Protecting People and the EnvironmentUNITED STATES NUCLEAR REGULATORY COMMISSION
UNITED STATES
NUCLEAR REGULATORY COMMISSION
***
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
***
476th ACRS MEETING
***
Conference Room 283
Two White Flint North
11545 Rockville Pike
Rockville, Maryland
***
Friday, October 6, 2000
The above-entitled meeting commenced, pursuant to
notice, at 1:30 p.m.. MEMBERS PRESENT:
Dr. Dana A. Powers, Chairman
Dr. George Apostolakis, Vice-Chairman
Dr. Thomas S. Kress, ACRS Member
Dr. Robert L. Seale, ACRS Member
Dr. William J. Shack, ACRS Member
Dr. Robert E. Uhrig, ACRS Member
Dr. Mario V. Bonaca, ACRS Member
Mr. John D. Sieber, ACRS Member
Dr. Graham B. Wallis, ACRS Member
Mr. Graham M. Leitch, ACRS Member
OTHERS PRESENT:
John T. Larkins, Executive Director, ACRS
Noel F. Dudley, ACRS Staff
Michael T. Markley, ACRS Staff
Howard J. Larson, ACRS/ACNW Staff
Ralph Beedle, Senior Vice President, Nuclear Energy
Institute
Alex Marion, Nuclear Energy Institute
Lynette Hendricks, Nuclear Energy Institute
Doug Walters, Nuclear Energy Institute
Doug Walters, Nuclear Energy Institute
Tony Pietrangelo, Nuclear Energy Institute
Mike Mayfield, Office of Nuclear Regulatory Research, NRC
Ed Hackett, Office of Nuclear Regulatory Research, NRC
Safish Aggarwal, Office of Nuclear Regulatory Research, NRC
Robert Lofaro, Brookhaven National Laboratory
Amarjit Singh, ACRS Staff
Jose Calvo, Office of Nuclear Reactor Regulation, NRC
Richard P. Savio, ACRS Staff
Paul A. Boehnert, ACRS Staff
Mark Cunningham, Office of Nuclear Regulatory
Research, NRC. P R O C E E D I N G S
[1:30 p.m.]
DR. POWERS: The meeting will now come to order.
This is the second day of the 476th meeting of the Advisory
Committee on Reactor Safeguards.
During today's meeting, the committee will hold a
discussion with industry representatives concerning industry
issues.
We will also discuss GSI-168, equipment
qualification; generic guidance documents associated with
license renewal; future ACRS activities; and, have a report
of the Planning and Procedures Subcommittee.
We will reconcile ACRS comments and
recommendations and discuss proposed ACRS reports.
The meeting is being conducted in accordance with
the provisions of the Federal Advisory Committee Act. Mr.
Howard Larson is the designated Federal official for the
initial portion of the meeting.
We have received no written statements 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 speakers use one of the
microphones, identify themselves, and speak with sufficient
clarity and volume so that they can be readily heard.
With that little piece of business out of the way,
Ralph Beedle, welcome to the meeting with the ACRS. As I
commented earlier, we have gotten just absolutely heroic
service from members of your staff here supporting us and
pursuing various individual issues.
We never get a chance to sit down and talk
philosophy, talk approaches, talk strategic plans, and we
have had a chance to examine your strategic plan and I
congratulate for a well written and very useful document.
We hope to be able to cover four or five areas of
great interest, but I'll turn it to you to give any opening
comments you'd like to make.
MR. BEEDLE: Thank you very much, Dr. Powers. I
appreciate the opportunity to join you today.
Let me, I guess, follow-up on your admonition. My
name is Ralph Beedle. I'm the Chief Nuclear Officer for the
Nuclear Energy Institute, here in Washington, D.C.
With that, I would like to start with kind of a
brief overview of NEI. NEI, Nuclear Energy Institute, is
about six years old. We were a result of the consolidation
of NUMARC, the technical group within the nuclear industry
that interacted with the NRC on a variety of licensing and
technical issues, as well as the ANEC, the lobbying
organization.
So we tried to combine those and provide an
industry focus to the collection of lobbying, advertising
and the technical issues that were being dealt with by the
industry.
There are about 126 people at the NEI
organization. I believe you have an organization chart that
shows the basic organizational structures, divisions that
represent nuclear generation, basically the power reactor,
the technical side, business operations that are looking at
issues associated with benchmarking of perhaps new plant
construction in the future. Also, their focus is on Yucca
Mountain and the activities associated with high level waste
disposal. Then we have the typical communications and
administration.
More importantly, the organization was -- there
was a recent change in it, where we created a division under
Angie Howard to focus on policy. That, in large measure, is
a result of the consolidation with the industry and we see
our roughly 200 members decreasing in number as both the
utilities consolidate and as the NSSS vendors consolidate.
We expect we will see considerably more of that
activity as time goes on.
So with that, the question is how do we provide
the services and support for our members and as a result of
trying to answer that question, we created this group under
Angie Howard to provide some clear focus on that.
We are involved in a great many activities, and
they, as you can imagine, range from activities on the Hill.
We support Congressional information trips to Yucca Mountain
and to other nuclear facilities, to help inform our elected
representatives of what the nuclear business is all about.
We also spend a great deal of time in the
communications area. As you drive to and from work in the
Washington, D.C. area, you probably hear some Nuclear Energy
Institute ads on the radio, tune into WTOP-1500 on your AM
dial and you'll hear those.
And, again, those ads, I think, have an effect in
the long haul. They, combined with the performance of the
industry, the increased focus on electricity and demand,
energy issues are becoming more and more a subject of
conversation, and the net result is that nuclear is looked
at differently today than it was four years ago, even three
years ago, and, in large measure, that's a result of the
effort on the part of a lot of people to try and educate the
policy-makers.
We have often that it would be almost mission
impossible to try and educate the public. It's the desire
to go into the public school system and provide educational
devices and programs. It's almost impossible to do that.
So our focus has been on the policy-makers, the
people that are making these decisions that affect our
day-to-day life. It's the decision to release or not
release oil from the strategic reserve. Those things have a
profound impact on whether or not you're going to have home
heating oil in New England next winter or whether or not
you're going to have gasoline next summer.
And part and parcel to that is a better
understanding of the entire regime of energy supply within
this country.
As a philosophical approach to this whole process,
NEI has been endeavoring to make sure that people understand
the value that nuclear brings to this country. Nuclear is
not a religious event. It's a business. We operate
businesses based on a technology called nuclear and if we
can't make money in that business, you're not going to stay
in business. It's as simple as that.
One of the reasons that we have not built a new
plant in this country for roughly ten years is the fact that
the cost of construction is high and the rate of return
isn't all that great.
Not only that, the tremendous uncertainty
associated with the regulatory process makes it a tremendous
risk on the part of any business venture that you might
undertake.
So our effort has been to try and develop a stable
regulatory environment, one in which there is predictability
and surety for the licensees as they go forward in the
operation of these plants.
I would argue that the license renewal program, as
successful as we deem it to be, is successful only because
of the surety that we have in the licensing process and in
the operational process that we have today through the
revised reactor oversight process. A little later, we'll
talk about both of those programs.
So this approach that we have of trying to get
clarity, surety, objectivity in the regulatory process and
environment is a necessary ingredient if the nuclear
programs are going to survive in this country.
So that, coupled with the significance of 20
percent of the generation in the country being produced by
nuclear generation is rather significant. We look at the 20
percent capacity basically and almost every year we see that
that is exceeded in terms of generation, and we generate
somewhere in the 23 to 25 percent, simply because of the
reliability of these power plants.
We find more and more effort on the part of the
utilities to provide that reliability. I'm sure you're all
familiar with the problems associated with power supply in
the Illinois area, with ComEd, and the relatively low
capacity factors, and then with Oliver Kingsley and his new
approach to management of that company. It was all ten all
summer, this past summer, and they did a very good job at
doing that.
The energy supply issue in the center of the
country isn't near as critical as it was last year, simply
because of the performance of the plant. But nonetheless,
it points out the value that nuclear adds to that mix.
The other element of the nuclear equation is the
clean air and with the Kyoto Accord and the focus on clean
air, there is only one way that you and I are going to
achieve clean air in this country, and that's through the
use of nuclear as a key ingredient in the energy mix.
If we were unable to keep license renewal on
track, we would have a very significant problem in trying to
meet those accords.
So the combination of clean air, regulatory, and
the emphasis clean air puts on this energy source, the
license renewal process, improved environment, all that has
caused nuclear to become more and more of a subject of
discussion within the Administration, within the political
leadership in this country, and I think we're going to see
eventually a major shift in that focus, such that we will
have an environment in which we can produce new nuclear
generating facilities.
Let me ask, are there any questions about NEI and
what we're trying to do?
DR. POWERS: I guess I'd like to understand a
little more what you mean about stability in the regulatory
process, because we may be working at cross purposes here.
We've got a committee that historically is, for as long as
I've known the ACRS, and it's getting to be pretty close to
26 years that I've appeared on one side of the table or the
other in front of this institution, has endorsed a move
toward a risk-informed regulation or even a risk-based
regulation.
That's a pretty profound change and that doesn't
seem to be consistent with the idea of regulatory stability.
MR. BEEDLE: Well, I would take a little bit
different view. I think risk-informing the regulation is
very much consistent with stability.
When we look at the reactor oversight process, and
I think that is really the fundamental change that's taken
place in the last couple years, and it is the single most
significant thing that has occurred that's produced some
stability in the regulatory environment.
And it's an interesting situation because we went
into this revised reactor oversight process with the
objective of having clearly defined performance criteria
through the performance indicators and the various
categories and from that, then, a predictable outcome.
I would say that prior to this revision, there was
not a chief nuclear officer in the industry that didn't feel
that he was only one event away from the watch list. One
event away from the watch list. And there wasn't one chief
that didn't feel that way, because what they saw happening
in the industry is the event occurred and then the NRC would
send their inspection team in and they would spend as much
time as necessary to develop a catalog of problems and we
had an uncanny ability to make mountains out of molehills
and we said every problem is important and you have to fix
every problem that we've identified before you can ever
start up again, and we were automatically in a $200 million
fix, get well program and about two years to accomplish it.
It was predictable in that sense. You had an
event. If we could have just figured out how to write a
check for $200 million and be done with it, I think most
chief nuclear officers would have been glad to do that.
So that unpredictability and inability to look
long haul and say I can guarantee that I will be able to
operate, it's not to say you aren't going to have some
problems, but to have some reasonable outcome when you do
examine some of the problems, that was what we were after.
So we go into the oversight process and we have
the set of performance indicators, on the one hand, and we
said, well, that's going to really help us out because if
I'm in the green band, you kind of leave me alone, let me do
my job; if I'm in the white band, you increase the
inspection hours and we try and determine what caused you to
be in the white band, and we accept that, we'll go figure
out how to make things -- improve the performance and get
back in the green band.
So, I mean, if you give me a target, I know how to
achieve performance to match that target. It's when you say
shoot at the wall and then you draw the target that we have
our problems.
So the oversight process gave us some targets that
yielded predictable outcome on the part of the inspection
group.
DR. POWERS: Maybe it gave you more. It gave you
trends so that you didn't have this roll-off-the-table kind
of phenomena, that you can see when things are becoming --
maybe your performance was beginning to lag a little bit.
MR. BEEDLE: Absolutely. Absolutely. And with
those trends, though, as you cross thresholds, you knew what
would occur as a result of that change.
The other component in this oversight process was
the inspection and there is no question about the fact that
there are many, many things that take place in this power
plant operation that don't lend themselves to performance
indicators.
And, indeed, the idea of the performance
indicators was developed with the idea that once you have a
set of indicators, if the indicators indicate a declining
performance trend, it then gives the NRC the ability to use
those indicators to focus their inspection resources and go
examine those areas that represent the declining
performance, and you focus on that as a way of getting
better understanding of the problems and ultimately improve
the performance, as opposed to the just send in and inspect
everything.
You could have a problem with a steam generator
and they'd be in there inspecting the rad waste system. You
say what's the purpose of that; well, we're just looking for
problems, you know. So anyway, you end up with this
performance indicator process.
In order to evaluate the significance of the
findings that are developed during the inspection process,
the NRC constructed a process called significance
determination. And so the SDP or significance determination
process is probably the single most important element to
come out of the revised reactor oversight process, because
it helps the NRC and the licensee understand the
significance of an event, of a finding, of some disturbance
in the normal order of business.
It tells you whether or not something is of safety
significance or it is not, and, based on that, you can
disposition it with reasonable surety that you're doing the
right thing.
Now, the value of that significance determination
process as it applies to the inspection program has been
recognized in many areas and we have significance
determination processes for each of the cornerstone areas in
the oversight process.
We find that fire protection has one now. We've
got one in just about every area where you need to answer
the question of how significant is the finding or the
concern that you're trying to address, and that, I think, is
the item that has helped us understand better how to treat
the various discrepancies that may be uncovered in the
inspection program or identified by the utility, as well.
DR. WALLIS: I thought Dr. Powers had a somewhat
broader question about the effect of a move to change the
regulations in some fundamental way using risk information.
MR. BEEDLE: I'll get to that. I have not
forgotten that.
DR. WALLIS: Thank you.
MR. BEEDLE: But I think this is important that
you understand what we have managed to do in this revised
reactor oversight process; this ability to define the
significant versus the insignificant, the relevant versus
the irrelevant.
DR. SEALE: Is it fair to say then that the real
message in that process is not in the color, but in the
determination of the significance of the event?
MR. BEEDLE: I think the answer is yes.
DR. SEALE: And that's what drives the actions of
the utility and of the inspectors. And as long as the
outside process continues to focus on colors, then you don't
have the adverse consequences of over-sensitivity to the
relationship between the NRC inspection process and the
operators of the plant.
MR. BEEDLE: I think they are very closely linked,
because the construct of the performance indicator is such
that its thresholds are set to tell you how far you are, on
a relative scale, from something that looks to be
unacceptable from a risk point of view.
DR. SEALE: Yes.
MR. BEEDLE: Presumably, the margin to safety is
less if you're in the red area than if you're in the yellow
area than if you're in the white area, and the green area
represents kind of the nominal performance of the industry
and, in fact, the definition says that if you're in the
green area, you are operating within the industry norms.
If you're in the white area, you're starting to
depart from the industry norms. If you're in the yellow
area, you have some reduction in the margin to safety or
margin of safety, and then if you're in the red area, you
have a significant reduction.
DR. SEALE: Is there an event significance
determination for something that allows you to remain in the
green area?
MR. BEEDLE: When we have an event or a finding
through an inspection process, then that event is evaluated
through the significance determination process and a color
is assigned. So you can have lots of events where it's
green.
DR. SEALE: But you're still worried.
MR. BEEDLE: The fact that you're green does not
mean that you don't go address the problem. What happens is
if you have, let's say, a valve fails to operate the way
it's supposed to and you evaluate it and it's green, we
don't just walk away and say the valve broke and let it
remain that way.
DR. SEALE: No.
MR. BEEDLE: We put that into our corrective
action program and we go about the process of correcting
that deficiency, just as though we had discovered that thing
ourselves, which is more than likely the case.
Similarly, if you have an event that would place
you in the white area or in the red area or yellow area,
then what that -- I mean, the utility is going to go ahead
and resolve that problem.
DR. SEALE: Yes.
MR. BEEDLE: But that would then bring with it a
change in the inspection routines that the NRC inspectors
would conduct, and they would go increase their level of
inspection as a result of that.
Now, this whole process is not possible unless you
have an understanding of risk calculation and the ability to
predict and calculate these things. So the risk-informing
of the regulation I think is a necessary follow-on to the
application of these risk insights that we are using in this
oversight process.
DR. APOSTOLAKIS: Isn't it also necessary, though,
for each unit to have a decent PRA?
MR. BEEDLE: Yes, but even with that statement, we
have an oversight process that is built on the results of
the IPEs that were developed back in 1988 or so.
Now, I'm not suggesting that we stop the clock and
just stick with that, and I'll cover that in a minute, but
what I'm saying is we don't have to have a detailed PRA in
order to use risk insights in the operation of these plants.
Now, if I am going to go to a risk-informed
regulatory basis, yes, we need the use of the PRA, a good,
solid, well founded PRA to do that. There is no question
about that.
DR. APOSTOLAKIS: Yes. I'm not talking about a
detailed PRA, but maybe, for most units, something that's a
little better than the IPE, because the IPEs are, after all,
12 years old now.
MR. BEEDLE: Yes.
