terms of the four performance goals, our
pillars, what are the merits and demerits of these
tests from the standpoint of our four performance
goals? Does it increase safety? Is it an increased
regulatory burden? Does it improve public confidence?
We asked ourselves those questions too.
DR. KRESS: Trouble is the answer to those
are qualitative.
MR. MARSH: They are qualitative. If they
all aligned in a particular direction, that would
guide you in a certain decision, but they don't. They
don't. Some of them conflict. And from the
standpoint of safety, which is fundamental, we could
not disagree with the Licensee's proposal from that
standpoint.
MR. ROSEN: The two tests we're talking
about are the generator load reject and the main steam
isolation valve closure.
MR. MARSH: Right.
MR. ROSEN: You would agree, would you
not, that those two occurrences are anticipated
operational occurrences within the life of the Plant?
MR. MARSH: True. Yes, sir; they are.
MR. ROSEN: So that in fact the Plant is
going to run these tests one day.
MR. MARSH: At Hatch.
MR. ROSEN: And the only difference
between doing it now and doing it then is that the
Licensee and the staff get to choose the time of the
test, rather than letting the Plant choose the time of
the test.
DR. KRESS: And you can set the boundary
conditions a little better, I think.
MR. ROSEN: Yes, and the Licensee could
have additional staff available, additional monitoring
equipment, management awareness --
MR. MARSH: True.
MR. ROSEN: -- could assure that no other
activities are going on in the Plant at the time that
could distract operators from their -- so we're going
to have a tests, we just don't know when it's going to
be.
MR. MARSH: That's probably true. Hatch
had one, and in fact that's a part of the staff's
observations, that the test was not required for
Hatch, yet they had one. And the Plant performed as
it was projected to perform. There weren't any
surprises. So you're probably right. These are AOOs.
It's not a Condition 2 or a Condition 3 or Condition
4 event, which are the big challenges to safety. The
issue is do you have to mandate when it's going to
occur with all the bells and whistles that go with
that and to what end?
MR. SHUAIBI: To clarify that, we're not
saying that Hatch did the test. Hatch had an event.
They had several load rejection events. And we went
back and we looked at those, and there was really
nothing of significance there that would tell us, you
know, that these Plants ought to do the --
DR. KRESS: What code was used to predict
the results of these tests? Is that a ODYN --
MR. SHUAIBI: ODYN code, yes.
DR. KRESS: And you look at ODYN code and
say, "We know the parameters in there within a certain
level of confidence, and if we run this test, it won't
increase that confidence enough to say that the
predictions would give me better confidence." That's
kind of the judgment type call you make.
MR. SHUAIBI: Well, validation of the
codes is done, like we said earlier, through --
DR. KRESS: It's already done other ways
--
MR. SHUAIBI: -- many, yes, other ways.
DR. KRESS: -- so that gives you a certain
level of confidence.
MR. SHUAIBI: Right.
DR. KRESS: And the only reason to run
these other tests is to see if the code's missing
something or to improve its confidence level. And
you're saying that the test just doesn't --
MR. SHUAIBI: I want to add something on
this issue. I think we discussed this in detail in
coming up with our conclusion here. We do not
believe, although it's been stated before by others,
we do not believe that theses tests are necessary for
the codes, and we did not ask that these tests be done
for validation of the codes. We have other ways of
validating codes. And we have a draft SRP and Reg
Guide on all the kinds of things that you need to do
to validate codes and how you would run plants within
the boundaries established on those codes and
correlations, et cetera.
These tests are really component response
tests to ensure that these components that you're
relying on are going to respond in a way that you
expect them to. So it's not a code issue; it's a
component issue. It's how you model the components
when you run the codes. Are you modeling them
correctly? Is the valve going to shut the way that
you expect it to?
DR. KRESS: I have trouble with separating
that out as not being a code issue, but I guess --
MR. SHUAIBI: Well, I guess, it's -- you
know, there's the code issue of correlations and
things like that, and then there is the modeling of
the components that go into the decks that run the
codes.
DR. KRESS: Well, I consider that's part
of the code.
MR. SHUAIBI: Okay.
MR. MARSH: Jerry Wermiel is here from the
Reactor Systems Branch. Jerry, do you want to add
anything to that?
MR. WERMIEL: No. I thought Mohammed did
a -- this is Jerry Wermiel, Chief of Reactor Systems
Branch. I thought Mohammed's response was right on
the mark.
MR. MARSH: Okay.
DR. WALLIS: The arguments for not doing
the tests are that it challenges the Plant in some way
so there's some risk involved? Is there a risk of the
equipment won't work so well the next time around
because it's been through the test or something? And
then there's the burden. Is that that they have to --
they don't produce power for a period of time?
There's a cost? How big are these things?
MR. SHUAIBI: We didn't actually -- as we
said earlier, we didn't actually go through the back
fit process. First, I think it would be very hard to
quantify the benefits of this test.
