Protecting People and the EnvironmentUNITED STATES NUCLEAR REGULATORY COMMISSION
UNITED STATES OF AMERICA
NUCLEAR REGULATORY COMMISSION
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
***
MEETING: PLANT OPERATIONS AND FIRE PROTECTION
U.S.N.R.C., Region III
801 Warrenville Road
Lisle, IL
Wednesday, June 14, 2000
The committee met, pursuant to notice, at 8:30
a.m.
MEMBERS PRESENT:
DANA A. POWERS, Chairman
GEORGE APOSTOLAKIS, Vice-Chairman
JOHN J. BARTON
MARIO V. BONACA
ROBERT L. SEALE
JOHN D. SIEBER
GRAHAM B. WALLIS. P R O C E E D I N G S
[8:30 a.m.]
CHAIRMAN BARTON: Good morning. The meeting will
now come to order. This is a meeting of the ACRS
Subcommittees on Plant Operations and Fire Protection.
I am John Barton, Chairman of the Subcommittee on
Plant Operations, and Jack Sieber is Chairman of the Fire
Protection Subcommittee.
ACRS members in attendance are George Apostolakis,
Dana Powers, Mario Bonaca, Robert Seale, Robert Uhrig, Jack
Sieber, and Graham Wallis.
The purpose of this meeting is to discuss
selective technical components of the plant operations and
fire protection issues. The subcommittee will gather
information, analyze relevant issues and facts, and formally
proposed positions and actions, as appropriate, for
deliberation by the full committee.
Jit Singh is the Cognizant ACRS Staff Engineer for
this meeting.
The rules for participation in today's meeting
have been announced as part of the notice of this meeting
previously published in the Federal Register on May 24,
2000.
A transcript of the meeting is being kept and will
be made available as stated in the Federal Register Notice.
It is requested that speakers first identify
themselves and speak with sufficient clarity and volume so
they can be readily heard.
We have received no written comments from members
of the public.
We will now proceed with the meeting, and I call
upon Mr. Jim Dyer to begin.
MR. DYER: Thank you, Mr. Barton. Good morning.
Welcome to Region III. I'm Jim Dyer, I'm the Regional
Administrator for the Regional Office. With me here today
are Mr. Marc Dapas, who is the Deputy Director of the
Division of Reactor Projects; Mr. Jack Grobe, who is the
Division Director, Division of Reactor Safety; and, Mr. Jim
Caldwell, who is Deputy Regional Administrator.
Also, throughout the day, we've scheduled an
agenda, which copies are available for the public over by
the coffee pot and as you come into the conference room, and
the various members of the staff will be addressing the
subcommittees today, based on the information we understand
that you request, and if you want additional information,
we're very flexible. We'll try to get anybody who is here
on the staff to answer your questions or present anything in
particular you wish to address.
I think, going to my first slide, a little
background about Region III. This was the recent addition
to the package, so we'll have copies made. But just Region
III encompasses an eight-state area involving, on the
reactor side, involves 16 operating sites, 24 operating
reactors, and those are the people sitting at the table
right now, particularly Mr. Grobe and Mr. Dapas, have the
principal responsibilities for safety oversight in those
areas.
We also have a Division of Nuclear Materials
Safety and Division of Resource Management and Assessment.
Our reactors are relatively close to each other in
the eight-state region and particularly in the State of
Illinois, and it makes convenient travel from the Region III
offices here in Lisle.
CHAIRMAN BARTON: Something that can't be said
about travel to here.
MR. DYER: Yes. Just a little overview about the
regional organization, and I can make some introductions of
the folks, the staff.
What I really want to just focus on is the upper
half of the chart we provided you here. For our
presentation here, what we plan to do is I was going to go
through the overall regional organization and then allow the
division directors, particularly the Division of Reactor
Safety and Division of Reactor Projects, to go into their
more detailed reviews of their staffing and how we're
organized to manage our safety responsibilities here in the
region.
I guess, first of all, we are organized with four
divisions, three technical divisions; that is Division of
Resource Management, the Division of Nuclear Materials
Safety, Cindy Pederson is out today, and we didn't plan on
her participating. They do have some responsibility for the
decommissioning reactors. So if you have any questions that
go into that arena, we'll bring somebody down to discuss
that with you.
Additionally, Mr. Grobe, Division of Reactor
Safety, and in my oversight and role, the way I look at the
way the region operates is somewhat of a matrix organization
between DRS and DRP. I view the Division of Reactor Safety
as the functional experts in the various areas. So their
responsibilities are in the operations, engineering, plant
support areas, radiological protection, fire protection, and
that.
In those areas, they're responsible for looking at
specific areas across all of our 16 operating reactor sites.
So for the case of operator licensing, they're responsible
for overviewing of the operator licensing and operations
inspections, team inspections, and calibrating safety
assessment at all 16 sites across that one functional area.
Then separately from that is the Division of
Reactor Projects, which is organized by reactor assignments
to the various sites, under Mr. Dapas and Mr. Grant, and
these are organized more in lines with projects.
They're our generalist inspectors and basically
they are responsible for everything that goes on at that
site. So within the region, if a particular event occurs or
a particular issue comes up at a site, there should be two
points of contact that have cognizance over that area.
One, the DRP point of contact from a generalist
view, because it affects that site and you can integrate the
impact of the assessment across all the functions at that
site and put it in that proper context.
The second would be from the functional area
review and taking a look at, from Mr. Grove's DRS point, how
does this -- what are the lessons learned, how are we
consistent across all of our 16 sites in the way we're
treating that area.
So that's the general oversight of how the region
is orchestrated and integrated. In particular, the key
aspect of regional activities that establishes and
identifies the issues we're going to follow up is at 8:15
every morning, we conduct a review of plant events and plant
status. Normally, it's in this room. This morning it was
taking place in our other conference room down the hall on
the third floor.
But in that meeting, we will go through any
reported events for the night and any emerging issues that
come from the sites, from the resident inspectors, we bring
them up and put them on the table and discuss what is our
response going to be to those activities.
CHAIRMAN BARTON: So is this organizational
structure you described pretty similar in all four regions?
MR. DYER: It's identical in all four regions.
It's just my concept of operations, if you would, as to how
they -- other regions may decide to do things differently.
We all have morning meetings, but we all have some
differences as to how we would approach a morning event or
an emerging issues.
CHAIRMAN BARTON: Thank you.
MR. DYER: I think a little bit about the Regional
Administrator staff; in particular, this is also similar to
all the regions. Of most interest to you is probably our
area of enforcement and allegations. If there are any
questions that would come up regarding -- by the
subcommittees today. Mr. Brent Clayton is here this morning
and he is available, if you have any questions, or he is
going to spend some time this morning and we'll bring him
back or we'll get a member of his staff, if there are any
questions about the allegations or the enforcement
activities that we have going on here in the region.
Additionally, we also have an Office of
Investigations, which is similar in all regions, a Public
Affairs staff, and then a Regional State Liaison Officer.
Mr. Roland Lickus had to take his son to a doctor's
appointment this morning. He is going to come in a little
bit later.
I think what is unique to Region III is our
relationship with the Illinois Department of Nuclear Safety.
I'm convinced there is no state that has the extent of
nuclear oversight that the Illinois Department of Nuclear
Safety has with their resident inspectors at all the six
sites that are operating in Illinois and their extensive
emergency planning and incident response capabilities.
If you care to discuss our relationship or how we
interact with Illinois Department of Nuclear Safety, Roland
is probably the best person to talk to with that.
Additionally, we have a regional counsel, who is
in our -- spends a lot of his time involved with reactor
enforcement cases, and particularly, now that recently we
have had a lot of discrimination issues that have taken a
lot of our time and have been a challenge.
So that's the basic overview of our organization
here, from a regional administrator's level. I guess I
would ask if there are any questions.
DR. POWERS: I guess one thing that has just
emerged for the committee is we're anticipating getting a
power upgrade application from Guianardo, rather substantial
one. So any thoughts you have on that power upgrade that
you think we ought to know about would be useful, if there
is a chance during the day.
MR. DYER: Okay. You're going to have the senior
reactor analysts later on the day and I know they may be
more informed. I know Mike Parker was out there with
Research and did some walk-downs.
DR. POWERS: I think we would be interested in thoughts
about are there synergistic effects associated with going to
power upgrades and high burn-up fuels in an aging plant.
Things like that. It's the first of what we see of many
rather substantial power upgrades. I hesitate to quote the
exact amount, but it's about 15 percent power up rate, which
would mean they're about 20 percent of what they had in the
past.
MR. DYER: We can certainly comment on the impact
that has on the inspection program. But the technical
viability, NRR reviewers get involved.
DR. POWERS: Sure. And insights that you have
that are peculiar to you that we would be most interested
in.
MR. DYER: And Commonwealth Edison is also looking
at what I consider to be rather substantial power up rates
for both the Quad Cities and the Dresden stations.
DR. POWERS: I think there's going to be a covey
of them coming in.
DR. SIEBER: Speaking of coveys. You've had the
privilege or honor or whatever of not being involved in the
first sub-group of license renewal activities and perhaps it
would be more appropriate to address this question for
Regions I or II, but I don see any of them here.
How do you anticipate a license renewal
application would impact the regional activities?
MR. DYER: Quite frankly, I don't think it is
going to impact. We'd love to have one. Right now, we
don't have any takers. I think Commonwealth Edison is --
both Commonwealth Edison and the management company, which
formed Duane Arnold, Monticello and that, are both talking
about it, but --
DR. SIEBER: No one has committed yet.
MR. DYER: Nobody has committed and I think
they're at least six months away from doing that. I know
that we really haven't taken a look at that for license
renewal.
MR. GROBE: We can talk in a little bit more
detail when we get into the details of how my division
operates.
MR. DAPAS: I think, in summary, though, it's
probably relatively transparent to the new inspection
program.
DR. SIEBER: I had one other question. Are any of
the people out here representing the public as opposed to
members of your staff?
MR. DYER: They are all our staff.
DR. SIEBER: Okay.
MR. DYER: All of which I believe may be giving
you presentations later today.
DR. SIEBER: Okay. I was just curious.
MR. DYER: Okay. Next slide, please. Following
up, I think that a few of the activities that we've recently
completed or are in the process of completing that may
provide some areas for later discussions, of course, is some
of our more recent regional accomplishments.
We did implement the pilot program at both Quad
Cities and Prairie Island. I think in particular, it was a
unique relationship, particular with the Quad Cities sites,
in that it involved integrating the Illinois Department of
Nuclear Safety into this program.
We conducted the training here in this room, in
fact, and brought all the Illinois Department of Nuclear
Safety folks in to cross-train them. Secondly, Quad Cities
had some unique performance indicator verification issues
and it really opened up, I think for the industry and the
NRC, an understanding as to just how many different ways you
can calculate performance indicators.
And as a result of that, Commonwealth Edison
really took the lead, I believe, for the industry to
solidify and come up with a common way of doing it.
I think Oliver Kingly, at our last review, made
the comment that he says he never realized that they had
seven different ways of calculating EFTY within the
Commonwealth organization, and depending on which
organization you asked, as to how much reactor burn they've
done, they have a different way of calculating it.
So it was those kinds of things, and the same
thing with how they recorded availability. It was
interesting.
DR. POWERS: The NRC seems to have about seven
different ways of calculating availability, depending on
what rule you go to.
MR. DYER: We have transitioned to the new
oversight program at all our sites, with the exception of
D.C. Cook. I would like to add that while you were in
transit yesterday, I signed the D.C. Cook 0350 closure
letter. So D.C. Cook is -- the closeout has been done and
now they're in the process of heating up and testing their
systems in mode three and trying to wrestle with a problem
with the turbine-driven aux feedwater pump this morning.
But they will be the final plant to transition
after the restart of Unit 1, and that will be later this
year.
CHAIRMAN BARTON: When do you see them fully under
the new oversight process?
MR. DYER: Jack probably has the best -- I was
asked that at the Commission meeting, and I would say about
six months after startup.
MR. GROBE: One of the things that we have to
consider is how effective the performance indicators are
before we transition them back to the regular oversight
program. That's been shut down for almost three years. So
there is no valid performance indicator data, with the
exception of maybe in the health physics and emergency
planning areas.
So we'll be looking at the performance indicator
data and turn the plant back to the routine inspection
program as soon as we feel comfortable that the way the
program is structured, we can effectively monitor the plant
performance.
CHAIRMAN BARTON: Thank you.
MR. DYER: We completed our PPR reviews for the
end of cycle on the pilot plants and also did some mid-cycle
reviews for the other plants, just to get them going in. Of
note, as a result of the review, we, believe, are the only
region, and we have two yellow performance indicators within
the region, Kewaunee, alert notification system and siren
system is in a yellow status, and we completed the 95-002
inspection, which is the supplemental inspection at
Kewaunee.
Additionally, Quad Cities, the HPCI system went
into a yellow status because of availability on an auxiliary
oil pump, and we can discuss those. We have not done any
supplemental inspections or held the public meetings yet
with respect to the Quad Cities plants. That was just a
recent issue.
Again, implementing the revised enforcement
process, and if there are any questions, Brent Clayton is
available in that arena.
Some of the areas -- one of the areas that's been
a major shift here in the region and a major focus is -- I
don't know if you know of the RIT system, which is our cost
accounting system, which is used as the basis to budget our
resources. We have found that we have not been accurately
recording our costs and things that we thought were going in
one of the cost bins, such as follow-up inspections or plant
assessment, were, in fact, going in a completely different
bin, some of our SRA training time.
So we've wrestled with our cost accounting system
and it's clear that under the new budget constraints and
that, that we are going to have to become better managers of
our resources and understand what our budget resources are
and what the plans are that we're doing.
DR. SIEBER: Does that affect the licensee
billing?
MR. DYER: It turns out that licensee billing was
about the only thing we did right, as far as the inspection.
We were very good with inspection reports, but there's a lot
of non-direct costs. That would be plant assessments,
follow-up on technical issues, things like that that we were
getting coded to other administrative duties and things like
that.
So it sort of skewed our model and didn't capture
accurately what the costs of how the region did business,
and we've subsequently gone back and cleaned it up. So
hopefully for the rest of this fiscal year, we should.
But fee billing was it -- the inspection efforts,
the direct inspection, as well as prep and doc for the
inspection reports was pretty much -- that was done well.
DR. SIEBER: Is it fair to say that the net effect
of all of this was to tend to put more time or more pressure
on the administrative rather than the programmatic side?
MR. DYER: Yes. The real impact was on -- we
receive resources for plant assessment. By and large, those
were under-billed, those resources, and administrative was
over-billed.
DR. SIEBER: That can be embarrassing in the long
run.
MR. DYER: As you'll find out later on, we've had
-- when I first got here a year and a half ago, we had, I
believe, six plants that were receiving enhanced oversight
under 0350. Every one of the managers at this table was
overseeing either Commonwealth Edison or at least one or two
of the facilities that were preparing to restart.
And when the budgets -- and the staff was
similarly supporting all those activities. And when the
cost data came back and we were budgeted six and a half FTE
for plant assessment, and we spent two and a half, which
just didn't make sense.
So we knew something was up. Everybody was
spending all their time in 0350 panels and oversight and
when the cost data -- that's when we started looking as to
why we did it and what it was was we had some old cost codes
that we had been using for years and they were translating
to some sort of different -- so it's caused a -- it's been a
rather substantial effort.
Again, we made also a focus on improving our
communications, enhancing them, particularly to get the
implementation of the new oversight program. There's more
rumors flying around about the program, as any time you go
through a significant change.
We've held monthly meetings, enhanced meetings,
with the divisions and have done some very good training. I
think it's paying off now. I think the folks at the working
level that are actually leading the change and the
transition and they are the ones that have the best concept
of what's going on at the plants.
CHAIRMAN BARTON: I want to ask you a question.
MR. DYER: Sure.
CHAIRMAN BARTON: Regarding that. If I were a
"good plant" in this region, as defined by you folk, now
under the new oversight program, with the baseline
inspection program, would I be receiving more or less
inspection hours?
MR. DYER: Absolutely more. I have a slide. I
can diverge from that, if you want to.
CHAIRMAN BARTON: We just heard that yesterday
loud and clear as a complaint. So we wondered whether it
was true or whether we were just hearing a story.
MR. GROBE: We are going to talk about that
specific aspect in some more detail.
CHAIRMAN BARTON: Okay. Good.
DR. POWERS: The question has some things to it in
that it may be true now, but it is going to be true once
you're in a more steady-state on the inspection program.
MR. DYER: Right. We are probably the extreme
region for that concept, but --
MR. GROBE: The reason for that is that under the
old program, we had some flexibilities, and we'll get into
that in detail. We had a number of problem plants and I
don't remember the total numbers, but it was upward, over a
period of years, 20,000 inspection hours at D.C. Cook,
similar at Clinton and other sites.
So a plant like Davis-Besse, which was one of our
better performers, under the flexibility of the old program,
got significantly fewer hours.
The baseline, the risk-informed baseline is
intended to establish not a ceiling, but a floor, and that
floor is higher than what Davis-Besse got in the past.
MR. DAPAS: And we'll explain why there was that
flexibility under the old program and relative to the new
program.
DR. POWERS: I mean, I guess the question that
comes to mind is why shouldn't there be that flexibility. I
mean, if you're going to have problem plants, and you are on
occasion going to have those, why shouldn't you put your
resources where the squeaky wheel is and let the guys that
are doing a pretty good job --
MR. DYER: Well, I think it's a little more
complex than that. We're going to get into it. We have
about an hour set aside for this.