DR. APOSTOLAKIS: Or not 12, but a number of years
old, and I think the process would be much more efficient if
both the licensee and the NRC inspectors have access to
probabilistic models that people can believe and do these
determinations.
MR. BEEDLE: The NRC, when we put together these
SDP processes, they started out using the information that
they had here in their offices in the form of the IPEs,
which it was our obligation to submit as a result of a
generic letter in the 1988 timeframe.
But they went out and visited the plant and
validated the SDP process with the plant's PRA as it stood
as recent as a year and a half ago.
So they have looked at those PRAs. I mean, it's
not a case of the utility has this PRA and they're not going
to let anybody look at it. In fact, that PRA has been used
by the inspectors and, in fact, the inspectors will go back
to that PRA when they're trying to determine the outcome of
an SDP process.
The question, and this has come about over the
course of the last several months, about quality of PRA and
the comment that the NRC staff doesn't have all these
current PRAs at their disposal, we're trying to examine ways
to develop an improved or increased level of confidence on
the part of the staff in the development of these PRAs.
One of the things has been the peer review process
that we've got ongoing. We expect to have that completed by
the end of 2001. That's kind of coupled with some issues on
ASME code development and so forth.
But there are some very robust PRAs out there.
There are probably a half a dozen plants that probably need
to go spend some more time in their PRA. We know that and
we're working on those plants.
But by and large, those PRAs are pretty solid and
well constructed and are providing good results.
DR. POWERS: So I take it, just an anecdote to embroider
what you're saying, I think, the committee had a chance to
visit Davis-Bessie and I think they were one of those plants
that maybe people thought hadn't espoused risk-informed
regulation as much as some of the others.
When we got there, we found that these guys are
doing an incredible amount with the resources that a single
unit has, which is not an infinite number of resources, but
they were taking maximum advantage of things like owners
groups' efforts and things like that and doing great things
to improve their PRAs to come up to snuff, so much so that
they raised questions, in my mind, whether this SDP process
might not be running into difficulties because of the
quality of computational tools available to the staff.
That if, in fact, the PRA capabilities of the
licensee were vastly superior than cruder tools available to
the NRC staff, you might get conflicts based on differences
in technology.
It's a question that came to our mind during this visit to
the Davis-Bessie plant.
MR. BEEDLE: Well I don't think that that is
really a major issue. I mean, as you get further and
further down the SDP process, I mean, you could get to the
point where you're really trying to get down to the third
decimal place to figure out something.
But by and large, the band in which you're
operating is so huge that you never really get to that
point.
Let's just take one parameter that we measure, and
that's the reactor scram or reactor trip. We've got a
threshold that sits around the four mark to get you into the
white band, but to get in the red, you've got to have about
21 of those things. It's like diesel reliability. We're
shooting for a 99.99 percent reliability, but the truth is
you can operate with about an 80 percent reliability or 75
percent and still not be in red.
And what that ought to tell us is that the margin
of safety, the robustness of the design is tremendous.
We're operating like, you know, a tenth of a second is going
to make the difference between core melt or not, and that's
called science fiction, gentlemen.
DR. SEALE: But, Ralph, aren't we getting -- you
come back to the IPEs and Dana says, yes, but they're 12
years old and so on.
And by inference, you certainly led me to
recognize, and I think I knew it before anyway, that the
guys who are really on top of it are so much beyond the
status or the capability as reflected in their original IPE
that there's really very little relationship between that
IPE and what they have now to evaluate.
They've sharpened their tools. They've gone back
when they had a problem that had to do with this particular
train and represented that train more discreetly, so they
could really look at what the interactions were in it and so
on.
Granted, there are still some utilities who
haven't looked at the fourth page in the original IPE yet,
but most of them have done much, much beyond that.
Aren't we under-representing the capabilities, the
understanding that now exists in these utilities, when we
refer to their database as being the IPEs?
MR. BEEDLE: I agree, absolutely. Absolutely.
DR. SEALE: But, see, there are a lot of people on
the outside who also know that IPEs were pretty rudimentary.
So they say, well, your representation of your PRA is
primitive.
MR. BEEDLE: Well, I think that comes from our
regulatory focus on things. I mean, you can't deal with
regulatory posture unless you have things docketed and on
the record and submitted in triplicate, one of those kind of
things, and that's what we're dealing with.
All those IPEs were submitted. There was no
requirement to submit your PRA.
Now, in fact, a lot of plants, this thing we're
calling IPE, they're actually PRAs.
DR. SEALE: Yes.
MR. BEEDLE: Because when we looked at what it
costs to do a PRA versus what it costs to do the IPE, we
said I'd be foolish to do an IPE and make the static
analysis, let's do a PRA so we can take some advantage of it
and learn something about our plant.
So we developed the PRA and we made an effort to
try and keep the thing up-to-date, and that's what was
submitted to the NRC.
So a lot of these things that we're calling IPEs
are, in fact, PRAs.
But set that aside for a moment. The fact that
we're able to, I think, with a great deal of validity,
address the SDP processes on the basis of that 1988 analysis
ought to tell us something about the robustness of that
analysis technique.
It tells you a lot about how you can analyze the
comparative risk of one sequence versus another, this
component's failure and its effect on the outcome and the
performance of the plant. These are really pretty good
systems that have been developed.
Granted, the details and the level of detail that
you put into the thing, whether or not you're using third
decimal place data for reliability or fourth decimal place,
turns out to be, in my opinion, really on the margin. It
has not had a lot of difference. It hasn't produced a lot
of difference in outcome.
But as a result of this whole process of analysis,
we've seen a lot of improvement in the plant performance.
We've looked at putting in systems that weren't originally
designed that have improved our results of the PRA.
So it's been, I think, an extremely successful
venture on the part of the NRC and the industry.
So all of that said, in order to take advantage of
that process, we need to risk-inform the regulations so that
we are focused on the right things.
I would suggest that spending thousands of dollars
on a pump that is no different than the $150 pump only to be
covered by a piece of paper does not make a lot of sense,
and that is only a small cost, because it then gets layered
with maintenance routines and operational tests and
performance checks and qualifications of people.
We've put all these trappings around it and it
really doesn't buy us anything in terms of reliability. And
the way you gauge that is through risk-informing the
processes that help you categorize and determine the level
of attention needed for those various systems, and that's
precisely what we're trying to do with the risk-informing of
regulation, and we'll talk about that in just a moment.
So in answer to your question, Dr. Wallis, that
really is where the value in this comes and if we can figure
out how to separate the significant from the insignificant
in terms of our treatment, both from a procurement and an
operational and a maintenance point of view, I think the
industry will be much better off.
DR. WALLIS: I was wondering, though, in the
long-term, if you didn't see a bigger kind of benefit from
risk-informing, where actually a lot of the excess margins
were better understood and could perhaps be reduced and so
on, where there was actually a more realistic regulatory
process, and if you didn't see real advantages maybe five or
ten years down the road from that.
MR. BEEDLE: Well, I think that that would maybe
be a follow-on activity for us. For example, if we look at
some of our design criteria, there's a tremendous amount of
margin in there. Now, how do you play this margin in design
versus risk in terms of performance of the system is
something that really would have tremendous benefit in terms
of construction costs, time and so forth.
But our focus at this stage of the game in trying
to risk-inform the regulation is saying do you need to have
regulatory purview over that system, that component, and
that structure, and, if not, then what is the proper
treatment for it, and that's the struggle we're going
through right now in trying to deal with the Option 2
issues.
DR. BONACA: Mr. Beedle, I have a question. If we
pointed out the significance of the corrective action
program, really it's almost like a cornerstone element,
together with the significance determination process.
One question we've been wrestling with is should
there be indicators regarding the performance of the
corrective action program. We realize that the indicators
on the corrective program, or some indicators, would be
almost like some more indicators of culture, to some degree.
If you have, for example, an insignificant event
and you have a repeat, a repeat, and a repeat, it tells you
something about the fact that your corrective actin program
may be not be working right.
I would like to have your thoughts about that.
The reason why I'm asking is that the utilities pay a lot of
attention and they look at a corrective action program, the
way it works, as an element of the health of the
organization and does it have a worth in regulatory space.
MR. BEEDLE: The corrective action program is one
that I think has matured over the last roughly ten years.
INPO has had a significant -- has been significant in
influencing the utilities in how to construct an effective
corrective action program. They follow-up with their
periodic inspections of the plants to look at the execution
of that program.
More importantly, I say more importantly, it's a
device and a mechanism used by plant management to ensure
that problems are resolved. So it's a very significant
program. It's not just something that sits on the side and
collects things that the NRC inspectors find.
It's an active, very useful, necessary mechanism
employed by the utility to improve the performance of their
facilities.
Now, if the inspector finds a problem and it's
judged to be green, and it is a problem and it needs to be
fixed, so it goes in the corrective action program, how many
of those green problems do I have to have to make a white?
That's the question that I think is underlying a
lot of this angst over the corrective action program.
DR. BONACA: No, no. I said they repeat, another
repeat of another insignificant issue. I'm only saying that
you do have an issue, you say it's not very significant.
There may be some intermediate significance.
I mean, I'm not specific about the measurement of
this significance determination process. You put it in the
corrective action program and later on you find that you
have another problem identical to that. The root cause of
the first one hasn't been fixed yet, and so on and so forth.
Now, that's one issue that is significant, they
pay attention to those.
MR. BEEDLE: That's part of the corrective
program, is whether or not the problem you're trying to
address today is one that existed in the past and whether or
not that root cause had been addressed.
DR. BONACA: That's right.
MR. BEEDLE: And if the problem comes up today and
you judge that you obviously have missed the target, you
didn't quite get the root cause, you go fix it again and you
keep working that. There's a little feedback loop in there
that says what have you had in the past.
So that's part of the corrective action program.
DR. BONACA: I'm saying in some cases it may not
work, the feedback, and I'm saying that should there be
something there for which there is an indicator.
MR. BEEDLE: I think that's the whole process that
we're trying to address here, is what should you take
regulatory purview of and what shouldn't you. And I would
suggest that if it's green, it's green, let the utility deal
with it.
You have inspection modules that look at the
corrective action program and if you find that there is some
failure in the corrective action program because they don't
have that feedback mechanism, then you address it through
that process.
DR. BONACA: Okay.
MR. PIETRANGELO: But there's another issue. Your
question, Dr. Bonaca, was specifically on whether the CAP
should have its own EIs. There's the PIs on the equipment
that are in the revised oversight process, but I would also
add that the maintenance rule sets performance criteria on a
lot bigger scope of SSCs than just are in the revised
oversight process and if you hit the performance criteria,
you're required to do root cause analysis, take corrective
action, and set a goal to determine the effectiveness of the
corrective action.
I would argue that just based on the equipment
alone, under the scope of that rule, there's ample
indication to monitor the effectiveness of the corrective
action program and until that's proven that it doesn't work
very well, and I don't think it has yet, I don't think
there's a basis to -- I understand your point very well.
It could, I suppose, have some indication on the
culture of the organization, but I think just on the hard
equipment itself, there's a lot of indication already.
DR. BONACA: The reason why I asked the question
is that clearly there is -- we're going towards an
environment where we rely much more on the instruments of
the utility to repair problems than on the regulatory
intervention.
As we do so, then we have to be confident that
those instruments work and the corrective action program is
a fundamental instrument and that's why we're asking the
question.
I agree with your comment that, in fact, the
maintenance rule provides another window.
MR. PIETRANGELO: And that's in the regulatory
purview. That's not just the licensees looking at that.
MR. BEEDLE: There has been a question raised
probably once a year or so, somebody says what's the
connection between economic performance of a utility and the
safety performance, do they go hand-in-hand.
INPO has attempted to answer that question a
number of times. I think the is yes, there's definitely a
linkage, because the same people, the same management
philosophy that's driving your good economic performance is
also driving your safety performance.
And there is not one utility executive that wants
to have a safety problem at the plant, because a safety
problem means failure on the economic front, and everybody
understands that. No question about it.
So I think that the -- you know, if you look at
the economic performance of these, I'm not suggesting we
have an indicator here, yet --
DR. POWERS: You're among friends. This committee
is dead set opposed to economic indicators.
MR. BEEDLE: The first six months of this year,
the industry produced six percent more electricity than they
did last year for the same six-month period. You don't
achieve -- and we're already at about 90 percent capacity
factor.
I mean, you don't achieve that kind of performance
just on a whim and you don't do that setting aside all your
focus on safety.
So these things really work hand-in-hand and I
think Tony's comment that the performance of the plant will
be reflected in the way the plant material condition is
carried out, and that is going to tell you something about
that corrective action program.
The corrective action program and self-assessments
that have been fostered through the INPO process over the
last ten years have probably done more to improve
performance at the utilities than any other thing.
They've caused the utilities to become more
familiar with their equipment. They look at problems. They
take that problem, correct it, feed it back into their
maintenance systems, and, as a result, you see improved
performance across the board.
I hope that answers the question.
Dana, let me turn to a couple other areas here, if
I may.
DR. POWERS: Please.
MR. BEEDLE: I've got Tony Pietrangelo, who is the
Director of our Licensing Strategic Group.
MR. PIETRANGELO: Not anymore.
DR. POWERS: We hardly know Tony at all.
MR. MARION: Whatever is says on the chart.
MR. BEEDLE: Alex Marion, Lynette Hendricks, and
Doug Walters. Tony is going to talk a little bit about the
risk-informing regulation and some of the interactions we've
had with the ACRS and the staff and give you some insight
into that.
Alex is going to discuss a little bit about the
oversight process. Lynette is going to touch on
decommissioning activities, and I know that was an issue
this morning when you were talking with the Commissioners,
and then Doug will talk a little bit about the license
renewal.
But before they do that, let me just make kind of
an observation on ACRS. ACRS is, I think, a relatively
unique element in the regulatory process and I talk about
the regulatory process as opposed to advisors to the NRC
Commissioners.
You are an integral part of the regulatory
process. You're important to the Commissioners and I think
they underscored that with some of their comments this
morning. And you're important to the industry.
Your judgments, or lack thereof, your definitive
recommendations or ambiguous suggestions create problems for
the industry and we have to address them, deal with them in
some fashion, because they have an effect on the staff and,
consequently, an effect on us.
DR. APOSTOLAKIS: They never solve any problems?
MR. BEEDLE: They do solve some problems.
DR. APOSTOLAKIS: Some problems.
MR. BEEDLE: I'm saying they're on both sides of
that fence.
DR. APOSTOLAKIS: Now you say it.
MR. BEEDLE: The direct ones, some of the not so
direct ones.
DR. APOSTOLAKIS: Fine.
MR. BEEDLE: And my comment, I think, would have
to be considered to be somewhat in synch with the feedback
that you got from the Commissioners today, and say, hey,
what do you expect me to do with that recommendation. I
can't do anything with that recommendation. What do you
want me to do?
So that's part of, I think, your challenge, to
deliberate on issues that you think are important, that the
Commissioners need to hear about, and give them some good,
solid recommendations. I don't think this half of the ACRS
can give them recommendation A and this half can give B and
you send that to the Commissioners and ask them to
auctioneer. That doesn't help them.
DR. WALLIS: I think we can't go too far. We can
point out things, but we can't do everything. Resolving a
problem sometimes involves capabilities from elsewhere, not
just the ACRS.
MR. BEEDLE: Well, I guess I would take a little
bit different view of that. I don't think that it's
sufficient to just point out there may be a problem there or
I'm uncomfortable with that result or I think that's a
little bit too high or that's a little bit too low.
That doesn't help them. If you feel that there is
an issue in which the staff has not come up with the right
answer, you have this gut feel, your intuition tells you
that there is a problem with the fidelity of the process,
then you need to go ferret out that problem.
You can't just walk in and tell the Commissioners
I've got a stomach ache. It doesn't help. They don't know
what to do with that.
I would submit that you do that on almost every
issue. Somebody could have some angst over just about
everything that goes on in this industry, but I don't think
that's particularly helpful to the Commissioners and it
certainly doesn't help me.
DR. WALLIS: But somebody else is working these
problems, as well as us.
MR. BEEDLE: That ought to be part of your answer
then. Anyway, I just want you to know that you are an
important element in this regulatory process and you're
important to the Commission, obviously, and you're important
to the industry.
We end up frequently up here making comments and
observations, we the industry, and I think that ought to be
testimony to the value that we put in it. We would
encourage you to continue to look at lots of issues, but
focus on providing good, sound advice to the Commission that
is not ambiguous.
