465th Meeting - September 1, 1999
UNITED STATES OF AMERICA
NUCLEAR REGULATORY COMMISSION
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
***
465TH ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
U.S. Nuclear Regulatory Commission
11545 Rockville Pike
Room T-2B3
White Flint Building 2
Rockville, Maryland
Wednesday, September 1, 1999
The committee met, pursuant to notice, at 8:30 a.m.
MEMBERS PRESENT:
DANA A. POWERS, ACRS, Chairman
GEORGE APOSTOLAKIS, ACRS, Vice-Chairman
THOMAS S. KRESS, ACRS Member
MARIO BONACA, ACRS Member
JOHN J. BARTON, ACRS Member
ROBERT E. UHRIG, ACRS Member
WILLIAM J. SHACK, ACRS Member
JOHN D. SIEBER, ACRS Member
ROBERT L. SEALE, ACRS Member
P R O C E E D I N G S
[8:30 a.m.]
DR. POWERS: The meeting will now come to order.
This is the first day of the 465th meeting of the Advisory
Committee on Reactor Safeguards. During today's meeting the committee
will consider the following: Safety Evaluation Report related to the
Oconee nuclear power plant license renewal application; proposed
resolution of Generic Safety Issue-145 actions to reduce common cause
failures; proposed final source term rule and associated draft
Regulatory Guide and Standard Review Plan; proposed revision to
Regulatory Guide 1.78 evaluating the habitability of a nuclear power
plant control room during a postulated hazardous chemical release; ACRS
plans for revising the RETRAN-3D Thermal-Hydraulic Transient Analysis
Code; proposed ACRS reports.
The meeting is being conducted in accordance with the
provisions of the Federal Advisory Committee Act. Dr. John T. Larkins
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.
A transcript of portions of the meeting is being kept and it
is requested that the speakers use one of the microphones, identify
themselves, and speak with sufficient clarity and volume so that they
can be readily heard.
There are some of us that might think there is no justice in
the world, but it has turned out that there is at least a little bit. I
am pleased to announce that Professor Apostolakis received the American
Nuclear Society's Tommy Thompson Award at the PSA '99 Conference on
August 24th, 1999. I would like all of you to join with me to
congratulate Professor Apostolakis for receiving this prestigious award.
[Applause.]
DR. POWERS: I also note that Professor Apostolakis is the
second member of the ACRS to receive the Tommy Thompson Award, and
furthermore Tommy Thompson himself was a former member of the ACRS and
the Chairman in 1961, I believe.
This is clearly a banner day and shows that the American
Nuclear Society is an astute and perspicacious organization.
[Laughter.]
DR. SEALE: That's on good days.
DR. POWERS: I will also note that the members have before
them a draft calendar for meetings in the coming year. I would hope
that you could look at that calendar and compare it to your own
obligations so that we can finalize that this evening.
Members also have before them an agenda for the preparation
of reports and the priority we attach them. I will hope that all the
cognizant members can be prepared in approximately the order of priority
for that process.
I will ask now are there any comments that other members
would like to make before we begin today's session?
[No response.]
DR. POWERS: Seeing none, I guess we will turn to the first
item of business, which is the very important process of license
renewal. Jack?
MR. SIEBER: As far as the Oconee report, I must recuse
myself from this presentation to avoid any possible conflict of interest
which might arise from my stock ownership in Duke Capital Corporation.
DR. POWERS: Okay. We will expect you to remain mum, then.
It is, however, perfectly legitimate for you to pose questions as you go
through, but you will otherwise not be involved in that process.
The license renewal process -- Mario, I guess you are the
cognizant member.
DR. BONACA: Yes.
DR. POWERS: And do you want to lead us through this
process?
DR. BONACA: Yes. Thank you, Mr. Chairman. On June 31st
and July 1st on this year, 1999, we met to review the application by
Duke Power for life extension of the Oconee units.
We heard presentations from Duke Power. We also heard
presentations from Babcock & Wilcox, which has developed four topical
reports supporting the application in the area of the vessel and ability
to go to life extension.
We also heard from the Staff. There was a significant
presentation on the SER. Both the application and the SER were thorough
and quite extensive. There are still a number of open issues and
confirmatory items to be addressed. However, notwithstanding those
items, there seems to be significant progress in this application and
therefore we have asked the applicant to come and give us an overview of
the application itself and then for the NRC to give us an overview of
the SER with special emphasis on the open issues and confirmatory items.
With that, I will move on to Mr. Grimes --
DR. POWERS: This will include the discussion of fuses and
things like that?
DR. BONACA: Yes, we specifically have identified the number
of issues and communicated them to the Staff so that they are prepared
to answer those questions regarding issues such as fuses, for example.
DR. POWERS: Good.
DR. BONACA: So we will have, first of all, a presentation
from the licensee, Duke.
MR. ROBISON: Good morning. I am Greg Robison from Duke
Power. I am the Project Manager for the Oconee License Renewal Project.
I want to take just a few minutes this morning to give you an overview
of our application.
With me assisting this morning is our Lead Structural
Engineer, Deb Ramsey and Paul Colaianni is with us also -- he is our
Lead Electrical Engineer.
Just a smidgeon on Oconee. Many of you are very familiar
with Oconee and where it is located. It is in northwestern South
Carolina. We have three units on site, a little over 2500 megawatts.
It is the largest nuclear installation we have on the Duke system.
Construction obviously finished in the early '70s. We are here for
license renewal since our license is expiring in 2013 and 2014 and we
have about 1300 people employed on the site -- that might give you a
feel for the magnitude of the site itself.
We will do two things. We will do a view of the application
and then an overview of the application. Here is the view of the
application. Through the wonders of getting to play with a digital
camera, we took a picture of the application. It made it a whole lot
easier to transport than the volumes themselves.
[Laughter.]
MR. ROBISON: Some of you have seen the application. There
are four volumes to the application. Three of the volumes are
technical, involve the technical matters. The fourth volume involves
the environmental report, and I will overview each of those this
morning.
I thought it would be interesting to give you a little bit
of the organizational criteria of the first three volumes, the technical
volumes associated with the application.
The thought process that went into making this puzzle piece,
this application puzzle piece, fit were we wanted to present the
information required by the rule, address generic safety issues, follow
our discipline results because we designed our answer on a discipline
basis -- mechanical, civil, structural, and electrical. We also wanted
to present the information such that it paralleled the current FSAR we
have for Oconee so that you can see matches between the application
itself and the current licensing information that we have at the plant,
so that was another criteria that we put into developing the
application.
Finally, we wanted to make sure that commitments were clear
in the future, so we created a new chapter to the UFSAR, Chapter 18,
where we hope to house and we propose to draft a formal for the UFSAR,
Chapter 18 to capture the aging management programs, so those are some
of the criteria that we put into the design of the application itself.
I will walk through in a little more detail how we created
the thought process that went into creating the technical information.
I won't dwell long on this. There are a good many notes in the
handouts, and I will just hit the high points of the next few slides.
The integrated plan assessment portion of the application,
we took the regulatory words and really worked to get it down to an
engineering equation, if you will, something that we could be mindful of
continually as we went through the process of putting together the
technical documents that back up the application, as well as the
application itself. And that equation is C plus A plus P plus D equals
R.A.
DR. WALLIS: What are the units of this equation?
[Laughter.]
MR. ROBISON: Time. Lots of time and lots of money.
That's a great question. I'll have to think about that.
[Laughter.]
It did give us a shorthand on being able to ask -- when you
ask a question or when you're in the middle of the technical work, where
are you? Where are you in the development of this work? Are we
describing aging effects and how the aging effects may manifest or
describing the screening and scoping process and the derivement of the
components that it's important to be looked at? Are we talking about
the programs, which is the "P" portion of the equation?
The second area I'll speak just a few minutes on is the time
limited aging analyses.
Let's just walk through the equation at a high level and
show you how it fits into the application. The C is the structures or
components that require aging management review. Again, I say that
we're divided by traditional engineering discipline.
There's a special feature of note in our application. We
separated reactor building containment and the reactor coolant system
out from the remainder of the structures, the mechanical and electrical
items in the plant. We did that on purpose, for two reasons.
Historically they've received additional treatment in our design and our
licensing basis. Certainly they are the focus of safety at our plant.
And additionally we were working with the B&W owners' group on primarily
the reactor coolant system components, so it allowed us to make a match
line between the work that the owners' group was working on and our
application and not get it somehow convoluted with the other materials.
So we could very clearly see how the owners' group work marries into the
Oconee work. So this worked out well, and the results of that are
captured in volume 1, chapter 2 of volume 1 of the application set.
Moving on to the applicable aging effects, here again we
worked to develop a set of information working with the owners' group so
that we could understand the potential aging effects or the range of
aging effects that could be possible given certain material,
environment, and stress conditions.
A little bit of basic material science review. We created
that and then used that as a template or a menu, if you will, to look at
the actual plant conditions at Oconee, the materials of construction,
the environment we had those materials exposed to, to determine the
applicable aging effects set that we needed to make sure we were
managing. That information is captured in the application in chapter 3,
which is in volume 2 of the notebook set.
The third portion of the equation is the P or the
programmatic aspects. Here in our application we defined a series of
attributes for a program so that we could make sure we had a measuring
stick for each of the programs, for each of the programmatic actions we
were using at the plant or we were going to need to use at the plant to
manage aging. We wanted to make sure we had a standard to hold that up
against and find a way to consistently present that information. We did
provide that set of attributes in the application, initially in chapter
4, which is in volume 3, and then each of the programmatic actions after
that, follow that template, follow that format. So again, that's a
little organizational process that we use to help us stay on track and
to make sure we presented information in a consistent fashion.
DR. WALLIS: So looking at your equation again, and really
if any one of these parts that you're talking about is zero, then the
result is zero. So we really ought to multiply these things instead
of --
MR. ROBISON: You are absolutely right. It dawned on us a
year after we developed the equation that it was really a multiplication
equation. So you're right.
DR. WALLIS: You are taking logs.
MR. ROBISON: Yes.
DR. APOSTOLAKIS: It's really the Boolean intersection.
[Laughter.]
MR. ROBISON: The fourth variable here is demonstration.
Here it was important I think through the course of the number of years
we developed this thought process in discussing this with the NRC staff
that we find a way to work the demonstration or the reality check of our
programmatic actions into the application. It wasn't good enough, and
we agree it's not good enough just to assert that you have a program and
look at a notebook on a shelf. You need to pull it off the shelf, look
at the historical results of how that programs been run over time, and
ask yourself is it a learning program, does it have features built into
it, has it been executed at a certain frequency that makes it
justifiable, not just one time 20 years ago, but is it robust enough to
be able to do the job that we're asking it to do. The demonstration
portion of this equation allowed us to ask those questions, and we
presented the results of those findings in volume 3 in chapter 4 with
the programs themselves.
And then the final variable, which I don't have a slide for,
is reasonable assurance, which is what we're after here, reasonable
assurance that we're managing aging of that important set of hardware.
DR. WALLIS: What's the measure of reasonable assurance?
MR. ROBISON: We had a sort of an engineering feel when we
got through the process of what we believed was a reasonable answer. If
it didn't feel right, again, I don't believe there's a prescriptive
criteria for that. If it didn't seem right, it didn't have the fullness
to it, we went back into the --
DR. WALLIS: Do you mean something like 90 percent or 50
percent or, I mean, what --
MR. ROBISON: We asked ourselves does the component set --
is the component set fully described? Is the aging effect characterized
in a manner that industry literature can help us understand the rate the
phenomena occur so that we can look at our programmatic actions and make
sure they're happening at a reasonable frequency? Do we have operating
experience at Oconee that may tell us we've not operated a program at a
correct frequency or perhaps not focused in detail on the correct set of
hardware?
DR. WALLIS: So it's all qualitative?
MR. ROBISON: Very much so; yes, sir.
DR. WALLIS: So it depends very much -- our evaluation
depends then very much on our assessment of your judgment?
MR. ROBISON: Yes. Yes. And many of the programs
themselves have industry standards to back them up, a good portion do.
So I think there's a collective wisdom that comes back to the table.
But in many cases where you don't have rigorous criteria out of a code
book to hold up, it does come down to engineering judgment, and I think
having multiple sets of eyes in this case, the owners' group developed
to where we have a number of utility contract people, then our own folks
at Duke, and then the staff review, it's the combination of all of those
things that help us feel that we've reached reasonable closure on each
of these topics.
And certainly as this is a first-of-a-kind activity, there
will be an opportunity to learn and standardize and perhaps quantify the
measuring criteria for when you solve an equation like this. But the
good thing is I think we went in with our eyes open, not biased to not
go put a program in place or not ask ourselves if there was an important
activity we needed to do. We went in looking for opportunities to put
programmatic actions to make sure the plant stays safe as we move
forward in time. So that was our bias.
DR. BONACA: I have a question regarding this overhead here.
The existing programs, activities to be enhanced actually, you showed in
many cases that they consist of one-time inspection, okay? For the
existing programs that you list under the number of 28 there, if I
understand it, those are actually programs in a broad sense. They
involve many procedures. Or are they simply -- I'm trying to understand
the extent of the content of those 28 programs there. And the reason is
very clear. For the application we had for Calvert Cliffs, there were
on the order of 400 existing programs, and there were a very small
fraction of new programs which were similar to what you had, one-time
inspections. So I'm trying to compare the 350-odd number for Calvert
Cliffs with your 28 presented here.
MR. ROBISON: That's a very good question. The slide we're
looking at is a compilation of 50 programs that we found necessary to
implement or have implemented at Oconee for license renewal. The
difference in the Calvert work and our work is as an engineer I don't
think in terms of multiple procedures to accomplish an activity. I try
to think in a holistic sense of the programmatic solution or making sure
that the system and component set can continue to do its job.
An example in the existing program area where there were
multiple procedures that we grouped under a program was fire protection.
We have a long list of fire-protection procedures that check the
condition of the pumps, the valves, the pipe. I took all of that and
grouped it and called it the fire-protection program with a variety of
attributes and controlled it at a high level to make sure that I didn't
miss something within the framework of that program. I want the whole
program that touches the important hardware that we're looking at here
to work, the workings of the program.
We did look at the procedures, but when we reported the
results, we tried to look at a higher level and group the information.
So that's the difference in the Calvert work and our work. You'll see
Calvert very meticulously looked at each of the procedure sets. Nothing
wrong with that, but they reported it back in the application at the
procedure-set level. We just simply grouped those, because in many
cases that's the way we operate our station. We group that procedure
set up into a higher-level program, and then try to --
DR. APOSTOLAKIS: So the corrective action is a program
then?
MR. ROBISON: The corrective action step is a program
itself; yes.
DR. APOSTOLAKIS: Which will have preventive maintenance,
corrective maintenance, and those things?
MR. ROBISON: Yes, sir. It'll instigate whatever the
appropriate --
DR. APOSTOLAKIS: So if we start counting those, the work
processes under the program, then the number 28 may become 300 or --
DR. SEALE: This is the log again.
MR. ROBISON: Yes.
DR. WALLIS: I have a question. It seems to be kind of
self-assessment. You're inventing a process and inventing the equation
and so on and satisfying yourselves you've done something. But this is
a license renewal application. I would think you'd start out by saying
these are the NRC requirements listed. Those are the NRC criteria that
we have to meet, and this is how we're going to do it. Maybe it's
because these don't exist that you've done it yourself your way. Why
aren't you doing it in the way of meeting something that's already been
laid down by the NRC?
MR. ROBISON: I think your point that this is a really early
or first of a kind is an appropriate comment.
DR. WALLIS: But you start with some objective, to meet some
NRC criteria.
MR. ROBISON: Yes.
DR. WALLIS: Surely they have framework laid out.
MR. ROBISON: They very much do. We have both NRC guidance
materials in both final and draft form as well as industry materials in
final state as of 1995-1996. But we have to understand that in this
initial interpretation we're all going to continue to struggle to get
perspectives on the answer. And so what we wanted to do is create a
process that used the guidance -- certainly we're based on the materials
that are there -- but not limit ourselves to the guidance that's there,
so that as new ideas and new thoughts came along, we were going to be
inclusive of those.
I thing you'll see I have a slide in just a few moments
where I relate the application structure back to the license renewal
rule and the findings, and you can see very much our thought process on
how we were linking it back to the regulations. What you've seen in the
equation here is really our interpretation of what's below the
regulations, how do you take the words in the regulation and go to your
plant and identify the correct hardware, pull the correct material
science results out and pull all of that together in order to meet that
rule requirement.
DR. BONACA: But isn't it true that for existing programs --
existing programs essentially compare with what program you have
implemented against the requirements of the regulation.
MR. ROBISON: Yes, they do.
DR. BONACA: So the judgment you're using is that for -- in
the extended life certain components will be under those programs, which
means that the programs are adequate. I mean, you're still making a
judgment of compliance with current regulation, aren't you? So the
judgment really is whether or not the current program, which is how you
comply with regulation, meets the requirements also during the extended
life or if you need a new program. That's the judgment your making.
MR. ROBISON: We are making both that judgment, and we tried
to be objective when we made that judgment. Because some of the
programs were put in place based on a knowledge and awareness level that
maybe is not as acute as where we are today with understanding the aging
of the plant. So we really did a -- I say reviewed again for the first
time of the existing programs to make sure that, given our material
science understanding of how the plant is aging, that those programs
really are doing the right job, and even in the area where we say the
programs perhaps require enhancement, some of the enhancements here
could be due to the fact that the program's not quite focused at the
right level or depth.
To deal with the aging phenomenon, given the state of the
world, the knowledge of the world back in 1978, or '80 or '82 when that
program was established, it may not focus exactly where we needed it to
focus for the remainder of the six year period.
DR. APOSTOLAKIS: Why, I mean just on my own information,
why was there a need for programs because of what happened?
MR. ROBISON: There are a few components that are receiving
the spotlight on them for really the first time. The reactor internals
package is a good example of where we are putting a new program in
place. We were so focused on the reactor vessel shell for all of these
years, and rightfully so, that when we began to look at functional needs
to operate the reactor, and we looked at the aging of the metals in the
internals, we realized we really didn't have a real good handle on what
was going on with thermal issues and neutron issues. It just sort of
raised itself as an area we need to focus on, it caused the creation of
a new program.
DR. POWERS: I guess that is really surprising. I mean the
history of the last 10 years has focused on reactor internals.
MR. ROBISON: I think the focus has been, my understanding,
and I could stand corrected, would be more on the bolting issues
associated with the internals and our understanding of specific aspects.
What we were trying to do was take, again, a more holistic view of the
entire internals package to make sure if there are plate materials that
are somehow having an aging phenomena that we need to look at out into
the extended period, that perhaps have not been focus of the research of
the last 10 years. We didn't want to miss that. So we established a
program to make sure we could be inclusive of that.
DR. APOSTOLAKIS: So should we go back then to licensees who
do not -- were not requesting a license extension and demand that these
programs that you call new also be implemented there?
MR. ROBISON: I think to the extent that -- you have to
qualify you are specifically staying with the reactor internals. One of
the real struggles there is the difficulty characterizing the aging that
is actually going on. We recognize, understanding material, environment
and stress, something could be going on. There is a little bit of data
coming out of the research world that tells us something may be going on
and, because we can't quite characterize it, we wanted to put a frame
around it and make sure we are consciously looking at that.
Ten years from now, if there is a good understanding of what
is going on, aging-wise, and it looks like the right smart thing to do,
then, yes, I think we would go back and make sure that folks that be in
that condition would make sure that that problem is not occurring or
being managed in their plant.
Today, I can't say that I would recommend that, because one
of the struggles in the new program is to really characterize the aging.
And there are a number of steps that we have laid out in program, both
the B&W Owners Group Program and the Duke Program to characterize the
aging, do some initial inspections, come back and see if we have learned
something new, and then determine what additional activities would need
to be taken. That is going to happen over maybe a 10 year or more
period.
DR. POWERS: I get the impression that you would have felt
benefited by a more aggressive research program in this area.
MR. ROBISON: I don't think I am qualified to address that.
DR. POWERS: Did you see significant gaps in the existing
knowledge?
MR. ROBISON: I saw an issue raised for the first time in my
awareness, I don't know if that is an initial first step or a large gap.
I don't know that I could characterize it. It was just new information
to us that we felt like we needed to be able to work with and manage
over the course of time.
DR. UHRIG: When will these new programs be put into effect?
With the issuance of the new license or some other time?
MR. ROBISON: That point is under discussion right now.
From a regulatory perspective, we saw the need to have the program in
place at the end of the 40 years, that is my regulatory answer. From a
technical standpoint, it may be beneficial to us to go ahead and
instigate them now and let them begin to grow up, so that we have them
in place at that 40 year point. So there is really two answers that are
being discussed.
One of our new programs that we credited is an old, oil
sampling program at our Keowee hydrostation. That program has been in
existence for 30 years, and it is now a matter of just simply
formalizing what we have already done. I don't need to wait till 2013
to formalize that.
For the reactor internals program, as it grows up, you know,
the commitments we were trying to make were to have us at a certain
point in that program at a future state. How that grows up between here
and 2013, we have yet to design.
DR. WALLIS: It is interesting to me that there is always an
emphasis on programs and activities rather than on results.
DR. APOSTOLAKIS: I think it is because of the rule. The
rule says you should manage aging, right.
DR. WALLIS: It just seems to me that you go through a lot
of motions, and I don't really care what the motions are. It would seem
to me there are certain results to be achieved. It should be definite
and hard.
DR. APOSTOLAKIS: The rule says that you should demonstrate,
as I recall, you are managing aging.
DR. WALLIS: Yes, but demonstrate something substantial.
DR. APOSTOLAKIS: But you are managing aging, and that is
what they are trying to do, demonstrate that they are managing it.
DR. BONACA: Either that, or, in effect, it is not going to
have an impact for the extended life, or that a program is in place to
detect the impact and to address it when it is detected. I mean that is
my understanding of how you have covered that.
One final question I have, and I don't expect that you made
a direct comparison, but maybe the staff can answer that. Going back to
the issue of the metrics here, would you say that the ratio of new
programs to existing programs is comparable just as an order to
magnitude to the one for Calvert Cliffs.
MR. GRIMES: This is Chris Grimes, I am the Chief of the
License Renewal and Standardization Branch. And my reaction to that
question is that I think they are generally comparable.
DR. BONACA: Okay.
MR. GRIMES: Obviously, there is a different counting
structure.
DR. BONACA: Sure.
MR. GRIMES: And we attempted to explain that in the context
of the broader generic issue on credit for existing programs. And we
did not see substantial differences, apart from it is a different plant,
the utility manages this plant a little bit differently, but I would say
they are comparable.
DR. BONACA: Okay. Thank you.
MR. ROBISON: And I will close the slide by making one
comment. A number of the enhancements, as you pointed out, were one
time programs that we felt were needed to characterize aging, if it is
there, with a corrective action loop to continue to investigate it or
programmatically manage it should it be there.
The other is many programs were not learning programs, they
didn't require a documentation step at the end, so that you understood
you had done the inspection but found no problems. The procedure said
do the inspection, but it never told you what to do if there were no
problems, so the next person coming along could not learn from you the
fact that you had not seen anything in the past, so there were document
enhancements. But we also called up --
DR. WALLIS: That is an emphasis on activity rather than
achievement, rather than outcome.
DR. SEALE: In those areas.
DR. WALLIS: Yeah, but this is a concern I have with the way
the agency does business. The outcome is the only thing that matters.
Activities, 10 activities or 100 are irrelevant if they don't achieve
the outcome.
MR. ROBISON: The time-limited aging analysis portion was
the next step in the technical matters in Volumes 1 through 3. Again,
these are boundary conditions on the design. We went through a
multi-tiered step to find these particular items. There is a criteria
in the license renewal rule, a six part criteria to help you to
understand how to define a time-limited aging analysis. It included a
review of exemptions. We did not find any Oconee exemptions that were
time-based. The results are presented in Volume 3.
The environmental area was the fourth volume of the set.
You can see the organization again tried to write it in a style, lacking
any better guidance, that told a story so that there were completeness
in thoughts and we could provide information. I will note here that the
good work the NRC had done with the Generic Environmental Impact
Statement meant that the plant-specific work was simply a complement to
that and added to that. You take the results that the Staff's already
provided us in the Generic Environmental Statement and you supplement it
with your information, and the Staff will write the supplemental
environmental impact statement.
Our environmental report is aimed at providing information
to create that supplement and show a match line between the generic work
and the plant-specific work, and that is what we have done.
DR. WALLIS: Is there anything in the environmental issues
which really is important for license renewal, any issue that matters?
MR. ROBISON: We have really found none that rose to the
level that we had not already been taking care of in our other
environmental work and our commitments to the community -- in
permiting -- we were taking care of for thermal permiting and what-not.
Those were already pretty well being focused on at the plant.
These were the conclusions. I think I'll move on for the
sake of time -- I'll let you read it. The conclusions of our report
were pretty much substantiated in the draft Environmental Impact
Statement that I am sure you have seen for the plant, and I will move
right to the table slide and get back to a previous question.
What you see on this table is a matrix breakdown of the
equation itself for the technical areas. You will see the environmental
report out in the far right column, but for the particular items in the
equation the five focus areas within the report itself and each of the
chapters of the application, so you can see how the equation itself maps
into the application.
In turn, you can see how each of those items maps onto the
regulations down along the lower part of the matrix. Exhibit A just
happens to be the name of the particular tab in the book that captured
all the technical items. You see the regulatory requirement in the next
to the last item for the 54.21 area, so each of the regulatory required
steps we map back onto the application.
DR. APOSTOLAKIS: So what are the numbers in the matrix
there, 2.3?
MR. ROBISON: I'm sorry -- the 2.3, 2.4 are chapter numbers
or section numbers within the application itself. This was my attempt
to show you how each of the five technical focus areas and each of the
equation steps map to each other and it makes it fairly easy.
You will see a progression of thought, for example on
containment, from 2.3 to 3.3 over into Chapter 4. You will see RCS
components 2.4, 3.4 -- again we are trying to create shorthand patterns
that are easy enough to be able to jump section to section, lacking a
better way to organize the application.
DR. WALLIS: Well, if you have, say, electrical components
here is one line, does the demonstration end up with some clear evidence
that electrical components will satisfy the requirements for the next
license period?
MR. ROBISON: Yes. The Electrical Component Aging Review is
done in Section 3.6. In that area we presented information to show that
we were not going to have any electrical components that fall within the
scope of renewal.
DR. WALLIS: You don't have to renew anything?
MR. ROBISON: We did not feel so at the time. Since then we
have had an open item on the SE where we have investigated it further
with the NRC Staff at Oconee and we are now of the opinion that there
are a few limited areas in the plant where cabling could experience --
water's in a trench is one isolated location there and there may be, we
haven't finished resolving the issue, there may be one area where cable
may be exposed to temperature extremes.
The general -- the conclusions we drew in the application
based on field survey, work that we had done through the plant, were
there were going to be no areas that had aging effects that were going
to cause detrimental effects on the hardware, the electrical hardware,
over the 60-year period.
Finally, I will conclude and address any questions. I put
in your package this morning the conclusion and again we put this in the
very front end of the application so that you begin reading the
application with the confidence that the complete application is there
and the findings can be made from the materials that are presented, so
you have seen the view, you have seen the overview, and you get a feel
for the conclusion that is there.
This is the materials that the Staff has not taken and
driven through their SER process, so that just gives you a feel for what
the Staff encountered when they opened the front of the notebooks. With
that, I will conclude. Thank you.
DR. BONACA: Thank you. Any --
MS. RAMSEY: I'm sorry -- Debbie Ramsey, Duke Power. I
would like to address two points that Mr. Wallis made.
The first one is the standards that we used. The equation
that Greg put up there is really a simplified equation of the Staff
guidance and information that is in the rule and I think it showed very
clearly in the previous table he put up the "C" in our equation is equal
to the 54.4 criteria for defining the scope, so there is a very good
correlation between our simplified equation to what was in the rule.
The other point I would like to make about programs is that
the rule does focus on the programs but you have to remember that in
54.21 we are not required just to identify the programs but to
demonstrate that the programs are effective. I think that is where we
focus on the outcome of the programs that you are concerned about,
because we had to demonstrate that we are looking at the correct scope,
that we are looking at the aging effects that are associated with those
components, that frequency and all of that is satisfactory, to identify
the aging effect prior to any kind of loss of intended function, so I
think that we do -- are required by the rule to address your concern
with outcome of the program.
DR. WALLIS: Can I ask my colleague, Dr. Bonaca, something?
DR. BONACA: Yes.
DR. WALLIS: Are we going to get this simply as a
presentation by Duke Power that they have made this submission and then
there is going to be NRC come here and say we have evaluated it and we
have reached some conclusions, or is there some effort by this committee
to look maybe randomly at pieces of this and say is the quality good
enough?
DR. BONACA: That is exactly what the subcommittee has done.
We as a subcommittee assigned portions of the applications, all relevant
portions, to individual members and also portions of the SER. We did
that review and so I think we had quite a thorough review.
DR. SHACK: Nobody mentioned whether there is any program --
there are a number of open issues we sort of left in July. Have any of
those been resolved?
MR. ROBISON: We have not made a formal submittal of
responses but what we have done is there have been a number of topics
that we felt like we needed further discussion with the Staff on so we
provided draft responses to the Staff and we have held public meetings
to work through those responses to make sure we were of the same mind
and there's probably 60-70 percent of those open items, the 43 open
items and 6 confirmatory items, that we have already had some type of
discussion with the Staff on.
We are progressing well to get them all closed.
DR. SHACK: Okay, so you don't see any particular hard spots
there?
MR. ROBISON: I don't feel that there are any hard spots
that we can't resolve.
DR. BONACA: And I understand the presentation from the
Staff is going to focus on the open issues.
MR. GRIMES: Chris Grimes. I would like to point out from a
procedural standpoint that we are going to describe the results of the
Staff's evaluation that led to the initial issuance of the safety
evaluation and based on the feedback that we got during the subcommittee
meeting we are prepared to talk about the results of our review in
specific areas to illustrate the nature of the Staff's review and
particular issues, but Greg is correct that the process will proceed by
continuing dialogue between the NRC Staff and Duke to make sure that
they understand the nature of the open and confirmatory items so that
they can respond by October 15th in a formal way with their position on
those topics and then we will use that information to start preparing a
final Safety Evaluation.
We would intend to come back to the ACRS with an explanation
of how the open and confirmatory items have been resolved, but today we
are going to make a presentation on particular areas that were of
interest to the subcommittee and that we feel illustrate the quality of
the Staff's review and provide a technical explanation for how we
reached conclusions relative to the programs that Oconee will
demonstrably manage aging effects through a period of extended
operation.
DR. KRESS: Did the Oconee IPE play any role in this
evaluation of the license renewal application?
MR. ROBISON: We did not directly use the IPE. Certainly we
have talked to the plant engineering staff and we are looking at many
programs, not just license renewal activities but maintenance rule and
other activities to make sure that we have an integrated set of
programmatic solutions for our plant and as the IPE played in we
understood how the results fit in, but it doesn't have a direct
criteria, if you will, that applied into license renewal.
DR. KRESS: Just out of curiosity, what is the CDF and LERF
for Oconee?
MR. ROBISON: I knew you'd ask that. I don't know. I
didn't bring the staff member. Maybe there is someone else here that
can help with that. I don't walk around with that particular number in
my head. I apologize.
DR. WALLIS: Will the aging of Oconee have any effects on
the CDF?
MR. ROBISON: At this point we think not. When we took at
look at the items that were there that really are high risk items, we
believe that they are not really age-driven type risks, and we couldn't
see a correlation there between those two.
Again, it was not a formal criteria that we investigated but
we did ask that question in a qualitative sense.
DR. BONACA: Irrespective of the CDF, but a complete PRA
would likely identify some components which are safety important that
are not identified by the deterministic process and it seems to me that
the existence of an IPE would offer an opportunity, even if the IPE may
not be as complete as one wishes it to be, you have not looked at all to
IPE?
DR. APOSTOLAKIS: I thought it was the opposite, that the
deterministic approach they are following includes many more components
than the risk-significant ones. In other words, if you used the risk
information you would probably have to do less work.
MR. BARTON: That's true.
DR. BONACA: That is true, but if you remember, for example,
the South Texas application where they identified a lot of components on
the deterministic process which are not safety-significant, but then
they define only a small family of components which are highly
safety-significant but they are not in the deterministic process, so
from a perspective of completeness there may be a number of components
which are not merged under the aging program that are
safety-significant.
DR. KRESS: That was the nature of the question I was trying
to ask.
DR. POWERS: I think both, you are both correct that in some
integrated view it is true that the PRA could focus you activities and
reduce the amount of work time, but it is very likely that the PRA would
find a different subset including things that are not found by the
deterministic process. I think you are both correct.
DR. APOSTOLAKIS: Did the deterministic process limit itself
to safety-related components?
MR. ROBISON: No, it did not.
DR. APOSTOLAKIS: No -- so we don't know.
MR. ROBISON: Right.
DR. APOSTOLAKIS: It is true that South Texas identified I
think about 300 nonsafety related components that deserve to be elevated
to some risk significance level, but they may be part of the group that
the Oconee people are looking into because they didn't use any
classification as safety-related or non-safety-related in identifying
what they have to look at. Right?
DR. BONACA: The issue of scope still is a significant issue
that is open and the Staff will discuss. I understand the Staff looked
at the IPE, so we should hear from them what the experience is because I
think that what is important about it is going forward with the scope
issue regarding what falls under the aging programs for plants in life
extension.
If there is in fact a family that only PRA would identify,
however small that may be, you know, that is an issue.
