460th Meeting - March 11, 1999
UNITED STATES OF AMERICA
NUCLEAR REGULATORY COMMISSION
***
MEETING: 460TH ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
(ACRS)
U.S. Nuclear Regulatory Commission
Conference Room 2B3
Two White Flint North
Rockville, Maryland
Thursday, March 11, 1999
The committee met, pursuant to notice, at 8:30 a.m.
MEMBERS PRESENT:
DANA A. POWERS, Chairman, ACRS
GEORGE APOSTOLAKIS, Member, ACRS
JOHN J. BARTON, Member, ACRS
MARIO H. FONTANA, Member, ACRS
DON W. MILLER, Member, ACRS
THOMAS S. KRESS, Member, ACRS
ROBERT L. SEALE, Member, ACRS
WILLIAM L. SHACK, Member, ACRS
ROBERT E. UHRIG, Member, ACRS
GRAHAM B. WALLIS, Member, ACRS
MARIO V. BONACA, Member, ACRS. PARTICIPANTS:
MR. BOEHNERT
MR. MARKLEY
MS. MCKENNA
MR. MATTHEWS
MR. BELL
DR. LARKINS
MR. NISSLEY
MR. NEWBERRY
MR. AKSTULEWICZ. P R O C E E D I N G S
[8:30 a.m.]
DR. POWERS: The meeting with now come to order. This is
the second day of the 460th meeting of the Advisory Committee on Reactor
Safeguards.
During today's meeting, the committee will consider the
following: the proposed rule for event reporting requirements,
reevaluation of the generic safety issue process, licensing framework
for fuel burnup extension, and NRC participation in the Cabri Reactor
Fuels Research Program, and we will be working on an extensive number of
ACRS reports.
The meeting is being conducted in accordance with provisions
of the Federal Advisory Committee Act. Mr. Sam Duraiswamy is the
designated federal official for the initial portion of this meeting.
We have received no written statements or requests for time
to make oral statements from members of the public regarding today's
session.
A transcript of portions of the meeting is being kept, and
it is requested that speakers use one of the microphones, identify
themselves, and speak with sufficient clarity and volume so that they
can be readily heard.
Do any of the members have any comments they would like to
make at the opening of the session?
Fine. Let's move on to our first topic, which is proposed
rule for event reporting requirement.
Mr. Barton, I believe you are the cognizant member on this.
MR. BARTON: Thank you, Mr. Chairman.
The purpose of this session is to hear presentations by and
hold discussions with representatives of the NRC staff concerning the
proposed rulemaking to modify the event reporting requirements for power
reactors which are contained in 10 CFR 50.72 and 50.73.
The staff has issued an advance notice of proposed
rulemaking, held public meetings to discuss the advanced rulemaking last
August and has conducted tabletop exercises with the industry.
The objectives of the proposed rulemaking are to better
align reporting requirements with the NRC's reporting needs, to reduce
reporting burden on licensees consistent with NRC's reporting needs, and
to clarify reporting requirements where needed.
The committee intends -- is expected to prepare a report on
this matter at the closure of this -- prior to closure of this 460th
session.
At this point, I'll turn the meeting over to Scott Newberry
of the staff to make a presentation.
DR. POWERS: Before the presentation starts, let me say that
in examining the materials that were sent to us, I had a hard time
understanding how we had defined what the NRC's current reporting needs
are. So to the extent that you can clarify that for me, I would
appreciate it.
MR. NEWBERRY: Let me start out -- Dennis Allison to my
right is going to do the briefing. I am Scott Newberry. I'm in a new
position right now. I've been there about four weeks. I'm now David
Matthews' deputy in the division that has a number of projects where we
--
DR. POWERS: Should we congratulate you or pass on our
condolences?
MR. BARTON: Condolences are in order.
MR. NEWBERRY: As you wish. But we have a number of
projects, as you know, where we're reexamining how we're doing business
-- yesterday, 50.59, today, reporting requirements, and there's quite a
few others.
But Mr. Chairman, we would be glad to answer that question.
I was reviewing the viewgraphs last night. Your question is not
answered in the viewgraphs, and I think, Dennis, if you make a point, we
need to talk about why we need reports, what the objective and what the
use of the reporting information would be, and we would be glad to get
into that in the briefing. It's a good question.
But just to amplify on the context of where we are, we're in
a situation where we have a schedule of issuing a proposed rule early in
April; that is, getting it to the Commission early in April. I would
like to meet that schedule, if we can. That is a proposed rule.
I think we aren't quite done yet. We'll tell you where we
are today. We've tried to be responsive to the comments we have
received as a result of the ANPR. So --
DR. POWERS: You might remind members what ANPR stands for.
MR. NEWBERRY: Advanced Notice of Rulemaking. So we put a
proposal out into the arena and have a considerable feedback on that
proposal; yet, we're still going out with a proposed rule. So Dennis
will talk to you a bit about the comments we have received.
What we're talking about here is a reduction in reporting
requirements. I mean, that's pretty much what it is across the board.
And so the debate one has in-house here is, are we going too far? You
know, that's --
DR. POWERS: Well, it seems to me that equally well is, have
you gone far enough?
MR. NEWBERRY: Well, primarily, the conversations I've
participated in --
DR. POWERS: I bet they have, but I with raise the other
issue.
MR. NEWBERRY: Okay. We will look forward to that feedback.
And with respect to that, one of the things that we're
looking at is, you know, reporting rule is not our only source of
information. Yesterday, we talked about 50.59. There's an obligation
for utilities to come to the staff for reviews when necessary, and we're
talking about that threshold.
We have a new assessment and a baseline inspection effort
where we're going to get some much better information, and so as I go
through these discussions, I think -- as I think about the issues, I
think about this not being the only source of information, and I think
that's significant.
So I think I'm going to get out of Dennis' way now. I had
hoped to be in a shape where I could have told you, you know, what
Dennis is putting in front of you is the staff position. A snowstorm
prevented that. We've still got some work to do, to talk with our
internal stakeholders, to look at some of these issues that Dennis I
think will mention, where we aren't quite there yet. But I think we're
still planning on meeting our schedule.
So Dennis.
MR. ALLISON: Thank you, Scott.
My name's Dennis Allison, and I'll ask the reporter if you
can hear me all right? Okay.
Okay. This is -- I have passed out copies of these slides,
and I may stop showing them sometime during the presentation, but the
first one is just the title sheet. I think I'll stop right now and just
talk.
Okay. On page 2, you have the objectives laid out, which
Mr. Barton has already read to you, and they're simply to better align
it with what we need reported by this mechanism at this time.
The second is to reduce the burden consistent with what we
have to know, what the NRC needs to know.
The third is to clarify where needed.
Some examples are listed there on the slide, and I would
note just one thing. You can shuffle those examples around, you can
take some of the examples under two and put them under one instead and
so on. They're not mutually exclusive objectives.
On the third slide, I've provided the schedule, and that's
to publish the proposed rule May 14th, conduct a public workshop very
shortly thereafter, May 28th, and at that time, the public, including,
of course, the industry representatives, will have had two weeks to look
at a proposed rule, come to the workshop and talk about -- we get a lot
of good out of these workshops because it's a lot easier to have an
interchange in that format than it is comments and formal responses.
So we'll do that early in the process, and that will allow
the industry to get a leg up on making their comments and the public.
David Lochbaum from the UCS probably -- he might come, I don't know, but
he has shown some interest in this.
Then the comment period ends 77 days after publication, and
we'll address the comments, provide the final rule package to the CRGR
and the ACRS in October with a goal of publishing it in January of 2000.
On the next slide, number four, this is just a list of what
I considered to be the significant or the principal changes that are in
the package, and I'm going to discuss each one of these in subsequent
slides, but I'll just read the titles now: Initial reporting times;
design issues; late surveillance tests; unanalyzed condition; condition
outside coverage of procedures; and system actuations.
Okay. Then getting into the first slide about required
initial reporting times, what we have proposed in the package is that
one hour reports -- and actually, I should say -- I want to just stop
and point out the requirement is really to report immediately after the
declaration of an emergency class, and in no case more than one hour.
So -- but these are -- the one hour reports then would be limited to
declaration of an emergency class or an emergency action taken pursuant
to 50.54(x), and the reason 50.54(x) is still there is what that rule
allows licensees to do is to take an emergency action contrary to the
technical specifications if they can see no way to provide protection to
public health and safety and comply with the tech specs. So clearly,
that's something, if it arises, the NRC ought to know about right away.
DR. SEALE: What is the NRC's take on a situation in which a
utility might declare an emergency class and then, before the end of
that first hour, determine that in fact that declaration was
inappropriate? Is there a withdrawal mechanism or do you have to go
through the full range of agony associated what the reporting and so
forth to clear that issue?
MR. ALLISON: Well, there's not a withdrawal mechanism, but
the agony associated with reporting an emergency class is not great.
You have to pick up the telephone.
DR. SEALE: Yes.
MR. ALLISON: You don't even have to dial it. It --
DR. SEALE: I understand.
MR. ALLISON: And you have to say, I declared an emergency
class at the following time for the following reason. And it's
perfectly acceptable to say, not only that, that was a mistake, and I
undeclared it 20 minutes later.
DR. SEALE: Okay. Okay.
MR. ALLISON: But it's not an agony.
DR. SEALE: Okay. Well, I think that's important because
you can see people dancing to decide whether or not they want to do it.
MR. ALLISON: But once they declare it, it becomes
reportable.
DR. BARTON: That's the way it is now, and that's the way
licensees implement it.
DR. SEALE: Once they do it, it's important.
MR. ALLISON: Well, the rule requires reporting if you
declare an emergency class.
DR. SEALE: Okay. Once you've reported it, though, then
it's -- but you can withdraw it before the report.
MR. ALLISON: Sure.
DR. SEALE: Okay.
DR. BARTON: If you withdraw it as soon as the plant
conditions are such that you get out of that situation.
DR. SEALE: Sure. I understand.
DR. BARTON: And you just report I'm out of it.
DR. UHRIG: It really doesn't make any difference whether
it's before or after the one hour is up, really.
MR. ALLISON: That's correct. There are all kinds of things
that can happen, but you could change your mind in ten minutes --
DR. BARTON: Right.
MR. ALLISON: Before you've called the NRC.
DR. BARTON: Right.
MR. ALLISON: And you could just report both things.
DR. BARTON: Right.
MR. ALLISON: At the same time.
DR. BARTON: Right.
MR. ALLISON: Okay. So this one-hour reporting is intended
to capture things where immediate NRC action might be necessary, like
fully activating the response center or -- with a lot of these things
tending up toward the alert category we will partially man the operation
center and follow the event, although we don't fully activate.
Then the next category, all of the other 50.72 reports have
been placed in an eight-hour category, which is as soon as practicable,
and in no case later than eight hours after the occurrence of the event;
50.72 by the way is based on the occurrence of the event, not on the
discovery of it. That is the timing.
Okay. This is intended to capture events where NRC action
might be needed, but not immediately. So something we need to do today,
like call the plant, find out more, understand what really happened
here, is there a need for a special inspection, should we request the
licensee to quarantine equipment or not start up or what have you until
it can be figured out what happened. And so that's those kinds of
things.
This also -- 50.72 has a second purpose besides taking
immediate action, and that second purpose is to obtain information to
alleviate heightened public concern. And this then captures things like
low-level radiation releases, which is a concern, of course; injured,
contaminated workers transported to an offsite medical facility; press
releases, reports made to other agencies, and things of that nature.
Now on the eight-hour reports, there are two issues that
we're still discussing in the staff. The first one is that part of
this, if we go ahead with this, overrides some requirements of Part 2.
There are two paragraphs in 50.72 that say report a certain kind of
release. Oh, I'm sorry, there's one in 50.72 and one in 50.73. You
report low-level releases at the following level. And if you do this,
you have satisfied the requirement of paragraph (X) in Part 20. That's
proposed to be an eight-hour report. In Part 20 the report's required
in four hours. So we're talking about how to address that conflict.
And that's the simpler thing. That's really a mechanical
problem, I think.
The second issue that we're talking about is whether eight
hours is too long for some of these reports. And people have raised
concerns about the kind I was just mentioning, low-level releases,
injured worker transported, things of that nature. There's a concern
that perhaps the NRC ought to know about these things sooner than eight
hours.
And then another kind of event is some of the other events
which are not there for public concern but for possibly safety, like
safety injection into the reactor coolant system due to a valid signal.
Some of these reports are considered to be more urgent than others. So
we're talking about which ones -- whether eight hours is really soon
enough for some of these reports.
DR. BARTON: What is your concern here, that on an injured
contaminated worker, since the licensee may have an obligation to report
local or goes to a local hospital that you're not going to know it for
eight hours but it would be out there for the public? There would be
public knowledge of it? Is that your worry?
MR. ALLISON: In that class of events, yes. Right. With
the contaminated worker or the low-level releases.
DR. BARTON: Right.
MR. ALLISON: Because, you know, if the releases are
significant enough that we might have to act they'd be in the emergency
class.
DR. BARTON: Right.
MR. ALLISON: So that's -- yes, that's one concern, is just
knowledge in that area.
And the other concern about the safety injection, that's
more is that an event that we need to react to faster than eight hours.
DR. UHRIG: What's the difference in the type of event
that's required to be reported under Part 50 and Part 20? Why the
four-hour and eight-hour difference?
MR. ALLISON: Well, that's the problem.
DR. BARTON: Different people wrote the different rules.
MR. ALLISON: That's the problem. See, right now there is
no difference because they're both required in four hours. But in this
proposal, the way it was drafted up --
DR. UHRIG: I see.
MR. ALLISON: It's just changed it to eight hours. So the
question is well, what's going on?
CHAIRMAN POWERS: What was the rationale for the four hours
originally?
MR. ALLISON: In Part 20? I don't know, but it was probably
consistency with the 50.72.
DR. UHRIG: It's just four hours -- it is still four hours.
MR. ALLISON: It's four hours right now.
So that's kind of a -- shoelaces were tied together or left
hand and the right hand. But we need to address whether they should be
consistent or not. That's what we're talking about.
DR. BARTON: Well, your feeling is that eight hours is
adequate.
MR. ALLISON: Yes, that's what we've proposed, and -- but we
don't have all the staff concurrences that we need.
DR. UHRIG: Does the staff feel that certain events should
be reported as soon as possible even though you don't have all the
details, or is it better -- would you rather that they get more details
before they send in a report?
For instance, a person is injured, you don't know whether
he's contaminated or not. Is it better to wait until you know, or is it
better to report and then have a followup from your standpoint?
MR. ALLISON: Well, in all of these 50.72 reports I would
say it's better to report and then follow up. That's -- but in the case
of the injured worker, I guess that we -- one option that we might
choose to deal with this is to emphasize that the requirement is to
report it as soon as practical, and not more than eight hours. And --
DR. UHRIG: Well, that one has to do with public relations,
because the hospital -- the word is going to get out from the hospital.
MR. ALLISON: That's right. Yes. And --
DR. BARTON: Or it would get out as soon as that ambulance
heads down the road to the plant.
MR. ALLISON: Yes. That's heightened public concern. And
people will call and say what's happening.
Okay. On the 60-day reports, those are the LER, of course,
and that includes these things here, but it also would include other
things where the NRC might need to review the event in order to learn
the lessons of operating experience basically.
Okay. This is a good place to talk a little bit about the
NRC's reporting needs. I guess the purposes of these rules would not be
changed by this rulemaking. So when I showed the objectives, that's --
those are the objectives of this amendment, but they're not the
objectives of the rules.
There are two ways to define -- two ways that I see to
define the NRC's reporting needs. One is the objectives that are
stated: Is immediate action needed by the NRC to protect public health
and safety? And I would say that in the broad sense. So the eight-hour
report we might be talking about a special inspection or something. But
we need -- the NRC may need to do that.
The second is where accurate information is needed to
respond to heightened public concern. We've talked about that. Those
are the objectives of 50.72.
The objectives of 50.73 are basically to learn the lessons
of operating experience. That is, the purpose of the rule is to require
the reporting of events that are considered significant and useful to
the NRC in its efforts to identify and resolve problems, and they're
also suitable for other studies. So the objectives stand the same as
they were in '83, and my summary of it is that's to learn the lessons of
operating experience.
The other way that I look at the NRC's reporting needs is
this way, that is, what are we going to do. And for the one-hour
reports there are things that we would want to activate, the response
center, or partially activate, or take some action very soon. The eight
hours -- it could be four, it could be 12, it could be six -- but those
are things that we want to take actions that the NRC ought to take
today.
MR. ALLISON: The third is just we ought to -- this event
should be reviewed in due course. So those are the two ways that I have
of looking at it. And the industry has urged us in their comments to
think about what, in all regards, what is the NRC going to do with this
report, both about the timing and whether it is needed at all.
DR. POWERS: I guess I am curious of what the empirical data
on past experience shows. When you have gotten what was a four hour
report, did you do anything for four hours or not?
MR. ALLISON: Yes, sometimes.
DR. POWERS: Well, now, the partially manning the emergency
center is probably an easy thing.
MR. ALLISON: Right.
DR. POWERS: Because they are geared up to do that. Okay.
It is the subsequent steps that I am interested in.
MR. ALLISON: Well, --
DR. SEALE: Well, go ahead, finish. I'm sorry.
MR. ALLISON: Okay. Well, I mean it is not that common, but
among -- because we get a lot of four hour reports, but for some of
those, some of those are like complicated trips, or loss of HRH cooling
capability, things that we are concerned about both from -- now, yeah,
we might be concerned about following that, but the objective of the
four hour report is not really to man the operations center, it is to,
in my view, it is -- is it something we need to do today? And that
frequently happens. I'm sorry, not frequently, it is rare, but --
DR. POWERS: I guess I am asking, what is the history on
those events that you have gotten reports in the past, and what are the
activities that took place? And, similarly, when the report was made,
did it constitute a burden to the licensee at that time? I mean you
have one guy that, I presume, picks up the phone and calls you, and it
is a little hard to imagine one guy out of a hundred makes that much
difference, but maybe it did. Maybe the decision making process that
you have to make this call in this active period of time was, in fact, a
burden, in these historical events.
MR. ALLISON: I just haven't done that study.
DR. POWERS: Well, I think you need to have some empirical
data here to make decisions on how you categorize things, because, in
fact, if the NRC did nothing except things it is geared up easily to do,
but, in fact, said, okay, well, let's -- we will talk about it on Monday
when everybody is here, and then we will decide whether to send out this
special inspection team or not, then you know that there are classes
where you didn't need an eight hour report. Otherwise, there is a
strong speculative component to categorizing issues.
MR. ALLISON: Without the empirical study?
DR. POWERS: Yes.
DR. SEALE: To put it another way, what do you lose if you
give up the time between four hours and eight hours? Were there actions
you took that presumably would be unavailable to you if you waited until
eight hours?
MR. ALLISON: True.
DR. SEALE: And did they make a dime's worth of difference?
MR. ALLISON: Yeah, you have got a half-full/half-empty
situation here.
DR. SEALE: That's right.
MR. ALLISON: That's right.
DR. SEALE: Exactly. Exactly.
DR. BARTON: You know, I would tell you from personal
experience that what it would affect, if you change it from four to
eight, it would delay NRC's decision by four hours on whether they want
to assemble an AIT, and that is what would be the impact on the NRC.
DR. SEALE: Okay.
DR. BARTON: On a utility, by extending the four to eight
hours, it would enable a more accurate report because you would have
more facts and figures, et cetera, et cetera.
DR. SEALE: You would take the noise out of the process.
DR. BARTON: Yes.
DR. POWERS: So, let me ask --
DR. BARTON: It helps on one hand, the licensee, and you can
also see where it helps the decision making on the part of the NRC by
doing a shorter report?
DR. POWERS: Well, if you have given me eight, I am going to
ask now for 24. I mean doesn't that just make it all the more both
ways.
DR. SEALE: Sure.
DR. POWERS: And as soon as you give me 24, I am going to go
for seven days, okay.
DR. BARTON: Maybe an annual report.
[Laughter.]
DR. SEALE: Let me ask --
DR. POWERS: Once I get seven, I am going to go for a
quarter.
DR. SEALE: Well, let me ask a slightly different question
here. You mentioned that you had run across this inconsistency or this
necessary -- the necessity to do some remediation with respect to Part
20 if, indeed, you made this change. Have you gone through and
systematically looked at the whole body of regulation to see if there
are other reporting requirements that you might have similar problems
with?
And I can tell you, last week I attended a briefing that the
Commission, one of the Commission offices had with some states and
people from other federal agencies, FEMA and so forth, on Dr. Kress'
favorite, potassium iodide issue. And it turns out there, pre- or
post-administration is a serious question in terms of what is to be
attempted with KI and, furthermore, that the problem is really a, quote,
state problem and not a national problem in terms of the community
relations. And, certainly, you have an awful lot of fingers sticking
out on this thing that can get mashed if things turn out wrong, so you
want to be very, very careful that you don't run into difficulties with
your state constituencies and so on.
It seems to me a pretty systematic run-through on what these
other reporting requirements might be could be pretty helpful right now
to let you know how many strings you are pulling whenever you make these
changes.
MR. ALLISON: Well, I haven't personally reviewed all the
regulations, but NMSS has looked at a bunch of them, and I know there
are quite a few other reporting requirements that some say to report
immediately, and immediately is not defined. Some say to report within
one hour, some within four hours, some within 24, some within two
working days, and so on.
Now, there is a separate project associated with this. The
Commission told us to ask for comments on other rules that could be
improved, or made more risk significant, or clarified, and some of those
other comments captured a few of the rules you are talking about. They
said you should change Part 20 to make it consistent with this, or you
should change this other requirement that is not --
DR. SEALE: Yeah.
MR. ALLISON: -- even directly tied to 72, but if immediate
-- why isn't immediate eight hours, or one hour, or something there? So
that is another project that we are going to deal with after we make
this change.
The two paragraphs that I mentioned in Part 20, though, are
kind of a direct conflict. We would have two separate requirements with
different timelines for the same event. The others are a more general
question, should we define what immediate means? And should it be one
hour, or eight hours, or what? But we will have to deal with that, it
is a different project.
Okay. Going to the next slide, significance levels for
design issues. Now, when the Commission approved the rulemaking plan
for this rulemaking and directed the staff to consider using criteria
similar to those that were used by the Millstone Special Projects Office
in sorting design issues out according to significance, and these are
those criteria. The Millstone project people used it to sort things
out. So, outside the design and licensing bases and the system can't
perform its safety function, that is the first -- that is the highest
level. The second level is the same thing, but it only applied to one
train of a multi-train system. The third is that is something is
outside of its -- or the system does not meet its design and licensing
basis, but it is still able to perform its function. And the fourth is
things that don't affect anything. So those are the levels.
So in the next slide, this proposal is related to those
levels. The first level is reflecting -- well, first, I should say, I
think you saw it, the mechanism that is used in this package is to
delete the requirement for a report, an event where the plant is outside
the design basis of the plant. That is the current requirement. That
would be deleted, but the reporting of design basis issues would not be
eliminated because of what is on this slide. You would have to make ENS
notification and an LER, submit an LER for a design or analysis defect
or deviation, or anything else, by the way, it doesn't have to be a
design issue. It could be an equipment problem, it could be operator
error or what-have-you, if it is significant enough that the system
can't perform its specified safety function.