So with that, let me turn to Tony and ask him to
talk a little bit about risk-informing regulation and some
of the things that he's been involved with, when he was the
licensing manager.
DR. APOSTOLAKIS: What are you now, Tony?
MR. PIETRANGELO: I'm the -- go ahead, Ralph, my
title.
DR. POWERS: Risk and Performance-Based Regulation
Director.
DR. APOSTOLAKIS: So this is up-to-date.
MR. PIETRANGELO: Yes, that's the new one. I
actually had responsibility for our risk activity several
years ago and had a lot of interaction with this committee,
and then we were kind of to the point we are now, I think,
with a lot of the activities and then we went through what I
term our blue period, from 1996 through 1999, where I think
some of the things Ralph talked about before, trying to get
stability back into some key regulatory processes was of
greater importance than trying to reform the regulations
with risk insights.
So I tried to get out and they pulled me back in,
George, to risk-informed regulation. So here we are.
DR. APOSTOLAKIS: They did that very well.
MR. PIETRANGELO: Let me start just by saying that
there's already a lot of industry interest in risk-informed
regulation, and it's present today in the regulatory
process. Just to name a few items, the monitoring, the
level of monitoring done under the maintenance rule is
risk-informed. The new A-4 configuration risk management
process is certainly risk-informed.
The reactor oversight process is risk-informed in
terms of the significance determination process, as well as
the areas that the inspection process focus on. Sixty
plants are getting ready to submit risk-informed in-service
inspection requests and almost all plants have gotten some
kind of allowed outage time extension through risk-informed
tech specs.
So the fact is that risk insights are already
being used in the regulatory process to a great extent
today, even though, and I think this was mentioned before,
but maybe not this way, there is no requirement for PRA in
the regulation.
All we have is that IPE generic letter, which was
just do you guys want to do one and if you do, can you send
it to us. And you're right, we've come a long way since
then and I think as part of the effort going forward, and
that's kind of the where are we now part, I think, as an
industry, we recognize that that old information that was
docketed as part of the IPEs is insufficient as a foundation
for moving forward with risk-informed regulation.
So both on a macroscopic level, from the fleet of
plants, risk information, from the risk information that's
submitted on a plant-specific level, we've got to put
something on the docket that's available to the public,
that's transparent, so that to make some of the changes
we're talking about in risk-informed regulation, that has to
become part of the licensing basis and has to be very
visible.
We're analyzing several alternatives now on how to
do that best and I think you'll be hearing from us in the
near future on that.
DR. SEALE: Tony, could I ask a question? This
morning, you were there, I believe.
MR. PIETRANGELO: I was not there, Bob.
DR. APOSTOLAKIS: Ralph as.
MR. PIETRANGELO: Ralph was there.
DR. SEALE: One of the comments that was made was
that in several contexts, we find that the process of
docketing something has a tendency to ossify what's done.
So when you say you want to put the PRA in the docket, yes
and no, in the sense that we sure would like to have the PRA
on the docket, because that's something that everybody then
has some access to.
And so the unfounded impugning of the level of
consideration of risk issues would be hopefully diminished
if someone was willing to look at the PRA that would be
available. But at the same time, we would hate to think
that if you put the PRA on the document, that you would then
be reluctant to upgrade it at appropriate intervals so that
it continued to be a living document.
So somehow -- and I don't think there is a
provision for this right now.
MR. PIETRANGELO: There is not.
DR. SEALE: But somehow, between yourselves, and
clearly you're a party to it, the Commissioners and, I
guess, this committee is an honest broker, hopefully honest
anyway, we have to come up with an idea of how we docket
something that also satisfies the idea of being a living
document.
MR. PIETRANGELO: And that's one of the things. I
don't think we said we're going to put the PRA on the
docket, but in the same vein, though, I think some of the
things we're thinking of, for example, even with the ALWRs,
there was a summary of the PRAs put in the FSAR.
The FSAR is updated on a regular basis. You could
have the main contributors. You could have some of the
metrics that are associated with the PRA. It could be
updated on an ongoing basis.
I think as part of the application, if you're
going to be a player in risk-informing the special treatment
requirements in Part 50, you're going to have to put
something on the docket and have to keep that updated over
time.
There's already mechanisms in the current
licensing basis to do that, to put it in your CLB, whether
it's the FSAR or a commitment or somewhere else in the
licensing basis, and they already have mechanisms to update
and even change control processes for those kind of things.
DR. SEALE: Okay.
MR. PIETRANGELO: And I think that's -- because we
really haven't had much discussion on that yet, that may
have been driving some of the issues that have come up in
risk-informing Part 50, particularly in Option 2. We've had
a lot of talk about the PRA standards and the
prescriptiveness of it, the detail of the Appendix T for
50.69.
And because we don't have something that the
regulator has direct oversight over in the licensing basis,
kind of using these longer levers to try to get control over
the PRA quality, and maybe that's not the best way to do it.
So the other thing I think that I want to talk to
you all about, and I think your letters and deliberations
have been helpful already, is that we need to understand the
process of how this is going to work to be able to put the
individual pieces in, like the PRA standard and like the
peer review process, and, yes, the prior staff review, prior
to the application.
I think in some of the documents you've seen, the
objective for the detailed Appendix T and the detailed
standard being no prior NRC staff review and I think that's
probably a false premise for the kinds of applications we're
talking about in risk-informing Part 50, Option 2 and 3.
There is going to be a need for prior staff
review. Reg Guide 1.174 is about the staff's review of
licensee submittals to change their licensing basis with
risk insights.
I think the process we envision is that some of
these pieces, like the standard and the peer review process,
can help facilitate the staff's review. We can even get it
down to a template such that the thing doesn't take the time
that South Texas is taking right now to process their
exemption request and that those could be done in a very
efficient way.
The risk-informed ISI submittals are down to
templates. This committee has looked at a lot at
risk-informed ISI over the years. One of the specific reg
guides is on ISI and IST, and we've got the submittal down
to a template. It's not the FSAR-like volumes of work that
were done to support ISI, but it lifts the pertinent
information out that the staff needs to conduct its review,
and I think we can get to that same point on risk-informing
Part 50, Option 2.
DR. APOSTOLAKIS: I think we're running out of
time real quick. So I h tow questions that I would like to
ask you.
We hear -- in fact, the Commission had asked this
morning, but we heard it, some in the industry think that
perhaps we're going too far in trying to risk-inform Part 50
and you hear things about low-hanging fruit and so on.
MR. PIETRANGELO: There is no -- we'll have to get
that out of the lexicon. There is no low hanging fruit in
the regulatory process.
DR. APOSTOLAKIS: But do you -- is NEI's sense
that we are going too fast?
MR. PIETRANGELO: No.
DR. APOSTOLAKIS: Okay. Good. Second question.
I think sometimes one learns from one's not so successful
efforts and -- how to put it in a more straightforward way
-- do you think that any of the views that the ACRS has
expressed in the last several years have, in fact, not been
helpful or they missed the mark or whatever?
MR. PIETRANGELO: No. I think the views the ACRS
expressed, that have been expressed by the ACRS have been
pretty good and on point. I just wish it wouldn't take you
so long to do it.
I mean, it took almost four years to get that
series of reg guides out.
DR. APOSTOLAKIS: So you're not just blaming the
ACRS for that.
MR. PIETRANGELO: No. But I think this is to
underscore Ralph's point. You're not just deliberating here
off-line. You are part of the process and I think as an
industry, our sense of time needs to change a little bit.
Some of the generic safety issues that the staff and the
ACRS reviews have been in that log for a long, long time and
you start to ask yourself, well, how important can that
thing be if it's been there for 15 years.
So I think in the industry, with deregulation and
other things, that our timeline has to change a little bit
in terms of how long it takes to do some of these things,
because you can't sustain that level of resources to focus
on that problem for that many years.
We've got to have something that's either
step-wise growing up to the end goal or decide it's not
worth it and get out of it, because it's a lot of resources
for us to support your meetings. I know it's a lot of
resources on the staff to support your meetings.
I wanted to compliment you on waiving your review
of the final reg guide on 50.59. I think you recognized
that you done a good job on the preliminary reg guide, there
weren't that many changes on it, and there was no -- was
going to be no value added by another ACRS review.
I think you should look at that more often in the
future. And on the risk-informed stuff, your insights are
vital to the process and for public confidence in the
process.
But I think what I'm trying to get at is we can
find a way to get them in or at least think about the times
that you really want to weigh in. Do you want to weigh in
on the front end or do you want to let the staff kind of
bake something and interact with us and then bring you
something or when?
DR. APOSTOLAKIS: What do you think of that?
Because that has been a criticism of the committee from some
quarters, that if we get involved from the beginning, then
we are not independent.
MR. PIETRANGELO: That's right.
DR. APOSTOLAKIS: On the other hand, if you look
at the history of this committee, up until maybe ten or 15
years ago, it was very common that the staff would invest
two or three years into a project, then appear before this
committee, but the committee would disagree and you would
have all sorts of acrimonious debates.
It's unfair to the staff because they have already
invested a lot of time, but it's unfair to the committee,
too, to say, at the end, well, gee, you know, we've invested
so much time, you better go along.
On the other hand, you are accused that you are
not independent.
MR. PIETRANGELO: Sure.
DR. APOSTOLAKIS: It seems to me you can be both,
but I would like to know what you think.
MR. PIETRANGELO: And I think it depends on the
situation, of course. There's a balance between -- you're
not line management for the NRC.
DR. APOSTOLAKIS: Yes, we are aware of that.
MR. PIETRANGELO: I don't know how many meetings
I've walked into where the first slide was ACRS concerns.
Well, you know, the thing is just starting out and I think
it's something to think about with each of the issues the
committee deals with, about when you interact, is it the
right time, and not to just assume some rote schedule for
that interaction.
I think I've already seen some things about
backing off some of the process type issues and focus in
more on the safety standards type issues. I think that
makes perfect sense for this committee.
But overall, and there's no right or wrong answer
to your question, George, I think it's got to be a balance
and it's got to be a consideration in the back of the
committee's mind of when you weigh in and is it the right
time in the process.
DR. APOSTOLAKIS: And sometimes, of course, that's
beyond our control.
MR. PIETRANGELO: That's right.
DR. APOSTOLAKIS: When we get an SRM, we have to
respond.
MR. PIETRANGELO: Absolutely. Absolutely.
DR. POWERS: One of the motivations for this, we
were so anxious to have this meeting and, of course, we do
do a self-assessment of ourselves.
MR. PIETRANGELO: Right.
DR. POWERS: And, in fact, one of the assessments
that we did, we looked at the number of issues that we were
trying to tackle and said it's too many and set up some
criteria for when to, as you say, not weigh in, because it's
outside of our expertise or we really didn't have anything
to contribute.
We do keep metrics on ourselves and I'm proud to
say that whereas our metrics are not as good as the -- the
performance metrics are not as good as the nuclear
industry's, we are on an effector.
MR. BEEDLE: I hope most of them are green.
DR. POWERS: Since we get to set the color bands,
we're white in two and green in the rest.
MR. PIETRANGELO: Doug is going to talk about
license renewal in a second, but I know this committee is
thinking about license renewal and not too long, probably
every month, you're going to be deliberating on someone's
application for license renewal, and that's a statutory
requirement for this committee to do that and you are part
of the process, make no mistake about that.
And it's worked well thus far. We know we can
process two at a time or three at a time now and in a couple
of years, that's going to be eight to ten. Again, if you
don't get on top of this when you interact in the process
thing, when you're having that kind of workload coming down
the stream, I think you're going to put yourself in a
critical path position on some of these. So it's something
to think about.
MR. BEEDLE: Dana, we have until three on your
agenda, is that correct?
DR. POWERS: Sure.
MR. BEEDLE: I just wanted to try and make sure we
didn't run over.
DR. POWERS: We got a little slop.
MR. BEEDLE: Why don't we touch on license
renewal, Doug?
MR. WALTERS: Good afternoon. I'm Doug Walters.
I have responsibility for renewal. As Tony said -- well,
let me start off by, I guess, saying that we've been fairly
successful, as you know.
I think you've had the opportunity to review at
least two applications. We have three that are under review
now and by our own assessment, there's somewhere on the
order of 30 more units that will come in with applications
between now and about 2003-2004.
So the interest in renewal is swelling and as Tony
said, you do play an important role in that process.
I know that you're going to talk about the GALL
and the standard review plan and the other guidance
documents, and that's really our focus in the industry right
now. We're reviewing those documents.
With the exception of the applications that have
been submitted, I think these documents represent probably
the most significant thing that's come out of renewal in a
while, and these are extremely important documents for the
follow-on applicants because they intend to rely on them.
And I would point out that in the review of
Calvert Cliffs and Oconee, the number is somewhere between
85 and 90 percent of all the programs that were credited
were existing programs.
So what we really need to focus on is that
remaining, whatever the number is, ten to 15 percent and
make sure that we understand what enhancements are needed or
what new programs might be necessary.
To pick up on a point that Ralph made about ACRS
and the role you have in that process and the impact that
you have, I just thought of two examples that I will share
with you.
One is the environmental effects of fatigue, which
has probably been debated to death and nobody wants to
regurgitate, and I understand that. I feel that way myself.
But I would just point out to you that it's not
that we're whining about where that issue came out. I think
what we would like to see, though, is perhaps more
consistency.
For example, if you read the staff's letter, they
indicate that this is not an issue for ALWRs. Yet, they use
the same, at least as I understand it, ALWRs use the same
methods for evaluating fatigue as we did on the operating
plants.
And somehow, on the operating plants, for renewal,
we come out with a different result. And when we know an
issue like that is coming to an independent technical body,
we're anxious to see how you look at that and is there
consistency across the issue.
And I'm not so sure, in this case, and I'm not an
expert in the fatigue area, but based on what I read and
what I do know of the issue, I'm perplexed by that.
The impact of that, though, is we now, as an
industry, have to go out and figure out a way to deal with
environmental effects only from year 40 to 60 and we go to
the ASME code, they haven't changed the curves, I think
something is underway perhaps, but it presents a real
problem for us, because it's up to the applicant, the
renewal applicant to deal with that issue and we're not
quite sure how we do that.
We'll get there. It's not to say it's not a real
phenomenon, but when a decision is rendered that, yes, this
is something you need to address in renewal, we need to do
that.
Another area is rule changes, when there is a
change to a regulation, and I will mention 50.55(a) and the
adoption if IWE and IWL, and there is an example where we
have an opportunity to look at that rule change and say,
okay, if we adopt this change to IWE and IWL, how does that
impact renewal.
I think in this case, the staff concluded that
what they were doing in the rule change would cover the
license renewal period.
Now, the renewal staff came back and published a
reg guide and disagreed with that. So, again, we're in a
predicament that we've got a regulation that says IWE, IWL,
and, specifically, the focus is in accessible areas, and we
have a regulation that says that should be okay for renewal,
but we have also a reg guide or a NUREG that says, no, we
don't agree with that.
We need somebody besides ourselves to kind of look
over the broad spectrum and say is there consistency there,
and I think these are two examples, not necessarily -- it's
being candid with you. I'm not trying to be critical with
you.
But we need that kind of broad overview to say are
we being consistent technically. The process issues we can
deal with.
So as we go forward, we hope that that's the kind
of keen eye you will give to the GALL, principally the GALL.
I think that's where all the technical meat is. And we need
that document to be technically correct so that when these
30 or so renewal applicants come in over the next couple of
years, we know we can rely on that document and there will
be some success there.
That's all I have, unless you have any questions.
MR. BEEDLE: Okay.
DR. APOSTOLAKIS: Just out of curiosity, if we
have 30 units in the next four years, would risk-informing
the process help a little bit or you don't want to touch
anything? The process is working.
MR. WALTERS: I'm going to defer to the end of the
table for the specifics. I will tell you that this is
another consistency issue, in my view.
If I may put the risk-informing aside for a moment, it's
curious to me that the maintenance rule, for example, has a
scoping criterion safety-related and that scoping criterion
is the same in license renewal for safety-related. Yet, we
haven't figured out a way of how we can give credit for what
we do under the maintenance rule.
If we go to risk-informed scoping, it seems to me
that you ought to be able to apply that to license renewal.
That would be, it seems me, the prudent thing to do. That
makes it consistent.
But as Tony can explain to you, the way that
Option 2 is working, if you do renewal, you can't do Option
2. I don't know if you want to expound on that.
MR. PIETRANGELO: Yes. There is a perverse
consequence of a staff position on not applying the
risk-informing Part 50, Option 2 to Part 54.