DR. SEALE: Could I ask, one of the programs that was
already in existence at Oconee before you started on this particular
thing was your implementation of the maintenance rule.
MR. ROBISON: Yes.
DR. SEALE: And I assume that the maintenance rule is
represented by one of the programs or perhaps a subset of two or three
of the so-called programs that went into the list that made up your pie
chart?
MR. ROBISON: The maintenance rule itself really does not
fit into the pie chart, since the maintenance rule in most respects was
the complementary program to handle the dynamic equipment in the plant
and the license renewal efforts were focused on the static equipment, if
I can do a gross characterization.
DR. SEALE: Okay.
MR. ROBISON: You end up having --
DR. SEALE: Fine.
MR. ROBISON: -- having a complementary set. Now the way we
did link them is we created an engineering functions set for the
plant -- systems and structures -- and it was from that common set that
the maintenance rule drew and the license renewal drew --
DR. SEALE: Okay.
MR. ROBISON: -- in order to establish the functional
criteria, so it has a common foundation that exists side-by-side --
DR. SEALE: But let me ask you this. As you then went
through and looked at this overall array, and I agree with you it is
license renewal but it is a little more besides when you talk about what
it takes for the next 20-year period, but when you looked at that
overall array, did you find yourself coming up with some amendments,
extensions, elaborations, whatever, to the things that were covered
under the maintenance rule as well as the things you found for your
aging management programs?
MR. ROBISON: We really did not explicitly find additional
items for maintenance rule. Because of some additional maintenance rule
scoping criteria that are separate from renewal, many more aspects of
the plant for the emergency operating procedures were already drawn into
the maintenance rule and again it is focused on the dynamic equipment.
It had already picked up a good bit of the other equipment, so we really
didn't find license renewal driving any changes to the maintenance rule.
DR. SEALE: Okay.
DR. WALLIS: Can I ask you a very general question on public
perception. The public perception is an engineering system, as it gets
older, is more likely to malfunction. That is the experience of life,
washing machines, you know, no matter what. What is different about
reactors where we are told that it doesn't matter as it gets older?
MR. ROBISON: Well, I think -- that is a nice question. If
you look at your washing machine example, we typically don't go into the
washing machine and look at the hoses and the pumps and the belts and
the windings of things on a periodic basis. We put our clothes in, we
turn the controls and it works. Until it doesn't work, and then we go
complain and we fix it.
In the case of a nuclear power plant, we do go look at the
motors and the pumps and the belts and we --
DR. WALLIS: And we fix them.
MR. ROBISON: And we fix them, or we maintain them at a
certain high level. And if you look at our operating and maintenance
expenses, a significant difference between perhaps our fossil brethren
in the nuclear is that additional maintenance, insight and focus that we
have to keep the plant at a high order, so that, if you will, our
washing machine is maintained at a high level.
DR. WALLIS: So it is more like an airplane.
MR. ROBISON: It is more like an airplane.
DR. WALLIS: It is maintained much more strenuously than the
washing machine.
MR. ROBISON: Yes.
DR. BONACA: Any other questions for Mr. Robison?
[No response.]
DR. BONACA: If not, thank you very much for your
presentation and we will hear now from Mr. Grimes.
MR. GRIMES: I would like to introduce first Joe Sebrosky,
who is the Project Manager for the Oconee license renewal application,
and also Barry Elliot, and John Fair of the staff, and Chris Groton.
The team is still growing. And Alan Heiser.
We are going to start off, because of the broad interest in
the reactor vessel, with a discussion of the staff's review B&W Topical
Report 2251, and we are going to describe aging management programs
relative to the reactor vessel. And then after that presentation, then
we will move into the specific area of interest for the Oconee
application.
DR. SEALE: Mr. Chairman, I wasn't aware of the fact that
ACRS was getting its license renewed.
DR. POWERS: You haven't seen some of the SRMs coming down,
have you?
[Laughter.]
MR. ELLIOT: Good morning, my name is Barry Elliot, I am
with the Materials and Chemical Engineering Branch of NRR. My
discussion today is going to be about our view of the BAW Topical Report
-- B&W Topical Report, BAW-2251, which has to do with the reactor vessel
and our review of license renewal. Helping in this review was Herb
Conrad, Simon Sheng, John Fair and Mohammed Razzaque.
Our review is complete, there were no open items and there
were no confirmatory items. The principal aging management programs for
the reactor vessel are, first, the ASME Code, Section 11, Inservice
Inspection Program, in which critical components are inspected to
determine whether there is cracking, where or loss of bolting integrity.
Another critical program is the Boric Acid Wastage
Surveillance Program, which is used to manage whether there is any
corrosion of the carbon steel reactor pressure vessel on the outside.
A third program is a Technical Specification Leakage Limit
Program where leakage is monitored to determine whether there is loss of
bolting integrity.
And the fourth program is the B&W Owners Group Reactor
Vessel Integrity Program, and this program is used to manage the aging
effect of embrittlement of the reactor vessel, neutron embrittlement.
DR. SHACK: Did they have a program for the Alloy-600 vessel
penetrations?
MR. ELLIOT: Good point. I left that out. ASME Code,
Section 11 -- I just changed my glasses, so now I can see what I wrote
here. The ASME Code, Section 11 program, we found it acceptable except
for one particular area, which was the vessel head penetrations which
are Alloy-600. The ASME Code program needed to be augmented with an
additional inspection of the Alloy-600 vessel head penetrations. Oconee
has been participating in an industry-wide program and has inspected
some of their critical vessel head penetrations in the CRDM nozzles.
Thank you.
DR. SHACK: Barry, just on that, everybody says Alloy-600, I
mean that must be Alloy-600 with an 82-182 weldment.
MR. ELLIOT: I am not sure of how the nozzles are welded to
the --
DR. SHACK: To the vessel.
MR. ELLIOT: To the vessel. But we are looking at the welds
and the adjacent material to see if there is primary water stress
corrosion cracking.
When we reviewed the safety evaluation -- when we made our
safety evaluation, this was a Topical Report and we identified 13
plant-specific renewal applicant action items. At the time we discussed
this with the subcommittee. All the action items had been completed
except for one. The only one that was still open was a flaw evaluation
and that has been completed now. So now all the plant-specific action
items for the reactor vessel are complete.
There are two --
DR. SHACK: What is the nature of that flaw evaluation? Is
that evaluating flaws that they have identified?
MR. ELLIOT: Yes, this is a flaw that they have identified
in the reactor vessel during their Section 11 inspection program. The
flaw is left in the vessel and it was originally evaluated for the 40
year life, and now it needed to be evaluated for a 60 year life, and the
applicant has completed that evaluation.
There are two license renewal issues that are generic
license renewal issues that affect the vessel. One is vessel
surveillance and the other is fatigue of metal components.
DR. WALLIS: Can you tell me about the flaws? Are these
flaws sort of seen and you watch how they grow and you decide they are
not going to grow significantly during their life, or is that the flaws
aren't seen at all?
MR. ELLIOT: No, what happens is the flaws are observed
during inspection and it is part of the Section -- if they see a certain
amount, which in this case one of them did, they are required to be
evaluated and to see if it can remain in the vessel without repair, it
was done. Now, after that is done, the applicant or licensee, in this
case, is required to go back and monitor the flaw growth.
The original evaluation assumes a certain amount of flaw
growth. And then over the next 10 years the licensee goes back and
looks -- in fact, over the next 40 years, the next 60 years, the
licensee will go back and look at those flaws to determine whether the
actual flaw growth used in the evaluation is actually true. Okay.
DR. SHACK: But he does no probabilistic exam of the flaws
that he might have missed?
MR. ELLIOT: No, we don't -- well, it is not --
DR. SHACK: There is no flaw tolerance kind of argument
here?
MR. ELLIOT: There is no flaw tolerance argument that what
he missed is the -- we have a requirement that is being implemented now
to qualify inspection procedures and, as a result, if you pass that
qualification procedure, you are capable of finding, detecting and
monitoring flaws with some high probability.
The vessel surveillance program, I will start off with that.
Our goal here for the license renewal is that every vessel should have a
vessel surveillance program where the irradiation data is sufficient, is
far enough out that it is equivalent to the end of life. At the end of
the license extension period, the neutron fluence of the surveillance
data is equivalent or greater than what is expected for the vessel at
the end of the license renewal term.
The second goal is that if -- and this affects Oconee in
particular, is if the surveillance capsules are removed prior to end of
the license extension period, that the licensee define the critical
environmental characteristics of the reactor vessel, neutron
environmental characteristics and that it maintain those based upon the
surveillance results, based upon when the capsules were removed, and
maintain those, that neutron environment throughout the license
extension period.
In both these cases, the applicant has complied. They are
participating in an integrated surveillance program in which they will
be received at equivalent to end of license, end of life extension
neutron fluences.
The integrated surveillance program for Oconee, the actual
capsules are not irradiated in Oconee, therefore, Oconee doesn't have
any capsules. So they have established a neutron environment which they
must maintain. If they violate that environment, they have to
reestablish the program and tell us how they are going to modify their
embrittlement estimates.
As far as fatigue --
DR. SHACK: On that one again, all their materials are in a
surveillance program similar or are they relying on the Reg. Guide for
some evaluations?
MR. ELLIOT: Okay. I can get into that, it is a lot of
detail. Oconee 1 has plate materials in their beltline, so they have
six axial welds and three circumferential welds that make up. Of that,
there are six different heats of material that were used to fabricate
those beltline welds. It turns out that for B&W fabricated vessels, the
welds are the critical locations. Of the six heats that are in the
Oconee 1 vessel beltline, four of them are being monitored by the
integrated surveillance program. Two of them are relying upon the Reg.
Guide 1.99 Rev. 2 chemistry calculations exclusively.
But I would say that if you have four -- if four of the
heats are following the Reg. Guide, it is a pretty good chance that the
other two are also following the Reg. Guide. And so that you may miss a
few, even though we missed two heats here, there is sufficient data to
confirm that everything is going according to what we think it is
supposed to be.
Oconee 2 and Oconee 3 are -- I am going from memory -- are
forging fabricated beltlines. So they only have circumferential welds.
I think they only have two circumferential welds in the beltline. In
the case of Oconee 2 there are two heats, both of those heats are being
-- are in the surveillance program, and in the case of Oconee 3, there
are three different heats making up the beltline welds. All three of
those heats are in the surveillance program.
So the thing is that in a typical, in a normal vessel, you
are lucky if you get one weld that is in the beltline and you examine
it. In this case we have multiple heats in the beltline and we are
examining the majority of them, because of the integrated surveillance
program.
Any more questions on surveillance?
DR. POWERS: Well, I guess I need to understand better this
concept that if we examine n heats, and find them satisfactory, that the
n plus 2, the plus 2 is also going to be satisfactory with an
examination.
MR. ELLIOT: Yes.
DR. POWERS: I mean what is the basis for confidence that
that is the case?
MR. ELLIOT: What is the basis for the confidence?
DR. POWERS: Yes. The difference is we know the chemistry
of those two other heats, so we know how to predict, you know, what is
going to happen to them. We know the chemistry of the other heats and
we know what is going to happen to them. We can compare the predicted
values for all the other heats that are evaluated, and if the prediction
method works for those heats, then we think they will work for the two
other heats.
MR. ELLIOT: So you are confident that you are working on an
interpolative scheme and not in any extrapolation?
DR. POWERS: No, we are not extrapolating here. One reason
we are not extrapolating is because we are requiring that the actual
surveillance data be exposed to a neutron fluence equivalent to the end
of license, expiration license.
Fatigue of metal --
DR. SHACK: You would typically find that the Reg. Guide
1.99 Rev. 2 predictions for the welds that you are monitoring actually
gives you conservative predictions, wouldn't that be the case?
MR. ELLIOT: Yes. In fact, we do that all -- that is our
main job is vessel embrittlement, is we constantly get from all PWRs and
BWRs, we get surveillance reports and we compare the test results to the
currently prescribed method for determining embrittlement. And I would
say 99 percent of the time the actual data conforms with what the
embrittlement equation is predicting.
Fatigue of metal components is an issue that is concerned
with the impact of environmental fatigue on the usage factor. The
licensee has done an analysis based on the environmental factors
described in NUREG/CR-6335. The staff has determined that the B&W
Owners Group has adequately addressed GSI 190 regarding environment
fatigue of reactor vessel components and the fatigue of the Oconee
reactor vessel will be managed during the period of extended operation.
That is an open issue then only in the sense that you
haven't gotten the final report or --
MR. ELLIOT: No, that is not an open issue. We have
completed the fatigue of the reactor vessel. So all the issues with
respect to the reactor vessel as far as fatigue are complete.
As far as time-limited aging analyses, fatigue of components
is a time-limited aging analysis; pressurized thermal shock, which is
related to embrittlement, is a time-limited aging analysis. Part of
that, the embrittlement is dependent upon the material composition and
the neutron fluence, and as part of this review, the applicant submitted
a report on neutron fluence, Topical Report BAW-2241P. We reviewed it
and approved it.
Another time-limited aging analysis is the effect of
embrittlement on Charpy Upper Shelf Energy. A topical report was
submitted on this, and staff has reviewed it. This was an extension of
the previous topical report, which had done for 40 years. This just
modified it for 60 years.
A fourth time-limited aging analysis was the growth of
intergranual separations in low alloy steel forgings heat-affected zones
under stainless steel weld deposit cladding. This is an analysis as it
describes of the growth of these separations during the 60-year life of
the vessel, and also the impact of embrittlement to determine that these
separations will not grow to a size which will affect the integrity of
the vessel.
DR. SHACK: Recently the French did a reexamination of some
of those cracks and sort of discovered they were bigger than they
thought they were.
MR. ELLIOT: Our analysis assumes that they're much
bigger -- and they started out -- the analysis we originally -- when
they were first discovered they were only one-tenth of an inch. Our
analysis started with -- assumed that the crack not only was -- it was
.165, which is larger than a tenth of an inch. But also we assumed that
it went right through the clad. So we assumed both that it was --
actually it isn't our analysis, the applicant's analysis -- assumed like
.365 inches in depth, which is much, much -- three times larger than
originally assumed. And then we grew the crack from there based upon
the operating cycles of the vessel.
DR. SHACK: So they are no longer embedded cracks, they are
surface cracks in this analysis.
MR. ELLIOT: In the analysis there, they are surface cracks.
But in reality it doesn't represent that at all. That's just a very
conservative analysis.
DR. WALLIS: Well, what was the purpose of picking a size of
crack which was much bigger than it really was?
MR. ELLIOT: We wanted to have -- you know, there could be a
flaw that's larger than a tenth of an inch, so we want to make sure we
encompass everything that's possible here.
DR. WALLIS: Did it also analyze the real tenth of an inch?
MR. ELLIOT: You could analyze a tenth of an inch also.
DR. WALLIS: Did you? It seems to make sense.
MR. ELLIOT: The applicant chose to make a larger floor, and
we agreed that it was conservative and appropriate to do that.
DR. WALLIS: I'd like to know how conservative it is. I
mean, is the tenth of an inch very different from the .165? You have a
gut feeling it is, but it sort of makes sense to analyze it to find out
how different it is. Maybe it's a factor of a million or something. I
don't know. Nice to know.
DR. SEALE: Actually there are two elements of conservatism.
I would imagine that the assumptions that go into the evaluation
methodology are tilted or skewed or whatever the word is so that if you
had a tenth -- if you started with a tenth of an inch, you'd still be
conservative in the behavior of a tenth-of-an-inch crack. You then
assume .165. There's another conservatism that comes from that.
MR. ELLIOT: These are just separations at the boundary.
DR. SEALE: Yes.
MR. ELLIOT: That's all they are. They aren't totally
cracks. We're assuming that, and we assumed in the analysis not only
did they -- they grow right through the clad. And that's not true.
That's the biggest assumption we make -- they made.
DR. SEALE: Well, it would be worthwhile to know if a
significant fraction of your claimed conservatism is in the basic set of
assumptions that go into the crack-growth analysis or are those
conservatisms in the size of the initial crack that you assumed to begin
the analysis.
MR. ELLIOT: I think we're doing both of those.
DR. SEALE: I didn't say that. I said it would be
interesting to know how much is -- or at least on the size, what the
magnitude of that conservatism is, and I think that's what Dr. Wallis is
driving at.
DR. WALLIS: Can you supply that information?
MR. ELLIOT: I don't have that right here. We have to do
the analysis ourselves, or --
DR. WALLIS: We hear a lot about conservatism and margins
and things like that in these presentations, and it would be very useful
if they could be quantified so we know how close we are to stepping over
some margin of safety or whether we've got an enormous margin.
MR. ELLIOT: Well, let me just say this. The ASME -- we
have an enormous margin in this case for the flaws, I'll tell you that,
and the reason I say that is not only do we when we describe it -- we
made these flaws cracks and they aren't, and we made them go through the
wall, but we also put additional criteria on if the ASME code requires
margins on that, that's the square root of 10 safety margin above and
beyond that. So we're talking margins on top of margins here to assure
that the vessel is not cracked.
DR. WALLIS: You think of that as virtuous. Now someone who
is saying let's be realistic and let's not be overconservative might
review that as not virtuous at all, to have something which is
conservative by maybe a factor of 1,000 or something.
MR. ELLIOT: If we were asking them to repair those vessels
based upon these flaws or to do additional inspections based upon these
flaws, that would be a, you know, they would have a legitimate complaint
because of all the margins that are here. But we're not asking them to
do anything here. We're just asking them to evaluate the integrity of
the vessel. And they've done it. They've evaluated the integrity of
the vessel, and it's acceptable with these flaws in them.
DR. WALLIS: But I think there's a different audience there.
There's the public or even people like us looking in saying yes, your
conclusion is we don't have to make them do anything.
MR. GRIMES: This is Chris --
DR. WALLIS: But the other need is if you could sort of show
the sort of materials community that, you know, when you do the
realistic analysis how that compares with the one you -- and that would
give us more perspective on what kind of regulations you're enforcing.
Because maybe your regulations are far too conservative.
MR. GRIMES: This is Chris Grimes. I'd like to point out
that this is an area where Barry is addressing an explanation about how
the current licensing practice relates to a conclusion relative to
extending the license period for 20 years. We do not nor did we attempt
to change that practice for the purpose of developing a license renewal
conclusion that attempted to segregate out the degree of margin in the
calculations that we rely upon today within the existing regulatory
framework, nor did we conclude that there was a need to do any
additional analysis or a change in the practice simply for the purpose
of developing a license renewal decision.
DR. WALLIS: Well, I understand, but do you understand my
point of view? What you're doing is you're doing this for internal NRC
purposes. There's no other audience than your own folks making a
licensing decision.
MR. GRIMES: I'm not sure that I agree, because I believe
that we're doing this from the standpoint of the ASME community and the
plant operators who have an established practice that we rely on as an
existing program that manages an aging effect or set of aging effects
associated with the integrity of the reactor vessel. And I believe that
that established practice is something that, like we explained to the
Commission, we're relying on the credit for that existing program as a
practice that would continue on into the period of extended operation.
MR. ELLIOT: This is my final slide in the presentation, and
I talked about the master integrated surveillance program. It has data
both from Oconee and also the other applicants who are participating
in -- other participants in the integrated program, as well as
supplementary capsules.
The last issue affecting the reactor vessel is the PTS
analysis, and initially Oconee's -- the limiting weld for all three
Oconee vessels is the Oconee 2 upper shelf to lower shelf
circumferential weld. When we received the original report, BAW-2251,
when we did our evaluation of this weld, it turns out its reference
material properties turn out to be higher than the PTS screening
criteria by four degrees.
We pointed this out in our safety evaluation, and the
applicant has made a plant-specific evaluation using more up-to-date
neutron fluence evaluation as well as chemistry results, and has
determined that the actual circumferential weld in this vessel is not
above the screening criteria. Its RT PTS value was 297. We have
confirmed by looking at surveillance data that this is a conservative
estimate for the RT PTS value for this weld.
DR. WALLIS: Well, in this case you don't make an absurdly
conservative assumption, you're actually so close to some limit that you
do something different.
MR. ELLIOT: Now what we did, we do exactly what we do with
everybody when we review a PTS evaluation. We look at the surveillance
data and we determine whether or not, based on the surveillance data,
the PTS evaluation is adequate. We did that here. We looked at --
since they have a substantial amount of surveillance data, we looked at
it, and we determined that the methodology that is used in determining
this number is conservative.
DR. WALLIS: Well, you may well be doing the right thing.
It's just that I'm struck by the difference between the approaches in
the two cases.
MR. ELLIOT: Well, I just want to say that this methodology
has conservatism built into it.
DR. WALLIS: A different kind of conservatism.
MR. ELLIOT: We have --
DR. WALLIS: You're not assuming something is the way it
isn't in order to be conservative. You're doing a different kind of
conservatism.
DR. SHACK: He makes an extremely conservative screening
limit, and then he calculates precisely whether he's above or below --
DR. WALLIS: That's very -- yes.
DR. SEALE: Yes.
MR. ELLIOT: Anyway, that's the end of my presentation --
DR. SEALE: Could I ask a related question? There's a
generic issue on fatigue that's been laying out there for an awful long
time, and I was wondering, we've been told we're going to get a chance
to look at that sometime in the near future. I was wondering what the
status of that was.
MR. ELLIOT: That's John Fair's area.
[Laughter.]
DR. SEALE: Well, I saw John flinch when I started.
MR. FAIR: Thank you, Barry. It's not exactly my area.
It's Research's program, and I believe that they're working hard to get
it done by the end of the year. But I don't want to speak for them.
DR. SEALE: That sounds familiar.
DR. WALLIS: Well, does it need to be resolved by the end of
the year?
MR. SEBROSKY: Well, we have an open item on GSI-190 that
will come to Dr. Wallis, and we'll discuss how we intend to resolve that
on a plant-specific basis.
DR. WALLIS: How long has GSI-190 been in existence?
MR. SEBROSKY: I don't know.
MR. WESSMAN: Yes, this is Dick Wessman from the staff. I
think GSI-190 itself has been in existence for maybe two or three years,
but of course it has precursors, as you know, going back to GSI-166, and
was it 157?
MR. FAIR: 166 is the precursor.
MR. WESSMAN: But I think we should let Research describe
when they're ready to come to the Committee and talk about the program
that is ongoing there, I think John and Joe will address the aspects
specifically related to license renewal and how the applicant and the
staff are addressing this issue in the light of a GSI that is not fully
resolved.
MR. SEBROSKY: My name is Joe Sebrosky, and I am the Oconee
license renewal project manager in NRR, and before we get into the
slides, just a high level on what I'll be presenting.
We were asked to present high-level differences between the
Oconee and Calvert applications, and also unique plant configurations
that affected the review, and I'll have a slide on that.
The packet that I gave you also contains a synopsis of all
43 open items and six confirmatory items. Because we are limited by
time, we are going to focus that, as Dr. Shack had mentioned earlier,
into the ones that are rising to the level of management attention
today.
I don't know October 15th what that list is going to be, but
I can tell you today what our list is, and in that list contains some of
the things that we were asked to look at. Specifically, the ACRS wanted
us to talk about scoping. We have an open item on scoping. They also
asked us to talk about void swelling, and we will talk about that a
little bit later. So the list that I am going to go through of our
management items today encompasses some of the items that you wanted.
And the list that I will talk about is scoping, complex
assemblies, form and content of the FSAR. I am then going to turn it
over to Barry to talk about the open items associated with our Section
3.4, that is where void swelling is, and 3.4 is on the reactor coolant
system. So that is -- those open items are receiving a lot of
management attention now. And we will also discuss the open item
related to fatigue.
There is one item that is not on our management attention
radar screen right now, but we will talk about the reactor coolant pump
oil collection system. We have an open item associated with that.
And in addition to that, I do not have a presentation on two
subject matters that the committee had expressed an interest in, and
that is on fuses and boroflex. We don't have a presentation on that,
but we have people here to answer questions when we get to that.
So I will go ahead.
MR. GRIMES: Actually, before you begin, Joe, I would like
to point out that Joe's presentation is organized in a way that we are
going to describe matters that relate to our review of the Oconee
application. And as Joe mentioned, to the extent that the committee has
any other questions related to the staff's generic approach to license
renewal, or the practices, or positions that we have sent to NEI
relative to how we intend to apply Part 54, we would be prepared to
discuss those at the conclusion of the presentation on where we stand
with the Oconee review.
And we also have a meeting planned, I believe September
23rd, to continue to pursue the subcommittee's interest in generic
license renewal matters. So there will be an opportunity to continue a
dialogue finally with any of the topics of generic interest that you
have. But we want to try and keep the Oconee specific matters focused
and the we will address any concerns that you have relative to the
generic aspects.
MR. SEBROSKY: As far as the application differences go,
Calvert Cliffs used a vertical approach by system or structure. And by
that, I mean if you look at Calvert's application and you compare it
with Duke's application, the format is totally different. For example,
you would pick a system, reactor coolant system for Calvert, and what
you will do is you will find for that particular system, in one section,
how they scoped it, how they applied the screening process, the aging
effects, the aging management programs and TLAAs all in that section.
That is what I mean by Calvert using a vertical approach.
Oconee, on the other hand, used a horizontal approach, and
their approach is more similar to the SER formats that we have for both
Oconee and Calvert Cliffs. And, specifically, if you look at their
Chapter 2, they have identification of structures and components subject
to an AMR. Chapter 3 talks about aging effects. Chapter 4 of their
application talks about aging management programs and activities. And
Chapter 5 is TLAAs.
Oconee, also, to contrast the differences, referenced
several topical reports and the topical reports are meant to be written
in such a way that they can be relied on by another B&W licensee. And
the topical reports that Duke relied on for Oconee are one on RCS
piping, pressurizer reactor vessel, reactor vessel internals and fluence
methodology. And Barry gave you the discussion on one of those, the
reactor vessel.
The next bullet on the slide talks about the safety
evaluation reports for Calvert and Oconee have similar formats. That is
the format that the SER and we expect future applications to be
converging towards. And if you look in Chapter 2 of our SER, you will
find the discussion about the scoping and screening process. Chapter 3
has aging effects and aging management programs, and Chapter 4 is the
TLAAs. So there is not a one-on-one correspondence with Duke, but it is
similar.
As far as plant differences go, there is obvious plant
differences in the respect that the Duke plant, Oconee, is a B&W plant
where the Calvert Cliffs is a combustion engineering plant.
I tried in this slide to just give you some high level
differences and unique features at Oconee that the staff had to spend
extra time on. Specifically, those are Oconee has a standby shutdown
facility. And the standby shutdown facility is meant to provide an
alternate means to achieve and maintain hot shutdown conditions
following a fire, sabotage, turbine building flood, station blackout and
tornado missile events. That is a unique facility that did not exist at
Calvert. And the other obvious plant difference was that Oconee relies
on the Keowee Hydroelectric Station as the onsite emergency power
source. So both of those fell into our review.
Now, turning to the open item discussion, on page 3, the
first two bullets involve the open item with scoping. The first bullet
has actually been completed, and that is that Duke was to provide us a
discussion in writing of what they told us in a May 11th meeting. And
where we are at right now with this issue is the second bullet, that
subsequent to the receipt of the above information, the staff will
determine whether additional inspection activities will be needed to
verify the adequacy of the applicant's process for identifying
structures and components that are within the scope of the rule.
We had a meeting with Duke, actually, we went to the site
from August 16th to August 18th and at this point, I would like to turn
it over to Chris Grimes for further discussion on that open item.
MR. GRIMES: We brought this topic up at our monthly
management meeting, which was last Friday, and we have found we are
still overcoming terminology differences and trying to avoid falling
face first into challenging the current licensing basis. We are still
attempting to understand what the current licensing basis is relative to
how Duke has scoped their safety-related systems, structures and
components that are relied upon to perform safety functions for design
basis events.
And therein lies the problem. There is a definition of
current licensing basis in Part 54 that has a very sweeping and broad
approach to what is a design basis event, and Duke has been maintaining
the license basis for Oconee with a different perspective on design
basis events, separate and apart from other commitments and responses to
Generic Letters, for example. And so we need to sort through that in
order to establish a clear understanding of what the current licensing
basis is and how it has been captured for the purpose of license
renewal, and explain that in a way that both -- that there is a shared
understanding between us and Duke on what that scope really is. So we
are going to continue that dialogue until we achieve that shared
understanding.
DR. WALLIS: That sounds a little odd. That you don't know
what the current licensing basis is, yet the plant is licensed.
MR. GRIMES: That's not --
DR. WALLIS: That not unusual?
MR. GRIMES: It is not unusual for the staff to not know all
of the details of a current licensing basis that is maintained in
accordance with the regulatory process that permits the applicant to
change -- excuse me -- permits the licensee to change the design and to
maintain it, and to respond to NRC inquiries and take actions that
result in design changes.
DR. WALLIS: I thought you said there was a difference
between the licensee's interpretation and the agency's interpretation of
the current licensing basis.
MR. GRIMES: No. There is a difference in the way that they
use the term "design basis event" and the way that we understood design
basis event to be a much broader term. We had to overcome that just to
be able to start communicating with each other in terms of what events
are appropriate or applicable for the purpose of scoping.
In addition, it is further confused by -- Duke has
approached this from the standpoint of using one approach for
electrical. It recognizes that there are no clear system boundaries for
electrical like there are for piping systems. There is a different
approach for structural that relies more heavily on a commodity approach
and the fact that structures are, again, like piping systems, they don't
have clear boundaries.
And then there is a mechanical approach, and in the
mechanical approach, there is an extensive reliance on what are design
basis events relative to safety functions. And we are now mucking
around in the details of, well, what is the difference between the
safety function and the current licensing basis and a design basis event
that is explicitly included in the FSAR and analyzed to perform a
particular function. And we are pulling that apart now in such a way as
to clearly understand what the design basis events are for Oconee
relative to what the safety functions are that they rely on in their
current licensing basis. Those two are a little different.
DR. WALLIS: This has nothing to do with license renewal.
MR. GRIMES: This has everything to do with scoping for
license renewal, to identify the passive systems, structures and
components that are relied upon to perform safety functions or
non-safety functions that could affect safety functions and our ability
to make a conclusion on the methodology for scoping without having to
list component-by-component, item-by-item, each and every thing in the
licensing basis that is subjected to an aging management review.
The rule is predicated on a process approach, that in order
for us to make a finding on the process we have to understand enough
about the current licensing basis and design basis events in order to
establish reasonable assurance that the process captures all of the
safety-related systems, structures, and components.
DR. BONACA: The issue is the Part 54 again has a very
sweeping definition in the scope definition, and it doesn't say look at
all your Category 1 components. It doesn't say that.
MR. GRIMES: Right.
DR. BONACA: It says components that perform a safety
function. It doesn't tell you how you are going to go about that. That
is one issue that needs to be clarified.
MR. GRIMES: That's correct. As a matter of fact, we look
back and we see when the rule was constructed in 1991 it was intended to
be a much more sweeping rule, and at that point we put a definition of
current licensing basis in the rule that was very broad. In 1995 when
the scope was narrowed and we ended up with 54.4 defining a set of three
different kinds of design basis events, the safety functions, the
nonsafety functions whose failure could affect safety functions, and the
regulated events, we maintained that definition of CLB, but then we
didn't make a clear distinction about what the expectation was about
narrowing the event scope, if at all, so that is the area now where we
are trying to focus on how big is the population that was intended when
the rule was modified in 1995.
DR. BONACA: And the bottom line is that you want to make
sure that passive components which have to perform a safety function
will be in the aging management program. It's as simple as that.
MR. GRIMES: That's correct.
DR. BONACA: And so that is the important objective.
MR. GRIMES: Yes, and I want to emphasize we are going to
try to do that with a Commission expectation that was clearly
articulated in the Statements of Consideration that the Commission
expected us to be able to do that without compiling this CLB.
If you recall many years ago there was an initiative that
would have caused -- the Commission considered whether all plants should
go back and compile their current licensing basis and reconstitute their
design, and that led to a tremendous number of initiatives and some
special inspection programs to go out and verify that the current
regulatory practice and process was maintaining the current licensing
basis in a satisfactory way without requiring compilation of that
information simply for the purposes of communicating it to the NRC and
the public.
MR. SEBROSKY: The next issue I would like to talk about is
represented by the second bullet here -- on page 4 -- and that is
passive, long-lived skid-mounted equipment are excluded from an aging
management review.
What the Oconee site has in the standby shutdown facility is
a diesel generator and this example is best represented by looking at
that diesel generator.
On a high level, that diesel generator is considered to be
within the scope of the license renewal rule, but there is a screening
process that is then done after you determine the systems within scope
and active-passive determination.
Duke considers the diesel generator to be screened out for
purposes of license renewal, and the Staff has no objection to the
diesel generator. It says right in the rule that diesel generators are
outside the scope of the license renewal rule.
What the Staff is objecting to, though, is the definition of
diesel generator. Duke considers when the diesel generator was supplied
it was supplied on a skid. Any of that equipment that was supplied with
the diesel generator on that skid is part of the diesel generator and
does not have to be looked at to determine if an aging management review
needs to be done.
The Staff doesn't agree with that and specifically in our
SER we mention that there's components on that skid such as the diesel
engine jacket, water heat exchanger, and portions of the diesel fuel oil
system and starting air system that would we believe fall within the
scope of the rule, so we have not seen an answer to this open item yet
but it is something that we expect because of the generic ramifications
of it that we expect it will be elevated to management -- so that is the
issue on complex assemblies.
I'll skip ahead a little bit here. The next issue that
right now is receiving management attention for Calvert, and we expect
that it will also receive management attention for Oconee, is
represented by the first bullet on page 6, and that is form and content
of the FSAR supplement.