DR. KRESS: Yes, I don't know how you --
MR. SHUAIBI: I mean regardless of whether
it's code validation or anything else. I mean you
could probably get some estimates on cost, but what is
-- that's the burden of this test. But what is the
benefit of this test?
DR. WALLIS: No, look at the burden. I
mean there is some benefit and it's in public
confidence, and here's a test which would suggest --
which they propose to do, and if they had done it,
they could say, "Well, we've gone out of our way to do
the test, and we've got more confidence, which is good
for public relations."
MR. SHUAIBI: True, true. But in doing a
cost/benefit, you would need to look at both sides.
DR. WALLIS: So what's the cost to them?
Is it that they don't produce power for some period of
time? Is that the big thing that makes them reluctant
to do the test?
MR. SHUAIBI: Well, I think in terms of
burden -- and maybe they'd want to talk about this a
little bit -- but what we looked at as being a burden,
it's not just the fact that they're going to be down
when they trip the Plant; it's they're going to be
down, they're going to have additional staff that's
going to be at the Plant, there's going to be a lot of
evaluation of the data. You don't run the test and
come right back up. We looked at all that in terms of
burden, but that's only one aspect of it. I mean I'm
not here to say that because of burden we're not doing
these tests.
There's a lot of stuff that we looked at
that's convinced us that these tests are not
necessary. We looked at the equipment and how that's
going to be affected by the power uprate. We looked
at -- and that's what we looked at to convince
ourselves that it's not necessary. And then we also
looked at the burden, and we also looked at the risk
associated with it.
I mean the risk argument is -- it's kind
of -- it's pretty balanced on the risk side, because
you could say that there's risk associated with it,
but you can't come up with a definitive number that
says, "But here's the benefit from it," so that you
could compare it to. I mean it's really qualitative
in that. So you have a number on the risk associated
with it, nothing on the benefit.
So I think what we're here to say is we
looked at the components and how they are affected by
the power uprate.
DR. KRESS: Have these tests been run at
these Plants at the previous power level?
MR. MARSH: I believe they have. I mean
I believe that's part of the start-up testing program.
Please refute that if that's wrong, but I believe
that's part of the start-up testing program, the
initial start-up testing program.
DR. KRESS: So you have the equivalent of
these tests at the old power level.
MR. KLAPPROTH: We do have tests, and we
do have transients. I believe within the last two
years we've had one of each of these at Dresden and
Quad units.
DR. KRESS: So you do have a lot of
information.
MR. KLAPPROTH: We do have a lot of
information at the current power level.
DR. KRESS: And so all you're doing is
extrapolating to a new power level.
MR. KLAPPROTH: Right.
MR. MARSH: It leaves you with a question,
though, and the question -- one of the questions, what
is the original purpose of the test? I think we've
answered that. But it leaves you with the question
of, well, I've increased power by about 20 percent.
Suppose they increase the power by 30 percent or 40
percent? At what point would the staff say, "You know
that's enough of a challenge to a system or enough
questions about the code or enough issues about plant
modifications that I want to run the test." It leaves
us with that question that's not answered.
DR. BONACA: Well, that's exactly why I
was going back to the original tests. They were not
done superficially, okay? They were plants. I
remember the detail. There were justifications,
specific reasons why you're running the test. And I
just am surprised that that information hasn't been
developed to justify why it's not being right now if
the information is available. Now, clearly, it's a
long time since the last plant was started up, and
maybe information has been lost, I don't know. But
something -- I mean that would have kept us from being
here for a long time discussing it among ourselves
what may be in the mind of the people who recommend
that we don't perform the test.
MR. MARSH: I agree.
MR. ROSEN: Specifically, with respect to
the main steam isolation valve test, the steam flow
rates through those lines will be 20 percent higher.
MR. MARSH: Right.
MR. ROSEN: So those valves will never
have closed against that much more steam flow. Now,
my recollection of those valves is that they are
assisted by flow. The closure is assisted by the
flow. So more flow may actually be better.
MR. MARSH: You still have a restriction
of timing.
MR. ROSEN: Right. You can't go too fast,
and you can't go too slow.
MR. MARSH: It can't be less than three;
it can't be more than five, right.
MR. ROSEN: But it is a complex system,
that valve, and it's a huge valve which may be
assisted by increased flow in terms of its closing,
but it may be assisted too much in the sense it may
close too quickly.
DR. KRESS: What's the downside of closing
too quickly? Are you getting water --
MR. MARSH: Too fast of a transient on the
reactor.
DR. KRESS: Too fast.
MR. MARSH: Too much of a power feedback
into the reactor. The pressure increases too fast,
and it causes too much fuel feedback. But they're
testing the MSIVs.
MR. SHUAIBI: They do MSIV testing in
accordance with the tech spec surveillance
requirements. Now, they're not testing the MSIVs with
20 percent additional flow.
MR. ROSEN: But the point is that -- that
was exactly the point.