For the near-term applicants in the queue who are
already in the process of preparing their applications, it
has absolutely no effect, because this rulemaking is going
to take a couple of years. It's not even slated to go out
for public comment until late next year.
So it will not have an impact on the near-term
applicants. Rather, if one of those near-term applicants
wanted to also take advantage of Option 2, you would -- the
way the process is going to work is the safety-related, low
safety-significance box, the RISC-3 box, there would be some
kind of minimum requirement in lieu of the special treatment
requirements for those components.
In the license renewal process, there is a
demonstration that the effects of aging are being managed
effectively, and a lot of those come from the special
treatment requirements.
So you're asking -- unless the staff wants to
stipulate up front that whatever that minimal requirement is
would be adequate for the demonstration, you're going to
undo a lot of the things you did in risk-informing Part 50,
Option 2, and it's going to be a disincentive to moving
forward.
So we think that Part 54 should be included in the
special treatment requirements. We think it will not have
an impact on the applicants that are in the queue. It will
make the process more efficient and coherent across Part 50
and Part 54 for the remaining 60 to 70 plants that are going
to come through the license renewal process in the future.
I just think it's just not consistent. This was
Douglas' point that says we can us risk insights to focus
our resources on the right things in Part 50, but we don't
want to do that for Part 54. To me, that's not consistency
and coherency in the regulatory process.
MR. BEEDLE: I think it points out some of the
difficulties that the staff has and the industry, as well,
in trying to figure exactly how you take this risk-informing
and lay it across all the regulations.
Short of just taking a clean sheet of paper and
starting to develop these regulations, I think you're going
to continue to run into these sort of problems. But we
would hope that there would be more thought given to the
statements that you find in SECYs, for example, because
those things carry a lot of weight. They carry an awful lot
of weight.
That's, in part, the challenge that you gentlemen
have in trying to provide advice and counsel to the
Commissioners. You make a statement and the staff uses that
and they act on it. They think that that's --
DR. POWERS: We are well aware the Commission
today asked us both questions, are we going too fast and
should we be working on 50.54.
I think, in sympathy to lots of people suffering
from future shock, they have to be a little patient here.
I'm sure there's going to be some dislocations.
But you're talking to a committee that, if they
had their druthers, would have taken out a clean sheet and
written the whole thing in risk language and moving 54 in
there would have -- I mean, when that decision not to have
license renewal be risk-informed was made, I think we
understood the practicalities and the time pressures at the
time, but I think we were disappointed that that didn't
happen.
MR. PIETRANGELO: I think the concern is upsetting
the apple cart and I think that's a false --
DR. POWERS: And I think that's a good point,
Tony. I think I'd like to have you come down here and make
that point to us a couple more times.
MR. PIETRANGELO: Ralph made this point to the
Commission in a briefing last week.
DR. APOSTOLAKIS: It takes us three times.
DR. POWERS: It is a problem which we have right
now, which is an inconsistency between one end of the table
and the other, and we've been whistling along saying, well,
we'll evolve toward it, and I think what you're telling us
is let's help Mother Nature evolve a little faster, let's do
some genetic engineering here.
MR. WALTERS: Well said.
MR. BEEDLE: If we could, let's turn to -- do you
have anything else, Doug?
MR. WALTERS: No, I don't.
MR. BEEDLE: Turn to Lynette and a little bit of
discussion in decommissioning activities.
MS. HENDRICKS: I'd like to take probably what
little time I have and talk about risk-informing the
decommissioning regulations. There were some other things
on your agenda about cask process and decommissioning
process, but if I had my druthers, I guess I'd like to say a
few things first about the spent fuel pool risk study and
how that's being used to risk-inform the decommissioning
regulations.
I did not, unfortunately, have the opportunity to
attend your briefing, but I do have the benefit of your
slides and I concur with the position you're taking here,
that when you start to risk-inform this special event over
here with the pool, you're in different space, and I see
that in your recommendations that the LERF may not be as
appropriate as it had initially appeared to be when you
start to look in more detail at the consequences.
I think the ACRS has been very helpful and very
astute at trying to direct the staff's attention to refining
the consequences, but I guess I would hope that since we are
in different space, we would give the same emphasis to
refining the probability.
What I mean here is that at this point, we are --
the whole event, as you know, is driven by the single,
somewhat bounding seismic event, and that's different than
what we have in operating space.
In fact, the same event is a background risk
factor and I think we're going to have some unintended
consequences if we are not equally careful in balancing the
consequence refinement with the probability refinement.
I'll give you an example of that. It's in the
SECY-00-145 that's before the Commission today, and
understanding that that's going to change, but I guess what
troubles us with that is it seems like we've missed some
opportunities to apply real risk insights on a practical way
by basing all of our concern on this single bounding seismic
event, for which there are no mitigating factors.
I can say that more in pragmatic terms. What you
end up with is a rulemaking that essentially layers
deterministic requirements; in other words, you don't get
out of your EP and you don't get out of financial protection
until you can demonstrate the event cannot happen.
At the same time, the rulemaking plan proposes to
layer these new requirements that we worked constructively
with the staff to come up with to address all the other
sequences.
We propose some real practical solutions to
identifying and mitigating and responding to these slow
leaks and loss of cooling and we also put forward a seismic
checklist that we were hoping would help screen out this
event, which, actually, if you look closely at the implied
requirements, really you're talking about a big imposition
in a regulatory sense.
What it ends up with is you have a voluntary rule
that industry is not likely to take advantage of at all. So
you're going to have a status quo. All this rulemaking, all
these risk insights for nothing, because the industry will
have to look at that rule and say, jeez, you know, I've
operated with this out here seismic risk and I have
financial protection, I have EP, I shut down, I have one
thing to worry about, just the pool, and I have this
background event that's now predominant and I have to, in
order to still have EP and financial protection, incur all
the costs of these additional requirements.
It's kind of an untoward outcome in risk space.
So to sum up, we certainly appreciate your --
DR. APOSTOLAKIS: Is there a good risk analysis
that the industry has done on this issue that I can look at?
MS. HENDRICKS: If you will recall, we
participated early on. We had ERIN Engineering provide
input to when we were predominantly looking at these events.
Remember, the staff initially said this risk was
equivalent to an operating plant risk because there was very
little credit for the long duration of the events and
they've since refined their HRA analysis and now we're
working very diligently on the seismic risk and we had hoped
to sort of put that to bed.
DR. APOSTOLAKIS: If you do that, you deserve a
statute. A seismic issue? In this context.
DR. POWERS: And one of the points we tried to
raise with the Commission was that this bounding seismic
analysis was strictly inconsistent with the whole philosophy
we were pursuing everywhere else and it was expedient, but
it was not useful. So I think we're sympathetic in that
area.
I think we're looking for a much more stately
approach to this than maybe is what you're looking for. We
don't get the phenomenological things squared away and then
proceed into the risk analysis and going, it seems to us,
completely to risk, rather than a damaged frequency or
something like that.
MS. HENDRICKS: Thank you.
DR. KRESS: We recognize from the start that the
bounding seismic analysis that was used, the CFCLP concept,
was very conservative and we knew if you were going to get
any relief at all, that that's the end you would have to
work on. I'm glad to hear that somebody is addressing the
real risk, not just bounding it, because actually there's
two parts, there's the consequences and there's the
frequency and you've got to have both of them.
And we really had no way to focus on the
frequency, other than in our evaluation, all we could look
at is consequences. Now, somebody has to look at the
frequency and balance the two and I think that's what you're
trying to say.
MS. HENDRICKS: We are and I think, to a certain
extent, it poses a challenge because we're sort of changing
horses, too. For operating plants, we use the sort of a
more deterministic approach to seismic, with the safe
shutdown earthquake and you look at the tectonics and say
this is worst case, with some dealing with the uncertainty,
whereas now we're deciding, well, let's embrace probability.
So we pick up either the Livermore report or the
EPRI report. When you try to look at the whole scheme,
you're always going to have an over-focus on these very
large, very low probability events.
So I know this is coming back to you and I know
the staff is looking at it, but if we would think even
beyond, well, kind of what we've gotten from the staff is,
well, you wanted probability and here's the number you get,
but that's not looked at in the consequence of one event
driving the whole sequence. We may come out with funny
screening numbers.
DR. KRESS: I think what you're saying there is if
you have just one event driving your whole risk, that your
risk acceptance criteria might be viewed differently --
MS. HENDRICKS: Exactly.
DR. KRESS: -- than if you had a whole lot of
things, and I think there's some truth to that.
DR. POWERS: Not only that. I think you run into
the challenge of interfacing defense-in-depth concepts.
DR. KRESS: That's an additional depth to this
thing.
DR. POWERS: You've got to think very carefully
about how you're going to be consistent with the
defense-in-depth philosophy when you've got one dominant
initiator.
DR. KRESS: Absolutely.
DR. POWERS: It's a strange thing.
MR. BEEDLE: Well, it's not an issue that's going
to be resolved this afternoon, which I judged from the
Commissioners' comments this morning.
Let me turn to Alex and ask him to provide a few
comments on the revised reactor oversight process.
MR. MARION: Thank you, Ralph. My name is Alex
Marion. All I can tell you at this point is I think I work
for Ralph.
DR. POWERS: It says licensing and programs.
DR. SEALE: He thinks so, too, I take it.
MR. MARION: Just very briefly. I know we're
pressed for time. Let me just indicate that overall, we
think the process has been very successful in that we have
been working very closely with the NRC staff and other
interested stakeholders to make sure that the entire
community of interest understands what the indicator system
is all about, understands what the thresholds for regulatory
action that Ralph alluded to earlier in terms of inspection
is all about, and that entire process is scrutable.
As you all know, the first year of implementation
is getting close to being completed. They're into the third
quarter. The second quarter data has been already provided.
Third quarter data is going to be submitted to the NRC, I
believe, the 21st of October.
It doesn't indicate that everyone is in the green
area. It's starting to identify a mix and I think as we go
through that process, we'll be able to evaluate the
indicators and the thresholds and the action and draw some
conclusion as to whether or not it's doing what we expected
it to do in terms of the overall scheme of things.
The NRC has just recently announced the formation
of a FACA panel, similar to the pilot plant evaluation
panel, that will start the effort to begin the lessons
learned review of the process during this first phase, this
first year of effort.
That panel is scheduled to have the first meeting,
I believe, early November, November 1st and 2nd. Again,
that body will, as necessary, include various stakeholders
to provide the overall assessment of the process in terms of
internal NRC benefit, understanding scrutability to external
stakeholders, as well. So we find that to be very positive.
There is one performance indicator that's being
changed. It begins with the letter S, dealing with counting
of automatic and manual scrams.
We have reached an understanding with the NRC on
alternate indicator and we're going through the
implementation phase of benchmarking or testing that
indicator through a pilot process. We have 21 units
involved and that process began this month and will continue
through March of next year.
And fundamentally, if it's found to be acceptable
by NRC, then we expect it to be implemented third quarter of
next year. We find that to be positive.
So it is a dynamic process. It seems to be
working well. We're, quite frankly, pleased. We believe
that the changes that have been identified and agreed to are
necessary changes and there will likely be changes in the
future, and we're just pleased to see everyone working on it
and applying their best efforts and making sure that it's
accomplishing its intended objectives.
DR. POWERS: One of the issues that we have
debated around this table, to no successful resolution,
except an accommodation, was that the cross-cutting issues
that include things like human performance, will they, in
fact, be revealed adequately early by the other performance
indicators.
I think the accommodation the committee came to
within itself over that was to say this is an untested
assumption and it needs to be considered as we get into this
assessment and whatnot.
Is that one of those things that you're paying
attention to at all?
MR. MARION: Yes. It is being considered. I wish
I could provide you a definitive answer at this point in
time. But I suspect as we go through this cycle, we will
probably have a more confident feeling on how that
particular element is being considered in the overall
program. But it's something that everyone is aware of.
DR. POWERS: Yes, that's good. The other area
that the committee has not spent any time on, because of a
variety of things, is the significance determination
process. We understand what it is, but we have not gone
through the details of it, especially for those things where
there is not a PRA capability available, things like the
significance determination process for fire protection
issues and things like that.
MR. BEEDLE: Well, there are a number of them;
fire protection, security.
DR. POWERS: Security.
MR. BEEDLE: And if you'll recall, security, EP,
rad protection, are not elements that are amenable to the
PRA focus.
DR. POWERS: The current PRAs can't do that.
MR. BEEDLE: But they were nonetheless important
elements in this overall assessment process for the facility
and we aged with the inclusion of those pieces as
cornerstones and the necessary outcome was to have some sort
of a significance determination process.
We have, just for an example, an SDP in the rad
protection area, talks about if you have a planned
maintenance evolution with a target exposure and you are
approaching that and you go over that exposure calculation,
then that puts you in white. That makes that a white
finding.
Well, safety significance is zero, but we agreed
that that is something that is important and we understand
the ground rules. Again, if I know what the target is, I
can figure out how to deal with it.
Well, there is a provision to go in and change
that target and then you get into the philosophical argument
of whether or not it's best to change your target or not
change your target, what's the effect on people and how do
they deal with that whole issue of management control.
So we've got that behind us. Now, that's a good
example of where a very deterministic pre-determined course
of action would be taken, given a certain set of
circumstances. We understand that, and there are a lot of
areas like that.
DR. POWERS: I guess what we were concerned most
about is it seems like some of those non-PRA things, the
thresholds are a good deal more stringent than the
corresponding ones that are treated with a probabilistic
risk assessment, and we worried are we losing the advantage
of focus by putting in very stringent thresholds on some
cornerstones and not so stringent on the others.
MR. BEEDLE: I think that's a very valid question
to be asking and I guess maybe the corollary to that is
what's the effect on does it divert management attention and
are you really looking at the right thing and so forth.
So I think we'll have to continue to look at that.
We'll get our FACA panels to look at that.
DR. POWERS: I think it would be interesting to pursue that,
because I think there's -- my own feeling is there is
discontinuity when you go away from those that are
subjective to quantitative assessment to those that are more
qualitatively or traditionally assessed.
MR. BEEDLE: I would suggest that the underlying
principle involved here is don't let go of anything. So
we've got to have that cornerstone. I can't let go. I know
it's not safety-significant, it's trivial in the grand
scheme of safety, but I don't want to let go of it.
I've always managed that, I've always had to look
at it, I've always had some inspection activity there. So
the staff is looking for some element.
Now, the other part of it is public perception.
It doesn't serve the industry or the NRC well to not have
something in there that speaks to protection of the public
to radiation exposure limits, along with emergency planning.
DR. POWERS: I think I don't want to provoke
concerns by the public. But I think when we set thresholds,
that those are a little more arcane than whether you have a
cornerstone or not.
MR. BEEDLE: That's true. Well, back to your
question on the cross-cutting. The cross-cutting issues are
important to the industry, no question about that. They're
important to the NRC.
I think that over the course of the last year, the
cross-cutting issues of safety-conscious work environment,
corrective action program, human performance have taken on
increased significance.
There has been more discussion of those issues and
those programs and focus by the NRC staff on those issues
today than there was a year ago, two years ago, three years
ago.
And I do believe that they are self-revealing in
the performance of the facilities.
DR. POWERS: Commissioner Diaz argues that the
corrective action program is the unspoken third element of
the new risk-informed regulatory process. Three attaches a
great deal of significance to that cross-cutting issue.
MR. BEEDLE: Yes. I don't have anything else to
add.
DR. POWERS: Well, this has been truthfully
excellent, in my point of view. I especially like it when
Tony can come in and tell me things I'm supposed to be
doing, and that's good, because I need to be reminded often
and I'm glad to have him do it, because he does it in a very
gentlemanly way.
And I appreciate all the presentations and --
MR. BEEDLE: And he's not even Irish.
MR. SIEBER: That helps.
DR. POWERS: I appreciate all your program
directors and their comments and whatnot and I hope that
they will feel free to come to us and say, hey, these are
the things we need you to pay attention to and give us the
high sign occasionally, because we are very much in the
situation of having a lot more issues come to us than we can
possibly handle and we have to do some picking and choosing.
And I kid you not, we do track our own performance
metrics on things that we are declining to review and things
that we do accept to review, and we'll take all the help we
can in making that selection, because it's sometimes hard to
do.
So I appreciate all the comments you made and
taking time to come down here and talk to us. It's been
very, very useful, for me at least.
MR. BEEDLE: Thank you very much, Mr. Chairman.
DR. POWERS: With that, I will recess us for 15
minutes.
[Recess.]