And let me just close out. Part of the issue is
that -- I'm quite pleased and, Jack, you couldn't wipe the
smile off his face, but the fact that yesterday was the
final closeout of our 0350 process and our formal restart
0350 process for D.C. Cook is -- that has been a -- that is
a significant impact on the region and that's the final one.
As I said when I got here, we were doing it with
LaSalle, Quad Cities had just started up, we had Clinton, we
had Cook, Peach Bottom, Point Beach wasn't that far away
from restart. So there was a number of -- we have literally
been focusing from plant to plant.
And last year at this time, the great fear was
that if Clinton kept delaying and LaSalle kept moving their
schedule up, it looked like both of them were going to
restart within a week of each other. They subsequently
restarted about a month apart. So that was a great relief,
because a region literally cannot handle two restarts
simultaneously of problem plants coming up.
So now we're poised to do the D.C. Cook restart
and we are getting resources from all the other regions in
order to support the final closeout of the inspections, as
well as the actual startup.
CHAIRMAN BARTON: But with Cook coming back, that
will only help the stability question in this area.
MR. DYER: I believe it actually helps more the
northeast, because it's the tie lines. When Commonwealth
Edison came back, Chicago was flush and the last time I
talked to Oliver Kingsley, it looks like they could actually
have excess power. What they want to do is it get to the
northeast, where there's a need for power, and the tie has
been right there at D.C. Cook. They have been able to route
power through that intertie out of the main grid.
So they've actually been wheeling power south and
then back up.
DR. SIEBER: Or it would go through Canada.
MR. DYER: Right.
DR. SIEBER: To what extent does headquarters hold
the region accountable when a plant -- I'll speak louder.
To what extent does headquarters hold the region responsible
or accountable if a plant emerges as a problem plant?
MR. DYER: Well, you have to take a look at how
did it occur and it's more you do a root cause analysis, if
it's caused by an event; you know, should we have found it
earlier, and done that.
I don't think it's any kind of fingerpointing or
blaming as a result of that, but it always causes you to
reflect. And I can say it's not only just the region that
has the problem plant.
When the Commonwealth Edison problems came up and
the Cook problems came up, and Millstone, even when I was in
Region IV as Deputy Regional Administrator, we were all
looking could that happen here. It's a general --
DR. SIEBER: Do you think the oversight process
will help you identify precursors to problem plant issues
more so than the old inspection program?
MR. DYER: I don't that the oversight process will
help the NRC identify it. I think the deregulation is going
to force the commercial nuclear industry to take a greater
role in fixing, and the cost, the main cost in production,
those areas, the pressures that they now feel are far more
than what the NRC used to put on them.
They have to be a much more demanding manager now of their
plants in order to accomplish the shorter outages, in order
to bless the less than one reactor trip per year, on an
average now, in the industry.
That's not NRC-driven. That's economics-driven,
in my mind. And no matter how much I, as a regulator,
challenge the licensee to improve performance, it's going to
cost them a couple hundred thousand dollars a day now when a
plant goes off-line, that's making the difference.
So I think our critical focus is shifting to make
sure that they follow the prescribed processes and that
they're playing by the rules, if you would; that when a
system is inoperable, they declare inoperable and do the
right thing, as opposed to how are they fixing it. That's
the emphasis.
MR. GROBE: The new inspection program is more
indicative than it is predictive and that's one of the
concerns that we have in how we implement this, to retain
the ability to identify the early precursors of more
significant problems.
We're going to get into that in a lot of detail
with lessons learned on the new inspection program to date.
DR. POWERS: And if you find routes to prediction
under the new inspection program, we're going to be real
interested, because it really is an indicative program.
DR. SIEBER: One more short question. With all
the emphasis on cost-cutting and economical production, do
you see things like the plant material condition going up or
down, or programs being eliminated or consolidated to the
detriment of the whole program, or other issues that are not
being attended to that otherwise, in a more generous
economic situation, might be attended to?
MR. DYER: I guess from my perspective, I've seen
an investment in the plant. The thought of looking at
extending the life cycle, the prospects of doing that and
whatever they run, they've got to run well. Those are the
key things that we've seen.
Particularly, what we saw was a total focus, I
believe, from some of our plants is when they were shut down
under the 0350 process and trying to get restart, they took
a focus away from operations and they were focused on
getting the plant fixed, whether it was reconstituting the
design basis, modifying the plant to fix a long-term
problem, or doing whatever is necessary to get their
procedures and infrastructure effective.
There had been a lack of focus on maintaining the
operating crews and maintaining the plant in an operating
status net. So now that we've seen the plants once they
start up, there has been a shift toward that operational
safety focus, an increase in number of licensed operators.
In Region III, and I think Jack probably has a
better handle on the budget numbers than I do, but I think
we were looking at typically we were running between 30 and
50 exams a year and once Cook, Clinton and ComEd got up and
running, in the past year and a half, two years, there now
-- our number of licenses are upwards of 160, demand for us
to give 160 licenses.
So it's literally tripled our workload in a short
period of time. That's put a pressure on the region to get
a lot of qualified license examiners and borrow them from
headquarters and management, which is what Jack has done,
but that kind of a ramp rate, if you would, has put a severe
strain on the regional resources for that program.
But that's what we're seeing now, is an enhanced
focus on operations and an investment in the plant. So I
think almost all the plants are --
DR. POWERS: I was just going to comment in
response to your question about material condition. I think
under the new program, when you look at unavailability,
performance indicators, if the licensees are maintaining a
material condition, you would expect to see that manifested
in transients caused by equipment problems and challenges to
the operator.
So I think the new program has carved out a role
of ensuring that material condition is being maintained or
at least flagging to us that there are problems in that
area, and then we would go in and look at the licensees'
root cause evaluation and corrective actions as part of our
supplemental inspection of a particular performance
indicator threshold, for like system unavailability.
MR. GROBE: Jim and Marc are focusing primarily on
reactor operations and those issues that directly make
money. In some of the peripheries, we've seen some
problems; for example, in the security and safeguards area.
Commonwealth Edison substantially changed their approach to
event response and protecting the plant from a physical
threat and we just recently completed what is referred to
our OSRE, operational safety response evaluation, at Quad
Cities and they performed poorly.
They changed the strategy also at LaSalle and
Braidwood, significantly reducing the number of armed
responders, for example. And we have exercises there later
this month.
DR. SIEBER: So that was an issue involving the
security organization as opposed to operation involvement in
security.
MR. GROBE: And I think Jim's point on the
financial demands is really key. Those things that can
produce power and ensure equipment reliability are getting a
lot of attention.
DR. POWERS: I think we can say the same thing in
fire protection, because it doesn't generate kilowatts, it
may be getting less attention than some of the other things,
as well.
DR. SEALE: Not very well.
DR. SIEBER: Well, this is apparent or has been
apparent for some years. I've worked with LaSalle for a
couple of years and they had a lot of fire protection work
orders that had aged substantially and I see the same thing
on division valves at other sites and people say, well, as
long as the valve is open, we're okay, but if you rupture
the main, you may put the your whole system out, because you
can't isolate.
So I think that that often needs attention,
because it somehow jumps outside the risk-significant
portions of the plant, which are the CAT-1, structures,
systems and components.
MR. DYER: One other, on the same spin, I was just
thinking, you know, in the case of Clinton, was one of the
plants that was really run on a shoestring. It was a single
unit utility. I think we have seen a significant commitment
of resources and improvement and a change over there,
particularly since Amergen took over and purchased the site,
and it was shortly thereafter that they came out with a
business plan that included looking at license renewal as
opposed to the mentality when it was Illinois Power, which
was get the plant restarted itself.
So it was do what was necessary to restart the
plant, which it did not include training new operators.
DR. WALLIS: Do you find that consolidation of
plants under single owners is helpful then, in general?
MR. DYER: We've had limited experience with that.
The Amergen is the first one under, and now the management
company is just trying to formulate and they really haven't
had an impact yet.
Commonwealth Edison, we've had a seesaw
relationship with over the years. Right now, it's riding a
wave up and it's doing better. So I'm waiting to see.
MR. CALDWELL: I think the real answer to your
question, though, is it's going to be case specific. I
don't think you can make a generic statement about how
deregulation is going to affect all the plants. The
single-unit sites, if they don't have a lot of resources, it
may have a major impact. These sites that are now being
taken over by large companies, they can't afford to have the
kind of shutdowns that we've seen in the past, the
multi-year shutdowns.
So they're going to have to focus on making sure
the plants are properly maintained. So it's going to be up
to us to look at the different facilities and the different
situations they're in and to try and understand it. But I
don't think you can make a statement across the board that
it's going to have the same impact.
MR. DAPAS: That's one of the things the agency is
looking at is industry consolidation and there is a working
group that I'm involved in to understand what changes may be
necessary in certain program areas as a result of industry
consolidation.
DR. SIEBER: I don't want to ask too many
questions and get you off schedule.
MR. DYER: I think I've blown my schedule.
DR. SIEBER: I'm sorry.
DR. POWERS: We have a tradition of doing that.
MR. DYER: Yes. But what I was going to do is now
turn the meeting over to Jack Grobe and Mark Dapas and let
them got into more of the details of how the DRS and DRP
organization goes, consistent with our program.
MR. GROBE: We had some donuts delivered and, Dr.
Barton, do you want to just take three minutes?
CHAIRMAN BARTON: No, we're behind schedule. If
you want a donut, get up and help yourself.
MR. GROBE: Excellent. We've laid out an agenda
that I think that I think, we had coordinated with Jit, that
hopefully meets your needs. We've got about 65 slides to go
through, which our ability to do that is probably limited,
but our goal is to make sure that we answer all your
questions.
So I'm going to try to be a little bit of a
gatekeeper on the clock and move us along as we go.
But the first thing we're going to do is talk a
little bit about how we're structured, how we're
implementing the new program and some lessons learned on the
new program, and then invite Sonia Burgess up to talk about
our senior reactor analyst program. She's one of my SRAs.
MR. DAPAS: I thought I'd start out with kind of a
broad overview of our geographic responsibilities. You can
use the slides of you can go through the handouts we
provided, whichever is easiest for you.
But we are responsible for 24 operating reactors
at 16 sites, and that consists of 13 pressurized water
reactors and 11 boiling water reactors. As Jim said, our
responsibility encompasses an eight-state area. We've got
six sites in Illinois, two sites in Wisconsin, three sites
in Michigan, two sites in Minnesota, two sites in Ohio, and
one site in Iowa.
And as Mr. Dyer mentioned, it's relatively easy to
travel to any site. We can get to Prairie Island and
Monticello, which is near Minneapolis-St. Paul, Twin Cities
area, in a day; same thing with Duane Arnold, near Cedar
Rapids, Iowa. So that doesn't present the challenge that it
does to some of the other regions in terms of being able to
get to the sites.
The Division of Reactor Projects, or DRP, has
roughly 75 professional and administrative staff. Most of
the inspection staff in DRP has an engineering background or
a technical science degree. So we have a fairly
professional staff.
And we've organized the branches to provide
additional oversight to D.C. Cook; D.C. Cook, of course,
being an agency-focused plant. We've got one branch
dedicated to Cook, which results in the other five branches
have three sites apiece, and we thought that was appropriate
considering all the inspection activities and coordination
of our technical issue resolution that's associated with
restart preparations by the licensee. And I'll go through
more specifically how we're organized in a minute.
Next slide, please. I thought I would talk a
little about the functional responsibilities for the
Division of Reactor Projects. One of the most important
functions we have is inspection program management. DRP is
the clearinghouse for the inspection program. We're sort of
a gatekeeper for regulatory activities associated with the
specific sites. We manage the site-specific inspection
plan.
I expect the branches to be cognizant of all NRC
activities. That means specialist inspections that are
ongoing by the Division of Reactor Safety Inspectors, DRS,
allegations, status of enforcement actions. The branches
are knowledgeable of all the inspection findings,
performance indicator information, and any outstanding
inspection follow-up items.
So all regulatory activities and issues that
impact on inspection responsibility are pretty much
processed through DRP.
We maintain a continuous on-site inspection.
Specific inspection activities are carved out for the
residents on a periodic basis, and that's, of course, within
the context of the new baseline inspection program.
But there is a premium placed on that on-site
inspection and the ability to observe activities firsthand.
DR. WALLIS: Excuse me. Continuous to me means it
goes on all the time. That can't quite possible.
MR. DAPAS: We don't have 24-hour coverage.
Continuous meaning that we have a day-to-day presence.
DR. WALLIS: That everyday there is a presence.
MR. DAPAS: Yes, correct. Daily on-site
inspection would probably be more appropriate.
MR. CALDWELL: They also, they live in the general
area and are available to go in for event response, or if
there is a particular issue.
DR. SIEBER: Do you have any problems filling
those jobs, are you shorthanded?
MR. DAPAS: I was going to talk a little about
some of the staffing challenges we have in maintaining the
resident positions fully staffed and give you an idea of
where we're at.
DR. SIEBER: When you do that, you can also talk
about rotation, there is a certain rotation that's supposed
to occur that sometimes doesn't because of lack of
personnel.
MR. DAPAS: I was going to comment on that
specific item.
DR. SEALE: I would also like to hear about
growing those positions in the sense that the revised
inspection process, the interest in risk-informed regulation
and so forth seem to be adding to the challenges that the
inspectors face, having to operate in a slightly different
environment, knowing when to inquire of the risk analysts
about appropriate information concerning the operations at
the moment and so on.
I would be interested in how you are growing those
people in that sense.
MR. DAPAS: I think we will touch upon that. If
we don't, point that out, please. The residents are the
focal point for agency interface with the licensee. Of
course, there's the routine exit meetings and where the
resident staffs discuss their specific inspection results.
They maintain cognizance of the results of any DRS
inspections. When the licensee identifies any type of
degraded equipment, which would result in like a technical
specification limiting condition or LCO entry, that's
communicated to the resident inspector and reportable events
are communicated to the residents, any notice of enforcement
discretion requests that are developing.
Basically, the resident is the information conduit
and that includes licensing issues. Certainly, there's
discussions between the NRR project manager and the specific
licensee representatives involved in licensing activities,
but the residents are cut in on that and they inform the
region of outstanding licensing issues.
So they're clearly the focal point for that
communication between the NRC and the licensee, which
underscores our goal of assigning mature, professional
individuals to the sites, because they are the eyes and ears
of the agency, in many regards.
Also, the resident staff serves as first
responders for incident response, as Jim Dyer mentioned and
Jim Caldwell. The resident inspector would respond to the
control room and the senior resident inspector would respond
to the technical support center for any type of emergency
event declaration, like an unusual event or an alert.
Anytime the licensee mans their emergency plan.
And they provide NRC management with information
to determine the appropriate agency response, monitoring,
standby, or initial activation, and they ensure the licensee
is following their emergency operating procedures and
actions for each emergency event classification.
One of the central things that the residents
communicate early on to regional management and headquarters
management is, is the plant in a safe condition, what are
the licensee concerns, what are the principal areas that
they're focusing on. So that first communication is very
important in terms of the agency responding appropriately.
Next slide, please. I thought it would be
informative just to discuss briefly some of the specific
inspection activities that a typical resident encounters.
There is clearly a focus on operations. We target
activities where the plant is configured with the greatest
risk impact. As an example, if the licensee is going to
perform an integrated test of the emergency core cooling
systems, that involves a lot of coordination between the
operators, both in the control room and in the plant, valve
and switch manipulations.
That may be a risk-significant evolution that we
would want the residents to observe.
Event follow-up, as I mentioned before, that could
be a reactor trip, a partial loss of off-site power, plant
transient, any particular event that challenges the operator
response and the residents are there to follow-up on that.
DR. POWERS: Let me legitimately make a point
about this response to any event that occurs, that the
resident has to do. He becomes literally the eyes and ears
in those cases, at least for the first hour or two, he is
the eyes and ears of the agency.
But one would hope that that's an activity that he
doesn't get to practice very often. How does he practice?
How does he develop skill in that area?
MR. DAPAS: We will talk about the detailed
qualification program that a resident goes through, but
there's a lot of mentoring. The senior resident inspectors
have experience in event response. There's, of course,
simulator courses that the resident staff takes in
Chattanooga, where the plant is put through -- the simulator
is run through different emergency transients, and the
residents clearly understand what EOP should be implemented,
emergency operating procedures.
And there is a specific procedure for event
follow-up, which gives the inspectors guidance of particular
things that they should be looking for. And one of the
things that I think is effective when we have our oral
qualification board, which we'll talk more about, it's not
uncommon to ask a question, you'll be walking into the
control room and there's this, this and this going on, what
areas are you focused on, what information are you trying to
ascertain.
So we try, to the extent we can, to prepare the
residents to be able to provide that event response and
communicate the information.
MR. GROBE: The other thing is the resident
inspectors in Region III, the folks we've tried to place out
there, are experienced and, as Marc, said, mature people,
extensive experience as system test engineers, integrated
test engineers, folks that had come through the Division of
Reactor Safety. For the operator licensing program,
operator licensing folks have extensive knowledge and
appreciation of what's going on in the plant.
And within a very short period of time,
approximately a half an hour, they're going to have a ton of
support from the regional office.
DR. POWERS: Yes. But it's really that they're
working on their own and having to use their own judgment.
Of course, nothing schools judgment better than experience.
And the number of events we have, I mean, we just don't have
very many.
So experience -- it did remind me of the
simulators in Chattanooga. That of course, would be a good
thing, having a proceduralized thing, that's a very good
thing.