And so a paraphrase of the proposed rules is given in
italics here, and ENS is required, for an event or condition, that at
the time of discovery -- and at the time of discovery is a new feature
in this proposed rule, but at the time of discovery could have prevented
fulfillment of the safety function. And you have to submit an LER for
an event or condition that, during the previous two years, and during
the previous two years is also a new element in this proposed
rulemaking, could have prevented fulfillment of the safety function.
Now, we go down to the second level where only one train of
a multi-train system is affected, and this event would be reportable as
an operation or condition during the previous two years, again, but it
is an operation or condition prohibited by the plant's technical
specifications. And the idea there, of course, is if it is a design
thing, it is going to last longer than the allowed outage time, so you
will have had one train out of service, or inoperable for longer than
allowed by the tech specs.
So that is the way design issues would be reported.
MR. BARTON: What is the point about the previous two years,
which is the new requirement?
MR. ALLISON: The previous two years was in response to a
public comment where licensees commented that we should put some kind of
a statute of limitations on one of these criteria. They only asked for
it I think on Number 1, but I thought it applied equally well to Number
2.
We need a statute of limitations because they will come up
with an event where you find out that you might have had two trains out
of service at the same time if certain things occurred. Now if in most
cases it is immediately obvious based on knowledge and experience that
this did or did not occur, but it may be a rare circumstance and the
operators just don't know if it occurred, so they have to search through
operating records to find out.
To search 30 years of operating records to see if some rare
event -- that's just not warranted, so the past two years, which is an
operating cycle seemed right.
MR. BARTON: Okay, so it is tied to an operating cycle,
basically?
MR. ALLISON: Right. Two years would cover at least an
operating cycle, so that was the thinking there.
A couple of the issues we are talking about are with regard
to design basis. One issue is that we sometimes get LERs that indicate
a component is outside of its design basis but that didn't make a train
or a system incapable of performing its function, so everything was
operable, but there is something about a component that it can't deliver
what it is supposed to do. Maybe it is an MOV that doesn't deliver
enough torque or whatever, but everything can still perform its required
function.
We get those reports sometimes. We certainly don't get all
cases of component problems reported but we get some, and those are
useful to the Staff who are following the component problems, people who
keep track of what is happening with pumps and valves and so on.
The second issue that we are talking about here -- now there
are other sources of component problems that we can tap, so that is the
discussion there.
The second issue is just a concern that if the licensee
finds everything operable, although something might be outside of its
design basis, he can then continue to operate until the next refueling
outage without reporting it, and could we be missing something there
that we would want to know about? That is the second concern.
That is the discussion about design basis that is going on.
On the next page, page 8 --
MR. NEWBERRY: Just to pause there, that is a significant
point. We are still talking with Staff about that. Dennis has even got
some data about how many -- this is an area where he does have empirical
data and we are looking to see what the true experience is, how many
hundreds of reports have we gotten that just are not significant or they
are, and there is another criterion I don't think you have mentioned yet
in terms of backing this up in terms of a degraded, seriously degraded
situation.
Was that on the slide? But anyways, the idea is to really
look at the safety significance of those reports versus learning from
operating experience and to come up with the proper balance there. We
are getting good feedback from our technical folks and we are going to
have to sort through it.
CHAIRMAN POWERS: One of the thrusts or comments that this
Committee has made several times is that the empirical data constitute a
resource that can support the Agency's move toward risk-informed
regulation.
MR. NEWBERRY: Yes, sir.
CHAIRMAN POWERS: And you have got to balance that need
against the burden that you impose both on yourself and the licensee on
getting reports that are -- on the surface of them result in no action.
MR. NEWBERRY: This is an area where I say this isn't the
only source of information. If a plant is outside the design basis they
have an obligation under other processes as well as the oversight
process where we are getting data on performance and there is going to
be a risk-informed baseline inspection to look at certain things where
the indicators are not covering what you are interested in so we are
looking at that right now and this is probably one of the more
significant issues we are struggling with before we get the proposed
rule out.
MR. ALLISON: Okay. On Slide 8, late surveillance tests, in
the past we had constructed a logic that said if you fail to perform a
surveillance test on time, and that goes on for longer than the allowed
outage time for the equipment, we will assume that it was inoperable
during that time and therefore you have to report it as an event or
condition -- or condition or operation prohibited by the tech specs.
We made that assumption even if it was in fact operable
during the whole time and so this change is simply to say you don't have
to do that. The surveillance test was performed late but you went back
and did it and the equipment still worked, then that is not a
condition -- you do not have to report that as a condition prohibited
by tech specs.
DR. FONTANA: Is that a potential early indicator of a
management system breakdown though?
MR. ALLISON: Oh, yes, it may be, and -- but it hasn't
typically been that in the past.
The other thing is that of course the resident inspectors
are very well aware when anything like this happens.
DR. UHRIG: What has it been an indication of?
MR. ALLISON: Well, I haven't been able to determine that
the reports have been useful in any way.
MR. BARTON: It's probably been because it is a scheduling
glitch when a surveillance was due and somebody forgot to put it on the
work schedule for that week. That has been the majority of these late
surveillances.
MR. ALLISON: Pardon me, yes. That and then there's another
thing too that we do get and we are getting a lot of reports like this.
I have reviewed this surveillance procedure and discovered a
glitch where maybe there wasn't overlap or neglected one of the criteria
so for all this time I have been doing this surveillance test but I
didn't really show it was operable.
We get a lot of those, and so therefore although I did the
test and it passed, the test wasn't quite proper, so they will report
that as a condition prohibited by tech specs.
MR. NEWBERRY: This is an important change in philosophy as
well as just their reporting system. Just because you haven't tested it
each month doesn't mean that it is not going to work. In fact, you can
look at the data to give you a pretty good feeling, and in the past we
have done some things that maybe for even the proper safety alternative
because of a mistest to go ahead and now put the plant in a different
condition in order to perform the test.
MR. BARTON: To have a scram and then go do the
surveillance --
MR. NEWBERRY: Right, that's right.
MR. BARTON: -- because you didn't do it two days ago like
you should have. Right. This is a good change.
MR. ALLISON: And there is an issue we are talking about
here too, and it is not whether -- it doesn't seem to be an issue really
in the Staff whether we ought to make the change, but the package
proposes to do it by changing the guidance and this particular issue is
with OGC -- why not change the rule so that it is reflected in the plain
language of the rule as well as in the guidance, and that is usually a
good idea if you don't have to make the rule long to do it.
Okay. Moving on to Slide 9 is an unanalyzed condition that
significantly compromises plant safety. Now the package proposed to
delete this requirement, this criterion. The reason was that other
criteria would capture events of this type if they are significant, and
specifically if you look at the kinds of reports that we get here, they
are all design basis issues.
They are the same kind of thing you find when you review
design basis reports. There was a switch here that was set wrong and it
didn't meet some limit in the FSAR only in this case perhaps -- and the
system would still work, okay? -- so that kind of report would be
eliminated and if the system wouldn't work because of it, the report
would be retained, but that is all we get here is design basis problems
but they don't, for some reason or other they don't consider it outside
the design basis so they report it under this.
This is an issue that we are talking about. The question
is what they are reporting now would be captured by other reporting
criteria -- in fact, everything our guidance tells them to report, if we
make this change I'll just take those same examples and move them over
to the other criteria where they'd be captured. It's not a problem. It
seems to be redundant, but the question is, well, could something happen
that we can't foresee that we don't understand or we can't foresee that
this criterion would catch that we need to know about, so that is the
question with regard to that.
CHAIRMAN POWERS: How do you pursue that question?
MR. ALLISON: My approach is to look at what has been
reported and say that doesn't happen and if you're worried about
something that we can't foresee, that we can't define in these reporting
rules happening that has some safety significance that we need to know
about, then we would have two or three backstops available.
One is 10 CFR 50.09, which requires the licensee to report
anything that he determines -- the licensee determines is significant,
has significant implications.
The other of course is our resident inspectors really do
know a lot. They are in the control room every day. They read logs.
They see what the operability and terminations are. They review all of
the corrective action systems.
Then the third is the common sense of the licensees -- that
is, that if something is really very significant, and this is a
significant thing, significantly compromises plant safety --
MR. ALLISON: So those are the kinds of backups we have.
DR. POWERS: It sounds like -- it sounds pretty rational.
You've got an empirical database and you said I can't find any examples
that wouldn't have been caught by my other criterion on the proposed
rule.
MR. ALLISON: Correct, if they are significant.
DR. POWERS: And if -- but that could be a fallible
database. Maybe there's something out there, but I have a couple
backstops that seem to cover this as well.
MR. ALLISON: Right.
DR. POWERS: So now you get into real flights of imagination
to find something that somehow alludes all of these things, and there's
a point where you just say, and if it happens, we'll change the rule
again, I guess.
[Laughter.]
DR. POWERS: I mean, it seems silly to try to cover the
inconceivable or the unconceived.
MR. ALLISON: Unconceived.
DR. POWERS: That sounds like a pretty powerful argument to
me.
MR. ALLISON: That's the argument. But we do have to settle
the issues within staff.
DR. POWERS: Well, I mean, the optics are unfortunate
because it's an unanalyzed condition that significantly compromises the
plant safety, and to say, well, NRC doesn't want that reported anymore,
that doesn't sound like a good optics.
[Laughter.]
DR. POWERS: I think what you want to say is, it's reported
elsewhere.
MR. ALLISON: Right.
MR. NEWBERRY: In fact, one of our comments on the design
basis issues is the optics and public competence issues. You know, why
shouldn't the NRC receive a report any time you're outside the design
basis regardless of the significance?
DR. POWERS: True. That's a fair question to ask.
MR. ALLISON: Okay. The next slide is a condition not
covered by the operating and emergency procedures. The package proposes
to delete this criterion.
It sounds similar to the first, but it's a little different.
Right now, this criterion doesn't result in any reports that are
worthwhile. It results in about four erroneous reports a year. By
erroneous, I mean I discovered a mistake in a procedure, and so -- and
decided to report it under this criterion. But that's not what this
criterion requires. This is when the plant is in a condition that's not
covered by its procedures. The thinking is that other criteria would
identify this kind of an event if it's significant, those other criteria
being a broad scatter.
In the case of unanalyzed events, I know just what is
reported and what would catch significant events. Since nothing is
reported here, it's kind of like -- I think the other criterion would
catch it.
DR. WALLIS: Well, it seems that there is a possibility of
something getting into the syringe, though, which is very interesting.
You've got to catch it.
MR. ALLISON: Well, I think I would have to --
DR. WALLIS: Just the fact that you haven't got significant
evidence yet doesn't mean that it might not occur. You don't want --
MR. ALLISON: Yes.
DR. WALLIS: -- a plant getting into the conditions.
MR. ALLISON: Well --
DR. WALLIS: You want to know if it does.
MR. ALLISON: Yes. I think everyone agrees if somebody gets
a plant in this condition and the other criteria didn't catch it, we
would like to know about it. The real discussion here is, do those
other criteria event in fact identify significant events, because we
have not been able to think of an example where this happens and it
wouldn't be caught by other criterion. But anyway, that's the question:
Could it happen?
Okay. Moving on to the next slide. That's slide 11.
System actuation. Okay. Currently, the rules require reporting the
actuation of any engineered safety feature, and the package proposes to
stop using that term and to say report the actuation of any of the
following systems, and it lists the systems there.
This would provide some consistent reporting for the most
risk-significant systems in general. Some plants don't consider the
emergency diesel generators or the auxiliary feedwater system or the
RCIC to be ESS, but they're all very highly risk significant systems.
On the other hand, frequently we get reports of control room
ventilation actuation, and although that may be an important system,
it's not too highly risk significant, or even a fan in some penetration
room, you know, we'll get people reports. But we don't -- so this is an
attempt to rationalize that a little bit.
DR. POWERS: How did you go about making your list of
systems that you wanted --
MR. ALLISON: The way those were made was by -- we did --
had a consultant look through the PRAs for five or six plants and
calculate the Fussel-Vessley ratios and so on, made a table and lined
them up, and the systems that are in that last pretty generally came up
high for that sample of plants, plus the general knowledge of PRA
analysts who've been working with PRAs and have a feeling for the risk
significance.
Then, of course, we had to add things that wouldn't show up
in that kind of measure, like containment, because those measures I just
talked about apply to the first level of a PRA, not to containment
integrity.
So that's how they were made up.
DR. POWERS: I guess the only thing that distresses about
that is what was the sample of plants?
MR. ALLISON: Well, it was three PWRs and two BWRs.
DR. POWERS: Okay. So that means that there's at least one
class of BWRs that got no coverage.
MR. ALLISON: In that particular sample.
DR. POWERS: Yes.
MR. ALLISON: Yes.
DR. POWERS: And I don't know what the PWRs are. If there
are sufficient differences among PWRs, my suspicion is that I can
identify -- I could conceivably identify classes of PWRs that didn't get
covered.
MR. ALLISON: Oh, many plants, yes. Yes. And there might
be a system at Zion that is risk significant. It's possible. And it's
not on that list. But I guess what I -- you know, this is based in
large part on the experience of people who have done PRAs, too.
DR. WALLIS: Why should you make the list? Why shouldn't
the plant make the list based on their own PRA?
MR. ALLISON: Well, that's a possibility. My main defense
of this list is it's a lot better than the one that they're using.
DR. POWERS: I agree with you. I agree with you.
[Laughter.]
DR. POWERS: But since we're going to go to the effort to
change a rule, why not try to get as good a rule as we can within our
time constraints here?
MR. ALLISON: Well, I think that's the key, is the time
constraints.
DR. SEALE: You talk about things that impact other rules.
This is not the first time we've run across this problem; that is, that
there are systems that are not on the, quote, safety list which in fact
have risk significance. And the last thing I think we would like to see
is a proliferation of such lists which may reflect differences in
interpretation of what constitutes the threshold of safety significance
that you want to go up to.
I think the chaos that could come out of that could be
absolutely -- well, it just opens all kinds of arguments and so on. If
you want to go to this kind of approach, and I certainly wouldn't argue
against it, it seems to me there's a large constituency for this and you
need to get all of the members of that constituency involved.
MR. ALLISON: That's a good point because this is an issue
that the industry --
DR. SEALE: I'm sure it is.
MR. ALLISON: -- consistently opposed, and I don't understand
why because they call me frequently and say this fan in something room
that went off, and I didn't report it, and the resident's citing me, and
did I have to report it? And the answer is yes, you had to report it.
It was an ESS. And you shouldn't have to report those kinds of things,
but you do. So I would hope to, maybe at the workshop, to try to
persuade industry that they would be better off with this certain list
than -- but anyway, that's my argument to use, but --
DR. WALLIS: Is the list negotiable between you and the
licensee or is it something decreed by you?
DR. POWERS: Once it's in the rule, it's decreed.
[Laughter.]
DR. WALLIS: Well, it seems to me that this doesn't make
much sense because it would -- you want to rely -- the licensee has
something to contribute about what is not risk significant. They may
reveal to you something you don't know which would be beneficial.
MR. ALLISON: Yes. Yes, that's true.
DR. WALLIS: You might miss something the licensee knows
about.
MR. ALLISON: Absolutely. Absolutely. But the time, I think,
is my big constraint there because -- and we've been through this before
on the reliability data rule, is when you -- when you write a
requirement and say, well, I know these systems are pretty generally
risk significant throughout the industry, so you have to include these,
now look at your PRA and determine what else might rise to this level at
your plants, we're going to spend a long time talking about that one, I
think.
DR. SEALE: Well, are service water systems on your list?
MR. ALLISON: Yes, but in this case, we're talking about
wanting to know about the actuations, so we're really only asking them,
when you have a system like Palo Verde that's just a standby system, and
it goes off like an ECCS does, report those actuations. But for a
common service water system that you turn on and off all the time, we
didn't ask them -- and keep running all the time, we didn't ask them to
do those, not because the service water is insignificant, but because
for a constantly running system, we don't really need to know when it's
actuating.
DR. WALLIS: Yes, but some configurations are sensitive when
they're suddenly stopped and started and things like that.
MR. ALLISON: Yes.
So anyway, well, that's what we've proposed here, and I
think it's important to get reasonably consistent reporting of the
actuation, and in part, this is part of the estimation of reliability
parameters we're doing.
Okay. My last slide is the principal effects on the burden,
and I guess I would call significant reductions in, number one, the
reporting of late surveillance tests in -- you know, based -- based on a
review of the LERs that we do have, it looks like we would get maybe 160
or so less LERs per year in that category; and eliminating reporting for
design issues that don't rise to the level of affecting operability,
that would also be around 160 LERs per year eliminated, and it might be
more, but I don't want to claim too much credit there.
Minor reductions -- the extension of reporting times is not
so much an FTE or man-hour kind of -- it's not a big thing there. In
fact, it doesn't show up in the estimates. But it seems to be very
important to the people who have to actually make the reports and that
is when they have to make these things -- do these things in a rush when
it might not be warranted, it's a real burden.
Okay. Eliminating these two criteria, the redundant -- what
I call the redundant criteria on analyzed condition and condition not
covered by emergency procedures, that would -- neither one of those
would, of course, by itself create a significant burden reduction
because we don't get a lot of reports in those categories that wouldn't
be captured by other criteria.
And final, minor increase in the content of LERs doesn't
affect the number of reports, but in the content of LERs in the
guidance, we're asking for some additional information and particularly
for shutdown events in what was operable and what was not operable when
you're reporting a shutdown event. The purpose of that is to support
how PRA looks at the significance of events.
So that's pretty much the story of what's in the package and
the open issues that we're still discussing. The schedule is fairly
aggressive, but we still -- and we're late in the game, but we still
hope to meet that schedule if we can, but I can't tell you --
DR. POWERS: I guess in your schedule the only thing that
surprised me was that there's a two week -- two weeks after you put the
notice of proposed rulemaking out, you're having a workshop?
MR. ALLISON: Yes, sir.
DR. POWERS: Has that been coordinated sufficiently that, in
fact, the licensees can participate effectively?
MR. ALLISON: Yes, they've told me they can.
Now, the schedule I showed you doesn't do the schedule
justice. It's very crowded in that time. You have to do it fast
because --
MR. NEWBERRY: I wouldn't let him put all those milestones
up there.
MR. ALLISON: Within 30 days, the licensees have to submit
their comments to OMB, and so this gives them a chance to come and have
this discussion and say, what did you mean or why did you write this
that way, et cetera, and find out so that they can -- and some -- that
helps them with their comments.
DR. POWERS: Sure.
MR. ALLISON: Sometimes they'll find out, no, that wasn't
what we meant or -- and they don't have to waste time commenting on
that.
DR. POWERS: They just change the comment from --
MR. ALLISON: Right.
DR. POWERS: -- don't do this to say what you mean instead
of what you said.
MR. ALLISON: Right.
DR. POWERS: Yes. Sure.
MR. ALLISON: So anyway, it's a kind of an aggressive thing,
but we hope to still make it, and, of course, it might be -- well,
that's the end of the presentation.
Scott, do you have anything?
MR. NEWBERRY: Yes. Just -- I wrote down three takeaways
that I think in one way or another we would have touched on, but I think
it's fair to summarize here that we would commit to touch on them, and
that's the issue of going to four to eight hours, taking a systematic --
call it empirical look to have a good understanding to support a
decision one way or the other, and that was an issue that was on our
plate anyway, but just to confirm that we're going to look at that, and
then a comment on systematically running through, you know, all the
rules reporting requirements for consistency. Knowing Dennis, he's
probably done a lot more than I'm aware of there. But we'll talk about
that one.
And then this last issues which you run into over and over
again because of significant plant-specific differences when you take a
risk-informed look at something and try to get system specific --
there's the benefit there to try to pin things down, but then you could
have a problem, and I think that will be an issue that we can talk about
and even raise in the workshop to see how --
DR. POWERS: It may be a subject that you would want public
comment on.
MR. NEWBERRY: Yes. And we could highlight these in the
Federal Register Notice to really pinpoint comments. That's something
that I've always liked to do, to really pinpoint areas where you want to
get focused comments.
DR. POWERS: You run into this -- I mean, from your
licensees, you run into this problem of is it fair to go out and ask
them to do a risk assessment on their systems. On the other hand,
life's a lot easier if you tell me exactly what systems I have to do,
you know. They may have a preference.
MR. ALLISON: And I have to say with the current situation,
they do have a list of systems. They can look in their FSAR, and if
it's not an ESF, it's not required; if it is, it is. So they do have a
certain list right now, and --
DR. POWERS: I mean, the problem with that list is it's not
risk informed, and so you --
MR. ALLISON: Correct.
DR. POWERS: -- get information on ventilation fans that are
not risk significant, and that's an imposition both on the licensee and
you.
MR. ALLISON: Yes.
DR. POWERS: And what you'd like to do is focus both of your
efforts on things that have some safety significance.
MR. BARTON: Now, what was your first time? You said you
had three.
MR. ALLISON: The first one was going from four to eight
hours --
MR. BARTON: Okay.
MR. ALLISON: -- in terms of --
MR. BARTON: Got you.
Anyway other comments or question from members of the
committee to the staff?
DR. SEALE: I really have the mental image of a bunch of
rocking boats in the wake of this thing because you're going to interact
with a lot of other requirements and --
DR. POWERS: It's certainly clear that they're going to have
an interfacial problem with the state requirements.
DR. SEALE: Yeah. Yeah. With the states, you mean the
agreement states? Yes.
DR. POWERS: Well, all the FEMA --
DR. SEALE: Yeah.
DR. POWERS: -- activities.
DR. SEALE: And all the other --
DR. POWERS: I think you've got that covered.
DR. SEALE: Yes. You know, there's this PT boat going
through the marina at full speed, and there's going to be a lot of
debris left in the wake of this.
MR. BARTON: Don't you have that covered in a proposed rule
where you said something about there were other reporting requirements
that were, you know, more frequent or in different time frames, you
would still have to report to the NRC? I thought I saw that in a
proposed rule some place.
MR. ALLISON: I don't remember if --
MR. BARTON: For example, if you had a local -- if you had a
requirement to report to your local community something that happened
immediately, but you didn't have to do it here for eight hours or
something, you would --
MR. ALLISON: That's right. That's --
MR. BARTON: You would still need to notify the NRC if you
had to notify other government agencies.
MR. ALLISON: Yes.
MR. BARTON: That's still in there someplace, right?
MR. ALLISON: Yes. And -- yes. There are two things. One
is in the response to comments in the State of Illinois. So that is
discussed in the response to comments.
MR. BARTON: Okay.
MR. ALLISON: Basically, we saw two -- a couple of things.
One is that we've asked as much a we can for state input on these
things, but if the State of Connecticut insists on a one-hour report,
that's your deal with them. This rule doesn't control what Connecticut
wants you to do.
MR. BARTON: That's true, but if I have to report something
to Connecticut in one hour, I now have to tell you.
MR. ALLISON: That's right. So it's as soon as practical
and in no cases later than eight hours. Well, if you reported it to
Connecticut in an hour, it's probably going to be practical to call the
NRC --
MR. BARTON: Yes.
MR. ALLISON: -- without waiting eight. It does say that.
MR. BARTON: Hearing no other comments at this point, I
understand NEI is here and would like to make some comments regarding
the proposed rule.
MR. DAVIS: Good morning. I'm Jim Davis, Director of
Operations at the Nuclear Energy Institute, and I think you'll
appreciate that my remarks will be exceedingly brief.