DR. POWERS: Let's come back into session. We are
now going to turn to the subject of GSI-168 and equipment
qualification. Professor Uhrig, I think you're going to
provide the leadership here.
DR. UHRIG: Today we're talking about the
safety-related electrical cables, primarily low voltage
associated with transmission of signals from plant to I&C
devices and instruments, as well as medium voltage that
conducts power to safety-related devices.
The business of environmental qualification
started about 1971 or thereabouts. There are three
categories of plants today; one, those that started
construction before '71; those that started construction
after '71, they had the IEEE Standard 323, the 1971 version
involved.
The trial standard did not address aging and
required dynamic -- required a systematic program of
analysis, testing and quality assurance, whereas those that
were before '71 were held to high industrial quality as the
primary standard. And after the plants with CPs after
7/1/74, there was a '74 version of the standard.
So that's a little of the background. The 1982
equipment qualifications rulemaking process led to the 10
CFR 50.49. The Commission did not backfit this to the older
plants and today, we have, again, three separate categories
of plants, the largest being a group that were
grandfathered, some 70-odd, if I understand correctly, to
the original standards.
So we'll hear more about this and with that, I
guess, Mike, do you want to say a few words before you
start?
MR. MAYFIELD: Yes, just briefly, to frame this.
We had hoped to come to the committee at this meeting with a
proposed resolution to this generic safety issue. However,
as you'll hear, sort of the last bullet in the slide
package, we're still considering some of those closure
options.
So we're not prepared today to talk about the
closure. What we wanted to do was to come and present to
you what we have done, what we have found, and looking both
at the testing that was done and at the risk evaluations,
the scoping studies that have been performed, and to solicit
any input, any questions you might have, any areas that you
think we should specifically be prepared to address when we
do bring the closure package.
So it's something of a status briefing, but at the
same time, we're hoping to see if there are any insights you
particularly want us to make sure we address when it comes
to the committee.
DR. UHRIG: Are you asking for a letter in
conjunction with this?
MR. MAYFIELD: I think at this stage, a letter
would be premature, because we don't have the closures to
propose. If the committee chose to write a letter, of
course, that's obviously your prerogative, but we are not
specifically seeking one.
When we bring to you the proposed closure package,
at that point, we would need a letter.
DR. SEALE: When do you expect to have closure?
MR. MAYFIELD: I had hoped to have it for this
meeting.
DR. SEALE: I understand.
MR. MAYFIELD: At this stage, hopefully yet this
year. We are on a list that goes down to Capitol Hill that
shows closure by December 31. So I am hoping to make that
milestone, but I also wouldn't want to prejudge this. I'm
not trying to give you a flip answer. It's just while the
staff is still debating this, I'm hesitant to venture a
strong guess.
DR. SEALE: We appreciate your incentives.
MR. MAYFIELD: With that, I'd like to turn it over
to Ed Hackett, who will start the briefing.
DR. POWERS: Ed, before you start, I noticed that,
in looking through your viewgraphs, that some of the
research was done by Sandia National Laboratories.
I have some vague association on a very episodic
basis with that organization, and so I may have a conflict
with this.
MR. HACKETT: Okay. I guess, like Mike said, on
the schedule, we're actually revising our operating plan
milestones just today and we'll hopefully still be able to
live with some of the ones we put forth there.
I guess some introductions are in order here. The
folks at the table here, myself, I'm Assistant Branch Chief
for Materials Engineering Branch, in the Office of Research.
On my right is Dr. Bob Lofaro, from the Brookhaven
National Laboratory, who is the actual researcher who has
performed a lot of the research under contract for NRC.
To my left, immediately, Satish Aggarwal. Dr.
Aggarwal is the RES Program Monitor for the program at
Brookhaven and has done the bulk of the work in this area,
along with Dr. Jit Vorah from the Office of Research, and
Jit is over there behind Mike.
And to the far left is Mark Cunningham, who is the
Chief of the Probabilistic Risk Assessment Branch in
Research, and Mark will be addressing some of the risk
significance aspects of the work.
I'm sorry. Satish just reminded me. Jose Calvo
just joined us, also, from NRR. He is at the table next to
Mike. So with that kind of talent at hand, we ought to
be able to hopefully -- don't want to get too cocky here,
but hopefully we ought to be able to address some of the
issues.
DR. APOSTOLAKIS: Finish it up today.
MR. HACKETT: In terms of a purpose or objective,
I think Mike covered that, but just to restate it here, we
had hoped to have sort of a resolution path or approach to
discuss with you in more detail today.
Obviously, as Mike said, we're not going to be
able to do that today. So what we want to do is basically
review the technical background and the regulatory
background to the issue and then just give you sort of
status of where we are in a technical and regulatory sense.
I think Dr. Uhrig covered an awful lot of the
background, but I'll go into some more of that in a block
diagram.
Before that, I had one I was going to put up that
is not in your package. I debated whether to put this up or
not, but since there are a variety of different folks and
backgrounds represented in the room, I thought I would at
least mention some of this in the way of background, and I
hope I'm not offending anybody here.
But, basically, what we're looking at here, as
Mike Mayfield has characterized many times, is less of an
electrical engineering problem, actually more of a polymer
materials engineering problem in terms of degradation.
You've got the polymeric materials, basically the
jackets and insulators that are degrading. They are
subjected to obviously thermal and irradiation environments
in the nuclear plants.
DR. POWERS: Can you give me some sort of feeling
about the total doses that these jackets are getting?
MR. HACKETT: I cannot, but maybe Bob, Satish.
DR. AGGARWAL: We plan to cover that when we
discuss the test program with you. Would that be all right?
DR. POWERS: That would be fine.
DR. WALLIS: These are not subjected to
ultraviolet or something?
MR. HACKETT: Not at -- don't worry about it, Jit,
I'll do it from here.
DR. WALLIS: There's no ultraviolet radiation
effect.
MR. HACKETT: Not that I'm aware of.
DR. WALLIS: They're away from bright lights.
MR. HACKETT: They're in the containment, mostly.
DR. AGGARWAL: The cables are qualified to all TID
source term and that's what these require for all nuclear
power plants, TID source term.
MR. HACKETT: This was, again, just, also, in
terms of setting the stage, but like a lot of these types,
we just briefed the committee yesterday on neutron
irradiation embrittlement of reactor vessels.
Like that, this degradation is progressive with
age, like you would expect, results in embrittlement. I
think Dr. Aggarwal has some samples here we'll pass out at
an appropriate time to show you the extent some of this
embrittlement can take. It can be pretty dramatic.
And, obviously, the embrittlement increases their
susceptibility to cracking, and that almost goes without
saying. A lot of times, the failure mechanisms are then
from moisture intrusion, which can either lead to gross
failures or shorts in the cables or potentially leakage
currents, undercurrent type situations, which could give you
misleading information in terms of instrument and control.
So that's just in the way of sort of setting the
stage.
DR. WALLIS: So there's no humidity effect or
anything. You say thermal and irradiation. It seems to me
a lot of the things degrade --
MR. HACKETT: Yes, absolutely. I didn't mean to
imply there's not a humidity effect. That's obviously there
and it's synergistic with the other effects, too.
So with that, what I would do is move on to the
next slide in the package, sort of an attempt to run through
how this is set up and I think Dr. Uhrig covered a lot of
this, so we don't need to be overly redundant.
But, obviously, this goes back to the regulatory
basis, including an EQ rule, which was promulgated in 1982.
There's a Regulatory Guide 1.89 which is part of the process
that endorses IEEE standard, that basically set a qualified
life or a qualified envelope type of approach to
qualification of these cables.
However, on the slide you see here, the EQ aging
research that was conducted at Sandia, as Dr. Powers noted,
showed some failures of low voltage I&C cables and in some
cases here, you'll see, Dr. Aggarwal is going to go into
greater detail on an upcoming table here, but there were
percentage failures that increased with time in different
cable categories, and we'll go into some more details there.
This was about vintage 1992, that this research
was completed. There were information notices that were
issued by the Office of Nuclear Reactor Regulation to cover
this area.
There were also issues associated with connectors.
DR. APOSTOLAKIS: Are these accelerated tests?
MR. HACKETT: They are accelerated tests, and
that's part of what will be covered, too, the aging.
DR. APOSTOLAKIS: It will be covered.
MR. HACKETT: Yes.
DR. SEALE: You say 18 at 20 years, 23 at 40
years, 32 at 60 years. Are those aggregates? That is, is
the increment between 20 and 40 the difference between 18
and 23?
MR. HACKETT: I'd turn that over to Satish.
DR. AGGARWAL: The numbers are actually -- they
did the tests on 20, 40 and 60 years.
DR. SEALE: Yes.
DR. AGGARWAL: When they did the testing at 20
years, 18 percent of the cables failed had low IER. When
they were doing it for 40 years, it was 23 percent failure.
DR. SEALE: Twenty-three percent more.
DR. AGGARWAL: Of the set.
DR. SEALE: Of the set. So that included the 18,
plus five more.
DR. AGGARWAL: No. We studied the independent
test. They are doing three tests.
DR. SEALE: Okay.
MR. HENDRICKS: One for qualified life equivalent
to 20 years.
DR. APOSTOLAKIS: It's cumulative.
DR. SEALE: There are separate ones, but --
DR. UHRIG: Separate tests.
DR. SEALE: Yes, okay.
DR. UHRIG: Not the same cables.
DR. SEALE: Not the same cables, but presumably
comparable sensitivities.
DR. UHRIG: Yes.
DR. APOSTOLAKIS: And clearly it's not linear.
DR. WALLIS: The rate slows down.
DR. SEALE: Well, it's saturating.
DR. SHACK: You can only get so dead.
DR. SEALE: That's right. That's the truth.
MR. HACKETT: Dr. Shack's comment is probably
particularly appropriate. When we pass some of these
samples around, I think you will see that some of these,
that the failures can be fairly dramatic, particularly with
the bonded jacket cables.
DR. SEALE: Now, let's don't make bad comments
about geriatric activities.
MR. HACKETT: Okay.
DR. APOSTOLAKIS: You will talk about these tests
later?
DR. AGGARWAL: Yes, sir, if you'd like to hear.
DR. APOSTOLAKIS: Well, if you talk, I'll hear.
MR. HACKETT: Trying to move across the chart
here. Out of the research work and the regulatory summaries
that were issued, that coupled with some knowledge of what
was going on in terms of operating experience is what you
see on the center of the chart, led to some additional staff
actions.
There was an EQ task action plan which was
promulgated in 1993, which was at least one of the issues
that was to be addressed in that task action plan was
looking at accelerated aging and was accelerated aging
indeed a valid way of coming at this thing.
And for those of you who have been involved in
this, this involves the Arrhenius methodology. I think the
bottom line is what comes out of that is, yes, that's a
valid methodology, as long as you're not changing the
mechanisms of the cable aging or failure as you're going
through that process.
The EQ task actin plan was actually closed by the
NRC in 1998. The long-term research aspects of that were
transferred into a new GSI, 168 at the time, which is the
subject of why we're here.
Then there was ongoing research, obviously,
associated with that GSI. Forth-three technical issues,
you'll see, at the bottom of the slide, that were identified
and that was in NUREG-CR-6384 in 1996.
Those research issues have been ongoing or
addressed in ongoing research for us at the Brookhaven
National Laboratory.
I should also point out that the industry has also
been active in this area, even more so now with the EPRI/DOE
NEPO program. There is continuing work at Sandia and
elsewhere under EPRI funding working in this area, too.
Some of the cable types that are addressed you can
see under the box that has the EQ aging research under
Brookhaven identified. We've seen failures across many
different types of manufacturers.
Again, Satish and Bob maybe will get into some
better definition here as we go through what is or isn't a
failure, some of that is problematic in how you define these
things.
But to the -- maybe just the final point, the top
of the chart, obviously, considering license renewal, I
think one of the things we'll come to in summary is that the
situation for the operating plants I don't think we feel is
any dire situation by any stretch.
For the license renewal period, on the other hand,
we think something is going to need to be addressed. The
something and the process and the approach for doing that,
like Mike was saying, remains to be determined or proposed
by the staff, but that's sort of where we are now.
DR. WALLIS: Excuse me. Your 18 percent at 20
years is the result of some research program.
MR. HACKETT: That's correct.
DR. WALLIS: There have been plenty of cables out
there for 20 years already. So how do they do? Do they
fail at the same rate?
DR. AGGARWAL: When you're talking about 20 years
cable, you're talking about the operating plant and their
operating environment. When we are talking about the test
here, we're talking about an axial cable and LOCA test. We
are not reporting here these are expected failures in the
operating plant in normal life.
DR. WALLIS: How do we relate these numbers to
reality out there in the plant?
DR. AGGARWAL: I'm going to talk about it as we
proceed.
MR. HACKETT: That part will come. As a matter of
fact, what I will do at this point is just outline the
remainder of the presentation. I will turn this over at
this point to Dr. Aggarwal and he will summarize some of
these tests, and Bob Lofaro is here, also, from Brookhaven
to assist with that, if needed.
In the part of your package that goes to the risk
considerations, Mark Cunningham will address those.
At this point, I guess I would say, though, if
there are any overall questions on this background for any
of the staff that are here present, this would probably be a
good time to hear those. Otherwise, we'll just go ahead and
move on.
Hearing none, I will turn this over to Dr.
Aggarwal.
DR. AGGARWAL: Thank you, Ed. Before I talk about
the program, and what we have done or not done, I would like
to take the opportunity to pass the board, on my right side,
which shows you the typical I&C cable which are used in a
nuclear power plant and these are new cables.
On my left, I am going to pass some photographs of
the cables which are degraded conditions during the test.
Anther example I am going to hand you is the Okonite 40 year
post-LOCA cable, and please look at the condition what we
found.
The next sample I have is the anaconda cable,
three-conductor, No. 12, unbundled cable.
DR. UHRIG: These are used primarily for signals,
I&C signals.
DR. AGGARWAL: That's right. All the cables which
you are looking at are used in a nuclear power plant for
instrumentation and control only. We are not discussing any
power cables at all.
DR. UHRIG: What voltage?
DR. AGGARWAL: Normally, they are rated 600 volts,
but in a plant, they may be 24 volts.
DR. UHRIG: Twenty-four volts mostly.
DR. AGGARWAL: Right. I am going to pass another
sample of a Rockbestos cable, which is a two-conductor XLPE
neoprene. This is the 40-years aging and you will see the
cracks which are visible for -- without glasses you can see
them. At the same time, you will see the jacket is brittle.
I am going to pass two samples, one on this side
and one on that side. These are the Samuel Moore cables,
which are two-conductor, No. 16, EPR Hypalon bonded, which
fail during our look at test No. 4.
Before I go further, that we were, in other
program, looking at only three types of these cables, which
are most popular and commonly used in nuclear power plant,
and the sample which I've passed on to you, the big board
has all those three kind of cables and the other sample
which I am passing on are essentially the cables which
failed or did not meet our acceptance criteria.
Now, this two-sample, Samuel Moore, which I passed
on, you will see some cable mark, black spot. This is where
the cable failed. One rad conductor was punctured.
The next sample I have is a Rockbestos XLPE
neoprene, 60 years, which is similar to test No. 4. And if
you look at it, you will see how the jacket is cracked and
is brittle.
Another sample I have here to show to you is after LOCA test
and, again, this is Rockbestos XLPE neoprene jacket, which
undergone 60 years of aging and post-LOCA.
DR. UHRIG: I want to make sure I understand.
This is aged to 60 years and then --
DR. AGGARWAL: To a LOCA.
DR. UHRIG: Then to a LOCA.
DR. AGGARWAL: Bob, if you like, I can run down
basic program how we do the testing.
DR. UHRIG: I think it would be useful to at least
give us a quick review of that.
DR. AGGARWAL: Let me do that. Let me finish one
more sample and I go back to that.
DR. UHRIG: All right.
DR. AGGARWAL: This cable is the Okonite cable, 60
years of aging, post-LOCA, and look at the -- this thing
completely failed. You could see the conductor through all
the length.
DR. WALLIS: If we handle this, it will fail some
more.
DR. AGGARWAL: I just wanted you to have a feeling
what you're going to look at.
This is the Anaconda cable, which is 40 year and
post-LOCA, and you could see, again, the cable here.
DR. WALLIS: Is it the LOCA that does this to it
or is it in pretty lousy shape before?
DR. AGGARWAL: It depends on what kind of
insulation you are looking at.
DR. UHRIG: Do you have any test where you apply
the LOCA to brand new cable?