MR. GROBE: And that's the primary focus of our
requal training. They get extensive systems training
initially, but the requal is primarily focused on the
simulator.
DR. SIEBER: And it's been my experience, also,
that resident inspectors participate in licensing drills.
They are either observers or actual players, and that's
really good experience for them, because they not only learn
what the licensee is supposed to do, but they see how the
licensees act and how to communicate with them.
MR. GROBE: When we get into the new inspection
program, you're going to see that we have less flexibility
to do that.
DR. SEALE: You can't essentially tag along when a
plant operator is going through -- or a plant operations
team is going through a simulator exercise with a
plant-specific simulators.
Do your inspectors get to, if you will, watch this
and ask themselves what their role would be as they go
through that?
MR. GROBE: Once every two years, we have a
requalification inspection, where we observe the licensees'
simulator examinations, and a few years ago, we made a
decision, for that exact purpose, to include one of the
residents on the requal team, and we try to do that whenever
we can.
But we wouldn't be in a mode of interfacing with
the people that are in the midst of an examination.
DR. SEALE: I understand that's a very careful
line there.
MR. GROBE: It gets the operators into the
simulator.
DR. SEALE: Exactly.
MR. GROBE: And it gets the resident inspectors
into the simulator on some periodic basis. The one area
that I'm concerned about, and we're looking at trying to do
something about, is that we have very limited training on
CMG, the severe accident management guidelines. All the
licensees that had training on the CMG materials and our
emergency responders have limited training in that area, and
we're looking at trying to do something to familiarize the
staff and management on the severe accident management.
MR. DAPAS: When I was talking about event, I
talked about it in the context of a significant event. Of
course, event can cover a broad spectrum, certainly.
One of things that we engulf with our event
response procedure is an assessment of the risk associated
with that particular event to determine should we initiate a
special inspection, and that's pretty clearly defined.
DR. APOSTOLAKIS: How do you do this? How do you
assess the risk?
MR. DAPAS: We look at conditional core damage
probability. We look at what was the particular equipment
configuration, mitigative systems, et cetera, and what is
the risk associated with that challenge.
Obviously, when you have an event, if it's a loss
of off-site power, reactor scram, you had the initiating
event, now what's the consequence of that, what systems were
available.
DR. APOSTOLAKIS: So for each unit, you have a
PRA?
MR. GROBE: No. We have very limited tools
available to the residents, broad guidelines on what are the
most risk-significant systems and things of that nature.
MR. DAPAS: But it's different than what we did
have in the past, which was more deterministic. I think as
a result of the Indian Point 2 event, we incorporated more
risk perspectives into our event response procedure.
DR. POWERS: I'm not sure we can get into it right
in this presentation, but one thing that you might comment
on, we have discussed this issue of tools, risk tools
available to the residents and the wisdom of whether they
really want tools, to have more tools or not, because
they've got a full-time job as it is, that's maybe adequate
if they have risk information resources available to them,
the role that normally is played by your senior reactor
analysts.
But asking a guy a question and being able to look
it up yourself are two different things. So this balance
between information directly available to them and resources
available to them is interesting. I don't know how you make
the decisions. If you have thoughts on that, it would be
interesting to hear.
MR. GROBE: Truly, I don't believe we want the
residents doing risk analysis in an event response. They
need to be aware of what's going on at the plant, what are
the precursors to further severity of the event, making sure
that the licensee is focusing in the right areas and
providing information to us.
But both of your risk analysts are on-call. We
got into this just recently with an event. It's difficult
to provide risk analysis on any sort of short timeframe.
We're trying to develop a concept where within a few hours,
they can provide the agency some risk insight, but not any
sort of analytical or very technically defensible risk
analysis on a period of a couple of hours, to determine
whether or not that could provide further insight on the
extent of the team that we should send out or the type of
response the agency should take.
Within a matter of 24 hours, we should be able to
provide some fairly defensible risk analysis of what's
happened.
From a responder point of view, 24 hours is not
terribly useful. So there is an interesting conundrum
there.
DR. POWER: That's really incredibly useful
information there, because I'm wrestling with how fast we
should be able to do risk information and I think you've
given me a key. Clearly 24 hours is too long. Now, what is
the appropriate time? It sounds to me like an hour or two
is the kind of rate you'd really like to be able to do
things in.
MR. CALDWELL: Let me clarify something here.
What Jack is talking about is the type of follow-up event
response we would conduct. The inspector, the resident is
still going to go to the site on an event response and
they're in the mode of observation. They'll go to the
control room, they'll observe operator actions, they'll
observe plant conditions, and that information will be fed
back to us.
But they will not be constrained by some sort of
probabilistic review. But our follow-up event response
would -- our special inspection or AIT or whatever we decide
we might need will depend on the risk of the event itself.
MR. DYER: I think the residents need to have a
general understanding of the risk models, what are the
vulnerabilities at the plant. As they go in and they
initially respond, they're not in an inspection mode.
They're in a protect public health and safety mode in the
incident response, as we all are in that role.
And so from that perspective, when they go in,
they need to know what are the critical assumptions, what
are the vulnerabilities, what are they going to check on,
what are they doing, are they following their EOPs, are they
staying in their modeled assumptions and that.
DR. BONACA: I have a question. RES has been
developing plant-specific models, PRA models, they are
simplified, or apparently they're getting into a more
complex presentation of the plants.
Are they available at the region level, those
models?
MR. GROBE: Sonia.
MS. BURGESS: Yes. The models that we are talking
about are available in the region. Mike Parker and myself
are the ones that have the models here in the region. The
residents at the sites do not have the models.
MR. DYER: They're going to make a presentation
and talk to you later on. So I think the answer is yes.
MR. GROBE: The residents understand the
risk-significant systems and they understand that their
principal focus is do you have the ability to move water, do
you have the ability to provide electrical power where you
need it, do you have containment through piping systems. So
that's what they're focused on, what the licensee is
prioritizing as far as their response to the event, and
that's where they need to be focusing.
MR. DAPAS: I think that's best illustrated -- we
had a recent example here with Palisades, where they had a
problem with the diesel generator output breaker, where the
breaker failed and they could not open it. They had lost
control power. The residents responded to the control room
to understand what was the impact on emergency A/C power
availability and communicated out to the branch chief, and
then we had Sonia Burgess involved looking at what's the
ongoing risk impact of not being able to open the output
breaker and what damage may have been -- when you motorized
the generator, was there a problem.
So that would provide us a perspective, what's the
risk significance of the plant continuing to operate in this
condition and should we provide any augmented support to the
resident staff.
DR. WALLIS: As the technology advances, one could
imagine that inspectors in the future could have some
handheld computation device which would give them a SPAR.
MR. GROBE: We get very anxious when we start
talking in that area, because a lot of this is instinctual
on how you respond to an event. Let's just say we get
anxious.
DR. POWERS: Well, I think my own view was that
inspectors have more to do with providing the input to risk
modeling on a pump than they do running the pump.
DR. SEALE: They need to be able to communicate.
DR. POWERS: And you'd be -- I mean, all of these
things. One of the biggest concerns that I have about the
oversight program is it's taking away from hours in the
plant to hours at the desk, and that's a tradeoff which
ought to be consciously made.
And having risk tools to play with, it quickly
becomes risk tools that you have to play with and that is
just another detraction from eyeballs on the plant.
But I'm looking at, at the same time, this guy
should have all of the support he thinks he needs in
answering questions, in his mind, about risk. So it's
really tools for Sonia and her team that I think we're
talking about here.
DR. BONACA: On the other hand, my question was
more in the direction of just part of the maintenance rule
now, the operators can take out-of-service multiple
components and, of course, there is a requirement for the to
evaluate the risk significance and to what extent a resident
does a spot-check for a given configuration that he may
consider risky enough for him to ask a question, without
having to depend entirely on the plant staff.
I think that is an important objective long-term,
it seems to me.
MR. DAPAS: Nora Collins was smiling. She is
going to be talking later about on-line risk and I think can
provide some insights in that area.
MR. CALDWELL: There are a couple of issues
associated with the SRAs availability of having the analyst.
So we're looking at succession planning for the SRAs, but
integral in that is there's a task group they're putting
together with NRR and the regions to look at that question.
But integral to that is a discussion on training
and what types of training that the various levels need and
one of the -- the regions, I guess, got together and decided
one of the aspects of training that all the inspectors need,
including the resident inspectors, was risk inspection
planning, which would go to what you're talking about; what
things should you look at and when should you look at them.
So there is a task group that's going to look at
the types of training that should go to the residents, the
type of residents that should go to senior inspectors here
in the region, and succession planning for the SRAs.
DR. POWERS: I think that speaks to the issue of
how detailed and how high quality we have to have the risk
resources, not necessarily the turnaround time, but the
quality and detail, which is an issue in itself, whether the
SPAR models are adequate or we need something more detailed,
because inspectors tend to look at things at at least one
level down on the level of modeling PRAs.
I mean, it's the same problem the engineer at the
plant has. He tends to work on things that are a level
down.
MR. DAPAS: For the sake of timeliness here, I'm
just going to kind of go through examples of each activity
here, but I'll just point out a couple of things.
Operability evaluations, clearly, the residents get involved
in evaluating the impact of degraded equipment.
If a pump is supposed to deliver X amount of flow
for the surveillance procedure, it doesn't pass the
surveillance test, and then the licensee does an evaluation
and says, well, the pump can still perform its intended
function, that can lead into a 50.59 evaluation, because the
pump operation may be different than described in the final
safety analysis report, et cetera. So they get involved in
that.
Severe weather preparations --
DR. POWERS: We're going through a substantial
change of 50.59.
MR. DAPAS: Correct.
DR. POWERS: And there's a high judgmental
capacity content to this on what is a minimal change in the
impact assumptions, things like that.
MR. DAPAS: I think our safety system design
inspections get more intrusive into the quality of the
50.59. The role of the resident is the licensee conducts a
50.59 and they kind of look at does this make sense. If
they need more additional help, they can engage DRS
inspectors.
But looking at it from the programmatic aspect, I
think select samples as part of your design inspection.
MR. GROBE: I think, if I understand your question
correctly, it was what's the staff's reaction to the
judgment and the subjectivity that might go into the new
decisions in the rule.
I think the staff truly was uncomfortable with
some of the Draconian outputs of using the rule as it was
written before. Some unreviewed safety questions that were
really insignificant would result in enforcement action.
So on that specific issue, while it involves more
judgment, I think the staff is more comfortable. There are
a number of areas with the new inspection program that the
staff is not as comfortable as what we used to have and we
can get into some of those.
But that's an area I'm not sure we have a lot of
concern with. The implementation we haven't actually seen
yet, so we're going to have to walk through that.
DR. SIEBER: I think we would like to hear your
concerns later on that, so we know what they are.
MR. DAPAS: Now I more fully understood your
question. Severe weather preparation, with the plants we
have located here in northern climates, we get involved a
lot in that. In fact, we had an issue at Point Beach
regarding freeze protection for a safety injection recirc
line, tangible example of where inadequate freeze protection
resulted in problems.
And problem identification and resolution. An
integral part of each inspection procedure is ten to 15
percent of that is dedicated to follow-up for problem
identification and resolution, and that, of course, is the
foundation of new program, corrective actions.
And there's two aspects to that. Of course,
annual review and then follow-up on issues specific to the
area being covered by the individual module, like
surveillance testing.
DR. SIEBER: In that regard, under the new
oversight process and significance determination, they
aren't writing as many violations. On the other hand, we're
probably writing more non-cited violations, and all those
are supposed to go into the CAT.
Do you folks follow-up inspecting CAT to make
sure?
MR. GROBE: Not all of them. There's two. One is
that we do a regular inspection of the effectiveness of the
corrective action program and that's run out of Merck's
division, and we have people on that inspection.
In addition to that, we sample a portion of
non-cited violations as part of that inspection, but we
don't look at all of them, and that's part of the new
inspection program that actually makes sense, because the
violations we identify are a very small portion of the total
number of issues that need to be corrected on a yearly
basis.
So we'll select a portion of the violations we
identify and that were non-cited, as well as a large number
of other issues that we focus, from a risk perspective, on
trying to get the more important ones.
DR. SIEBER: I guess my personal feeling is that
NRC gave up something when it moved from deterministic
systems into risk-based systems and significance
determination. What you gave up was the ability to write a
violation and get a written response and a commitment from
the licensee that you could follow-up up on and for a given
unit that could have been anywhere from five to 20 items a
year.
On the other hand, once you give that up, you have
to put a little more emphasis and follow up with a
corrective action program to make sure that it didn't
disappear.
MR. DAPAS: You're right. That's a balancing act,
obviously. The crux of the new program was what's the
appropriate amount of regulatory burden. You're writing
violations, the licensee has to respond, what is the
threshold for that.
That's why we -- we put great stock in our problem
identification and resolution inspection. We think that's a
critical aspect of the new program.
DR. SIEBER: Even the Commissioners see that as a
key. They're very adamant about that.
DR. POWERS: Well, I think the Commissioners see
it more than the headquarters staff.
MR. DYER: Well, I don't know that. I think it's
we -- a lot of the violations, I think, as Jack said
earlier, a lot of the violations that we wrote, we were
spending a lot of time on correspondence that didn't improve
the safety of the plant.
DR. SIEBER: Yes. We were on the other end.
MR. DYER: So I think the new program does allow
-- what we have to do is take significant actions when we
find a licensee is not -- when they break that trust.
And one of the things we get through here, when we
start looking at the new program, that is the importance of
the cross-cutting issues, in my mind, as a regulator, and,
in particular, the corrective action program.
As you said, we are turning a lot over. This will
make for a more efficient and effective way of regulating
and allow the licensee to prioritize, but they have to have
a good program.
MR. CALDWELL: There's a major challenge to the
licensee that comes out of this. In the past, when we wrote
a violation, it came out in our report, they had to respond.
Typically, they had to get senior management to agree with
the response, so that the managers were heavily involved in
those activities, at least the inspection activities that we
conducted.
Now, it's included in their corrective action
program. So the licensee's management has the challenge of
staying involved in those issues that occur. They are going
to have to be asking more questions and getting more
involved in their corrective action program. So it is a
challenge.
MR. GROBE: Our ability to cause licensee
management to engage in issues is diminished under the new
inspection program. One of the things that we got good at
and our staff gets very good at is appreciating a broader
perspective and focusing on root cause.
Now, as Jim indicated, the licensee has to take
that burden completely on themselves, which is appropriate,
but our ability to direct that, unless it results in a
risk-significant issue under the SDP, is limited.
DR. SIEBER: One final question, which you can
answer yes or no. You have exit meetings when you conclude
an inspection, either a resident or a specialist inspection.
Since the new oversight process and the burden has changed,
do you have any idea whether the level of management that
attends those exit meetings has changed to a lower level
since there is less management involvement?
MR. DAPAS: I can actually comment on that
specifically. I think there's actually been a higher
engagement of management, because we communicate at that
exit meeting some issues that may not be documented in the
report, and that's a program office policy decision that
some issues that don't rise to the threshold of an
inspection finding or a green issue, the licensee is
interested in hearing about those and those are communicated
at the exit meeting.
Many times, a site vice president or plant
management wants to hear those firsthand.
DR. SIEBER: That's good input for me, because I
would have expected, just human nature being what it was,
that it would have gone the other way. So that's good.
Thank you very much.
MR. DYER: I think the other dynamic in that is,
again, the economic pressures. Licensees realize that the
NRC inspection findings that are below the threshold for
being documented in the report can, in fact, affect their
operation, you know, may provide them an insight or
something maybe to address before it -- it's a precursor.
In today's environment, that's necessary.
DR. SIEBER: Thank you.
CHAIRMAN BARTON: Gentlemen, we're going to have
to move this along a little bit. Maybe we can have some
more questions during the lunch break. We're one-third
through item three, which was supposed to be completed at
this point. So I think we need to kind of hold questions
and have maybe some discussion during lunch. Otherwise,
we'll never get through today.
MR. DAPAS: The last point I was going to make is
performance indicator verification, obviously an important
activity the residents are engaged in.
We had a number of lessons learned from the pilot
program that have been communicated to licensees, and that
underscores the importance of consistent application of the
performance indicators, and I think we're going to talk more
specifically about those a little later on.
Next slide, please. This is just a slide showing
how the division is organized, as Mr. Dyer said, relatively
consistent across the regions. We are currently only one
site is staffed at N+1, that's D.C. Cook; of course, our
agency focus plant, and we're actively recruiting to fill
the reactor engineer vacancies that exist.
I'm going to talk a little bit more later on about
the challenges that have been presented to DRP in trying to
fully staff in the context of the new inspection program
requirements.
DR. WALLIS: You have four vacancies here at the
reactor engineer level?
MR. DYER: That's correct.
MR. GROBE: What we've done is added overage
positions. Several of those positions are overage, and
we've done that in operator licensing and both engineering
branches and in the reactor engineering DRP. And the goal
is to minimize the amount of downtime we have, when we lose
a number of the staff.
So we're trying to fill those up. Once we fill
them, we're going to have a substantial buffer, we hope.
MR. CALDWELL: These reactor engineers are not
intended to be overage positions. We do have overage
positions elsewhere, but we're trying to stay ahead of our
-- unfortunately, we never meet our ceiling. And so we're
trying to get ahead of the ceiling so that we at least have
utilization of all the FTE who are left.
MR. DAPAS: We bring the reactor engineer on board
and a vacancy occurs at the plant and that's got to be our
primary focus, is making sure the sites are fully staffed.