I think Mr. Allison was somewhat surprised when he issued
the ANPR in that normally those things go by and there's not much
reaction, but in fact the industry started getting involved right at
that point, and aggressively involved, because our analysis and our
attempts to get the reporting requirements sorted out over the last few
years we feel that probably the process for coming to closure on this
rulemaking is almost as important as some of the changes that are being
made. And there hasn't been any discussion of that. I wanted to make
sure that I pointed that part of it out.
We think that part of the difficulty is that we have three
parties that have been playing in this particular area. We had AEOD at
one time who was collecting data and information and a valid
requirement, we had the licensee who was trying to report, and a totally
different group doing enforcement activities. And we weren't well
coordinated in that particular area.
So we're the ones that asked for a tabletop exercise this
fall as an additional step and asked the staff to bring in some of the
draft language they were going to use. We sat down with some people
from the region and with licensees to see if we could make any sense out
of it, and if -- we ran a bunch of scenarios against that draft language
to see if you really could figure out what you needed to report.
We're also the ones that asked for the workshop that you
talked about in that we said (1) we want to see the rule, and we also
need the NUREG, the implementing guide, at the same time, and we can't
comment on the rule if we don't have the implementing guide in the hand.
So we've got to understand the entire process, what we're aiming for in
the reporting in the higher level and how that's going to be
implemented, and we in fact will be ready to participate in that
workshop.
And again the idea is we're going to come in and okay,
here's what you've said you want, here's what the implementing guide
says, here's a scenario, does it make any sense, can we walk through
this and could we make this consistent so that in the future the
licensee knows what he's reporting and the NRC knows what is going to be
reported. Because I think that is a much more important argument and
discussion item than the ESF action, what's on that list, from our
standpoint.
Now, I mean, I picked that one because Mr. Allison is
correct, the industry is concerned that the plants have been licensed in
one arena, we've defined what the engineered safety features are, and
suddenly we're going to change that rubric and say no, the real list of
engineered safety features is something else, and we've got to be
careful how we're approaching that. And that's our concern, that
suddenly in reporting you want a consistent list of items to be
reported. We also need to make sure what we're doing is consistent with
the license and the basis under which the plants have been licensed.
And we'll engage in that discussion when the notice comes out in that
area.
The other comment, as we looked at it we saw two other
problems that we think need to be solved in this arena that maybe
haven't been addressed. There is the need for timely information to be
given to the NRC, particularly in the old four-hour reports, which were
things that would have been of media interest, outside interest, things
reported to other agencies. One of the problems is that in fact the
event reports have become also the public window to the industry, and
for most media and for most outside organizations, that event report
also is the alerting or the initial indication that they get.
So there needs to be a balance. And we feel that the eight
hours allows, one, the facility to in fact react and provide accurate
information in an initial report, and still provides the staff with
adequate alertment on what's going on. Just like the contaminated man
in the ambulance, what you really want to do is put the right
information in front of the world the first time and not the second
time.
I have had several events that are reported with the acronym
Y2K in it, and in each case, the information in the initial report has
not been -- has not conveyed the whole story and been totally accurate
in the situation. But in the national media those two have gotten a lot
of attention, and therefore we've always wasted a significant amount of
time going back and feeding the right information into the system. So
that's our cut on the eight versus four hours. And obviously there's a
balance. The staff will look at that, because there are other factors
that need to be considered in that arena.
DR. BARTON: But you make a good argument for the eight
hours.
CHAIRMAN POWERS: I mean, it's a really interesting point.
I think you're right, that obviously things are not requiring
instantaneous action if you're going to make either a four or an
eight-hour report, and if that report is in fact the window on the
industry to the rest of the world, it serves everybody's interest to be
as accurate as you can.
MR. DAVIS: And the rest of the world has a problem when we
come back later and say well, that wasn't really the story, the story is
thus and such. And so there is some need -- and actually there's a
burden on the industry in that arena, that those eight-hour reports need
to also be carefully staffed out so that you provide --
CHAIRMAN POWERS: So let me ask you this question.
MR. DAVIS: The right information.
CHAIRMAN POWERS: If eight gives you more accuracy than
four, wouldn't 24 give you more accuracy than eight?
MR. DAVIS: There's a balance. And I heard your earlier
comment. One, this is published on a daily basis. It's put into the --
onto the Web site on a daily basis. So I think eight is a rational
balance between the daily summary requirements and the need for rapid
reporting of the material. If we went to a --
CHAIRMAN POWERS: I mean, I can understand why you might
think -- that your intuition and your judgment, which is undoubtedly
better than mine -- on this says that's a good balance.
MR. DAVIS: I could argue --
CHAIRMAN POWERS: But it's not very persuasive.
MR. DAVIS: I could argue the first point on judgment, but
the industry is, you know, collectively we looked at it, and we think
that eight hours -- we could live with eight hours and that it
provides for the scenario --
CHAIRMAN POWERS: That's not my question. You could live
with four I'm sure. I mean, you're not going to shut down plants
because they make it four hours.
MR. DAVIS: Yes, sir. That's correct. We've lived with
four before.
CHAIRMAN POWERS: The question really is what is optimal.
If I look at the past reports and the corrections to those reports,
would it make more sense to have 24 or 8?
DR. KRESS: It might be a good idea to ask why is there a
time limit at all.
CHAIRMAN POWERS: That's right.
DR. KRESS: And work from that direction. You might end up
with next day's okay or --
CHAIRMAN POWERS: Um-hum. I mean, in some way -- I have no
doubt in people's judgments and I -- one thing I'm confident is that
their judgments are better than mine on this. But to make a scrutable
rule I need to look at empirical data here.
MR. DAVIS: And I have no doubt that I could contribute that
would answer the eight versus 24 question. We have not looked at it in
that --
CHAIRMAN POWERS: It wouldn't matter. As soon as you gave
in to 24 I'd ask for seven days.
MR. DAVIS: The final point that we're -- and I think the
design-basis issue is the best pointer in this area -- is we're trying
to get wording that is self-contained within the reporting requirements
and doesn't get what I'll say contaminated by other issues that will
come and go within the arena. As I look at the rule as it's written,
the requirement to report things that are outside the design basis makes
a lot of sense. It's when you redefine design basis to be changes of
color of pipe that we get into a problem. And what I've seen is -- and
I don't know if this sort of analogy will get me in trouble -- but I've
see this, it's sort of been the broken-taillight syndrome, you know,
that you find problems in other areas, you don't know exactly how to
write it up, so you write it up as a reporting violation, and suddenly
we get ourselves into the position that we're reporting a lot of things
to avoid problems in other areas and not to provide the information that
the NRC and the public need.
So I'm hoping that with the changes we get the criteria
focused back on the real issues and what needs to be reported so it's
sort of self-contained and as we -- because, you know, we're looking at
PRA and a whole bunch of other issues, and things are going to come and
go in that area, and hopefully the reporting can maintain its integrity
in that area. And I think, you know, from our viewpoint those are
probably the two most important aspects of it.
And we will argue around the periphery on the ESF actuation
list and some of the examples and that kind of stuff, but I think we're
moving towards a significant improvement in the process, and we plan to
stay heavily engaged with the staff when they bring this out into the
public arena.
CHAIRMAN POWERS: I agree with you on that. It looks like a
significant improvement, and I'm delighted that you're devoting
resources to this. I think that helps.
DR. BARTON: Thank you.
MR. DAVIS: Thank you, sir.
DR. BARTON: Comments? Questions?
If not, I'll turn the meeting back over to the Chairman.
CHAIRMAN POWERS: I will promptly recess till 10:15, I
think.
[Recess.]
CHAIRMAN POWERS: Let's get back into session, and everybody
is very anxious to get on to this, our next topic, which is the
re-evaluation of the generic safety issue process.
DR. MILLER: Just a bit of background. I may overlap what
John Craig has prepared, but a little background for the committee, to
get us on track here.
The generic safety issue is one that precludes my time by a
long time on this committee. It's been around a long time, and Bob
Seale maybe has the most lengthy experience with it.
Most recently we've dealt with this issue in several
meetings in the last year or so, and March 16th of '98, almost one year
ago, we wrote a long letter to the EDO, followed up by another letter in
October, in which we identified as a committee a number of concerns, and
our most notable concern, most notable, I guess, recommendation was that
the committee stated in the October letter there's a need for a
systematic re-evaluation of the process, the generic safety issue
process.
The staff heard us and maybe others on that issue and, with
the assistance of a consultant, Arthur Andersen, conducted a
self-assessment, of which we have, of course, a rendition of that.
We've had that for a week or so now.
In that self-assessment, the staff made a number of
observations, and from that, they derived a number of recommendations,
and they then proposed to implement those recommendations in what's
called a management directive, and that's number 6.4, and associated
with that is a management directive handbook, which really gives all the
implementation details.
The management directive, as I said, looks at the
self-assessment recommendations and then puts through an implementation
plan.
I want to remind the committee, the management directive was
not derived specifically to respond to the concerns we raised.
However, the staff has done an assessment of our concerns,
and through a cross-referencing, which we all have now, have indicated
how they believe the management directive does speak to our concerns,
and I don't know if they're going to address that today or not, but I
just wanted to remind the committee that they did that, although the
management directive specifically responded to the recommendations of
the self-assessment.
With us today, John Craig will get things started, and Mike
Marshall, I guess, is on the hot seat to see that we have this process
on a track that hopefully will resolve a number of these issues over the
next several years.
With that, I'll turn it over to John.
MR. CRAIG: Well, good morning. It's a pleasure to be able
to come and make this presentation after so many months of working on
the generic issues program.
A number of the staff that have spent a long time with the
GI program and process are here to support it, and as occasion arises,
I've asked people like Ron Emrit, who's been the historian for the GI
process, to be here to help answer questions.
Let's go to the next slide.
This is a progress report on a work in process. The
management directive has not been reviewed and concurred in by the other
offices. We have brief senior managers from the other offices, NMSS and
NRR. We have gotten input from some of the key staff members within
those offices who are responsible for generic activities, but it is by
no means final.
Our goal is to publish this month a draft management
directive for comment that will be both office comment, and we plan to
have stakeholder meetings, public meetings with NEI and others, who also
have an interest in the generic issues program.
So, this is a status report.
I'm not going to go down through each and every item on the
slides.
The letters that the committee wrote reflected comments
about the generic issues program that a number of other sources had
identified. A number of your comments went beyond their comments, and
so, we had a large pool of comments.
I think the point that was most significant perhaps was the
discussion last year about the need for a self-assessment, and rather
than trying to respond to individual comments or criticisms of the
program, there was a realization that so much had changed that it wasn't
a matter of minor changes to the program, we needed to stop, review the
program in a systematic way, and make a determination as to whether or
not the program, in fact, would meet the goals that we thought needed to
be met.
Let's go to the third slide.
One of the handouts that I believe -- you all have received
copies of the self-assessment, and in the back of the self-assessment,
attached is some background information that goes through the history
and evolution of the generic issues program, and the historical
highlights reflect the items that are discussed in there.
I'll note that, in the early '80s, which is the starting
point, I think, for the current system in 0933, it was really an NRR
tracking system, and it was a work item tracking system, much like the
NRR directors -- I think it's a bimonthly report now, it has task action
plans for issues, those kind of things, and so, there was some -- there
is today and there will be some duplication if NRR has responsibility
for one of the generic safety issues, and they have responsibility for a
couple, they're included in that task action plan, in that directors
highlights report, as well as in 0933, and as a result of the system
being moved to research but not fundamentally changed, we perpetuated, I
suppose, a number of problems.
The thought that the licensing link to these issues, as that
was previously defined, didn't exist anymore continued to be part of the
program, and as you read 0933 and you get into a discussion about a
generic issue versus a generic safety issue, you can appreciate, I
think, rather quickly some of the problems that we've encountered and,
indeed, some of the things that we've discussed with the committee as
we've talked about individual issues.
The generic safety issue was, in fact, a potential
cost-beneficial safety enhancement. It's not a safety issue that needs
to be resolved to make a decision about whether or not a plant is safe
enough today. So, it's a safety enhancement.
In earlier times, when there were a lot of license
applications being processed, applicants had to come forward and address
unresolved safety issues, and for generic safety issues in the pipeline,
they could simply commit to implement whatever the staff's ultimate
resolution was, so that there was a priority associated with GSIs as a
result of the need to address them in licensing applications for plants
coming on-line with OLs.
That need obviously doesn't exist anymore, and as a result
of that, generic safety issues have received a lower priority over time
than more safety significant activities in the Office of Research, and
as a result, as I don't need to remind you, as you know, a number of the
GSIs that are open today have been with us for some time.
We have made significant effort over the last year to
complete the prioritization and re-prioritization of the GSIs that are
open.
There are approximately 15 open GSIs today, and we've
developed task action plans for each one, and we're driving them to
closure as quickly as we can, and you were briefed a week or two ago on
B-61, I believe.
I'll note that the comments that we received on that issue,
some of the comments about the risk methods, etcetera, we hope to fix
with this new process.
Another thing that has happened, after we briefed the
Commission in December 1995, was that the Commission directed us to have
a transparent system for the agency so that a member of the public could
ask a question about a generic issue, and if it was an NMSS issue, they
could figure out what the issue was, what the schedule was, etcetera,
and we've been working with the CIO's office to make that transparent
via the computer and link tracking systems.
I think the reasons for revising the process and the
comments in our self-assessment -- we are prepared to go over those.
Mike can talk comment by comment. What we'd like to do is cover the
process and walk through it, and as you have questions about the
process, let's ask them now in a give-and-take, because I think that
will be more meaningful for us, more helpful for us as we make the
revisions to the management directive so we can then publish it for
comment.
As I said, our goal is to get it out in a couple of months
final, and I would -- I don't know if we've already scheduled, but we
would schedule another meeting as we got the final management directive
to go back to the committee.
CHAIRMAN POWERS: The only caution I'll introduce at this
point is that I do have to cut this session off at exactly 11:45.
MR. CRAIG: Okay.
CHAIRMAN POWERS: Several of the members have another
obligation at 11:45.
MR. CRAIG: I will note, on slide 4, that DSI-22, as you
know, the Commission made the decision that Research doesn't do
rulemaking anymore. In the previous generic safety issues, generic
issues process, the resolution stage included development of a draft
rule. And so one of the changes that you will see made in the new one
is that, after the research is done to identify a solution or two, we
would then make a recommendation to either NRR, NMSS, if we thought a
rulemaking was appropriate, or a Generic Letter, and it is up to the
program office then, with our support, to take the lead and run with
that.
So with that as background, I will turn it over to Mike
Marshall.
MR. MARSHALL: Thank you. Good morning. Just a couple of
things we would like to touch on just to start with. There's a couple
of improvements we are trying to seek in a program. We wanted to
provide clear guidance to the staff on how to handle a generic issue.
We wanted to revise the tracking system so it would become more
transparent, not only would it track the beginning of the process, but
follow issues all the way through implementation and verification. A
couple of other things we want to do is update some of the guidance in
the process.
This program overlaps with other programs such as rulemaking
and the inspection program where there's good a bit of guidance for that
part. At the beginning of the process, there's less guidance, or the
guidance was contained in this program, as far as how to screen issues
to decide if you should work on them or not. And most of my -- sorry,
wrong one. And when I am talking, I will be spending most of my time
talking about the items highlighted in red because that is the part of
screening which is contained within the generic issue process, and
usually the black part is contained in other documents, other Management
Directives or other NUREGs or NUREG handbooks, describing to the
staff --
DR. WALLIS: I have to complain, you guys give us a huge
piece of paper, but the most minuscule typing in boxes, in some of the
boxes. It doesn't make sense to use a big piece of paper and then use
the smallest available typing.
MR. MARSHALL: I am at fault for that one. I got criticized
many times, and I thought just going to a bigger sheet of paper would
fix it, but apparently it didn't. I apologize for that. Okay.
But, overall, the flow chart describes the overall process.
You will probably see a lot of similarities to the current process in
here. And the goals for the overall program we have, I just want to
make sure we have a process for identified generic issues, and we have
process or procedure for screening them. We make sure that identified
generic issued are analyzed, meaning we establish a technical basis for
taking some sort of action and we identify cost effective solutions.
Also, for the generic issues that are resolved, we identify
what regulatory vehicle the solution should be implemented in, verify
that implementation has taken place. And for all the work that occurs
in this process, we want to make sure it is documented, one, for people
outside NRC to see how we handled the generic issue, also for the staff.
As issues happen to come up again, we will be able to go back into the
documentation and see how the staff handled issues in the past or
handled similar issues, or to evaluate whether an issue was resolved
adequately if questions arise about how well an issue was resolved.
And, again, all the documentation would be shared with anybody that is
interested in how we handled the generic issue. And that's the overall
goals we hope to implement through the Management Directive.
DR. POWERS: I take it from this flow chart that the word
"resolved" associated with the GSI is going to take on a different
shape, form and meaning.
MR. MARSHALL: Yes, it will. Before --
DR. POWERS: That, if nothing else, is a step in the right
direction. Make sure that we highlight that in any letter we write on
this.
MR. MARSHALL: Before -- right now, what "resolved" will
mean is that an issue has gone through pretty much all these stages,
that the NRC has taken the time to verify that an implementation
occurred. Previously, "resolved" meant that back up here, right along
in this area here, the NRC decided how to address a generic issue. We
decided we needed a new rule. We decided we needed a new Regulatory
Guide. It didn't necessarily mean we issued a new rule, we issued a new
Regulatory Guide, or anybody verified whether licensees implemented the
solution.
And this was one of the biggest criticism from, not only
internal to the NRC, but external groups had with our process, was
"resolved" didn't mean "resolved," it meant we had a plan, essentially,
for resolution.
DR. SEALE: Mike, I hate to beat a dead horse, but --
MR. MARSHALL: I have a bigger, hopefully, --
DR. SEALE: Graham raised a question. I finally figured out
what all the diamonds are except the top one.
MR. MARSHALL: Okay. Let me go through them. The other
side isn't any better than this one, really.
DR. POWERS: Classification.
DR. SEALE: Thank you.
DR. POWERS: We got it resolved, Mike.
[Laughter.]
MR. MARSHALL: Okay. Class -- oh, sorry.
DR. WALLIS: I didn't go through all the steps to resolve
that one.
MR. MARSHALL: Okay.
DR. POWERS: Sure we did. We came up with a candidate
solution and we implemented it.
DR. MILLER: Did we verify it will work now?
DR. POWERS: We haven't verified.
MR. MARSHALL: Oh, another thing we are going to get rid of,
as we went through technical -- what we are calling technical assessment
now, that used to be called prioritization. There used to be a box
outside of prioritization, where you use -- assign a "nearly resolved"
to an issue at this stage. Essentially, it meant it didn't go through
the formal prioritization stage, yet we decided we are going to work on
it anyway. I guess that wasn't the proper way to phrase that. But we
decided we are going to work on it. It didn't mean it was nearly
resolved, just at the very beginning of an issue, you would label it
"nearly resolved," and that was merely to indicate to us that it didn't
go through the formal prioritization stage. We will no longer be using
the term "nearly resolved."
And, also, on that point is recently one of your comments
was, you took all the nearly resolved and you just changed them to high.
One thing we are stressing in a Management Directive is any decision you
make with regard to a generic issue should be documented, even a
decision of why it was important, why it was not important, why you
decided to work on it. And, again, we would document that in a central
location so everybody could see why the NRC is working on an issue, or
not working on an issue, also, which is probably just as important.
DR. WALLIS: I think it would be useful if you had sort of
an expected mean timeline for these activities.
MR. MARSHALL: Yes, we do. For the beginning -- for the
very beginning, we do. For the initial screening --
DR. WALLIS: For the beginning.
MR. MARSHALL: Let me -- for the initial screening. And
this is essentially taking a candidate to generic issue and evaluating
whether it is a generic issue or it should belong in another program.
Is it plant-specific and should be directed to another agency -- I mean
not another agency, another part of the agency? If there is another
ongoing program which this issue fits in, generic compliance activity,
directing it to that program. Making sure it is duplicated -- already a
generic issue on the books. And then defining what scope the issue
would have.
One problem we identified during the self-assessment was
that the scopes of some of the issues weren't defined until much later
in the process. It made it hard to decide how much work you actually
had to do to solve a problem if you weren't sure what problem you were
solving. And we are going to give ourselves a month to do that, decide
if it belongs in the process and decide what scope the issue has before
we send it on to the technical screening stage. And we gave ourselves
three to six months to do all technical screenings. And, essentially,
the purpose of the technical screening is to decide is this issue worth
working on, essentially, is what it boils down to. Will we get the
benefits we think we are going to get from them, and are they
substantial? And for certain issues, are they cost effective to
implement? We will try to get a handle on that before sending it
through the rest of the process.
Just a couple of issues. We are still on slide 9 now.
DR. WALLIS: You want to make sure that the plan doesn't
change before you get out of the first process.
MR. MARSHALL: What?
DR. WALLIS: The plan, the plan.
MR. MARSHALL: Oh, the plan.
DR. WALLIS: If you have too many boxes, you may find that
the whole system changes before you get out of the bottom box.
MR. MARSHALL: Okay.
DR. WALLIS: Why does it take three to six months to do
this?
MR. MARSHALL: Three to six months, one --
DR. WALLIS: Aren't some of them done quicker?
MR. MARSHALL: If you wouldn't, could I address that a few
slides later.
DR. WALLIS: Sure.
MR. MARSHALL: Around slide -- once we get to slide 12, I
will pick up that question, I will make sure.
But I would just like to touch on a few things about
tracking documentation before we get much further into this. Right now
our tracking system, and it is a central tracking system, we track the
beginning of the process. We tracked how we initially screened it. The
technical screening, which used to be formerly called prioritization,
and we tracked the development of the resolution, reaching the technical
basis. And after that, there was no longer a central tracking system.
Now, we propose tracking issues from one they entered the
process, all the way through verification. So if you have a question of
whether an issue was implemented or not, you will still have to just go
to a single point, you won't have to go a different office. You can go
to what we are going to -- we are going to create a position for a
project manager position, or you can check with the project manager for
the process and he should be able to tell you what stage in the generic
issue process any issue is, whether it has been implemented, how many
more plants will have to do implementation, because all information will
be funneled to the generic issue process project manager.
We are also going to establish minimum documentation
requirements, again, just telling the staffers, or telling us what
things we have to document, what technical decisions, what management
decisions need to be documented. And, again, all these would be
funneled to the generic issue process project manager, so it will be
kept in a central location.
One other item, I guess just to note on here, and we have --
this isn't available to everybody yet, on Item 3, we have begun
developing a web page for generic issues. And from that web page, it
should send you to all different tracking systems for generic issues.
That is under works right now. Generic Issues Status Report, GIMCS,
will continue to be updated on a quarterly basis, and, hopefully, we can
increase access to it and get it out on a much more regular basis than
we have recently.
One thing that hasn't changed is the identification. Pretty
much all these balloons here represent sources where candidate issues
can come from, the public, industry, internal to the NRC, and most of
these bullets over here represent sources within the NRC where generic
issues, candidate issues can come from.
And we are going to request for issues, or organizations
internal to the NRC to complete a candidate issue form. We had
something similar to this before, but we never really required the other
parts and agencies to complete this form. We figure by having the
submitter complete this form, we will get a better grasp on a scope. We
are sure everybody is talking about the same problem.
Again, the problem with the scope is, what the agency that
-- or the submitter that submitted the issue was thinking the problem is
one thing, the person that got the problem to work on is thinking
something else, and sometimes we end up solving the wrong problem, or it
might be the right problem, not necessarily the problem the submitter
wanted us to solve.