MR. SHUAIBI: Yes. But they do test the
MSIVs. It's part of the IST Program. They test
MSIVs, and they have limits on both sides in the tech
specs. They have a limit for how fast they can close
and how slow they would close. And looking at that
from -- you know, there's a three-second limit and a
five-second limit. On the five-second limit, these
valves are supposed to shut against steam line breaks
and areas which remain the same for this case. On the
other side, we looked at what exists today to require
the Plant to go back and do testing to make sure that
it will close. There's not a 20 percent additional
flow test that release from 100 percent power.
DR. WALLIS: I wonder if it's time to move
on to the other question.
MR. MARSH: Yes. Can we do that? Move on
to the other -- the recirc runback test? Okay.
Stewart?
MR. BAILEY: Okay. This is Stewart
Bailey. I'm the Project Manager for Quad Cities, and
as you requested, I will give you a brief overview of
the recirc runback system and the testing that the
Licensee has proposed to do that. First, to make sure
that we're on the same page, the recirc runback system
that the Licensee is adding is really for trip
avoidance only. That is its function.
Currently, the Licensee, in operating the
feedwater system, runs three of four condensate pumps
and two of three feedwater pumps. As a result, they
have installed spare of each pump. The way their
system operates right now is if one of these pumps
should fail for any reason, there is an auto-stop at
the standby pump, and that will recover the feedwater
flow and prevent the reactor scram, okay?
If you look at the way the Plant will be
operating following EPU, when they're greater than the
current power level, they need to run these installed
spare pumps. They'll be running four out of four
condensate, three out of three feedwater pumps. And,
therefore, they've lost the ability to auto-start an
installed spare to prevent the scram. What the
Licensee has done is they've added a recirc runback
system that will recover, essentially, some of the
ability that they're losing to prevent that scram.
Basically, what they're doing, I think we
know some of the details of it, is the flow will be
reduced to about 70 percent. And if you look at the
flow control line, that corresponds to roughly the
current full power level. So the runback system
brings them back into the -- basically, back into
their current operating conditions and allows the
feedwater system to recover reactor vessel level.
The system that they put in place is very
similar to the runback system that's used at other
plants. I believe it's very similar to the system
that's been into -- that's already in service at Peach
Bottom. There is some additional logic to it, because
it's only required at high steam flow rates, when
you're at the point where you cannot install -- just
rely on the installed spare or rely on the reduced
number of pumps.
Now, the Licensee has proposed and is
going to be conducting a significant amount of testing
on the recirc runback system, okay? This testing is
being done essentially in accordance with their post-
modification testing procedures. On the feedwater
system, they're doing a number of instrument
calibrations, they're taking a look at the logic for
the staggered pump trips, they'll be tuning the level
control. They have a series of tests where they
adjust feedwater position and/or reactor vessel water
level input and check for the stability of the
feedwater system. And they'll be doing tests on the
feed reg valve position changes and what not.
On the recirc system that they're
installing, they will be doing tests on its circuitry
and its alarms. They have calibrations of the recirc
pump runback speed limiters of the scoop tube final
position, and they will be doing full logic functional
tests on that. A series of overlapping tests so that
there is some overlap between the systems that are
being tested, and this gives them the confidence that
this system will perform as intended.
The Licensee does not intend to conduct an
integral test of the system. They do not intend to,
for example, trip a pump and exercise the runback
system and see whether or not it actually avoids
scram.
MR. ROSEN: But you'd agree that the Plant
will conduct that test some day.
MR. BAILEY: I think it's highly probable
that they will lose one of their condensator feedwater
pumps, and the system --
MR. ROSEN: And the runback system will
come into play.
MR. BAILEY: That is correct. Whether I
would characterize it as a test, I might stop short of
that, but, yes, I think they will exercising this
feature.
MR. MARSH: I agree.
MR. BAILEY: When the staff took a look at
this and the need for the testing, as you mentioned,
the fact that this is a new system would lend some
weight to the Licensee performing an integrated test
of this. Similar to the logic we used for not doing
the large transient testing where they haven't done
significant new mods or trips, here they have done
what could be considered a significant mod and added
logic into it, okay?
But the staff's fundamental consideration
when looking at this is that the recirc runback does
not perform a safety function, okay? The recirc
runback is really there for trip avoidance only, to
recover some of the ability that the plant used to
have in responding to a loss of a pump in the
feedwater system.
MR. ROSEN: But you'd agree that
preventing initiating events, for example, a reactor
scram, is a worthy goal.
MR. BAILEY: I believe that it is a worthy
goal. The staff has considered, though, that they can
approve the extended power uprate without this feature
installed, okay? The licensing topical report, ELTR1,
acknowledges the fact that some plants will be running
their installed spares, and therefore they will no
longer have this capability.
MR. ROSEN: So the staff would be willing
to accept --
MR. BAILEY: That's correct.
MR. ROSEN: -- without the recirc runback
system and let the Plant trip for a condensate pump or
a feedwater pump.