I will give you a high level description and then I will
again turn it over to Chris Grimes for discussion.
What this open item represents at a high level is how the
Staff is going to capture the commitments for posterity as far as the
license renewal application goes. The license renewal application
itself becomes a historical document at the time we give them a license.
It is not something that gets updated.
The rule requires though when a license renewal application
is submitted that the applicants supply the Staff with an FSAR
supplement and the form and content of what rises to the threshold of
what will go into that FSAR is what this open item is about.
I'll ask Chris if he wants to add anything to that.
MR. GRIMES: Yes. I would like to point out that the
desired outcome in this instance is to make clear what the Staff's
expectation is relative to what additional things will be managed under
50.59 as part of the continuation of the current licensing basis into
the future, and so our end-game for this issue is to establish an
approach that will ensure that the specific details that the Staff
relied upon in commitments and explanations of program attributes for
the purpose of license renewal will be appropriately captured in the
regulatory process.
MR. SEBROSKY: As far as the next portion, I would like to
turn it over to Barry. The next couple of slides start on page 9 and
we'd like to talk about the reactor coolant system aging effects open
items.
There was one open item that we have that the ACRS had
expressed some interest in, and that was on void swelling. I will go
ahead and turn it over to Barry.
MR. ELLIOT: Before I start out, I would just like to give
you our engineering view of how we did this evaluation.
It comes back to the discussion before a little bit. We
look at the existing program. We look at what we think are the
mechanisms or effects that affect each component, and then we decide
whether or not the existing program or the proposed program is adequate
for managing that aging effect.
Now for the reactor coolant system, outside of the reactor
vessel there are three effects that we were concerned about that we
thought existing programs may not be adequate and we needed more
discussion with the applicant, in this case Oconee and also Calvert
Cliffs.
The aging effects are: the effect of irradiation and how it
affects the material properties such as void swelling, fracture
toughness, irradiation of system stress corrosion, cracking of
internals. Many of these things have to do with that. We have an
irradiation effect that the existing program may not be adequate.
The second effect would be thermal embrittlement of cast
stainless steel. This is an aging effect that reduces the fracture
toughness of a particular component called cast stainless steel.
The third aging effect which would affect materials in the
reactor coolant system is primary water stress corrosion cracking of
Alloy-600.
Then we have to look to determine whether the existing
programs are adequate for those mechanisms and aging effects, so I will
go through each one of these.
The first one is the pressurizer spray head which is a cast
stainless steel component, and it will have a reduction in fracture
toughness, so we have begun discussion with the applicant on how they
are going to inspect this and evaluate it since it has a reduction in
fracture toughness due to the thermal embrittlement.
The next issue is void swelling. This is an irradiation
issue for the internals. There is a lot of literature on this subject.
Some of the literature says void swelling is a problem for PWR
internals. Some of the literature says void swelling is not an issue.
At the moment --
DR. WALLIS: Excuse me, what voids are these voids?
MR. ELLIOT: This is a void in the internal material crystal
structure.
DR. WALLIS: The void is what? What is in the void?
MR. ELLIOT: It's hydrogen expands the lattice structure and
causes --
DR. WALLIS: Gas.
MR. ELLIOT: Yes, hydrogen gas.
MR. HISER: This is Alan Hiser of NRR. It is mainly helium
bubbles that form.
DR. WALLIS: They don't diffuse out?
MR. HISER: No.
DR. WALLIS: Doesn't close off.
MR. HISER: No, they are constrained within the lattice.
They nucleate and then grow during additional irradiation. The main
concern here is that the growth of these voids can swell the overall
dimensions of the structure and in particular what we are concerned
about with the internals is either loss of dimensionality through, say,
tight fit components or maybe constriction of float levels that are
important.
DR. WALLIS: Can these bubbles link up in some way?
MR. HISER: I don't know that they so much link up. It is
just that the growth of them, just overall, distorts the structure.
DR. BONACA: And the phenomenon is also very much
temperature-dependent, is it?
MR. HISER: Yes, that is correct. The main parameters of
importance are the flux temperature and also the accumulated fluents.
DR. BONACA: If I understand the issue, it is that for
temperatures in PWR is they are borderline insofar as there's periods
that are applicable.
MR. HISER: Yes. The industry citations indicate that void
swelling may be on the order of 4 percent or 14 percent at end-of-life
PWR conditions. That is a fairly broad range and part of the problem is
that the definitive data at PWR conditions doesn't exist so there's a
lot of extrapolation from breeder reactor conditions and how one does
the extrapolations and what assumptions one makes drive whether you get
a low estimate or a high estimate.
DR. WALLIS: Did you say there was a 14 percent of what?
MR. HISER: That would be a change in dimension. If you
have a dimension that is one inch, it would become 1.14 inches.
DR. WALLIS: A piece of steel becomes 14 percent bigger?
MR. HISER: That is correct.
DR. WALLIS: I would think this might have quite a few
effects.
MR. HISER: That is what we are concerned about.
DR. WALLIS: If it becomes bigger in all directions --
becomes longer?
DR. SEALE: Three dimensional.
MR. ELLIOT: Yes.
DR. SHACK: Yes. For fast reactors it is very dramatic,
where there is no argument over the existence of the effect. As Allen
said, the real question is at these temperatures how important is it.
MR. ELLIOT: Our approach to handling this issue is to
look -- is to find out where the critical locations are on the
internals.
DR. WALLIS: If any piece of my house became 14 percent
longer, the effect would be dramatic.
DR. BONACA: Drop in price too.
MR. BARTON: Real estate values would go up --
DR. SEALE: -- 14 percent.
DR. WALLIS: A door that becomes 14 percent bigger, it's
sometimes embarrassing.
MR. HISER: Well, my understanding is there are problems --
I am not sure if it is in the breeder environment with fuel rod growth.
DR. BONACA: Yes.
MR. HISER: And, you know, it's just along those same lines
but just fuel rods you are able to remove and replace --
MR. ELLIOT: That's right.
MR. HISER: -- but the internals generally are not replaced
and that is a real concern that we have.
DR. SHACK: Well, in the breeder you design for the fact
that it is going to get bigger and that really wasn't a design
consideration.
DR. BONACA: As well in the fuel assemblies. Typically you
have spread because you have growth, but isn't it something that
inspections would cure insofar as looking --
MR. HISER: At the current time, the only ASME required
inspections on internals are a VT-3, which is basically a fairly gross
visual inspection. If we are looking for, say, growth of a one-inch
component by 140 thousandths of an inch, I don't think that that is
observable.
DR. BONACA: Okay.
MR. HISER: Sort of the frame of mind that we were on this
would be to try to locate some critical dimensions and maybe monitor
those through the standard ASME inspection process.
DR. WALLIS: Well, I am really puzzled. I must be naive.
You have fuel elements and things that slide in and out through spacers
and all kinds of devices. Do those spacers actually swell by this
amount?
MR. HISER: They could.
DR. WALLIS: They could?
MR. HISER: One of the discussions --
DR. WALLIS: Then you'd have things rattling around
tremendously when it's new or it's jammed in there when it's old?
MR. HISER: That is some of the response we have had from
the applicants is that things like rod insertions would indicate whether
there is a problem.
Our concern is that if one needs to insert rods and maybe
they went in properly the previous time, but now you have got an
accident condition and you need them to insert and you have had
additional void swelling and now they don't insert. That is not a good
condition.
DR. WALLIS: That would seem to be an embarrassing
situation.
MR. HISER: It could be more than embarrassing.
MR. GRIMES: Yes. Dr. Wallis, that is really at the heart
of this issue is the fact that, you know, the industry now is
approaching this from the standpoint of they are not convinced that the
cost and personnel exposure associated with trying to inspect
dimensional changes inside the reactor is justified by the data that
suggests that void swelling might or could occur.
So we're at a point where we have to confront this with is
the data sufficiently persuasive to require particular inspection
requirements, and if so, what can be the appropriate amount without
being unnecessarily burdensome.
DR. BONACA: We read a report in which -- again I believe it
was supported by EPRI -- in which a recommendation was made for further
research in this area. It was not clear at all in the report on what
research and who would be performing this research. Do you have any
comment regarding that?
MR. HISER: The industry has put together a group called the
NPR, Materials Reliability Project -- I guess it's MRP, sounds better.
And they're tackling a lot of issues, basically the ones that Barry
mentioned, neutron embrittlement of stainless steel, void swelling is
one of the issues they're looking at. This is an industrywide program.
They also are obtaining materials from overseas, and a lot
of collaboration overseas. What they're basically -- my understanding
is what they're trying to do is assess the effects of things like void
swelling. Since there is limited data applicable to PWR conditions,
they want to try to get as much of the material and information that's
available and assess the significance of the issue.
DR. WALLIS: Void swelling is progressive with age, is it?
MR. HISER: That's correct.
DR. WALLIS: So there must come some time in the life of
these things when you cannot insert the control rods.
MR. HISER: That may be 60 years. It may be 120 years.
DR. WALLIS: You seem to be very vague about it. I would
like to be reassured that it isn't going to happen.
MR. HISER: So would I.
MR. ELLIOT: But that's the point of having a research
program, to look into issues like that, and that's what the program --
DR. WALLIS: The program doesn't give the results always.
You need the results now, don't you?
MR. ELLIOT: No, we don't need the results now.
DR. WALLIS: Relicense and then there will be some method
for monitoring this?
MR. ELLIOT: The fast breeder program determined that for
the fast breeder, which is a much higher temperature than we receive,
the incipient point for a significant amount of void swelling was like 4
times 10 to the 22nd neutrons per centimeter squared. That's way out in
the life of this plant. And it's not now, it's many, many, many years
from now before they ever reach that point. So we have time to research
this issue.
MR. WESSMAN: Let me add to Barry's comments briefly -- this
is Dick Wessman again from the staff -- I think to keep the operational
aspects in perspective we need to remember that things like the control
rods are subject to tech spec testing requirements and the BWRs have a
periodicity of moving them in and out, and there are required scram
times and this sort of things for them all. And so I think, you know,
if there is an unexpected situation where the rods aren't going to work
properly, we've got enough surveillance history and enough requirements
to give this high level of confidence in the operating reactors.
I think I should also point out some of this work by the
MRPs is relatively in its infancy in our interaction with them. I think
we've only been really working with them and NEI on several issues for
the last year or so. And so, you know, specific work on void swelling
may still be some time in the future. But yes, there is work on it, and
there is attention being given to it.
MR. SIEBER: It would seem to me that for insertion of
control rods the dimensions there were related to the thimbles and the
fuel and the control rod guide tubes as opposed to the structure of the
internals. So that's probably not a big issue. On the other hand, an
issue that is important is the bolting that holds it all together. But
a small dimensional --
MR. ELLIOT: That's true. We agree with that 100 percent.
That was one of the critical locations is about the bolt. That's part
of the research program to see that in fact the void swelling and high
neutron fluence, how it affects those.
MR. SIEBER: And it seemed to me that some bolting has
failed in the past for other reasons. On the other hand, the bolting is
susceptible to a lot of different things. This is just another effect
on bolting that can cause an early failure.
MR. HISER: Recent inspections at three plants in the
U.S. -- Farley, Point Beach, and Ginna -- have in at least two of the
cases indicated that some of the baffle-former bolts are totally failed.
MR. SIEBER: Yes, that's what I would suspect, as an
aging --
MR. HISER: These may be the kinds of precursors that we
need to look to. The failure analysis of the bolts is ongoing at
Westinghouse, but that's something that we're closely monitoring to see
whether it's more IASCC-oriented or if there's some other sort of
mechanisms going on.
MR. SIEBER: But that investigation has been going on for
years. Is that not correct?
MR. HISER: The baffle-former bolts themselves of U.S.
plants is -- oh, this is within about the last 12 months. But the MRP
again has taken over some of that evaluation, and there is more
extensive experience in Europe that they are tapping into.
DR. BONACA: What MRP? What does it stand for? I'm sorry,
I lost --
MR. HISER: Materials Reliability Project.
DR. BONACA: And that's sponsored by --
MR. HISER: I think it's EPRI is I believe maybe the
contracting agent for that. Greg may be able to provide a little more
specific information.
DR. SHACK: But it's basically for PWRs.
MR. ELLIOT: Right.
MR. ROBISON: Greg Robinson, Duke. Yes, it is EPRI. EPRI's
the contracting agent for that, and it's an attempt by the utilities to
pull a variety of materials programs together, the overall intent, so
that we don't have materials programs going on in the Westinghouse
owners' group, the B&W owners' group, EPRI, other contractors. It's an
attempt to put some oversight over all of that and focus on topics so
that we can apply industry resources in a focused manner.
DR. WALLIS: Is this a potential area for sort of the
unforeseen event? I mean, we have this tremendous emphasis on design
basis accidents which are sort of scenarios concocted by various
experts. This looks like some situation where the bolts are stressed
and then one pops and then something moves which jams A, which then
prevents the motion of B, leads to some long thing which eventually
becomes serious, which is not in anybody's design basis or previous
analysis. Is this that sort of thing with that sort of potential?
MR. WESSMAN: Dr. Wallis, Dick Wessman again. I guess let
me take a crack at answering that and then perhaps Al can supplement a
little bit.
I think there's been fairly extensive work going on in the
last several years, and Al referred to the three facilities where
there's been inspection of the bolts, and the staff has looked at
topical reports and done safety evaluations for the Westinghouse
community with various postulated failed bolting patterns, and have
determined an allowable quantity of bolts that could fail, and what that
would mean, and, well, the plants still satisfy the accident analysis
and asymmetric LOCA loads and this sort of thing. And it's a fairly
considerable quantity.
For example, at a plant like Point Beach or Ginna there may
be 600 or 700 bolts and the analysis can show a pattern of failed bolts
of maybe 100 or 200 or something like that, a number that's considerably
in excess of what has been seen so far from the domestic inspections.
Coming a little closer to what we're discussing here today,
and that's the Oconee facility, we've also met with the B&W
representatives regarding the postulated baffle bolt issue and failures
in the B&W reactors, and they've provided us some information that leads
us to believe that the configuration is somewhat different and the
situations that were experienced in the foreign reactors or what has
been seen at Farley or Ginna are quite unlikely, at least at this time,
in the B&W facilities. For example, the fluence is different and some
of the materials of construction are different, and the water flow at
one of them is upflow and one of them is downflow, and I don't remember
which, but some things that contribute to a pretty high confidence level
on the B&W plants.
They have not dismissed the issue entirely, and I think
there is still consideration in planning by the B&W community as their
plants get a little older, to make a decision on whether an inspection
is needed. And I think at Calvert the baffle-former plates are welded
construction, and so the bolting issue is not relevant there.
Al, if you have anything else, go ahead and add it.
MR. HISER: No, I think the only thing I would add is that
the B&W owners' group, as Dick pointed out, does have an inspection as
part of their program. I'm not sure, I think the timing is a number of
years off at this point.
DR. BONACA: Okay. And we need to move. We have a little
more than 15 minutes left, so --
MR. ELLIOT: Continuing on, the next issue is the Alloy 600
heater sheath-to-sleeve plate and heater sleeve-to-bundle diaphragm
plate weld. As part of the Alloy 600 program, this was one of the five
most susceptible locations to primary stress corrosion cracking. So we
want to make sure that they're given an adequate inspection.
The next issue is provide an assessment to ensure that the
repaired letdown coolers are operating to preclude failure from thermal
fatigue.
The next issue is a CASS stainless steel issue which for
Oconee affects valve bodies, pump casings, and the pressurizer spray
head, and we're trying to determine based upon the evaluation procedures
in EPRI topical report for cast stainless steel whether materials are
susceptible or not, whether these components need additional inspection.
The next issue is an internals issue, that is, we're
concerned about irradiation-assisted stress corrosion cracking, and look
at nonbolting wrought materials in the internals, which is the most
limiting, where should we be looking in the future for potential
cracking from irradiation-assisted stress corrosion cracking.
The next issue is to provide a plan for managing aging
effects of the baffle-former bolts, also an internals issue.
The one after that is another internals issue, but this is a
synergistic effect of thermal embrittlement and neutron embrittlement,
that is, some of these internal components are also cast stainless
steel, so they get both the thermal embrittlement, a reduction of
fracture toughness, plus the neutron embrittlement fracture toughness,
where these components, how much do these components get as far as
reduction in fracture toughness, do we need some more inspection of
these components.
And then the last one is also internals. It's we want to
look at the pump and valve in-service test program for the in-valve
bodies and returning rings. These have cast stainless steel, and we've
got to look at that program to see whether the inspections in this
program are adequate for irradiation-assisted stress corrosion cracking
and thermal embrittlement.
MR. SEBROSKY: That ends the discussion on the three four
open items.
The next open item that I'd like to talk about is on page
12, and it's the second bullet. It talks about the basis for the
acceptability of the scope of the reactor coolant pump motor oil
collection system inspection.
I'll give you some background on what the open item is, and
if you have any questions, we have a reviewer here to hopefully answer
that.
This open item, the basis for this open item is Duke is
saying that they would inspect one of 12 tanks to characterize the loss
of material from corrosion. The reactor coolant pump oil collection
system, there are four tanks in each unit, for a total of 12 tanks. And
the staff is concerned they would like to have the basis for the one
inspection being bounding for the other 12 tanks. And they're also
concerned that the inspection is focused on the carbon steel tank, the
collection tank, and there's other materials present within that system.
So the second part of that open item has to deal with how the inspection
bounds, what would happen to other materials in that system.
So that is the open item the that staff has on the reactor
coolant pump oil collection system.
DR. BONACA: One question I raised, or we raised during the
subcommittee meeting was regarding the remaining portion of the
collection system which is mounted on the reactor coolant pumps and is
subjected to significant vibration. We heard that that was not
inspected. Now, I am surprised about that because the concern would be
fire. I would expect that that system would be inspected as part of the
fire protection program.
MR. SEBROSKY: Actually, I will turn that -- we understood
that the concern was with the mechanical vibration of the system.
DR. BONACA: Yes.
MR. SEBROSKY: That is not an aging effect that was
identified in the application.
And, Greg Robison, hopefully, you can address the concern
about mechanical vibration not being identified.
MR. ROBISON: Can I borrow your spot for a second?
I thought it would be good to use a schematic, so what I
have done is pull the schematic from our mechanical diagrammatics.
Highlighted in green are what we term "enclosure boxes" that were
installed to capture oil should a line rupture occur. Highlighted in
red are flex hoses that isolate those boxes, or those areas, actually
from the oil tank you see at the bottom of the figure, so there is a
vibration isolator, if you will, between the pump and the oil collection
boxes and the tank.
The one line that you do see, the hard line that you do see
here is the original installation for oil being drained from the tank,
that line is no longer in use and, in fact, is valved off inside the
tank. So the lines that are in question that could contain oil are
isolated in a vibration sense from the tank and the equipment here.
You have to put this in additional perspective, and I really
apologize for the shape of this figure, this is a scaled down version of
a scaled drawing, and you can put those large green boxes in
perspective. This is the outline of the pump here. I should have
darkened that in. The little small green boxes are actually the
collection boxes associated with the oil. And I have been able to
identify, although it is straining my eyes to do so, two of the three
flex hoses. You can see where they are located in association with the
pump, so that we have actually been able to isolate the pump from the
oil collection system via these flex hoses.
DR. BONACA: I still have a question. That is, is the
system inspected periodically, I mean, or are you waiting just to see
some splashing of oil on surfaces before you are going to inspect in
this system?
MR. ROBISON: The aging effects that were identified for the
system are being inspected for, and Joe has alluded to the tank
inspection. We don't believe there is really any aging going on in the
tank. We do know that from a vent valve off the tank here, as you can
see on the diagrammatic, there is an opportunity at this vent valve from
some spray and decontamination washdown to get into the tank, and we
have seen in the tank, some of the tanks, over time, water. Although
the tanks are lined with an oil film in the air space, and they are
going to full of oil when you drain the oil from the pump, which is done
during the refueling outages, the likelihood of any kind of corrosion
occurring is very low.
We wanted to make sure that there really were no problems
going forward, so in order to assure ourselves, we wanted to do the one
time inspection. Beyond that, we didn't identify mechanical vibration
as an issue, you know, so we have to kind of separate the two problems
here, because, literally, we are going to reach 10 to the 7th, 10 to the
8th cycles on pump operation within days. So were the lines to be
challenged by some sort of high cycle problem with endurance limit
challenges, we would have seen that long ago.
DR. BONACA: But those hoses, for example, are they being
inspected?
MR. ROBISON: Oh, yes. Yes, they are. The piping system
here and the enclosure boxes are all Class D, which is part of our
scoping criteria there. Seismic-related, require pressure boundary
integrity. So the piping system and the tank and the external surfaces
certainly get the examinations.
DR. BONACA: Okay. I didn't get a sense. I asked the
question and I didn't get a sense, and I wanted to hear that.
MR. ROBISON: Okay.
MR. GRIMES: Dr. Bonaca, I would like to point out that we
have -- you know, for the sake of trying to accomplish our conclusion
relative to aging management programs, we have concentrated primarily on
looking at particular aging effects and then a series of inspections or
other programs that would look particularly for those aging effects.
We haven't really delved into the details about the extent
to which there are normal plant walkdowns and inspection activities that
check for configuration and conforming conditions.
DR. BONACA: Yes, I understand.
MR. GRIMES: And so there are a lot of things that we might
not be aware of that occur and aren't credited or aren't focused enough
to look at, you know, how the current licensing basis is reflected in an
operability sense.
DR. BONACA: But the reason why this issue came up was
because I am less concerned about corrosion in the tanks and the
consequences of it than I am about vibration of a system which is
mounted on the pump, that may fail and cause fire, because these events
have occurred before. Now, all I wanted to hear is that, in fact, these
programs exist, because I was surprised to hear at a subcommittee
meeting they didn't exist. And that is why I asked that question and
that was the reason for it.
MR. GRIMES: No, and I understand that. You know, if we had
pursued a concern about vibration as a potential aging effect, it could
have been addressed that way, but we didn't delve into that area.
DR. BONACA: Okay. Thank you.
MR. SEBROSKY: The next open item I would like to go to is
on page 14 of your handouts, and it is GSI-190, it is represented by the
third bullet on page 14. And the open item is the time-limited aging
analysis of the reactor coolant system is not adequate to address
fatigue concerns for operation beyond 40 years.
In the SER the staff gave Duke one of three options to
address this open item, and the three options that we gave them were to
develop an aging management program that incorporates a plant-specific
resolution. So, in other words, come up with a plant-specific
resolution for GSI-190. The second option was provide a technical
rationale for the staff on why it could be deferred. And the last
option is to adopt a GSI-190 resolution when it becomes available.
And we had a meeting with Duke on August 25th, and in that
meeting, Duke basically gave us Option Number 2. They gave us a
technical rationale for why they think it could be deferred. And what
the staff said at that meeting was that there is some regulatory
uncertainty associated with that option. If they had given us a
plant-specific resolution for GSI-190 and had done additional analysis,
that the staff would write off on that, and GSI-190 would not be
revisited at Duke.
What Duke said -- we had some feedback back and forth. What
Duke said is they would consider what the staff told them at the August
25th meeting and then tell us by October 15th what option they intend to
pursue.
I will turn it over to you, John, if you want to add
anything.
MR. FAIR: No, there is no more need to add to that, unless
you have some questions.
[No response.]
MR. SEBROSKY: Now, out of the open items and confirmatory
items, that was the last issue that I had as far as being highlighted.
DR. BONACA: I would like to ask just one more issue that
was raised was the issue of fuses, and the fact that the staff agreed
with the applicant and the NEI position that they should be considered
active components and, therefore, excluded from the scope. However, the
staff still expressed concern about potential malfunctions of fuses and
their intent to pursue it potentially as a Generic Safety Issue. I
would like just to hear more about that.
MR. GRIMES: In particular, what more would you like to
hear? We pursued it from the standpoint of -- we wrestled long and hard
with this issue, much like we are still wrestling with the issue of
complex assemblies. What is the right borderline for declaring a
component that needs to be subjected to an aging management review?
And we looked at fuses long and hard and concluded that it
was hard for us to take the fuse out as a component by itself and say
here is a component for which we need to identify aging effects and
establish an aging management program. And so we relied on the language
in the rule that talks about components that perform their functions
with a change of state. But deep in our hearts, it was more a matter of
a fuse is a part of a circuit.
DR. BONACA: Sure.
MR. GRIMES: Which is, you know, more like a piece part.
But we do recognize that, while we haven't seen any experience that
fuses age and therefore are affected -- and their intended function is
affected by particular aging effects, it seems like an area where that
might warrant some more research, and so we are going to -- we know that
there are industry initiatives that are developing now. There are
manufacturer recommendations on, you know, how fuses should be treated
in terms of age. So we would like to sort of see those things come
together, but we didn't view that as a necessary part of a conclusion
for license renewal.
DR. BONACA: Yes, the reason for my bringing it up is that,
first of all, I mean it should be excluded from the rule, it should be
excluded because it is not a concern and not because it is active or
passive. Now, I agree that if it is in fact determined to be active,
then it has to be addressed in a generic sense, and GSI may be the case.
I wasn't surprised that this issue would come up for the first time, at
least from my perspective, at this time, and that the issue has not come
up yet in this industry, because I mean plants are aging anyway, even
within the 40 year life they are licensed to operate today. And if
there was, you know, more information on this issue than anything else,
I think that would be valuable.
MR. PRATO: My name is Bob Prato, I work in License Renewal
and I had a hand in developing that position. The reason it hasn't come
up before is because there is no operating history, and we were just
looking from a theoretical approach, whether or not that this was a
feasible failure mode. As a result, we initiated a generic safety -- we
put the issue into a generic safety process and we asked Research to
take a look at it. Okay.
The reason it hasn't come up is because it is not, in our
minds, a credible failure mode, but we just wanted to be thorough and
make sure that we covered all bases.
MR. GRIMES: Okay. As Bob pointed out, we put a lot of
attention in this as we started to go through and try and categorize
electrical components. There is no operating experience that suggests
that there is a credible aging mechanism here that warrants an aging
management. But there is intuitively a thought that there are probably
aging effects that would apply to the materials. There are practice
issues that IEEE and the fuse manufacturers are kicking around. And so
it is something that we didn't dismiss lightly, but at the same time we
made a conclusion that nor is there anything that we could point to
right now that justifies having to impose a particular requirement for
fuses.
DR. BONACA: Okay. Thank you.
MR. GRIMES: And Jose or Paula, do you want to add anything
to that?
MR. CALVO: No, I think it has been covered very well.
MR. GRIMES: Jose said I covered it very well. Thank you,
Jose.
DR. BONACA: With that, I understand the presentation is
completed. I would like to ask the members if they have any additional
questions for the presenters.
[No response.]
DR. BONACA: There are none, so I would like to thank the
presenters for very information presentations, and the applicant, too.
And with that, I will pass it on to you, Chairman.
MR. MATTHEWS: Dr. Bonaca.
DR. BONACA: Yes.
MR. MATTHEWS: Hi, I am David Matthews, the Director of
Regulatory Improvement Programs for the NRR staff. I wanted to see if
Greg did have any concluding remarks, and if he didn't, I had a couple.
Is it appropriate to offer those at this time?
DR. POWERS: Yes.
MR. MATTHEWS: I just wanted to indicate with regard to
outcomes, which were discussed earlier today, there is an overriding
outcome that the staff is driving towards with regard to this process.
And in regard to our evaluation of methods or methodologies and
individual programs, we are striving towards outcomes that primarily
relate to four areas. And, obviously, the overriding outcome that we
are looking towards is one that will maintain safety through the period
of extended operations. You know, the findings that we make in the
safety area are outlined in detail in the regulations, but, in general,
it is just that, that we maintain safety and have reasonable assurance
that safety will be maintained through the period of extended operation.
We want to do this in such a way that there is public
confidence in the process for us to establish that. And, as you know,
we had a discussion in the past with regard to the amount of credit to
be given for existing programs, and one of the drivers towards the
staff's view on that issue was this issue of ensuring that there was
public confidence in the staff's process for reaching that conclusion
that there was reasonable assurance that we would maintain safety
through this process.
You know part and parcel as we proceed in this new arena, of
course, is that we want to reduce unnecessary regulatory burden when
reaching our conclusions in this regard. So we want to strive to
perform this licensing action in an efficient and effective manner, and
I wanted to thank the Subcommittee and the full Committee for their
efforts in looking at how we're going about that because I think you
have, as well, the desire that we do this in an efficient and effective
manner in the hopes that we'll continue to improve that for subsequent
applicants. So I think this interaction's been very beneficial.
My expectations are primarily, as I understand it, that an
interim letter will be generated with the hope that in the February
timeframe, we'll be able to come back to you following the publication
of the final SER and address these open and confirmatory items that we
described in summary detail, and a few in some depth, leading of course
to a final letter that will permit timely issuance of the recommendation
of the Commission. So with that, again, thank you for your efforts.
MR. GRIMES: I'd like to clarify one point. We haven't
talked yet about the logistic detail about the timing of a conclusion
from the ACRS and publication of a final safety evaluation. And I would
expect that Dr. Bonaca and I could work that out in a way that you can
conclude your review at about the same time we conclude the safety
evaluation, try and time that, you know, optimally.
MR. BONACA: Yeah, I don't see any trouble with that. In
fact, I think that would be desirable that we are informed of your
progress, and then we're not, you know, faced with a long review after
that. So that, I agree with it. I think the meeting we have scheduled
for September 23 should be a good step in the direction of setting up
some process. Yeah. And with that, we intend to write an interim
letter this week.
MR. GRIMES: Thank you.
DR. BONACA: Thank you.
DR. APOSTOLAKIS: Thank you, gentlemen. We'll recess until
ten minutes past 12.
[Recess.]
DR. APOSTOLAKIS: The next issue is the proposed resolution
of generic safety issue 145, Actions to Reduce Common Cause Failures.
I'm the cognizant member, so I'm turning it over to me. And I believe
Mr. Rossi has some opening remarks.
MR. ROSSI: Yes. I'm the director of the Division of
Systems Analysis and Regulatory Effectiveness in the Office of Research.
And I just wanted to make a few comments before Harold VanderMullen goes
through his presentation.
I would urge you to focus on as he goes through the
presentation on whether the NRC and licensees currently have
sufficient robust processes in place to look for and address common
cause failures, because that's the important safety issue associated
with this generic safety issue.
And then also, consider whether additional generic
requirements are needed of licensees and if so, as you know, those would
have to meet the test of being a substantial increase in the overall
protection of public health and safety, and also the cost would need to
be justified by the increased safety that's obtained.
I want to point out that the NRC has a systematic process.
We're looking for accidence sequence precursor events, and part of that
process, they can identify where common cause failures have been a
significant contributor to risk-significant events. The NRC also has a
systematic process for maintaining a common-cause failure database,
which I know has been discussed with the ACRS on a number of occasions.
Now, that systematic process includes a review of all the licensee event
reports and a review of the IMPO equipment performance and information
exchange system, as well as other sources that might indicate where we
have seen or experienced common-cause failures.
It should be noted also that our risk-based analyses of
operating experience, which I know you've been briefed on also, is
showing that the risk from common cause failures is generally consistent
with that shown in licensees' individual plan examinations.
Licensees have, over the past few years, become sensitized
to common cause failure issues, and they have been given the
common-cause failure database and supporting reports for their use. And
finally, the NRC's overall process for review of operating experience
addresses generic common cause failures in accordance with their risk,
as they are found, and also generic insights related to common cause
failures will be a part of the risk-informed inspection process. So I
wanted to make those few comments, and now I will turn things over to
Harold to make a presentation on closure of this issue.
MR. VANDERMOLEN: Thank you, Ernie.
My name, as some of you know, is Harold VanderMolen. I'm in
the Generic Issues Team in the Regulatory Effectiveness Assessment and
Human Factors Branch of the Division of Systems Analysis and Regulatory
Effectiveness in the Office of Research. We're here to talk about
Generic Issue 145. The title of the issue is "Actions to Reduce Common
Cause Failures". This issue goes back a long way, and the title is
fairly obvious, is wide-open.
It's important to go over some of the historical context and
scope of this issue to really understand what it is. We'll take a look
at
DR. APOSTOLAKIS: Are your slides
MR. VANDERMOLEN: Slide 2.
DR. APOSTOLAKIS: Is the projector out of focus, or is this
deliberate?
[Laughter.]
DR. APOSTOLAKIS: It's not a focus problem.
DR. SHACK: No, I don't think that's a focused problem.
MR. VANDERMOLEN: No, we're recycling everything around
here; I think they're using recycled material in our slides now.
[Laughter.]
MR. VANDERMOLEN: I didn't realize that until it tore loose
from the paper. It's it readable from back there?
DR. APOSTOLAKIS: We have a hard copy here. That's all
right.
MR. VANDERMOLEN: Generic Issue 145 actually originated in
the Agency's response to Davis-Bessey loss of all feedwater event back
in 1985. This is one of 37 issues that came directly out of the well,
it's one of 37 issues that came directly from the event. Some of those
issues in turn generated daughter- and granddaughter issues. So there's
quite a few issues generated
DR. APOSTOLAKIS: What, what happened to Davis-Bessey? Is
that the one where the operators were caught not being alert?
MR. VANDERMOLEN: No. This was a loss of all feedwater
event. The steam generators actually dried out.
DR. APOSTOLAKIS: So it was a common cause failure?
MR. VANDERMOLEN: At the time, it was things were pretty
active around here. It was not that long after Three-Mile Island, so
there was a lot of attention paid to it. The plant obviously the core
was saved, but nevertheless a lot of issues were started. You learn as
much from the event as possible.