And just to show you some of the things we are going to ask
for on the candidate issue form, there is a quick list of them, a
listing here, and I will just let you read through them. And, again,
most of it is geared towards making sure we all have a common
understanding of what issue we are supposed to address. This is a
backup slide, and I will leave a copy of it here with Ji at the end of
the meeting.
DR. WALLIS: I just wonder how much of this has been done
before. Isn't there a long history of work on these issues, so that a
lot of the work has already been done? You don't want to start too far
back in the process.
MR. MARSHALL: Oh, no. No, no. One thing, I guess an
important thing I forgot to mention is that the process, the changes we
are implementing are for new issues. Any current issue would complete,
would follow through the current process. So the 15 issues John has
mentioned earlier would go through -- we would let them finish the
current process. We would document them in 0933, and this, the
revisions we are talking here are handling any new issues that arise.
DR. WALLIS: I see.
MR. MARSHALL: Any new material and reactor issues that come
up will go through this process here.
DR. WALLIS: So, this has no bearing on getting rid of the
present issues.
MR. MARSHALL: No.
Next slide.
The initial screening -- and that's the beginning here --
we'll still do some of the things we currently do with the initial
screening.
Again, that's checking for duplication, checking to see if
this issue has been addressed before, checking to see if this issue's
being addressed in some other program in NRC.
One big -- one change we're going to have this time around
is, number one, we'll make this a joint activity between the offices.
Originally, our initial screening was done within the Office of
Research.
Again, this is geared towards making sure that there is a
common understanding between the offices of what the issue is and what
the resolution of the issues is supposed to accomplish.
CHAIRMAN POWERS: It strikes me that frequently I see things
that are said, well, this is a joint activity and we'll get all the
offices together to work on this and that is the death knell to
progress, that all those offices you list up there are fairly busy
people, you come in with this additional activity, it is not top on
their priority list, and so, it languishes there, and it's very
difficult to get them all in one room at one time or to even do it in a
sequential fashion.
How do you avoid that becoming the slow step in this kinetic
process?
MR. CRAIG: The way we do it is to have it be clearly
defined as part of a management directive, and as we get office buy-in
on the management directive, there will be specific individuals that are
identified to participate on this and make it happen.
The fact that we're reporting these to Congress on a monthly
basis is clear evidence to the agency that some external stakeholders
think these are very high priority and it is, in fact -- that our
activities reflect recognition of that significance.
So, I think the comment's fair, and it's a caution that
we're going to have to guard against.
In a similar manner, the research effectiveness review
board, where we have people from all the offices -- there was some
jerkiness getting out of the gate, and it quickly went to the office
directors, there was some discussion, and then we got the support. The
same thing would happen here.
But it is a caution that we're going to have to be sensitive
to to make sure that these things don't get delayed.
DR. MILLER: You mentioned that there will be individuals --
they're the ones that are going to be on this generic issue panel or
committee, I believe you called it.
MR. CRAIG: Right.
DR. MILLER: And then there's going to be a chairman of that
group that's kind of the manager of this? That person will have real
designated to manage this, or it will be just an add-on?
MR. CRAIG: No, it will be part of their assigned functions,
it will be a senior executive, I believe it's going to end up being Jack
Rosenthal in the new reorganization.
He can't be here today because he's on the west coast
looking at some grid reliability issues, but he's fully fluent in the
presentation and on-board with what we're doing.
DR. MILLER: Just one more comment. One thing that might be
valuable, Mike, when you go through -- going back to the assessment
program, very clearly said that set certain goals and then we set
certain success criteria, and they were risk reduction, burden
reduction, public confidence-building.
You might just remind us, as we go through this process --
some are pretty obvious and some may not be so obvious, but I think
those are very high-level goals and very valuable to this entire
process.
MR. CRAIG: You'll see some of those reflected on the other
slide specifically.
MR. MARSHALL: We hope to achieve the public confidence by
having this very transparent process that's easy to delve in and get a
question answered.
DR. MILLER: A web-site is certainly one of those key
issues.
MR. MARSHALL: Right. And for the risk reduction and burden
reduction, we'll touch on that right now in this slide here.
The issues -- the generic issues that we're going to handle
in this process we're going to put in one of these three bins here,
either adequate protection, substantial safety enhancement, or burden
reduction initiatives, and I'll just give everybody a moment to read
over those definitions.
The screening of an adequate protection, substantial safety
enhancement is different. There's a big difference between how we would
handle a burden reduction issue and an adequate protection concern,
issue. That's the main reason for the difference.
USI will also be part of the system, and I'll touch on that
on a later slide.
DR. WALLIS: How do you do cost-beneficial analysis? Do you
put a price on various aspects of public safety?
MR. MARSHALL: Yes.
DR. WALLIS: You do.
MR. MARSHALL: Yes. And we touch on that later in the
guide. That's one of the sections that wasn't complete when we sent out
the management directive to you.
DR. WALLIS: So, you have a formula for various sorts of
cancers and costs and land all the other things?
MR. MARSHALL: Can I touch on that later, in the same slide
I told you before. Sorry about that.
DR. WALLIS: No, that's all right.
MR. MARSHALL: Okay.
Is there any questions about the definitions?
One point I would like to make on the slide, though, is the
generic issue process won't handle all generic issues. For instance,
generic compliance issues -- we don't envision them going through this
process.
What we used to call -- what are still called licensing
issues, which used to be part of the generic issue process -- they will
not be handled through this process.
DR. MILLER: And all the current issues will not be handled
through this process.
MR. MARSHALL: And the current issues will not be handled
through this revised process. The current issues will be handled
through the current process, and so, there are some -- I think there's
at least one issue in the current process. That will be worked and
completed under the current process.
DR. UHRIG: What about the re-licensing issues that are
associated with --
DR. SEALE: -- license renewal?
DR. UHRIG: Yes.
MR. MARSHALL: If one of those issues meet one of these
three definitions, they could be handled through this process.
DR. UHRIG: Well, there a number of issues that have been
declared generic issues that are not addressed by the applicants.
MR. CRAIG: I was with you until you said they weren't
addressed by the applicants.
Currently in the generic issues program, there are two
technical issues, environmental qualification of electrical equipment
and fatigue associated with license renewal.
In addition, NEI put together a list -- and I forget how
long the list was -- of technical issues related to license renewal, and
you may have been briefed on those.
The decision was made that those were not potential generic safety
issues and shouldn't get dumped into this program, and that was the
point that Mike was making.
DR. FONTANA: How about the first two that you mentioned?
MR. CRAIG: They're in this program, and we're working them
on pretty aggressive schedules.
DR. FONTANA: I read something to the effect that the
industry intended not to do anything more on those two. Is that
correct?
MR. CRAIG: I believe that there's a difference between what
industry is going to do to look at fatigue or EQ technically, industry
collectively, versus plants individually, as opposed to what is being
discussed with Chris Grimes and his staff as to what has to be included
in an application, what level of detail, and I don't know if that's come
to closure yet or not.
DR. MILLER: Well, I think every plant is addressing the
cabling issue a little bit differently. They all develop different
databases, and we haven't seen that, but I've seen that at different
technical meetings. That's kind of aside from this issue.
CHAIRMAN POWERS: Neither of those two issues are a part of
this.
MR. CRAIG: No.
CHAIRMAN POWERS: They're continuing on their own pace.
DR. MILLER: But you have a program to address the current
15, I believe, issues that are in this.
MR. CRAIG: Yes.
DR. MILLER: And you said earlier you're going to drive
those to completion.
MR. CRAIG: Whether it's a new rule or no new requirement,
however they come out.
DR. MILLER: And you'll have a time-line on those?
MR. CRAIG: For each one we do, yes.
DR. SHACK: And you have some license renewal generic issues
that have now been renamed something else.
CHAIRMAN POWERS: That's what they were discussing.
DR. SHACK: No, no. There's many generic -- you know,
there's EPRI technical generic issues, then there's these generic
license renewal policy issues, then there's these GIs.
DR. MILLER: I think we ought to get new names for all this.
It's confusing.
DR. SEALE: Don't encourage him.
MR. MARSHALL: I'll probably spend a bit of time on this
slide here.
DR. KRESS: Does this slide mean that, for purposes of
generic issues, we defined adequate protection in terms of CDF and LERF?
MR. MARSHALL: That's one of the comments we got. That's
not consistent with our definition of adequate protection. So, this is
probably subject to change.
I was actually going to skip this first flow chart on this
slide.
DR. KRESS: But if you're going to quantify it in terms of
risk or dollars or something, you have to use something.
MR. MARSHALL: Right.
DR. KRESS: And you can't use the standard adequate
protection definitions, because it's non-quantifiable.
MR. MARSHALL: Right. It's something we're going to have to
work through. The initial cut was -- to some of us, it seemed
desireable to do this, but we're not sure it's consistent with how --
legally with how the NRC handles --
DR. KRESS: For generic issue, do you think you have to be
legally -- I mean this is a process internal to the NRC to help you do
your job.
MR. MARSHALL: Right.
DR. KRESS: I applaud you trying to put something that's
quantifiable as a surrogate for adequate protection. Let's call it a
surrogate. I'm just not sure CDF and LERF incorporated all the things
you might be interested in, that could be quantifiable. That would be
the only comment I have.
But I would encourage you to continue in that direction,
maybe look for other things, like person-rems.
DR. WALLIS: This might be a starting point. So, the
minimal thing is to calculate CDF and LERF, and then you might also want
to do something else.
DR. KRESS: Well, I think I read somewhere in there you
talked about frequency -- product of frequency and person-rems, which
might be another thing that would capture some other adequate protection
issues that maybe LERF and CDF doesn't capture.
I wouldn't think, legalistically -- I'd think what can I do
to help my process.
MR. MARSHALL: We thought this would be nice to help the
process. It was raised during our internal review as a concern that we
had to be aware of.
CHAIRMAN POWERS: On your previous slide, one of the items
in your four bullets that you had at the top of that slide was to define
the scope of the issue, which is done up in this flow chart that you had
marked in red.
MR. MARSHALL: Yes.
CHAIRMAN POWERS: And then it comes down into the group that
actually starts working to develop the resolution on this.
If they work along and said, well, this guys defined the
scope of the issue all right but they were wrong and it should be
different, the scope needs to be different or changed or augmented or
shrunk, something different, what do they do?
MR. MARSHALL: They take it back to the review board.
CHAIRMAN POWERS: At the top or someplace else?
MR. MARSHALL: Right there, wherever they happen to be, this
group, even though most of their work is done here, they follow the
issue -- they're supposed to follow the issues along. Same with the
generic issues project manager.
So, if there's a question that the scope should be changed,
they should bring it to the attention of the review board, schedule a
meeting, and get the scope formally altered and document why it was
changed.
CHAIRMAN POWERS: And that review board gets together and
says, nah, we like the scope it was, and the guys that are doing the
work are from NMSS, and they said, no, we don't agree. What happens?
MR. CRAIG: They'll work the management chains to make it
get up and over. The idea was that the board was roughly at the branch
chief level, and if, indeed, the individuals feel strongly, it will get
bumped up and then come over at a higher level.
CHAIRMAN POWERS: You understand what my concern was, and I
think you've given me the right answer, is that you've got these branch
chiefs making decisions and I've got the director of NMSS making
decisions, and he's going to pull rank on these guys, and I don't care
how many of them there are, he's still going to pull rank, and what
you're saying, well, if that happens in the management system and if it
can't be resolved at the office director, it goes to the EDO level, and
that's probably the only answer that there is to that.
MR. MARSHALL: The other thing we have done in the technical
screening is revised risk assessment guidance.
Essentially what we did here was we went through NUREG-1489.
That's a review of NRC uses of PRA, and we essentially
lifted out of Appendix B, if I recall correctly, the guidance they had
there for technical -- for using PRA for screening, and then we reviewed
that against Regulatory Guide 1.174 just to see if they were consistent,
and we believe that the guidance we have in there is consistent with
those two documents.
Some of the changes -- one of the major changes was to
recommend to the staff that they use the most current available PRA.
Another thing is also to use what's called risk transforms
or trying to figure what the early releases were, to use the more
current source information.
CHAIRMAN POWERS: I guess I don't understand, quite. I've
got a generic issue, means it's applicable to at least a class of
plants, and so, now I want to understand something about the risk
significance of this, and I want to use something from a PRA to do that.
What PRA do I use?
MR. MARSHALL: Well, the person doing the analysis should
look through the PRAs and pick the one that's closest to the affected
plants, and if there's not a single one, he should pick multiple PRAs.
CHAIRMAN POWERS: Whose PRA does he use?
MR. MARSHALL: Essentially a search through whatever is out
there.
CHAIRMAN POWERS: Oh, okay.
So, he goes and he grabs a PRA from some amorphous group.
This could be a PRA that was done back in the days when methods were, at
best, crude and whatnot? Well, that doesn't sound very good to me.
MR. MARSHALL: Well, he shouldn't have picked that one,
because it didn't meet the criteria of being a current PRA.
CHAIRMAN POWERS: How does he know?
MR. MARSHALL: Whoever is doing the analysis -- one criteria
we're going to have in the program is that they be familiar with the
PRAs they're using, that they're not just merely a black box to them and
they're just changing numbers in SAFIRE or something.
They should know what the assumptions were, what tools were
used to develop the PRA, and that's the person that should be doing this
analysis here.
Tom King's group will be -- if Research wants to do more of
this work in-house, it will be done out of that group.
CHAIRMAN POWERS: Okay. That's good, because those guys now
are subscribed two-and-a-half times their available working hours or
something like that, and so, now you've got to know another 10 percent
of that. So, you've got to 2.6.
Tom's a busy fellow here.
MR. MARSHALL: No argument there.
CHAIRMAN POWERS: The problem is that you've got these
people that are becoming a focus as you moved to risk-informed
regulation. I don't have a solution to this problem.
I can tell you the lengths and the breadths of it is that,
if I want a PRA for Mark I boilers, that's a pretty good class and
liable to have a generic issue on Mark I boilers.
I've got a lot of old ones, and I've got a bunch of things
that were used for the IPEs, and I know that those have never been
blessed, and I know the old ones are just exactly that, old, and may not
have the best techniques.
So, it strikes me you've got a hole here someplace, but I
got some resources, too, in thinking about a solution.
I got some SRAs in every single region, and they have to use
something for their risk assessments, and they might be a pretty good
source for what you should use in these generic issues, because they're
going to be as up to date as anybody I can think of.
MR. MARSHALL: That's a good suggestion.
MR. CRAIG: Tom and his staff have participated in this. We
have John Flack and others helping us, and he made the same point about
limited resources and how much he can support this.
CHAIRMAN POWERS: I mean every group that comes in here says
Tom King's going to help us on this.
DR. MILLER: Which Tom King?
CHAIRMAN POWERS: I've been walking out of this building at
seven o'clock and seen Tom walking in after dinner, so I know that he
puts in a few hours here, but I know there's also a limit.
DR. KRESS: You're going to end up identifying a set of
plants for a given generic issue, I'm pretty sure. Each of those plants
will already have calculated a CDF and a LERF. So, you don't really
have to re-calculate those, if you believe what you've already got.
So, that box really only needs to calculate a change in CDF
and a change in LERF.
Now, those changes generally are plant-specific for a given
issue, but there may be a way to conjure up an easier way to get a delta
than to get the actual. You don't have to recalculate the actual every
time.
This could be a place where a little research on how to do
the PRA could be very useful, how to calculate a delta with a
generic-like PRA that would be applicable to more than one plant. I
would be more apt to believe a delta, the generic number, than I would
the bottom line.
CHAIRMAN POWERS: I'd harken back to the IPE insights
document that's pretty clear that the assumptions and methods are so
different that inter-comparison even from sister plants is very
difficult.
DR. MILLER: Even on delta?
DR. KRESS: Yes, but what you kind of want to do here, Dana,
is -- this is a screen, and you want to pick -- I think you can pick out
a representative CDF and representative LERF for this class.
CHAIRMAN POWERS: I can take an average.
DR. KRESS: I don't think an average is the right thing.
CHAIRMAN POWERS: I can take a mode, a mean, from an
unreliable source, okay? I'm not wild about this.
DR. KRESS: Well, for screening purposes --
CHAIRMAN POWERS: I think you have to be careful about that,
because these guys are a little more aggressive than just screening.
They're also going to define the depth and breadth of the issue.
DR. KRESS: Well, I'm interested in what kind thresholds
they're going to have, also. Would your threshold for CDF be 10 to the
minus 4, for example?
DR. MILLER: Are you talking about threshold on CDF or
change in CDF?
DR. KRESS: Well, I'm looking at CDF and LERF as -- well,
actually LERF, and CDF, as a surrogate for adequate protection, and I
want to be sure that --
MR. MARSHALL: We intended to use the figures -- I think
they're three and four -- out of Reg. Guide 1.174 for the thresholds.
CHAIRMAN POWERS: That's good. I like those figures,
because they've got a plot right across the horizontal that says CDF,
not CDF during normal operations, not CDF just during external events,
it's CDF.
How do you find a -- where do you go on the horizontal axis?
DR. MILLER: Tom King will do it.
[Laughter.]
MR. MARSHALL: Just one moment.
DR. KRESS: In principle, your concept is logical.
CHAIRMAN POWERS: Yes, I think that's the take-home lesson.
I don't think we've said anything that impacts the outline of your
documents.
DR. SEALE: I do have a related question. In some of the
criticisms that we and, I guess, others have voiced of some of the past
generic issue resolution efforts, we observed that the solution or,
let's say, the considerations that went into the judgement that the
issue had been resolved stove-piped, if I may use the term, the breadth
of the issue as it was originally defined.
That is, you looked at a fairly narrow version of that
particular problem and there were other cases where things were really
under the same generic issue, but they weren't considered in the
analysis that made the judgement that it had been resolved.
I assume you're going to try to defend or guard against that
kind of tunnel-vision, I guess you could call it.
MR. MARSHALL: That's something we need to be aware of as we
try to keep a stronger hold on the scope, that we don't force ourselves
into more of that.
DR. WALLIS: I have a concern that we see a lot of plans and
things like that here, how things are going to be done.
My advice to students who are planning anything is to, as
soon as possible, immerse yourself in doing something and seeing if
something works along the lines you envisage, because you may change
your plan once you actually try to implement something like the plan you
have in mind. So, try it out as soon as you can on something. Get some
experience with doing.
MR. MARSHALL: Okay.
I'd just like to touch on the third bullet for a moment,
revisions to the cost estimation guidance.
One thing we're going to do -- and again, this section
hasn't been completed yet -- is recommend the use of the current OMB
discount rate, whatever that happens to be, to use the net present value
in calculating cost, and also, when you're assuming -- especially for
relaxation of regulations, right now the assumption is to assume that
all plants will take advantage of the relaxation -- is to try to get a
feel for how many plants would take advantage of the relaxation, to
decide if the NRC is willing to spend the amount of money to relax a
regulation.
The estimate would be different if you assume all plants
versus, let's say, only three plants are interested in a particular
relaxation of the regulations.
CHAIRMAN POWERS: OMB, in their guidance, don't they -- they
give you a discount rate out to a certain number of years and then the
cost of money after that is zero? Is that correct?
MR. MARSHALL: Not sure. Not sure at all.
CHAIRMAN POWERS: I think you want to go out 10 years.
DR. KRESS: Well, the thing is it's likely to change
drastically. It's very low right now, and the question is, if it
changes drastically, do you go back and redo the calculation?
MR. MARSHALL: Right now, in the current -- the guidance we
have, I believe they recommend we use seven as the discount rate.
CHAIRMAN POWERS: It's a pretty high value for their
recommendation relative to current environment, but the problem is you
frequently run into, when you plan long-term activities and, say, a mean
of 35 years for the existing plants, assuming the license renewal, the
value of money after -- I think it's 10 years -- is zero, and so, it
gives you a -- you push all decisions off to beyond that time because it
didn't cost you anything to do them.
DR. WALLIS: Are there really decisions which are like that,
which are so border-line that you have to worry? I would hope that some
of these decisions are clearly one way or the other, in fact that many
of them are. They're overwhelmingly beneficial or not.
MR. MARSHALL: The next bullet is just in the screening
stage. We're going to provide guidance for how to screen a burden
reduction issue. That was left more qualitative in the former process.
We would like to quantify that, also.
The biggest change in here -- and this goes exactly opposite
of one of the ACRS recommendations, and we should bring it up here -- is
number five.
Before in the prioritization, when we're calling this
prioritization, when we screened an issue, it came out as being high,
medium, low, or drop.
Essentially, high meant -- high and medium, we would work on
those. Lows -- they would just sit in a bin somewhere until they either
moved up to high or medium or they moved down to drop, and drop we just
didn't think about anymore.
And mainly, that was done to help us with our budgeting.
If it was high, you're supposed to seek funding for it.
If it was medium, you're supposed to wait till you're done
with your highs you had funding available, work on those, and if it was
low, it was at the discretion of the division director if he wanted to
seek funding for it or not.
One problem we have or we feel we have with labeling things
high, medium, or low, instead of being an indication of the priority it
should be in the budget, we thought it was -- it appeared to some to be
also a priority as the safety significance or, rather, the safety
deficiency associated with an issue, and so, we want to get away from
that appearance in the sense, since we know we're working on mediums and
highs, we just call those go's, we'll continue working on those, and the
lows and drops, since we just sat those aside anyway, we're not going to
work on those, and that's what the drop here represents.
Something before that would have been low or drop is now
just drop, and things that will continue through the process are mediums
and highs, and later on -- and it's indicated in this checked box here
-- I guess as part of the research self-assessment that's going on now,
they're coming up with a prioritization scheme for all research
activities as part of the PBPM process, and these issues, the
substantial safety enhancement and the burden reduction issues, they
would be prioritized against all the other research work in the Office
of Research.
Is there any questions?
CHAIRMAN POWERS: I assume that there will be a historical
record that's maintained so that I don't keep putting in an issue that
is no-go that you have to go through this work on over and over and over
again?
MR. MARSHALL: Yes, there is. We'll have that documented.
That's one thing we definitely want to keep is the documentation of the
system.
DR. WALLIS: If your flow chart is right and there are lots
of candidate issues which are going to emerge, it's going to soon become
pretty clear that you'd better work on the most important ones, and the
payoff on those is going to be pretty obvious.
So, you don't need to be worried about splitting hairs. You
need to be worried about getting the job done.
MR. MARSHALL: And that's what the technical screen is
supposed to help us do, pick out the ones that are most worthwhile, the
biggest bang for the dollar, and work on those.
DR. KRESS: You'll have to have some other sort of screen
for material license issues.
MR. MARSHALL: The screening here -- I should have stayed at
that -- the three flow charts on slides one through 12 -- those are for
reactors. The materials issues -- and we're working with NMSS -- they
said that's probably going to be more qualitative.
I'll just pick out one of these flow charts. Let's just
pick the first one on slide number 12 and just talk through that. I've
already touched on some of this already.
The person or team responsible for doing the screening of an
issue would calculate the CDF, change in CDF, calculate the LERF, change
in LERF, then compare that to Reg. Guide 1.174 and decide if we should
continue working on it or drop it.
Now, if it continues, they would assume a solution -- you
have to assume a solution anyway when you want to calculate the change.
You calculate the cost associated with implementing that
solution, developing that solution. You also calculate the reduction in
estimated public dose. We don't take the cancers and translate them
into a dollar figure. We try to get a feel for public health with a
estimated public dose they might receive.