MR. BAILEY: That is correct. And when we
were looking at the risk evaluations associated with
the power uprate, the Licensee did not credit this
feature. The increase in scram initiation off the
feedwater transient just looked at the additional
probability of failing these pumps, and we didn't
credit any recovery from the recirc runback.
So these are some of the considerations.
In additional ELTR1 didn't require an integral test of
the runback system on plants that already have it,
okay? So these are decisions that have been made by
the staff, and due to the limitations on what this
system actually does, the result, if this system
should fail to perform its function, is that you get
a reactor trip, and you end up in a stable condition.
MR. ROSEN: ELTR1 is a GE document.
MR. BAILEY: ELTR1 is a GE document, yes.
MR. ROSEN: It's not a staff document.
We're getting a little confused here by saying ELTR1
doesn't require, as if it were some sort of regulatory
requirement. It isn't; it's simply a GE document.
MR. BAILEY: The staff has approved ELTR1.
MR. MARSH: You're right, though. It's
not a regulatory document; it's a proposal which the
staff has accepted.
MR. ROSEN: Right.
MR. BAILEY: Okay. So I guess the bottom
line is since this does not perform a safety function,
does not prevent anything from reaching any safety
limits, okay, that was one of the staff's primary
considerations in whether or not a test of the system
would be needed. In contrast --
MR. ROSEN: And here again you're using
the word "requirement" in this final slide -- there is
no requirement for an integral test.
MR. BAILEY: Correct.
MR. ROSEN: One could read that with a
bold heading, "NRC Staff Conclusions." I can almost
read like there is no regulatory requirement. Really,
I think what you're saying is you don't feel that on
balance it's needed.
MR. MARSH: Exactly.
MR. BAILEY: I don't think it's needed on
balance. I think there are several people who believe
it would be prudent to run this test to verify. What
they have done is they've run the ODYN codes to
predict reactor vessel water level and to see whether
or not they've lowered the reactor vessel water level
sufficiently to prevent the scram, or whether the
runback goes back sufficiently to prevent the scram.
Running this test, of course, would verify that, and
it would give them information that they could use to
tune their feedwater system response, tune the recirc
runback system or perhaps initiate a further reduction
in reactor vessel water level. But I think it stops
at the level of prudence, and as you've noted, the
probability is high that they will be exercising this
system sometime in the future.
MR. MARSH: Okay. Mr. Chairman, that
completes our presentation on these testing issues.
I hope you've gotten a sense of the staff trying to
weigh the pros and the cons for each of these, and
there are pros and cons associated with each of these
tests. And we've constructed for you, I believe, our
bases for coming to where we are. So thank you very
much.
DR. WALLIS: Does this change the SER?
Are we going to get a final SER?
MR. MARSH: Yes. You will be getting a
final safety evaluation for the large transient tests.
I believe that's true for the recirc runback tests or
is it --
MR. BAILEY: No, the recirc runback tests
we did not go into detail, because it did not perform
a safety function.
MR. MARSH: Okay.
DR. BONACA: That's right. Same
difference there.
DR. WALLIS: So we're going to be asked to
make an evaluation without seeing this final SER
version of the --
MR. MARSH: You will certainly see the
large transient test safety evaluation change. It's
a new section.
DR. WALLIS: When will we see it?
MR. MARSH: We're ready. I thought that
was already here. Did we not send that to you yet?
DR. WALLIS: No.
MR. MARSH: Okay. It will be here today.
MR. ROSEN: And it discusses the recirc
runback as well?
MR. BAILEY: No, it doesn't, because there
is not the safety function associated with the recirc
runback system.
MR. ROSEN: So we will get not further
information than that on the whole question of recirc
runback, will we? The information we have is what
you've provided today.
MR. MARSH: Right. Right. Exactly.
We'll get you the safety evaluation for the large
transient.
DR. WALLIS: Are there any other changes
to the SER?
MR. MARSH: I don't believe so. Singh,
any changes to the safety evaluation? Okay. I don't
believe so.
MR. ROSSBACH: Larry Rossbach, Project
Manager. We have resolved a number of small or open
items that were in the report, and we've done a lot of
editing in response to your comments on the Dwayne
Arnold report. So, yes, there are a number of
changes.
MR. MARSH: But wasn't that set down
before, Larry? Wasn't that before the last full
Committee report?
MR. ROSSBACH: No, it wasn't.
MR. MARSH: Okay.
MR. ROSSBACH: Just the original draft.
MR. MARSH: Okay.
DR. WALLIS: So we may be being asked to
write a letter without seeing the final SER. If there
have been changes as a result of our previous comments
on Dwayne Arnold, then they might be significant.
MR. MARSH: Okay. Singh, do you want to
say something?
MR. BAJWA: No. I was just going to say
that the final safety evaluation becomes final when it
is issued, so the only way we can provide you is once
we actually issue the document.