DR. AGGARWAL: Yes, sir. Each cable, we had a
controlled sample.
DR. UHRIG: Controlled sample.
DR. AGGARWAL: Yes. Let me go through the process
very quickly and that will give you some understanding.
As discussed earlier, you know that our
requirement is 10 CFR 50.49, the EQ rule, the reg guide,
they require that cables be qualified and qualification here
is a verification of design.
Namely, we want to ensure or assure us that these
cables will perform their intended safety function in the
expected environment.
It is a one-time test. Although the reg guide
makes it implicitly that this qualification should be
maintained for 40 years, qualified life. But there is no
explicit requirement in our regulations which require that
we do any kind of monitoring.
Keeping that in mind, let me try to explain the
qualification process to you. What we do here, we will take
a brand new cable, and our goal here is to bring that cable
to the end of life, qualified life.
If it is 40 years, then we want to bring that cable to the
equivalent to at the end of 40 years, and the reason being,
a LOCA can take place at 39 years and 364 days.
So we are trying to bring that cable to end of
life condition. How do I do that? Because we don't have a
nuclear power plant which have seen 40 years of natural
aging.
So we had to go and look at the Arrhenius theory
and see what does it tell me. Coupled with the activation
energy, we come out with a certain number. That for so many
hours, if you preheat the cable at this temperature, then we
will consider that this is equivalent to 40 years.
Now, NRC paid no role in arriving those
conditions. The manufacturers were told that go ahead, we
understand their limitation of the Arrhenius theory, but
come back with those parameters and those parameters were
picked up by each manufacturer as he deemed necessary.
And it is also interesting to note, many times,
you find the same kind of material, but different
manufacturers. They came out with different numbers.
Again, NRC was not dictating those numbers, they
were there. In our test program, we decided that we will
use the same parameters which were originally used by the
manufacturer. We will make no deviation whatsoever in terms
of the rate of aging.
So what we do now, we'll take a cask sample, brand
new, put in an oven at a certain temperature, and for so
long time in each case as dictated by the manufacturer.
When that is done, then we will put, on the same
cable, the operation type radiation and then we go to the
accident rating.
Once that is done, then we go into what we call a
LOCA test. The guidance for test, how you do all these
things, is provided in IEEE 323.74, which is endorsed by Reg
Guide 1.89.
Now, this particular standard for wide profile are
for B's and B's and they are typical and these standards
clearly state that you can follow them, they are typical, or
you can have plant specific.
My experience tells me that the industry chose to
use those profiles. The profiles were two peak profiles and
a single peak profile and, again, these profiles were
provided to us many, many years by manufacturer, vendors,
and they are typical in nature.
This is a two-peak LOCA profile and it's
definitely more conservative compared to a single peak.
Now, again, some manufacturers chose to qualify
their cables to two-peak rather than one peak and NRC
accepted both.
DR. WALLIS: You test in water. Are you going to
get to that?
DR. AGGARWAL: Yes, sir.
DR. WALLIS: Are you going to say what the water
is?
DR. AGGARWAL: Yes.
DR. WALLIS: Because there are all kinds of water.
DR. AGGARWAL: Then there are certain tests which
are called post-LOCA testing. What we are doing here, we
take the sample, which has come out of the LOCA test, after
doing some condition monitoring and so on, we apply what we
call the strain test. The cables are now submerged in
water.
This is a requirement of IEEE 383.1974 that these
cables must meet this requirement. The water is ordinary
tap water and it the cable is somewhat into it and we apply
80 volt per mil. Usually insulation is 30, so you see about
24 on the board.
Now, note it is not always necessary that cable
will fail at 2,400 volts. It can fail at any intermediate
range.
Any failure which cable, if it fails to withstand
water 2,400 or lower will be considered as failed.
DR. WALLIS: There's nothing about the chemistry
of the water. It's just some sort of tap water.
DR. AGGARWAL: It is tap water. This is what is
required under the standard and this is how the industry has
done.
DR. WALLIS: We don't have tap water in
containment after a LOCA.
DR. AGGARWAL: True. That's very much true, but,
again, we are just meeting acceptance criteria in the 1970
timeframe and -- I'm just telling you what industry doing or
what we have done.
DR. SHACK: Again, so you're not irradiating the
cables. You're simulating all the damage by some sort of
aggressive thermal history. Is that the idea?
DR. AGGARWAL: No, we are putting radiation doses,
too, as a part of the pre-aging. We go through the sequence
of thermal aging and then the normal radiation and then the
accident radiation.
DR. APOSTOLAKIS: So the normal radiation you use
some kind of a law like the Arrhenius law again to determine
the time?
DR. AGGARWAL: Generally, as I said earlier, the
industry came and 50 mega rad was considered to be standard
number that the industry has used, not only in this country,
but throughout the world, for qualification.
One may argue, and that is my next point, people
have argued that these tests are too conservative and I want
to present my personal point of view, not NRC staff point of
view, my personal point of view are these tests are really
conservative.
The first point is that pre-aging environment,
which we are preconditioning the cables to, as I told you
earlier, you take the activation energy, the Arrhenius
theory tells you so many hours equivalent.
DR. APOSTOLAKIS: But that's for temperature.
DR. AGGARWAL: That's the temperature, the time.
DR. APOSTOLAKIS: Not for radiation.
DR. AGGARWAL: No.
DR. APOSTOLAKIS: Okay.
DR. AGGARWAL: So now people say, well, in a
nuclear power plant, you never see those conditions. All
right. Sitting here, I don't know any better matter to
bring a cable to the end of condition 40 years life. Only
thing I know of the limitation, the Arrhenius theory, and
this is what we do --
DR. APOSTOLAKIS: I don't understand this. You
never see what conditions? The temperatures that you are --
that is not the idea of an accelerator test.
DR. AGGARWAL: The idea of accelerator test is to
bring the cable to the end of condition.
DR. APOSTOLAKIS: Right.
DR. AGGARWAL: Forty year. What industry argued
is that the tests are unrealistic. In real life, cable
never sees these kind of temperatures and the point I am
making there, the only criteria in the consensus standard
was to bring the cable to the end of life, 40 years, and
they came with a number.
Bottom line, I'm saying, in my opinion, those
numbers are not very conservative.
DR. APOSTOLAKIS: So if we put it in a different
way, the argument is that if you have a cable subjected to
temperature T-1 for 40 years and then you have an
accelerated burst at a higher temperature for 260 hours, the
result of the test, the damage to the cable is not the same
as you would get for T-1 for 40 years.
DR. AGGARWAL: Some people argue. Let me make
another --
DR. APOSTOLAKIS: Is there any basis for that?
DR. AGGARWAL: Yes. Another argument -- may I put
-- from the industry's point of view, in my 30 years in EQ
at NR, what I found, any failure in the EQ area can always
be explained and you can declare victory.
I have not found, it is not because of requirement
are not very clear, explicit, they require single prototype
be tested.
If it fails, the staff didn't know. Industry is not
required to report, but declare cables are qualified and
they are there in the plant.
DR. WALLIS: Can you tell me more about this
accelerated test? You don't have to just do it at T-1 for N
hours. You can do it at a lot of temperatures for a lot of
different times and show that this fits on a curve and it's
all -- that, therefore, your theory is kind of valid. You
don't just do one test.
Isn't that done?
DR. AGGARWAL: No.
DR. SHACK: Was it done once upon a time?
DR. APOSTOLAKIS: Well, when Arrhenius proposed
the law, I guess he did it.
DR. SHACK: That's a long time ago.
DR. APOSTOLAKIS: It was a long time ago.
DR. SHACK: I mean, for the particular cable, I
presume that somebody did a test over a range of
temperatures and demonstrated a kind of linear behavior.
DR. AGGARWAL: No. They used the activation
energy and the Arrhenius theory and came up with a number.
MR. MAYFIELD: This is Mike Mayfield, from the
staff. If you go back into the polymer science, you'll find
that that kind of work has been done and the Arrhenius
applies for it, just like it does in metallurgy, as long as
you stay within the same degradation mechanism. If you go
high enough in temperature, the mechanism changes, the whole
thing is over.
From the test reports I've seen that we've
acquired, when they qualified a cable, they did not run at a
series of temperatures. They set out to qualify it for a
set of conditions where you could operate the cable
ostensibly at this set of conditions for 40 years and the
tests were run to qualify the cable to operate at those
conditions.
There weren't a series of tests run, but this --
now we're talking about the practical implementation of a
piece of science. The series of temperatures, that type of
work was done in the research labs effort the standard was
written.
So there is science that says as long as the
degradation mechanism doesn't change, this methodology
holds.
DR. SHACK: But, I mean, you've established that
the range we're extrapolating over is within the range that
people had demonstrated.
MR. MAYFIELD: Yes. Now, there's been subsequent
work done at Sandia looking at -- because some of the
polymers exhibit, which called non-Arrhenius behavior, and
Ken Gillen out there at Sandia has done some work looking at
why that occurs, when it occurs, and how to predict it.
And for the work we have here, we didn't feel like
it was much of an issue, but it's something that is being
followed up both through the DOE/EPRI NEPO program, as well
as stuff that we're at least following what's going on.
So there is continuing work in this area sort of
at the basic technology. What we are presenting today is
the application end of it. But the underpinning, the
scientific underpinnings are there.
It's just they're not -- it's not done for each
qualification test.
DR. WALLIS: Where does the conservative adjective
come in? I would think the question is are these realistic
tests. What is the difference?
MR. MAYFIELD: Well, the concern -- I think your
characterization is exactly right, Dr. Wallis.
DR. WALLIS: Are they realistic tests?
MR. MAYFIELD: It's are they realistic versus
conservative, the argument being that, oh, they're too
conservative, but then are they realistic, and I think it's
a fair question.
The issue was to try and qualify cables within an
envelope, if you will, for unrestricted use in the nuclear
power plant.
DR. WALLIS: Conservative, to me, would mean that
you look at the errors in the Arrhenius theory, the
deviations from it. You've got some kind of a bounding
thing because of the uncertainties and then you go to that
bounding test, and no one has said that anyone has done any
bounding test.
It looks as if this is some sort of an average
realistic type test that's being done. So we don't know the
limits in terms of a bounding conservatism.
DR. AGGARWAL: That's correct.
DR. WALLIS: Maybe the word conservative is
irrelevant in describing these tests.
DR. AGGARWAL: I was just trying to pose -- well, concluding
that, we do the same task, where I explained to you, that
some is in the water.
DR. WALLIS: And this is just tap water.
DR. AGGARWAL: Yes, sir.
DR. SHACK: Let me go back just to the question
again. Now, we did the pre-aging and then the irradiation.
Have people examined the effect of the sequences?
DR. AGGARWAL: Yes, sir, they have and there are
different conclusions. Incidentally, I've been already
reminded twice about the time. Last night I told my wife
that don't wait for dinner for me. So I will be here as
long as you want me. The pleasure is all yours.
DR. APOSTOLAKIS: But we will not.
MR. MAYFIELD: Dr. Shack, this is Mike Mayfield,
from the staff. The tests done at Sandia, in fact, do
simultaneous temperature and radiation exposures, but the
IEEE standard has sequential exposures.
I guess I personally don't know if someone has
swapped irradiation versus temperature, but I know that we
got somewhat different results and I think you'll see some
of that in the next couple of slides that Satish presents,
when we do simultaneous thermal and irradiation exposures
versus sequential.
The sequential appears to be more severe than the
simultaneous.
DR. UHRIG: What is the radiation test? Gamma?
MR. MAYFIELD: It's gamma.
DR. UHRIG: Is it Cobalt-60?
MR. MAYFIELD: Cobalt-60 sources.
DR. UHRIG: So it's not the same energy spectrum
that you might expect in a plant.
MR. MAYFIELD: It is not the same energy spectrum.
It was a TID source term estimation and they take a
cumulative gamma dose and irradiate the cables using the
Cobalt-60 sources.
DR. WALLIS: I guess all these questions wouldn't
be so important except that the failure rates are a
significant percentage, and therefore, we do have to worry
about the meaningfulness of the test.
DR. AGGARWAL: Bill, to your question, I just want
to add that in certain part of the world, for certain type
of insulation, they have found that if you do the radiation
first, and thermal things later, it's more conservative for
some type of insulation, not all, and it is consistent with
the conclusion what Sandia reached some time ago.
And, incidentally, I may also point out, in this
program, we have produced that many reports which have been
circulated around the industry and public for their comments
and their input will all be incorporated.
Now, I would like to move on to page 3.
DR. WALLIS: I'm sorry. This temperature test is
uniform temperature?
DR. AGGARWAL: Yes, sir.
DR. WALLIS: Are these cables carrying current?
They're not subjected to --
DR. AGGARWAL: Yes, sir. They are energized in
the nuclear power plant.
DR. WALLIS: They just set an environmental
temperature. There's no heating because of the current in
the cable.
DR. AGGARWAL: Right.
DR. WALLIS: Is that right?
DR. AGGARWAL: The temperatures, we tried to
maintain, what we came out, we are not giving any kind of
credit for the heating created by the current going through
the cable, because these are very small cables.
All we have a very limited current. So really
there's not a heck of a lot of current.
DR. WALLIS: Right, as long as someone has
assessed it.
DR. AGGARWAL: Yes. Well, we have taken those
temperatures which the manufacturers had taken. We did not
make an attempt how --
DR. WALLIS: It's just if you have a lot of cables bundled
up in a tray and they're all generating small amounts of
heat, it might affect the temperature in a way which isn't
just the environmental temperature.
DR. AGGARWAL: You're right, but again, normally,
on a controlled cable, you only allow 60 percent and there's
plenty of space and these cables are hardly getting any
current. They are control and indication type cables.
In this, all I want to tell you, if you go to the
bottom line, you will note this is what they found in
Sandia, two out of five fail in the 20 year, one out of
eight fail in 40 years, and five out of 11 fail, then we are
talking about 60 years
DR. WALLIS: Show us on the table where those are?
MR. SIEBER: The two and five are -- two failures
and five -- what is the category here, marginal criteria?
DR. AGGARWAL: Right.
MR. SIEBER: A and B, down here at the bottom.
DR. UHRIG: You had two failures and five marginal
is what that means.
MR. SIEBER: Out of 39.
MR. LOFARO: So seven had a problem out of the 39.
DR. UHRIG: Seven out of the 39.
MR. LOFARO: Two of them were failures and five
were just marginal.
DR. WALLIS: So two failed in that 20 years and
one failed in 40 years. So one that was failed recovered.
DR. AGGARWAL: It's a separate --
DR. WALLIS: I know it's a different test, but if
you think it's meaningful, that would look as if one
recovered.
DR. UHRIG: It's within the error of the
experiment.
DR. WALLIS: But then it seems a very small number
to rely on, doesn't it? If there is that sort of behavior.
DR. AGGARWAL: Well, it all depends, again, how
you define your acceptance criteria.
DR. WALLIS: I'm not defining anything. I'm
asking, really.
DR. AGGARWAL: In that particular test, the
criteria was that if you have higher value, they're lower,
or you blew a one-amp fuse, then it would be considered a
failure or marginal.
DR. WALLIS: But if I drew a curve of sort of
failure rate versus time and I had tow and then one and then
five, with these numbers, I wouldn't have much reliance on
the meaningfulness of the result, would I? Very small
sample size and not even consistent with the trend.
DR. AGGARWAL: You're right. You're right. And,
again, we --
DR. WALLIS: So who decided these were adequate
tests?
DR. AGGARWAL: Who decided? They came out of the
IEEE Standard 323, which is a national consensus standard.
This is what EQ experts are telling is the right thing to
do.
DR. APOSTOLAKIS: Do they specify the size of the
sample?
DR. AGGARWAL: They don't talk, but it's implicit
that you take -- well, let me rephrase it. There's no
explicit statement in the standard what industry have done,
they test one sample and if that passes, you declare
victory.
DR. WALLIS: And if it doesn't?
DR. AGGARWAL: You test again.
DR. APOSTOLAKIS: We should risk-inform that.
DR. AGGARWAL: That is one of the concerns we
have, whether we should be doing.
DR. UHRIG: That's what the manufacturer does.
That's not what you did here.
DR. AGGARWAL: We reported what we found.
DR. UHRIG: You had about ten kinds of cables
here.
DR. AGGARWAL: Right.
DR. UHRIG: And I notice you had five specimens,
some you had two. Was there --
MR. SIEBER: Some you had one, the Anaconda.