And again, we take a chart that's similar to the one in
0933, essentially. We just want to check the values on that and decide
if we continue working on that or we drop it, and if it passed through
all those thresholds, it goes to the technical assessment stage, and
each issue -- type of issue would go through a similar process.
DR. KRESS: You'll have to calculate the first box before
you decide which category it goes in. Then you use that information to
say, well, this is essential to safety enhancement.
MR. MARSHALL: The decision of which box it goes in will be
made prior.
DR. KRESS: How can you do that without having that
information in the first box?
MR. MARSHALL: It will be based on what the intent of what
the issue is. If the intent is saying, okay, we think the plants are
safe enough and we just want to -- we think there's a substantial safety
enhancement we could make, we're going to label that as --
DR. KRESS: More of a judgement call.
MR. MARSHALL: But if somewhere along this process they --
and that section wasn't completed in the management directive we gave
you -- if they decide that -- for instance, that an adequate protection
concern was more of a safety enhancement, they should reclassify it at
that point to make sure it's treated properly through the rest of the
system and again document why it was changed from one thing to another.
DR. KRESS: Do you envision the different parts of this
being virtually identical to the regulatory analysis, using their rules
and their --
MR. MARSHALL: That was another source. Other documents we
used, we tried not to create anything new here. We went back and we
reviewed the backfit handbook, how backfits should be handled, the
regulatory analysis handbook. We again are borrowing from those to make
sure -- not only just not to create anything new but to make sure we are
consistent.
DR. KRESS: Let me ask you a question about that. When you
calculate costs then, are you -- what are you going to do with averted
on-site costs which of course they are talking about changing or at
least relooking at?
MR. CRAIG: That is a timely question, since I was with
Ashok late last evening talking about the plan to respond to an SRM, to
re-examine the Reg Analysis Guidelines including averted on-site costs,
and we put together a plan to move forward to develop that SECY and
whenever that comes out with the Commission's decision it will be
reflected on this one, so Mike was very correct in that that is an open
question right now.
DR. SEALE: I presume you have requested Divine guidance on
that?
MR. CRAIG: I think it is early yet, but that may be --
[Laughter.]
MR. CRAIG: -- since it is a very challenging issue.
DR. SEALE: It sure is.
DR. KRESS: Let me give you some Divine guidance. Don't
include averted on-site costs in your regulatory analysis.
CHAIRMAN POWERS: Mike, let me ask you about the flow chart
at the bottom of the page, the one you chose not to go through and it
may be just a clarification.
This is burden reduction and you go through your various
steps and you calculate the costs. You compare them against a threshold
and then you solicit industry interest. Is it the intent or is it
likely to happen that when the industry's interest is solicited that you
may have to revise your calculated costs?
MR. MARSHALL: Possibly, yes.
CHAIRMAN POWERS: I mean there is something funny about that
flow. You might want feedback into that because they may know their
costs a lot better than you do.
DR. WALLIS: I don't quite understand why industry doesn't
come first, why they shouldn't be in the position of soliciting --
CHAIRMAN POWERS: It was my presumption in looking at it was
that there are probably a lot of things that come up that you just don't
want to bother people with because you can quickly go through and say,
ah, this sounds like a good idea but it is worth $10.50 or something
like that.
DR. WALLIS: Isn't industry in the best position to assess
burden reduction potential?
MR. MARSHALL: I would say yes.
DR. WALLIS: Why don't they initiate this? Then you
respond. If they don't initiate anything, you don't do anything.
MR. CRAIG: Industry certainly could be a source at the top
to identify a burden reduction initiative. If the Staff identified it,
as we go down through and I was looking because on one of the other
slides, and I can't even read these with the bottom half of my glasses,
we had a slide that indicated where we would get public involvement and
recognizing that while we may have some good guesstimates about what
industry costs are, before we got too far down in the process of
devoting resources on what we thought was a burden reduction initiative,
we needed to get industry involvement upfront because it is quite
conceivable that what we think is a good idea that they agree with, but
they also say but that is not going to save us any money so we wouldn't
adopt it anyway --
CHAIRMAN POWERS: You don't want an Appendix S.
DR. KRESS: Does that category include reduction of burden
to NRC?
MR. CRAIG: I think we are going to have another process for
that one.
DR. KRESS: A process for that one? Okay.
DR. MILLER: Well, if you can reduce burden on industry,
oftentimes you reduce the burden on the NRC.
DR. KRESS: Of course. A lot of times you could do
something just to save NRC resources, not very safety significant.
CHAIRMAN POWERS: Right now we have got to save Tom King.
[Laughter.]
DR. MILLER: The Chairman of the Committee announced that we
have a definite hard timeline, so --
MR. MARSHALL: I could probably wrap up in a couple of
statements.
The only other thing, the one other change I would like to
notice, before the process was just sequential. It just went one box
into another. Right here we have a break here and the reason we put the
break here is to acknowledge that not all issues will necessarily result
in a new rule or new guidance, and also sometimes if industry steps
forward, like John mentioned, we will have public meetings and if they
come forward and say we have a solution, we would track it through this
"other activities" but we would still follow, we wouldn't just drop the
issue at that point. We would still follow it to see that it was
implemented and try to verify to make sure it addressed the concern we
had.
Like I mentioned earlier, much of the guidance, inspection
manual and others, are already developed to provide guidance on how to
perform or perform the functions in these other boxes here.
In closing, I would just want to mention some of the things
we think we are doing that's okay or pretty good -- resolved will be
resolved now --
CHAIRMAN POWERS: Hooray.
MR. MARSHALL: The tracking will be through all the stages
so you will just have one location or one person to call or one
organization to call to find out the status of any particular issue.
We are going to try to improve the documentation of
decision-making so you won't have to worry about why something was
switched from never resolved to high. You could get an answer to that.
CHAIRMAN POWERS: You'll learn how to spell "decision" too.
MR. MARSHALL: Yes -- oh. I thought I spell-checked it.
[Laughter.]
MR. MARSHALL: Hopefully we can improve the coordination on
defining the scope -- some issues at the very beginning between the
offices so we can avoid having problems toward the end of a resolution
or the end of working on an issue, and just revising the screening
criteria. We think those are improvements.
John has already talked about the schedule. I think we are
done except for any questions you might have.
MR. BONACA: The question I have and I realize you may
already provided the answer before but I am new here so I'll ask my
question anyway.
You are proposing a new process and the centerpiece of that
is the use of PRA to make certain determinations and already here we
have discussed what PRA and all that kind of stuff. As a minimum have
you attempted to try this process on some past issues to see how that
would come out? It's almost like trying to validate methodology, trying
to see if it makes sense. That would be the first question I have. I
think that would be probably a good exercise.
Second. I really would like to hear more about this PRA
model you are going to use because if you have a generic model in your
mind, you know, there are pet issues that each one of us have that tend
to really bias the evaluation we make. I mean we have strong beliefs
about certain contributors to risk and when they are allowed to prevail
over actual calculations they really mask actual effects, and there is
evidence of that all over the place, you know, in the past, so two
issues.
One is have you tested this process or are you planning to,
and second, is there any way to see some examples in fact on how you
would apply that?
MR. CRAIG: One of the things that you will note is that
this effort is the first effort in probably 15 or more years to take a
look at the generic issue process. We have been using risk assessments
related to prioritizing or determining whether or not to work on generic
safety issues now for some number of years.
One of the criticisms that we have received, some from some
members, is that the risk assessment, the PRAs that were used for some
of the issues were old and even some of the plants had had newer and
better risk models and methods that had been developed and used and one
of the problems with the current processes, the risk methods are not
clearly defined and it is sort of in a box, because as soon as we define
"the" risk method to use, it will be out of date or Tom's people will
come up with a better one, or something, so as Mike indicated we are
encouraging people to use current methods coordinating upfront with
whoever it is in the research staff working for Tom that is
knowledgeable in that area, recognizing that there are limitations and
key assumptions with different methods.
We have -- all the examples so far are examples of the past
which illustrate the need to change and use more current methods and
techniques. We haven't tried with a new issue to run it through the
process to go back and say which method or model should we use, what's
the criteria to make that determination, and the comment that Dr. Wallis
made a few minutes ago is a real good one.
We have got a new issue from NRR that is the boron dilution
question that came over. It's reminiscent of GSI 22 and right now this
morning we were talking about should that be the first surrogate to go
through this new process and it may likely be. As we come up with new
issues and identify them, we will be briefing the Committee on changes
to this process and the basis, and that will be one aspect of it.
DR. WALLIS: I have a simple question. The new process is
here because it's thought to be better than an old process. Where's the
problem with the old process, in the process or the fact that the work
just didn't get done?
MR. CRAIG: Well, the old process to the extent that we use
terms that were significantly different from common usage, that was a
process problem. To the extent that the old process stipulated that we
used out-of-date PRA methods and methodologies, that was a process
problem.
Compounding that was the fact, as I indicated, that
potential cost-beneficial safety enhancements over time as licensing
decreased lost their real significance, so things tended to drag out.
So that was a management problem. The fact that the tech staffs between
offices could argue for extended periods of time without coming to
closure was a management problem, not a process problem.
It was only a few months ago where we were here on a looped
LOCA and the Committee heard reasons why it should be closed and why it
shouldn't be closed from various interests. And I think that reflects a
management failure not to come to some definition of scope earlier on.
So there were a lot of problems to go around. Some were not getting the
work done; some were the process.
DR. MILLER: That was a question leading to my question. I
think overall from where I see it's been a management problem from day
1, in that when there was a priority problem of top management,
everything was moving ahead, even with maybe a process that wasn't that
good. Once the top management got onto other things and this became a
stepchild, so to speak, this thing got into trouble.
My question is, two years from now, the three who are
involved, and there are two involved, three of you involved, may be on
to new things, what's to give me kind of warm fuzzies that we won't be
in trouble three years from now, even though this process looks quite
good?
MR. CRAIG: I don't have a good answer.
DR. MILLER: Can't answer that one?
MR. CRAIG: I can't answer that one.
DR. KRESS: I thought your earlier answer on that was pretty
good, about the reports and the personal managers for each one of them
and their names will be on the reports.
DR. MILLER: That's true, but that still -- you've got this
committee setup which has designees from the top managers, but years
from now the top managers are going to have new issues.
DR. KRESS: But each issue will have its own manager. I
think that's a key.
MR. CRAIG: That's a significant difference that we've made.
In the past the input was entered into a computer data system and it was
a staffer talking to a staffer, and the first time I looked at it, there
were a number of closure dates that were two years old, and I queried
whether or not we'd be able to meet those dates. We were already past
it by two years, recognizing that we're good, some of them --
CHAIRMAN POWERS: Did you get any answers but yes?
MR. CRAIG: I didn't get any good answers. And by assigning
a specific staff person for each issue, that's a step in the right
direction for accountability. Each generic safety issue is specifically
identified in our operating plans as part of the budget that get
reviewed within the line in addition to reporting them to the Hill. So
as long as that process is in place I'm optimistic and I don't have any
reason to doubt that whoever is the manager responsible for it will be
able to meet due dates, because you don't slip due dates in the op plans
without office director recognition and review. So the dates may slip,
but there will be conscious decisions as to why.
DR. MILLER: We plan to write a letter on this report, and
do you want input on the management directive at this time from us, even
though you said it was fairly early?
MR. CRAIG: We would appreciate any comments that you would
have about risk methods, comments on the management directives, et
cetera. We have a window to go out for comment, and you'll have the
opportunity again to the extent that we can revise this to reflect
near-term comments. We'll do that. And we've discussed that with the
other offices, so it's truly a work in progress. But any comments that
you feel you want to make, we'd appreciate.
DR. SEALE: I have a question on the generic issue question
in general, and, you know, maybe you're ready to talk about this or not.
I don't know.
Back several years ago a large number of issues were -- I
won't say relegated, really, but were assigned to a category that says
well, when we get the IPEs in, we will have resolved those issues. And
that turned out to be at least in some cases a fairly hollow aspiration.
Is there an overall assessment of how many of the issues that were
allocated to the IPE program really did in fact achieve resolution?
MR. CRAIG: Yes. For each issue where there was an
assumption that it would be addressed we're going back to verify it, and
if in fact it was adequately addressed, that will be the basis for the
documentation of the closure or no additional action, and that closure
package then would come to the Committee.
DR. SEALE: You're going to do that for us then.
MR. CRAIG: Actually it's going to be Ernie Rossi that's
going to do it.
DR. SEALE: Okay.
MR. CRAIG: But it's -- I've already started the process.
The ones that are closed are the easy ones. The ones where the
process --
DR. SEALE: Sure.
MR. CRAIG: Did not close them, then we're going to have to
identify what steps needed to be taken to come up with a solution and
close the issue, and we'll have to develop a task action plan for each
one of those.
DR. SEALE: I want to congratulate you on not ducking the
hard ones, because that struck me as a hard one, and I'm really very
pleased to hear that you're not going to duck that, you're going to go
after it.
MR. CRAIG: We're going to do them -- I've told the staff
it's a lot like putting shingles on a roof. You do them one at a time.
DR. SEALE: That's right.
MR. CRAIG: And you don't move on to the next one till
you've put the number of nails in that one.
DR. KRESS: I believe you've put shingles on a roof. That's
exactly the way you do them.
MR. CRAIG: That's exactly right.
DR. MILLER: You have done that before, it sounds like.
Since I five or ten minutes ago implored us to move on and
get finished by 11:45, we're now slightly ahead, and I congratulate the
Committee for letting me do that. Any more Committee input here? We'll
discuss this letter later today; is that right?
CHAIRMAN POWERS: That's right.
DR. MILLER: So I encourage the Committee to be thinking
between now and later today to have input on the letter. We'll have a
draft or some sort of a skeleton started.
Anything else that staff here wanted to provide for us?
MR. CRAIG: No. We're glad to have the opportunity, and we
look forward to the comments. We think it's an opportunity to make this
process significantly better. Thanks.
DR. MILLER: Well, I think we at least feel like we're on
the right -- at least on first base or maybe at home plate and heading
for first base. It's a lot better than it was six months ago.
MR. CRAIG: Thank you. We think so too.
DR. MILLER: I'm sorry to use a metaphor. For a guy who
plays basketball, he doesn't understand what I was talking about.
Probably never heard of that game.
Thank you very much.
MR. CRAIG: You're welcome.
DR. MILLER: May I turn it over to the Chairman. I
congratulate the Committee on leaving a few more minutes for lunch, I
guess.
CHAIRMAN POWERS: There's no time for lunch, because several
of the Members have signed up to participate in a tour of the ops
center. I guess Jocelyn Mitchell is going to lead you on this
expedition, and so we will recess until one o'clock.
[Whereupon, at 11:35 a.m., the meeting was recessed, to
reconvene at 1:00 p.m., this same day.]. A F T E R N O O N S E S S I O N
[1:00 p.m.]
CHAIRMAN POWERS: Let's come back into session.
In this afternoon's presentation, we're going to cover one
of the topics that is, to my mind, extremely important both to the NRC
and to the industry as we move into an era of deregulated generation of
electricity, and that is the extension of fuel burn-up.
There has been a historical trend over the last 20 years of
extending the burn-up of fuel to the point that we're now up to
somewhere around 50 giga-watt days per ton, and in the course of making
that progression, we've discovered that there may be limits on how high
you can take the burn-up of fuel.
The NRC has imposed a limit currently of 62 giga-watt days
per ton.
In making that limit, they drew a regulatory judgement that
burn-ups to that limit would not pose any significant threat to public
health and safety, and they have initiated a class example of a
confirmatory research program to validate that regulatory decision.
In today's presentation, we're going to get a status report
on the progression of that confirmatory research program. We have --
were a part of planning on that research program and had commented on
that, those plans, and will be looking to see how the work has
progressed.
In our comments, we also suggested that the NRC staff needed
to look forward and determine what kinds of information would have to be
provided in order to allow the NRC to approve burn-ups that went beyond
the current limit of 62 giga-watt days per ton.
Even though the agency's decision was that the licensees
would have to provide any data or analysis necessary to make that
approval, it still seemed prudent to look ahead and understand what kind
of information would be necessary to support an application for even
higher levels of burn-up, and there is every indication that, in fact,
the industry wants to move to higher levels of burn-up.
Members will recall that, in our quadra-partite meeting,
that higher levels of burn-up are envisioned in every country, and
certainly, 75 giga-watt days per ton is considered feasible, especially
with some of the more modern clads.
So, we're also going to hear about how the staff plans to go
about finding out what kinds of information would have to accompany
applications for extended fuel burn-up, but as I understand it, we're
going to begin the presentations today with -- Margaret Chatterton is
going to discuss the perspectives from NRR on the high burn-up fuel
issues.
MS. CHATTERTON: Good afternoon. I'm Margaret Chatterton,
Reactor Systems Branch, NRR, and I'll be talking about the licensing
strategy for burn-up extension.
The presentation will be basically an update of where the
staff is on burn-up extension. I will review the process as we see it,
talk about the burn-up extension requirements that we see and then fill
you in on what's happened in the last year or so, the progress that's
been made, and the current status.
The approach is pretty much the same, has been explained
before and outlined in the agency program plan that was issued last
July.
The key points are that the NRC will be working with the
industry to develop a strategy and a plan, but it will be up to the
industry to actually develop that plan and then, with that plan, the
plan should point out areas that data will be necessary, it will be up
to the industry again to do the testing and to develop the criteria that
must be met.
You probably remember that, in the past, the NRC has done
the testing and developed the criteria. This is no longer possible
because of budget constraints.
So, at the point that the industry has developed the plan,
then the NRC would review it and hopefully endorse it as a regulatory
guide.
CHAIRMAN POWERS: When the NRC goes about reviewing it and
endorsing it, what we see are, as we go up in burn-up, changes in
phenomenology --
MS. CHATTERTON: Yes.
CHAIRMAN POWERS: -- things that we had not anticipated from
a straightforward extrapolation of past behavior.
I guess my question is how do you review something when you
don't have any experience in the phenomenological regime?
MS. CHATTERTON: Well, I think the approach is to have a
systematic comprehensive plan that all areas of operation are looked at
in order to determine where data needs to be obtained, and then the
industry will have to go ahead and get that data.
Some of that data may be from lead test assemblies, a lot of
it's going to be from other types of testing, and it's the plan that
industry will be putting together that should show us where we need to
go and what areas need to really be examined more carefully.
CHAIRMAN POWERS: I guess -- maybe you said it here.
My concern is that, in the past, when we moved to higher
burn-ups, the industry was very good about running lead test assemblies,
and we had data on the normal operations of these high burn-ups rods,
but we relied on the fortuitous experimental results for off-normal
conditions that came from abroad in order to make us aware that there
might be a problem with these high burn-up fuels.
Is the industry going to be willing to collect data for
off-normal and upset conditions?
MS. CHATTERTON: They'll have to get the data that's
necessary as part of -- that's outlined as part of the plan that they
come up with.
When we get a little bit further into this, I was going to
talk about what we expected -- what we expect as part of a systematic,
comprehensive plan, and I think maybe that will help.
CHAIRMAN POWERS: Sure.
MS. CHATTERTON: The burn-up extension requirements as we
see them and has been previously outlined are that the current licensing
requirements need to be addressed, it needs to be risk-informed.
The specific areas of LOCA, ATWS, and RAI must be
specifically addressed.
There should be a real emphasis on a lead test assembly
program and that they must be prototypical as far as power history,
types of fuel and cladding as intended to be used to higher burn-ups, it
must be based on burn-ups up to the requested burn-up, and they should
also be in near-limiting locations.
This is a change from previous work that's been done.
Usually, the LTAs were in very much non-limiting locations, and our last
experience has found that sometimes power history does make a very big
difference.
So, we're going to emphasizing the industry that we want
them in near-limiting locations so that they will have a very similar
power history to what the fuel will be operating in when it's actually
used. Better to find out that a few LTAs don't do well than to have a
whole batch not do well.
Also, as part of the requirements, we expect a good fuel
performance monitoring program.
DR. APOSTOLAKIS: Would you explain the second bullet,
please?
MS. CHATTERTON: The risk-informed?
DR. APOSTOLAKIS: Yes. What exactly do you expect them to
do?
MS. CHATTERTON: Well, I think it has to be looking at not
only accidents and the risk of the accidents to probability, but it also
has to be looking at the consequences, both parts.
DR. APOSTOLAKIS: So, they will do something like what
Regulatory Guide 174 specifies? I mean these are the criteria you will
use there?
MS. CHATTERTON: Yes.
DR. KRESS: Does that mean you need severe accident
information on the high burn-up fuel, such as how it affects the melt
behavior and the fission product release?
MR. COLLINS: Excuse me. This is Tim Collins from DSSA.
With regard to being risk-informed, there's currently a
whole set of deterministic requirements out there for fuel, and the
risk-informed is kind of acknowledging the fact that some of those may
be overly prescriptive or overly conservative, and if the industry wants
to propose something different than the existing criteria, they need to
use a risk-informed approach to do that, okay?
Now, whether the -- the specific criteria of Reg. Guide
1.174 -- I don't think we've thought hard enough about that as to how to
apply this particular fuel arena, but this is just kind of like a
high-level invitation, if you will, to the industry to say, look, if
there's things which you happen to think are overly conservative, if you
want to choose something different than what you've done in the past, it
needs to be a risk-informed approach.
DR. APOSTOLAKIS: So, the second bullet is an option.
MR. COLLINS: It's an option, yes.
DR. APOSTOLAKIS: It's not required.
MR. COLLINS: That's correct. It's not a requirement in the
sense -- we don't have a regulation out there that says you have to do
that, and if somebody wants to come in and say, okay, we can meet all
the existing fuel criteria and demonstrate that those criteria do apply
to the conditions which we believe are associated with high burn-up
operation, that's okay. It is an option.
CHAIRMAN POWERS: What if you found that the set of fuel
criteria that you developed so many years ago under such a low burn-up
regime really weren't an adequate set?
MR. COLLINS: Well, in fact, we're hoping to find that sort
of thing out by the systematic approach that we're asking the industry
to follow.
Now, if we find that it's inadequate, I mean we would
require that they address whatever -- on a risk-informed basis whatever
is found to have been previously inadeaquate, and we would have to go
back and look to see if we had to back-fit existing burn-up limits.
DR. WALLIS: What does "address LOCA" mean?
MS. CHATTERTON: It basically means go through, examine it,
examine all the phenomena, exactly what happens, determine whether and
where new data is needed, how much data is needed.
DR. WALLIS: Does this mean taking, say, fuel which has been
subjected to high burn-ups and then quenching it, doing tests of the
sort of response it would have to going through a LOCA?
MS. CHATTERTON: It may mean that, but it doesn't -- I don't
think it necessarily means that. I think, at this point, what we're
saying is it needs to be addressed, we need to look at it very
carefully, they need to look at it very carefully, and justify exactly
how they're going to treat it.
CHAIRMAN POWERS: Suppose in that regard the industry came
along and they said, look, NRC, you've done these tests on exposing fuel
to LOCA conditions up to about 60 giga-watt days per ton, maybe 62, and
we've got this magic computer code that allows us to extrapolate that up
to 75 giga-watt days per ton and it shows everything's fine.
MS. CHATTERTON: My next question would be how do I know
that the code is validated beyond 62? That's part of what the industry
would have to justify. As part of the process, every step is going to
have to be justified, and the justification is going to have to be
documented.
If the industry can come along and show that the code is
valid up there, I can't see how they can do it without data, but let's
assume they can do it, there's some -- like you say, a magical code that
can do it and that they can provide an adequate justification, then I
think we're in the position of having to accept it.