So it's an ongoing challenge to try and fully staff the
reactor engineer position while keeping the resident program
fully staffed.
DR. WALLIS: Because if you lose one more, you'll
have none, it looks like. Four out of five.
MR. DAPAS: We're heading the other direction.
DR. APOSTOLAKIS: I have a question. I'm looking
at the report from the web site regarding the maintenance
rule. It says that you interviewed two licensed reactor
operators and three senior reactor operators to determine if
they understood the general requirements of the maintenance
rule.
Is this something that you do routinely? I mean,
what if they don't understand it, what would you do?
MR. DAPAS: Which -- I'm not familiar --
CHAIRMAN BARTON: This is the follow-up to the
maintenance inspection report that was done in the regions.
Part of that was going in and asking various people on the
stations what was their knowledge of the maintenance rule.
Remember that part of it?
DR. APOSTOLAKIS: Is it still the situation that
we will interview people to see if they understand something
under the new revised oversight process? Is that part of
the baseline inspection?
MR. DYER: Not that I know of. That might have
been a special inspection. Was that done under a TI?
MR. DAPAS: I thought that was associated with
implementation of the new maintenance rule.
CHAIRMAN BARTON: That's what it was.
MR. GROBE: It was a special inspection.
MR. SINGH: It was a follow-up inspection to the
original inspection.
MR. GROBE: Right, where there were open issues,
and you go back out, and part of that, I think, was ensuring
that the licensee understood performance goals and on-line
risk assessment.
CHAIRMAN BARTON: A lot of that was going into the
control room to ask the SROs, the supervisors, how was their
knowledge of the maintenance rule.
MR. GROBE: We have one our maintenance rule
experts here, Any Dunlop.
MR. DUNLOP: The maintenance rule baseline
inspections and most likely what this was, there were some
open issues that came up during the baseline inspections and
what we did at each of the sites, when we had open issues,
we would go back and follow-up on them, and that's most
likely what this inspection report is discussing, a
follow-up inspection to address any open issues that had
come up.
I'm not sure, I wasn't part of the follow-up. I
was part of the original inspection.
DR. APOSTOLAKIS: It's not really this specific
thing that I'm asking about. I'm just asking, in the
future, with the new oversight process, is there room there
for us as an agency to see how much the licensee knows about
something? Aren't we supposed to be moving towards a more
performance-based system? Is there a cross-cutting issues
that says try to see how much this operator at the plant
knows?
MR. DUNLOP: I think the maintenance rule is
supposed to be one of our first performance-based rules that
we put into effect and I think the purpose of the baseline
inspections was to, unfortunately, have a programmatic
review of what the licensees know and how the program was
actually put together. I know as part of the new A-4 new
maintenance rule, there will be some PI developed and we'll
be doing some inspections at some of the sites.
How much we'll be looking into the programmatic
aspects versus the performance-based, I don't that's been
determined yet.
MR. DUNLOP: I believe that inspection was sort of
a -- the baseline and the follow-up was sort of to set the
groundwork to then go forward. In the future, I don't
believe we're going to be quizzing people on their knowledge
level.
I think part of the baseline inspection, if I
remember correctly, part of it was to see did the training
take. When you go in and you took it, when they had
implemented a change in the program, part of our inspection
is, okay, did the training take, do people understand their
responsibilities.
And as a basis for that, that was the nature of
the questioning and I think that was specifically called out
in a temporary inspection, which would be not part of -- it
would be a one-time inspection, not part of a routine
inspection that we would continue.
So it would take the headquarters, if they
decided, for some other reason, that we needed to go back
out and periodically reverify the training, then we could
look at it again, but it wouldn't be part of our normal
routine program.
MR. GROBE: I was going to say, by contrast,
whenever we observe an activity, I expect the inspectors to
be assessing the knowledge level of the people that are
performing that activity of the procedures and the specific
work they're doing.
So we would continue to evaluate, if we observe a
maintenance activity or a test activity or an operations
activity or talk to an engineer about a calculation, we'd be
assessing their understanding of what they're trying to
accomplish and their understanding of the procedures
involved in that.
So we will still be getting into assessing the
capability of the people to accomplish the work they're
trying to accomplish for those activities where we're
observing performance.
But as Andy pointed out, that was strictly a
programmatic inspection. It didn't involve actual
implementation of the program as much as on a day-to-day
basis, as much as the programs, procedures and training.
DR. APOSTOLAKIS: There are similar findings in
other places, and I'm not questioning you why you did this.
I'm trying to see what the future will be under the new
oversight process. We have the cross-cutting issues, of
course.
MR. DAPAS: We have an individual that actually
has probably conducted the resident inspector portion of
that and certainly can speak to what the new program entails
as far as the maintenance rule.
DR. APOSTOLAKIS: While you're getting the
microphone. I've made several findings here that really I
didn't expect to see. For example, The company nuclear
review board members were thoroughly prepared for the
September '98 meeting.
MR. DAPAS: Does that embody observations and --
DR. APOSTOLAKIS: Yes, but is it going to be in
the future, are they going to be observe whether people are
well prepared. There was a finding later that the expert
panel deliberations were not recommended, and so on.
And I thought that in the new oversight process,
what really matters is the decisions of the expert panel and
not whether they document what they're doing.
So the question is how much of this is going to
change in the future, if any?
MR. DYER: I think you'd have to look at either
the Quad Cities or the Prairie Island plant issues matrix to
get a better understanding as to what the new program is
going to look like. Davis-Besse was under the old program.
DR. APOSTOLAKIS: I understand. This is old.
MR. DYER: And there's a specific module. So the
resident inspectors, once every 18 months, had to go observe
an off-site review committee, and they do the best they can.
MR. COLLINS: I can talk to that. My name is
Laura Collins, and I was a resident inspector at the pilot
plant, Quad Cities, and did the maintenance rule inspection
portions for the residents there a lot.
The kind of observations that you're talking
about, unless they were to really result in a problem,
because we're more results-oriented, are not the kinds of
things, I don't think, we would be documenting anymore.
But we would still be, if we observed those
things, communicating them to the licensee, so that they can
learn from them.
So if we make those observations, we're going to
share everything that we observe with the licensee, but we
have higher thresholds for findings. There's got to be some
kind of a result of that improper implementation of the
maintenance rule.
DR. APOSTOLAKIS: So that in the future, then, you
would not particularly care about how the expert panel
conducts its business. You would just look at the results.
MS. COLLINS: That's right.
DR. APOSTOLAKIS: Is that the correct perception?
MS. COLLINS: We start with the results.
DR. APOSTOLAKIS: But you may get back into the
thing, I mean, if you want to understand --
MR. DYER: I think one of the things that the
utilities, the vice presidents, are particularly interested
in is if we said we observed the meeting, we have no
findings, a lot of times they'll ask you, what did you think
of the conduct of the meeting, and that's one of those
issues that may be provided below the line, but it's not
going to be documented in the inspection report, there's no
response required.
MR. CALDWELL: And it's not that we don't -- you
said we may not care about it anymore. We still care, but
we wouldn't document it necessarily. We would communicate
it to the licensee, if we felt that would give them some
insight.
DR. SIEBER: Unless you came away with the feeling
that the result was inadequate, and then you may go further
to find out why that is.
MR. DYER: Now, if we come out with an inadequate
safety review, we may take it back to there was an
inadequate safety review, it was not adequately reviewed and
people weren't prepared, something like that. But it would
be tied to the results.
MR. DAPAS: Or then the expert panel concluded
this system should be -- there should be performance goals
established for this system to review its importance and
risks, and the licensee didn't address that, that would be a
result.
DR. APOSTOLAKIS: I agree, but that is clearly
within the new rules of the game.
DR. BONACA: How do you know if there is no
implementation. What I'm trying to say, there are examples
there, some examples where the PRA defined some component
that's safety-significant, but determined it wasn't really
safety-significant and, therefore, they did not report this
activity.
Now, there is an importance also in the
documentation. You've got to make a determination that the
decision ultimately was the correct one. Performance-based
doesn't mean you're waiting until you have an event. It
means that you're performing the right things. So you still
have a burden on the processes that you have to inspect and
the show of the work.
DR. APOSTOLAKIS: See, what confuses me -- and,
again, I'm not referring to a specific thing, but is that in
Washington, we're being told time and time again that
managing the plant and the organizational aspects are really
the licensee's responsibility and we should not get
involved.
In fact, several of the research projects of the
Office of Research have been killed on that principle. And
then I come here and I see that an appropriate feedback
process was in place, operators responded conservatively to
plant transients, operators were prepared for the possible
closure of feedwater regulating valve surveillance testing.
All this is organizational management, isn't it?
MR. GROBE: No.
DR. SEALE: It's
CHAIRMAN BARTON: It's observation of plant
operations, George.
DR. APOSTOLAKIS: But there is a feedback process?
That's their business.
MR. GROBE: Well, it's also required pursuant to
Appendix B.
DR. APOSTOLAKIS: So what we are told there is not
entirely accurate. I'm trying to reconcile the views. It's
very fuzzy, isn't it?
CHAIRMAN BARTON: Especially when you're assessing
management's competence and safety culture versus
observation of plant operation.
DR. APOSTOLAKIS: That's an extreme, I agree. I
agree. But having an appropriate feedback process, it seems
to me, is an organizational issue.
MR. GROBE: I'm not sure what the context of that
was. But it's important, though. For example --
DR. APOSTOLAKIS: Plant issue matrix of
Davis-Besse, dated September 28, '99.
MR. GROBE: For example, within the training
context, the feedback process is absolutely critical,
because on a system-based training process, you have to have
that loop. In the training inspection, that's part of what
we look at.
Within the context of an oversight committee, the
engagement of the committee in questioning the quality of
the product and understanding it is critical to the outcome.
So if we only look at the outcome of the meeting, there may
be significant things that they missed because they weren't
well prepared for the meeting.
And it gets to root cause, really. If we're going
to have inspectors in the field observing the activities,
those are the kinds of things we expect them to look at. As
Laura pointed out, those issues wouldn't find their way into
a report today unless they resulted in a risk-significant
finding.
MR. DAPAS: And that's the key. Regulatory
engagement is a product of the consequence of that, but we
would still feed that observation back to the licensee.
MR. GROBE: Exactly. Both positive and negative.
If we found that the people performing a maintenance or a
test activity were very qualified and competent and
displayed that in their discipline, in the way they
approached their job, provide that feedback.
DR. APOSTOLAKIS: So the action matrix of the new
oversight process, that would not be triggered. That would
not be affected by these observations. You just provide the
feedback.
MR. GROBE: That's right.
DR. APOSTOLAKIS: Because there is nothing white.
MR. GROBE: That's right.
MR. CALDWELL: You also understand we're in the
initial implementation phase of this new process, this is
what we think, we may learn something as we go along and
change our approach, but right now, that would be the
outcome.
MR. DYER: What we found at the two pilot plants
that we've implemented the program in, when we first went to
-- we actually applied the SDP to it and we went through our
formal exit and said here's our formal observations and the
utility management look at us and say is that all, you've
been here for a month, you need to give us more feedback.
It evolved out of that --
MR. GROBE: Tell us what you really think.
MR. DYER: Yes. And evolved out of that is we
have a formal exit now where we say here is what is formally
going in the inspection report, here's our observations that
aren't going to make the report.
MR. GROBE: Dr. Barton, in the interest of time,
let me quickly go through the next six or eight slides, and,
Bruce, keep up with me.
In the Division of Reactor Safety, we really have
five major functions; engineering inspections, health
physics and emergency preparedness inspections, safeguards
inspections. We also have operator licensing and that
includes initial examinations, upgrade examinations, as well
as requal inspections, and incident response is one of the
major functions of the Division of Reactor Safety.
Let me just highlight a few things in the
engineering inspection area that are new and exciting. We
have a much stronger emphasis today on design inspections.
We have an inspection called the safety system design
inspection, or the SSDI. We also have an inspection that
focuses more heavily on the Appendix R design of the plant
and the ability of the plant to sustain a debilitating fire.
Those are two inspections that are new, much
stronger emphasis in the design area.
DR. POWERS: I attended the fire protection forum
and it was an interesting complaint. They said, gee, you
guys are focusing all your attention on this Appendix R and
the safe shutdown and neglecting all this other fire
inspection stuff, and it's just not right. The fact is we
haven't done the Appendix R safe shutdown inspections in the
past to the extent that they probably should have been done.
And now we're just bringing things back to some
sort of proper balance.
MR. GROBE: And we're not disregarding classical
fire protection either. That's part of the resident
program. But there's a summary on the slide of the types of
engineering inspections we get engaged in and we'll go into
some more detail later on some of those.
In the safeguards area, we look at contingency
response, access control and fitness-for-duty primarily,
and, as Marc indicated earlier, each component of our
inspection program, we look at problem identification and
resolution or the effectiveness of the corrective action
program.
DR. SIEBER: When you do an OSRE, though, that
also involves the operations people, right? With strategies
and so forth.
MR. GROBE: Exactly.
DR. SIEBER: But that's not part of your baseline
inspection. You're just looking at cameras in the field and
--
MR. GROBE: Contingency response is actually --
DR. SIEBER: Is that in there?
MR. GROBE: Yes. It's kind of in there in hiatus
right now. OSRE is suspended and we're trying to work with
the industry to come up with a better way to do
force-on-force drills.
DR. POWERS: One of the questions that I've had
about that is the extent to which we can use some of the
computational tools that have been developed by the national
laboratories, among other people, I think, for simulating
these force-on-force exercises.
They won't do everything that the OSRE does for
you, but they would certainly augment or maybe reduce the
need to do actual OSRE type activities. Have you looked
into this at all?
MR. GROBE: I don't know.
CHAIRMAN BARTON: It's a civilian industry
initiative at this point.
MR. GROBE: These are simulation type tools that
--
DR. POWERS: They were originally developed -- the
ones I know about, the ones that were originally developed
were Air Force bases in Europe. They became concerned when
the Red Army was running around, could they, in fact, defend
their weapons systems from an intrusion force, and that
would be different than an ordinary military fighting force.
And they had done a lot of exercises with these
guys with laser rifles and things that had sensors all over
them and they computerize it and out of that they come up
with what's the optimal strike force against it, what are
the vulnerable sites, locations on the facility and things
like that.
MR. GROBE: I'll look into it.
DR. POWERS: They eventually got very
sophisticated, but I don't know whether they've gone into
the commercial sector or not.
MR. GROBE: I have not heard about it.
DR. POWERS: They resulted in massive changes to
the way they the military protected their facilities. I
mean, they were shocked at how easy it was to break in.
MR. GROBE: Appreciate that insight.
In the rad protection area, three primary focuses;
plant protection of the people on-site, radioactive waste
and transportation, and protection of the public, effluents
and environmental protection.
DR. SIEBER: This is probably where Illinois
Department of Radiation Safety comes in quite a bit.
MR. GROBE: Well, they're much more intrusive.
They have reactor safety specialists that are resident at
the sites. They are very sophisticated, very impressive
organization. Not quite as good as us, though.
DR. SIEBER: Well, I knew that.
MR. GROBE: In emergency preparedness, we observe
exercises, as well as do programmatic reviews on a regular
basis. Operator licensing, I mentioned earlier, we do
initial exams. Sometimes those are SRO, instant SROs exams,
sometimes reactor operator exams.
We also do upgrade exams and requalification
inspections. In each area, again, problem identification
and resolution.
Incident response, we maintain and coordinate for
the region maintenance of our incident response capability,
and that includes exercises, training, equipment and
facilities, as well as interface with Federal, state and
local, and unique in Region III is some tribal interface up
at the Prairie Island plant.
The division is broken up into four branches. Two
engineering branches, one focusing primarily on electrical,
which includes environmental qualification, I&C, fire
protection, electrical engineering and analyses; mechanical,
which gets into mechanical, civil structural, as much as we
do these days, maintenance rule, in-service inspection,
steam generator replacement, steam generator tube
inspections, things of that nature.
An operator licensing branch, which is very busy
these days. Unlike some other regions, for example, Region
I puts emergency preparedness and operator licensing
together in one branch. This branch is strictly focused on
operator licensing.
DR. WALLIS: Does mechanical engineering include
thermal hydraulics?
MR. GROBE: Are you talking, for example, of -- we
do heat sink inspections.
DR. WALLIS: Heat and fluid, yes. Water and
steam, where they are and what they're doing, how well they
are performing their function.
MR. GROBE: Within the reactor, we don't do a lot
of inspection from a thermal hydraulic point of view.
However, from a heat sink point of view, heat exchanger
performance, we do some inspection in that area.
Plant support is health physics, emergency
preparedness and incident response in that branch.
MR. DAPAS: You can get into those aspects,
though, like with an operability evaluation, where the
licensee is using thermal hydraulic analysis to support a
particular conclusion. We might look at that.
MR. GROBE: We just completed safety system design
inspection at Point Beach and the focus was the service
water system, a lot of thermal hydraulic analysis involved
in that.
That was fired through the scrub oaks on Division
of Reactor Safety.
CHAIRMAN BARTON: Very good, you did good. At
this point, I'd like to break until 10:30.
[Recess.]
CHAIRMAN BARTON: We're back in session. Marc,
are you still on?
MR. DAPAS: Yes. The next presentation that we
wanted to address was the comparison of the new program to
the old program. I think some of you have raised some
questions. In the context of the new program, I think this
is an opportunity to more thoroughly address some of those.
As an example, I know Mr. Seale raised a question
about the resource expenditure tracking and we can talk
about what challenges that presents.