MR. MARSH: How can we help? We'll give
you the draft updated, the latest draft that we have?
DR. WALLIS: I think that would be useful.
MR. MARSH: Okay. We can do that. Which
includes the large transient parts too, the large
transient sections?
DR. POWERS: I'll remind you that our
Senior Fellow has given us a comparison of before and
after.
MR. MARSH: Okay.
DR. BONACA: Okay. Any further --
DR. WALLIS: Any further questions for the
staff at this time? Let's move on then.
MR. MARSH: Thank you.
DR. WALLIS: Thank you very much. We're
ready when you are.
MR. NOSKO: Thank you very much, Mr.
Chairman, and once again, thank you for allowing us to
come to the Committee.
We have been asked to provide a statement
of concurrence with the staff review. My name is John
Nosko. I'm the Project Manager for the Extended Power
Uprate project for Exelon Nuclear. And I am here to
do that.
Specifically, first, for the subject of
large transient testing, we, of course, do concur with
the presentation, as you heard from the staff. I
believe we've clearly laid out our position in
previous submittals. I would, however, like to make
sure that we clear up the apparent misunderstanding
that was laying there on the table about the no new
information. We did provide a letter the 18th of May
on this very subject, and we did note in there that it
was our conclusion that conducting large transient
tests will not provide significant new information
regarding transient modeling or the performance of
Plant components. And I would like to make sure that
-- I don't believe we have gone through the absolute
of saying "no new information," so just to clarify
this.
And as a final point of note, that last
bullet on this slide, of course, we would, with any
flat transient, we will be collecting data if and when
any of these transient events occur at the Stations,
and we will be making a very thorough and careful
evaluation of those results against the predicted
Plant performances. So, yes our conclusion to the
Committee members is that large transient tests are
not warranted.
MR. HAEGER: This is Al Haeger. There was
some question on the burdens, Dr. Wallis. The primary
burden is the thermal cycle on the Plant and its
components, and some of our components are sensitive
to that and require -- may require repair after that.
And so inducing two of these transients within a short
time, of course, would be an extra cost. And a
secondary cost, of course, is the down time, which is
about two days for each of them.
MR. NOSKO: Okay. Moving on then to the
recirc runback test and the should we or should we not
trip a feed pump as part of a performance test for
that new operating feature, I believe Mr. Bailey
summarized it very well, that the modification has
been designed to provide us with operating margin
only, help us ride out a low-level scram. We believe
that the scope of our plan testing does provide
adequate demonstration of the capability of this
modification to perform its intended function. And
our conclusion is, as the staff has concluded, that
the feed pump trip at power is not warranted.
And with that, the Committee also had, I
believe, three additional questions that they would
like Exelon to address, and we will be bringing people
forward to answer those three questions at this point
in time. The first speaker will be Dr. Jens Andersen
from Global Nuclear Fuels.
MR. HAEGER: Let me summarize the
question, as I understand -- as we understood it from
last time. It's regarding fuel energy deposition
limits and it was to justify the use of the 170
calorie per gram energy deposition limit that was used
for determining that the potential instability during
an ATWS event was acceptable in the light of the data
that some of the members, or one of the members had
seen regarding fuel failures at different exposures.
We have provided these in writing. I
would hope that you had seen those answers in writing.
Is that correct?
MR. SIEBER: Yes.
MR. HAEGER: Did you get those?
MR. SIEBER: Yes.
MR. HAEGER: Okay. Thank you.
DR. POWERS: I certainly have not seen
them, but I've been off in Europe looking at high
burn-up fuel assemblies getting exposed to power
insertions and blowing apart.
DR. ANDERSEN: This is Jens Andersen from
Global Nuclear Fuel. Al Haeger just reiterated the
question, and the issue that for the EPU we really --
we are not extending the burn-up limits. The burn-up
limits are the same as they have always been. We are
not increasing the duty on the fuel and really no new
phenomena introduced.
What we have seen in previous analysis for
the -- typically, when we looked at the ATWS
instabilities and the duties that could be imposed on
the fuel, we have seen enthalpy energy depositions in
the fuel that are significantly lower than the current
limit. As discussed in the report, NEDO-32047, that
was submitted a long time ago on the ATWS
instabilities, what we saw was maximum energy
depositions that was in the range of about 30 to 80
calories per gram. Eighty calories per gram was
typically off fresh fuel, high-power bundles. Highly
exposed fuel out in its third cycle were down in the
30 calories per gram. That is significantly below the
current failure limit.
DR. POWERS: Well, what -- I mean you keep
saying "current failure limit." The current failure
limit, I think, for BWR fuel is like 170 calories per
gram. And I haven't seen a fuel assembly -- or fuel
rod survive 170 calories per gram for burn-ups above
25, 30 gigawatt days for ton. I mean it's just not a
meaningful limit anymore. I mean you're telling me
something that's of no impact whatsoever.
MR. HAEGER: Well, we need to finish the
answer to the question, okay?