DR. BONACA: I had a different question, which is still
regarding the second bullet. Would you allow today risk-informed
approaches to the design of regular assemblies?
MR. COLLINS: Yes.
CHAIRMAN POWERS: We would allow it in principle. Nobody
has asked for it, have they?
MR. COLLINS: Oh, right. We would allow it in principle.
Nobody has asked for it.
MS. CHATTERTON: Yes.
CHAIRMAN POWERS: Margaret, let me ask you, down in
prototypic, you've indicated that power history makes a difference, and
certainly, this committee has had a presentation by the industry in
which they showed us that they typically have two varieties of high
burn-up fuel, one that burns up very quickly and one that's a more
leisurely high burn-up, and we've had presentations by you on some of
the difficulties that have been encountered with high burn-up locations.
MS. CHATTERTON: Right.
CHAIRMAN POWERS: Do we now have a documentation on the
magnitude of this power history effect on fuel behavior?
MS. CHATTERTON: I don't think we do. I don't think we have
a comprehensive -- I don't think it's something that's totally known.
It's fundamental, but exactly how they work in each case, I would not be
able to say we have anything, and I don't think there is anything.
CHAIRMAN POWERS: I'm thinking about the confirmatory
research program that's going on now. When you go about selecting
variables to investigate -- and clearly burn-up is one of them --
MS. CHATTERTON: Yes.
CHAIRMAN POWERS: -- and maybe this history is another one,
maybe the water chemistry that the fuel has been subjected to is yet
another one.
MS. CHATTERTON: Yes.
CHAIRMAN POWERS: I mean you get quite a few variables.
MS. CHATTERTON: Yes. And I don't think that you've covered
the area unless you've covered the areas that you want to have as far as
power history.
If one power history is all that's -- that there is data
available on, then maybe that's the limit of how you can approve a new
fuel.
CHAIRMAN POWERS: So, you're saying that there might be
multiple pathways here to getting approval. You may have I'll only do
fast burn-up fuel, these high burn-ups.
MS. CHATTERTON: Yes. Then that would be the only one we'd
know about.
CHAIRMAN POWERS: Yes.
DR. APOSTOLAKIS: Let me come back to the earlier comment
about the second bullet. The view-graph says that it's a requirement,
but now you said that it is not, it's an option, and I wonder how
optional it is.
In other words, you do all the calculations, you're
addressing LOCA, ATWS, and so on. Wouldn't you be curious after that,
even though the utility may not have asked or -- or the industry may not
have asked to use 1.174 -- what happened to LERF and the core damage
frequency?
Would you ask them to address that even though they are not
proposing to use the regulatory guide?
MR. COLLINS: No, I don't think we would ask them to address
that. We're talking about the fuel surviving all the basic -- all the
different design basis transients. For there to be a hidden risk, I
guess there would have to be some sort of a different physical response
to another transient of some significant probability, and the way the
design basis transients are set up, they're typically based on what
different phenomenology could occur, and they're grouped that way.
For example, you know, loss of heat, increase in heat
removal or decrease in heat removal, and the fuel response then is
calculated in such a way that the fuel doesn't fail or is limited in
failure to a certain value.
Now, I don't -- we'd have a hard time justifying, I think,
going back and asking for a risk assessment after that.
DR. APOSTOLAKIS: Are you saying that, if the deterministic
calculations show that the fuel can take whatever abuse we think of,
then the risk really does not change?
MR. COLLINS: Yes, I think so.
DR. APOSTOLAKIS: I'll have to think about that.
DR. WALLIS: These tests determine behavior of the fuel in a
reactor. Do they extend to what happens to it afterwards? You're going
to take it and put it in a spent fuel pool or something and things are
going to diffuse and some changes perhaps will happen post-irradiation.
DR. UHRIG: They'll be examined, won't they?
MS. CHATTERTON: Well, the LTAs will certainly be examined.
As part of the LTA program, there will be post-irradiation examinations.
They will be both on-site and there will be a good number that are
destructive in a hot cell.
DR. WALLIS: If you've seen a spent fuel pool, does nothing
more happen to this after that? It's harder to do a test of a long
duration.
MS. CHATTERTON: I'm not sure what more would be.
CHAIRMAN POWERS: Well, imagine that you've heavily hydrited
the fuel --
MS. CHATTERTON: I'm sorry.
CHAIRMAN POWERS: -- heavily hydrited the clad --
MS. CHATTERTON: Okay.
CHAIRMAN POWERS: -- then you stick in water and you wait --
and you wait a while, and metal hydrites and water are usually not a
compatible combination.
MS. CHATTERTON: Correct.
CHAIRMAN POWERS: And so, after a year-and-a-half, you get
breakthrough and you start getting reactions of the hydrites with water.
That's one conceivable thing that could happen and probably a spirited
event if it did, especially if you get localized high concentrations of
hydrites.
MS. CHATTERTON: I don't know.
DR. WALLIS: I guess your answer is that you're going to
look into it.
MS. CHATTERTON: Yes.
DR. WALLIS: Or somebody has.
MS. CHATTERTON: Yes, that will obviously be looked at.
DR. SEALE: And you're telling us that the NRC will have the
-- in its discussions with the definition of the program and so on --
will be in a position to include data points in the test matrix that
will allow an evaluation of these kinds of concerns.
MS. CHATTERTON: It may be bounded by other things that
they're looking at, and that may be the way it's handled.
DR. SEALE: Fine. But if it's not and you come up with
something which is a little corner of the world over here that they
didn't include in their test matrix, then, you know, prudence says
somebody has got to stand up for that.
MR. COLLINS: Ideally, where this needs to be included in
early on in the systematic assessment of what's most important and
what's not, okay, and if it comes out in that assessment to be
significant, then that's where we'll have to get data if we want to
address that issue.
MS. CHATTERTON: That's right.
DR. SEALE: So, this represents a much more aggressive input
into the definition of somebody else's text matrix than you've ever had
or even insisted upon before, and I'm not saying that's bad.
MS. CHATTERTON: Yes, that's absolutely true.
DR. SEALE: I think that's great. That's the best way to
spend your money, is to spend it on the problem.
MS. CHATTERTON: There's no question that this process will
be much more comprehensive and much better documented than we've had in
the past --
DR. SEALE: And aggressive.
MS. CHATTERTON: -- yes, and aggressive than we've had in
the past.
MR. COLLINS: It should be aggressive, but hopefully it's
aggressive in that it focuses on what's most important, not aggressive
in the sense of let's go do a whole lot, where the aggression is
supposed to be at whittling away what's not important, because it's
going to be expensive to get data in, we've got to make sure we're
focused on the important data.
CHAIRMAN POWERS: Can we move on to the fuel performance
monitoring program in there? It's the last line on your slide.
MS. CHATTERTON: Yes.
CHAIRMAN POWERS: What are you looking for there?
MS. CHATTERTON: How the fuel behaves under all kinds of
situations they might normally go through, exactly how well do the codes
predict a lot of the properties that may be very important in accident
conditions.
I look at this as being able to give us a real handle on are
the parameters going into the accident analysis really correct.
CHAIRMAN POWERS: You're actually looking for some
destructive examinations?
MS. CHATTERTON: Yes. There will definitely be.
CHAIRMAN POWERS: I don't think you cover it later on, but
if you do, just let me know.
There is a persistent concern that comes up over the new
source term we have, which has been developed, by and large, based on
analysis of 34-, 35-giga-watt day fuel and whether it's applicable to
these burn-ups within the current limits but high burn-ups, and even
that concern just gets worse and worse as you go beyond the current
limits.
What's the thinking now? Can they use the new source term,
or do you have to modify it?
MS. CHATTERTON: I'm not the right one to answer that.
Ralph has got that in his presentation. I'm sorry. We'll get to
Ralph's presentation quickly, I think.
The current status -- we had several meetings with NEI, and
I will probably interchangeably say NEI and the industry, because NEI is
actually coordinating this, we're working with them and the industry all
together, and they're working on a comprehensive systematic plan.
As a matter of fact, we met with them yesterday, and I'm
going to tell you a little more about what we got out of that meeting.
In parallel, the Office of Research is planning work in selected areas.
Now, at yesterday's meeting with NEI, NEI proposed a process
to use -- to determine limiting criteria as well as the combination of
methods, programs and data that will be appropriate to demonstrate
meeting those criteria.
I want to stress that right now, the process is still being
developed. The staff has just had a chance to look at it briefly. We
will be examining it a lot more, we will be giving our comments to NEI.
Initially, our reaction was favorable. They're really starting at the
big picture and working all the way down, looking at all areas of
operation, normal operation, the transients, the accidents, to see
exactly which areas need further work and which areas need a lot more
testing.
This is part of where the industry will be documenting the
justification for each step of the process, and for anything that's
eliminated or put aside so we don't have to work on, they will have to
document the justification for why that's acceptable, and we will be
working with them as that process works along.
The process will provide a, as I've said, a comprehensive
systematic plan such that we will know when we're to the end. It won't
be, well, maybe we should do some more here or there or -- and you've
gone on and never quite shown when you're finished.
We'll be having regular meetings with NEI probably every
couple of months or so, I believe, and we'll continue to work on this
process.
There's another area that I would like to mention that we
talked about yesterday, and that was that the Office of Research invited
the industry to participate in a PIRT process on specific accident
scenarios.
The industry's reaction was favorable. Of course, they have
to check resources and be sure that all that works out, and Ralph Meyer
is going to describe all of that process a lot more in his presentation.
DR. WALLIS: The last bullet here on research --
MS. CHATTERTON: Yes. Research. I meant Office of
Research. Sorry.
DR. WALLIS: Surely there's a history here. I mean, people
must have known that there was an interest in how far you could push
burn-up. I would naively expect research would be going on for a long
time. There's a lot of basis for it that someone could give us a
perspective on -- this is the current status of knowledge and this is
where the gaps are. It appears as if all this is starting from the
beginning.
MS. CHATTERTON: No, I don't think there's going to be a lot
of starting from the beginning, but I think where -- what we're looking
at is, let's get it -- let's this time take a little bit different
approach, take a broader approach, and maybe you're right, start at the
beginning, not to do research, but start at the beginning to document
everything as to how you get where you are and why what --
DR. WALLIS: Research is usually what you do ahead of time.
DR. POWERS: Maybe I can interject here and just recount the
history or ask Ralph if he can recount the history. But the truth of
the matter is that NRC used to have a very aggressive research program
in the area of fuels behavior, and as anyway research program that's
been ambitious, they quickly got into the state of diminishing returns,
and the press of other interesting areas for research arose and a
decision -- I think I'm fair in characterizing it -- was made that that
research program had to be allowed to atrophy and atrophy almost to
zero.
DR. WALLIS: Well, if there were diminishing returns, that
sort of implies to me that a lot had been achieved.
DR. POWERS: Oh, a huge amount had been achieved.
MS. CHATTERTON: Yes.
DR. WALLIS: Maybe sometimes, it's nice to have a
perspective on how big that achievement is --
DR. POWERS: Oh, it's very --
DR. WALLIS: -- and all that needs to be done.
DR. POWERS: It's very significant. The point was that the
research program was reduced to minimal subsistence levels or well below
that perhaps, and then along came some research results that suggested
-- well, what had happened was that fuel burn-ups in excess of the
database were being achieved in the reactors, and along came some
results from abroad that suggested that there were greater
vulnerabilities.
At the same time within this country, some operational
difficulties that Margaret would probably be glad to review with you
chapter and verse since she was the point person on that --
MS. CHATTERTON: Yeah. Several of them.
DR. POWERS: -- arose.
DR. WALLIS: You said that they were achieving higher
burn-ups in reactors than the existing research database?
DR. POWERS: That's right.
DR. WALLIS: That sounds extraordinary in terms of strategy.
DR. POWERS: I don't think there's anybody in this room will
disagree with you --
MS. CHATTERTON: Yes.
DR. POWERS: -- about the unusual character of it. It was,
again, a case of confidence you could extrapolate, and indeed, had there
not been a change in physics that occurred at some level of burn-up, it
probably would be possible to extrapolate the codes.
So the Agency has found itself in a position of having to
make quick regulatory decisions, which they have done, and I think we
have examined those decisions as a committee in fair detail and been
fairly supportive on that, and it's a classic -- I mean, this is a
classic example of how you use confirmatory research to validate
decisions.
In addition, they've gone on and said, if you want to go
beyond what the have as current level, you, the applicant, will have to
supply the data, and that's what Margaret has been outlining for us.
DR. WALLIS: It's also a classical example of how you fail
to do anticipatory research.
DR. POWERS: I think it's fair to say that the -- that it
would have been very useful if the --
DR. WALLIS: I don't know why you're answering my question
rather than --
DR. BONACA: I just have one comment I would like to make.
I made it before privately. For those as old as I am, I can remember 25
years ago, some experiments showing that the irradiated fuel had a much
lower threshold for failure than -- you know, in calories per gram, than
new fuel, and I don't know why that perspective did not end up in the
Reg Guides that supported the 280 calories per gram position for, you
know, reactivity insertion. But I don't know. It will be an
interesting exercise to go back 25 years and figure out. That must have
been illogic, because I still remember that I have -- I think I have
that paper somewhere in my garage.
MR. MEYER: This is Ralph Meyer from the Office of Research,
and I think I know exactly what you're referring to. There was one test
in the spurt reactor which produced a low enthalphy fuel failure at 85
calories per gram.
DR. BONACA: Correct.
MR. MEYER: And I can tell you that the rationale, which was
not valid but nevertheless believed around here, was that it was a
flawed test, that there was a preexisting defect and some waterlogging
in that fuel rod.
Many years later, when we went back and checked as carefully
as we could and actually were able to find some of the original
experimenters to talk to, we find that that's not the case, and it was
indeed an indication of things to come had we only recognized it.
DR. BONACA: Because I remember another test at 115 calories
per gram that was a failure.
MR. MEYER: Well, now, that didn't occur until late 1993,
and we learned of that in 1994, and that's the one that really got the
ball rolling.
DR. BONACA: Right.
MS. CHATTERTON: Are there anyway other questions or are we
ready to move on to Ralph's presentation?
DR. POWERS: I think we can move on. Thanks, Margaret.
MS. CHATTERTON: You're welcome.
DR. POWERS: Oh. Margaret, let me ask you one question.
MS. CHATTERTON: Sure.
DR. POWERS: You've mentioned NEI throughout this last
slide, and this committee is familiar with a research program being
conducted under the auspices of EPRI. Are these one in the same
activities or complementary activities?
MS. CHATTERTON: They're not one in the same, but they're
working together.
DR. POWERS: The usual kind of relationship between NEI and
EPRI?
MS. CHATTERTON: Yes.
DR. POWERS: One's kind of an experimental arm and the other
one is kind of a policy arm?
MS. CHATTERTON: One's a policy regulatory arm, yes.
DR. POWERS: Thanks.
MS. CHATTERTON: That's exactly it. And the various
utilities will be working in that as well.
DR. POWERS: So it's a fairly coordinated --
MS. CHATTERTON: Yes.
DR. POWERS: -- activity by the industry.
MS. CHATTERTON: Yes.
DR. POWERS: And I assume the fuel vendors as well.
MS. CHATTERTON: Oh, yes. Yes. If I didn't mention that,
all the fuel vendors are involved in this process, as well as utilities
and NEI and EPRI.
CHAIRMAN POWERS: Welcome, Ralph, and I, in case you hadn't
detected, do recognize that there are some new members on the committee
since last time you spoke here, so you may find it useful to give a
little more background that you might ordinarily have anticipated.
MR. MEYER: Okay.
I'm Ralph Meyer from the Office of Research, and Dana
mentioned a few minutes ago that there were some changes in physics that
took place at the higher burn-ups, and since there are some new members
here, let me just mention two of them.
You can keep them in the back of your mind and they will
underlie a lot of the discussions that are going on now.
One thing that happened was that, with the increasing amount
of fission gas that was being held up inside of the uranium oxide
pellets, there was a change in micro-structure in the pellet, and this
affected things like the rate of release of fission gas to the gap.
The other major change in physics that occurred in the
mid-burn-up range of 40 to 50 thousand mega-watt days per ton had to do
with the oxide layer on the cladding.
The oxide layer on the cladding begins to develop
micro-cracks in it, and the cladding loses its effectiveness as a
barrier for oxygen transport into the metal surface of the cladding.
So, these two things, one of them that affected the pellet
and one of them that affected the cladding, changed rates so that we
were no longer able to just extrapolate from our previous rate
equations, and those two changes then had big effects on lots of things.
So, now let me --
DR. FONTANA: The effect of the oxide layer on the cladding
-- is that due to water chemistry primarily?
MR. MEYER: No, I think it's just due to the density of the
oxide, and when it gets thick, it just can't support an adherent layer.
CHAIRMAN POWERS: The Bedford-Pellum ratio on these things
is dicey, and I think once you get beyond a certain thickness, that
you've got to relieve stress some way, and cracking up is nature's way
of relieving stress.
DR. WALLIS: Could you say a bit more about these two
changes? You said there were these changes in physics. Are these
incremental sort of slow changes of a differential nature where they
sort of slowly increase, or are they of the type where you approach
rather rapid deterioration so the fuel becomes mushy and porous and
bursts of gas come out and things?
MR. MEYER: As far as we can see from the kinds of things
that we observe, like rod pressure or gas release or just measured
amounts of oxide accumulation, the changes are gradual. You can see the
change in the rate, but I'm not aware of any precipitous change that
occurs. I don't think there is.
CHAIRMAN POWERS: The precipitous change is when you
saturate and start getting hydride precipitous changes when you saturate
and start getting hydride precipitations in the clad.
That one will be non-continuous, and in truth, physical
manifestation of the rim effect, where you can detect it, is a step
change.
Now, that development, I think you're right, is a gradual
thing, but when you can actually see a rim effect, looks like almost a
step change.
MR. MEYER: Yes. We've now mentioned some additional terms,
rim effect and the hydrides. It gets complicated. If we go into that,
we won't cover what I am supposed to cover.
So, you can go ahead and put the first slide up for me,
Harold.
The agenda says that we're going to talk about PIRT and
about the status of the Cabri program. Sometime prior to the meeting I
had some additional discussion with Med El-Zeftawy and with Dr. Powers
and was told of some additional subjects of interest, and this is the
list of the subjects of interest, and I'm going to try and cover all of
them.
We have until 1:30. Is that correct?
CHAIRMAN POWERS: 2:30.
MR. MEYER: From this list, I'm going to spend most time
talking about the first item and the list item, the PIRT and the source
term, and I'll try and move over the others quickly if you let me.
[Laughter.]
CHAIRMAN POWERS: Well, I'll point out that the first entry
and last entry are the areas where the committee has some obligations.
So, maybe we better let him focus his efforts as he sees fit.
MR. MEYER: It was already mentioned that we are interesting
in a PIRT.
What is a PIRT? I think there is at least one person on the
committee that knows a lot more about a PIRT than I do. So, I beg your
indulgence.
I have never done a PIRT. The acronym standards for
Phenomenon Identification and Ranking Table. We use these to try and
learn what is important, to learn what is the problem, so that we can
address that.
We think it will help us determine what data are needed to
assess or modify regulatory fuel damage limits and the codes that we use
to show that we meet the limits.
We call this a PIRT. We haven't done it yet. I know that
there are a lot of sophisticated techniques that have been developed in
the thermal-hydraulic area, and let me just say that we're trying to use
the concept.
We may not satisfy all of the detailed procedures that have
been used before, but we're going to try it. We discussed it yesterday
briefly in this meeting with the industry that was mentioned, and we
hope that they will pitch in with us and work on it.
I have a series of slides following this.
DR. WALLIS: If all you've got is expert opinion and there
isn't a good knowledge based behind it, you shouldn't take it too
seriously.
MR. MEYER: Knowledge base -- you mean hard experimental
data.
DR. WALLIS: Yes.
MR. MEYER: I agree, and one of the things that we're trying
to do is to see if this experimental base is adequate.
Let me try and -- with the next few slides, to think through
an example of a PIRT activity as we imagine it, and then you can see
what we have in mind and you can comment on it, and perhaps the
industry, who has at least one representative here, will see in a little
more detail what we were trying to talk about yesterday in outline form.
The PIRT is based on scenarios, and in a review of our high
burn-up activities that was performed in the last couple of years, we
developed a plan, and in this plan, we said that there were three areas
that were important enough from both a risk point of view and the
regulatory point of view that we were going to engage in research on
them, and they included the loss of coolant accident for both types of
reactors and then a reactivity accident of a different type for each
reactor.
In the PWR, the reactivity accident is a rod ejection
accident, where you postulate that the housing in the upper head breaks
and the control rod drive, along with the control rods, are ejected
under the pressure differential.
In the boiling water reactor, we had gone through some risk
and some probability arguments to conclude that we would be better off
looking at oscillations, power oscillations related to ATWS events than
looking at the control rod drop accident in the boiling water reactor.
So, these are the three major accident types that I think of
as sort of the gate-keepers, the thing that keep you getting into severe
accident regions, and we have regulatory criteria that are to keep the
fuel in a coolable geometry, to ensure long-term cooling and reactivity
control for these, and so, this is the focus of these research programs
and, therefore, the focus of the PIRT that we are hoping to do.
DR. WALLIS: Even without accidents, there are limits to
burn-up, presumably. Why this focus on accidents? That's part of it,
but it's not all of it.
MR. MEYER: Well, it comes from -- it's not all of it, but
it's the risky part of it. We have -- all this stems from our
traditional regulatory approach --
DR. WALLIS: It must be risky, for instance, if you
irradiated it and then stuck it in the fuel pool, and a year later,
because of some chemistry and physics which you hadn't anticipated,
you've got really exciting events in the fuel pool.
MR. MEYER: Yes.
DR. WALLIS: It's not an accident, this type.
MR. MEYER: Okay. Well, let me then back off and say that,
in my part of the organization, we are concerned with the reactor
events, and there is another group of people who are interested in the
spent fuel concerns. We do work with them. We have work going on right
now in that area, and it's not included in what I am trying to talk
about here.
DR. SHACK: I think it's probably also fair to say that, if
you get enough hydriding to cause you to worry about the spent fuel
pool, you're going to have problems in some of these other situations.
CHAIRMAN POWERS: And it's also true that high burn-up
impacts things in the transportation area, as well, and it's just
another area that we're going to have to address at one point or
another.
MR. MEYER: Okay.
So, this is the outline of the PIRT.
I'm going to try and walk through some example on the
reactivity accident for the PWR, and at the end of the PIRT activity,
then we will try and reach some conclusions on the adequacy of the
database and the particular licensing criteria and codes, and that step
I think of as a separate step from the PIRT.
We do the PIRT to identify the phenomena, and then decisions
are made based on that.
I make this distinction because we have to worry about how
closely do we work with the industry on certain things, and I think,
just like data collection, we can work objectively with the industry on
identifying and ranking phenomena.
When we then start reaching conclusions, we probably will go
our separate ways.
CHAIRMAN POWERS: Under phenomena on your slide, Ralph, you
have identify main phenomena and fuel behavior. I take it that's surely
broad, that it could include core behavior as well, things like boron
absorption on the clad and things like that.
MR. MEYER: I realized, in fact, as we get on through the
slides, as I was thinking about them before coming down here, I realize
that there are some obvious omissions in the lists and things that I
have here, but the idea was, not having done the PIRT, to sit and just
imagine what it might be like to do the PIRT in sort of a sketchy way
and at least give us a tangible example to look at.