Rather than continue to use overheads here, if we
can just go through the slides, if that's okay with you.
CHAIRMAN BARTON: That's fine.
MR. DAPAS: Great. Starting on page 20, when we
looked at the old program, that was pretty much broken up
into thirds between the core program, what was previously
termed the regional initiative, and special inspections.
Special inspections was our mechanism for following up to
specific events.
And as we talked about a little earlier, we use
risk as a gauge in determining what's the appropriate
engagement in terms of numbers of folks we send to the site.
The regional initiative, of course, involves some
subjective judgment about the declining licensee performance
in a particular area or aspect of plant operations, and we
would send some folks out to do a more intrusive review of
that.
Under the new program, it's pretty much baseline
loaded, and the baseline represents that minimum amount of
inspection required to verify that licensee performance is
within the licensee response band, whereas under the old
program, the core represented that minimal amount of
inspection to verify the plant was being operated safely.
As you know, whether you're in the licensee
response band or regulatory response band, there's still a
sufficient safety margin. So it's a little different
approach.
There is clearly greater flexibility in applying
inspection resources under the old program. An inspection
procedure could be closed using judgment on whether the
intent was met.
For example, inspection procedure 71.707, which
dealt with operational safety verification, that would
include observation of control room activities, an
engineered safety system feature walk-down.
The inspector could decide, based on reading the
inspection procedure, I met the intent of this procedure
with X number of hours. Under the new program --
MR. GROBE: Before you go on, that's exactly what
got Davis-Besse down in the 1,800 hour range. I'm sorry.
MR. DAPAS: When you look at the sampling size
under the new program, X number of surveillance tests need
to be observed, X number of operability evaluations.
There's a certain periodicity; for example, looking at maybe
a couple samples a month.
And under the one-size-fits-all approach within
the licensee response band, the baseline inspection program
is fairly rigorous in the scope and estimated number of
hours to complete the inspection procedure. And as Jack
pointed out, that can translate to what is perceived to be
an increased regulatory burden for a licensee like
Davis-Besse, where there was more flexibility in determining
was the intent of the procedure met.
In the new program, you have to implement the full
scope to satisfy the inspection procedure objectives.
MR. GROBE: Philosophically, what we've done is on
the side of the angels. We looked at risk, we picked out
what are the most significant risk-related activities.
Based on the impact on the risk of that activity, we
identified those attributes that were important to inspect
and we assigned, developed inspection procedures and figured
out how much resources it would take to do it, and it came
out to, whatever it is, 2,100, 2,200 hours.
The challenge, point number one, is that that
consumes almost 95 percent of our resources. So the
combination of things; the new thresholds to get to a white,
yellow or red finding are fairly high. So we don't expect
to have much supplemental inspection. But we also have much
less capability to respond to a problem of that nature, and
we're going to have to depend on other regions and
headquarters to supply us resources, whereas in the past,
that 33 percent regional initiative, we could target those
resources based on management judgment.
That made us less predictable, and that was one of
the concerns the licensees had.
DR. SIEBER: The real opportunity, and I realize
it may be a second or third generation in the application,
to achieve this is the extent to which you can make the
inspections plant-specific, with justification.
MR. GROBE: We make all the inspections
plant-specific.
DR. SIEBER: In terms of coverage, not in terms of --
MR. GROBE: In terms of amount of hours, is that
what you're saying?
DR. SIEBER: Yes.
MR. GROBE: We talked about modifying the baseline
based on performance. But that gets back to where we were
and there is a lot of reticence to do that very quickly.
If, after a few years, we find out that they're --
DR. SIEBER: Maturing.
MR. CALDWELL: But it also is the way the system
is set up, we're not capable of doing that right now,
because there is not a gradation in green band. That's
licensee response band, that's where we stay at. So those
folks that are in that band get the baseline inspection
program, whether they're at the top of the band or at the
bottom of the band. That's the way the new oversight
process works.
So to try to come up with a way of reducing
inspection of one licensee over another is not -- within
this current program, that's not possible.
MR. GROBE: The pendulum swung to predictability.
We are extremely predictable now. The question is whether
or not we've taken too much of the judgment out, such that
we can no longer predict problems.
DR. POWERS: What I worry about, especially this
point about the inspectors losing judgment capability, it
seems to me that the good inspector can quickly say in this
area, I've met the intent here, and there are enough
problems for me to worry about here, this other area is more
complicated for me to understand, me personally to
understand than the average inspector, and there may be
bigger issues here, and so I need to spend more of my hours
here.
That judgment seems to be something that I want
him to exercise very much.
MR. GROBE: It's been reduced in the new program.
DR. POWERS: And it seems like he's -- that that's
the flexibility that is a real loss.
MR. DAPAS: Let me comment on that. As I
understand the new inspection program, the sampling size is
intended to be risk-informed. Operability evaluations is
clearly going to be a risk-significant activity. If the
licensee doesn't adequately evaluate the impact of degraded
equipment, will the equipment perform its intended function.
That's clearly related to risk. So what is an
appropriate sample size to gauge how the licensee is
performing in that particular area.
In the past, under the old program, you may decide
to watch one surveillance test and you felt that you've met
the objectives of the procedure. Under the new program,
there may be two surveillance tests that you look at on a
monthly basis, and that's the risk-informed sample size.
So it's more prescriptive in that regard and
that's why the hours are more rigorous.
Now, I think after the first year of
implementation across all the sites, we may end up
revisiting the scope of a given inspection procedure and
we're also providing feedback on a continuous basis.
If the inspector is performing a certain
inspection procedure and feels that the scope of the
procedure needs to be refined, they provide feedback and
that's communicated to the program office.
CHAIRMAN BARTON: Well, the intent of the whole
initial implementation for one year is to make adjustments
after that one year.
MR. DAPAS: Correct.
CHAIRMAN BARTON: Right, sure.
MR. DAPAS: That's my understanding.
DR. WALLIS: It seems to very ironic that the
reason for all this is to get away from prescriptive
regulation. They seem to have moved to more prescriptive
inspection.
MR. CALDWELL: It's prescriptive in the sense that
the inspection size or inspection scope and type were
supposed to be, as best we could, risk-informed. In other
words, you're focusing your resources in the area where
there is the biggest bang for the buck.
The desire to make it such that each region and
each inspector does it essentially the same way for
consistency, but to answer Dr. Powers' question, if an
inspector feels they have to spend more time to accomplish a
given sample size or given objective, they would take the
time necessary to accomplish the objective.
So if the inspector felt comfortable in this area
and was able to get it done pretty fast, that's what they
would do to accomplish the objective of the inspection. If
they felt that they needed more time in another area, they
would do it, they would spend the time.
So the judgment in that respect is still there.
MR. GROBE: But they right now don't have the
latitude to say I'm going to do 18 operability evaluations
versus 24.
MR. CALDWELL: Right.
MR. GROBE: But in addition to that, there's
barriers. We have put -- depending on the types of
inspections, the error bands can be up to 25 percent as far
as number of hours. To go outside that band requires fairly
high approval.
So we need to get engaged in what it is that's
causing the inspector to have to spend a lot more time, as
managers.
MR. DAPAS: And that's because we've communicated
that the baseline inspection program is that minimal amount
of inspections necessary to verify licensee performance is
still within that licensee response band.
MR. GROBE: We spoke earlier about observing more
behaviors and you talked in the context of management.
Those are the types of things that would give you confidence
that you can make your sample size smaller.
If you looked at the procedures and the guidance
and you looked at the training and you looked at how the
people were engaged in their job, in the past, we -- and all
of those things were very positive, so you had a high level
of confidence in the competence of the people and how their
work activity is controlled, we would feel comfortable
scaling back on sample size. Now we don't have that
flexibility.
DR. APOSTOLAKIS: Do you think that the new
oversight process can be modified to accommodate some of
these concerns, without affecting its intent regarding
predictability, for example, too much?
MR. GROBE: It's difficult. One of the things we
haven't thrown on the table is that it's my sense that one
of the motivators of this predictability was the financial
community having confidence in a regulatory oversight not
influencing negatively the financial viability of the
company, from a stock point of view.
So I'm not sure how that would work and we'd have
to do that jointly with the industry.
DR. SIEBER: What do you do with a plant like
Zion?
MR. GROBE: Zion?
DR. SIEBER: Yes. They still do the -- they
haven't applied for decommissioning yet.
MR. GROBE: In the decommissioning area, our level
of inspection is directly related to the level of activity
that the licensee has on-site.
DR. SIEBER: So you would cut back on the number
of residents you have there.
MR. GROBE: There are no residents.
MR. CALDWELL: We have inspectors here in the
region that go up there, and Zion is not that far away, but,
yes, our inspection program is based on the decommissioning.
There is actual inspection plan for decommissioning
reactors.
MR. DAPAS: Which is outside the baseline program.
Moving right along. Certainly, under the old
program, we used deterministic processes in our enforcement
policy to guide our assessment of significance associated
with inspection findings.
Under the new program, we process findings in the
significance determination process, which is based on the
probabilistic risk type analysis. I'd just simplify that
down into two concepts. You've got frequency of initiating
event and then the defense-in-depth regarding mitigative
capability and if you have a particular piece of degraded
equipment or unavailable equipment, you look at what impact
does that have on the mitigative capability.
You look at the availability of redundant
equipment. You can credit operators for recovery actions.
And then there is a plant-specific phase two worksheet that
is supposed to bring to the table the specific
configurations unique to that plant in terms of equipment
redundancy.
CHAIRMAN BARTON: Are they all out and back now,
are the plants commenting on them?
MR. DAPAS: Yes. Sonia, you might be able to
speak to that. I'm not sure of the exact status.
MS. BURGESS: As far as the agency-wide, no. Our
region, yes, with the exception of D.C. Cook. We have put
our comments back to Research, who has, in turn, given them
to BNL.
MR. GROBE: We took a different approach in Region
III than some of the other regions took. We had either Mike
or Sonia out on each site visit to make sure that we had a
clear understanding of the SDP and the licensee effectively
integrated plant-specific issues into the SDP.
Some of the other regions had Research do that or
headquarters staff. As Sonia indicated, she and Mike have
finished all the sites, with the exception of Cook, and we
need to get on Cook pretty soon here.
MR. PARKER: But to be more specific, we don't
have the comments back from BNL and back from Research yet.
So they're not integrated into the current process.
MR. GROBE: We have hand markups of the SDP.
DR. APOSTOLAKIS: Are you comfortable with the
SDP?
MR. PARKER: Am I comfortable with the SDP? I'm
very comfortable with the site visits we accomplished and
the corrections and the adjustments we made to them, but
right now the difficulty we have is working with the
residents, because it's not integrated into the formal SDP
worksheets.
MR. GROBE: It would be interesting to march about
a dozen inspectors up and ask them that same question,
because there's a lot of -- we use the terms risk-informed
and risk-based. The SDP is primarily risk-based.
And an excellent example, and if Laura is still
here, she can provide some of the details, if I screw up on
the details, there was a finding at Quad Cities involving
motor-operated valves, where the licensee did not
effectively correct problems on a timely basis, the
motor-operated valve setup.
The end result was that they had a number of
deficiencies that, if you take together, made it clear that
their motor-operated valve program was not functioning.
When I say motor-operated valve program, the setup
of the valves to make sure that they could handle
differential pressures and all those things.
From an SDP point of view, though, at any given
time, there was not sufficient valves that were determined
at that time to be non-functional, such that you got out of
the green band.
So it was a green finding, yet, it was clear to us
that there were systemic problems in the way the engineering
work was done to set up the valves, and that was a green
finding. So those are the kinds of issues.
We're comfortable with the SDP. It clearly tells
us what it's supposed to tell us, and that is whether or not
that one specific finding is of risk-significance, given the
other situations that occurred at exactly the same time.
DR. APOSTOLAKIS: So what you're saying is that
the actual finding may be limited to one or two components,
when, in fact, there is suspicion that there is a common
cause failure that might affect many more.
MR. DAPAS: If you have information that there's a
common cause --
DR. APOSTOLAKIS: This is one possibility.
MR. DAPAS: -- that has to be explored as part of
the SDP. You have to have clear information that --
DR. APOSTOLAKIS: But why couldn't you do that for
the MOVs?
MR. DAPAS: This was more of a programmatic
concern.
DR. APOSTOLAKIS: A programmatic common cause
failure.
MR. GROBE: We have a task group right now working
on what we call cross-cutting issues and right now what the
agency considers cross-cutting issues are the effectiveness
of the corrective action program, the effectiveness of human
performance, and the safety conscious work environment,
which is really kind of hard to separate from the
effectiveness of corrective action program.
We've got some concerns in other areas. Being
from the Division of Reactor Safety, engineering is a big
part of my life, and effectiveness of engineering, we think,
is a cross-cutting issue.
We are trying to work through those things and we
will be, over the next year, trying to more clearly define
how you handle cross-cutting issues and this valve issue is
a cross-cutting issue.
DR. APOSTOLAKIS: But let's come back to the
common cause failure. Usually there is a suspicion that
there is potential for common cause failure. Very rarely
you find all valves down. You look at one or two failure
and say, well, gee, this mechanism could have affected the
others.
MR. DAPAS: That's right. If the torque switch
settings weren't set appropriately on valve X, the licensee
should try and determine extent of condition, is that the
case with other valves, and that could be a potential common
mode failure.
DR. APOSTOLAKIS: Then you would go to the SDP?
MR. DAPAS: Correct, if there is sufficient
information to indicate that that is the case. But the
example that Jack was talking about, where the licensee is
trending valve failures and it has programmatic
implications, under the new program, the licensee should be
putting that issue into their corrective action program and
addressing it.
In our annual PI&R inspection, problem
identification and resolution, that might be an issue that's
part of our smart sample, where we would go in and evaluate
did the licensee look at this from a broader context, did
they take appropriate corrective action.
MR. GROBE: On that specific issue, Mike and Laura
-- Mike, you were involved in that, weren't you?
MS. BURGESS: I was.
MR. GROBE: Pardon me? You were?
MS. BURGESS: I was. I sat on the SDP panel and
the SDP panel did not believe that that was a common cause
failure, that that was a cross-cutting issue thing, but that
was not a hardware, there was no evidence that other valves
were exhibiting those kind of failures.
So each individual -- or this valve had to stand
alone and go through the SDP process, which turned out to be
a green.
MR. GROBE: The threshold for a common cause from
engineering issues is very high.
DR. APOSTOLAKIS: So when you say the SDP panel,
is it you or the licensees?
MS. BURGESS: The SDP panel is the NRC. It's one
SRA from every region and a branch chief from every region,
also, plus the program office.
DR. BONACA: Also, if you had a significance
determination for a certain event and found it was not
significant enough, but you have evidence that it would
repeat again, the determination would not -- but then you
would refer back to your corrective action program.
MR. DAPAS: That gets to how robust is your
corrective action program. Each time there is an event, you
have to look at the significance, or each time there is an
issue or equipment problem, you look at the significance of
that associated with unavailability via the significance
determination process.
And if that reflects a repeat occurrence, that
calls into question the licensee's corrective action
program. But, again, degree of regulatory engagement is
based on the overall significance.
For example, you could have repeat issues that are
such low significance, it would be inappropriate for us to
engage. Now, we expect the licensee to address those,
because, of course, the whole premise is the licensee needs
to address those low level issues before they manifest
themselves in more significant concerns or events.
MR. GROBE: I don't want to leave anybody with the
impression that we're not committed to make the program,
because we are. I want to make sure that we help expose the
challenges.
DR. POWERS: Understanding that the team that set
these programs up were under an enormous time pressure, did
a heroic job and did a job under the understanding that
there were going to be rough edges.
I think these are the kinds of rough edges that
are anticipated in this program and getting them all out in
the air early is the only way they're going to get
corrected.
What we're seeing is some resistance to any
changes in programs on the licensee side, which is amazing,
but I think there are things that have to be done better and
managerial and inspector flexibility strike me as you're
really losing something if you take that out of the ballgame
where that judgment component comes in.
I mean, what are we paying these guys to be
educated for if they don't use their judgments?
DR. BONACA: The reason why I was pursuing that
issue before, also, is the fact that on the licensee side,
it's been a common defense for a long time that this issue
happened, but it wasn't of such significance.
And so although it is an important element of the
determination, it's also, at times, a defense and an attempt
to pick more -- there are other links to other events that,
in fact, make it significant because it's a repeat.
So I'm only saying that the significance
determination process right now doesn't lead you necessarily
to assess significance based on the fact that you have
repeats, and those are very important because then we have
programmatic issues.
MR. DAPAS: Well, if you recall, our enforcement
policy previously had an allowance to address inadequate
corrective action, which there are supposed to be actions to
prevent recurrence.
But in looking at this and taking a step backward,
one of the issues clearly that the industry challenged the
NRC on, and ultimately Congress, was that our regulatory
activities resulted in unnecessary regulatory burden. And I
think, as an agency, we determined the best approach in
trying to establish a uniform baseline to determine
significance is using risk, and we came up with the
significance determination process, and I think that needs
-- there's additional modifications that need to be made to
that.
But I think we concluded that going forward for
initial implementation, that exercising that process and
engaging, as a regulator, when thresholds were crossed, if
that had been sufficiently established to ensure that plants
are being operated safely while we continue to refine and
further exercise that.
MR. GROBE: Any other questions on number three?
Because that seemed to be a big focus of --
DR. APOSTOLAKIS: Well, I remember when we had a
presentation on the significance determination process. It
seemed to me there was a lot of room for judgment there and
that's why I asked the question whether you are comfortable
with it.