DR. ANDERSEN: What we are seeing is that
-- clearly, the point I'm making is that we are not
extending the burn-up limit as part of the EPRI
analysis. And the Robust Fuel Working Group, they're
looking at what the burn-up limit should be. And I
understand that they may want to change those burn-up
limits for high-exposure fuel.
What we are seeing is that the energy
depositions we are getting for fresh fuel in the order
of 70 to 80 calories per gram, which is less than half
of the current failure limit. And I believe it's less
than what EPRI is thinking about reducing the failure
limit to for high-exposure fuel. For high-exposure
fuel, what we typically see are energy depositions
that are in the order of 30 calories per gram, which
is about a factor of five less than the current
failure limit.
DR. POWERS: Now compare it to where fuel
rods are failing in reactivity insertion events.
DR. ANDERSEN: What our accidents for
activity insertion events are, we did an extensive
analysis of that following the complete event. And
what we do is with the rod bank withdrawal sequence we
are limited to the maximum rod thrust of less than one
percent delta K over K. For a rod drop accident of
one percent delta K over K for fresh fuel, we get
roughly 70 calories per gram. Forty gigawatt days per
ton, we are down in the order of 25 to 30 calories per
gram.
Again, high-exposure fuel, the energy
deposition is almost an order of magnitude less than
the current failure limits. Therefore, we believe
that there is no safety issue, even in light of the
fact that if we're considering to lowering the current
failure limits.
DR. POWERS: Well, since you're not going
to quote the levels at which fuel rods are coming
apart, maybe I'll just quote some numbers to compare
against. We have fuel rods disassembling at 80
calories per gram, 50 calories per gram and one
remarkable rod at, depending on how you count, either
36 or 18 calories per gram. I mean to say -- quoting
against this 170 calories per gram is of historical
interest only.
DR. ANDERSEN: Yes, I understand that.
And I also understand that that particular test -- I
presume that you are referring to the CABRI test -- is
really not representative of BWR and the way BWR fuels
operate and the power that BWR fuel could be exposed
to.
DR. POWERS: I think that's a very fair
statement, because, admittedly, a PWR rod, what not,
and, quite frankly, I'm not absolute positive any BWR
rods have been tested, which itself says something.
MR. HAEGER: Well, our point, in
conclusion, is simply that adoption of more
restrictive limits for an EPU, in which no other
effects are changing, is not warranted at this time.
DR. POWERS: I think what you're saying --
and I think I agree with you -- is that right now
you're in compliance with what the staff's required.
It's just not an issue here.
MR. HAEGER: And we do feel that it's low
potential safety significance based on the data that
we have. We understand --
DR. POWERS: Well, I don't know about that
one. That's kind of -- that one I think we could
argue over, maybe over a beer sometime. The next
statement that you make is there's no -- the EPU puts
no additional load on the fuel. That's one that's a
bit of a mystery to me. Could you explain that a
little more? Why isn't there a 20 percent more load
on the fuel here?
DR. ANDERSEN: The peak power level of the
fuel doesn't really change.
DR. POWERS: Change.
DR. ANDERSEN: The rod lines -- the
maximum rod line does not change. So peak power of
the fuel is really the dominant effect of --
DR. POWERS: That's what you're saying.
DR. ANDERSEN: -- what controls the load.
DR. POWERS: It definitely puts more load
on the fuel, but it's not changing the peaks is what
you're saying.
DR. ANDERSEN: Right.
DR. KRESS: Unless the instability
magnitudes change because of increased burn-up.
MR. HAEGER: Yes, and we discussed that
last time, and I believe --
DR. KRESS: And they did change some, but
best I remember, they --
MR. HAEGER: I believe that GE last time
presented a sensitivity study that said that the
generic ATWS analysis applied -- it looked like it
applied to the EPU condition, in fact, the MELLA Plus,
in fact, testing that they've been doing.
DR. KRESS: Yes. Okay. So you get the
same sort of --
MR. HAEGER: That's correct.
DR. ANDERSEN: The rod line doesn't
change.
DR. KRESS: Yes.
MR. HAEGER: Okay. We'll go on then. We
have the next group of two questions Keith Moser is
going to respond to. Keith is our Reactor Internals
Program Manager for Exelon. As I understand, the
second question was referring to the dryer lug
attachment inspection frequency, and we came back with
that last time but really didn't address the
significance of that frequency. Keith is here to do
that first.
MR. MOSER: Yes. First of all, what I
want to do is say we just got out of the Dresden 2
outage, we went back and we looked at the dryer before
we installed the mod, and then we went and looked at
the lugs. And then after we got the mod done, put the
dryer back in, we looked at all the four set points to
make sure we had contact.
Now, after 32 years, these were holding up
really well. We saw no degradation. So with that and
the fact that when we did the modeling and the
endurance limit was 10,000 psi, code limit of 13,600
psi, so you already have like a 36 -- I'm sorry --
yes, 36 percent safety factor inherent in that, and we
met the 10,000 psi endurance limit.