CHAIRMAN POWERS: You've quickly learned what the value of
this is, that it gets awfully complicated, and asking one person to do
this -- you always leave something out.
MR. MEYER: First we select the scenario, and again, just
guesses, just put on here to show that you're going to need to be fairly
specific about the details of the scenario that you're picking.
Hopefully, it won't just apply to one plant type, we can do
one PIRT for all PWRs, I would say, but this is something you have to
discover in the activity, and also, I put a limit on here of 62
mega-watt days per kilogram on the burn-up.
DR. WALLIS: I guess I'm naively asking -- you're doing this
PIRT because no one's ever done the test. Is that why? Because this
looks like something you could answer by doing it instead of asking
about it.
MR. MEYER: Which test?
DR. WALLIS: If someone had done this test, you wouldn't
have to ask experts what might happen if we did it.
So, then why do you have to ask experts? Why have a PIRT if
you've got the evidence?
MR. MEYER: Well, let's see. I guess the right answer to
that is that we've done it halfway.
We do have some tests at 62 giga-watt days per ton. They
are tests that were done with what we now believe to be the wrong shape,
the wrong coolant environment, and we believe some of those affected the
results.
We're not quite sure how important those effects are,
whether this database is bad enough that we have to have a better one to
go forward or, as some people think, it might be good enough to live
with if we just recognize the shortcomings.
In addition to that, there is the desire to go to 75
giga-watt days per ton, or thereabouts, and we don't have tests up
there.
So, if we arrange for some cooperative work on this PIRT
with the industry, then I think we would immediate move this figure up
to 75 and would then be able to, with the one exercise, to be able to
address both the NRC's rule of confirming the current approved burn-up
ranges and the industry's interest in providing a basis for going beyond
that.
Let's go on.
So, here I have taken a guess at what some of the phenomena
are.
Right away, you've got to break the problem into several
categories of phenomena, because in this accident you dump energy into
the fuel rod, the pellet expands and applies a mechanical loading to the
cladding, the cladding responds to the loading, the cladding eventually
fails, and then fuel particles can be expelled, pressure pulses can be
generated, and you have these post-failure events.
Now, I -- this list has evolved from what I just remembered
off the cuff about a number of discussions we've had regarding these
existing data.
I really expect the list to be two or three times this long
as soon as some group sits down and really examines it, but you can see
the kinds of things that you get into, and the one that I left off
that's so obvious under the loading is just -- is the energy deposition
itself and the reactivities and the reactivity feedbacks and the shape
of the power pulse.
All of these things are important some degree or another,
and to find out the degree of importance is, I think, going to be of
significant value to us.
DR. WALLIS: Is there no chemistry that occurs you're
interested in?
MR. MEYER: Chemistry?
DR. WALLIS: Something to heat this thing up.
MR. MEYER: This transient takes place in 40, 50
milliseconds, and there is chemistry involved in terms of the corrosion,
the absorption of hydrogen, the solubility limits on hydrogen, its
precipitation in hydrides that you have prior to the transients.
DR. WALLIS: No rapid chemical kinetics?
MR. MEYER: Well, I don't think so.
Now, you have to decide the relative importance of these
phenomena, and that's going to depend on what you have in mind to do
with the results, and there are several different items that have to be
addressed that might depend on the same database or involve the same
phenomena.
One is that we currently use a fuel enthalpy criterion and
will likely continue to use an enthalpy criterion, and so, one ranking
of the importance of the phenomena would be in relation to the fuel
damage criteria that are used.
We use a plant systems code to calculate the energy that's
deposited in the rod and the power shape and related characteristics of
the pulse, and so, that would be another application of these phenomena
in order to see what was important in the plant transient codes, and you
can see right away that some things are going to be important in one
case and not in the other.
For example, the failure criteria, the fuel damage criteria
have to know a lot of about cladding, mechanical properties, applied
stresses, ductility. The plant systems code could care less about
those. It needs to know about reactivities and feedback effects and any
hydraulic involvement in this rapid transient.
So, you know, it just depends on what your application is as
to how your ranking is going to come out.
So, we have to do these three times.
Dana wants to ask a question, but just show the next slide
quickly and then we can come back. So, you end up with --
CHAIRMAN POWERS: He won't forget.
MR. MEYER: -- three different rankings.
CHAIRMAN POWERS: Ralph, in this approach that you've
outlined here, you have a slide in which you said here are the
phenomena, and I wonder, do you have to put in yet another breakdown --
I know it's adding to your complexity -- in which you say, yes, thermal
expansion is going to be very important for those fuel rods that have
burned up quickly, but it's not so important for those that burned up
slowly, as an example, and similarly, your cracking of the clad will be
important for those fuel rods exposed to a high boron borated water
chemistry but not for those that have been in low-boron-content water?
In other words, is there another breakdown where -- within
the RIA category, where you have to break things down according to
exactly what rods you're talking about?
MR. MEYER: Well, I suppose, but I guess I should ask you.
This is what the PIRT does for you.
CHAIRMAN POWERS: That's right. I mean I would think that
that would come out of it.
MR. MEYER: In the example, I haven't been good enough to
think through all of these things.
CHAIRMAN POWERS: That may be something you want to think
about in posing the questions for your expert elicitation. Do they need
to take a split on this and break the question down further?
I mean you asked your experts to do this for you. I don't
know that you can do it in advance.
MR. MEYER: Well, certainly the question of power history
has come up, as to whether it's important or not important, and so, I
think, obviously, you have to delve into that and see what are the
possible effects of power history that might get involved in producing
the outcome of this kind of transient.
We have one pretty slide here that Farouk developed last
month.
This is after the PIRT is over with. In the yellow bubble
in the middle at the top is the PIRT, and this is a slide --
DR. WALLIS: Did Farouk try to read it from the back of the
room?
MR. MEYER: No, but he insisted that I give you a handout so
that you can read it.
DR. WALLIS: Even that's a bit of a task.
MR. MEYER: It's one of these flow diagrams that show how
you make a decision regarding adequacy, and it simply shows that the
PIRT is one element, not the whole thing, that the PIRT is what we can
do working together with the industry to give us some good insights on
the importance of the phenomena, and then we put that together with
other things, including the knowledge which comes from the data, and
make the decisions.
DR. WALLIS: At least you've told me something new, which is
that it's less than minimal, it's insufficient.
DR. POWERS: It's so inside that it must be only this part
of the table that understands it.
DR. WALLIS: No. No, there's a table up there: excellent,
reasonable, minimal, and insufficient.
DR. POWERS: Okay.
MR. MEYER: Moving right along, I can mention that we have
begun to collect our list of names of experts. We have drafted work
statements, and yesterday, we talked to NEI and the people from the
robust fuel program about their possible involvement.
I think they're interested. If they go through with it,
then they would provide some of the experts, we would provide some of
the experts. It would be a different mix. I think it would be really a
very good mix because typically, we would get people from the
laboratories and the universities, and the industry would go to vendors
and consultants and utilities, and I think, with some representatives
from all those camps, that it would make for a very good group of
experts.
It will take us a couple of months to organize this effort
and get contracting taken care of, so I don't see that it could start
before late summer or early fall.
DR. APOSTOLAKIS: What is the actual use of the PIRT? This
is to --
MR. MEYER: I'm sorry, say again?
DR. APOSTOLAKIS: The actual use of the ranking, that will
tell you what kinds of experiments to do?
MR. MEYER: The colorful slide --
DR. APOSTOLAKIS: Well, the colorful slide goes directly
into the decision.
MR. MEYER: Yes.
DR. APOSTOLAKIS: I don't understand that.
DR. ELTAWILA: Well, the use of the PIRT is that going to
the -- once you identify all the important parameters that control the
accident scenario of interest, you want to design your experiment to
develop the data to provide this information, and at the same time, you
want your code, if you're going to use code, to have a model to
represent that phenomena.
DR. APOSTOLAKIS: So if we go back to the slide, then, with
the adequacy decision, there is an important box that's missing. It
says a PIRT feeds directly into the decision. Decision for what?
DR. ELTAWILA: If you look at the question that is asked
there, which I cannot see it from here --
MR. MEYER: The three questions are about the adequacy of
the fuel damage criteria that we're using as the speed limit for this
event, the adequacy of the plant systems code, which is telling us how
fast we're going, and then possibly the examination of another type of
code that may well be used in some intermediate step in the definition
of the damage criteria.
So you look at each one of those --
DR. APOSTOLAKIS: So now I get the ranking.
MR. MEYER: Okay.
DR. APOSTOLAKIS: And then I go back to the other diagram,
and then I can make a decision regarding what? I mean, when do the
experiments come into the picture?
MR. MEYER: Oh.
DR. APOSTOLAKIS: Why do I rank phenomena?
MR. MEYER: The big decisions that we're grappling with
right now are what kind of data and how much data are needed to support
the licensing activities with these --
DR. APOSTOLAKIS: And where is that in the picture?
DR. ELTAWILA: No, the experiment is on the right-hand side,
you know, the -- relevance of the data.
MR. MEYER: Yes.
DR. ELTAWILA: So you are --
DR. APOSTOLAKIS: So the PIRT actually is feeding into that.
DR. ELTAWILA: That's correct, yes.
DR. APOSTOLAKIS: Not directly to the adequacy of the
decision. I mean, should I take this literally or is it just a picture?
Because also knowledge of physics, it seems to me that's an important
input to PIRT.
[Laughter.]
DR. WALLIS: And to PRA.
DR. APOSTOLAKIS: PRA is a part of physics.
[Laughter.]
DR. APOSTOLAKIS: Well, I mean, I realize that maybe you put
this together just to talk to people, but, you know, I don't know how
literally to take it. It seems -- I don't -- you see --
DR. ELTAWILA: I think the adequacy to reach an adequate
decision about the regulatory criteria that we have --
DR. APOSTOLAKIS: I understand that.
DR. ELTAWILA: -- we need to know what are the important
phenomena, so it feeds into that. The PIRT itself feeds into the design
of the experiment and the design of -- or implementing model in the
code, but you need to know what are the important phenomena, and you
look from the other side. Does the experiment that we are -- I am
running provide me with the information related to that important
phenomena? Do I understand all the physics that's happening regarding
this phenomena? All this collectively is going to be used to reach a
decision that if we run the code or if you develop criteria from the
experimental data, which will be the case, we will reach an adequate
decision.
DR. APOSTOLAKIS: I realize that.
DR. ELTAWILA: Okay.
DR. APOSTOLAKIS: I'm not saying that you don't know what
you're going to do with this. All I'm saying is that this particular
figure does not send that message.
DR. ELTAWILA: Okay.
DR. APOSTOLAKIS: And in fact, there are diagrams that
reflect what you just said; they are called influence diagrams. You
have to make a decision of adequacy at the end, and you're deciding in
between what experiments to run because you don't have sufficient
knowledge. So maybe that would be a better way, or just drop it
completely.
DR. ELTAWILA: Okay.
DR. APOSTOLAKIS: Because I -- I mean, either we take this
literally and say, well, gee, they don't need physics to identify the --
to rank the phenomena and where are experiments -- I mean, these things
are not in parallel, all right, are not in parallel, because I have to
look later and find out that you're going to use this PIRT to identify
what you're going to do.
MR. MEYER: Okay. I knew I was going to be in trouble on
this one.
[Laughter.]
DR. APOSTOLAKIS: You're not in trouble. You're not in
trouble. It's just an outsider's view.
MR. MEYER: Because several of you have done this and I
haven't.
DR. APOSTOLAKIS: I'm trying to understand this ranking of
phenomena. What exactly are you doing here? Saying, okay, phenomena A
is more important than B? What does that mean?
DR. KRESS: You look at the information you need to make
your regulatory decision, and you look at the phenomena that feed into
that information with respect to the physical --
DR. APOSTOLAKIS: Right.
DR. KRESS: The fuel. And some of them are more important
than others. Some of them -- you list all the phenomena you can think
of; some of them you probably can drop off the list because it doesn't
enter into it. So you have to rank them some way. You may end up with
several of them that are ranked high. I mean, it's not ranked one, two,
three, four; it's ranked high, low or medium. So you end up -- if you
have to limit the number of experiments you have, you want to be sure to
look at the ones you think are the most important. So it's a relative
ranking and it's not an absolute ranking.
MR. MEYER: I can give you two examples of things we're
struggling with, and -- but I'm not going to be able to give you a flow
diagram to show how we're going to get from A to B.
But one of the examples has to do with the experiment, and
it has to do with whether the pulse width of the pulse in the reactivity
experiment is important, because we have a couple of test reactors that
are being used for this which have their own natural pulse widths, and
they're not the same as the pulse width of a power reactor. You know,
this has to do with doppler feedback and the materials layout of the
core.
So one of the big questions we have is, are these tests
which were done in test reactors with narrow pulses representative
enough of what we're trying to do with that for power reactors or are
they not.
Another example of a big issue that we're trying to deal
with is in the codes. The loading on the cladding -- obviously the
thermal expansion provides a loading on the cladding. You're dumping a
lot of heat into the pellet. It expands rapidly and it exerts a stress
on the cladding. But there's also a lot of fission gas in small bubbles
all around boundaries in the fuel, and these expand rapidly as well.
Can you model the loading on the cladding adequately with a
code that only accounts for thermal expansion, or do you have to have a
model in the to account for the fission gas expansion? There are
questions of timing.
DR. WALLIS: The fission gas is all chemically non-reactive.
MR. MEYER: It's mostly noble gases. So it's not -- there's
not an obviously answer to the question because there's a difference in
timing of these things, and so hopefully, if we go in with a group of
people who know something about this and a body of preliminary data,
which we have, we might be able to figure out which are the most
important phenomena that we just can't live without.
If we can get that out of the PIRT, then I think it's going
to be successful. How we dot the I's and cross the T's, I don't know
yet.
DR. FONTANA: Of course, a PIRT would be iterative.
MR. MEYER: Perhaps.
DR. WALLIS: You haven't given us a time-line in the plan,
but I was curious if you had a plan, said we need to do these activities
in order to finish, and the PIRT seems to have taken a lot of time,
first say let's have a PIRT and then how are we going to go about it,
and then you have to solicit and write contracts, and it sounds like six
months before you finish this one thing, which then puts you in a
position to begin to make decisions.
MR. MEYER: Yes.
DR. WALLIS: Is it worthwhile to spend that time?
MR. ELTAWILA: The answer is yes, Professor Wallis. We are
going to embark in a very expensive research program, test program, and
we don't know if all the tests or the test that's going to be run in the
Cabri reactor are going to be addressing all the important phenomena.
So, whether we need the tests or not, whether we are going
to get all the data that's going to resolve this issue or not, we need
to be sure, before we embark on that program, that we are going to get
the data, we're going to get it in a timely fashion, and there will be
no question asked in the future about pulse width or about the coolant
or about all the different variables that enter in the experiment
program.
So, it's going to help us making the decision that we really
need the tests and we cannot make any decision without this data.
DR. WALLIS: How long is the program, Farouk?
MR. ELTAWILA: We're talking about five to seven years to
get all this data from the test reactor.
MR. COLLINS: I think it's almost important -- one of the
purposes of the PIRT -- we're not going to have a big abundance of data
when we're through, and there's going to have to be extrapolation of
that data to different fuel types, different fuel designs, and the PIRT
gives you a good idea of understanding what's important in that
extrapolation.
MR. MEYER: Okay. It's 2:14. You want me to skip to
somewhere or you want me to keep on going?
CHAIRMAN POWERS: It would be useful if you could give us a
status report on your ANL program, and I think your Cabri program is
self-explanatory in the view-graphs, and then if you could move on after
a status report on the ANL program, because we never hear about what's
going on at ANL.
MR. MEYER: Okay.
The program at ANL is to look at actual high burn-up
cladding from a couple of commercial plants and to do some tests that
are directly aimed at checking the validation of the embrittlement
criteria in 10 CFR 50.46 at high burn-up, and it also has a component of
the program to do a series of mechanical properties measurements, which
are largely aimed at giving us the basic data that we need to run our
analytical codes.
DR. WALLIS: Was there a PIRT for this program?
MR. MEYER: No. But we are going to do the PIRT on it.
DR. WALLIS: Post-PIRT.
MR. MEYER: Mid-stream PIRT, because again, this is a
three-to-five-year project. These are relatively long-term projects.
DR. WALLIS: I think you have to be careful that you plan in
a way that you make the right decisions at the right time, for the right
reasons. I'm trying to figure out how PIRT fits into that.
DR. UHRIG: These go out to 50 giga-watt days per ton. Do
you expect to go higher than that?
MR. MEYER: Well, the -- oh, yes. The real fuel rods that
we want to test are at higher burn-up. They're BWR rods from Limerick
and PWR rods from H.P. Robinson. They haven't arrived.
What we in the cell are some intermediate burn-up rods from
TMI. We're using them for some limited purposes.
The oxidation apparatus have been tested out in-cell, works,
we're ready to go with the oxidation measurements.
The mechanical properties test really could be done at any
time that's convenient. We're essentially ready for that.
The LOCA criteria tests are much more complicated, and there
is just design and construction work going on on that equipment right
now.
The big question is when are the rods going to arrive?
We've had some delays in this. It looks now rather certain that the
Limerick rods will be shipped this summer and will arrive at Argonne
around September.
DR. WALLIS: I'm sorry. I'm going to go back to the
scheduling. You're spending seven years on a program which is
presumably eventually going to benefit industry. So, they're going to
save so much because they can use higher burn-up fuel 10 years from now,
maybe.
MR. MEYER: That's partly true. The deal is that -- and
this was the policy decision of last summer -- that the NRC has already
approved burn-ups to 62 giga-watt days per ton and that we did that
without having a full database, and so, now, we, the NRC, are going to
do the confirmatory work necessary up to that point.
The industry wants to go beyond 62, the industry is going to
have to pony up the work.
DR. WALLIS: But what you do and how fast you do it and how
much you spend on it depends on the return you expect future -- I would
think that would have a big influence on what you do and how fast you do
it.
MR. MEYER: The need is to make sure that our licensing
criteria are valid.
DR. WALLIS: And that has no price-tag.
MR. MEYER: I wouldn't put a price-tag on that.
DR. WALLIS: That's my impression, but eventually we may
come to a world where sort of accounting becomes more important. You
actually get a payoff from a certain amount of research.
DR. SEALE: Well, they've told you that, if you build it,
they'll come.
MR. MEYER: Field of Dreams.
DR. SEALE: Right. I mean they've said that, if you have
the criteria that allows you to go to higher burn-ups, that there's a
lot of interest by a lot of people in utilizing those criteria.
MR. MEYER: But what NRR has said is that the industry is
going to have to propose the criteria beyond 62. But what we're doing
here is NRC in its research program is working -- up to the 62 limit is
our responsibility, and we're looking for opportunities to work with the
industry to higher burn-ups, because if we raise the burn-up, then we
can satisfy our own needs and perhaps the industry needs at the same
time.
The current plans at Argonne have fuel only up to the 60-,
65-giga-watt-day-per-ton range. So, these would not support an
extension, but they do build an experience base and a facility where
additional testing could be done rather easily.
DR. WALLIS: It would be very good politically, if no other
way, to be able to show that, by spending 50 million for five years, you
can potentially save a billion over the next 10 or something.
CHAIRMAN POWERS: Especially if you could spend 50 million
over the next five years.
MR. ELTAWILA: Professor Wallis, did you say 50 million? I
wish, you know, but I don't think this is the case.
We are spending a very moderate amount of money on this
program, but in addition to the benefit that they gain -- and I hope NEI
can correct me with that -- that it provide the industry with operation
flexibility, operating flexibility.
For example, if they have a bundle that have some leaker,
and instead of waiting for a fresh bundle to come in, they can go to the
spent fuel pool and get one that -- high burn-up and put it and continue
operation.
So, it provide them with additional benefit in addition to
the increased burn-up, I think, over the long run there will be some
benefit from keeping the fuel longer.
There are also the issue of spent fuel, that you will have
less fuel to store in the storage tank, and so, you cannot add a value
on all these things, so -- but in general, it's reduction in burden for
industry, and as Ralph indicated, we have a safety criteria, we want to
be sure that what we granted the industry in terms of burn-up limit meet
our existing criteria right now.
DR. WALLIS: But maybe we should ask industry, then, because
it may well help you, if industry gets this perspective and can see how
much they would benefit from your research, they might be willing to
come up with 10 times as much.
MR. MEYER: Terry, do you want to answer that?
MR. RICK: Good afternoon. I am Terry Rick. I am the
Chairman of the Robust Fuel Committee that is looking into response to
transients is our title but we are also looking at the higher burnup.
I think you have touched upon some of our concerns that we
have. That is, if the industry is going to go forward with helping
sponsor some research we feel that there must be some benefits -- that
is, you have touched upon the benefits that we are looking for and we
would like to get to higher burnup so that we can save on the economics.
We do look at that and see that there's maybe on the order of hundreds
of millions of savings down the road.
But one of our biggest concerns is if it is five or seven
out into the future a lot of those savings go away with the present
value of money as it is --
DR. WALLIS: -- depends on whether or not you spend six
months on a PIRT. If you want an answer this year do something these.
MR. RICK: That's right. That's right, and that is where
the biggest concern has been at the industry on going forward with some
of these researches, and I think it is in one of the slides I think that
we have skipped over here, but there must be representative fuel,
representative experiments. It must be done timely so that we can
benefit in a timely manner.
DR. WALLIS: Well, you do have to pay for it. If you want
it so you can benefit, then you have to understand what your benefits
are and what the payoff is.
MR. BONACA: Yes. I also would like to say one thing about
it, because I've the problem with thought process that comes behind all
this.
We have got a problem evidently with running the fuel to
where it is licensed today, which is up to 62 gigawatt per day per
metric ton. We have problems both in operations by excessive oxidation.
We have examples of that and in expected response to oxidants, and
now -- so typically, you know, a logical regulatory response would be
fix the problem first, run the fuel to the limits that they are licensed
and that will give you sufficient confidence now to move onto the next
step, which may be 75, you know, gigawatts right now.
What I see in the industry proposal is throw away -- throw
everything together and try to design to 75 gigawatt a day. Now the
concern I have with that is that it is now an evolutionary process in my
perspective, okay? There is no experience being gained at a higher
exposure level, like for example -- I mean burnup level like 62 before
you move on to 75, so, you know, somewhere, maybe in the regulatory
response, we have to see that interval.
I mean we have never seen this happening in fuel. Fuel has
been a very evolutionary area where we have done a little step at a time
and then we have used past experience to project future gains and here I
don't see it.
I see -- you know, I was looking at the presentations from
the industry. It's good but it throws everything in with an objective,
which is 75 gigawatt/day per metric ton, and that is -- I want to know
that we can run at 62 first.
MR. RICK: Yes, I agree with a lot of what you are saying.
In fact, I see the presentation you have is back from our March -- I am
sorry, our January 13th presentation.
I think right upfront in there we said some of the initial
concerns, the concerns we have with existing fuel must be solved first,
and that is where our priority has been set in the robust fuel.
We have an -- oh, boy, what do we have? -- like a 50 million
dollar program in robust fuel and the first brunt of that program is to
solve existing problems -- that is, taking some fuel to hot cells,
examining some of those problems, coming up with better methods of
analysis so that we might change our operational methods, and that is
our top priority. Solve that. And we are only getting into the 75,000
because we know that it is a long lead-time and it requires some
research, so we are jumping into that now too.