DR. UHRIG: How long were they out?
MR. VANDERMOLEN: That, I don't know.
DR. UHRIG: I mean, how long was the plant --
MR. VANDERMOLEN: Was it down, you mean?
MR. ROSSI: I don't recall, but they lost the turbine-driven
auxiliary feedwater pumps for a number of reasons. And then they went
down, they had a motor-driven non safety-related one that was down in
the bowels of the plant some place in a blocked room. And they went
down and got that started. But the whole thing progressed quite, quite
rapidly. It was like within 15 minutes to a half-hour, as I recall
and again, this was a long time ago, so, so that's my recollection.
DR. UHRIG: Fourteen years. The B&W plants have a less,
lower volume of water than the others.
MR. ROSSI: That's right.
DR. UHRIG: Do you remember
MR. ROSSI: I don't remember, remember the details of --
DR. UHRIG: It would happen within a reasonable time.
MR. ROSSI: Well, yeah they recovered within --
DR. UHRIG: Thirty minutes?
MR. ROSSI: Yeah, probably faster than that. But again, I
don't remember all the details. I happened to be the leader of the
first Incident Investigation Team, which was associated with that event,
but I still don't remember all the details without going back and
reviewing them. But things progressed very rapidly and the operators
did a lot of things based on their knowledge of the plant, to obtain
stable plants conditions.
MR. VANDERMOLEN: Well, there were a number of things that
happened in that incident that involved common cause failures, which is
why it was such a matter of interest at the time.
Generic Issue 145 was one of them. It was a general issue
intended to cover the overall scope of common cause failures. I'm going
to discuss the scope of it in more detail in a moment. I did want to
point out that there were specific common cause failures with
significant safety implications. These were not treated as part of
Generic Issue 145. They were, some of them became separate generic
issues.
DR. WALLIS: Excuse me. GI-145 was born in 1985?
MR. VANDERMOLEN: Approximately then, yes.
DR. WALLIS: And it's now ready for high school or college?
MR. VANDERMOLEN: At this age, it would be just about --
DR. WALLIS: College?
MR. VANDERMOLEN: A very brilliant issue would be ready for
college at this point, I would think. It's got 37 of its immediate
family.
DR. APOSTOLAKIS: The issue was not created immediately. I
mean --
MR. VANDERMOLEN: No.
DR. APOSTOLAKIS: Yes, but I didn't mean to speculate on the
thought behind the question. This may not be the best issue to point to
the age of the issue because of what would be described now as a major
research effort into common cause, which has ramifications in the PRA
technology of major impacts. So, where some of the GSIs are a bit
horrific for reasons that you don't understand quite, this is not the
one I would tag with that, tar with that brush right now.
MR. VANDERMOLEN: No, I would agree with that, certainly.
Going out a bit, a few others that came out of that
incident, these are just examples -- this is not intended to be a
comprehensive list. There's Generic Issue 125.I.5; it's title is
"Safety Systems Tested in All Conditions Required by DBA [Design Basis
Accidents]". Generic Issue A-17; that one actually dates back to 1978.
It predates this issue by quite a margin system interactions. Generic
Issue 123, which was looking for any deficiencies in our regulations
governing the design basis accidents. The reason I mention these is,
you really do have to keep track of the scope of these issues. There
are a lot of them that are inter-related.
Still other things related to, and significant to, common
cause failures and these are just a plethora of things that have
happened over the years in the Agency. These again, I'm going from the
Davis-Bessey incident now, to the
DR. APOSTOLAKIS: Excuse me, Harold. Excuse me. Have these
other issues been resolved?
MR. VANDERMOLEN: Yes.
DR. APOSTOLAKIS: All three of them?
MR. VANDERMOLEN: All three of them.
DR. APOSTOLAKIS: Okay.
MR. VANDERMOLEN: I can give you some details on them, if
you like.
DR. APOSTOLAKIS: No, that's fine.
MR. VANDERMOLEN: Other things the Agency has done or is
doing relevant to common cause failures include the IPE program more
about that in a moment. This, again, is not an exhaustive list. These
are just a few examples. The station blackout efforts, which require an
alternate AC supply; our work on Atlas, which in essence required a
diverse way of getting the reactor to shut down. A number of generic
issues and information notices, as Ernie mentioned a moment ago, on
specific CCF mechanisms as they were identified. And finally, a rather
major program in the, I guess the former AEOD, developing a CCF database
and methods.
Now let's get back to Generic Issue 145
DR. APOSTOLAKIS: Now, the methods are not really part of
the resolution, are they? I mean, the methods are really to support
PRA?
MR. VANDERMOLEN: The methods actually are part of the
resolution. I'd like to hold off on that question just for a moment to
describe why I say that.
DR. APOSTOLAKIS: Okay.
MR. VANDERMOLEN: The reason I say that it's important to
talk about the scope of these issues is, as we saw a moment ago, we have
a lot of generic issues. It's very important that we do not duplicate
efforts; we can't afford it. It's even more important that we don't
have gaps between our issues and have a safety problem fall in the gap.
Finally, whenever you have any of these programs -- these are not
intended to be open-ended; they are specific jobs we have to get done.
That is, it's just good management practice that when you start
something like this, you have a destination in mind and have some way of
knowing when you're, when you've arrived there.
Fortunately or unfortunately, the early documentation for
Generic Issue 145 did not include a specific description of its scope.
What the early writers did instead was they suggested four potential
solutions. These solutions are not intended to be alternative
solutions, but as all four of them were considered and had it been
justified, more than one of them could have been pursued. These four
together define the actually scope of the issue as we carry it on our
books.
The first area of Generic Issue 145 was to request licensees
to perform a systematic evaluation of common cause failures based on
their IPEs; now, this predates the IPE program. In this possible
action, we would request licensees to -- evidently, you do importance
measure on any identified CCFs and look at them more carefully.
The second bullet --
DR. APOSTOLAKIS: Well, they have done now the IPEs.
MR. VANDERMOLEN: Yes. I'd like to describe that on the
next line.
DR. APOSTOLAKIS: All right. The second time you do this to
me, Harold.
[Laughter.]
MR. VANDERMOLEN: As long as I don't think --
DR. APOSTOLAKIS: -- price list.
MR. VANDERMOLEN: -- the credit card will come.
DR. APOSTOLAKIS: Now, the second bullet, it's data or
dates?
MR. VANDERMOLEN: Well, it actually is --
DR. APOSTOLAKIS: Rates?
MR. VANDERMOLEN: Dates.
DR. APOSTOLAKIS: Dates?
MR. VANDERMOLEN: The idea of the statistical analysis, it's
simple in principle. The mathematics can be a little complicated
depending on whether you're using a shock or a non-shock model. But the
idea here is that you're having single failures all the time, just
random failures. At a certain rate, you expect to find multiple
failures by random chance. Two things may happen "simultaneously."
"Simultaneously" has to be defined a little bit more carefully when
you're actually doing this.
If you observe multiple failures more frequently than you
would expect for the single failure rate, the difference is presumably
the fraction that comes from common cause failures. To do this, you
would look at the dates of failures -- this is actually written up in
the old literature. I stumbled over that term a little bit myself. One
more slide.
The third bullet -- a suggestion was made that we perform a
detailed review of the most important systems. To elaborate on this a
little bit, the intent of this was to actually have licensees pick out
the most important systems based on their PRA or IPE or whatever.
Someone would have to then walk the system down looking for any possible
common cause failures, and address them. And here at the Staff we would
have to review the whole effort. This is quite a program, obviously.
The fourth suggested program from Generic Issue 145 was to
simply develop methods and data to go with them for everyone to use,
basically a large research program. This is why I say that, yes, the
methods development was also included as part of the resolution issue.
Now, let's go on to that next slide that we promised Dr. Apostolakis.
We're going to discuss each one of these in a little bit more detail,
and this time I will not go on to the next slide; I will stop and answer
questions.
When the IPE made systematic evaluation of CCFs, we did
indeed request licensees to consider common cause failures when they did
their calculations. We did review it, too. There have been discussions
with licensees where we felt that they were using beta factors that were
too small. Common cause failures were treated, however, just as any
potential vulnerability. There was not a special effort on CCFs as part
of that program. Of course, now the program is history. There are
still discussions about it, but we are not at the beginning point of it
any longer.
DR. APOSTOLAKIS: Does the -- I don't remember now, but does
the insights report identify the treatment of common cause failures as
something that we not, that was less than perfect?
MR. VANDERMOLEN: I'd like to ask Dr. Flack to address that
one, because he was involved in that program. He had the misfortune of
sitting where I could see him.
DR. FLACK: John Flack, Office of Research. there was --
the report itself, I'm not, it's kind of in the back of my memory. But
there were instances with certain methodologies and treatment of common
cause failure that did show optimistic use of beta factors and multiple
Greek letters. Now those were pursued as part of the review process.
These licensees were brought to task to justify and defend those
positions, and in some cases they had in fact came back with revised
IPEs, which were more consistent with the bulk of them. So, as part of
the process, yes there were bumps and problems, but they were kind of
worked out as part of that process. And we, at the end, when everything
was said and done, there wasn't anything really outstanding that we felt
needed correction or raised a potential backfit concern.
MR. VANDERMOLEN: Any other questions?
DR. APOSTOLAKIS: Were they, did they confirm the finding,
did the IPEs confirm the finding that common cause failures are an
important contributor to system vulnerability?
DR. FLACK: Confirmation-wise, I would say, in retrospect,
it would be consistent with that position, that common cause failures
drive risk at plants. I think you'll end up at that point when the
independent failure rates become a smaller contributor and you get the
point where common cause failures then begin to stick out, but it
becomes very plant-specific. It depends both on the analysis and the
design of the plant.
DR. APOSTOLAKIS: But they don't derive the risk though,
John. Maybe they are contributing to the unavailability. But I
remember a few lists of important contributors to core damage frequency,
and I don't remember common cause failures. I mean, you see things like
non-recovery of offsite power within 50 minutes, or you know, loss of
diesel. I'm not sure they are major contributors to risk. But system
unavailability -- I believe, yeah, that's probably true in most cases
because of the redundance. You've got your square, the round failure
rates from random events, so they go down very quickly.
DR. FLACK: Yes, yes. Right.
MR. VANDERMOLEN: And the second possible program and
statistical analysis, it was just judged to be impractical. To do that
sort of analysis you have to have a lot of data and get the error bands
low, because you're taking the difference between two numbers that have
wide uncertainty bands. And in addition, with the advent of the
maintenance rule, we're really asking for root cause analysis on these
things anyway, which we all believe, I think, would be a better way of
identifying specific means of common cause failures; there's a lot more
information there.
DR. WALLIS: Well, it's not impractical; statistical
analysis is easy. It's just a question of whether your data justified a
conclusion. I don't know that until you've done the analysis. If all
the events turned out to be common cause, this would show up pretty
clearly.
MR. VANDERMOLEN: Oh, definitely. I mean, in principle
it's, it's mathematics --
DR. WALLIS: -- data, which is perfectly data which is
perfectly suitable for statistical analysis. Until you do it, you don't
know.
DR. APOSTOLAKIS: I think what happens here is that most of
the relevant information comes from common cause failures that were not
completed, potential common cause failures. In other words, they look
at one element of a redundant system and they identify a cause -- maybe
it has failed already. And then they find that the same cause has been
active in the other redundant elements, but they have not failed yet.
DR. WALLIS: Yeah, but his method would identify the --
DR. APOSTOLAKIS: The complete cause.
DR. WALLIS: -- complete cause, without knowing what the
cause was.
DR. APOSTOLAKIS: That's right. And for those, you can --
DR. WALLIS: There's a correlation between --
DR. APOSTOLAKIS: And for those --
DR. WALLIS: That would be very useful.
DR. APOSTOLAKIS: But the statistical basis for this
analysis is -- I mean, statistical data are very poor for that.
DR. WALLIS: Well, that's true, but you'd have to, well, I'm
just saying it's impractical.
MR. VANDERMOLEN: We really don't have that data.
DR. APOSTOLAKIS: Well, there are some. There are some, but
--
MR. VANDERMOLEN: There are some.
DR. WALLIS: But it's so easy to do, but it's not
impractical. It's just that whoever tried to do it presumably found
there wasn't enough data. I hope that's it, rather than just discarding
it without any thought.
MR. VANDERMOLEN: Yes. No, it's not discarded without any
thought. I can assure you of that.
The third potential program to perform a detailed review,
which would be quite extensive, expensive, and a bit intrusive, we do
not feel is justified under our current regulatory analysis guidelines.
To require that of licensees, we would have to make a, come to a valid
finding quantitatively that there is a problem that we are likely to
find.
DR. WALLIS: If you can't find common cause failures by
looking at statistical analyses of all the plants, because there's
insufficient data, it's highly unlikely that one plant's gonna find
sufficient data or any conclusions either.
MR. VANDERMOLEN: Probably not. You know, it's just --
DR. WALLIS: So it's a very peculiar plan.
MR. VANDERMOLEN: We would hope that they would not have
that much data.
DR. APOSTOLAKIS: It also depends very much on what, how you
define it, because as Harold says, simultaneous occurrence -- I mean,
what does "simultaneous" mean?
MR. BARTON: They tried to find that NUREG --
DR. APOSTOLAKIS: It has to be some time interval between
the failures. Maybe we could use 50-59 terminology. It should be
minimal but not negligent.
MR. BARTON: I think the NUREG attempts to define that term.
It does talk about what is --
DR. APOSTOLAKIS: Yeah, but they are not really simultaneous
failure now.
MR. BARTON: No.
DR. APOSTOLAKIS: So the question is now what kind of time
interval do you --
DR. WALLIS: Does this exclude common cause failures caused
by operations? TMI could be called a common cause failure as the result
of several operator --
MR. BARTON: It does not exclude human performance. Human
performance is --
DR. WALLIS: That's hard to say that the common cause is
that the operator's responded inadequately. Is that a common cause
failure?
MR. BARTON: No. It's --
DR. WALLIS: It's the same person doing it, but it's not
quite the same thing as common cause.
DR. APOSTOLAKIS: No.
MR. VANDERMOLEN: I think the analysis, if you just looked
at data --
DR. APOSTOLAKIS: The right perspective here is the
following: We are doing systems analysis; we're doing the PRAs, we have
identified certain classes of failures that we model explicitly --
earthquakes are a major cause of dependence. Then we're asking
ourselves, we have this redundant system; we've done the best we could
with the kinds of things we've identified, like the operators forgetting
to reopen a valve and so on. But we know from experience that there are
many other things that can happen that can cause to loss of redundancy
and we are not modeling them explicitly.
So we'll define this new class of failures, which we'll call
common cause failures, that we'll try to put a number on. So
essentially what we're trying to do here is, we're trying to put a lower
bound on the unavailability of the system, because if you go with the
traditional methods, the more redundant elements you add, the lower the
unavailability, then we know that's not true. So that's the right
perspective, I think, for this.
It's not -- I mean dependent failures in PRAs are handled in
many different ways; this is just one small part of it. It's the causes
that are not treated explicitly. They are treated as a class.
Something will happen that will lead to loss of redundancy that I have
not modeled explicitly. That can be environmental causes or maybe human
error -- though there now, there's a problem because we are modeling
human errors separately.
DR. WALLIS: But they're not independent. I mean, the same
person who fails to open some valve during maintenance may fail to open
another valve at the same time.
DR. APOSTOLAKIS: That's right, and that's, that's more the
explicitly, so it's not part of this class. If it's modeled -- like
fives, they're model-explicit so they're not part of this. However,
humidity is not modeled explicitly.
MR. VANDERMOLEN: No, that definitely is not.
DR. APOSTOLAKIS: So it may be part of this.
MR. VANDERMOLEN: Right.
DR. APOSTOLAKIS: But deciding what is part of the class is
sometimes a subjective call. But I think that's the right perspective.
I mean, it would be really misleading to think that dependent failures
in PRA are handled this way; I mean, there are many other things we do
in PRA to handle dependencies, like systems interactions that you
mentioned --
MR. VANDERMOLEN: Right. Right.
DR. APOSTOLAKIS: Fires, earthquakes, and all the human
errors. So this that other class.
MR. VANDERMOLEN: The others are modeled intentionally and
explicitly.
DR. APOSTOLAKIS: Yes, exactly. If you're modeling it
explicitly, it's not here. It's not here.
MR. VANDERMOLEN: Right.
MR. SIEBER: It seems to me though that common cause
failures are generated during plant events, as opposed to random
occurrences from day to day. For example, Davis-Bessey, as I recall,
had a number of MOV failures.
MR. VANDERMOLEN: Yes.
MR. SIEBER: Surry, when they had the pipe rupture and the
fire system went off, caused a lot of electrical problems. We had one
at our plant where both diesels failed to load because there was a
design defect in the system.
MR. VANDERMOLEN: Yes.
MR. SIEBER: You can pick those out when you do your
post-trip analysis as to, you know, I've got four or five of these
things that have failed in this particular event, and that immediately
triggers you into some kind of common cause investigation. That seldom
goes to a database, unless it goes through INPO, or an information
notice, where everybody else will know that there's a susceptibility for
this kind of compound.
MR. BARTON: Except if it ends up in an LER, then it ends up
documented that way.
MR. SIEBER: Right.
MR. BARTON: That's sort of --
DR. APOSTOLAKIS: And that creates another problem for the
analyst here, because typically when a utility, a you just said, Jack,
identifies a cause -- I mean, it's not a class now; it's a specific
cause. It said the diesels failed because of this. They take some
corrective action. So the problem now is, is it legitimate then to take
that failure and put it in the database since corrective action has been
taken? And the question is, how much of that is a failure --
[Laughter.]
MR. SIEBER: That's right.
DR. APOSTOLAKIS: And the assumption is -- I mean, most of
the time, people do put it in the database, and they say, in another
plant maybe they didn't take this action.
MR. VANDERMOLEN: That's right.
DR. APOSTOLAKIS: The original argument of the NRC staff
when they came back to the ATWS issue, was that the actual cause really
is not relevant, but what we saw was the manifestation or a realization
of a failure due to this other class. Next time something else will
happen, so we should include it.
MR. SIEBER: Right.
DR. APOSTOLAKIS: But it is a little conservative to do
that. I mean, there was a fire in five cabinets -- there were five
fires due to the same cause in different cabinets of the plant. They
finally replaced all of them. Is it one fire, is it five fires, or is
zero fires in the database?
DR. WALLIS: Certainly not zero fires.
DR. APOSTOLAKIS: It is not zero, but to say it's five, it's
way too conservative, right?
DR. POWERS: It's just fire and fire progression.
DR. APOSTOLAKIS: Yeah, but if you eliminate them, as you
keep eliminating causes, there should be some impact on the rate. But
that's why they produced four or five volumes -- they had to deal with
that.
[Laughter.]
MR. VANDERMOLEN: As long as the volumes themselves did not
create a fire hazard.
[Laughter.]
MR. BARTON: Maybe that's why the rate trend -- there's a
graph in Maury's office that shows the rate trend's going down.
DR. APOSTOLAKIS: Yes, the rate trend is going down.
MR. VANDERMOLEN: Yes.
DR. APOSTOLAKIS: But I think the main value of all this is
that it really puts a lower bound on the numbers, so we don't come up
with ridiculous numbers that we used to see in the early '70s, like
unavailabilities of hydraulic systems of 10 to the --
MR. VANDERMOLEN: Minus nine.
DR. APOSTOLAKIS: -- minus nine for demand, and major
failure of a plant, 10 to the -41 -- I've seen that, 10 to the -41.
It's crazy.
MR. VANDERMOLEN: I must admit, I've never seen it that low.
DR. POWERS: Oh, yes. Yes.
DR. APOSTOLAKIS: Yes. So it's really putting a lower
bound, making it more real.
MR. VANDERMOLEN: it is, yes. It's routinely put in any
time you do a fault analysis.
DR. APOSTOLAKIS: Well, the age of the earth's crust is
three-ten to the ninth years by the way, so that gives you a
perspective.
DR. WALLIS: Put it in seconds, milliseconds.
[Laughter.]
DR. POWERS: It doesn't help.
[Laughter.]
MR. VANDERMOLEN: Well, we focused on the last solution --
that is, developing methods and data as the only practical
cost-beneficial solution left to do. Let me go on to the next slide --
DR. APOSTOLAKIS: No, don't do that yet.
MR. VANDERMOLEN: No -- really?
DR. APOSTOLAKIS: Provide information -- that's where I have
a problem -- question really, not a problem. I read the letter. Is
that the only information you can provide? The Administrative letter
98-04, dated July 30, 1998?
MR. ROSSI: The only -- well, all of the database and the
reports and the disks that they're based on were sent out, and that's
just to tell people that they're available. But the licensees have been
sent all of the information here in the report.
DR. APOSTOLAKIS: Oh, so each licensee received the --
MR. ROSSI: Each utility received it, right. And the
Administrative letter told them that they were going to get it.
DR. APOSTOLAKIS: Okay, so this is the only action then that
you have taken.
MR. ROSSI: Right.
DR. APOSTOLAKIS: Right.
MR. ROSSI: But it's been widely distributed and it's been
discussed in a number of meetings on PRA also.
DR. APOSTOLAKIS: The thing that perplexes me a little bit
-- being vice-chairman, I learn from the chairman that this work has
become now infamous in the PRA community.
DR. POWERS: Yeah. You can go from perplexed to puzzled.
DR. APOSTALAKIS: Well that's worse?
DR. POWERS: Well I don't know.
[Laughter.]
DR. APOSTOLAKIS: I want to use the stronger statement.
DR. POWERS: Perplexed.
DR. APOSTOLAKIS: Perplexed.
DR. SHACK: Befuddled.
[Laughter.]
MR. BARTON: Totally confused.
DR. APOSTOLAKIS: Yes. I think Harold waits for my
question. And I would like my colleagues with utility experience
perhaps to contribute to this a little bit. When you provide
information to somebody -- that's the age of communication, right? So
what kind of information would make sense and so on. And when I read
the report, I thought that what you have in Section 4 -- I mean, this is
a summary of 6268, Volume 1. You've got some very useful insights, you
know, from analysis of CCF data, in the form of bullets.
"A major contributor to CCF events is programmatic
maintenance practices. A vast majority of CCF events are not due to
multiple failures in response to an operational demand, but result from
the condition of the equipment." And there is an elaboration.
I'm wondering how many licensees would really take the time
to look at the five reports and then spot this and say this is useful.
DR. WALLIS: Well, I'm wondering if this is information. If
you don't have enough information to make a statistical analysis, how
can you reach the solid conclusions that George is leading up to?
DR. APOSTOLAKIS: Because as I say, they didn't look just as
the completed common cause failures; there were partial, potential
common cause failures. So a lot of these things came from those. But,
but wouldn't it be more useful in your communication to them to say,
yes, all these reports and so on, but in particular -- I'm putting now
myself in the shoes of a utility person who has to make a decision.
Here is a number of insights that may help you, you know, focus your
efforts. And I wonder whether John or Jack feel the same way. To just
say, here are 5 reports, use them --
MR. BARTON: The problem I have --
DR. APOSTOLAKIS: -- then they open it up and they see alpha
factors and MGL --
MR. BARTON: The problem I have with, with the
recommendation and the way you propose to close it is that you sign an
Administrative letter. You foldered all this stuff and, you know, what
means does the NRC have to assure that licensees are using this good
stuff for common cause failure analysis in identifying common cause
failures? There's no requirement; you've got a shot at the maintenance
rule, reg. guide, to make it a requirement in there, and you chose not
to there either. Reg. Guide 1.160. So, there is no requirement that
utilities follow this good stuff in here.
MR. SIEBER: That's right.
MR. BARTON: So how do you know that this is really being
done to satisfy the Agency? I may or may not choose to do all this
stuff.
MR. ROSSI: Well, the licensees come to us from time to time
with amendment requests and so forth based on risk, and they support
them with various PRA analyses. And when we review those PRA analyses,
we look to see -- or at least NR looks to see what sort of assumptions
have been made about the common cause failures. So we believe this
information is the best information available on common cause failures
and it's gonna be used. Plus the fact that we look at events on a daily
basis and so forth for things that show up. And we now have Dale
Rasmussen here, who is reaching for the microphone as I speak.
MR. RASMUSON: I'm Dale Rasmuson, Office of Research. In
addition to the volumes that you currently have, we have always had
plans to take the database and go through it and glean out engineering
insights. We are in the process of doing that now.
DR. APOSTOLAKIS: You've already reported some.
MR. RASMUSON: Those were some very general insights. What
we are doing is, we are going through, in a systematic fashion, looking
at these with the first report that we're about ready to release for
internal, you know, for peer review, dealing with diesel generator
insights. It's about, on the order of 100 pages. This is going into
the specific events and looking at them, trying to characterize them,
show trends in these things. And the purpose of these insights reports
-- we plan to have about eleven or twelve volumes on this, of the
information that's on the database to look at it. This will be
available to the licensees.
And also, we are structuring them in a way that it can be
used in the risk-informed inspection so that they have those insights
available to the inspectors and to the utilities.
May I say one other thing here. You know, as we delve in
and look at these events and try to characterize them, you know, there's
not a lot of events that happen the same way. Most of them are very
unique. And so, what we are trying to do is provide that information
out there so that people have it and can at least take it and evaluate
it and evaluate their practices and gain something from the lessons
learned from it.
DR. APOSTOLAKIS: Yes, the issues, the two issues that we
just raised, I don't think question the validity of your insights or
whether you're making an effort to derive those insights. I think what
Mr. Barton said is, you are informing the licensees without having any
means of finding out whether they are taking advantage of all this
monumental effort of the Agency. Mr. Rossi said that now we will start
seeing this as they come back with requests and maybe PRAs -- although
the main effort here is not PRA.
MR. ROSSI: Right.
DR. APOSTOLAKIS: The way I understand it, this is not PRA,
although the project itself has done a lot of things to help the PRA
analyst. The generic issue is not a PRA-related issue. My suggestion
was, can you summarize the information that is relevant to a utility
manager? And I brought up as an example the bullets that you have on
pages 13 and 14 of Volume 1 of the report. And maybe the additional,
more detailed reports that Dale mentioned, and communicate them to the
utilities and say, you know, there are these reports that are available.
In fact, you're sending them to them. But here is what, you know, the
insights are. And then, see whether they would do something about that.
DR. WALLIS: The real payoff would be if -- you know, this
is entitled "Action to Reduce Common Cause Failures". The real payoff
would be if, as the result of something you do, common cause failures
are reduced.
DR. APOSTOLAKIS: They are being reduced.
MR. ROSSI: We provided -- I think we provided curves
showing the common cause failure trend in there.
DR. WALLIS: -- without any sort of issue.
DR. APOSTOLAKIS: In the notebook --
MR. ROSSI: Well, recognize that this, this database is only
one part of the process the agency has for dealing with common cause
failures. We deal with it through our very robust events assessment
process on a daily basis, and when we find new insights of any sort that
rise to the level -- I mean, it has to rise to a level where it warrants
an information notice or a generic letter or bulletin. We put one of
those things out, and we put a number of bulletins and generic letters
out to licensees on common cause failures that they need to be aware of
and need to address. And we put information notices out also.
But we are limited in terms of what we can ask licensees to
do because we have to ask them to do things that are gonna be important
in the safety space and are cost-justified. And that's why, in terms of
closing out a generic safety issue, we have to ask ourselves the
question, do we need new generic requirements? And you have to
recognize that for new generic requirements, there has to be a
significant benefit to safety, and you have to be able to justify it by
cost. And I think that if you look at all of the Agency's activities
related to common cause failures, that there are not, there's not a need
for new requirements of any sort to further reduce them.
DR. WALLIS: Is there a way to get rid of a generic issue by
the process of showing that because of other things that have been
happening, not necessarily as a result of addressing a GSI, the issue
has gone away?
MR. ROSSI: No -- I think, I think that the question is,
when you look at all the things that the Agency does in the way of
common cause failures, is there a basis for new generic requirements
that are cost-justified? That's the question you have to ask with
respect to generic safety issues. And if the answer is "yes", then,
then you proceed. But here, the answer is no. Now, this issue began in
1985, and since 1985 we have done many, many things to address common
cause failures. And now it's our belief that there's nothing more that
needs to be done to close out this generic safety issue. And I do have
some continuing, ongoing things that we're going to do. Dale talked
about things. We have the events assessment process. But is there a
basis for leaving this generic safety issue open at this point in time?
And our answer is, we do not believe there is.
MR. BARTON: Let me ask you a question. There's a party
input to analyze CCFs. You've got LERs you're looking at. You're
looking also at NPRDS data. What happens when NPRDS goes away? All the
stuff here talks about using NPRDS data. That's being phased out. What
happens?
MR. RASMUSON: Can I answer that?
MR. BARTON: Sure.
MR. RASMUSON: The industry has put together a database
called EPIX.
MR. BARTON: Okay.
MR. RASMUSON: We will be taking advantage of that database.
In fact, it is much better, has much more detail in it than NPRDS did.
There are a lot of, there's warts and that on it right now. We're
working with industry to get these things taken care of in that. So
EPIX will be the main source for our, the data, for our risk-based
performance indicators and things like that.
MR. BARTON: Okay. Thank you.
DR. APOSTOLAKIS: Is this the new way of pronouncing it,
"EPIX"?
MR. RASMUSON: That's how I pronounce it. "E-PIX", "EPIX".
DR. POWERS: EPIX. E-P-I-X. Yeah, coming back to Mr. Rossi
to discuss his comments, I don't think that either Mr. Butler or I
question the validity of the overall conclusion, although of course
we're not speaking for the Committee. The Committee will have to
deliberate on this. But all we're saying is, is there anything in
addition to just providing the information that you are currently
providing that the Agency can do to do things a little better? That's
all.
And I was pointing out that here is a class of causes that
has been observed and this and that, and maybe you can do something
about it, and addressing the issue of feedback, that something actually
is happening out there. Now on the other hand, I myself would answer
that and say, yeah, obviously something is happening because the rates
are going down. So, whether that is a result of the NRC's efforts or
the general realization that this is an important issue and people are
taking action anyway, or a combination of the two, I don't know and I
don't know that it's worth trying to find out. But the truth of the
matter is that the rates are going down. And it's a good thing. It's a
good thing. So let's keep the questions in perspective as well.
I personally think that, yeah, there really isn't much else
we can do. We should really close it out. But on one of the ten.
MR. VANDERMOLEN: Well, we're using the avenues that
are normally used that are available to us.
DR. APOSTOLAKIS: You're using what?
MR. VANDERMOLEN: We're using the avenues that are available
to us. We're just essentially --
DR. APOSTOLAKIS: No, you're making things worse. The
avenue that you have does allow adding a few bullets.
MR. VANDERMOLWEN: hat can I say?
DR. APOSTOLAKIS: Just go to the next slide.
MR. VANDERMOLEN: Okay.
[Laughter.]
DR. APOSTOLAKIS: Sometimes it's a good idea.
MR. VANDERMOLEN: The next slide --
DR. APOSTOLAKIS: Yeah, we know that.
MR. VANDERMOLEN: You know that.
DR. APOSTOLAKIS: So is there anything else you would like
to tell us?
MR. VANDERMOLEN: That one, I just want -- actually, if
you'll indulge me for just a moment slide 7, I have used these a lot
myself. I wanted to give credit to Dale Rasmussen and his colleagues
because I do think there was a lot of very good work done on this.
DR. WALLIS: Can I ask something?
MR. VANDERMOLEN: sure.
DR. WALLIS: You put out these -- they're used only by the
Staff or do utilities actually use them?
MR. ROSSI: These are the reports that have been sent to all
of the utilities.
DR. WALLIS: Is there evidence that they used them?
MR. ROSSI: Well, that's the question that was raised a
while back.
MR. BARTON: That's my issue.
DR. APOSTOLAKIS: Has anybody in the IPE program said, or
cited any of these reports as their source for the analysis of common
cause failures?
DR. FLACK: This is John Flack. I believe 4780 was the one
we recommend they use --
DR. APOSTOLAKIS: Okay.
DR. FLACK: -- in the IP program.
MR. VANDERMOLEN: The others are more recent. 5485 came out
in 1998.
MR. BARTON: 6268s are real good, except they're, they're
recent also, right?
MR. VANDERMOLEN: Yes.
DR. WALLIS: These are older than the ones John was saying.
This good stuff, you meant the stuff you're putting out now, asking if
that has been used? This has been out there, and the question that I'm
asking is, is there evidence that what's been out there has been used?
MR. RASMUSON: This is Dale Rasmussen. I can tell you that
the database is being used by utilities, because -- at least when they
go to install it, they have to call me up and get a password to use it.
So I know who has it and, you know, who has installed it on their
computers.
MR. ROSSI: I thought, by the way -- and somebody may want
to check this out -- I thought that what we sent the ACRS on closing
this out indicated that the PRA standard that's being developed,
referenced --
MR. BARTON: Yeah, it's in your letter.
MR. ROSSI: Okay. That's very important because if the PRA
standard references this, the PRA standard, of course, that we're
continuing to develop will be the basis of how PRA is done in the future
and how risk-informed regulation is implement to a large extent. So if
it's referenced in there, that's going to push people to use it.
DR. APOSTOLAKIS: Now, I don't remember when we reviewed the
program. Dale, do you remember the ACRS, or the Subcommittee had the
review?
MR. RASMUSON: Yeah. We did. It was like two or three ago.
DR. APOSTOLAKIS: Two or three years ago. I don't even
remember whether we wrote a letter. We couldn't find it. I remember
writing a list of questions that were addressed at the at the
Subcommittee level.