Given a certain finding, is it a routine matter to
determine its risk significance or people are still learning
how to do that is understandable.
MR. GROBE: The level one and level two reviews
should be -- the staff should be capable of doing those.
Our risk analysts had primarily gotten involved at the level
two as we're learning and their workload has just been huge
to try to help the staff learn how to use these tools.
When you get to level three, and our risk analysts
are engaging with the licensees' risk analysts, you get, I
think, a very highly defensible risk position. It takes a
lot of effort to get there, several months worth of work has
been our experience.
But the tools are still in the stage of
development and as Mike and Sonia indicated, the level two
worksheets, we just have pencil markups on them right now.
But the tools should be effective and there is going to be a
growing period where the staff learns how to use them.
But those tools -- do you want to comment on the
adequacy?
MR. PARKER: Yes. I guess I'd say I agree with
you, George. There is a lot of latitude there and we need
to make sure that we apply the appropriate assumptions and
that we can validate them and support them.
But in a lot of cases, it's very positive for the inspector
because an inspector can sit back and say, hey, I've got an
issue, I'm going to assume this equipment is out of service,
and still results in an insignificant issue from risk, and
he can move on without putting more resources into it based
on that bounding assumptions.
So it could help the inspector out to move on,
where, in the past, we might have pursued an issue to the
end.
Now he can step back and say, hey, this is not
risk significant, the licensee is addressing it, and he can
move on to other issues.
But Sonia and I would work with inspectors, if they have an
issue they believe, with some of their conservative
assumptions, is going to come out to be potentially risk
significant, then we'll try to make sure that we can
validate those assumptions.
DR. APOSTOLAKIS: It appears, then, from your
answer, that item number two would be affected, as well, in
that you haven't really lost all the flexibility that you
thought you had lost.
MR. DAPAS: Let me comment on that. This gets
back to Dr. Powers' point about inspector flexibility. One
of the things that, in Region III, we have attempted to
communicate to the resident staff, as well as the regional
inspectors and DRS, is that we've put people out in the
field that we think have mature judgment, have experience,
and if an issue that the licensee identifies or that we
identify doesn't comport with your internal risk meter, you
think there are issues there, we should ask those questions.
And as you screen that through the SDP and you
look at the different assumptions, to understand why or why
that is not a risk significant issue, and that's feedback
that we would provide to the program office, if we think
that the SDP should have an allowance to ensure that this
issue screens out.
And that's got to be well supported, but that's
where, in my view, the inspector judgment is brought to the
table and says I think this is reflective of the licensee
performance and I think we ought to have a way in our
process to capture that.
Now, that might be in the context of a
cross-cutting issue, that might manifest itself in a change
to the SDP, but it gets back to we continue to refine this
and we look at lessons learned, is there a particular issue
that may be screened out as green that subsequently does
manifest itself as a problem before you see a performance
indicator threshold change.
We need to go in and look at that and say does
that mean that the SDP needs to be modified. So I look at
it as a continuing work in progress.
DR. APOSTOLAKIS: Now, this is done here, right?
The SDP.
MR. PARKER: The phase one and phase two would be
done at the sites or with the regional inspectors or with
the resident inspectors and if it screens out to be
potentially risk significant as far as the colors go, then
Sonia and I would be involved with those activities at that
time.
But we might be working with inspectors up front
because they have some questions or difficulty.
DR. SIEBER: We had heard testimony a couple
months ago about an incident at a plant, not in Region III,
where the significance determination process was used by the
staff and it screened green. On the other hand, there were
two orders of magnitude difference between the staff's
opinion of risk and the licensee's opinion of risk.
Are you prepared somehow or other to deal with a
contest like that?
MS. BURGESS: The agency is part of the process of
validating the SDPs. We've done the first phase, where
we've actually sat down with the licensee and looked do we
have the right mitigating systems down, have we implemented
everything in your updated PRA.
The second portion of the validation is we would
be going to the site with scenarios of a green-white
threshold, something that would be -- an issue that would
put it in a white issue, a potential risk significant issue,
and we will have the licensee run it through their risk
program, computer program, to see if they get the same
answer.
We will also be looking for things that trip the
green-white threshold from the licensee's computer program
and then use our SDP to say are we getting a green-white
threshold or are we still in the green and if we are in the
green, yes, we do have a problem, we have non-conservative,
and that's what we're trying to avoid.
DR. SIEBER: I think part of the problem there was
not so much is the model correct or the process correct, but
how the model was applied to this particular instance.
MR. PARKER: That's possible and that's what I
think the new process makes -- makes it a little bit more
comfortable, that we're supposed to be entertaining and
having dialogue with the utility more sooner than we would
in the past, where we would -- on a potential phase two, the
residents, the senior reactor analysts will be talking with
the PRA organization to try to understand how they've
modeled it, they have more sophisticated models, and what
did we miss or what perspective didn't we consider or that
we might have inappropriately considered.
So we're trying to have that before we get to any
escalated activities in those areas.
DR. SIEBER: Have you and the industry agreed on a
set of rules as to how these things will be modeled or is
this a case by case basis?
MR. DAPAS: Again, the SDP, I think, to answer
your question, is a tool that the agency is using to
determine the significance of findings, and we want that to
be sufficiently conservative that we don't screen out
something that has risk significance.
My experience with the pilot program and listening
to discussions with sites and other regions involved in the
pilot program is we concluded that an issue, say, was of
white significance based on our application of the
significance determination process. The licensee brought
more detailed risk information to the table, with maybe a
more sophisticated model, with different assumptions, where
they had concluded it's not that significant.
So I've seen more examples of that versus --
DR. SIEBER: This is the one I cited as an example
of that and I see that coming to a contest someplace down
along the line if you get into civil penalty areas.
MR. DAPAS: But before we go there, before the
agency is going to make a final risk determination, we
afford the licensee an opportunity to engage us and explain
here's the results of our analysis, and that's where the
senior reactor analyst gets involved in phase three.
It essentially affords the licensee an opportunity
to bring their risk expertise and assessment to the table
and we would consider that. But ultimately we would have
responsibility for rendering a decision on the significance
and then take appropriate action, per the action matrix,
which, again, be it a white issue or yellow issue, doesn't
get into civil penalties. It gets into is it a cited or
non-cited violation, if it's a regulatory requirement.
MR. PARKER: I think the burden is on us right
now, though, and we need to be very careful in using SDP.
As Sonia pointed out, we haven't validated it yet with the
licensees. So it's a licensee -- if we have differing
results, we need to step back and look at the reasonableness
of theirs and why we have that discrepancy and make sure
we're working with the program office and experts and the
practitioners back in headquarters.
DR. SIEBER: There is some uncertainty, which
could be quite large, going into all these things. The
question is, is it really different or is the uncertainty so
large that they actually overlap. That's the problem you'll
have to deal with.
MR. DAPAS: And I think that's one of the most
important aspects when the licensee brings their risk
assessment to the table, is understanding the bounds of
uncertainty and that gets back to the assumptions; that any
risk conclusion is a function of the assumptions and that's
something I think we wrestle with is the uncertainty.
DR. SIEBER: I see that as a challenge.
MR. DAPAS: Right.
MR. GROBE: I went to get back to the flexibility
question, because I think that's critical to the ability of
our programs to be predictive, and they're no longer
predictive, and I'll use a case study, one that I'm familiar
with, D.C. Cook.
D.C. Cook would have been green and for years they
would have been green. Yet, we were never comfortable with
their performance and particularly in the engineering area,
and we applied a number of -- and this also gets to, I
think, your question on lessons learned.
We applied a number of special inspections over a
period of three to four years, including an operations
safety team inspection, what we called a system operations
performance inspection, which had an engineering emphasis,
and then we re-allocated one of our architect engineering
inspections to Cook, because we still weren't comfortable.
And it wasn't until we did that that we found the
issues. Those wouldn't have been found and they wouldn't
have been revealed, I don't believe, through our PIs, at
least looking back in history.
There was a number of risk significant issues that
were found after the plant shut down. This is some of the
soul-searching we did and it was emphasized by Chairman
Jackson at the time that we do this.
And we did two things, the lessons learned
specifically on our inspection programs in the area of
surveillance, because we didn't find the problems with the
ice condensers at Cook, and it had to do with the way in
which we were doing some surveillance testing activities.
But more importantly, from a programmatic point of
view, we looked at how we were looking at engineering and
that really resulted in a safety system design inspection.
We did not have as strong a design engineering
emphasis in our program as we do today under the new
program.
So hopefully that new design engineering emphasis
will help us reveal problems like Cook that we didn't find,
and didn't find until we did the architect engineering
inspection.
MR. DAPAS: Just to clarify, we did do a
feasibility study that looked at the inspection issues at
Cook and what would that result in terms of the action
matrix, but as Jack said, taking that back one step, would
the baseline program have resulted in the identification of
those issues in order to assess the significance, and I
think that's, as he pointed out, the genesis of a more
comprehensive look at design via the safety system design
inspection, because there is the recognition that
performance indicators don't provide you the information you
need to really get a good assessment of engineering
performance.
DR. SIEBER: Now, one of the industry initiatives
is to change 303, I guess, so that you can change modes with
something inoperable. And if you had an incident at a plant
or a condition that's screens green and the licensee shut
down, you now would have lost another tool to keep them down
until they fixed everything, before they start up again.
What would you do in that instance?
MR. DAPAS: I'm not sure I fully follow the
question.
MR. GROBE: Right now, if the licensee finds
themselves in a situation where their technical
specifications cause them to do something that is
unnecessary, we have a process for dealing with that, the
enforcement discretion process, and risk is a big
contributor to that decision-making.
I'm not aware of this initiative to do away with
303.
CHAIRMAN BARTON: It's 304.
MR. CALDWELL: But that would require a change to
the tech specs. I mean, if the agency decided to allow them
to change modes without certain pieces of equipment, then
you're right, we would not have a dog in that fight. We
wouldn't be able to restrict them from starting up because
of that particular component.
But as far as I know, that hasn't occurred yet.
DR. SIEBER: I'm thinking about where we should be
coming from as this issue matures.
MR. DAPAS: The tech specs, as I understand, are
to prescribe which equipment is -- whose operation is
important to assure you can respond to any kind of transient
or impact on the plant.
So if equipment is included in tech specs, the
operability of that is --
DR. SIEBER: Where it is now is where it would be.
MR. DAPAS: Right.
MR. GROBE: Philosophically, it should be
risk-informed, right?
MR. DAPAS: Right.
MR. GROBE: In which case, mode changes with risk
significant equipment out of service shouldn't be committed.
MR. CALDWELL: I guess the big concern here would
be if we did it generically. I think each plant would have
to say they're -- not get rid of 304, but to actually pick
out the components they think are no longer required for
specific modes and then you would have to do a risk analysis
for each of those components.
And if the agency were to agree, if the industry
came in with a proposal that we shouldn't have mode
restrictions based on equipment, then that would be a big
concern, because you wouldn't have analyzed each component
to see if it had a risk significance.
DR. SIEBER: The problem there is that most of
those occur between the mode four and the mode three.
MR. CALDWELL: Right.
DR. SIEBER: Which there's not very many PRAs out
there for that. So what do you use for the tool?
MR. GROBE: It's an interesting question, because
most of the safety systems are required at mode four and yet
they're not necessary to mitigate an accident at that mode.
MR. CALDWELL: But they -- you're right. It would
be a philosophical discussion, because it is now a tool and
a lever to make sure the plant is completely back in
operation prior to changing modes.
If you allowed folks to wait until the exact time
when he component was needed, then you're running up against
clocks and some people would put it off to the last minute
and others wouldn't.
Right now it works pretty good because licensees
know, in their outage, that in order to come out of the
outage, they have to have everything back and working.
DR. SIEBER: Right. There's no way out.
MR. CALDWELL: It's been, I believe, successful in
terms of plants are operating better coming out of outages
now than they had in the past.
DR. SIEBER: I agree.
MR. DAPAS: Moving on to, I guess, insight number
four that we offer regarding the new program compared to the
old program. The old program involved more direct
observation of plant activities. Under the new program,
there is an increased emphasis on inspection preparation and
office review, with, of course, the exception of testing,
where we do continue to have a number of direct
observations.
I'll give you an example, like maintenance. Under
the old program, we might observe the maintenance activity,
like a pump rebuild, was the work procedure sufficiently
comprehensive, are the steps being followed, et cetera.
Under the new program, we focus on has the
licensee conducted a risk assessment for that particular,
say, on-line maintenance activity. We would evaluate the
effectiveness of that risk assessment and licensee control
of the maintenance activity.
And I thought Laura Collins, who actually has been
an inspector under both the old program and then involved in
the pilot program, could maybe give another example in terms
of the maintenance rule, because I know there were some
questions that.
MS. COLLINS: We actually have two procedures that
we look at maintenance. We have one that is called
maintenance rule implementation and we have one I will talk
about later, which is sort of our evaluation of their
on-line risk assessments.
Under the maintenance rule one, which is the
resident inspectors' largest number of samples and largest
number of hours, that is largely a review of equipment
problems that they have had and how they've dealt with them
under the maintenance rule, and that's quite a bit different
from our previous maintenance observation kind of inspection
that Mark talked about.
So to me, that's a big distinct difference right
there in the area of maintenance.
The other one is the area of operations, which we
largely reviewed routine operations. Now we focus more on
non-routine evolutions and don't look so much at the routine
operations.
So those are just two examples of how we're not
directly reviewing routine activities in the field.
DR. SIEBER: And that means much less observation
of activities and more going through papers.
MR. DAPAS: The focus has shifted a little. It's
understanding the licensee's evaluation of risk associated
with that activity, their control of that particular
activity.
Inspection preparation, the inspectors need to
understand the risk importance of a particular structure,
system or component, or evolution that's being selected for
the sample, and that's where there may be more preparation
involved in saying, okay, here is a specific testing
evolution I'm going to observe because it's important from a
risk standpoint, and then the preparation involved with
going out and reviewing that activity.
But where that presents a challenge, that I'll
talk about a little later, is the licensee may be planning
to do a surveillance test tomorrow evening. The resident
inspector spends time getting ready to observe that and then
it's deferred and the inspector was planning to do another
activity on Thursday of that week.
And we selected that specific surveillance test
because it's more risk significant, where, under the old
program, you could just pick another surveillance test and
observe that.
The risk importance was less of an issue, and
that's where it impacts inspection planning and resource
utilization.
MR. GROBE: We're getting way behind schedule. I
wanted to make one more observation regarding observation of
activities. In addition to some of the resident issues, in
the plant support area, EPHP and safeguards, it's had a very
significant impact.
You can do the new safeguards inspection program
from the guard shack. You don't even have to go into the
plant. In the area of health physics, much fewer activities
being observed in the plants as far as how they're
controlling the activities from a radiological protection
point of view.
In the EP area, during the programmatic inspection
it doesn't require you to go into any of the emergency
planning facilities. So you don't actually observe whether
the facilities are in a state of readiness.
A lot of these are compensated for through the
PIs, the performance indicators, but in some cases, not very
well.
So there has been a shift from reviewing
activities that have already occurred through looking at the
paperwork to -- and away from direct observation in the
plant.
DR. SIEBER: How do you feel about that?
MR. GROBE: Our inspectors are not as comfortable
with that as they were in the past.
DR. WALLIS: I'm wondering of the public would be
as comfortable with that.
MR. GROBE: It's a new program and it's dependent
on multiple prongs. One of those prongs is performance
indicators and another one is effectiveness of the
licensee's corrective action system. So we're putting our
eggs in different baskets and we need to see how it works.
MR. DAPAS: But, also, when you look at the
particular inspection procedure, there's associated
objectives which are supposed to result in our acquiring the
information we need, and that can be arrived at via direct
observation or review of, for example, the licensee's
control of the maintenance evolution.
The key is do you obtain the information you need
to make an informed judgment, from my perspective.
MR. CALDWELL: There is an ongoing feedback
process. These particular issues that Jack talked about are
issues that we've fed back to the program office and will
continue to feed back.
So I expect to see some changes to the program
after the first year of implementation. So maybe a year
from now, we can talk about it again and see where we come
out on this. These are just early observations.
DR. SIEBER: Have you made your thoughts known to
the headquarters?
MR. CALDWELL: Certainly.
MR. GROBE: We do that and we've been rather
proactive I that regard. I think we've pretty much covered
item number five. Why don't we go on to item six.
MR. DAPAS: Regarding inspection resources, as
we've touched upon, there was more flexibility under the old
program, in a couple aspects.
In addition to the inspection scope, where we
talked about how prescriptive that can be under the new
program, we had more opportunity with use of regional
initiative, we had N+1 inspector, where you could use that
particular inspector to conduct some regional initiative in
the area of operations. There was more flexibility with
tapping DRS engineering resources to go out and do some
regional initiative inspection.
Now, under the new program, that DRS resource and
that former N+1 resource, which now may be assigned to the
region, is fully encumbered by the new program. So there's
less flexibility in that regard, which, of course, again,
was by design with the new program and the inspection scope.
But when you have extended absences or vacancies,
that requires back-filling the complete program, and so that
results in a greater degree of sophistication in inspection
program management. The branch chiefs out in the audience
can tell you that they have to plan hours in detail for,
say, a six-week inspection period so they can readily
identify where there are holes and you can't -- you can only
defer some inspection to a limited degree, because that
creates the bow-way that you're going to have address during
the next inspection period.
And when you have sample size ramifications, the
number of activities that you need to look at per month,
that's where that becomes an issue.