MR. BONACA: Yes, but a few years time you won't do it in my
book. I mean to me it seems because if you look at some of the general
fuels being used there, there have been hundreds of, thousands of
assemblies and operations, and then only a few of those have shown real
problems, and because of the few problems we have got to understand that
there were fundamental issues with oxidation.
So you have to have a significant experience to identify the
problems and you know, the LTAs by themself I don't think are going to
resolve all the issues.
MR. RICK: And I think part of the PIRT process, why we are
agreeing that we think we should get into that is to try to identify
what are the needs -- you know, what type of experiments are, you know,
representative of our fuel and in fact we don't want to just test
existing fuel if it is going to go forward, because we are not going to
use existing fuel to 75,000. We are going to have to use Zirlo and
other advanced alloys and so test that fuel to the higher burnups. That
is what we are looking for, so it has got to be a combination of
existing fuel to make sure you can survive to 62,000 but also the
advanced alloys to get you to higher burnups, so we are working very
closely with the NRC to come up with what are the effects, what do we
need to look at down the road.
MR. BONACA: And I don't have a problem with the industry
drawing both things together, but I think the regulatory response should
be different in the sense that it should make a clear separation between
what needs to be done to get here, and I think you said
something similar but I haven't seen it in the presentation.
What do you have to do to license and support what is
already licensable today by versus what else do you need to do in an
evolutionary mode to go beyond that?
MR. MEYER: Yes. We haven't explicitly said it and maybe
Tim can hopefully say that what I am going to say now is correct, but we
probably should complete the confirmation at 62 before we give approval
to go to 75.
MR. BONACA: That's correct.
DR. WALLIS: I guess I am trying to get out of the
regulatory world into the business world to figure out what is the
overall strategy and where do you want to be and how much is it worth to
be at various places at various times? How do you need to get there in
order to be most effective in the payoff for the industry and the
consumer?
I don't think you can get away from that. You have got to
face that.
MR. MEYER: Yes. I want to interject here, because, you
know, listening to the underlying thread in your comments and I think
the missing piece of information on the table is that a couple of these
test programs that we see as essential for moving forward are running on
very slow schedules and you can't hurry them up.
I don't think we have told you what those schedules are.
DR. WALLIS: Is this because you need a certain number of
neutrons over a period of time -- is it a physical barrier or a human
barrier?
MR. MEYER: It is the ability to -- the Cabri program. They
have been running at an old sodium loop. They are going to shut down.
They are going to construct a water loop, make some modifications to the
core, relicense the plant, restart a new experimental program in the
water loop.
Testing in the new loop can't begin until 2002, just because
of the work that has to be done between now and then, but there is no
way you can speed that up and this is a problem for our industry because
the results then come a few years later and it pushes a number of
utilities back close to the time of eventual shutdown and lost
opportunity.
It's tough scheduling problems. The program at Argonne is
moving very rapidly. We put it in place very quickly. It has taken us,
you know, a year, year and a half to get fuel rods to test and you miss
a window -- we had fuel at Limerick. We had fuel in shipping casks.
DR. WALLIS: It takes a year to get a fuel rod? You can't
just buy a fuel rod?
MR. MEYER: Last June -- these have to come out of a power
reactor. They have to come out at a convenient time when you don't
interfere with an outage.
We had fuel rods in a shipping cask last June at Limerick
and there was a problem. They were taken out and we will not be able to
get back in there and ship it until this coming June -- one full year.
DR. FONTANA: Now the NRC has approved 62 megawatts per
kilogram. The H.P. Robinson fuel says here 68 to 73. Have they
actually run to that?
MR. MEYER: Yes.
DR. FONTANA: Weren't they breaking some rules or something?
MR. COLLINS: I think they are for lead test assembly.
MR. MEYER: Lead test assemblies. They typically approve
lead test assemblies, a few assemblies that can exceed the burnup
limits.
DR. FONTANA: Thank you.
MR. MEYER: In my opinion this is the real value of the lead
test assembly. You can put one or two of them in there and really push
them and then take them out and let them feed your test programs.
MR. RICK: Let me add something else, and that is we keep
talking about burnup and burnup as being the high criteria. We had a
large debate on it. Is it burnup or is it something else? Is it, for
example, oxide thickness -- and if we don't use the existing materials
for the high burnup but we use another material, but that material is
oxide thickness and other characteristics are the same at 75,000 as we
have tested at the older fuel that might only be at 60,000, we would
like to do those effects type tests and say we can go to 75,000 with
this new material and it can act like 60,000 at the older material, so
that is the type of test we want to do to also say that we can go to
higher burnup.
DR. SEALE: There is an advantage of the PIRT process too I
think we need to recognize, and that is that unless you do something
like that, you are very likely to be trapped into thinking about things
the way you have thought about them before, and so while you are not
necessarily thinking about the unthinkable when you do the PIRT
assessment, you are thinking about the as yet unthought about, if you
know what I mean.
You are looking at processes that aren't in the physics
picture right now, but appear to be appropriate under the next increment
or the next -- or may be appropriate. And I don't think you're going
to -- you know, it's as good a way as any to bring your analysts and
your planners back up to speed. I mean here the NRC dismantled
virtually the fuel program. How do you get that expertise back? One of
the ways you do it is you look at the problem and you look at the
problem in an innovative, thoughtful way. And this is about as
thoughtful a way that I've heard of that anybody's tried to look at this
sort of thing.
CHAIRMAN POWERS: Certainly I think I agree with you 100
percent, that it's the only way I know how to go about doing it, and
given that your testing program by the very nature of it is always going
to be limited. And it has the additional advantage that a guy with a
reasonable amount of thought can think of lots and lots of physics that
might be important, but in this particular context of these particular
scenarios that Dr. Meyer laid out one of them for us, but I think he's
got several of them up his sleeve.
DR. SEALE: Sure.
CHAIRMAN POWERS: You can say ah -- you know, expert people
who are smarter than I am on these subjects can say ah, yes, this is a
phenomenon that's operative, but it's not important here. And you don't
need to investigate it. But you do need to investigate the combination
of these two otherwise individually unimportant phenomena but because
they hook together, there could be a problem here.
DR. SEALE: Well, and it's -- and the other thing is that it
provides a mechanism for organizing the input from the intellectual
community in this discipline rather than just going around and
collecting a bunch of "sea stories" about what's happened. You put it
together in a logical framework, and I think that's very -- very
important.
MR. MEYER: Do you want to go on to the source term or do
you want to quit?
CHAIRMAN POWERS: I think we have to touch upon --
DR. KRESS: I would request he at least cover slide 19.
MR. MEYER: Okay. High-burnup effect on source term we
think is --
DR. KRESS: Let me tell you why I would like for you to
cover that. Maybe that will help.
Number 1, I'm extremely glad to see your last bullet on
there, because I think all the other bullets are huge leaps of faith
that are unfounded, to tell you the truth.
MR. MEYER: Okay. Well now that the comments have undercut
--
[Laughter.]
MR. MEYER: All of my thinking on this, I'll tell you what
my thinking is.
Well, we are of the opinion at the moment, subject to
further information which we hope to get, that the source term is
probably affected mostly in its gap activity component at high burnup.
DR. KRESS: Let me ask you about that.
MR. MEYER: Yes.
DR. KRESS: Do you consider that gap inventory to be a
risk-significant element of the source term?
MR. MEYER: In the following slides I was going to show you
what the gap inventory affected, and I was going to say it doesn't have
much of an effect on LOCA.
DR. KRESS: Right.
MR. MEYER: And the LOCA is the risk-significant event.
DR. KRESS: Right.
MR. MEYER: So probably not important.
DR. KRESS: Yes. So why even worry about the gap? Let's
think about the other things.
MR. MEYER: I was asked to talk about the gap.
DR. KRESS: Oh, okay. I'm sorry.
Well, let me ask you, you know, I know you're going to cover
your slide, but let me ask you another one. When you talk about the
effect on the source term, did you attempt to run FASTGRASS at the high
burnups to see what you'd get out of that, because it does have a burnup
correlation in it, even though it doesn't, you know, it doesn't have a
data base to verify it at those burnups, it does have a correlation.
MR. MEYER: What we did was we ran FRAPCON-3.
DR. KRESS: Yes, but that's not a fission-product release
code.
MR. MEYER: It's got the -- what do you mean? It's got the
fission -- the gaseous fission products in it.
DR. KRESS: Yes, but it doesn't have the concepts that
involve the effects of burnup on fission-product release that FASTGRASS
has.
MR. MEYER: But it does have empirically because the
model --
DR. KRESS: Yes, but you see, the trouble is there's no
empirical data base that goes up that high, so you can't really --
that's the same problem I have with CORSOR-M. It's an empirical code,
and you can't rely on it to extrapolate.
MR. MEYER: I want to get the data slide for FRAPCON. We do
have data for the fission gas modeling in FRAPCON, and there are two
kinds of data. There are data from steady-state tests and data from
tests with power bumps in them, and there is some variability, so, you
know, it's not -- he's going to show you when he gets it -- it's not a
nice, clean fit, but it's a good fit. It's been peer-reviewed. It's in
the code. And we used it to have a look at -- this thing is -- we've
got data up around 65 gigawatt days per ton and one higher one --
DR. KRESS: What you're talking about there is the fission
gas pressure inside unfailed fuel. Right?
MR. MEYER: Yes.
DR. KRESS: That's been operated at relatively normal
operating temperatures.
MR. MEYER: Right. Right.
DR. KRESS: Now when I'm talking a source term, I'm not in
that regime at all. I'm --
MR. MEYER: Your gap activity is.
DR. KRESS: Well, the gap is, but, you see, I've already
discounted the gap.
MR. MEYER: So forget the gap.
DR. KRESS: Yes. See, so FRAPCON doesn't tell you anything
about --
MR. MEYER: No. No, no.
DR. KRESS: About a LOCA severe accident where you go up to
fuel melt --
MR. MEYER: Well, neither does FASTGRASS.
DR. KRESS: Well, FASTGRASS will take you up almost to fuel
melting. It has models in it. And it's worth looking at to see what
effect it has on the source term. Although I guarantee those models
haven't been validating data, but they have physics that talk about the
effects of burnup. And it would be well worth just making a calculation
with FASTGRASS to see what effect it has on the source term to see if
1465 may or may not be --
MR. MEYER: I don't understand.
DR. KRESS: I mean, that would be an easy thing to do.
MR. MEYER: Okay. Charlie Tinkler wants to make a comment.
MR. TINKLER: The question about the FASTGRASS models.
While it's true the fuel-release models in FASTGRASS would allow you to
do this kind of calculation, during -- one of the issues that we would
consider before that is when we went through the peer review of the
Victoria code, Victoria at one time had the capability to use those
kinds of models, and we were dissuaded by the peer review committee from
using those models because of their lack of validation. And we also
did -- we, you know, we have looked at the data and primarily come to
the conclusion that the temperature effect during core degradation is
the dominant effect above and beyond the burnup. And that's the
underlying basis for most of the conclusion aside from the fact that the
1465 is a prescription for the boundary conditions of a global core fuel
release, okay? And your ability to account for the effects of burnup
within the general prescription of releases from a damaged core in
vessel could easily be overwhelmed by things considering -- issues
related to the basic core configuration that was assumed for the 1465
releases.
DR. KRESS: The meltdown process.
MR. TINKLER: Yes.
DR. KRESS: Yes. I agree. But that's information you
needed. The release -- if you look at separate effects tests outside of
the core, which -- you do have a kind of picture but you don't have the
mobile effects that you get --
MR. TINKLER: Right.
DR. KRESS: Well, look at those. There's a definite effect
of burnup on fission-product release. It's clear in the data. It's
increasing almost Arrheniusly with burnup. FASTGRASS has models in it
and it has a trend with burnup. And so these data that we have are very
limited. They go up to maybe 50 gigawatt days, somewhere around there.
It could --
MR. MEYER: Could you indicate what's underlying these
effects that you're talking about?
DR. KRESS: It's the effect -- if you look at FASTGRASS, it
has all these transport mechanisms for fission products.
MR. MEYER: Yes.
DR. KRESS: That have to do with linking of pores.
MR. MEYER: Yes. Yes, yes, yes, yes.
DR. KRESS: You know, the infusion of gases through the
grains, through the grain boundary.
MR. MEYER: Right.
DR. KRESS: Those things are affected by burnup.
MR. MEYER: Sure.
DR. KRESS: And the size of the grains and the size of the
pores and the distances of travel, and those things are built into the
FASTGRASS. And they're built in in such a way that you can link the
power history and the burnup through fission product release by the
calculations.
MR. MEYER: I'm rather skeptical of that, and this is the
point that I wanted to make with the reference to CORSOR-M. No thought
that CORSOR-M has the ability to do these calculations. It's the
underlying concept that went into the modeling in the first place, which
was based on observations. And the observations were that the high
volatile species seem to be rate-limited by their ability to get from
inside the UO2 to a surface. But the low-volatility materials seem to
be rate-limited by their vapor pressures. You looked at the measured
release rates and they lined up in the order of vapor pressures.
DR. KRESS: But I'm only --
MR. MEYER: This says to me --
DR. KRESS: I'm only interested in --
MR. MEYER: That the dependence on the structure, which is
the burnup effect, is only going to affect the low volatiles, not the
high volatiles. And it's the high volatiles that dominate your source
term --
DR. KRESS: That the empirical --
MR. MEYER: The lows.
DR. KRESS: That the experimental evidence refutes that. If
you look at all the test data for release of fission products that are
talking about iodine and cesium --
MR. MEYER: Yes.
DR. KRESS: The things that are important.
MR. MEYER: Yes.
DR. KRESS: And even some of the noble gases. If you look
at the effects of burnup in the data that exist, it's an Arrhenius
curve.
MR. MEYER: Okay. The --
DR. KRESS: And that's --
MR. MEYER: Sure, the iodines and the cesiums are high
volatile.
DR. KRESS: Well, those are the ones we're worried about.
The iodines and the cesiums.
MR. MEYER: Well, you've got essentially --
CHAIRMAN POWERS: If I think about just an ordinary booth
diffusion model, which is a gross simplification of GRASS-type modeling,
but it includes the effect of grain size and it can be made to include
the effect of interconnected porosity, I see a very clear burnup effect.
That burnup effect appears in the pre-exponentials on all of
these things. It doesn't affect the exponential term. So, I guess what
-- I think your statements may be entirely self-consistent here, and
that what you are saying in connection with CORSOR-M is things depended
on the vapor pressure as you looked at them with temperature, which is
what you do with CORSOR-M, because that exponential term is dominant at
any given burnup. But the pre-exponential factor changes with burnup,
and I think what you will find is the burnup effect will affect all the
fission product releases, and it affects in the pre-exponential, not in
the temperature effect.
DR. KRESS: Well, I think it will also, though, affect the
time and temperature.
DR. POWERS: Yes.
DR. KRESS: That will be a synergistic --
DR. POWERS: Yeah, well, what happens is at any given
temperature, if you fix the temperature and you fix the burnup, then the
pre-exponential carries equal weight with the exponential term now, and
so if one goes up, everything goes up.
DR. KRESS: But the question is how much does it? Is it
worth thinking about a new 1465 that is a function of burnup? Will it
cause you, in the first place, to violate your 10 CFR 100 criteria, but
more than that, is it risk significant?
Mr. MEYER: Well, look, in the underlying calculations for
1465, the Nobles and the iodines, and I don't remember cesium, but the
initial release from the high temperature fuel was essentially 100
percent. And then you had deposition and retention mechanisms that
knocked down the aerosol ones. So that is why I thought that, well, the
burnup isn't going to have any effect because you have already assumed
100 percent release from the fuel for these volatile species, or
essentially. The Nobles is 100 percent. And the iodines, I think the
release rates were about the same, and they got knocked down because of
all the phosphoresis.
DR. POWERS: It is pretty clear 1465 envisaged over the
course of gap and in-vessel release, essentially quantitative release of
the Nobles. I don't think it has quantitative release of either cesium,
iodine, and if it did, it would be in error because --
DR. KRESS: It focused on the low pressure accidents, which
ended up hardly any retention, and it was, the source term was source
term into the containment. It was about 50 percent of the iodines.
DR. POWERS: Well, see, there is a late release into the
containment that is -- its origin is really two things, it is a
revaporization of those deposited materials, and it is the continued
degradation of the residual fuel.
DR. KRESS: Yeah. Those didn't enter into a big part of the
low pressure accident, they were part of the other accident.
DR. POWERS: Well, that amounts to 30 percent of the
inventory.
DR. KRESS: But it certainly wasn't release of all those
elements. It was only about half that.
MR. TINKLER: Let me -- the ex-vessel and late in-vessel
releases for iodine are quite substantial in NUREG-1465.
DR. POWERS: Yeah?
MR. TINKLER: Yes.
DR. POWERS: They are very significant because they get
right to the containment.
MR. TINKLER: But, remember, the implementation of the
revised source term is limited to the gap in the early in-vessel. And
with regard to the pre-exponential term and the temperature-dependent
exponential term, I think it is also important to recognize that some of
the separate effects tests of limited duration show a little different
effect than what you might expect in a core that is hot for any
substantial period of time, when the temperature dependence would be
quite substantial, and may be more important than other issues.
DR. KRESS: I agree, there is --
MR. TINKLER: Okay. Because some of those --
DR. KRESS: There is -- rate effect in that data.
MR. TINKLER: Some of those separate effects tests are of
short duration, and it would show up larger than in a core that has been
hot for an hour.
DR. KRESS: I agree with that.
MR. TINKLER: And we think that -- you know, when we are
looking at the global characterization of source term from a core damage
event, where we have the release over roughly two hours, I guess we just
don't see that -- I guess we see the temperature effect as being the
dominant effect.
But, now, Ralph's last bullet there, that we -- you know,
there's surely merit in confirming these behavior. And I would even
throw in, if we can talk them into them, putting some high burnup fuel
in the PHEBUS. Of course, there are many more candidates for tests than
there are tests left, but, you know, we would suggest that that is an
excellent opportunity to look at integral behavior.
DR. WALLIS: Will this final bullet resolve all this
hypothesizing and rationalizing and debate? I mean there is nothing
like data.
DR. POWERS: Well, I mean it seems to me that --
Mr. MEYER: We are hesitating because it is a bench
experiment and --
DR. POWERS: Well, the concern I think you might have over a
VERCORS test is that VERCORS, in the past anyway, have been very small.
DR. KRESS: They are very small. They are almost a carbon
copy of the Oak Ridge tests.
DR. POWERS: Well, I think they are smaller than that.
DR. KRESS: About the same.
DR. POWERS: And you will surely not get anything from gap
out of this.
DR. KRESS: No. Although the gap is going to open by the
time you get into the test. But is Jeff Rest still at Argonne and does
Fast Crash still exist?
Mr. MEYER: Yeah.
DR. KRESS: I mean that is a pretty cheap code to run, and
you could at least explore some of these speculations.
Mr. MEYER: Maybe one of the first things we need to do is
just to go back and look at the 1150 calculations and see, you know, get
beneath --
DR. KRESS: I wouldn't do that because 1150 doesn't have any
concept of the effect of burnup in it, I don't think. I think you are
just barking up the wrong tree if you do that. I would ask you to go to
Fast Crash and see what it tells you. You know, you can't believe it,
because it doesn't have a database, but it is the best we have.
MR. TINKLER: Well, the last comment I made about Fast Crash
is we tried to convince Professor Olander about this, and we had a very
difficult time. And as a last point, at the end of the day, if someone
comes in with a high burnup core, and they need to redo their dose
calculation, do we believe they should redo their dose calculation with
a TID source term or with a revised source term? And in our view, there
is no doubt, they should be doing it with a revised source term, because
all these issues apply to the core. So what characterization do you use
for the source term for your accident analysis? So, you know, I realize
that is not --
DR. KRESS: Yeah. My point is --
MR. TINKLER: That is not the technical issue, but that is
part of it.
DR. KRESS: My point is there should be a revised revised
source term for high burnup fuel.
DR. POWERS: I think this is obviously a fertile ground for
subcommittee deliberations and, Ralph, as you knew well coming in, you
are running one of the most exciting programs around and it gets this
committee's blood going and whatnot, and we have more questions than you
possibly have time to answer. I retain my enthusiasm for your pert
approach and hope you can keep us informed as you progress through the
planning.
I certainly hope that you are successful in getting
cooperation in this venture from your licensees and the industry in
general. I would hope that we can also get cooperation from the members
of our quadra-partite group in assisting you in carrying this out, and
that you will keep us informed, and we will give you a forum with a
little more time.
Mr. MEYER: Well, we would certainly welcome any involvement
that you would want in that regard.
DR. WALLIS: Could I also suggest, Mr. Chairman, that
sometime when you have sorted things out a bit more, we see a plan that
says that we are going to achieve certain things by certain dates, and
here is the regulatory payoff or something from what we are going to
achieve by that date. It will enable us to make certain decisions we
can't make today. And there is going to be some kind of benefit or
whatever, see the --
DR. POWERS: I am not sure, I guess I have to just
interject. The regulatory decision has been made.
DR. WALLIS: Well, why aren't we doing any work?
DR. POWERS: Because we would like to know that that
regulatory decision was, in fact, the right decision to make. The
regulatory decision is that these burnups of the 62 gigawatt days do not
pose any significant threat to the public health and safety.
DR. WALLIS: We are just making a reality check on the
regulatory decision, is that what we are doing?
DR. POWERS: What we are trying to do is -- the regulatory
decision was made on the basis of an engineering judgment, and I
guarantee you the people that made that judgment spent a lot of time
pacing the floor and saying, do I really do it, in the face of some
experimental data that was fairly inflammatory, and fairly diverse
technical opinions. That is what was communicated to us in previous
meetings on this subject. So if anybody would like to correct me on my
perceptions of the decision making process.
Mr. MEYER: Well, I would only comment on the timing and
that is that the regulatory decision to go to 62,000 megawatt days per
ton was made prior to the performance of the REP NA-1 test in Cabri,
which was the first test to show that there might be a significant
problem with one of our Regulatory Guides.
DR. WALLIS: I guess what I am just asking for is a laid-out
plan which says what you are going to achieve and what will be the
effect, whatever it is, that is measurable in some way, or
understandable in some way. And if it is not quite what I thought it
was, then maybe you can explain what it is, when you get there. I don't
think we are going to get it today. I think -- I am sure good work is
going to be done. I would just like to know about why this is the
optimum way to do it and where are the rewards.
Mr. MEYER: I am looking over for my boss to see --
DR. ELTAWILA: Professor Wallis, have you see the agency
plan on high burnup? Maybe I will make a copy of it.
DR. WALLIS: I have seen it. I may not have realized that I
had to pay as much attention to it at the time.
DR. ELTAWILA: Well, I will send it to you and I hope you
pay attention to it. And if you need any information, that will give it
to you.
[Laughter.]
DR. POWERS: I think we can probably get it for you.
DR. WALLIS: You hope or you fear?
[Laughter.]
DR. POWERS: Well, thank you very much, and, as I say, this
is certainly an exciting program, and I hope you get your fuel
shipments.
DR. ELTAWILA: Thank you.
DR. POWERS: I am going to recess until a quarter after.
[Whereupon, at 2:57 p.m., the meeting was recessed, to
reconvene at 8:30 a.m., Friday, March 12, 1999.]
Page Last Reviewed/Updated Tuesday, July 12, 2016