DR. WALLIS: So what is the desired outcome? The outcome is
not producing paperwork; it's that something is actually done in the
field. What is the desired outcome from producing this paperwork?
DR. BONACA: Well, it seems to me, you know, that what I
would like to see is, when you have a failure of some type and you
perform a cause, the stuff will be able to identify in fact those which
result from common cause failures.
DR. WALLIS: The desired outcome is to help the Staff?
DR. BONACA: The Staff --
DR. APOSTOLAKIS: No, no, no.
MR. BARTON: It's to understand the common cause of common
cause failures and take the appropriate corrective action so you don't
have them again. That's, that's where you're trying to go with this
program.
DR. APOSTOLAKIS: the desired outcome is what's on page 110
of this notebook, on the top 3. The rates are going down.
MR. BARTON: That's right.
DR. APOSTOLAKIS: So there is no urgency to this anymore.
DR. WALLIS: So the outcome has already been achieved --
MR. BARTON: Yes, it is.
DR. APOSTOLAKIS: Exactly. The outcome is already there,
and I agree with Herman. Given this and given that the Agency is doing
many other things as well, is it worth pursuing this?
DR. WALLIS: So the outcome extrapolates to zero now? The
graph extrapolates to zero now.
DR. APOSTOLAKIS: No, it stops at .05, the way I see it. If
it continued a straight line, then probably you would hit zero.
DR. WALLIS: No, no. It's --
MR. ROSSI: We would not want to set zero for common cause
failures as our goal for closing out this or any other generic safety
issue, thank you.
DR. WALLIS: It looks very good though. It looks very good
-- it is extrapolating to zero.
MR. VANDERMOLEN: I'd be reluctant to extrapolate to zero.
DR. WALLIS: Yeah, but that's still a good sign. I'm not
trying to be --
MR. VANDERMOLEN: A decreasing trend is, I think, all we can
see from a graph like this.
DR. APOSTOLAKIS: Okay. Why don't you, uh -- slide 8 and
then we'll come back.
MR. VANDERMOLEN: Well, with the provocation reports, as
we've discussed, we feel that the issue is essentially complete and
should be closed. We will consider doing more work on the general
subject of common cause failures, but only in the context of a regular
research program. It would essentially have to compete for funds with
all of our other research programs. That we consider it as a generic
issue.
But I do want to make it very clear that we do, as Dr. Rossi
mentioned a few minutes ago, we do have a very robust program that is
vigilantly looking for such problems, such interactions, and will
continue. It does indeed use all available data, including what we get
from other --
DR. APOSTOLAKIS: But what would say to somebody, that
you've been so successful -- we're looking at figure 1 here that shows
that trend going down, that there really is no need to continue with the
program?
MR. VANDERMOLEN: well, I would say, perhaps it was the
vigilance of the program that has made it go down. If we stop the
vigilance, it may go up again.
MR. ROSSI: To continue with which part of the program?
DR. APOSTOLAKIS: The CCF analysis, all the stuff that Dale
mentioned.
MR. ROSSI: Well I think as we go along, we want -- I mean,
we have developed a very good, solidly based process for looking at
operating experience in a risk-based way. And you know, that's going to
be very useful to us in this. It's one part of it.
DR. APOSTOLAKIS: By the way, we are not reviewing today the
methods that you have developed. We are just reviewing the generic
issue and disclosure. I remember -- as I started saying earlier -- we
had a Subcommittee meeting I believe, maybe a full Committee, meeting
where we challenged some of the methods that are being used. And I'm
wondering whether there's been any impact from those comments on what
you're doing or have been doing the last couple years. I still see here
the alpha factors, the MGL. Has there been anything else that you did?
MR. RASMUSON: Well, we looked at your questions and
addressed them, and evaluated them, as such.
DR. APOSTOLAKIS: I think -- I mean, you're showing, say, on
figure 3, page 2, alpha factors and multiple Greek letter factors. Why
don't you also show the basic parameter, model parameters? It wouldn't
be difficult for you to do this, would it?
MR. RASMUSON: No. They're all related there. You know,
one's just a transformation of the other and so forth.
DR. APOSTOLAKIS: Sure. The thing is that --
MR. RASMUSON: But for the basic parameter model, I need the
failure rates of the components themselves and I don't have those.
DR. APOSTOLAKIS: But you have already looked at random
failures, and your tables do have those.
MR. RASMUSON: Right, but I don't have demands. I do not
have demands, George, or run times. I have independent failures but I
don't have the other part that I need for estimating that.
DR. APOSTOLAKIS: I'd like to know, when you find out that
the licensee used the alpha factor model, I would be curious myself to
read what you've done, because I think it's a bit complicated and I'm
not sure that people would spend the time to do this. Now, your
experience may be different, and this is not relevant to the generic
issue. So I'm not pursuing it any further. But I mean, simplicity is a
great virtue in these things, and I don't know whether a PRA person out
there is familiar enough with these alpha factors to do it. That's why
perhaps trying to give also basic parameter models by using some
examples would help. But again, I don't want to pursue that; it's not
part of the generic issue. What?
MR. VANDERMOLEN: I'll leave it at that.
DR. APOSTOLAKIS: So, do any of my colleagues have any
comments? On what they have heard?
DR. SEALE: I have one comment.
DR. APOSTOLAKIS: Sure. Go ahead.
DR. SEALE: The research program in general has gotten a lot
of gratuitous and not necessarily warranted criticism from various
sources. Here's an example where you are winding up something which has
made a contribution in the past and which, in the risk-informed mode of
doing things, certainly should make contributions in the future. I
don't see where you're hanging the head of this beast on the wall. You
need some, a little bit more visibility for having done the thing and
gotten it behind you. Clearly, the common cause failure problem is
recognized in the industry as much as it is by the Commission and its
staff. And it seems to me you need to figure out a way to get a little
bit more visibility for an accomplishment of the research program.
MR. ROSSI: Well, we will take a look at that comment. I
agree with you. You do recognize, of course, that we report each month
to Congress on the progress we're making on the generic issues, so there
is that visibility.
DR. SEALE: Yes.
MR. ROSSI: And it's probably true that there's more
visibility if we miss a date than if we make it, but that's the way life
is. But we will look at how to make --
DR. SEALE: When you make your report to Congress, if you
could make a citation in that report on how many utilities have now
accessed this thing, with the implied result that they are now using it
in their analyses of failure rates and so forth, it might provide some
of the kind of visibility that I'm speaking of.
MR. VANDERMOLEN: We'll certainly look into that, yes.
DR. WALLIS: On another note, is there a payoff as to
something in burden reduction as a result of having done this work.
MR. VANDERMOLEN: Well, it would depend on what the
particular CCF mechanism was.
DR. APOSTOLAKIS: This was inherited from the AEOD, yes?
DR. SEALE: Yes.
DR. APOSTOLAKIS: And we've seen many times that one of the
things they didn't do very well was advertise what they were doing.
DR. WALLIS: But is there a payoff to the utilities for
resolving this issue?
DR. SEALE: Well, the meaningfulness of their risk
assessments, you know. The quality of the product.
DR. APOSTOLAKIS: I think there is a general awareness of
the problem now that was not there fifteen years ago, the issue of
common cause failures.
DR. SEALE: But the Congress report might be a particularly
useful thing to do.
MR. ROSSI: Well, it is, but recognize it's a very short
report with a list of, you know, what we did and didn't do.
DR. SEALE: Well, maybe you ought to make it a little
longer.
MR. ROSSI: But we will take a look at the mechanisms for
making this work a little more visible.
DR. SEALE: Cite the utilities that are --
MR. VANDERMOLEN: I'm one of the individuals who actually
has to provide text to it, so yes, I'll certainly take that into
consideration. We do report when they are completed.
DR. APOSTOLAKIS: Are there any other comments? I have one
last question: Is this Administrative letter of July 30, 1998 the last
communication from you to the licensees, or do you plan to send other
letters as well, in the future?
MR. VANDERMOLEN: Dale, I'll pass that one to you.
DR. APOSTOLAKIS: I'm sorry?
MR. VANDERMOLEN: I have to pass that one to Dale.
MR. RASMUSON: I don't think we have considered one way or
the other on this. I think that when we finish our insights, it may be
very good to do another Administrative letter and provide the general
summary of common cause, and some things and --
DR. APOSTOLAKIS: I would really --
MR. ROSSI: I believe all of the reports themselves, the
total reports, are distributed widely to the industry.
DR. APOSTOLAKIS: I understand that, but you know how it is.
I mean, if I get five volumes, unless somebody puts a gun in my head,
I'm not gonna read them. But if you put in the letter, look, this is
the useful stuff that's in the report, I think you're helping people;
that's all. So, you would not be averse to sending another letter,
perhaps in the near future, maybe summarizing the insights to supplement
what you have already sent them, and without tying that to the closure
of the issue?
MR. ROSSI: We'll certainly consider it, but those kinds of
letters have to go out of NRR, and we'd have to coordinate it with them
and make sure that it meets their criteria for sending something like
that out. But we certainly will consider it.
DR. APOSTOLAKIS: Yes, okay. Jack?
MR. SIEBER: I think the bottom line is the fact that it's
going to go into the PRA standard. it's not clear to me that licensees
will recognize that they ought to save and start using this information
because it will become a de facto standard for them against which their
performance will be measured.
DR. APOSTOLAKIS: Well, the thing is that the project that
Dale is running is doing many things, and that's why I kept saying,
let's not just look at the PRA. We're not looking at the PRA issue
here. So the PRA guys or the utilities would probably become very aware
of it, or they already are. But are these the same people who would be
making a decision regarding potential couplings the diesel generators?
I'm not sure.
DR. BONACA: In some places.
DR. APOSTOLAKIS: In some places, perhaps.
DR. BONACA: It depends.
DR. APOSTOLAKIS: That's why I think summarizing issues that
are more important to management --
DR. BONACA: That's why, by the way -- I wasn't here -- you
know, look at an executive summary and see for attributes here, which
one specifically talks about PRA admission time. It's a concern
somewhat because already this documentation addresses itself to the PRA
groups, you see.
DR. APOSTOLAKIS: Look at Section 4. There are some very
good insights in Section 4.
DR. BONACA: Oh, I agree. In fact, this is incredibly
valuable to people doing essentially root causes, and the question is,
are they going to be tapping into it? In some cases, they will.
DR. APOSTOLAKIS: Exactly.
DR. BONACA: But in some cases, they won't.
MR. ROSSI: One of the things that we have been doing with
the risk-based analysis reports is, when they're sent over to NRR by
memo, the forwarding memo does give insights and ways they can be used
in the inspection program and that kind of thing. So it does give that
information to NRR for use in the inspection program. That's been done
with, I believe, all of these.
DR. APOSTOLAKIS: Are there any other issues?
DR. WALLIS: I have to ask -- clarification point. This
figure 1 is very impressive about occurrence rate for complete CCF
events per year. But the scale is .05. Now you cannot have a twentieth
of an event, so does this -- this must be per-reactor. You must take
all the events in the country and divide by the number of reactors?
MR. RASMUSON: No, it's per operating year.
DR. WALLIS: It doesn't say that -- no, it says "per year";
it doesn't say "per reactor year".
MR. RASMUSON: Oh, okay, that's --
MR. ROSSI: Let's correct that. But that's what it is, or
we'd have a twentieth of an event.
DR. WALLIS: Well, a twentieth of an event is a bit hard to
measure.
[Laughter.]
DR. APOSTOLAKIS: This is a rate.
DR. SLACK: Not if it's one over twenty years.
DR. WALLIS: No, but it's per-year.
DR. SLACK: It's a twentieth-per-year rate.
MR. ROSSI: That rate is per year.
DR. APOSTOLAKIS: Per calendar year or per reactor year?
MR. SIEBER: Per reactor year.
DR. APOSTOLAKIS: Any other comments or questions? Any
comments from the Staff? Well, thank you very much, Gentlemen.
MR. VANDERMOLEN: Thank you.
DR. POWERS: Thank you, George. I will recess until twelve
minutes after one.
[Whereupon, the meeting was recessed, to reconvene at 1:12
p.m., this same day.]. A F T E R N O O N S E S S I O N
[1:12 p.m.]
DR. POWERS: I will go ahead and bring the meeting back into
session. We are going to now pursue the use of the alternative source
term at operating reactors and the associated regulatory guide, and,
Tom, are you going to walk us through this issue?
DR. KRESS: Not very far, but I will introduce it. You
remember that the NUREG-1465 source term was developed to make it more
realistic, but it was meant for future plant uses, and the issue before
us has been the potential voluntary use of the same source term for
operating plants, to make changes.
We previously reviewed the efforts that were done by the
staff to evaluate and identify any issues associated with that, and we
wrote a letter on that, and I don't know if you recall or not, but we
thought that was a pretty good job that they did, and it made a good
basis for developing a rule on it. So what they now have done is
developed a final rule and a draft regulatory guide and the associated
standard review plan sections. And that's what we're going to review
today is those parts of it.
I guess I would particularly like for the Committee to pay
attention on some specific items, and the ones I would like for you to
make note of, to listen carefully to, are the redefinition of the source
term to make it more broadly applicable to all design basis accidents
rather than just LOCAs. That's one.
A second item would be the adjustments made and the
rationalizations for these for use of source terms for extended burnup.
The third item I want you to perhaps think about while
you're listening is the staff's intent to look at the effects of any
changes based on the new source term, the effects of these changes on
CDF and LERF.
A fourth item would be the question of potential reductions
in safety margins.
And a fifth item would be the draft regulatory guide
requirement to use 1.02 times the license power level to determine
inventories consistent with Appendix K use of 1.02.
And I guess another item would be to think about is the
question of selective use versus full use if a plant voluntarily decides
to use the source term.
So those are the issues I want you to think about as we go
on. I don't know, who on the staff wants to introduce --
MR. REINHART: I would just turn it over to Steve and let
him give his presentation.
MR. LAVIE: Well, good morning. Good afternoon, actually.
I'm Steve Lavie with the Probabilistic Safety Assessment Branch in the
Division of Systems and Safety Assessments in the Office of Nuclear
Reactor Regulation.
The effort I'll be presenting this morning started out
several years ago, as Dr. Kress mentioned. We've had the opportunity to
brief the full ACRS on this topic on several occasions, the most
recently being November 1998. And we briefed the Severe Accident
Subcommittee on August 10 of this year. This morning I'll be discussing
the status of this effort. The staff is expecting the ACRS to notify
the Commission of its position on publishing the final rule for use in
the draft guide for public comment.
In my briefing this morning I intend to discuss the status
of the final rule, and then changes made to the proposed rule language.
I will then briefly update you on some of the changes that have been
made in this document as a result of office concurrence, comments, and
CRGR review. The document you have in front of you when it was
transmitted pointed out that the office review and your review were
being done in parallel. So I'll update you on the changes that have
occurred.
I then get into the draft guide, and when we get to the
draft guide we'll look at the staff's objectives in the draft guide and
then a brief discussion of some of the more significant policy decisions
the staff has made in that guide.
We'll first get into the final rule. The objective for the
final rule was to enable the use of the alternative source terms by
operating reactors. It's not a mandatory approach. It's strictly an
enabling rule. In keeping with the direction provided by the
Commission, the total effective dose equivalent criteria were
incorporated in the rulemaking. The final rule is voluntary and it is
applied only to those facilities who desire to use an alternative source
term, and this will also include renewal licensees as well.
Now included in the rule are some conforming changes made
for the Part 52 and the Part 100 rules that were issued previously in
December of 1996. These particular changes are necessary for future
licensees under Part 52. We just use this opportunity to get these
things done without a separate rulemaking.
The proposed rulemaking was published for public comment on
March 1999, and the staff received seven public comment letters, none of
which raised an objection to publishing the rule.
We're working to resolve the comments we have received, and
one of the public comments forced us into a change in the definition --
actually two comments. One comment noted that the phrase in the
original language "from the reactor core to the containment," while
accurately describing the NUREG-1465 characterization, did limit this
definition for a LOCA. It could give rise to the interpretation that
the staff's intent was that selective implementations were not allowed.
Okay? That was certainly not our intent. We needed to do something to
change that definition.
The second comment we received, from one of the utilities,
was that the phrase "magnitude and mix" as used in the original language
could be interpreted in a way that would prevent licensees from
implementing desirable fuel management techniques without approval under
50.67. Changes such as changing their reactor power levels, as long as
they're still within their tech spec limits, moving the fuel around in
different regions, these type of things could be prohibited by 50.67,
since some were interpreting core inventory to be included in the term
"magnitude." The bottom line we had here is that we were trying to
control the gap fraction specification, not the core inventory.
The final language in the discussion of the statements of
consideration established the staff's intent, which was to control the
fractions.
Obviously the core inventory is controlled by other
regulations already in place, technical specifications.
The other change to the final rule, and this is something
new since we briefed the Subcommittee, is that in our CRGR presentation
it was identified that the language we had used excluded renewal
licensees issued under Part 54. This was not the staff's intent, so we
revised the language to bring this class of licensee into the
applicability.
You might be wondering, the January 10, 1999 date was the
effective date for Part 52, and also the changes to Part 100.
Applicants for construction permits, design certifications, combined
operating licensees under Part 52 are licensed against the TEDE criteria
and the revised source term already. Therefore, Part 50.67 would have
no effect for Part 52 licensees and additionally because under Part 52
the source term would be part of the design certification which would
require processing under 52.97 to be revised.
So we felt this language brought the renewal plants into the
fold without bringing in more people than was justified.
Some of the additional changes that occurred, as we already
discussed, we changed the definition of source term. There were
obviously several editorial clarifications and typos that didn't change
the intent. We revised the text of some of the public comments based on
discussions we had with the Office of General Counsel in the interest of
improving the responsiveness and establishing a clearer record of how
the comment was dispositioned.
There were some changes made to the environmental assessment
to provide a clearer, more legal conclusion by expanding on the draft.
We had said changes are covered by existing regulation and therefore
safe. This was expanded to demonstrate how the existing regulation
assured that the environmental assessments would be done.
We also had to add a section on a new administrative
requirement regarding the voluntary consensus standards. The bottom
line for this particular rulemaking was no appropriate voluntary
consensus standard was found. And as we pointed out just a moment ago,
we clarified the applicability of 50.67 to renewal licensees.
Okay. That's the final rule. We'll move into the draft
guide. This is where most of the technical meat is. As we pointed out,
the rule was intended to be enabling, with much of the guidance being
given in the regulatory guide. Our objective in preparing this draft
guide was given here. Representatives from NRR and Research served in a
working group to put this document together.
A preliminary draft was circulated to all the affected
technical branches, and a preliminary draft was made available to the
public in a meeting on June 2. We have had a good deal of discussion
with affected stakeholders in this.
One of the main points I need to emphasize in our objective
is that the draft guide was written to support the staff's traditional
design basis deterministic analysis approach. No attempt was made in
this draft guide to risk-inform the entire process. As you are well
aware, the staff currently has a separate effort in progress to
risk-inform Part 50. If it comes out of that effort, we'll be
revisiting this draft guide.
As we identified in the November 19 --
DR. POWERS: You say suppose they come through and they
risk-inform Part 50.
MR. LAVIE: Okay.
DR. POWERS: Okay? I'm not going to prejudice it by saying
what that's going to look like. A thesis that exists maybe more in myth
than in reality is that when we risk-inform things, we make the analyses
and what not much more realistic. Is this revised source term that we
find coming out of 1465 sufficiently realistic that you think it
qualifies within that myth, or do we do something different?
MR. LAVIE: Okay. The authors of 1465 indicated that their
objective was to provide a representative source term, that it met no
single sequence, nor did it envelop all sequences, that there was still
some need for defense in depth, which we believe we currently have in
the way we do these analyses.
The source term itself is risk-informed in the sense that
they looked at risk-significant sequences to develop it, but no way did
1465 envelop all possible sequences.
Did I answer your question?
DR. POWERS: You've set up a better preamble to the
question.
MR. LAVIE: Okay.
DR. POWERS: Hit me with the bottom line. I think what
you're saying is no, that whereas this is a step in the direction that
if you went the whole way, and we don't know what they'll do --
MR. LAVIE: Right.
DR. POWERS: In this, but if aspirations were met, no, that
what generated this source term, what was used to create this source
term would be used, not the source term itself.
MR. LAVIE: I have to think about that some more. Let me
try taking a different tack. Maybe I've -- when we refer to
risk-informing Part 50 of course you're aware that the Agency's position
is that we are risk-informing, not risk-basing, that the Agency's policy
on the use of PRA and risk information includes balancing both the PRA
analysis with the deterministic analysis. I believe what we're looking
there is is that right now we have licensees analyzing sequences such as
locked rotor accidents, fuel-handling accidents, accidents which are not
risk-significant to the overall plant. We believe that many of these
analyses can be eliminated.
There is also, the Committee I know is very interested in
moving to frequency-consequence curves instead of doing all the
calculations based on point values. These are the type of changes we
would see coming out of risk-informing Part 50 as regards these accident
analyses.
DR. KRESS: Yes, that's interesting that you said that,
because a lot of those sequences that you mention that are in the
deterministic world that are said not to be really risk-significant,
when we say that, in our mind we're thinking of risk as being CDF and
LERF. If we want the frequency-consequence curves --
MR. LAVIE: Right.
DR. KRESS: These things, those particular accidents may
have a significant effect on parts of those curves.
MR. LAVIE: That's correct.
DR. KRESS: And that's been a concern of mine, that we don't
really want to lose that, I don't think.
MR. LAVIE: You're talking to the choir a little bit on
that.
DR. KRESS: Okay.
MR. LAVIE: I feel very strongly about that myself, is that
our choice of CDF and LERF as risk metrics may not serve us in some of
these other events. You take the fuel-handling accident. Some plants
can estimate as high as 45 rem at the site boundary for a fuel-handling
accident. It's not risk-significant.
DR. KRESS: That's right, but you wouldn't want it to
happen.
MR. LAVIE: The public would really feel strongly about 45
rem at the site boundary. I would, if I was at the site boundary.
DR. KRESS: Yes. That's been exactly my concern about going
fully risk-informed, and --
DR. POWERS: Well, on the other hand you can predict 10,000
rem at the site boundary if you give me free rein to invent the
accident.
DR. KRESS: Oh, yes, free rein. I can invent that accident
too.
DR. POWERS: I mean, you've got to constrain me a little bit
about the probabilities. But I guess I'd come back and say yes, there
are some analyses that are currently done that may be eliminated. I
think that's not quite right. They'll be replaced.
MR. LAVIE: That probably would be a better
characterization.
DR. POWERS: Yes. And I think when you replace it, the
question comes up is do you want to replace it and still provide it a
fairly prescriptive source term, or do you want to take a source term
that is a function of the accident is the real question that comes
forward.
It really doesn't make any difference to me whether I have a
prescription out of 14-844 or a prescription out of 14-65. It is still
a prescription, okay, and so if we are going to go to risk-informed
regulation, do we want to retain a prescription on the source term? I
mean it is an approximation and there is nothing wrong with it.
I mean I am not criticizing it. I am just characterizing it
right now -- or do you want to say no, the real insight that was gained
in all the research that was done on the source term is they are a
function of the accident and the prescription is just not a right
characterization.
DR. KRESS: Yes, I think if you want to go that route, you
would have to be sure that you, if you are going to have accident
sequence specific source terms, which I believe in also, you would have
to be sure you have a look at the complete range of accidents then like
a PRA.
DR. POWERS: That leads to the question of when we do a PRA,
carrying it out to consequences, a full-blown PRA, we still have some
relatively stylized accidents in there.
DR. KRESS: There definitely are.
DR. POWERS: And the biggest feature of those accidents is
that typically actions that are not proceduralized are not included, but
when I go and look at accidents that have happened or accidents that
have nearly happened I find these non-proceduralized things affect them
a great deal and so the question comes up are we analyzing the right
accidents.
DR. KRESS: That is a good question, but I don't believe we
will ever -- for a long time -- do away with the design basis concept.
MR. LAVIE: No, I think you are correct. We won't.
DR. KRESS: So there is a need for a design basis source
term that accomplishes what you want and I think this happens to be a
particularly good one.
DR. POWERS: Well, I mean this one is based on a closer
approximation of reality than we have ever had before.
DR. KRESS: Right.
DR. POWERS: And has a diversity to it that is very
attractive when you are using it to assess the adequacy of safety
systems. It is not just all gas and it's not just all particles. It is
kind of a mix.
DR. KRESS: As a matter of fact, that is where its greatest
value is because, you know, the proposed use of it as a site qualifying
source terms and as a containment thing -- actually I could care less.
You could use the other one just as well. It is actually in these other
areas that it is most useful.
DR. POWERS: Yes, I think you are right because it doesn't
appear all at once. It appears over time. I mean there are lots of
things that make it very useful for assessing the adequacy of the
design.
DR. SEALE: Yes, in putting together your version of this
source term for your reactor you have to understand your system a hell
of a lot better than you would if you just used the 14-844, and that is
the virtue of the change.
DR. POWERS: Sure. Sure.
MR. LAVIE: It also puts a lot more burden on the Staff
because everybody has its own way of doing it. Back in the old days we
could say you will assume this, you will assume that, and it was very
easy to review.
DR. POWERS: That's what puts such a burden in the
Regulatory Guide of articulating carefully what the philosophy is behind
that.
MR. LAVIE: Right.
DR. POWERS: Rather than just the specific numbers.
DR. BONACA: One place that I see an enormous benefit coming
from this opportunity is in some cases, for example, the timing of
release. That timing of release has been used as a criterion for design
on systems that are not easy to operate -- the HVAC systems -- and as
you know, I mean it challenges almost every site of their meeting the
one minute requirement for either pulling a vacuum in an enclosure
building or for realigning HVAC systems and so on, and I feel it is
really a detriment to safety, the fact of imposing requirements that
force the equipment to do things the equipment can't do.
DR. POWERS: I think if we were to characterize the one
thing we have learned it is that in taking something that looks on the
face of it to be a very bounding thing, we have created something that
may be a detriment to safety.
DR. BONACA: Absolutely.
DR. SEALE: Yes.
DR. KRESS: That's right.
MR. LAVIE: Well, that's why the Staff feels that this
rulemaking will not only reduce burden but the results are definite
possibilities for improvement in plant safety as well as worker
protection.
If we don't need to have that main steam valve shut in 14
seconds or have the leakage of 14 CFM, we are going to spend less dose
in repairing that valve.
As we mentioned in the November, 1998 briefing of the ACRS,
the Staff chose not to simply endorse NUREG-1465, but we established the
concept of an alternative source term, recognizing that there could be
additional source terms generated in the future, and then we would have
the situation where we would have the revised revised source term, so we
decided to call it "alternative" and the draft Guide contains some
minimum standards for what constitutes an acceptable alternative source
term and that also provides the source term we currently feel is
acceptable. Obviously in any Regulatory Guide they can pose
alternatives.
Based on our discussions with the subcommittee last week, we
changed the language in the standard slightly to emphasize the Staff's
intent that an awful lot of work and effort was put into 1465 and we
would expect any alternative to have the same level of pedigree.
We have also addressed accidents other than the LOCA here
because they part of the traditional deterministic approach and in some
cases vendor analyses project some amount of fuel damage, for instance,
with a steam generator rupture or main steam line break, so it was
decided we would need to address all of the accidents at this point.
I am going to shift now to the scope of implementation. Dr.
Kress mentioned this briefly in his charge to you on things to listen
for. The Staff's original view on the alternate source term was
recognizing that the source term has its tentacles over a large portion
of the plant design and procedures. Initially the Staff had this
feeling that if you went with the new source term it was going to be an
all-or-nothing situation. We were going to let the plants pursue timing
only applications, but the philosophy was if you are going to do a dose
calculation you are going to do all of them.
The Commission rejected this approach and directed the Staff
to allow the maximum flexibility while maintaining a clear, consistent
logical design basis -- a very, very tall order to try to comply with.
Much of the complexity you see in the draft Guide is directly associated
with that sentence -- being not too prescriptive, giving a great deal of
flexibility but also maintaining the consistent logical design basis.
Now the Staff does believe that regardless of what type of
application he comes in with that the applicant must evaluate all the
impacts of the proposed change. That would include risk.
The rebaselining study provides a basis for discounting most
source term related impacts, for instance, increased cesium. Now,
however, the other aspect of the impacts here is that a licensee who
uses the new source term is likely going to do plant modifications and
there will be impacts associated with that modification. Obviously, the
licensee will need to address those as well.
Now, we have decided with regard to the minimum number of
analyses is that if the licensee performs the full LOCA analysis as
outlined in the Draft Guide, the staff can make a finding with
reasonable assurance that the alternate source term implementation is
safe and that future modifications of that facility can use the design
basis that has been updated for the AST and TEDE. Thus, the licensee
would be able to pursue future extensions under 50.59.
Now, if the second implementation, since the staff may have
not had the opportunity to evaluate an extensive evaluation, for
instance, a licensee who comes in and says that he would like to delay a
particular valve closure time from two seconds to five seconds, we
wouldn't necessarily do a very large amount of analysis associated with
that application. However, we would feel very uncomfortable if that
licensee was then to decide that he could extend this to removing the
filters or containment or something of this nature.
So, therefore, under the selective implementation, the
current language has the licensee must come back in if he is going to
change the other aspects of his design. This, incidentally, is an issue
that has got a great deal of interest amongst our stakeholders.
Now, the scope of reanalysis, depending on the pedigree of
the plant, you can find anywhere from a couple of hundred to maybe 500,
600 radiological calculations associated with the design basis.
Obviously, to recalculate all these would be a great deal of effort and
no one would pursue the new source term. So we needed to put some
guidance on what we thought was the minimum requirements, and I pointed
out a minute ago, we expect that all radiological and non-radiological
impacts will be addressed.
Early on in this process we heard people saying we could
eliminate containment sprays on the basis that we don't need to scrub
the iodine, perhaps forgetting that the containment sprays also serve a
pressure transient function. So we are looking at all impacts.
We don't expect a complete recalculation of the design
basis. We do expect licensees could use scaling or scoping analyses for
certain evaluations. For example, in the EQ area, you very typically
have one calculation that provides the source term and then 75
additional calculations use that. If you can show that that is
conservative, then you can make a conclusion regarding the rest of them.
And we point out if a calculation based on the old source
term, a TID14844 and the traditional dose methods, can be shown to bound
the results that would be obtained if you updated the analysis, then
there is no need to update it. However, in this situation, the staff
feels that if you subsequently have a reason to revise that calculation
for whatever reason, that you would update to what was currently in your
design basis, that would be alternate source term and TEDE.
DR. POWERS: Suppose you did that, you had some reason to
update a calculation that you had done previously because --
MR. LaVIE: Yeah, it was bounded.
DR. POWERS: -- it was bounded, whatnot, and you did so. Do
you have to bring that to the staff?
MR. LaVIE: It depends on the application they are going
for. If it involves a tech spec change or is going to end up with a
license amendment, then, yes, it would have to come to the staff. The
staff would have to make a finding about the appropriateness.
DR. POWERS: That is because it involved a tech spec change
or a licensing amendment.
MR. LaVIE: Right. But if he can do it on a 50.59, there is
no reason to come to us.
DR. POWERS: But suppose that it didn't involve those, then
you could invoke 50.59.
MR. LaVIE: That is correct. If he can meet the criterion
of 50.59, no increase in consequences, no increase in the risk, no new
accidents, whatever ends up in the Reg. Guide, then they would not
necessarily have to come to us for review. Now, we believe that that
situation is safe, as you will see when you read the regulatory analysis
and the Environmental Impact Statement, because the type of changes that
can be approved under 50.59 are not expected to have significant offsite
impacts.
DR. KRESS: Yes. The likely risk-informing 50.59 is, in my
opinion, going to allow probably more severe changes in terms of effects
on risk than the current one does. That is just an opinion, because I
don't know that. But in my view, the changes that can be made as a
result of adopting the new source term in a plant, and still meet the
dose requirements that are in, say, 10 CFR 100 --
MR. LaVIE: Oh, certainly.
DR. KRESS: -- and things like that, you are still going to
have to meet those. But in my opinion, those changes would probably
qualify under a risk-informed 50.59 and that you could almost say any
changes resulting from the source term, you look at it from the
standpoint of 50.59 and you don't even have to bring it to the staff.
Would that be a good perspective?
MR. LaVIE: I am not familiar with the language on 50.59
regarding risk-informing it.
DR. KRESS: Yeah, I don't think they have that.
MR. LaVIE: I have read the most recent version and we are
still in deterministic space in the current version.
DR. KRESS: It hasn't been risk-informed yet.
MR. LaVIE: So I am not familiar with it. I would hope that
when we get ready to do that type of an action that we would be
considering the types of things you are addressing. We would have, in
radiological space, we would have a problem. As we mentioned earlier,
what is our metric? If it is CDF or LERF, we probably are going to have
a problem. If it is frequency consequence curve or actually going to
the qualitative health objective or something of that nature, we may be
more comfortable with it. But that is -- I am talking off the top of my
head here because I haven't seen the language.
DR. KRESS: Well, that is an interesting concept.
MR. LAVIE: As I alluded to a couple of times, there are
some impacts of using the new source term. If you are familiar with the
tables, you will realize that not all the nuclides went down. Some went
up.
I do want to mention something that Dr. Kress just reminded
me of is that one of the reasons we even have the enabling rule 50.67 is
the concern that the new source term in most cases will reduce the dose
and as such 50.59 criteria for increasing consequences would never have
been met. This is why the Staff has got the enabling rule that the
first implementation you do we want to see.