DR. POWERS: Yes. And the question is how was
that number determined. On the basis of the amount of that
cable that was used?
DR. AGGARWAL: Not really. The number of the
cable size, then we had -- This will go, again, a little
more in detail. Some of the cables were around the
mandrill. Some were in the tray. And it's different --
DR. WALLIS: But does the failure rate depend on
how long the cable is? Is there one-inch cable or one-meter
cable and they're looked at the same way?
DR. AGGARWAL: I believe you use a ten-foot long
sample.
DR. WALLIS: The whole thing seems somewhat
strange.
DR. UHRIG: It also looks like certain cables,
like Anaconda, unbounded very well in this first test.
DR. AGGARWAL: Right.
DR. UHRIG: And on the other hand, it didn't do
very well over in the 40-year test on the second page.
DR. AGGARWAL: Correct. The second page you are
looking at are the tasks which we did at Wylie Testing Lab.
DR. UHRIG: It says Brookhaven here.
DR. AGGARWAL: Yes. Brookhaven don't have the
lab, the testing. They're actually Wylie Lab.
DR. UHRIG: So Brookhaven supervised it.
DR. AGGARWAL: Right. They have a contract on
this program. If I turn over to the next slide, again, this
is a summary of six tests which we did. And you see the
results. And I will be willing to answer your questions and
explain to you what those failures were
DR. WALLIS: Tell me. If you do five feet of one
foot of cable, is that different from doing one length of
five foot of cable? What's the difference? So why is it
five tests?
MR. LOFARO: What we did in the test program is we
used links of cable as recommended in the IEEE standard,
which is they suggest you use about ten feet so you get the
proper effect.
Obviously, if you took one foot of cable, you
might miss something. So you want to take a sample that's
at least representative of what the cable might actually see
in the plant.
DR. WALLIS: So they do specify a length.
MR. LOFARO: They recommend a length of about 10
feet, and that's what we use in our program. And what they
do is our program we actually wrap it around the mandrill,
so that there was a good ground plane.
DR. WALLIS: These five samples are put in the
same testing apparatus or they're tested at different times?
MR. SIEBER: They were all tested exactly the same
time. What we did is --
DR. WALLIS: So if I had 50 feet of cable, it
would be like having five pieces of ten foot.
DR. AGGARWAL: Right.
DR. WALLIS: Because they all put through the same
test.
DR. POWERS: They are all put through the same
test.
MR. PIETRANGELO: They go the
DR. AGGARWAL: Right. They go through the same
process.
MR. LOFARO: Well, what we wanted to do is we
wanted to have multiple samples in there so that we could do
some -- have some statistical accuracy on the samples. We
put control sin there that did not get aged. We also put
samples in there that were pre-aged to different years.
Some tests were 20 years, some tests --
DR. WALLIS: I'm sorry. The cable, the ten-foot
cable fails, does it fail in one place or in many places or
all over?
MR. LOFARO: Well, that's one of the things that
we were looking for and what we actually found is that some
of the cables had a global degradation. For example, the
Okonites I think that will be touched on.
We found some cables that only failed locally,
where most of the cable was in excellent shape, but when we
subjected it to a subjected it to a submerged high pot test,
there were single point failures in the cable.
So it depends on the type of cable you're looking
at and the materials they're constructed of and what
conditions they're exposed to.
DR. WALLIS: But you never seem to get five out of
view. Still, if you put these links in and one survives and
the others don't.
MR. LOFARO: If we had a case where there were
five out of five failures, I think that would clearly point
to a problem.
DR. WALLIS: I'm just kind of surprised that you'd take five
lengths of identical cable and do exactly the same test,
you'd think that in many of these tests, you either get the
all pass or they fail.
You wouldn't get many cases where one fails and
the others don't.
MR. LOFARO: Well, unless they're reaching a point
where the degradation is just borderline. Maybe there's a
problem, maybe there's not.
DR. APOSTOLAKIS: That means you have alliatory
uncertainty. That's exactly what you have.
DR. AGGARWAL: I think it's very difficult to predict how
many will pass or not. For example, Okanite cable, three
sampled or three failed out of three.
DR. WALLIS: But these cables are sort of
identical. The manufacturer --
DR. AGGARWAL: Yes. They are coming out of the
same reel. Mike wants to add something.
MR. MAYFIELD: I was just going to reiterate
something that I guess Satish had said earlier and ask you
to please keep in mind the qualifications standard is based
on one successful prototype test. The issue that we were
addressing with these cables is, is that realistic. If you,
in fact, test more than one sample, do you get all passes.
The notion that one prototype are ten foot length,
nominally, tested and successfully passes the IEEE test
qualifies all cable of that particular type is one of the
things we were challenging through this research.
So whether we got all passes or all fails, and,
Dr. Wallis, I think you talked earlier about are the tests
conservative or are they realistic, the part of the notion
is to put the cables through their original qualification
and with multiple samples, whether it's three or five, to
determine whether or not this notion that qualify one sample
and, therefore, all cables are qualified, is that really a
legitimate notion.
And I think at least these test results raise some
questions about that and it's a dialogue that we will
continue to have with IEEE.
DR. WALLIS: It may well be that to answer the
question, you need to test 100 of them.
MR. MAYFIELD: That's the question, is how large
would the sample have to be. Conversely, do you accept the
notion that this qualification standard doesn't get you to
all cables pass and, therefore, you test perhaps more than
one and then try and establish some failure frequency.
DR. WALLIS: Don't you have just experimental
theory that you're relying on to reach conclusions here?
MR. MAYFIELD: Say again, please, sir.
DR. WALLIS: Don't you have some kind of a
theoretical background about failures and statistics which
tells you whether or not this is a significant experiment,
what it means? I mean, this is a well worked area of
planning experiments and deducing things, making conclusions
when there are random variables.
DR. AGGARWAL: I'm aware, in Sweden, they use some
statistics and they came out with a sample size of ten, but
in U.S. it is one.
I think, as I recall, in that time arena, what
IEEE was worrying about, that if you do ten samples, or
hundred, for that matter, and they argued that nine out of
ten pass or 95 of 100 pass, do you declare victory or you
don't. And there were no criteria for them. So they
decided you just do the one test and the sense of the
standard now and with the whole work to do that.
Now, we can debate the numbers and --
DR. WALLIS: Well, victory should be based on some
logic.
DR. AGGARWAL: When we look back now and think,
you're right.
MR. MAYFIELD: Again, Professor Wallis, our
objective here was to challenge what was being done and to
see if it seemed to hold together. I believe the test
results here suggest there are some concerns. I think the
kinds of things you're raising, we agree with you. We're
not here to try and defend the standard, simply to report
this is what we followed and it seems to be raising some
questions.
DR. AGGARWAL: If I might turn to the next
viewgraph and give you some basic data out of the operating
experience. This is an EPRI report, July 1994, where they
collected some data from '68 to '92, and they found there
was 87 events which identified degradation of in-containment
cables.
We know that we have problem with the splices at
operating power plants. We also know that in our test, we
had the problem with splices in all the tests, test number
one, two and three.
Two passed without a problem, but first and three,
they said there was a moisture intrusion.
DR. WALLIS: Where did the moisture come from?
DR. AGGARWAL: Steam.
DR. WALLIS: There's a steam environment?
DR. AGGARWAL: Yes. It's a LOCA test.
DR. WALLIS: It's a LOCA test. I'm sorry. They
had actually had LOCA in operating experience?
MR. SIEBER: No, they tested it.
DR. AGGARWAL: I'm just making two points. One
was that there had been a problem --
DR. WALLIS: This is to support an LER, so it
can't be a LOCA test.
DR. AGGARWAL: No, no. Two points. There are
real life problems.
DR. WALLIS: But you said moisture intrusion in an
LER. So the moisture came from somewhere --
DR. AGGARWAL: No, I jumped a gear.
DR. WALLIS: I'm sorry.
DR. AGGARWAL: I think went to the second point to
make, that in our test, we also found problems with the
splices and there we have the LOCA.
DR. UHRIG: But you also have problems with
splices at the operating plants.
DR. AGGARWAL: That's right. That is the point
I'm making.
DR. UHRIG: There was not a moisture problem
there.
DR. AGGARWAL: No.
DR. UHRIG: That's probably a faulty splice.
DR. AGGARWAL: One of the reasons. But you know
when many of the LER do not go to root cause analysis and
tell you what really happened. But I know on a personal
experience that you ever try to splice cracked jackets,
splice it, you'll never succeed.
MR. CALVO: Jose Calvo, from the NRR staff. The
problem that we found with the splices in the operating
plants was just the way they was done, in the installation.
The procedures was not followed, they didn't make it long
enough.
Sometimes moisture was found in the terminal boxes
and things like that.
I think it was mostly installation, they didn't
follow procedures and I guess the same question was raised
when we were doing this test with Research and maybe the way
the splice was put together, it was not appropriate and
that's why the issue of failures.
MR. LEITCH: How were the 87 events identified?
Is that 87 LERs or how did they come up with the conclusion
that there were 87 events?
DR. AGGARWAL: This is a proprietary report of
EPRI, which I cannot really go more in detail. But the
conclusion, it is reasonable to conclude that those 87
events were independent events in different nuclear power
plants over the period.
MR. LEITCH: I would think there could be a
considerably larger number.
DR. AGGARWAL: Again, everyone is not reported.
You have in a plant, an MOV fails. What do you do? You go
there, you look at the cable, and you see, well, it doesn't
look very good, you disconnect the cable, put a new cable
and connect it up.
Do you report it to NRC as related? No.
MR. LEITCH: That's my point. There could be many
more than 87.
DR. AGGARWAL: Exactly. Exactly. In the LER and
the NPRDS data, we also found that there were specific
problems with the neutron monitoring systems and failures
were due to high temperature and moisture intrusion.
DR. WALLIS: Where did the moisture come from?
DR. AGGARWAL: In an operating plant, you might
have --
DR. WALLIS: Are these outdoor plants?
MR. SIEBER: It's in containment.
DR. AGGARWAL: It's in containment.
MR. SIEBER: If you get a little leak someplace,
humidity goes to 100 percent and --
DR. WALLIS: So these are associated with some
other event, like a leak.
DR. AGGARWAL: Right.
DR. UHRIG: Why was the neutron detecting cable so
sensitive?
DR. AGGARWAL: Radiation.
DR. UHRIG: Neutron monitoring cable.
DR. AGGARWAL: Radiation. And as I was talking to
you earlier today, my basic concern is every plant had hot
spots.
DR. UHRIG: Right.
DR. AGGARWAL: And my concern is do the plants
really know where their hot spots are, are they really
monitoring those environments, are they really doing
something, I have no idea.
Then we talked to the industry informally, they
say, yeah, many utilities do, but are they all doing, I
cannot answer that question.
MR. SIEBER: I don't think that you survey to find
every place where there is a neutron beam. A lot of times a
cable would fail on a power range monitor at the connector.
DR. AGGARWAL: Right.
MR. SIEBER: Because it's exposed there and it's
in a pretty high neutron flux. Find the same things on the
top of the reactor head and a lot of utilities will end up
shortening and shortening and shortening the cable until
they run out of cable because it ages right where the
connector is.
DR. AGGARWAL: And they made choose to make a
splice there.
MR. SIEBER: Well, other ones may have. Splices
are more trouble than you can imagine.
DR. AGGARWAL: I would like now to turn over the
rest to Mark. The only point I want to make, that cable
aging can be this significant and let the expert take over.
MR. CUNNINGHAM: Thank you, Satish. I've got a
couple of slides to talk about the context of what the
implications are of aging I&C cables might be to risk
assessment.
Over the years, there's been about three studies
done, risk studies done to look at this issue. All of them
have really been in the form of sensitivity studies, because
you get to -- basically, get to a -- you require a level of
information to go beyond the sensitivity study, which we
just don't have.
We don't have the information to work with. These
studies have generally shown that the risk of aging I&C
cables could be significant under a certain set of
conditions.
The studies looked at basically a spectrum of pipe
breaks or LOCAs in the containment and tried to deal with
the fact that the LOCAs would create an environment, a hot
spot, if you will, either very locally in a small break or
very broadly in a large break that could compromise these
cables, as we've seen.
Then the question becomes, well, if the cables
fail, what does that do to the function that's associated
with those cables or does it give a set of indications to
the operator that could mislead or confuse them.
So the cable failures, either the gross cable
failures or the leakage types of failures have the potential
to certainly complicate the accident or remove equipment,
make equipment unavailable, or to confuse the operators.
To go beyond anything like kind of a general
sensitivity study, that you need a fair amount of
information that we don't have, and that's kind of itemized
on slide seven.
In the context of the cables themselves, you need
to have a sense of what types of cables are associated with
what types of equipment in the containment and what are the
more critical pieces of equipment in the containment. You
also need to have a sense of the extent of aging of those
cables, those particular cables and then relate that to the
cables and the location of the cables to the potential
environment.
Again, the offer of what's the relationship
between the hot spot from a pipe break or something like
that with a cable of a certain type for an important piece
of equipment. The other part of that is given these types
of things, how would the operators respond. Do partial
failures of cables cause more problems than complete
failures and that sort of thing.
So the risk study in this case is, in one sense,
simple. It could be important. We don't have enough
information on the details of where the cables are and the
functions and the types of cables and that sort of thing to
take it much more than saying it could be.
DR. WALLIS: You're saying this is what you would
have to do if you were to do a risk assessment, but
essentially a risk assessment has not been done.
MR. CUNNINGHAM: A risk assessment has been only
done in a very general sense, gross sense, that's right.
DR. UHRIG: So one of the considerations you have
under study is to seek this information from the utilities.
MR. CUNNINGHAM: That comes back to Mike's
discussion earlier of the options for pursuing this generic
issue.
DR. UHRIG: This is one of the options then.
MR. CUNNINGHAM: That's one of the things we talk
about, yes, is trying to obtain this information from
licensees.
MR. LEITCH: Would you expect that BWRs would be
significantly different than PWRs in this case? I'm
thinking about under the bottom hood of the BWR, there's
just a real rat's nest of cables under there.
MR. CUNNINGHAM: I couldn't tell you whether that
would be more important or not.
MR. MAYFIELD: One of the scoping studies looked
at BWR versus PWR and the B's came in somewhat lower, but,
again, it went to -- I wouldn't want to put a lot of
credibility in the number simply because of the level of
uncertainty and the level of information we don't have to do
it.
But they weren't -- it wasn't night and day, but
the B came in lower than the P.
DR. UHRIG: What about international studies, what
else is going on? I remember hearing a report about five or
six years ago on some work in Sweden, I believe it was. A,
do you have access to it, and, B, is it comparable results?
MR. MAYFIELD: We have had and continue to have
dialogue with the international community. There was work
in Sweden. In fact, we had talked with the plant manager
from Oskarshamn about the possibility of getting -- they
re-cabled the plant when they basically rebuilt it.
We've had some dialogue with them about trying to
get aged cable samples for testing. We've had dialogue with
the Germans, fairly active dialogue with the Germans, as
well as with the French. The Japanese have ongoing programs
and we've had some dialogue with them.
I think, in general, the observations are
consistent. I wouldn't want to say that there have been
strict comparison tests.
There was a coordinated research program with IAEA
that Bob Lofaro was involved in looking at this. So there's
been a fair bit of international dialogue. The staff does
have access to, I guess, most of that information.
DR. UHRIG: Are they carrying out any of the kind
of analysis that you would -- a risk analysis for which this
type of data would be sought?
MR. MAYFIELD: I guess that I don't know. I'm not
aware of any of those studies, unless Mark is.
DR. AGGARWAL: If I may, different country have
different approaches. When you look at Germans, they don't
go in PRA. What they do, they put sample, what they call
cable deposits, and then they do the testing every five
years and then they determine whether they have level of
confidence or not.
When you talk to the French, their LOCA profiles
are somewhat lower than ours and they qualify the cable
using IEEE standard.
Sweden, we have all the reports you describe in
our office here and we have looked at them. We have visited
Sweden several times and our conclusion is theirs are
totally consistent with ours.
IAEA program, which Mike just mentioned, they are
coming out with a guidance document, a guide and the
recommendation is this is the international feeling that
these cables you qualify one time, you simply can't walk
away from these cables and do nothing. They want to see
some kind of condition monitoring be done.
IAEA is going to come soon, the guide will be
printed there recommending condition monitoring in certain
areas.
DR. UHRIG: And yet we have 70 percent of our
plants grandfathered to where no additional monitoring or
any additional work is being done.