We have a pretty good basis for saying
that the ten-year interval established by Section 11
is appropriate. But because we know that there's BWR-
4s, BWR-6s that have this problem, we've made the
determination that after one cycle we're going to go
back and take another look for all four units and make
sure that we don't have a problem and then we'll go
back and reassess if the ten-year interval is an
appropriate length for us to go. Does that answer
your question?
DR. FORD: Yes. The reason for the whole
topic to start with was that during one of the
presentations that were made you said that there would
be an increased vibratory stress transferred to the
lug, which originally made we think of the fact that
the failure mode is not necessarily fatigue on which
the core is based, just stress corrosion cracking of
the lug. And it is generally the incidence of stress
corrosion cracking will increase with a small
vibration load regardless of fatigue as being the
phenomena of failure.
And I wondered how your inspection
periodicity might be changed because of that
particular phenomena, that is stress corrosion
cracking accelerated by a small superimposed and
increased vibration load? And that was the physical
origin of the question. The fact that you are going
to maintain your inspection periodicities, as I
understand it from your reply, less than ten years, it
should mitigate that problem.
MR. MOSER: I believe it will. Also, I
talked to Sam Ranganath about this and Dave Randall,
and in these cases, for the BWR-4 and BWR-6, it was
fatigue.
DR. FORD: Okay. It was transgranulas?
MR. MOSER: Well, they didn't take a vote
sample, per se, but it was classic. It may have had
a small starter point with IGSCC, but it's pretty
clear that it was fatigue in this case.
MR. HAEGER: The final question was on
flow accelerated corrosion. We had a predictive model
that showed an increase in wear rate in some portions
of the feedwater system, from a projected 19 mils per
year to 21 mils per year. And we didn't have at our
fingertips what the thickness of the pipe was at that
time and then some replacement frequency information,
so we've provided that here.
MR. MOSER: Yes. And once again, we're
talking about some fairly significant pipe -- 120 for
the schedule, 18-inch diameter and 24-inch diameter.
And what we're looking at with the CHECWORKS model,
we're looking at a place where 2010 is where we're
saying we think we may want to be thinking about
replacement, not necessarily replacing. But we're
going to inspect in 2008.
Now, prior to the EPU, we were saying we
could go all the way out to 2010 and then thinking
about replacing in 2012, but, you know, we're talking
about something that's fairly thick, fairly
significant. The projected wear rates are actually
overestimated, if you will, and we feel like we're
very conservative in this manner.
MR. HAEGER: We should emphasize these are
localized components. These are things like reducers
or elbows, not entire pipe sections, obviously.
MR. MOSER: Actually, the component was --
DR. POWERS: Actually a hole in any part
of the pipe is a hole.
MR. MOSER: Well, yes, but we --
DR. POWERS: It doesn't matter whether
it's local or --
MR. MOSER: We thought you may be
envisioning entire pipe replacements or something, I
guess.
MR. HAEGER: And the component that was
most susceptible was the concentric reducer.
DR. WALLIS: What are the actual wear
rates? You have had inspections of those components.
MR. MOSER: Yes. They're --
MR. HAEGER: Thirteen to 16. I think the
pre-EPU actual rate was 13.
DR. WALLIS: So they're comparable but
lower.
MR. HAEGER: Yes. Right.
MR. MOSER: And then the thing you have to
do with -- we have two inspection data points and with
CHECWORKS you want to do the predictive module rather
than just count on two data points. So we're being
somewhat conservative. Any other questions?
MR. HAEGER: Okay. Well, again, we want
to thank the Committee for allowing us to come back,
and we'll look forward to your deliberations. Thank
you.
DR. WALLIS: Are there any other questions
for the staff or Exelon or GE?
DR. POWERS: Well, let me ask a question.
The answer may be it's covered by a different part.
And that is we're going into a combination of power
uprate and inevitably using our fuel to hire burn-ups,
and we have a concern in this power uprate about the
possibility of getting into an oscillatory regime.
And we have a process -- a procedure for recovering
should we ever get into that oscillatory regime that
is -- it's not trivial.
I mean it does involve dropping the level
of water, injecting some boron and bringing the water
up. Do we understand what kinds of stresses, lateral
stresses that puts the fuel under, and are we
confident that that fuel survives those lateral
stresses?
MR. HAEGER: Dr. Andersen, I don't know if
you can help with that. That's something I wouldn't
be prepared to discuss.
DR. ANDERSEN: I'm not prepared to discuss
the stresses on the fuel.
MR. HAEGER: I guess the only response we
can give is that the ATWS instability study that --
and help me out if this is not right -- but that study
was a generic study that since GE has confirmed that
they believe applies to power uprate. Now, I don't
know if that covers lateral stresses on the fuel or
not.