So we have to have contingency plans in place if
we're going to support a rotational assignment to another
program office or we've got a vacancy at a particular site
because the individual left for a promotional opportunity or
reassignment to the region.
In order to implement the new program, we've got
to have two fully engaged people at the site. There is some
flexibility there, but not a lot. Frequently, you will hear
a branch comment that I need some help during this time
period because inspector X is going to be involved in this
activity, and it causes is to continually focus on what are
our priorities and what we can support, because we don't
have the latitude right now of saying that we have completed
the baseline program with this amount of inspection, like
you could under the old program with the core inspection
hours.
MR. GROBE: I think as far as public awareness, we
are greatly aware that the public is taking opportunity,
taking advantage of the web site information that's
available to them. The PIs are on the web. Our inspection
reports are in the web, and that is a significant
improvement over the --
DR. WALLIS: It's on the web. Do you have a way
of counting how many people -- how many times it's actually
looked at?
MR. GROBE: Actually, Augie Specter counts it and
reports on it regularly, in thousands of hits. I can't
remember what the numbers are.
DR. WALLIS: They actually stay with it. They
don't just hit and leave.
MR. GROBE: The question I got is how many of
those were Augie logging on. But he's counting those. And
I headed a public meeting out at Cook, a lady who called
herself Auntie Nuke, who had downloaded a lot of stuff off
the web. So the public is taking advantage of it.
DR. POWERS: One of the things I find -- items
that show up that say, in effect, management is very well
prepared for the safety review, to be as helpful for me to
understand the plant as those that say the operators didn't
handle the jumper control very well.
The upside and the downside are very valuable to
me. Now it sounds like the upside is going to be
disappearing.
MR. GROBE: No, it's gone.
DR. POWERS: It's gone. And somehow I worry about
the communication aspect, to me and everybody else.
MR. GROBE: We all shared your concerns, but it
was the view of the industry that that's what they wanted
from the standpoint of communication in our inspection
report, and, by definition, that's what goes into the PIM
and goes onto the web.
MR. CALDWELL: Well, our observations and findings
that go into the PIM are supposed to be risk-informed and
it's very difficult to risk-inform the positive. So you
wouldn't be able to do what you might like to do, and that's
come up with a balance. But a positive comment would weigh
as heavily as a yellow or a white finding, in which case a
positive comment may have little or no safety significance.
There is no way to evaluate that.
So the decision was made to just --
DR. POWERS: Philosophically, George, I think he's
hit upon a flaw in this PRA technology.
DR. APOSTOLAKIS: No, it has not been used.
DR. POWERS: It only gives us good ways to
quantify the negative and no good ways to quantify the
positive.
DR. APOSTOLAKIS: That's what we have done so far,
but one can actually say that because they're doing such and
such, the human error probabilities that were assumed in the
past are actually lower, so there's a positive impact on
plant safety, or that the failure rates are expected to be
on the lower side.
DR. POWERS: Your problem is one of communication,
George.
DR. APOSTOLAKIS: Why?
DR. POWERS: That I can understand, well, a number
going from three to four, as in
times-ten-to-the-minus-fifth.
DR. APOSTOLAKIS: But not from three to two?
DR. POWERS: But the other way, the positive -- I
mean, how do I understand going from 99 to 99.9?
DR. APOSTOLAKIS: It's just that we've never used
it that way.
DR. POWERS: That's right.
MS. BURGESS: But I think you can understand that
if a licensee puts -- adds another diesel, then I think
everyone can understand they have decreased their risk. So
those kinds of things can be put into our report.
DR. POWERS: He tells me all the time that I can't
assume they've decreased their risk.
DR. APOSTOLAKIS: I think that's a good point, but
we can say something. The thing is we've never attempted to
say how improving things, if we're finding the good things.
I wanted to say something, but Dr. Powers destroyed my
thinking.
DR. POWERS: I've been successful again today.
CHAIRMAN BARTON: Yes. Before this deteriorates
further, do you want to continue?
DR. APOSTOLAKIS: He probably can't even remember.
If everything is green, that is a message, right?
DR. POWERS: I insist that that's a degraded
message.
DR. APOSTOLAKIS: And that's why people are trying
to --
DR. POWERS: When everything is green, then you
start looking at what are the shades of green and you see
these things where guys plot where they lie on the green
band and people start paying attention to that and not
paying attention to the fact that it's green.
MR. GROBE: What's interesting is green is not
good. A green finding is a finding. If you have 100 green
findings, that's not better than having one green finding,
that's worse, because that might be indicative of a systemic
problem.
And the colorization, I have a lot of problems
with these colors.
DR. SEALE: Amen.
DR. APOSTOLAKIS: So the ideal is no findings.
MR. GROBE: Well, no. If we have no findings, my
concern would be that the inspection program is not
functioning effectively.
MR. CALDWELL: The idea should be that we're an
active regulatory body, we're inspecting, we're having
findings. The findings are not such that it's outside of
the industry response band, which means it's staying within
a band that we're allowing them to correct their problems.
That is a plus or minus, however you want to look
at it. If they drop out of that band, then people can ask
questions about their safety.
DR. APOSTOLAKIS: But this raises, again, an issue
that is a favorite of mine. I've raised it several times,
but I don't know that I got a response.
CHAIRMAN BARTON: So you're going to try again
anyhow.
DR. APOSTOLAKIS: Yes. What is the purpose of
these inspections? I mean, there are two alternatives, in
my mind. One is to make sure that the risk profile of the
plant, as we're understanding through the IPE and PRA,
remains the same, especially hasn't shifted upwards. So
that's a plant-specific finding or determination.
The other is to look at it as one unit in the
population of 103 units and see whether you are -- I mean,
that particular unit is within the industry norm or it's a
percentile. These are two very different things.
And the third one, I guess, is to make sure that
the licensing basis is still met, which is not -- it is
related to the risk profile, but it's not the same thing.
And I'm not sure that the designers of this
process really articulated very well what their objective
was. In some instances, I get answers that, yeah, it's
industry-wide, we're very interested in what's happening, is
this an outlier or not. In other cases, no, we really want
this plant to remain the way it was risk-wise.
So what, in your opinion, is the objective of all
of this? I mean, we have a risk profile, we have in the
PRA, you do all these determinations such as PIs and the
action matrix and so on, because that's related to the green
now, because if everything is green and I can conclude that
the risk profile has not changed, then things should be all
right.
Because then I get into the business of how many
greens do I have, how many findings, one versus 100.
MR. GROBE: Possibly. I wouldn't suggest you
count findings, but what's important is to understand the
root cause of the findings and what that root cause can do
to the risk profile.
DR. APOSTOLAKIS: So the potential for getting out
of the green.
MR. GROBE: Exactly.
DR. APOSTOLAKIS: That's what you worry about.
MR. GROBE: Exactly.
DR. APOSTOLAKIS: But have you any idea as to what
the intent of the oversight process is?
MR. DAPAS: Both aspects are addressed. When you
have a particular inspection finding, that's got to be
placed in the appropriate context of a given plant
configuration. You have to bring plant-specific PRA
knowledge to bear.
I think the performance indicators address that
across the industry, where if we set a threshold for number
of scrams that would result in regulatory engagement, that
threshold is a function of overall industry performance.
DR. APOSTOLAKIS: And it shouldn't be, in my view.
MR. DAPAS: That may be a few, but that's at least
my understanding of the intent of the program.
DR. APOSTOLAKIS: The inspection findings are
plant-specific, but the PRAs are --
MR. DAPAS: Well, the PI is plant-specific, if you
will, in terms of you had scram X, you had transient X, but
the threshold --
DR. APOSTOLAKIS: It's an industry --
MR. GROBE: And the same thing with inspection
findings in the SDP. The base risk profile of a plant might
be five-ten-to-the-minus-five, it might be
one-ten-to-the-minus-seven, but the threshold for a green
finding is ten-to-the-minus-six, no matter what the base PRA
of that plant is.
DR. APOSTOLAKIS: But, you see, the fact that the
thresholds are so high has made the utilities themselves to
have more stringent plant-specific thresholds for internal
use.
MR. DAPAS: Right. And the reason for that is
because we told the industry they shouldn't be using our PIs
to manage their plant. I would expect them to have more
restrictive, if you will, indicators so that they can
address problems before it does cross the threshold.
MR. CALDWELL: That goes back to what Marc had
mentioned earlier. The basis of this program is an
effective problem identification and corrective action
program on the part of the licensee. So they have to have
in place their performance indicators or whatever they think
is necessary to identify their problems early and resolve
them before they become bigger issues.
That is what we are relying on. We have to see
now if that works or not by implementing this program and
see how well the licensees' corrective action programs --
how effective they are.
DR. BONACA: But you said before that D.C. Cook
would have been all green.
MR. GROBE: It was all green.
DR. BONACA: So there would have been no signal
coming from the indicators for action. So does it mean that
the action at D.C. Cook was successive or does it mean that
the indicators really have been a big help?
MR. CALDWELL: I missed that conversation. I
think Jack is saying the performance indicators may have
been all green. I'm not sure our inspection findings would
have been all green. Our inspection findings likely would
have been something other than green.
DR. BONACA: So you didn't check for that.
MR. GROBE: No, we did. We ran all the LERs and
findings prior to the outage through the -- at that time, it
was a very preliminary draft SDP, and didn't come up with
any significant findings.
I don't know if we came up with any whites, but it
wasn't until after the outage that you started seeing
yellows and reds.
The point I was trying to make was that the level
of resource expenditure that we put into Cook, we would not
be able to do today. And somebody earlier mentioned that
the program is more indicative than predictive, and that's
true. We have less capability of being predictive, unless
the thresholds are crossed with a specific finding.
MR. DAPAS: And that gets back to, if you recall,
our discussion with the Commission. One of the fundamental
premises that the industry proposes is that performance
indicators would be crossed, threshold changes before there
is a significant programmatic concern that manifests itself.
Right now, I think there are some differing
schools of thought and that's why the role of cross-cutting
issues, I think, has played such -- the importance of that
has been elevated.
There is a task force that's looking at human
performance and corrective action programs and safety
conscious work environment, cross-cutting issues, because
not everyone full ascribes to this tenet that you will see
performance decline clearly manifested in the PIs before you
see risk significant inspection findings.
DR. POWERS: The committee has advised the
Commission that we consider that an assumption that needs to
be validated. You're only reinforcing that opinion.
MR. GROBE: The lunchroom across the way gets busy
at around noon.
MR. CALDWELL: What we're doing is we're having --
they're bringing over sandwiches and some salads.
CHAIRMAN BARTON: We'll just keep going then.
MR. CALDWELL: So I can let you know, it's $10 a
person, and we should be bringing -- we'll bring a table in
right behind here and you can go over and pick up and eat as
you wish.
CHAIRMAN BARTON: Excellent. I'd like to get
through the SRA process before lunch, then we can take a
break, if we can get to it.
MR. DAPAS: I've just got one point left to make
on the public awareness. I think clearly there has been a
public outreach effort associated with the new program,
industry workshops, et cetera, which I think is a positive
initiative.
We have touched upon the DRP --
DR. WALLIS: Well, public outreach, how broad is
the public that gets involved? Public outreach, how broad
is the public involved?
MR. DAPAS: We've invited, like, for example, when
we've conducted meetings on the new program and we're going
forward with meetings at each of the sites within six months
of initial implementation. Certain officials, et cetera,
we're inviting, but it varies, the degree of public
attendance.
We're trying to advertise that via web and other
communication forums, but it does vary.
MR. GROBE: We don't see a lot of public awareness
-- public involvement.
MR. DAPAS: It depends on the site.
DR. WALLIS: Public should not consist only of
people with some personal interest, like an economic
viability of their plant.
MR. DAPAS: Right. Right.
MR. CALDWELL: It's strictly -- I think it's
strictly related to how interested the surrounding area is
in that plant and most of our plants do not have active
public involvement. So when we have these meetings, they
are not widely attended.
But we do put out a lot of announcements to that
effect and people could attend, if they wanted. And I
suspect if there was an interest, like one of our
facilities, Prairie Island, there's an interest in dry cask
storage. And so we always get a pretty good attendance at
those. But it's really related to how well the public - I
look at it this way. If you don't get a lot of public
attendance, that means that they feel comfortable with that
plant as it is.
Otherwise, they would be coming to the meeting to
try to understand or express their views.
MR. DAPAS: My comment was more in the context of
the old program, where really the only public outreach, I
would offer, was a meeting to discuss SALP results, versus a
more concerted effort.
I've touched upon some of the DRP challenges here.
One of the challenges we face, of course, is feedback and
dissemination of lessons learned on the new program as we
attempt to further revise that, and there's a number of
forums for doing that.
We've got feedback forms, weekly conference calls
with the program office, inspector seminars, and then, of
course, DRP/DRS counterpart meetings, where Jack and Mike
and Geoff Grant attend to discuss some issues with the new
program.
DR. WALLIS: One measure of success might be that
there were lessons learned which were useful when you
actually look back at it.
MR. DAPAS: Right. Which gets into the
self-assessment area. We have been given an opportunity to
weigh in and comment on the self-assessment plan
development, which includes appropriate metrics, and this is
in support of the IOU we have to the Commission to evaluate
the new program and report to the Commission in June.
And headquarters is currently involved in our
inspection report review to help ensure consistency and we
do plan public workshops to obtain feedback, which was
fairly well received in the pilot program.
Unless there are any questions, that pretty much
summarizes DRP's involvement in the new and old programs.
MR. GROBE: Let me just highlight one challenge
that we're going to be talking about a little more later, I
hope, in the Division of Reactor Safety. There's a number
listed here, but the one that's most significant for us is a
change in required expertise.
We depended heavily on contract resources when we
needed design expertise in the past. We no longer have the
financial resources to procure contract resources in that
area.
So that's a challenge for us. It's a staffing
challenge. It's a recruiting challenge, and we're trying to
meet that and we'll get into some more detail later.
The other issue is risk analysis capability and
why don't we just go right into the risk presentation that
Sonia has prepared.
MS. BURGESS: Here's a little background. In
October of 1995, the SRA position was developed to assist
the agency in transitioning to a new risk-informed arena in
the way we do business.
I don't believe that in 1995 the Commission
realized what a large leap we were going to make ultimately
into getting our whole process into the risk-informed arena.
Fortunately, when the transition, the pilots, the
new reactor oversight pilot program started, the SRA program
was fully staffed in all of the regions and we were fully
trained and qualified and certified.
I think that has been a big asset in the success
we have had in implementing the new reactor oversight
process.
Some of the bullets highlighted here are just some
of the key things that we do here in the region. Our
biggest role right now is to support the new oversight
program.
We were very much involved in the development and
the implementation of a pilot process here in the region and
we sat on a lot of committees, helped in reviewing many
procedures, things of that nature.
Our main support now is in the SDP arena. As has
been brought up, Mike and I have visited every site in our
region, because we think it's imperative that these SDP
tools that we have been giving to the inspectors are
accurate, that the licensee agrees that they're accurate,
and that they are -- although simplified, they are the best
tool that we have produced to date.
DR. POWERS: The question that often comes up, to
my mind, is the scenarios they have are very simplified.
Are they simplified by intent or out of necessity?
MS. BURGESS: The scenarios on the SDP worksheets,
like the loss of off-site power?
DR. POWERS: Right.
MS. BURGESS: I think, yes, they're definitely
simplified out of necessity. We certainly do not have the
resources of the capability to model 50 initiating events
and that's typical of a licensee's own PRA analysis. So we
have narrowed it down to probably ten to 12 initiating
events. Has there ever been a demonstration that -- with
some rigor -- that narrowing it down to these ten or 11
events constituted an adequate description of the risk
profile of the plant?
DR. POWERS: Yes. And in our site visits, along
with the other regions, these scenarios, these initiating
event scenarios have captured the majority of the risk
contribution from their PRAs.
MR. PARKER: I would also add that we started out
with, I think, four to six initiators and we did some pilot
activities with the program office. One of them was one of
our plants in the region.
We went there and tried to do some V&V by taking
some scenarios, some major systems and correlating it with
the licensee's PRA and we found some non-conservative in
ours, where the licensee identified it as a fairly high risk
activity.
And that's where we had to step back, as an
agency, and I think it set us back several months, trying to
identify additional initiators that were necessary to truly
capture the majority of the risks, as Sonia says, that we
are right now, that we were able to pick that up.
DR. POWERS: I might be willing to concede they
captured the CDF. The question is, did they capture the
risk.
MR. PARKER: That's some of the -- I mean, right
now, what we're looking at is internal events and some of
the difficulty we have in using the tool is we don't have an
effective took in place for containment, for shutdown, for
external events. So there's a lot of -- the majority of the
risk is still being captured through screening tools that
we're trying to put in place right now and when we have
those type of issues, that Sonia and I have to get involved
with it, we have to get involved with the licensee's IPEEE,
and we have to work with headquarters in a lot of cases if
it involves external events, it's just a screening basis in
IPEEE.
So we might not be able to capture all that
ourselves.
MR. DAPAS: A good example of that is a recent
issue we had at Quad Cities with -- what is it, Marc -- safe
shutdown makeup pump and that being unavailable and how you
bring the external event fire risk into play. There's not a
tool used. We used risk achievement worth, I think, and CDF
to come up with an overall risk assessment.
We discussed it as part of the significance
determination panel. We communicated that to the licensee
as the most appropriate tool we have right now and then the
licensee is going to come to the table with their assessment
of the risk impact in terms of fire risk.