On the alternate source term impacts, as opposed to the
plant modifications, the rebaselining study done by Research showed us
that the analyses based on the containment atmosphere, okay, that the
alternate source term results would be less limiting, a conservative
situation -- less limiting in comparison to TID-1484.
However, the study also showed that the increased Cesium in
the containment sump water would increase long-term integrative doses.
The crossover between the TID-1484 source term and the NUREG-1465
results would occur between 30 to 40 days post-accident.
With the exception of the equipment environmental
qualification calculations, the post-accident integrated dose
requirements in Part 50 are based on a 30-day exposure period, thus
there is no significant impact.
For EQ doses, the Staff has a difference of opinion.
However we now have a generic issue to resolve it. It believes that
this impact must be considered. However, the Staff also believes that
this impact could apply to all reactors whether or not they use the new
revised source term.
As a result the generic safety issue has been identified to
determine whether or not regulatory action is justified in this area.
DR. BONACA: Before you move on --
MR. LAVIE: Certainly.
DR. BONACA: -- this could be a very significant impact on
existing power plants, right?
MR. LAVIE: Yes. It is going to vary largely on the plant
and how the analyses were done. Several of the architect engineers
routinely throw in what they call "designer margins" into their
calculations, very often a factor of two. If you can do a scoping
calculation based on the stated source term and show that the source
term MEV per second in the various entity groups only went up by 2
percent, then you are well within that factor of two margin and you
could by an engineering evaluation conclude that your EQ calculations
are good.
That particular facility may not have an impact. Another
facility that didn't have that designer 2 margin could have some
impacts.
DR. KRESS: What equipment are we talking about? Is it the
sump pump and --
MR. LAVIE: This would be the equipment that is relied upon
to operate beyond 30 days that would be most likely the residual heat
removal system components, pumps, heat exchangers, seals, valves --
DR. BONACA: Cables.
MR. LAVIE: Pardon?
DR. BONACA: Cables.
MR. LAVIE: Cabling -- anything that is reliable to work
longer than 30 days.
DR. KRESS: Are those close enough to the sumps -- you know,
you are assuming this stuff is in the sumps --
MR. LAVIE: No, what we are looking at here is not so much
what is in the sumps. What is inside containment has already been we
believe adequately addressed. Most of the components in containment are
designed to a 10 to the 7th rad -- very, very high numbers.
What we are looking at here is there's a lot of components
outside the containment that recirculate containment sump water --
DR. KRESS: The stuff they recirculate --
MR. LAVIE: Right, and that dose, that activity being pumped
around can irradiate components in those areas usually in the auxiliary
building and continuous areas.
It may be in some plant designs be, particularly in the
older plants where TMI lessons learned type of things were put in, I
remember one plant I was associated with had put their ATWS
implementation panel right above, right adjacent to the RHR system
because of lack of space to put it anywhere better. It was the best
location they could find. So it is going to vary from plant to plant
what the impact would be.
You are dealing strictly with the circulation liquids.
DR. BONACA: That would affect, for example, back-flowing
into RWST after a period of time and therefore also site --
MR. LAVIE: Right --
DR. BONACA: -- limit, impact on the control room?
MR. LAVIE: The backflow in the refueling water storage tank
is addressed in the draft Guide as a pathway the licensee needs to
consider.
DR. BONACA: I understand that. I am saying that it has
extreme ramifications. I mean it could affect significantly a lot of
things. All right.
DR. KRESS: But this is just not part of the real rule or
anything. You are just saying that is something you are deferring to a
generic letter possibly.
MR. LAVIE: Generic safety evaluation to find out what
happens. If the generic safety issue comes back and says it is a valid
issue, then the draft Guide will be revised. In addition, there will
undoubtedly be some sort of generic communication to the other licensees
for which it would not be voluntary.
That is a decision that has to be made on the generic safety
issue -- is there a basis for backfitting the other utilities.
Obviously the plant that is going to after a new source term is
expecting to get some cost benefit, and you get a plant that is going to
cut its budget by $2 million due to deferring some maintenance on
components because they no longer need to do it as stringent standards
or something, for that licensee the cost of doing a couple more EQ calcs
is kind of balanced, but for a licensee who wants to have nothing to do
with the new source term it is all cost to him and burden to him, and
perhaps unnecessary burden, and that is what the generic safety issue
will try to determine.
Extended burn-up fuel -- those of you familiar with 1465
will recognize that 1465 source terms were based on fuel behavior
pertaining to burn-up levels lower than -- there's a typo in my notes
but it is correct in the slides -- 40 gigawatt day per metric ton
uranium.
As a result 1465 contained a disclaimer regarding the
applicability of the results, particularly gap activity, to extended
burn-up fuels.
DR. KRESS: Let me ask you about that.
MR. LAVIE: Okay.
DR. KRESS: The NUREG-1465 used calculations and expert
opinion that was based on, quote, "currently operating plants."
MR. LAVIE: That's correct.
DR. KRESS: Were those people not aware that those plants
have high burn-up fuel in them at the peak and places, or was it just
ignored or was it accounted for in some way? That is the question I
have.
MR. LAVIE: I don't believe it was accounted for, otherwise
there wouldn't have been a need for the disclaimer. I believe the
situation is that the people developing the 1465 -- maybe Charlie from
Research could help me out here -- is that the data they had to work
with was from the plants with that pedigree and no one was thinking
about operation beyond 40.
DR. KRESS: Well, let me ask the question another way. In
currently operating plants, about what percentage of the fuel has
burn-ups higher than 40?
MR. LAVIE: The way we are currently regulating, we define
burn-up to be the average over the peak rod. We assume that that peak
rod has got the maximum burn-up and also is in the peak power position,
okay? That is an impossibility. It can't be.
DR. KRESS: Yes. That is a different issue.
MR. LAVIE: Okay.
DR. KRESS: No, my real question is if I go to a plant and
measure the whole burn-up of every fuel in there and average it, what
would that average end up being, roughly?
MR. LaVIE: I think I have heard numbers, I am trying to
recall now, that the 62 gigawatt day metric ton uranium equates to
approximately 45 averaged.
DR. KRESS: Yes, that was the number I am after. That's
right.
MR. LaVIE: Okay.
DR. KRESS: Is it not a good concept to think that the
source term, a realistic source term, which involves supposedly a
significant melt fraction of the core, would be more dictated by the
average than the peak rod?
MR. LaVIE: Okay.
DR. KRESS: That is the question I am trying to get to.
MR. LaVIE: The question, okay. The staff used that
rationale in looking at the extended burnup issue. For the loss of
coolant accident, we realize that the average is the appropriate thing
to use because all the fuel elements are equally potentially affected.
Where we kind of have a problem now is when we go to an accident like
the fuel handling accident where you are pulling out one assembly.
Okay. You could pull out the worst assembly. It is unlikely because
usually that is the third region, it has had a chance to be outside the
peak power region for a while.
So the analyses methods and the way the source term has been
applied to these other accidents takes into account, we believe, the
fact that it is appropriate to use an average for some accidents and not
for others, and that is part of the justification for why the staff
believes we can use the 1465 numbers as high as 62 gigawatt day per
metric ton uranium. And that is explained in the rationale, it is a
part of the regulatory analysis.
DR. KRESS: Okay. But you did adjust the gap inventory?
MR. LaVIE: Yes. For the non-LOCA.
DR. KRESS: But you did that for all accidents.
MR. LaVIE: For the non-LOCA accidents, yes.
DR. KRESS: Oh, I guess I missed that.
MR. LaVIE: Okay. We used the fraction specified in 1465
for the LOCA.
DR. KRESS: Okay.
MR. LaVIE: It was the non-LOCAs.
DR. KRESS: I think I missed that.
MR. LaVIE: Okay.
DR. KRESS: The adjustments to the gap was for the other
accidents.
MR. LaVIE: Other accidents. Because part of the rationale
was we really didn't need to address in the LOCA space because the gap
is such a small portion of what we are assuming is going to get out
eventually.
DR. KRESS: Right.
MR. LaVIE: It affects the timing of things slightly.
DR. KRESS: Of course it affects some of the valve isolation
stuff.
MR. LaVIE: But, significantly, we didn't feel it was an
impact, that we could take the data from 1465 and use it up to 62, 62
being what we are currently licensing people to.
DR. KRESS: Thank you. That helps a lot.
MR. LaVIE: Okay. I am sorry, I didn't understand the
question.
Okay. In developing that rationale for extending 1465 to
the current plants, we had meetings with representatives from Research,
Reactor Systems to address how the data could be extended, and we came
up with what we believe is an appropriate rationale. One has to keep in
mind that this rationale is based on deterministic design basis
considerations. We are certainly not suggesting that what we have put
in there is accurate, however, we do believe it is conservative.
Okay. The staff, with regard to the reactivity insertion
accidents, rod drop accident, rod ejection accident, we are handling
those pretty much on a case-by-case basis as they come in, although the
Reg. Guide does give some default numbers.
This particular event is being considered under the agency's
high burnup program. As you are aware, Ralph Meyer of the staff
presented the program to the ACRS in a meeting earlier this year. As
the data comes out, if there is a need for us to change, we will change
the guidance we are putting out.
One of the other extensions we needed to do in getting from
1465 to something we could put in to address all the accidents in the
design basis is to look at the iodine species. NUREG-1465 provided
excellent data regarding iodine species for the release of activity in
the containment atmosphere, and, basically, to summarize, 95 percent
cesium iodide, 4.85 percent elemental and .15 percent organic. Okay.
We had no reason to feel that that was inappropriate and we obviously
decided to stay with it.
Although 1465 did identify a few other accidents, they
provided no additional iodine species value. So the staff, in our
meetings, developed what we currently believe to be the approach on
iodine species. The data are generally based on adjustments to existing
regulatory guidance. This is deemed appropriate given the traditional
design basis deterministic use of these data. They may not be 100
percent accurate, but we believe them to be conservative.
Now, late in the preparation of the Draft Guide, industry
provided a technical position of iodine species for a fuel handling
accident, advancing cesium iodide as a predominant form other than
elemental iodine. We will be considering this position as we consider
other public comments that come in during the public comment period.
I want to also point out that in many of the deterministic
analyses, the assignment of the iodine species is largely irrelevant,
because most of our design basis analysis assumptions make no
distinction on species. For instance, dose conversion factors do not
change if it is elemental or organic.
We may have decontamination factors in various process
streams that are iodine species consistent, but most of the assumptions,
they are not affected by the iodine species.
We did take the opportunity, although we didn't go all the
way to risk-informing it, we did take the opportunity to do some updates
in the guidance we have been giving licensees on doing these analyses.
A lot of this was based on the work done in the pilot projects. Jay Lee
is going to talk about the pilots when I finish. The guide also
contains some improved assumptions and methods developed outside the
pilot reviews. This slide kind of summarizes some of them.
We have a new containment spray removal model, new to the
regulatory space. We allow credit for main steam line deposition in
boiling water reactors. We have allowed licensees to propose fuel
damage estimates based on integrated enthalpy rather than DNBR. This is
for the non-LOCA accidents, except for the rod drop accident, rod
ejection accident, because there is significant doubt right now
regarding the criteria which are used for those accidents.
There is -- we reduced the iodine spike multiplier from 500
to 335, not a big change, but --
DR. KRESS: What was the basis behind that change?
MR. LaVIE: It was worked on on the steam generator,
alternate repair criteria effort. There was some research done
apparently that justified reducing it from 500 to 335. I do not know
all the fine details.
DR. KRESS: Yes, I recall that.
MR. LaVIE: Okay. I know there was a lot of industry work
done on that. There were several papers by a gentleman named Iglesias
who did some work on that. I think at point the industry was advocating
a multiplier of 20, but when the staff looked at it, they found that the
best they could support was 335. Of course, that only applies to the
steam generator tube rupture.
DR. KRESS: Yes, I remember now. It was based on actual
data that they -- they bounded it to some extent.
MR. LaVIE: Right. Okay. We also put in the guide that
there will be no rad analysis required for the locked rotor accident
unless you postulate fuel damage. And this change regarding integrated
enthalpy ought to pretty much do away locked rotor accident analysis for
radiological purposes.
DR. KRESS: Did you, in this allowing credit for that main
steam line deposition of BWRs, does that include the suppression pool?
MR. LaVIE: I am going to defer to Jay.
MR. LEE: Yeah, this is Jay Lee. No, we do not.
DR. KRESS: Did you put any new guidance in the guide on how
to deal with suppression pool contamination factors or not?
MR. LEE: The Draft Regulatory Guide does.
MR. LaVIE: Okay. The Draft Guides identifies reduction by
the suppression pool and says the staff will consider it on a
case-by-case basis.
DR. KRESS: Okay.
MR. LaVIE: The problem with the suppression pool is that
the three reactor -- the three containment designs all have different
performance. I believe they determined in the case of the MARK 3 types
that it could have been appropriate, but the licensee chose not to use
it, so we didn't analyze against it.
In a MARK 1 design, maybe taking credit for suppression pool
might not be adequate. So that is why we are treating it strictly as a
case-by-case basis.
DR. KRESS: Okay. This new containment spray removal model,
is that the one developed at Sandia?
MR. LEE: Yes.
MR. LaVIE: Yes.
DR. KRESS: Okay. I am familiar with it.
There were some changes that occurred during the office
concurrence and the CRGR. I will highlight these for you.
The most significant ones are originally in section 1.4 of
the draft guide there was an exclusionary statement regarding
downgrading of control-room habitability features. We indicated that
the staff would be reluctant to consider such changes. The statement
has been removed on the basis that if the licensee can come in and
provide sufficient justification, we will consider it. Of course the
burden of proof will be on the licensee.
DR. KRESS: Is this new stuff --
MR. LAVIE: There's a risk impact there as well.
DR. KRESS: Is this new stuff beyond the subcommittee
meeting?
MR. LAVIE: Yes, this occurred out of CRGR.
DR. KRESS: Yes. So this stuff we haven't heard --
MR. LAVIE: You haven't heard it yet. That's correct.
There weren't that many changes, but these are the ones.
Footnote 12 to section 3.6 of the guide contained a warning
regarding the inapplicability of fuel damage estimates for rod-insertion
accidents based on methods other than DMBR. This was the Cabri issue,
where some foreign research reactors feel that damage occurs earlier
than we previously thought. The applicability of that to United States
designs is kind of questionable. However, it's being evaluated by the
Agency.
We put a footnote in there as a warning to say that we may
not be willing to accept enthalpy for these events. The decision was
made that Section 3.6 enables the licensee to propose these things. It
doesn't tell him he has to. And therefore there was no reason to having
a specific exemption -- exclusion for these events. But the staff will
review whatever the licensee proposes.
There were four footnotes in Appendices E through H. They
were changed to clarify the applicability of Draft Guide 1074. This was
with regard to the steam generator alternate repair criteria. That's
what this Draft Guide 1074 is. That that draft guide contained a
different method for calculating -- a slightly different method for
calculating the doses, and also had different dose criteria for steam
generator tube rupture and main steam line break accidents. That's why
it was referenced in the footnote, to realize that there may be other
criteria out there that we need to use in conjunction. We clarified
that slightly to point out that all we're referring to in the draft
guide is the radiological assumptions and not the other material in the
draft guide. That was asked for by the Department of Engineering.
We had a reference to USQ in the draft guide that hadn't
been caught. That's been taken out and the language made more generic
to fit the current 50.59 or the current final not-yet-published 50.59.
Response to request, we had references to alternate source term impacts.
The language used to say the impacts of plant -- used to say the impacts
of plant modifications and the source term. We've changed that to the
impacts of the plant modification in the context of the source term.
We've already concluded that with the exception of that EQ issue that's
being handled generically that there is no impact from the alternate
source term. So therefore the only impact comes from the plant
modification. So that language clarified the staff's intent.
Okay.
DR. KRESS: Okay.
MR. LAVIE: With that -- I've tried to give you a brief
overview. It's a big document. We'd be here for days if I went through
every word.
DR. KRESS: No, I think you did a good job of hitting the
important highlights.
MR. LAVIE: Okay.
I'll be happy to entertain any more questions.
DR. KRESS: If there are any more questions, we'll go to the
pilot studies then.
Seeing none, I guess --
MR. LAVIE: Okay. Thank you for your attention.
MR. LEE: Good afternoon. I am Jay Lee, and I'm also with
Probabilistic Safety Assessment Branch, as Steve Lavie is in NRR.
My presentation this afternoon will deal with studies of
pilot plant review programs, and I will touch on some of a few
alternative source term technical issues.
DR. KRESS: How did this happen to end up in the PRA Branch?
You guys were available? Is that it?
MR. REINHART: I could make a comment on that. The hope is,
looking down the road at Level 3 PRAs is having the source term
radiation people with the PRA people that we'll be in a better position
to make that integration.
DR. KRESS: That's a good idea.
MR. LEE: Implementation of Alternative Source Term at
operating reactors really involves three source term tasks. One is
rebaselining. The second one is pilot plant application reviews that
I'll be discussing this afternoon. And the third and last item is of
course rulemaking and development of a new regulatory guide that Steve
Lavie just did. So pilot plant application review is just a part of the
overall effort for implementing alternative source term at operating
reactors.
For this program we received five pilot plant applications.
By the way, Dr. Kress, this account is slightly different from what I
gave you last August 10 to the subcommittee meeting. Since then we did
complete one more pilot plant review for the Grand Gulf, so now we
completed two pilot plant programs, the Perry and Grand Gulf. One is
still under review. That's Indian Point 2. One on hold in response to
the licensee's request to be hold. And the last one, the Browns Ferry,
they just withdrew their application.
DR. KRESS: Oh, they did?
MR. LEE: Yes, they did.
DR. KRESS: I see you've got large dry, a Mark 3, and a Mark
2.
MR. LEE: Right.
DR. KRESS: What's Oyster Creek? Mark I?
MR. LEE: Perry and Grand Gulf both are Mark III. And
Indian Point, Unit 2 is of course PWRs.
DR. KRESS: Yes.
MR. LEE: For the Perry Unit 1 pilot program, they requested
to eliminate the main steam isolation valve leakage control system.
Also they wanted to increase allowable MSIV leakage rate from 25 CFH per
line or per valve to 100 CFH, and 100 CFH from all four steam lines to
the 250 CFH.
We completed this review last February, and we issued the
license amendment in March this year. And I'll be discussing more in
detail about how we did it for the Perry station.
For the Grand Gulf Station, their request was just limited
scope application. This is selective implementation of fission product
release time. They requested to increase onset of fuel rod gap release
timing from 30 seconds currently specified in NUREG-1465 to 121 seconds.
Now we're not sure we're going to go to 121 seconds.
Probably we'll go to draft guide something like 120 seconds, which is 2
minutes.
We completed review just a few weeks ago, and we have not
issued the license amendment as yet. I think that is in the process
right now by the project managers. We do expect another submittal from
Grand Gulf for the full scope application using these alternative
accident source term sometime this month.
For Indian Point Unit 2 they requested to remove the
in-containment filtration system, which is a bit unusual for the PWR to
have such a filtration system inside the containment. So therefore
using these alternative accident source term they would like to
eliminate this in-containment filtration.
We just received -- as a matter of fact I think we received
yesterday more information from Indian Point that we requested through
our request for additional information. And we didn't really have the
chance to look at the review, but we expect -- we certainly hope to
complete review of Indian Point Unit 2 by the end of November this year.
DR. KRESS: Was that intended to knock down the elemental
iodine concentration in containment?
MR. LEE: Yes.
DR. KRESS: So they can meet the dose criteria a little
earlier?
MR. LEE: Right. This is engineered safety feature system
inside containment. Not very many PWR have these features, and
certainly the Indian Point, they would like to --
DR. KRESS: So it consists of a prefilter and charcoal
filters?
MR. LEE: And hepa filters.
DR. KRESS: And hepa filters also.
MR. LEE: Yes. Whole filtration systems.
DR. KRESS: So they're going to take the new source term and
take credit for natural deposition of aerosols and show they can still
meet the TEDE criteria without this thing. Is that --
MR. LEE: Correct. Yes.
DR. KRESS: Okay.
MR. LEE: As I mentioned earlier, the Oyster Creek is on
hold, and the Browns Ferry unit withdrew their applications.
DR. KRESS: Do you know why Browns Ferry withdraw theirs?
MR. LEE: Pardon?
DR. KRESS: Do you know why Browns Ferry withdrew their
application?
MR. LEE: They said they have some problem with the
resource. They just --
DR. KRESS: Oh, they just can't put the resources on there.
MR. LEE: Right. And so they withdrew.
For the Perry nuclear powerplant of course we used the
NUREG-1465. Now this NUREG-1465 is actually severe accident source
term, and for the design basis accident, for the fission product release
magnitude we used only the releases from the gap and in-vessel releases.
We assumed the vessel itself is intact, along with the containment.
Therefore, we did not use ex-vessel release and late in-vessel specified
in the NUREG-1465.
DR. KRESS: That was based on an assessment that the
probabilities of the accident going that far in the first place is
pretty small?
MR. LEE: Yes. We really like to differentiate the DBA from
severe accident. DBA definition itself is the reactor vessel itself is
intact.
DR. KRESS: Um-hum.
MR. LEE: And therefore we used fission-product release
portion that generated through in-vessel releases.
DR. KRESS: Does that look like a little bit of an
inconsistency to you in the sense that if you had this extent of core
melt that is represented by this source term, that means you have melted
core in the bottom of the vessel --
MR. LEE: Yes.
DR. KRESS: And it takes what, another 15, 20 seconds for it
to melt through and end up in the cavity below, or --
MR. LEE: Well, we're thinking about like a TMI-type
accident --
DR. KRESS: Oh, I see.
MR. LEE: Yes, the reactor core is melted, but it didn't go
through the reactor vessel.
DR. KRESS: Boy, that's one in a million.
MR. LEE: That is just the definition of the design basis
accident for radiological assessment, and we didn't want to go beyond
the vessel failure or core concrete interaction or containment failure
or containment bypass. Those are really severe accident sequences.
DR. KRESS: But it really wasn't based on probabilities?
MR. LEE: No.
DR. KRESS: That you would go further.
MR. LEE: No.
DR. KRESS: Because of --
MR. LEE: It was deterministic.
DR. KRESS: The likelihood you would go further.
MR. LEE: Yes.
DR. KRESS: Okay.
MR. LEE: And also we used the new rules Steve described
earlier, that's 10 CFR Part 50.67, which has 25 rem TEDE for site
boundaries, and 5 rem TEDE for control room operator.
And we gave credit, like Steve briefly discussed earlier,
the credit for the aerosol deposition in the main steam line. Also we
gave credit for dry well aerosol deposition inside the dry well.
DR. KRESS: Let me ask you about main steam line deposition.
MR. LEE: Yes.
DR. KRESS: I thought that credit was already in 1465 source
term.
MR. LEE: No -- 1465 source term really deals with release
magnitude, timing and chemical forms.
DR. KRESS: That's out of the core. It makes its way
through this main steam line and ends up into the containment.
MR. LEE: Well, main steam line is really -- it is not part
of containment and the source term that defined itself earlier, the
fission product release from the reactor coolant system into the
containment.
DR. KRESS: Right.
MR. LEE: And it is not part of release into the main steam
line or --
DR. KRESS: Or a part of the reactor coolant system?
MR. LEE: Yes -- it's really from the reactor vessel itself.
You are right, main steam line is a part of reactor coolant system.
Well, it is really steam line.
DR. KRESS: My impression was that that position was already
accounted for in 1465.
DR. SEALE: No.
MR. LAVIE: This is Steve LaVie again. The BWR design basis
LOCA is generally involving with the recirculation piping and not the
main steam line so that would be the direct path into the containment
and the dry well would be from the recirculation piping.
I mean steam piping is a pathway bypassing containment.
MR. LEE: I guess NUREG-1465 dealt with any deposition
inside the vessel itself but did not include any deposition in the main
steam line. Yes, Charlie?
MR. BOEHNERT: Charlie, identify yourself, please.
MR. TINKLER: This is Charlie Tinkler from the Office of
Research.
NUREG-1465 does already account for deposition in the RSC.
That is because it is released to the containment.
The main steam line aerosol deposition that is being
referred to in this analysis is deposition of aerosols between the
isolation valves, between the MSIVs. That is another section of the
main steam line.
DR. KRESS: Okay.
MR. TINKLER: So there is an opportunity if there is a small
leak break past the first isolation valve, a large volume, and then the
outboard isolation valves, so there's ample time for additional
deposition.
DR. KRESS: I understand what you are talking about.
MR. LEE: Now aerosol deposition in the main steam line as
well as inside the drywell, we use the methodology developed by Sandia.
It is NUREG-6189, a simplified method for aerosol removal by natural
processes in containment. We use that methodology.
Also we gave full credit for the containment spray and the
methodology we used at Perry we also used the methodology developed by
Sandia.
DR. KRESS: Is that a change from the previous design basis
giving full credit to containment spray? I thought you used to give
only partial credit.
MR. LEE: Well, chemical forms are different. Now we are
dealing with the mainly 95 percent is aerosol form. Only 4 percent is
elemental iodine. Back in the TID source term 95 percent is in the
elemental iodine.
DR. KRESS: Right.
MR. LEE: And we still use the SRP method for giving a
credit for elemental iodine but aerosol deposition or aerosol removal by
spray we used this new method developed by Sandia. Yes, that is
different.
DR. KRESS: You don't assume some failure of that spray
system?
MR. LEE: Pardon?
DR. KRESS: You assume the spray system is available --
MR. LEE: Yes, typically they have two trains.
DR. KRESS: I thought previously you assumed one of those
trains had failed and only used the other one.
MR. LEE: Right.
DR. KRESS: For the elemental.
MR. LEE: Right. Perry, I believe we used one train with
the two trains designed and they are engineered safety features system.
DR. KRESS: Are you telling me now you don't, you just
automatically use both trains?
MR. LEE: No, one train.
DR. KRESS: Okay. That is what I didn't understand. Okay.
MR. LEE: For suppression pool water pH that Perry proposed
to use, the sodium pentaborate they have in their existing stand-by
liquid control system, and we got some help from Oak Ridge to calculate
the resulting pH using the full amount of sodium pentaborate in a
standby liquid control system at Perry, what the resulting pH might be,
and Oak Ridge calculated and we reviewed it to the basic above 7 and
actually the number is like 8 -- 8.3 pH.
DR. KRESS: I have a question about that. It may sound a
little oblique. The question I have is what does that have to do with
the use of the alternative source term?
MR. LEE: Not directly related, but in BWR the licensees
didn't really claim that the pH control or pH additive while the PWR
used chemical additive to control the containment water sump.
DR. KRESS: Is that a condition in the alternative source
term that in order to use this source term you have to control the pH?
Is that in the rule somewhere?
MR. LEE: Well, they don't have to control pH but we don't
think they can meet the dose criteria without controlling pH.
DR. KRESS: That is my question. That doesn't show up in
the alternative source term.
MR. LAVIE: This is Steve LaVie. Where it shows up in the
alternative source term is in the specification of the iodine species --
the 95 percent and the 4.85 percent and the 1.8 percent species apply
only if the pH is maintained above 7.
DR. KRESS: Does it say that in the rule?
MR. LAVIE: It says that in the Reg Guide.
DR. KRESS: There is a place in there that says you have to
control the suppression pool pH in order to use this source term.
MR. LEE: That is not conditional. If they can meet the
relevant dose criteria, they don't have to control, but we don't think
they can meet without the controlling pH>
DR. KRESS: I don't think they can either. Okay.
MR. LEE: You know the Regulatory Guide is always guides,
but they can propose alternative to it, and if they can show us they can
meet those criteria without controlling pH we will certainly deal with
that case by case but we really don't think they can do that without
controlling pH.
DR. KRESS: But if they did not -- my point was that if they
did not control pH, then they have to use some other source term.
MR. LEE: No, no. They still can use alternative source
term, even if they don't control pH, if they can meet --
DR. KRESS: I don't understand -- because the reason for the
pH is to keep additional iodine from coming back into the containment,
which is another source term.
MR. LAVIE: This is Steve LaVie. The way the draft Guide is
written is that they can use the assumed, the 95 percent cesium iodide
only if they can control the pH to 7. If they can't control the pH to 7
then we need to look at the iodine species over again. They would be
forced into some work like -- Oak Ridge has got a code that will do
this.
DR. KRESS: And would have to calculate a new source term?
MR. LAVIE: Right, given the different species. Right.
Because the species would affect the efficiencies of the sprays --
DR. KRESS: So basically I was right. The source term in
1465 is conditional basically on controlling pH, otherwise the Guide
says you are going to have to do something else. Okay.
MR. LEE: Also we used updated ARCON96 called for
determining control room dispersion or control room factors. This is a
new code to estimate the dispersion in the vicinity of buildings to
calculate relative concentration at the control room intakes.
DR. KRESS: Is it something like HABIT?
MR. LEE: Well, it is like HABIT code, but it is a complete
different code. Then we used the new, more updated dose conversion
factors given in Federal Guidance Report 11 and 12, which is really
based on ICRP-30.
Using those parameters and assumptions the relaxation to the
operating license in this case changes in the tech spec. They were able
to delete main steam isolation leakage control system. Without this
system they can still meet the new rules and also they can meet the new
rules with increased allowable MSIV leakage rate.
Here are some relaxations to the design basis. I just
listed some of them here for Perry Station. There are annulous effluent
gas treatment system, this is equivalent to, typically, BWR standby gas
treatment system, charcoal absorber not used. They didn't have to use
this filter. The control room HVAC --
DR. KRESS: I guess one of the problems I have had with this
is that such a charcoal absorber could be a very useful device for
long-time accident management, well beyond the time when it is risk
significant and when you start converting iodine into organics and
rereleasing the other forms of iodine that are not aerosol. But it
would be nice to have a charcoal absorber in there, even though it is
not risk significant in terms of CDF and LERF. This is one of those
places where it would probably affect the FC curves, if one had FC
curves.
I was wondering what -- I see you say it is not used here,
but it is used some places.
MR. LEE: Physically still they have this particular
absorber.
DR. KRESS: They have in there, that they could turn it on
if they wanted to?
MR. LEE: Well, they are still using it as a matter of fact,
but in our calculation, and also their calculation, we just assumed that
this particular charcoal absorber does not exist. It is no credit.
DR. KRESS: Yeah, it wouldn't affect your calculation
anyway.
MR. LEE: But because it is a relatively small amount of
elemental iodine and organic iodine, no.
DR. KRESS: Well, my concern about is having it removed from
the design basis when actually it may be a very useful device to have if
one had FC curve criteria.
DR. APOSTOLAKIS: That was my question. Has anybody done
these calculations to see, you know, how sensitive the curves are to
changes of this kind?
DR. KRESS: No. I don't think so. I haven't seen it, I
have looked.
DR. APOSTOLAKIS: Well, we do have a set of curves from
NUREG-1150, so I don't know how --
DR. KRESS: How sensitive they would be to something like
that.
DR. APOSTOLAKIS: -- feasible it would be to do the
calculations. Bob says no.
MR. LEE: You mean with and without charcoal absorbers?
DR. KRESS: Yes.
DR. APOSTOLAKIS: Yes. Or other changes. I mean we have
been talking about the FC curves.
DR. KRESS: I don't know how sensitive they are. That is a
good question, because that is the question. If they are not sensitive,
who cares?
DR. APOSTOLAKIS: Yeah.
MR. SIEBER: Well, I think from a practical standpoint,
Perry at one time had a history of fires.
DR. KRESS: Well, yeah, it may have other -- it may have
risk implications that are different, you are right.
DR. APOSTOLAKIS: That is an interesting point.
DR. KRESS: But that has been my problem with CDF and LERF
all along. These kinds of things don't really get captured in it.
DR. APOSTOLAKIS: They are summary measures as you said. I
think we need to understand that a little better.
DR. KRESS: Yes. But, anyway, it doesn't impact this at
all, it is just a side issue.
DR. APOSTOLAKIS: Yeah, that is why he is happy.
MR. LEE: You know, of course they do have HEPA filters and
this NUREG-1465 source term itself, more than 95 percent of iodine is in
the form of aerosol, which is subject to be removed by HEPA filter.
Control room HVAC initiation, now they were able to delay
that for 30 minutes. Also, control room charcoal absorber efficiency,
they were able to reduce from 95 to 50 percent, they degraded charcoal
absorber in the control room emergency filtration system. They were
able to increase allowable engineered safety feature leakage from 10 to
15 gallons per hour. This is leakages outside the containment, such as
ECCS system.
They were also able to increase allowable containment bypass
leakage. There the primary containment is designed to leak .2 percent
per day leak rate. From that .2 percent per day leak, I think they used
7 percent of that will bypass the annulous or shield building, go
discharge directly to the environment. Now, they were able to increase
that to the 50 percent from previously assumed 7 percent, so it goes
from 7 percent to the 10 percent.
Now, for the control room unfiltered air inleakage, I set it
here high. We used -- the licensee and NRC both used unfiltered air
inleakage of 1,300 CFM, which is very inleakage indeed.
DR. KRESS: Yeah, but it matches some of the measurements
that have actually been made.
MR. LEE: This inleakage, what they do is in the current
tech spec, they are required and that they are doing this, a positive
differential pressure test 18 months. And they do not have any toxic
chemicals stored at the Perry site. And with this high inleakage rate,
they were still able to meet 25 -- excuse me, 5 rem TEDE in the main --
the control room operator.
DR. KRESS: That's good.
MR. LEE: So I guess this is my feeling that not every
reactor need such as trace, I guess, test, for example. This is really
highly specific to the plant design, the way they design the control
room. So a plant with such a high inleakage 1,300 CFM, and if they have
a positive differential pressure test in the containment relative to the
surrounding buildings and with no toxic chemicals, I believe they
certainly can get by with differential pressure testing rather than
going through such as more detailed trace gas testing.