DR. ANDERSEN: What the study showed was,
as I discussed earlier, that the energy depositions
that you had in the fuel was substantially lower than
what the current failure limits are. As I said
earlier, the energy depositions, the max energy
deposition we saw on fresh fuel was in the order of 70
to 80 calories per gram. For highly-exposed fuel, it
was down in more like 25 to 30 calories per gram.
What we did see that in the absence of an
mitigating event, that you could get translations on
the fuel and some fuel failures on a very small
fraction of the fuel in the core. Those failures were
oxidation failures, not stress failures.
DR. POWERS: Let me ask a question about
that and power input. Originally, you quoted power
inputs for a rod bank withdrawal. In the oscillation,
we're talking about a little bit different; that is,
a series of cycles that each -- which is putting power
into the fuel. Seems to me that the power and input
then depends on the number of oscillatory cycles you
go through.
DR. ANDERSEN: That is correct, but if you
look at the net energy deposition in the fuel, that is
equal to zero, because the heat removal rate to the
coolant is equal to the energy deposition into the
fuel.
DR. POWERS: Oh, yes? Temperature doesn't
go up at all?
DR. ANDERSEN: If you fail to relieve it,
then the temperature goes up, and that's correct. It
will go up and you will get a failure in a small
fraction of the fuel. If you look at the old report,
the NEDO-32047, we showed that that could happen to a
small fraction of the fuel that had the highest power
oscillations. When we analyzed how large a fraction
of the fuel that could be exposed to that failure
mechanism, it was less than half a percent of the
fuel. And that was deemed acceptable at that point.
DR. POWERS: And if we use a lower failure
criteria, what does that percentage go up?
DR. ANDERSEN: The failure rate was really
not associated with the energy deposition. The
failure rate was associated with the fuel during the
periodic oscillation exceeding the minimum fuel
burning temperature and failing to relieve it. And
then the failure was really an oxidation failure at
high temperature. So it was really not associated
with the energy deposition.
DR. POWERS: And I don't know if it's
appropriate in response to your question to comment on
operator mitigation, but, certainly, that's a
consideration.
MR. HAEGER: Yes. That's a question that
we continue to wrestle with. And I never know how to
come down on this. I mean you test the operators with
the current configuration in the Plant, and they do
very well.
DR. POWERS: That's right.
MR. HAEGER: And so now we're going to say
that the higher power, where they have a little less
time, they'll still be able to do very well.
MR. HAEGER: Tim, do you want to comment
on that?
MR. HANLEY: My name is Tim Hanley from
Quad Cities. The way we test and train the operators
really is that they can initiate it based on the
parameters they're monitoring, not a specific time
period. So their reaction to the ATWS event in a
uprated core versus a non-uprated core is not
significantly going to change, because it's based on
parameters that they're monitoring, not specific time
criteria.
DR. BONACA: I mean I understand what
you're saying, but if the parameter reaches a certain
point at which he has to take action in ten seconds,
right, it would make a difference --
MR. HANLEY: It certainly would if it was
making changes in that time period, but we're talking
on the order of minutes, not in the order of seconds
here.
DR. BONACA: But this argument has been
made many times that way. I think there is still a
sensitivity to the timing.
MR. HANLEY: Well, certainly, certainly.
If it was changing it to a matter of seconds or from
ten seconds to five seconds, that would make a
difference, but we're talking in the order of minutes
where the operators have an opportunity to assess the
situation, monitor the parameters and take the correct
actions.
DR. WALLIS: These extreme oscillations
that you gathered in ATWS, your understanding of that
is based on codes and theory or are there some sort of
experiments?
DR. POWERS: Well, a couple of events. A
full-scale event at La Salle.
DR. WALLIS: Okay. So does that give
confidence that you really understand what would
happen with these power uprates with the ATWS
oscillations?
MR. HAEGER: Again, we have covered this
in the past, but Dr. Andersen?
DR. ANDERSEN: Well, these are complicated
coupled thermal-hydraulic neutronics oscillations. We
do have full-scale reactor test data for thermal-
hydraulic coupled neutronics oscillations. We also
have some unplanned event. We have full-scale testing
that has been conducted on several European plants.
Particular tests were conducted at the
Leibstadt, the KKL Plant. We also have the event that
happened in 1988 at La Salle. We have used those
events to qualify our codes, and we actually predict
those events very well.
DR. WALLIS: Anything else? Then I would
thank you very much. And I will hand the meeting back
to the Chairman.
DR. APOSTOLAKIS: Thank you, Dr. Wallis.
We're 11 minutes early. Very good.
DR. WALLIS: Well, with thermal-hydraulics
things go quickly.
DR. APOSTOLAKIS: They go very quickly,
yes.
(Laughter.)
DR. POWERS: Because a momentum equation
doesn't come up.
DR. APOSTOLAKIS: Okay. We'll recess
until 10:30.
(Whereupon, at 9:58 a.m., the ACRS
Advisory Committee Meeting was recessed until 10:30
a.m.)
Page Last Reviewed/Updated Monday, August 15, 2016