DR. AGGARWAL: Actually, 100 percent plants are of
that view. We do not require any kind of monitoring. If
you turn over to my next slide --
MR. MAYFIELD: If I could -- I want to come to
that. It gets to be an important distinction, I think,
between what we require versus what the industry may or may
not be doing.
DR. UHRIG: Some of them are doing something and
some are not.
MR. MAYFIELD: I think that's an important
distinction. We are not in a position today to tell you
exactly what they are or are not doing. We know that some
have aging management programs, if, for no other reason or
for no other examples than the Calvert Cliffs and Oconee
plants in their license renewal applications.
So we know that there are programs in the
industry. We can't tell you definitively how widespread
they are or how effective they are, but I would not want to
leave the committee or the public with the impression that
nothing else is being done just because it's not required.
DR. UHRIG: But you do have resident inspectors in
every one of the plants.
MR. MAYFIELD: Yes, sir.
DR. UHRIG: Do they undertake to report this type
of thing?
MR. MAYFIELD: No, they do not.
DR. UHRIG: They do not.
MR. MAYFIELD: It's not part of their regular
inspection program.
DR. UHRIG: It's not part of their responsibility.
MR. MAYFIELD: No, sir.
DR. POWERS: Can we come back to the risk
significance? Give me a feel for what happens in a sequence
where instrument aging becomes -- loss of instrumentation
becomes important? I can imagine it in a long-term station
blackout for a boiler. What PWR sequences get me in trouble
here?
MR. CUNNINGHAM: You can take -- I guess I'd start
with LOCAs, LOCAs where you have some -- if you have some
unfortunate interplay between the location of the pipe
break, the type of cable that might be in that vicinity, and
the function that that cable is associated with, that if you
could first compromise the function by not causing actuation
or something like that.
DR. POWERS: Sure, I understand. But what am I
trying to actuate?
MR. CUNNINGHAM: Again, it depends on the
different types of plants, what type of equipment could be
important, but, again, small LOCAs, you'd be interested in
the relationship of cables that might actuate high pressure
injection, for example, or perhaps, more importantly, cause
signals to either the automatic signals or the signals in
the control room on that equipment to be either wrong or
confusing or that sort of thing.
MR. SIEBER: But the valves and the pumps are in
the aux building.
MR. CUNNINGHAM: Yes.
MR. SIEBER: What you're really trying to do is
trip the plant right away and usually you can see that on a
loss of pressure before the containment environment gets so
bad that it will wipe out the cables. Then the other thing
you have is you get a phase A and B containment isolation,
and almost all the isolation valves are air operated valves
with solenoids, but they fail closed if it were to short out
or the fuse would blow.
But you're talking a hundred of those that have to
operate probably. And beyond that, everything else is
outside containment, except you probably -- if everything
went, you couldn't monitor pressurizer level, pressurizer
pressure, reactor pressure, steam generator levels.
You would be able to monitor pressure because the
isolation valve for the steam generators is outside
containment.
DR. POWERS: What you are concerned about is an
increasing fraction of accidents ordinarily terminated
benignly that go on to core melt, is what you're concerned
about.
MR. CUNNINGHAM: That's right. That's right.
MR. DUDLEY: This is Noel Dudley. Just to put it
in another perspective, each step in your emergency
operating plan is based on an indication, it's better
supplied by one of these cables.
MR. SIEBER: That's right. It's either a go or a
no-go to the next step.
DR. WALLIS: So something like pressurizer level,
if it was malfunctioning, gave a wrong indication, might
really upset the operator's response.
DR. AGGARWAL: Yes, that's correct.
MR. CUNNINGHAM: Either change it or delay it or
something like that, and that's a big aspect of the risk
part of it that we don't have a good handle for.
MR. SIEBER: But that's practiced in the simulator
a lot. That's part of operator training to simulate failed
instrumentation and have the operator detect it and decide
what to do. You've got several different level instruments.
MR. CUNNINGHAM: Just to be clear, we're not
suggesting that this is a risk issue. We're saying it could
be under a certain set of conditions.
MR. SIEBER: That would be hard.
MR. CUNNINGHAM: We're kind of caught and not able
to go much further without a substantial increase in the
amount of information that we have.
DR. POWERS: It's not going to increase the number
of initiators and it's not going to change those accidents
that we think go to core melt now, which is increasing the
number, the sequences to go to core melt, that's all.
MR. SIEBER: A homemade success path.
DR. POWERS: Yes.
MR. HACKETT: Let's go ahead and see if we can
move to our summary here, of what is obviously not the
prettiest of stories, let's say. I think that's been
covered in significant depth here on some of the points Dr.
Wallis made and I think they're very well taken and there's
going to be more work to be done, suffice it to say.
But I think the first bullet we really covered,
the regulations, as Dr. Aggarwal emphasized, don't require
periodic inspection or monitoring and in some cases, we
could get into a discussion of that being problematic, given
access and so on and some of the techniques that are
available, maybe push the state-of-the-art.
But there are periodic spec surveillance tests
that do evaluate operability that these systems are
energized, but, of course, they're not of the capability to
evaluate aging and the degraded state of the cables and any
pending failures that might result.
The operating experience that Dr. Aggarwal
summarized does indicate some service-induced degradation
that was attributable to the elevated temperature conditions
and some moisture intrusion, but I think, also, as was
pointed out, I'm sure there are plenty of other conditions
that were not captured here.
In addition, there are things, cable failures
related to other than specifically aging degradation.
Maybe the most important point so far is that the
-- and I think Mike Mayfield summarized this pretty well.
The research results from both Sandia and Brookhaven and
Wylie, which are separated by at least five or six years,
and maybe emphasize some slightly different aspects of the
qualification, both show that not all these cable types
would do this qualified life type of function, even for 40
years, considering a LOCA.
And at 60 years, this becomes more problematic,
but, again, this is assuming that you're operating these
things at the rated temperatures.
The caveat at the bottom says, of course, that the
realistic conditions are in-plant service generally less
severe than the parameters used for these qualifications.
DR. WALLIS: How about 20 years? You said 40 and
60, but I thought many of your significant number failed in
the 20-year.
MR. HACKETT: That's a good point.
DR. WALLIS: That's before license renewal.
MR. HACKETT: That's correct. I guess we're
jumping ahead a bit. There's at least one of the summary
pieces here that covers that.
DR. WALLIS: Okay.
MR. HACKETT: The next slide, Mark, again, just
summarized pretty well. The risk studies can give you
relatively high CDF values, again, conditioned on some of
these situations, are all the cables failed during a LOCA.
The bottom line is I think we're in need of
further research data to support a sufficiently low failure
rate or to get the CDF values to an acceptably low level.
We are planning currently on Mark's branch doing
some work in this area over the next year or two and we have
budgeted for that in our budget cycle.
The risk studies, Mark, again, emphasized, are not
as definitive because of lack of detailed information. One
of the things that was discussed is that may be one of our
-- at least one of the options under consideration by the
NRC staff is to -- how to go about obtaining that
information, if it's decided that that's needed to be done.
That's one of the issues under consideration.
More data is obviously needed to really more
carefully define the risk significance.
I think it's pretty clear, one of the messages, I think, we
wanted to leave with the committee, it's pretty clear that
these research data, at least as limited as they are, as Dr.
Wallis pointed out, I think this is definitely one of those
cases and experimental endeavors where you would certainly
like to have more data, but these LOCA tests are expensive,
we are resource limited.
There is certainly a strong indication there that
something will need to be done for aging management for
license renewal, some way this particular phenomenon will
need to be addressed for the license renewal period.
And at the bottom, that the staff is, at this
point, still considering resolution options, I think, one of
which was discussed here for the current license term, but
the staff has not reached consensus on the resolution
approach.
I guess with that, that summarizes what we had --
again, as Mike mentioned, when we came down here, we wanted
to be able to tell you more about this resolution path and
hopefully we'll be able to be back soon with a story for you
there.
But at this point in time, that kind of summarizes
what we had planned to say and be glad to take any
questions.
DR. UHRIG: What other options are you looking at,
other than a reporting requirement?
MR. HACKETT: Did you want to take a crack at
that, Mike?
MR. MAYFIELD: We literally have been discussing
things that range from no action, other than supporting what
would -- what has gone in the license renewal world, up
through the potential for a rule change to 50.49.
I would not want to suggest to you that any one of
those is a favored approach at this point, but it's that
full range of options that have been being discussed, ways
to solicit more information, generic communications, all of
those regulatory devices are being discussed and we are just
not in a position today to say yes, this is -- even the
direction we're leaning.
But virtually all of those options have been
discussed in one form or another.
DR. UHRIG: I gather from reading the document
here that the Commission has faced this issue in the past
and backed away from it, in the sense that they did not
require backfitting.
MR. MAYFIELD: They did not require backfitting
when they imposed 50.49.
DR. UHRIG: And the rule came along and they had
not really imposed and the grandfathered plants.
MR. MAYFIELD: And when we saw the data, the
Sandia data in the early '90s, the information notices went
out, the staff went to the task action plan. A number of
things were evaluated at that point.
We came to these 43 issues through a series of
public meetings we held with the industry. We got a lot of
additional information from the industry, additional test
reports that helped us resolve most of those 43 issues.
We got it down to six that were being addressed
through the specific research program.
So as we've laid this information out for the
industry, they very quickly provided significant additional
information that we were able to evaluate and resolve most
of those 43 issues.
DR. UHRIG: Could you identify the six issues that
remain?
MR. MAYFIELD: Not off the top of my head, but I'm
betting Satish can.
DR. AGGARWAL: I can. Bob, one more point I
wanted to clarify before I go to the six issues.
Number one, when we were writing the rule and reg
guide in 1981-82 timeframe, we did not have any data which
indicated failure. The issue before the Commission of
backfitting was not whether you require condition monitoring
or not.
The issue at that time was that in '79 and 1980,
in response to the Commission memorandum and order, plants
were required either to meet the guidelines or NUREG-0588
requirements.
What the Commission ruled or decided at that time,
that billions of dollars were spent in qualification, so
they don't want to go back and ask the operating plants to
do more requalification based on new requirement.
But there were no issues at that time that
condition monitoring required or not required. When we
wrote the Reg Guide 1.89, the staff was concerned and we did
put in our discussion portion that consideration should be
given for some kind of monitoring, condition monitoring,
because we cannot really bring any test lab a cable at end
of life at 40 years.
So we must do something. If you read very
carefully IEEE standard 323.74, it also encourages people to
do something. The Commission was informed by a SECY paper
that the staff feels, at that time, like '97, that we think
some kind of condition monitoring should be done, and this
is a part of the SECY paper.
Now, let me turn over to your question about those
six issues.
Ed pointed out to you that we have 43 issues and
our plan is to look only at the six.
The first issue we wanted to look at was basically
a comparison between the accelerated aging techniques and
also what we find in real life, how these properties
compare.
The staff has concluded that the accelerated aging
techniques are acceptable.
Issue number two was the people are using
different numbers for the activation energy and we want to
go back and look at what the limitations are, what the
industry has done is right or wrong. Our conclusion is
that, yes, they are using reasonable correct assumptions and
the staff is satisfied.
The number three was that do the multiconductor
cables have different failure mechanisms as compared to
single failure, single conductor, and if so, what are those
mechanisms.
The staff has concluded that really there is
nothing that measure, but we have problem with the Okonite
cable which I told you earlier. They were bounded cable,
which was the next item that do they have a failure
mechanism, bounded cable compared to unbounded cable.
In multiconductor or single conductor cable, we
did not find any major concern. But under the fourth issue,
we did have the problem with Okonite and we are dealing with
that problem or rather NRR with industry as to what to do.
DR. UHRIG: That's a one-brand problem.
DR. AGGARWAL: Right. We have found the next
issue is what we looked at or referred to as any condition
monitoring for determining cable condition, and our
conclusion is yes, there are several techniques they are
promising. They cannot be used for all kind of insulation,
including visual examination, walk-downs during refueling
and so on.
And the last issue was do we know if condition
monitoring cable, which not only tells you the health of the
cable, but can it predict whether the cable will survive a
LOCA, and the answer is we really do not have a definitive
answer.
In a nutshell, we found there were problems with
one of the cables. We found problems with the Samuel Moore
in test No. 4, where two out of three cables failed, 60
years, pointed out to you several times, almost every cable
had problem, either low value or simply failed.
In a nutshell, this is the story.
MR. HACKETT: Let me add one clarification, based
on Dr. Uhrig's question. Bullet number four, basically, the
bottom line there is that problem exists with any cable that
has Hypalon bonded to EPR, or could exist, depending on what
the aging parameters are.
So I guess I don't want to leave anyone with the
misconception that it may be only an Okonite problem. With
Hypalon EPR, that specific polymeric combination could lead
to problems, depending on what the aging parameters are.
MR. LEITCH: I'm concerned about the confusion
that could be caused in an operating environment post-LOCA
and it seems to -- by misleading or failed cables.
It seems to me that there is a rather small subset
of instrumentation in the control room that was especially
-- that was environmentally qualified, and I'm remembering
Reg Guide 1.33, I don't know if that's right number or not.
MR. SIEBER: 1.97.
MR. LEITCH: 1.97, was it?
DR. AGGARWAL: Post accident monitoring.
MR. LEITCH: Post-accident monitoring, yes. It
would seem to me that that subset of equipment would take a
higher priority in resolution than some of the other
instrumentation.
MR. SIEBER: But I don't recall that it had
anything special in terms of cable. It had independent
circuits and independent power supplies and you were able to
disconnect the control room part of that. And that goes to
the shutdown panel and you can control the plant from the
shutdown panel, but I don't recall anything special about
the cabling.
MR. LEITCH: I'm not saying there's anything
necessarily special about it, but a failure in that cable
would be particularly distressing to the operators. They
have been trained that that particular subset of
instrumentation is dependable post-accident.
DR. AGGARWAL: Your observation is correct. In
the rule, all post-accident monitoring equipment, cables
included, must be qualified to the requirement.
MR. SIEBER: Right.
DR. AGGARWAL: Because those instruments are not
treated safety-related, but the rule require if they are
variable one and variable two, as defined in the reg guide,
must be qualified.
So this is the regulation.
DR. UHRIG: Any other questions from members?
DR. POWERS: It seems to me one of the issues that
you need to chase down maybe a little more is the footnote
on one of your slides that says the service conditions are a
lot less hazardous, because you are working with an
Arrhenius expression, which is an exponential expression.
That could be a very, very significant change.
I'm not sure what your activation energies are here. I
myself have never been satisfactorily persuaded that
Arrhenius works that well for low temperatures, but I
understand the database that exists for polymers here.
But it seems to me that I would look and see if
perhaps there may be not too demanding in the aging
condition that we're presuming takes place at the beginning,
because I think you've got orders of magnitude there and
everything else is factors of two.
DR. WALLIS: I'm wondering about the real cause of
failures of cables. You seem to treat it as a chemical
phenomenon, but if you take a cable and bend it, put this
stuff under stress, then that changes the failure mechanism,
doesn't it? There are all kinds of things that contribute
to failure of a cable.
It's kinked or it's bent or it's under some other
kind of load or something, as well as the chemistry.
DR. AGGARWAL: That is true. Again, we were not
trying to reinvent the wheel and find some new mechanism.
DR. WALLIS: The real question is when do these
things fail or might they fail in a real plant, not some
abstract test somewhere.
DR. AGGARWAL: True.
DR. POWERS: I mean, what happens in the polymers
is that you start getting cross-linking and you lose
plasticizers. A combination of things makes them --
MR. SIEBER: Brittle.
DR. POWERS: Brittle, yes. Real brittle, as a
matter of fact. So anything that happens to it subsequently
cause fractures.
MR. SIEBER: In fact, one of the questions I
wanted to ask is when you wrap the wire around a mandrill to
test it, it then is subjected to the radiation and the LOCA
spray, do you unwrap it to do the voltage test? Because if
you do, it's going to crack to smithereens. Or do you test
it on the mandrill?
MR. LOFARO: We tested it on the mandrill. WE did
not --
MR. SIEBER: So you didn't even get that stress.
DR. POWERS: I'm also reminded by Dr. Kress that
Hypalon evolves hydrochloric acid as it ages, which has its
own ramifications.
DR. UHRIG: Any other comments? Any from the
staff? Industry?
If not, Mr. Chairman, back you.
DR. POWERS: Thank you. I'm going to recess us
for ten minutes. We can bring the transcription to a close
at this point.
[Whereupon, at 4:51 p.m., the recorded portion of
the meeting was concluded.]