DR. WALLIS: As long as it only an air inleakage and not
that the net inleakage is a result of in and out, because if you have in
and out, there is a net and something coming from a place where the air
isn't clean, --
MR. LEE: The Perry --
DR. WALLIS: -- it will make -- it will contaminate.
MR. LEE: Right.
DR. WALLIS: It is the just net effect, there is a net
balance in, it doesn't mean they are okay.
MR. LEE: That's right. Well, Perry control room have the
ability, envelope design is such a way there is no other, any duct work,
for example, going through that that envelope, so there is no other
contaminated source potentially leaking into the envelope. So, as I
say, this is very specific to each control room ventilation design. In
the case of Perry, we felt that this positive differential pressure test
is sufficient with such high inleakage assumed with no toxic chemicals
stored onsite.
DR. KRESS: The transient toxic chemicals that go by is
sufficiently low in probability that you don't have to worry about
those?
MR. LEE: Right. I believe there are no major highways or
railroad or any such transportation pathways near or through the Perry
plant.
DR. KRESS: We can't assume that this would be a general
thing for other plants, though, because they have the chlorine stored
onsite.
MR. SIEBER: Roads.
DR. KRESS: Roads going by.
MR. BARTON: Near highways where you have got chemical
trucks.
DR. KRESS: So this would be plant-specific?
MR. LEE: Right. Very plant-specific. And then we feel
some plants do need trace gas testing, for example, but not every plant.
It really depends on design-specific nature.
DR. WALLIS: When we heard about control room habitability,
there seemed to be a tremendous uncertainty in just where the flows came
from and went to, and how much leakage there was. The uncertainties
were enormous.
MR. LEE: Once you are able to maintain a positive pressure
during normal operation in the control room habitability envelope, I
think you are safe to say that --
DR. WALLIS: Assume nobody opens the door.
MR. LEE: Well, the do open the door, but here you will see
the tolerance they have, like with 1,300 CFM inleakage, their control
room operator dose was 4.1 rem TEDE. Now, they still have room to go to
the 5 rem TEDE.
DR. KRESS: Not much.
MR. LEE: About 20 percent or so. So, in fact, you are
saying then, 1,375 CFM plus minus 20 percent. And their recirculation
rate is I believe in the order of 20,000 CFM, so you are talking about
10 percent inleakage in the air. That is a lot.
DR. KRESS: When they calculate the TEDE, for the control
room, do they take credit for masks?
MR. LEE: No, they do not.
DR. KRESS: No credit for donning protective equipment.
MR. LEE: No. No KI either. This is strictly --
DR. KRESS: No KI either. Okay. It is strictly if they
were standing there breathing this stuff in?
MR. LEE: Right. We did not count that aspect in the dose
calculation to come out with 4.1 rem TEDE, for example.
Last -- during the subcommittee meeting, Dr. Kress, you
asked me about the dose numbers using a TID source term against the
NUREG-1465.
DR. KRESS: I did?
MR. LEE: Yes.
DR. KRESS: Okay.
MR. LEE: So these are the numbers then, using a TID source
term, the site boundary dose was 140 rem thyroid, not the TEDE, and the
dose criteria is, of course, 300 rem, so they were about halfway. Now,
with the NUREG-1465, they calculated it to be, site boundary to be 22
rem TEDE and they are more closer to the limit, which is 25 rem TEDE.
For control room operator using a TID source term at the .29
rem thyroid and, of course, 30 rem is those criteria and so they are
very close to the limit.
DR. KRESS: The questions I have about this is if the 22 rem
is to site boundary?
MR. LEE: Yes.
DR. KRESS: Is that without any of the changes? I mean, is
that -- does that --
MR. LEE: With all the relaxations.
DR. KRESS: That's actual relaxations?
MR. LEE: Yes.
DR. KRESS: Rem 40 is without any relaxation?
MR. LEE: Without any relaxation and using the TID source
term.
DR. KRESS: And using the TID?
MR. LEE: Yes.
DR. KRESS: So, actually, I guess I was -- okay. I
appreciate that; thank you.
MR. SIEBER: It almost appears to be on the same calculation
as in --
DR. KRESS: Well, I thought you would have gotten something
like --
MR. SIEBER: Yes, 250.
DR. KRESS: Yeah, that's what I would get.
MR. LEE: For site boundary, yes, they had the larger margin
than now. The margin just --
MR. BOEHNERT: You have to speak closer to the microphone.
MR. SIEBER: Okay.
MR. LEE: On the other hand, the control room, as you can
see, they were almost up to the most acceptance criteria and now, it's
more -- there's about 20 percent or so difference.
Now, here is -- Dr. Kress, you expressed about the Ph
control. The iodine source term management, you know, this TID source
term, we deal mainly the iodine for thyroid dose and normal gas for
whole body. Now, we went to the TEDE concept, which considers not only
the thyroid and whole body dose, but other organ doses along with it.
So, therefore, I suppose the importance of iodine is perhaps a little
bit de-emphasized; but, nevertheless, we believe iodine is still major,
critical isotopes that we have to deal with, because iodine has such a
unique characteristics of it being a volatile nature.
DR. KRESS: Yes, very biologically effective --
MR. LEE: Right.
DR. KRESS: -- because, it goes straight to the thyroid.
MR. LEE: They can -- they can petition from the liquid
phase. They can, also, petition from deposited sulfates. So, it's
really -- if it can control the iodine source term, we believe really
you can control the efficient part of that behavior and transport in a
reactor accident. So, our first line of defense is then the Ph control.
We do that by adding -- or using the chemical additives, such as
trisodium phosphate or sodium pentaboric, in the case of a petty
station.
DR. KRESS: Which reminds me, I need to detract a conflict
of interest on the Ph control part of this, because it's based on some
work done at Oak Ridge that I was involved in. That's just for the
record.
MR. LEE: Okay. So, the purpose of this Ph control is, of
course, to mitigate iodine petitioning into the containment atmosphere
form either PWR sub-water or PWR suppression pool. In other words, you
just keep iodine in the water.
The second line of defense is then its containment spray,
even though containment -- main purpose of having a containment spray,
is, of course, containment pressure reduction. But --
DR. VALLIS: That's to put it in the water. First of all,
you keep it in the water and then the spray puts it in the water.
MR. LEE: Yes; right. The spray puts into it; the Ph will
keep it there. And in the case of spray, of course, removal of iodine
in the aerosol and elemental forms, it doesn't remove organic. On the
other hand, organic portion of iodine is relatively small.
DR. VALLIS: Is it advantageous to put something in the
spray that absorbs or interacts with iodine or is just the salability
good enough?
DR. KRESS: No, no, you need something in there for it to be
effective.
DR. VALLIS: Something to mop up --
DR. KRESS: Right.
DR. VALLIS: What do you put in the spray?
MR. LEE: The additives, such as sodium hydroxide. The
stuff will increase the removal efficiencies.
The last line of defense is the engineer safety features
filtration, of course. This filter will remove all forms of iodine. It
is a safety-related system and we're depending on these filters to
remove all iodines.
DR. VALLIS: It removes all forms of iodine?
MR. LEE: Including organic iodine, yes.
DR. VALLIS: Gaseous forms of iodine, combined with gas?
DR. KRESS: These are activated charcoal filters.
MR. LEE: Yes, they are activated.
DR. VALLIS: Not just filters, they're absorbers.
MR. LEE: Well, charcoal absorbers and hepa filters and they
are impregnated filters.
My last slide is some of the items that -- for our future
consideration, all the iodine source to management in the reactor
accident. These are the areas that we are looking into and for the
future studies and following this technical topics. The chemical form
of cesium. Of course with cesium, we think cesium will be in a cesium
iodine or a cesium hydroxide or cesium carbonate, all different form of
chemicals, including cesium molybdate. Cesium is really much more
abundant and efficient compared to iodine. So, we think --
DR. KRESS: Ten times as much.
MR. LEE: Ten times, yes. So, cesium iodine will represent
about 10 percent and the rest of the 90 percent will be in some form of
the chemical forms.
DR. KRESS: How do you expect to control the chemical form
of cesium?
MR. LEE: Control them?
DR. KRESS: Anywhere?
MR. LEE: Cesium is in aerosol form.
DR. KRESS: Yes. But the time it gets to the control room,
it's probably carbonate.
MR. LEE: Yes. Cesium carbonate or, of course, even cesium
bicarbonate reacted with the carbon dioxide in the air.
DR. KRESS: But, I don't know what you do with it then.
It's pretty well fixed?
MR. LEE: But still in the form of aerosol and subject to be
removed by filtration.
DR. KRESS: Oh, yeah, you can do that.
MR. LEE: Yeah; right.
DR. KRESS: You didn't intend for that bullet to mean you're
going to do something to change the chemical form of cesium; you're just
going to deal with it?
MR. LEE: Right.
DR. KRESS: Oh, I see.
MR. LEE: Yes.
DR. KRESS: Okay.
MR. LEE: And, also, we are looking into formation of
organic acid in containment water and the suppression pool due to any
organic compound going into the sump. For example, painted surfaces or
that type of a material going into the sump and get irradiated and --
DR. VALLIS: You put something in the sump, which could be
almost as bad as some of the mixes in the tanks at Hanford.
DR. KRESS: Not quite.
DR. VALLIS: All this stuff in there.
MR. LEE: Right. And this will include some formation of
weak acid. And, also, the other area we're interested is silver
reaction from the typical most of the Westinghouse type reactor control
route, which consists of silver, and the silver reacted with iodine
become insoluble silver iodine. So, it, again, retains the silver in
the liquid phase.
DR. VALLIS: Why do you do all these things with the cesium
and iodine, which are fine for short-term? There are other things in a
really severe accident, like actinides, which come out. Now, if you --
MR. LEE: Actinides --
DR. VALLIS: -- the chemistry for cesium purposes, you make
it worse for the actinides purpose.
MR. LEE: Well, if you look at the NUREG-1456, the magnitude
of cesium release is much, much greater than --
DR. VALLIS: Oh, this is the beginning; yeah, that's right.
Eventually, you'll have to worry about what happens to other things,
too.
MR. LEE: Such as, right, tellurium or --
DR. VALLIS: Right. If it's like the tanks at Hanford,
everything is in there, in some form or other.
MR. LEE: That, we have really no source for.
DR. KRESS: I think from a long time, cesium will probably
control your problems from the long-term and the iodine early on.
MR. LEE: But, particularly, cesium has such a long
half-life than iodine.
DR. KRESS: It's got a long half-life. And it's active and
it's in there in abundance, so I think it's probably the one to worry
about for long-term.
MR. BARTON: It's the one you worry about during TLI cleanup
for four or five years --
DR. KRESS: Yeah.
MR. BARTON: -- with cesium.
DR. KRESS: Sure.
MR. SIEBER: Just to clarify in my own mind, I had the
impression from somewhere that a charcoal absorber really doesn't absorb
iodine; it just delays emission through the absorbent path; is that
correct?
MR. LEE: It's really one we call chem absorption. It does
react --
MR. SIEBER: To some extent.
MR. LEE: Some extent with impregnated charcoal. And, yes,
you are quite right, it delays, also, yes. That's the benefit, as well.
DR. KRESS: You want to delay it long enough for it to
decay.
MR. SIEBER: Well, from the licensing standpoint, you're
interested in TEDE for 30 days.
MR. LEE: Yes.
MR. SIEBER: The fact that you're actually delaying the
emission, is that significant from the standpoint of that 30-day
licensing basis for --
MR. LEE: Charcoal absorber will not give you 30-day delays.
You will delay somewhat, but --
DR. VALLIS: Picks it up again?
MR. SIEBER: To some extent.
MR. LEE: It stays with the impregnated charcoal absorber.
DR. VALLIS: That must be interesting, then, charcoal filter
saturated with iodine. It must be interesting to deal with later on.
MR. LEE: The removal capacity is in such a way that it has
a great exchange capacity and it will reactivate iodine and hold on to
it supposedly. It's a chem absorption process; chemical, as well as a
physical exchange taking place in the impregnated charcoal.
DR. KRESS: Especially for elemental, it latches onto it.
Organic iodines will tend to move on through; but the elemental, it will
latch onto it and stay there.
MR. LAVIE: Excuse me, my response to Mr. Sieber's question,
I believe -- the treatment systems, where you have low fluorides, it
typically takes credit for the absorption of hold up of xenon and noble
gasses in the filter. But the filters we're talking here have fluorides
of 30,000 cfm through them; the hold up is going to be negligible,
compared to the chem absorption.
MR. SIEBER: Okay; thank you.
DR. VALLIS: And then once you've got the iodine in the
charcoal and other things come along, you have all sorts of secondary
chemical reactions, as other things come along and interact with the
iodine that may be freed up.
MR. LEE: That's a potential problem, you are quite right,
such as a chlorine gas, for example. Chlorine will degenerate it, as a
decomposition of organic material inside the containment. So if
chlorine comes along after you remove and retain the iodine in the
charcoal, then how would that react? It appears that they are looking
into that. We are certainly following it with an interest.
DR. KRESS: Are these mostly part of operative programs with
--
MR. LEE: Yes, such as PHEBUS test, for example. And the
other item is zinc polymer. Zinc is, also, reacting with iodine and it
will retain in the containment -- I mean, iodine in the containment.
Zinc will react with iodine, will hold on to it on the paint surface, if
you have a zinc polymer paint, for example. So, these items --
DR. VALLIS: It just seems to me, you refer to Europeans as
if they were one species.
DR. KRESS: You resent being locked in with the French, do
you?
DR. APOSTOLAKIS: Certainly not with the Greeks, though.
[Laughter.]
DR. SEALE: Mr. Churchill said give the Americans 50 years
of occupation time.
DR. VALLIS: And?
DR. SEALE: And --
MR. LEE: So, I listed these items to show you that even
though we do have the text and source term and we will have sometime
next year final Regulatory Guide, but that's not really the end of it.
We will continue to improve and will continue to interact with
stakeholders. And as more additional information come along, we will be
revising at times the Regulatory Guide and sometime in the future even
source term. And we will have branch technical positions or somebody
will communicate with our licensees and the applicant, if we are sure we
have more additional and technical information on these subject areas.
That ends my presentation on the pilot plant review program.
Do you have any questions?
DR. KRESS: Are there any questions for Mr. Lee?
DR. VALLIS: I think one result of the pilots will be to
identify these areas where you need research. That's one of the
purposes, isn't it?
MR. LEE: Well, yes, but I don't think the pilot program
review, itself, will give any answers to these.
DR. KRESS: I don't think so either, because you're not
going to have any tests in those pilots for you to actually put the
source term in.
MR. LEE: We do have interest, though, Dr. Kress, that if
you can give us some -- some suggestion or guidance, perhaps in which
item is perhaps more cost beneficial --
DR. KRESS: Oh, yeah.
MR. LEE: -- we ought to more closely follow. It's sort of
setting a priority. And I think Dr. Powers, during the last
subcommittee meeting, he expressed great interest on these areas.
DR. VALLIS: I think we have suggested that when you do
pilots, it's good to have a list of the things you expect to learn from
the pilot -- you know, which questions to keep asking; you know, which
areas are cost effective or which areas are most important and so on.
MR. LEE: Right. Yes, we could develop such lessons learned
--
DR. VALLIS: Lessons you expect to learn.
MR. LEE: Right.
DR. VALLIS: Even then some already.
MR. LEE: We'll look into that.
DR. KRESS: More questions?
[No response.]
DR. KRESS: Seeing none, I wish to thank Jay and Mr. Lee,
very good presentations, very informative, and we appreciate it very
much.
MR. LEE: Thank you, very much.
DR. KRESS: I'll turn it back to you, Mr. Chairman.
DR. APOSTOLAKIS: Thanks. Okay, we'll recess until a
quarter of.
[Recess.]
DR. POWERS: Let's come back into session. The next subject
we have has to do with control room habitability during a toxic gas
release event. The members have in front of them a digest of some of
the information that came out of the subcommittee meeting on this --
reviewing this Regulatory Guide. The safety issue is that accidental
release of noxious chemical on or in the vicinity of a nuclear power
plant could debilitate the operators. Noxious chemicals are indeed on
many of the nuclear power plant sites and noxious chemicals are used
near nuclear power plants.
The regulatory objectives of this Regulatory Guide certainly
include providing the licensees with less burdensome and more realistic
information on acceptable ways to analyze the magnitude of the threat
posed by noxious chemicals and acceptable ways to mitigate the threat.
They are anxious, also, to reduce the shutdowns due to spurious alarms.
And finally, they're trying to combine the guidance that's available now
for offsite chemicals and on-site chemicals. One of the important
issues to bear in mind in looking at this item is its risk significance.
And I notice in looking in the speaker's view graphs, that he'll go into
that in some detail.
Some of the key elements in the approach that the members
will want to pay attention to is they do provide screening criteria for
when the licensee needs to consider chemicals offsite, based on distance
away from the site, the frequency of which chemicals are shipped, and
the mass of chemicals that are available. They are endorsing an
approved model for the dispersal of chemicals on the site, the HABIT
code. They are providing some limits on the control room
concentrations, which actions have to be taken to protect the operators.
And the final sheet provides the members with some indication of the
points that we might want to raise in some report to -- on this
particular piece of work.
With that introduction, I'll turn to the speaker, I guess
it's you, Sud, and introduce Sud Basu from the Office of Nuclear
Regulatory Research, who has been given responsibility -- I don't know
how he's taken responsibility -- but he has been given responsibility
for this Regulatory Guide.
MR. BASU: That's a very correct accurazation. A comment:
we do anticipate writing on this particular subject.
Good afternoon. I'm Sud Basu. Can you hear me? I'm Sud
Basu, the Office of Research. I was actually hoping that your previous
presentation would run late.
[Laughter.]
MR. BASU: I didn't hope to speak too much.
MR. BARTON: Well, when you see 7:00, you'll get off.
[Laughter.]
DR. POWERS: You see, Sud, this is part of our performance
improvement program.
MR. BASU: I'm really impressed.
Okay. Well, what I want to do is give you a rundown on the
proposal to revise Regulatory Guide 1.78, dealing with the evaluation of
control room habitability, given a toxic chemical release -- toxic
chemical gas release. The presentation was given before the
Subcommittee on Severe Accidents. It was about three weeks ago, I
believe August 9th and 10th. It is going to be pretty much the same
presentation, with some very minor changes. Those subcommittee members,
who heard my presentation previously, please excuse. If you are getting
bored, you have the permission to snooze.
DR. POWERS: And they certainly don't have the permission.
We'll be looking for any slight discrepancy between what you said before
and what you say now.
DR. VALLIS: I think you should look for the hazardous
chemical release, if we all start snoozing.
[Laughter.]
MR. BASU: Talking about discrepancy, I was reading the
transcript from the subcommittee briefing and I notice the transcript
says, at some point when I was introducing myself, that I'm not a
chemist and then I went on to say I'm not a toxicologist, but the
transcript says I'm a toxicologist. So, please make that correction,
I'm not a toxicologist.
DR. POWERS: So, we shouldn't be asking you about MLH and
IDLH and the differences between the two.
MR. BASU: You can certainly ask me; but whether you get an
answer, that's another thing.
The presentation outline: I'm going to give you a brief
introduction and background; and then, I'll -- you're going to get the
post-revision to Regulatory Guide 1.78. I will attempt to give you a
reasonably clear regulatory analysis of the post-revision. I know we
talked about it in the subcommittee briefing and I certainly expect that
we'll discuss more about it in this meeting. And then, finally,
schedules and milestones, with the conclusion of the activities.
Before I get into this, I should also acknowledge that Mr.
Chris Parczewski from NRR is here in the audience. Please, if you'll
raise your hand. He is the user of the product. So, if you have any
regulatory questions, concerns, he's certainly the person that they
should be targeted to.
Regulatory Guide 1.78 has to do with the control room
habitability regulation during a postulative release or accidental
release of hazardous --
DR. VALLIS: The accidental term, I'm not sure that's good,
because if I wanted to sabotage and there are chemicals on site, this
might be a way of doing it. I mean, someone who willfully -- since the
probability is so low anyway, I think you shouldn't ignore a willful
release.
MR. BASU: You mean accidental and non-accidental?
DR. VALLIS: Not non-accidental, the willful release.
MR. BASU: Or intentionally accidental --
DR. VALLIS: Yeah.
MR. BASU: -- so to say? Okay.
DR. VALLIS: They may be made to look like an accidental.
MR. BASU: Your point is well taken. The title of the
Regulatory Guide says during a hazardous chemical release. It does not
distinguish between accidental and other form, as in my description.
Thank you.
And, of course, the list of hazardous chemical does include
chlorine. And I mention that because there is a separate Regulatory
Guide, which deals with the on-site release of chlorine. That's 1.95.
That was published about three years later, three years from the
application date of Regulatory Guide 1.78. And one of the purpose of
this revision is to combine 1.95 with 1.78, those appropriate changes
and revisions.
By way of background, you are familiar with the generic
safety issue 83, control room habitability, in a broader context. You
heard a presentation -- subcommittee presentation, also, about a month
or so ago on that. That issue led to the -- led to reassessment of
control room habitability and control room exposure to toxic chemicals.
And of course, the work that went on resulted in new information on
toxicity limits and new information on atmospheric dispersion modeling.
NRR identified subsequently a need to revise Reg Guide 1.78,
incorporating this new information on toxicity limits and atmospheric
dispersion modeling and asked Research to carry out that task.
DR. POWERS: This raises the issue of when we look at the
probabilities that are recorded in the Reg Guide, they're all down in
the 10 to the minus seventh per year for plant kind of event. That's a
pretty low probability of occurrence --
MR. BASU: That's correct.
DR. POWERS: -- especially if you think the conventional
estimate on the probability that the reactor vessel will shatter and
fall apart is 50 times that -- those numbers. I guess I can't help but
wonder why didn't NRR just decide to dispose of this Reg Guide.
MR. BASU: Good question. I'm not sure if I'm really that
qualified to answer that, NRR being the user of the Reg Guide.
DR. POWERS: Now, you told us you got a Reg Guide -- reg
person here.
What I can tell you -- and I think you have it in your
handout, if you flip to page -- I'm jumping to page 14, and that kind of
gives you a synopsis of operating license amendment applications over
the last 10 years. There, of course, many more applications. These
five were taken out as an example to illustrate that there are
applications from the operating licensees for amendment of the toxic gas
monitoring system or surveillance of that or something that has to do
with the detections, especially for the chemicals, the isolation of
control rooms, etc. So what I'm leading to is that there is -- these
applications do come to NRR and NRR has to review these applications and
make some judgment as to whether or not the amendment should be a code.
DR. VALLIS: Well, I'll ask you though, in relation to
Dana's point now, I'm not sure that's the right measure. The spill of
noxious chemicals, which led to debilitation of the operators, would be
a significant event, whether or not it led to core damage. I'm not sure
that core damage frequency is the right measure of concern that should
be given to this sort of an event. I mean, it just seems to me that
there may well be noxious chemical spills, which seem to me far more
likely than vessel failures, but they won't lead to core damage. The
fact they won't lead to core damage doesn't mean to say you shouldn't
defend against them.
MR. BASU: Well, actually, maybe I can jump to the risk
analysis part of it. But, if you look at maybe a couple of view graphs
before --
DR. POWERS: Well, don't let us disrupt your presentation,
but we can come to risk analysis when -- as you want to. Let me -- let
me -- well, we'll come to that. Let me ask this question: suppose I am
a licensee and I say, gee, I'm tired of these toxic gas measuring
systems and all the headaches they cause. Can I come in under 1.174 and
this Reg Guide and say, I've looked at my site and, indeed, the
probabilities calculated in the Reg Guide are bounding for my particular
site and the delta risk I have is one times 10 to the minus seven; I
have done my IPE and I know my CDF is three times 10 to the minus four;
so, you've got to let me do this?
MR. BASU: But, that's what these amendment applications
are. They don't -- I mean, I haven't seen in these direct reference to
IPE or IPEE, but they do go as far as giving you the CDF for this kind
of events that goes to 10 to the minus seven or even more of that 10 to
the minus seven. And then they say, well, because that's what it is,
could I sort of do away with PGMS or, you know, frequent surveillance or
whatever is currently required.
DR. POWERS: To be fair, they don't really get rid of it.
They take it out of the licensing basis.
MR. BASU: Yes, that --
DR. POWERS: I bet your chlorine tanks still has a detector
on it.
MR. BASU: And physically taken out of the system. They
just don't do --
DR. POWERS: Yeah.
MR. BASU: -- the surveillance as currently required.
DR. POWERS: I guess -- I mean, what all of that leads up to
is that it strikes me, in reading the Regulatory Guide, that it hasn't
been written with a great deal of recognition that this could be the
chain of events that somebody followed; that they -- the risk is low,
the delta CDF is low, so I can put in an amendment under 1.74 and get
rid of the thing.
An alternative approach to writing 1.78 would have been to
say this is what the licensee is going to do. And so, the guidance
would provide them with here are the kinds of things we expect to see
when they do this and essentially alert the licensees, we're going to
ask this question, this question, this question and we expect to see
these kinds of numbers and if you want to get rid of it, you can do it
this way.
MR. BASU: So, would you go one step further saying that you
subsume all these things into 1.74?
DR. POWERS: Well, the next question is, do we really need
it at all.
MR. BASU: I don't know. Chris, would you like to field
that question?
MR. PARCZEWSKI: Chris Parczewski, the Chemical Engineering
Branch. What was exactly the question?
DR. POWERS: Well, the probabilities that are quoted in
1.78, example probabilities are extremely low for CDF. There are a few
times 10 to the minus seventh. With the 1.174 Reg Guide, people can
make arguments for eliminating -- or amending their license, based on
small changes in CDF. This certainly qualifies as small. You got rid
of the whole thing. You're making a small change in CDF, almost
regardless of what the current CDF is that would qualify as small. And
so I'm asking two questions really: why we need the Reg Guide, if we're
dealing with small risks like this; and the second question is why
wasn't -- if we're going to keep the Reg Guide, why wasn't it written in
a fashion that would facilitate licensees, making applications to
eliminate their toxic gas measuring system based on 1.174?
MR. PARCZEWSKI: Well, I think that what we are saying is
that it's probability of core damage due to the incapacitating the
operator is very small. As a matter of fact, this argument is being
said by several of the licensees and we are accepting this type of
analysis. However, several of them are still submitting the whole
analysis. And this is -- I think that the only application that would
be -- the probability of incapacitating the operator is very high --
much higher. It's probably very few cases it's going to happen; but,
you know, as a precaution, in case it happens, you know, we should have
the guide. And some -- there are a lot of licensees still submitting
their probability -- the deterministic analysis. I mean, they're still
using that, you know, approach.
DR. POWERS: And they're asking for changes in their
licensing basis, based on -- for what? What are they using this Reg
Guide for, other than getting rid of the PGMS?
MR. PARCZEWSKI: Well, they are using it to determine their
-- if their operators are incapacitated. In most of the cases,
apparently, they have enough margins, so the operator is not getting
incapacitated. You know, there are two -- there's two minutes criteria.
As long as the toxic gas doesn't reach the limit of toxicity, in two
minutes, the operator has probability to get the equipment -- breathing
equipment and he's not incapacitated.
DR. POWERS: I guess I'm still a little lost. I look at
these events back here and I see since 1991, the ones that are quoted,
they all eliminate expect requirements for PGMS, for surveillance
requirements. So, it appears that this particular Reg Guide is used
largely for amending the license to get rid of some requirements.
MR. BASU: If I may just interject here, that simply is
wrong use, at least that's what's illustrated in this view graph. There
is -- it's used where you are actually providing guidance for the
operator protection in the control room. You have a spill and you won't
know what sort of concentration level an operator is going to be
subjected to, so you need to do some computation, calculation, or what
not. And the Reg Guide is providing some guidance for you to be able to
do so. It's providing guidance to the licensee. It's providing
guidance to the NRC staffer to review the license applications. So does
that answer your question?
DR. POWERS: Well, I think I'm going to go on. Maybe we'll
come back to this after we get to the risk analysis.
DR. VALLIS: I'm sitting here wondering why you're
concentrating only on chemicals in the control room. If you have
on-site major chemical release, it might well be a result of a vehicular
collision, in which case it might well be fire as well. You'd have to
worry about the effect of the toxic on the fire brigade and everything,
as well as the control room people.
MR. BASU: Absolutely. There is --
DR. VALLIS: It's all tied together.
MR. BASU: There is a separate Reg Guide -- separate Reg
Guide -- 1.120 on fire.
DR. VALLIS: But, they actually overlap?
MR. BASU: Yes. I mean, in the broader context of control
room habitability issue, all these things do tie together. I'm
addressing just a very narrow aspect of control room habitability.
DR. VALLIS: But the event might well involve then several
Reg Guides simultaneously --
MR. BASU: I would --
DR. VALLIS: -- analysis of such an event?
MR. BASU: -- certainly think so, yes.
I think we're done with this view graph. I'm going to go
through very quickly, and I think we've kind of touched upon it already
in our discussion, as to how the Reg Guide 1.78 process works, as to the
screening measures for determining releases that must be considered for
control room habitability evaluation. And I have given you in the
following bullets these screening measures.
The first one is the distance screening. All it says is if
this release occurs at a distance more than five miles from control
room, then you not consider -- you do not need to consider it. It's
screen out.
The second --
DR. VALLIS: Is this based on analysis of typical events,
say train collisions or derailments and things like that?
MR. BASU: I think it's more of the former. It's based on
analysis, if you have a spill of some magnitude of that distance and if
you do use the dispersion model to see what the concentration is going
to be five months downstream and you find that the concentration level
is either zero or very small --
DR. VALLIS: Five months doesn't seem much for a major
chemical spill, as a result of a train accident.
MR. BASU: We -- well, it does not; but then when you use
the Gaussian dispersion model, the exponential decreases. So, that's --
I mean, we didn't --
DR. VALLIS: So, five miles is justified by analysis, rather
than being placed --
MR. BASU: At the time. Remember, this is back in '74,
based on analysis that was prior to '74. That's how the five miles was
made. I think the good news perhaps is that with the improvement in
models and all that, it basically confirms that data.
The second screening, and that is after you have gone
through the first screening, that is you've screened out anything and
everything that occurred beyond five miles, whatever is left, then you
look at the frequency. And this is from a lot of sources, like trucks,
trains, and barges. And there's some frequency screening measured
there. For truck accidents, that's 10 accidents per year. So, in other
words, if you have 10 -- more than 10 accidents per year --
DR. VALLIS: Accidents or shipping?
MR. BASU: I'm sorry, 10 shipments per year --
DR. VALLIS: Yeah.
MR. BASU: Thank you, 10 shipments per year. So if you have
more than 10 shipments per year, then you consider that for further
analysis or computation. If you have less than that, you screen out.
DR. POWERS: Let me come back to the distance screening
criteria, I think it's also recognized and established in that screening
that if something occurs, it's very large outside five miles, that
there's going to be some warning available to the operators that it
occurred. The use of Gaussian analysis and what not for many of these
chemicals is often suspect, because they're typically not very good for
these denser than air chemical sources, which do tend to be directional
in terms of ground cutting. But, you still -- there probably would
indeed be a warning for things outside of five miles.
And I would just comment that in an independent examination
of exactly this issue that was done for the DOE sites, they came to
basically the same conclusion, that things outside the site boundaries,
which were at least five miles, sometimes more, they can never get a
hazardous level at the installation of interest --
MR. BASU: That's right.
DR. POWERS: -- from chemicals. It just turns out the
dispersals with Gaussian plume models is always too great. And if you
denied the Gaussian plume model, then you've got warning and they've got
reaction times.
DR. VALLIS: Well, Terrain matters, too. I mean, if you're
in a valley and it's heavier than air, you can get the flow patterns,
which are far worse.
DR. POWERS: But, you always get -- you always have the
warning.
MR. BASU: Right. As a matter of fact, between the
subcommittee meeting and this meeting, I did go through some -- looked
at these onsite locations, their geography, and what I found for many,
but not for all sites, they are -- I mean, they have either the
exclusionary zone or site boundary or something that are more than five
miles. So, I mean --
DR. VALLIS: I'm wondering how you know something like
frequency per year. Here, you've got a reactor and there, you've got a
highway. How do you know that some chemical company isn't going to
decide next year they're going to ship a truck a day of some chemical
along that highway?
MR. BASU: In the future?
DR. VALLIS: Anytime. How do you know it's going to happen
or not going to happen? How do you -- do you monitor everything that
goes along that highway?
MR. BASU: The interstate highway, the shipment --
particularly the hazardous shipments are monitored.
DR. VALLIS: So, you might then prevent them from shipping
along that highway?
MR. BASU: Well, in some interstate highways or some
segments of some highways, there are warnings that you can't simply
transport hazardous chemicals.
DR. VALLIS: So, this may not be under the control of the
licensee.
MR. BASU: Right. This is an external event.
DR. POWERS: The current hazards on most of the state and
federal highways is not under the control of the licensees, but he has
access to the information.
DR. VALLIS: He'll know if it changes?
DR. POWERS: The frequency can change. I mean, you can --
everything can be fine and then next year, there's a new route for some
chemical shipments. It may be worse than that, different chemicals get
shipped.
MR. BASU: Different chemicals.
DR. POWERS: Life does change.
DR. KRESS: It's just like Bill's shutdown risk, we don't.
[Laughter.]
DR. KRESS: We've just got to account for all of those
changes in your estimate.
MR. BASU: If we pass the screening criteria, then we move
on to weight criteria. When I say "pass the screening criteria," I mean
the screen out, those that are not frequently shipped and then we take
those that are frequently shipped within five miles at the weights --
you know, the weights. So, this is the weight of chemicals --
DR. VALLIS: It seems to me you're indirectly using a
probability of a certain dose rate of the site of whatever this stuff
is; that you're indirectly doing this.
MR. BASU: You're absolutely right. Yes, there's a
probability context here. There's a risk context here. So, we go into
the weight or the mass screening criteria and that, of course, depends
on the nature of chemicals to its toxicity. It depends on the control
room time and the core meteorology of the site and all that. But, they
are, in the long term, compared. And as long as the weight criteria is
met, you're going to screen out. So then you are left with a fairly
small subset of those chemicals and the shipment that need to be
considered for the control room evaluation. So that's the process.
Now, what we are proposing in this evaluation is to observe
the atmospheric dispersion model that is currently reference in the Reg
Guide 1.78 to an improved model that is in the HABIT code. And I'll
come back to it in a little more detail. We're, also, proposing an
update of the toxicity, and, again, I'll explain on that. And finally,
we're proposing changes in the shipment frequency and exploring the
definition of shipment frequency than anything else.
Aspects that are covered in Reg Guide 1.78, but which are
not subject to this revision, and so I'm not going to talk about it, the
detection of hazards and isolation of control room -- control room
characteristics, including in-leakage and detection system. They are in
the Reg Guide, but they are not changes proposed. So, that's why I'm
not going to talk about these things. And I said a couple of times
earlier that the most of the revisions subsumes 1.95 and that deals with
the control room operator protection against on-site chlorine release.
DR. POWERS: When I look at Reg Guide 1.95, as it stands
now, there's a lot of useful technical information on how to set up
detection systems and what they should look like and things like that.
Much of that information seems not to have been carried out in the
unification. Is there a reason for that?
MR. BASU: If the relevant information is not carried over,
and I really don't know where you're pointing to, but if indeed it's not
carried over, that certainly should have been done. It's an oversight
and it will be done. Relevant information from 1.95 to be carried over
to 1.78, that is the objective. And at least, in my mind, I thought it
was done.
DR. POWERS: Well, there's a table and a couple of figures
that didn't seem to come over.
MR. BASU: Oh, if you are referring to the different control
room characteristics, there's a table there. There is, also, a similar
table, not identical, a similar table in 1.78 -- the existing 1.78. We
retained that table as an illustration. We're going to compute the
control room flow of various other control rooms, using very simplified
model that we have provided in the appendix. So one can reconstruct the
table in 1.95, if one wishes to.
DR. POWERS: When I look at this analysis that you go
through to find out what their in-leakage is for control room, it says
you take the information from design characteristics construction
drawings. We previously had a presentation here that said, gee, these
guys designed these control rooms to have in-leakages that are very
small; when we measure them, they're huge; can't infer that from the
construction drawings. I'm wondering why does the Reg Guide seems to be
so confident in the ability to calculate in-leakage, based on either
design or construction drawings?
MR. BASU: Yeah, I'm not sure that I can answer that
question. I know that we, in an ACRS meeting, talked about -- there was
a couple of presentations on that issue and I think there's an ongoing
debate on that, as to what is appropriate in in-leakage. I'm really not
qualified to address that.
In this Reg Guide, what we basically said is that if you
have this kind of in-leakage characteristics, that this is going to be
your concentration level. So, calculations -- or the example
calculations have been provided. If you have a different kind of
in-leakage characteristics, you'll have different kind of concentration
level. To argue whether or not this is the right in-leakage for a
particular control room type, I really don't know if I can do that at
this point.
DR. VALLIS: It was very impressive when we heard about
control room habitability, how these in-leakage characteristics could be
off by orders of magnitude from -- the actual measurements could be off
orders of magnitude from this equation model. It's a very striking
piece of evidence.
MR. BASU: Yeah. If you're referring to the analyzed
magnitude versus the observed measure of magnitude, yes. I was here, so
I certainly saw the same thing. And I don't know if we can resolve that
particular issue here. But, I thought you were talking about if you
take one particular in-leakage characteristic, as opposed to another
one, within, of course, reasonable bounds and you compute the
concentration level in the control room, whether you get an order of
magnitude difference, I cannot seriously doubt that you will do that,
because you're using the same model anyway.
DR. SEALE: Well, that bothers me, because presumably a Reg
Guide contains within it a statement of position, which we would like
for the regulated community to have confidence in. And certainly what
you are attempting to do with the rest of this Reg Guide is to update
and integrate the information that goes into it, so that you're able to
present your best assessment of the problem. And here, we have a whole
set of -- or a whole part of this thing, which is terribly suspect, as a
result of measurements that have been made by various people and we
don't even caution the regulated community that they may be walking out
into a mine field and that when they really get into a situation, they
may have an environment which is much more severe than what is presented
here, as a credible level of that environment. I -- that's almost not
ethical.
MR. BASU: That's a fairly valid comment. I'm not sure --
I'm not really qualified to address that comment. Let me just mention
one thing: of the three sub-bullets that I indicated there that are not
subject to litigation and that are not subject to this discussion here,
there's only one on here that I know of -- correct me if I'm wrong --
there's only one of them that I know of that is kind of suspect, that is
the in-leakage characteristic --
DR. SEALE: Sure.
MR. BASU: -- of control room.
DR. SEALE: I think that's correct.
MR. BASU: So -- and that, I have already considered.
DR. SEALE: But two out of three bad is not the kind of
criteria you want to use in something like this.
MR. BASU: No, I like two out of three. But, I can't -- I
consider that in-leakage is a continuing debate. I don't have an
answer. So, I really don't know what I can do about it, other than
basically telling the community that if you have this kind of control
room, this is how -- this is how you compute and instead of saying that,
okay, this is the in-leakage characteristic.
DR. SEALE: There's somebody back over there standing up.
Maybe he knows.
MR. CUZEN: This is Curt Cuzen from NEI. I maybe stepping
beyond my own expertise, but based upon the experts that I have talked
to concerning toxic gas, we need to be careful about the definition of
terms. The term which we have been concerned with from the tests that
we have seen from control rooms is "unfiltered in-leakage;" whereas
during the radiological event, the control room frequently will isolate
and recirculate. And it is this unfiltered in-leakage during that
recirculation that causes a radiological concern.
During a toxic gas event, which is my understanding, and I
qualify this, maybe somebody else more knowledgeable than me, that the
control room only isolates. Therefore, the unfiltered in-leakage, which
we're talking about with the tracer gas test, is not of concern here,
because you are not recirculating, you did not have that flow
characteristics, and, therefore, that is not of the direct concern for
tracer gas, certainly for toxic gas.
Now, I hope I've got that right. I'm calling on my memory
from discussions I've had and it seems to fit with the questions that
have gone on and possibly that can be validated later. But, that's
maybe one way you might want to think about it, before you go too far
down that path.
DR. SEALE: That may be very true, but it strikes me then
that the differences of the circumstances that characterize the response
to a toxic release, as opposed to a radiological release, ought to be
clear in this Reg Guide that deals with toxic releases.
MR. CUZEN: That may be incorrect. I've not seen the
document yet, so I cannot comment on that. But, it sounded like there
was a question about terminology that we could get confused on, if we're
not careful.
DR. SEALE: You are certainly -- that's an appropriate
observation.
MR. BOEHNERT: And I'll just comment, we have a subcommittee
meeting on this coming up on the 17th. We're going to get -- we're
going to have people there, who are intimately familiar with the control
room layouts and how they work and they can -- we can get to the bottom
of this issue at that meeting.
MR. BASU: Okay. It looks like this subject is generating a
lot of interest.
DR. POWERS: You haven't heard nothing yet.
[Laughter.]
MR. BASU: I've seen nothing yet, huh?
DR. POWERS: You've seen nothing yet.
MR. BASU: Okay. The simple one first, frequent shipment,
the definition. If you look at the Reg Guide 1.78 text, that's on the
top of -- the top, and it says, 10 per year for truck traffic, 30 for
rail, and 50 for barge, something like that. And I don't know if I --
it's in the Reg Guide. But, the important thing to look at here is in
the draft Reg Guide -- revision to Reg Guide, it sort of qualifies those
shipments as total shipments; and by "total shipments," what is meant
here is the totality of the hazardous chemical shipment. And
considering the shipments in the singular category, we ask for the
distinction to what chemicals are being shipped.
And let me further expand on that. When we said 10
shipments per year, there is a -- there's always a provision to
interpret that as 10 shipments per chemical per year. And so, here,
we're saying there is 10 total shipments.
DR. KRESS: Trucks appear to me to hold less than trains and
trains hold more than barges.
DR. SEALE: Not necessarily.
DR. KRESS: I am surprised that there's not a correlation
between what the -- unless they can carry in these frequencies. I don't
--
DR. POWERS: It's frequencies of accidents.
MR. BASU: These are frequencies --
DR. POWERS: You have to weigh --
DR. KRESS: You multiply this by the frequency of accidents?
MR. BASU: Yes.
DR. POWERS: There's an inherent weighting to the frequency
of accidents. Trains have fewer accidents than trucks and barges have
very, very few accidents.
DR. KRESS: That would explain it. Thank you.
MR. SIEBER: It seems to me that the original documents in
guidance use the individual chemical commodities as a way to determine
what you should monitor for. For example, at our plant, chlorine was
the only one, even though we had a major railroad line and a major
highway and a river, and probably under those circumstances, we would
have exceeded the frequency limit. But the fact was that the cargos
were so dispersed, you might have one or two shipments a year of certain
chemical and, therefore, you would knock that out, because you would not
want to set up a separate monitoring system just for that. So, I'm not
exactly sure what the rationale is for making the Reg Guide much more
restrictive than it used to be.
MR. BASU: Well, it's more of a clarification and I believe
here the rationale is a very simple way. If you take 10 shipments per
year of truck and if you take the liberty to interpret that as 10
shipments per chemical per year, it is, at least in theory, possible
that in a particular site, you have in a particular year, you have some
15 different chemicals shipped. So, you'll get 150 shipments per year.
That's going to raise your CDF by an order of magnitude, because these
are all connected ultimately to the calculation of risk. So that's
where the line is drawn.
DR. VALLIS: I'm wondering about barge traffic. I mean,
there are some reactors on estuaries, where there's a lot more than
barge traffic on the water.
MR. BASU: Yes.
DR. VALLIS: Or is restricted to barge traffic?
MR. BASU: Why it is not --
DR. VALLIS: Well, there are ships. There are other things
than barges that come by. Out on the estuaries, there are some
estuaries that serve major ports and have reactors on those estuaries.
MR. BASU: That's correct.
DR. VALLIS: More than just barge traffic that goes by on
the waterway. Ships serve as chemical -- refineries and so on. There
are not just barges on the water.
MR. BASU: Yes. I think I understand what you're saying.
Let me ask for clarification, so I can respond to your question. Are
you saying that maybe some chemicals are transported in ships?
DR. VALLIS: Yes.
MR. BASU: I think when we say barges here, the U.S. Coast
Guard keeps statistics of barge traffic and barge accidents. And I may
be wrong, but I don't think that if you have an accident in the
waterways caused by a ship collision rather than a barge collision, I
don't think that they're going to say, okay, no, that's not barge, so
that's out of the accident scenario. I think they will count that as an
accident and I think they will consider that as a barge accident.
DR. VALLIS: You think?
MR. BASU: I think, yeah. I cannot be absolutely certain.
I have to check with the U.S. Coast Guard. But, I --
DR. SEALE: Well, ocean going vessels go up to Mississippi
all the way to Baton Rouge.
MR. BASU: Yes.
DR. SEALE: And that's above waterford.
MR. SIEBER: In the case of Calvert Cliffs --
DR. VALLIS: That's what I was thinking of, Calvert Cliffs.
DR. POWERS: I think that if I was waterford and someone
wanted to lump together ships and barges, I would strenuously object,
because the probability from the collision of the ship is going to be
much, much lower than it is for a barge. Now, I don't think that
obviates the need to -- maybe you need to stick a word in here and say
"ships," as well as barges. But --
MR. BASU: Yes. I think when the counting is done, they
count the number of accidents.
Okay, so, now onto the modeling, atmospheric dispersion
modeling, Reg Guide 1.78 has -- or makes reference to the Murphy Kemp
model, which is -- the Reg Guide is 74 and this one was developed, I
think, in '68 time frame. It's about 30 years old at a minimum.
THE SPEAKER: I'm having a hard time hearing you.
MR. BASU: Oh.
THE SPEAKER: You could move that mike up further in your --
DR. POWERS: If you could just move the mike up a little
higher on your tie.
Can you hear me now?
DR. KRESS: Yes.
MR. BASU: I am not a toxicologist. The age of that model,
that's one thing, but I think more importantly it is what that model
contains. It is a Gaussian plume model all right, but it does not have
the diffusion coefficients in that model as basically normal diffusion
coefficients it does not incorporate the effect of wakes in the vicinity
of building and structures, something Dana was referring to in the
beginning of this presentation.
It also does not include the effect of sort of near-field
effect or the near-ground effect.
So what has been done in the interim on the time the GSI-83
was formulated to now is that some improved models were developed and
these models are put into the HABIT code.
The revision to the Reg Guide makes reference to the HABIT
models and that is what this proposed revision is all about and I have
given you some indication of what this improved model does.
DR. KRESS: Quite often NRC develops codes, particularly in
the severe accident area, and they have a fairly robust peer review
process. Has HABIT been subjected to that peer review?
MR. BASU: You know, I really don't know. I would think so.
What I can tell you is that the model, the improved model that went into
HABIT, was subjected to peer review.
DR. KRESS: Okay, that is a --
DR. POWERS: That is only a step unless whoever coded up
HABIT has written the only code without error.
DR. KRESS: He gets a bottle of champagne if he does that?
DR. POWERS: Most of the codes with the notable exception
that maybe we will hear about later have physical equations in them that
have a certain universality of approval in them. The problem is when
you put it into a numerical vehicle for doing the calculations and have
to interface it with other things is where all the difficulties come
from.
I mean just having good equations is a necessary but not a
sufficient step for having a decent code.
MR. BASU: If you ask that question to Jay Lee, the previous
presentation, you have got an answer because I think it was during his
time that this model was put into the code --
[Laughter.]
DR. KRESS: I asked him but he uses a different model.
DR. POWERS: It is a new concept.
MR. SIEBER: Does the HABIT code include the effect of the
density of the released gas?
MR. BASU: Yes, it does.
DR. POWERS: It has density terms in it. It does not
include this ground-hugging kind of thing and it presumes ipso facto at
the beginning that it is a Gaussian release. It does not have the
transition from the release point to a Gaussian plume, which is where
the problem comes.
MR. SIEBER: That's right.
DR. KRESS: I don't recall them having density terms in it.
MR. BASU: I think it does.
DR. POWERS: I think it does.
MR. BASU: That will make a difference only for the
near-term -- if it does. It will make a difference in the near-term
concentration calculations, and I think I mentioned in the subcommittee
briefing that the HABIT model was compared with Wilson-Lamb and
Wilson-Chui models, which are non-Gaussian models for near-term
concentration, and they came out right on top of each other.
DR. KRESS: Very good.
MR. BASU: So it is really --
MR. SIEBER: So this would be useful for an on-site release
that is close to a control room intake?
MR. BASU: Yes. Let's see now. We are talking about
revision, toxicity limits. Again the toxicity limits that were listed
in Reg Guide 1.78 are taken from Sax's "Dangerous Properties of
Industrial Materials" that was published way back in 1968 based on the
knowledge of toxicity limits of however many chemicals we had at that
point identified or we could measure toxicity limits and if you want to
know the number of chemicals listed in Reg Guide 1.78 that is 27.
Now since then of course there has been a lot of work that
went on on toxicity or the toxicological area and which sort of
culminated among many documents into one document that I have cited
here, and that is the "NIOSH Pocket Guide to Chemical Hazards." In
1997 -- they update this guide every so often or every five years or so.
I think the last update was '97. We are proposing to use the
immediately dangerous to life and health exposure level that is included
in this NIOSH Pocket Guide and I have given you a definition of what
immediately dangerous to life and health exposure level is, and that is
basically that it is basically that it is the level which is likely to
cause death or immediate or permanent adverse effect if no protection is
provided within 20 minutes.
That is important, so if you subject an operator in this
environment, in the IDLH environment for 30 minutes without providing
any protection, then it is likely to cause --
DR. KRESS: If he dons his protective mask at 29 minutes is
he okay?
My next question is going to be how about 31?
MR. BASU: That is an interesting question. I don't know if
anyone is going to volunteer to do it, but I think in fairness to that
issue in the Reg Guide we are describing a two-minute donning time.
When you take two minutes in that context you have plenty of time to don
the protective gear and verily the next bullet says that because the
IDLH level can be tolerated for two minutes without any physical
incapacitation.
DR. POWERS: What I am hoping that we get a chance to do is
let the committee members attempt to don protective equipment in a 300
PPM ammonia environment in two minutes.
[Laughter.]
DR. POWERS: I think we had better move right along here and
not pursue that flippant remark because we need to move right along on
this presentation.
MR. BASU: I think I am going to skip the next viewgraph,
which is actually an example of comparison between the IDLH and the Reg
Guide --
DR. POWERS: Well, I guess we really have to understand both
why things are lower under Reg Guide 1.78 and the IDLH and why things
are higher.
DR. WALLIS: Why isn't there a criterion for how quickly the
operators have to be made aware of this. They have to know that they
are. If there is no physical incapacitation they don't know that there
is a toxic atmosphere unless something has happened. There has to be
some criterion for their awareness of that being appropriately quick and
everything.
MR. BASU: The alarm goes off or goes on, whatever -- I mean
one could detect, say, a certain level of concentration.
DR. WALLIS: There has to be criteria for detection and
everything.
MR. BASU: There are, in 1.78. If you recall, I said, you
know, that is an area that has not been revised but there are criteria.
Yes? Did you want me to project this?
DR. POWERS: Just explain a little bit to me. Let's take
acetaldehyde. You have Reg Guide 1.78 -- it says it has a 200 PPM; IDLH
is 2000. Why is the Reg Guide different from the IDLH or this the old
Reg Guide?
MR. BASU: This is the old Reg Guide -- based on the
information that was available back in '68 -- 2000 is the limit that is
prescribed for 30 minutes of exposure. In two minutes though you may
not have 2000 PPM in the control room.
DR. SEALE: What are you going to go to? I mean this is
your new -- the first column is your new listing?
MR. BASU: First column is the old Reg Guide, the existing
Reg Guide.
DR. SEALE: Okay. What are you going to go to in the new
one?
MR. BASU: IDLH.
DR. SEALE: IDLH.
MR. BASU: That's right.
DR. WALLIS: I'm sorry, but I have to ask you where in the
control room? If this is leaking in through a window or door or some
port, some leak, it doesn't immediately mix in the control room. It may
be, you may have 10,000 Parts Per Minute near the source and 10 at the
other end of the control room.
MR. BASU: You have 10,000 at the source level --
DR. WALLIS: Near the leak -- near the inflow -- and you may
have 10 on the other end of the control room.
MR. BASU: This is measured --
DR. WALLIS: It may take minutes to mix up.
MR. BASU: Yes, but the detection is at the control room
intake.
DR. WALLIS: How do you know where it is coming in if it is
leaking all over the place?
DR. POWERS: The ventilation to the control room. This is
what they are pulling into the control room.
MR. BASU: As long as there are intakes in the control room
and you have detectors placed at the intake.
DR. WALLIS: As long as you have your flows controlled well
enough.
MR. BASU: Okay, moving then on to regulatory analysis, and
I think that probably the first bullet answers something that you
raised, Dr. Wallis, chemical spill -- what has that got to do with the
core damage, so I think -- I don't know if that answers your question.
If the chemical spill or the release by itself is not going
to result in a core damage event, but the operator incapacitation may --
human error may at some point lead to a core damage event.
DR. WALLIS: Well, it could be that the system is so robust
that if all the control room operators die, the reactor is perfectly
safe. If this might have happened at TMI they would be better off than
they would be otherwise --
[Laughter.]
MR. BASU: They we don't need to worry about it, right?
DR. WALLIS: Do you not care about the death of the control
room operators?
MR. BASU: I think that kind of rings a bell that somebody
from the SES committee mentioned that we don't care about the control
room operators, do we?
DR. WALLIS: Is that true?
MR. BASU: No. No, of course that is -- we do care about
control room operators very much, otherwise the Reg Guide is not going
to be here, but I am just saying that the chemical release in itself is
not going to lead to core damage.
MR. SIEBER: Is there a probability that would tell us what
the relationship is between a toxic release greater than 1.78 and core
damage?
MR. BASU: There is an IPEEE insight report. It probably
has not come out for publication as a NUREG report yet, but it is in the
making and what I can tell you is that there are I think 78 plants
studied in that report and only one plant showed anything in terms of --
they looked at various attributes, if you will, for the core damage
frequency, like earthquake, fire -- these are external events -- and
then of course there was some classified as "other events," which
included the chemical release and so on and so forth.
There was only one plant that came out with a measure of CDF
for chemical release and that is actually less than 10 to the minus 7
and I think the cutoff is 10 to the minus 7.
DR. WALLIS: What is the incapacitation of control room
operators' probability as opposed to CDF? What is the probability of
them being incapacitated?
MR. BASU: That is also 10 to the minus 7 or less.
DR. WALLIS: So it is more likely that there is a CDF
relating from this, resulting from this than the control room operators
are incapacitated?
MR. BASU: No. That CDF was for -- no. I'm sorry --
MR. SIEBER: Go ahead.
MR. BASU: No, it said that the CDF that came out in the
IPEEE is due to chemical release and chemical release was associated
with operator incapacitation and what they did is they said once an
operator is incapacitated the probability is one that it will lead to --
DR. POWERS: The two are identical numbers?
MR. BASU: Yes. Now that may not be the case, but for the
sake of submitting the report, that is what they used.
MR. SIEBER: I guess that I would express my concern because
my personal experience, even though it is anecdotal, is that on-site
chlorine releases occur more frequently than these numbers would
suggest.
I am not aware of any significant one at a nuclear plant. I
am aware of them in coal fire plants and oil plants and water-treating
plants.
DR. POWERS: I think if you look at their analysis that they
have an appropriate measure for the frequency of that.
But then they multiply it by what's the probability of
incapacitated control room operators and that's what gets the numbers
down.
MR. SIEBER: Well, yes, our Cheswick plant, which is a coal
plant, they incapacitated the control room operators there, were lucky
to trip the plant. Put them in the hospital.
MR. BASU: I think the release -- onsite chlorine release is
also for nuclear plants also predominant amongst other release. Well, I
think we have to take it one step further. It's not just the release,
it's the release that will then raise the concentration level in the
control room to the point that it'll incapacitate an operator.
DR. WALLIS: So you're probably saying that a control room
in a nuclear plant is much better protective than his coal-fired plant?
DR. POWERS: That's true.
DR. WALLIS: So this has at least happened once in a coal
plant. So there's 10 to the minus 7 probability a thing has happened
once.
MR. BASU: In a coal-fired plant; yes.
DR. WALLIS: Were you aware of that?
MR. BASU: I didn't look at coal-fired plants.
MR. SIEBER: You see, I don't know where you would get the
data, either, because that kind of stuff, unless it results in serious
injury, would not be reported.
DR. POWERS: You can actually start -- there's actually a
mechanism to start getting it now, because the American Institute of
Chemical Engineers is starting to keep track of those things. But if
you wanted to go back five years, you're dead.
MR. SIEBER: Right.
DR. POWERS: They just -- things have in the chemical
business, chemical risk analysis has vastly improved, starting with the
Bhopal accident, but it didn't really get serious until there was a
fairly dramatic event that occurred in Philadelphia a few years back,
and now there are institutes that are really trying to do chemical risk
assessments and what not, and they're even now being required in
California and a few other localities. But, you're right, the data base
on this is really anecdotal, you know, I remember when.
MR. BASU: So anyway, the bottom line from the previous
slide is that the baseline risk is in the noise range. And so now we're
going to talk about the change in baseline risk because of this proposed
revision, and we're saying that factors that will potentially affect the
baseline risk are the revised toxicity limits and the revised
atmospheric dispersion modeling. And the other value at revised
toxicity limits may be viewed as relaxation, because you see in that
illustrative viewgraph that for some chemicals anyway that the limits
are higher than the Reg Guide 178 limits, but keep in mind that the
operators are not going to be necessarily subjected to the IDLH values.
They are going to have the opportunity to don their protective gear
within two minutes, and the IDLH values may not be reached for many
chemicals within two minutes. So change in baseline risk if any on the
revision would be minimal in the worst case.
Use of atmospheric or improved atmospheric dispersion models
will likely result in some changes in concentration levels in the
control room. For some chemicals, the levels may be higher. For
others, they may be lower. But even for those chemicals, and if they
happen to be higher, the risk is only going to be imposed -- additional
risk is only going to be imposed if either the IDLH values are exceeded,
because that's the reason that we are setting, or if the protection is
not provided within the prescribed time, and this is two minutes. So I
think the control room operators are not afforded any protection or the
IDLH values, then you're going to see additional risk. So given those
arguments, we're saying that the total change in baseline risk with the
proposed revision is not considered significant.
DR. WALLIS: The measure is change in CDF?
MR. BASU: Yes.
DR. WALLIS: So dose CDF is 10 to the minus 6, and that's
what you --
MR. BASU: Hopefully it will be less than 10 to the minus 7.
Baseline is 10 to the minus 7, we don't want -- well --
DR. WALLIS: You've made a conclusion that it's not
significant, and it's based on some numbers.
MR. BASU: The cutoff number is 10 to the minus 7, yes.
Well, if the change in risk is not significant, at least if we are able
to conclude that far, then what I'm showing here in this viewgraph is
that you are likely to have some burden reduction, by combining 178 with
195 you are going to have some administrative burden reduction on the
part of licensees as well as on the part of NRC review process.
Revised guidance is likely to reduce plant shutdowns from
spurious actuation of detection systems, and in one of the viewgraphs
earlier you have seen some examples where licensee amendment
applications came because there were some spurious actuations of
detection systems, whereas the hazardous chemicals were not detected,
not at least at the level that would be a concern. So by reducing the
plant shutdowns, you are definitely going to reduce burden. It
certainly will result in consistent and comprehensive control room
habitability evaluation.
I think there was a question or comment in the subcommittee
briefing about the list of chemicals not being exhaustive, and I do
agree it's not exhaustive, but it is certainly much more comprehensive
than 27, because we have listed 406 chemicals in this revision. And as
and when more chemicals are included in the NIOSH pocket guide, they
will be included also by reference.
The atmospheric dispersion model will provide best
estimate --
DR. WALLIS: What do you mean by best estimate?
MR. BASU: It certainly will provide you more realistic
computational --
DR. WALLIS: I would think it would be better to say
improved dispersion will provide better estimate? No way that you can
say it's best without some kind of criterion for best, and you get into
a morass there.
I think what you're saying is that this is to provide a
better estimate, more confident estimate, you think, and this probably
adds to public confidence, because you've got better representation to
what's happening physically. Best estimate --
MR. BASU: Has a different connotation. I agree with you.
It's a wrong choice of --
DR. SEALE: If it's a morass, he has a lot of company.
DR. WALLIS: So there are some -- this is just your sense
that the dispersion model is better, or is there some measure of its
being better?
MR. BASU: The comparison between Murphy-Campe model and the
improved model and a comparison with the experimental data and the
improved model has done significantly better in terms of predictive
accuracy.
DR. WALLIS: I think this does help at least professional
people's confidence. I don't know if it's public confidence.
MR. BASU: Yes. I don't need to dwell on this one. I just
want to emphasize that there is a plan to issue this draft revision to
reg guide for public comment by the end of this month, so there is a
line item there for resolution of ACRS comment, that is, if ACRS chooses
to provide comments to us in a fashion that we can incorporate them
before releasing it for public comments, we'll do so. If not, there is
always going to be an opportunity to provide comments even after
releasing for public comment. So it's your call.
And I think that about does it.
DR. POWERS: When you say you now have 480 chemicals --
MR. BASU: 406.
DR. POWERS: 406 chemicals listed, and I look at Table 1 in
the Draft Guide, I see 27.
MR. BASU: Well, the Draft Guide, that is exempt in the
Draft Guide. The Draft Guide makes reference to a NUREG/CR Report 6624
and that lists 406 chemicals. Does that answer your question?
DR. POWERS: I suspected that was the answer.
DR. WALLIS: I am interested in something you didn't talk
about, which is accidents per mile for barges seem to be sometimes worse
than for trucks. It doesn't seem to be consistent.
MR. BASU: Say that again.
DR. WALLIS: Oh, you are saying the spill probability is
less, even though there has been an accident. I see.
MR. BASU: Yes. Yes.
DR. WALLIS: Okay. Sorry.
MR. BARTON: I am concerned that this guide has contained
within it some requirements on licensees that I don't think are
appropriate, or this is not the way to impose requirements on licensees.
For example, on page 2 of the guide, under "Discussion," you talk about
the fact that "This guide does not address protection of individuals
outside the control room." However, you say, "Breathing apparatus
should be provided and readily accessible throughout the plant in order
to eliminate the need for personnel to seek shelter in the control room
during release." That tells me you have got to provide SCBAs for
everybody onsite so they don't all run to the control room because that
is the only place onsite that is safe during a toxic release.
And that -- I can just see inspectors going and saying,
well, where is all these SCBAs for all these 600 people that are onsite
when you have got a chlorine release? So I think that is a problem.
MR. BASU: Chris, would you like to respond to that
question?
MR. PARCZEWSKI: I think you are prepared to answer his
question.
MR. BASU: I think you are raising a valid concern. I don't
know, though, if there is a data point in all the LERs submitted, LER
database, data event report and whatnot that you would submit to that
schematic.
MR. BARTON: Well, I think you need to do something to take
that out here, which could be implied to be a requirement. You have got
the same problem on page 11, Item -- Section 9, Isolation Systems, where
you make a statement that says, "A control room exit leading directly to
the outside of the building should have two low leakage doors in
series."
I can name a whole bunch of plants where you walk out of the
control room and you are in an office building, administrative building
or turbine building and you don't have any airtight door other than the
control room door itself. Now you are saying -- now you are changing
the design of all the plants where they have got to have two airtight
doors in series once you exit the control room.
MR. BASU: I don't know if I am going to be able to respond
to your concern. I might just say, though, that this is one of the
challenges when you combine two different Reg. Guides written at two
different times, at two different -- or different orientation, and try
to import relevant features of one into the other. That particular
statement was taken out of Reg. Guide 1.95 and imported into this
revision.
DR. BONACA: But to me the examples suggest that certainly
there should be a reading.
MR. PARCZEWSKI: Parczewski. Reg. Guide is only
recommending -- gives only recommendation, there is no requirement. I
agree, maybe the language should be changed.
MR. BARTON: I have lived on the other side of the fence,
and don't tell me a Reg. Guide is only a recommendation when the
inspector comes and says, you know, you guys signed up for Reg. Guide
1.78, now where is your two watertight doors? You know, stuff like this
is dangerous putting in the Reg. Guide because it is -- it can be
interpreted to be a requirement.
MR. PARCZEWSKI: Maybe the language should be changed.
MR. BARTON: Well, I agree.
MR. PARCZEWSKI: Maybe instead -- state less definite
statement maybe.
MR. BASU: Less prescriptive, less regulatory language.
DR. BONACA: As a minimum it seems to me that this Reg.
Guide should have a thorough reading to see what kind of requirements it
identifies or suggests that do not exist right now in regulation,
because these are just two examples.
MR. BASU: I think I will echo what Chris said, that the
Reg. Guide is meant for guidance and not sort of requirement. But I do
sympathize with your comment.
DR. APOSTOLAKIS: So you are not going to change it, is that
what you are saying?
MR. BARTON: He is not going to change it.
MR. BASU: No, no, no. I mean it is not that we are not
going to change it. There is always -- this is going to be published
for public comments and we will have to incorporate whatever we receive.
MR. BARTON: Well, you just received a comment from the
public.
MR. BASU: Yeah.
[Laughter.]
MR. BASU: Good. So page 2 and page 11. Is that -- I'm
sorry. It is not page --
MR. BOEHNERT: Twelve.
MR. BASU: Page --
MR. BARTON: Page 11, Item 9, Section 9.
MR. BASU: Section 9.
MR. BOEHNERT: The second paragraph.
DR. POWERS: Are there any further comments that members
would care to make on this Regulatory Guide and presentation?
DR. SEALE: I have a comment, it is a more general one,
though. We have heard that these are the kinds of things that are
difficult to assess because the CDFs are suppressed, if you will, in the
sense that the difficulty lies with other things besides core damage,
worker peril and so on. And the risk from 1.78 are low. Why does it
get NRR endorsement as a task item?
DR. POWERS: Well, I think that is what we began this
discussion with. Why are we doing this?
DR. SEALE: That's right. And what I am saying now is,
instead of asking Research to prioritize their research program, perhaps
we ought to ask NRR, for their set of priorities for task orders that
they send to Research, what is the basis for putting this thing at the
level it is relative to some of the other things?
DR. POWERS: Good point. Any other comments?
[No response.]
DR. POWERS: Thank you, Dr. Basu.
MR. BASU: You're welcome. Thank you.
DR. POWERS: We are going now to turn -- I think we have
come to the end of our need for a transcript.
[Whereupon, at 4:42 p.m., the meeting was recessed, to
reconvene at 8:30 a.m., Thursday, September 2, 1999.]
Page Last Reviewed/Updated Tuesday, July 12, 2016