DR. POWERS: So you don't even have things like
five available to you.
MR. PARKER: No.
MS. BURGESS: No.
DR. POWERS: One of the -- an anecdote, to which
I've never had a resolution, is I believe it's Brown's Ferry
that uses ORAM for outage management and they were showing
me how it worked. I know a little bit about ORAM.
And they said, well, look for this particular
outage, we set up a configuration that had this red region
and by manipulating things around, we were able to change
the way we did our outage, so that instead of having a red
region and everything else green, we had two orange regions
and everything else green.
And I have puzzled and puzzled to understand how
one concludes that two oranges is better than one red.
MR. PARKER: That, I think, is some of the
difficulty in ORAM, is it's mainly a deterministic tool and
you're looking at defense-in-depth and most utilities don't
have a probabilistic shutdown model.
I think some of the plants are going there and we might be
able to look at it a little closer, but you pointed out some
of the difficulties we have with our tools. The licensees
are trying to suppress and reduce their overall risk and
from their perspective, they didn't enter a red, which was
prohibited, and it's very subjective and that's
decision-making.
DR. WALLIS: When you compare with the licensee's
PRA, you just compare with the results or you compare with
the details?
MR. PARKER: You're talking about SDP?
DR. WALLIS: Yes.
MR. PARKER: When we're looking at findings?
DR. WALLIS: Looking at your model versus the
licensee's. You have a simplified model. How much of his
PRA do you have access to?
MR. PARKER: We have very little access to most of
the PRAs, but when we did some of our benchmarking, we
wanted to get the cut-sets and the importance from there so
we can extract that and figure out what were the dominant
cut-sets that were affecting our SDP model.
DR. WALLIS: It's a peculiar kind of detective
work, or maybe there are some assumptions made that you
don't know anything about.
MR. PARKER: That's right.
DR. WALLIS: Which is reducing the licensee's
result. Don't you have a way of finding out what they are?
MS. BURGESS: Only if there is an issue or a
finding in that. I mean, we don't have a PRA inspection.
MR. PARKER: I think you're stepping back to what
I would call the infrastructure. We still haven't even
established a PRA certification. But on the other hand, we
are basing our SDP as closely as we can to the licensee's
IPE or their updated PRA model, and we haven't validated
that model yet.
So I understand and appreciate your comment and I
think the agency is pursuing that, but, again, we're
progressing slowly. Maybe there's different things we need
to prioritize in this arena, too.
MR. DAPAS: There is a conceptual issue here,
though. I think we -- if a piece of equipment is failed or
unavailable, we run that through the SDP, we communication
the results of that, then the licensee can bring to the
table more risk-specific information from their PRA.
Now, obviously, when we've got an issue and we're
running it through the SDP, the licensee is doing the same
thing, because they understand the SDP, we've communicated
to them, 0609 defines specifically what that SDP tool is.
If it looks like this is going to screen out as a
white finding, they're rather proactive in communicating to
us their assumptions and what their PRA model says. So
there is that dialogue.
DR. WALLIS: Assumption is the key word, because
assumption really is not worth anything unless it can be
challenged and defended. And if there is some mysterious
assumption you don't know about, that's like magic. It's
just like getting whatever you want.
MR. DAPAS: We should challenge that.
MR. GROBE: Your point is very good, and that is
that we don't know what the assumptions are in the model.
The IPE that the staff reviewed a number of years ago was
many generations earlier than what is currently being used
at the sites.
So to a large extent, we have to depend upon the
-- that there has been an intelligent evolution of the model
that the licensees use.
DR. BONACA: On the other hand, the event,
whatever you're evaluating, it's a fact. So you know what
you're going to check inside the model. It's not
hypothetical issues.
In general, you may question their assumptions in
the model to represent the --
DR. APOSTOLAKIS: Do we --
DR. BONACA: But now the fact that you have a
specific event happening, it allows you to go back and
verify the assumptions.
MR. GROBE: But they don't have it here. Is that
part of the SDP, the phase three?
MS. BURGESS: Phase three. Phase three will
challenge the licensee's assumptions, where we're different,
and take a look at what their program does, what their
assumptions are, and the validity of those assumptions.
DR. APOSTOLAKIS: What, in your opinion, would be
the ideal tool that should be available to implement a
risk-informed regulatory system, especially the oversight
process? What would you like to have?
MS. BURGESS: Personally, I think that some kind
of standard for a PRA is just essential.
DR. APOSTOLAKIS: But you would also like to have
a plant-specific PRA on the computer.
MR. PARKER: Right now we have safety monitoring
and I guess my perspective is to be able to have access to
the licensee's plant models and be able to manipulate them
and understand them. But we need to start where Sonia says,
that we certify your PRA or have some level of certification
to say this PRA meets certain thresholds and standards.
DR. APOSTOLAKIS: Let's take a specific plant,
like Davis-Besse. What PRA information do you have?
MS. BURGESS: In fact, I was there two weeks ago
to do their SDP worksheets. They have gone through an
extensive PRA update. Prior to my visit, the only thing we
had was what was documented in late 1980s.
DR. APOSTOLAKIS: But do you have --
MS. BURGESS: We have the docketed IPE here, which
is --
DR. APOSTOLAKIS: The PRA as they changed it.
MS. BURGESS: I was able to bring back, from my
visit of two weeks ago, the executive summaries, some of the
system notebooks that are used in the service water systems,
component cooling water. I was able to get risk achievement
worth, a lot of importance measures of systems, things like
that. They give us a better idea of how they have changed
their --
DR. APOSTOLAKIS: I don't understand why they
don't give you the whole PRA.
MR. PARKER: Because we haven't mandated it. It's
not required through the regulations and no utility --
DR. APOSTOLAKIS: The risk achievement worth is
not required either.
MR. PARKER: I understand, but I guess what -- you
said this is our chance. I would like to see us have some
type of requirement or standard where the utilities are
providing us their routine updates, no different than they
would on an FSAR. That's a difficulty we're having right
now with our SDP tool.
The SDP tool was put together by BNL, Brookhaven
National Lab, using the IPE and the SRAs are having to go
out and reevaluate that based on the licensees' current
models. So significant changes are taking place.
DR. APOSTOLAKIS: We have been told by some
licensees that they have -- especially the ones who have
risk monitors -- they have PC versions of their PRA, they
can see the impact of the change within a minute.
MR. GROBE: On-line risk monitor.
DR. APOSTOLAKIS: Sure. Would you like to have
something like that?
MS. BURGESS: Yes. Now, we do have -- like Mike
said, we do have safety monitor. Unfortunately --
MR. PARKER: We have the program.
MS. BURGESS: We have the program and we have the
eight models, which are like the Westinghouse tool for a
Westinghouse four-loop or things like that. We do not have
plant-specific models.
Now, some plants in our region -- as a case in
point, Kewaunee has given Research their program, their
model, and Research has given it to INEL and INEL is in the
process of converting it to SAPHIRE. So we have their
actual model.
DR. APOSTOLAKIS: Now, wouldn't the SPAR models
eventually meet the needs you have when INEL completes --
MR. PARKER: I think there is a potential that it
could meet most of our needs. The difficulty is going to be
they're working on low power shutdown models. They're
working on some containment and those have -- a lot of that
activity has been deferred because of the SDP activities in
progress that we can't -- we weren't able -- there are
competing resources.
So I don't see us getting there for several years.
MR. GROBE: We're significantly resource
constrained.
DR. APOSTOLAKIS: But you mentioned that the
licensee is under no obligation to give you the PRA. But
isn't it in their best interest to do that?
MS. BURGESS: We believe it is.
DR. APOSTOLAKIS: I mean, if they want
risk-informed regulation, we can't do it without risk
information.
MR. GROBE: We've been able to encourage several
licensees, just from an efficiency point of view, of
interacting with the staff, encouraged them to give us some
of their risk analyses.
The problem is, as Sonia and Mike have pointed
out, one, is that there is no standard. So you have widely
differing approaches, and second is there is no requirement
to provide it.
So it's only a phone call from Steve or myself
that says, listen, our interface would be much more
efficient if we had such and such and then we'll get some
documents.
DR. WALLIS: I'm not sure you need the standard.
If I look at thermal hydraulic codes, it used to be that the
staff would simply look at some codes provided by licensees.
But now in reviewing thermal hydraulic code, the staff is
moving to the position we want the code, we want the source
code, we want to be able to run it, we want to be able to
try things with it and see what it does.
MS. BURGESS: Many licensees are very reluctant to
put their updated PRA on the docket.
DR. WALLIS: But ideally that's what it should be.
It should be completely open.
MS. BURGESS: They just do not wish to have it on
the docket.
DR. POWERS: If you can think about the headaches
it would involve when it's updated, it's a significant
process.
Let me ask you. You've mentioned this need for
certification a lot and there is an activity going on with
the standards committee to set the standard for the PRA, and
I think NRC has a limited voice in that committee setting
that up.
Do you have a voice with those representatives on
that committee?
MS. BURGESS: The regions?
DR. POWERS: Yes.
MS. BURGESS: No, we don't have a particular
voice. Research is the member of that committee and I would
characterize their participation as much more than just a
minor committee member.
DR. POWERS: Mary Drouin and her troops.
DR. SEALE: That confirms what we found out from
them last week.
MR. DAPAS: We're not precluded from providing
input there. If Mary Drouin is the representative, I've
worked with Mary, I know Sonia. We'd have no problem
calling her up and saying, hey, we think this needs to be
considered.
So we are not precluded from that opportunity, but
there is not an outreach effort, if you will.
DR. SEALE: You're not getting timely information
on what the status of that -- the evolution of that
so-called certification process.
MR. DAPAS: Nobody else is either. Other than
what I read in the PRA implementation plan updated
Commission paper.
DR. POWERS: It seems to me that -- I think
there's a wealth of information at that tend of the table on
what the minimums ought to look like, just because of the
pain, it's knowledge that's been gained by pain.
I'm wondering if we can't find a mechanism to do a
download so that there is some hope that maybe that gets
represented in the standard, because the last thing you want
to do is get a standard back that's no good to you, that
doesn't standardize the things that you want standardized.
DR. SIEBER: It's harder to undo that kind of a
thing than it is to write it in the first place.
DR. APOSTOLAKIS: Will you have an opportunity to
comment on the ASME standard? I mean, the public is
welcome, so you are welcome, too.
MS. BURGESS: I believe the region will have a --
MR. PARKER: More than likely, Research has been
very accommodating in requesting our resources to comment
and provide feedback to all the new inspection processes and
generally the NUREGs that are coming out, too. So I would
see no difference in this regard.
DR. APOSTOLAKIS: There is a workshop, as you
probably know, on the 27th of this month. Do you plan to
attend?
MS. BURGESS: No.
DR. POWERS: They've got more than they can keep
up with as it is.
MS. BURGESS: Yes. We've been very busy.
MR. DAPAS: But, George, not to convey we don't
think that's an important activity. Like verification and
phase two workshops we think is a high priority, as well, so
that we can ensure we're capturing the licensee
plant-specific information.
So there's competing priorities we're trying to
wrestle with.
DR. APOSTOLAKIS: Is it fair to say that we risk-inform the
regulations with very limited risk information on our part?
MS. BURGESS: Yes.
DR. POWERS: When you look at this risk-informed
regulation, only a third of it is risk-informed. The rest
of it is something.
MR. GROBE: It's all risk-informed, it's to a
degree.
DR. APOSTOLAKIS: It's not quantitative.
DR. POWERS: This is the argument I sometimes make
with the gentleman to my left and say we've always done
risk-informed regulation, we didn't write these regulations
because we didn't think there was any risk there.
DR. APOSTOLAKIS: That's right, and I have been
persuaded, as always when I hear a reasonable argument.
CHAIRMAN BARTON: All right. Where are we here?
DR. APOSTOLAKIS: I think Sonia is telling us --
the last four bullets, we understand that you're doing that.
Do you want to move on to --
MS. BURGESS: One initiative that we actually -- I
did want to make a point, the initiative that we are doing
that we are going to -- we're doing outage risk assessments.
The plant is in an outage, Mike and I will go out to a site,
sit down with the scheduling people of the outage from the
licensee, understand where their risk significant evolutions
are and helping to focus the resident staff on what to look
at out, what to be observant of, what the most risk
significant issues and evolution is.
DR. POWERS: Do you have an understanding of what
the risk significant evolutions are during an outage, can
you tell me?
MS. BURGESS: Quite honestly, I think that our new
inspection procedure for outage work is pretty good on
hitting PWR/BWR risk significant evolutions, from a broad
perspective, to give, I think, excellent guidance to the
resident staff.
DR. POWERS: I'll look at it.
MR. GROBE: What we found is that the licensee's
risk analysts aren't getting involved early enough in
looking at the outage plan. We have been prepared to go out
and look at the outage plan and the risk analysts, in some
cases, haven't even started looking at it.
Are you asking the question because we haven't
really developed a shutdown risk model yet?
DR. POWERS: The committee has had the chance to
review a proposed rule in the area of shutdown regulation,
and rejected it, fairly sternly, on the basis that we didn't
feel like we had risk information about shutdown sufficient
to know what to regulate, and asked that Research undertake
a study to develop a risk profile during shutdown
operations, not only planned outages, but unplanned outages,
as well, and that has not progressed.
So as a result, I don't have the kind of
information base of what constitutes risk-significant
evolutions during outages that I have for normal operations
gained from things like the beginning of WASH-1400 and up to
NUREG 1150, and even the IPE insights document I find a
wonderful source of information about what is risky in a
plant during operations.
But I don't have that for outages. I've got a
huge inventory of, which I seem to now have a hobby of
collecting, of incidents that occur during various types of
outages and I know the kinds of things that get you in
trouble and I'm sure I could write a regulation to make sure
those things never happen again and I find, in general, they
don't ever happen again, people correct things.
But I don't have a feeling for how you get into
these problems and what kinds of things to look for.
MR. PARKER: And you bring up a good point.
That's what we're trying to do is look at those issues,
those risk insights that we have some knowledge on, but
we're using the tools defense-in-depth and some of the NEI
guidance to say, hey, mid-loop operation and different
operations like that are highly risk significant conditions
and that's the one tool we have.
But to go back to your point, the one opportunity
that we have is Perry is developing the shutdown model and
they intend to put that on their safety monitor, where they
will be able to have a probabilistic on-line risk monitor,
and it will be very interesting to be able to tie that into
their outage coming up next February.
But they hope to have it in place so they can use
it for their outage planning activities and that will be a
unique opportunity for us in the region to be able to see if
there's any insights that come out of that and share it with
other plants.
DR. POWERS: I think these things are all good. I
wish that you would have the kind of data that's in the PRA
community about the details of these models, because I know
that we have substantial questions about how you go about
modeling human error in these kinds of situations, which are
very different from operational situations.
And I don't see the kind of debate between
gentlemen, such as on my left, and his peers on how you go
about doing that modeling that I have seen in connection
with operational events and see the way that you set up the
structure, the fault trees and event trees for shutdown
events and the detailed discussions and the philosophy that
I see for operational events.
And so these things get created, I'm glad, and
they're going to help a lot, just like you said, but I would
-- I'm not sure they raise my comfort level an awful lot.
MR. PARKER: Well, that's what stirred up my
interest as far as certification. When we went out and did
the SDP activities, to look at some of the human performance
that we're crediting in our SDP that we have generic values,
ten-to-the-minus-one for a high stress and
ten-to-the-minus-two, and then we see the utility call it a
ten-to-the-minus-four for the same thing, we haven't
validated that and we're very uncomfortable and headquarters
is stepping back and looking, is it appropriate to use the
licensee's numbers versus ours.
And when we have an issue that results in a human
performance, how do we deal with that and where do we go; do
we step back and look at the licensee's assumptions and
their basis and validation behind that.
So there's a lot of questions in that area where
human performance becomes a real issue.
MR. DAPAS: That underscores the need for some
type of standard, in my view. From my perspective, your
comments are clearly valid about we have limited
risk-informed our processes. You're attempting to use the
tools you have. If the licensee is proactive, like they are
at Perry, you want to learn from that.
I think in the interim, though, we've tried to
come up with the SDP, recognizing its limitations, and we
have some tool to use to assess significance until we maybe
develop some standard where the licensee says here is my PRA
and we have confidence that it's sufficiently rigorous and
we can use that in our determination of risk.
Right now, we have this --
DR. APOSTOLAKIS: But will the licensee say here
is my PRA?
MR. DAPAS: They don't have to right now.
DR. APOSTOLAKIS: So does the Commission know that
you are a little bit constrained in your efforts?
MS. BURGESS: Yes.
MR. DAPAS: I hope so.
DR. POWERS: They should understand the
limitations of the SDP.
DR. APOSTOLAKIS: But, I mean, in order to
understand -- if we are the only ones, it doesn't work.
DR. POWERS: They have asked for us to talk to
them on the SDP, on whether the PIs are truthfully risk
significant. I don't think they're ready for the answer
we're going to give them. And since I get to be the
messenger, I may be dead next week.
MS. BURGESS: Slide 28 just highlights three
bullets, that the SRAs in the region are extremely involved
in the new process, very active and very busy just resolving
findings and issues that inspectors from DRS and DRP are
bringing to the table, running through the SDP process.
Since these worksheets are not yet completed, done
with the revisions, the SRAs are involved in almost every
issue.
DR. POWERS: I understand people are looking into
expanding the workforce of SRAs.
MR. GROBE: We can talk about that a little bit.
MR. DAPAS: That's one of the staffing challenges
Jack mentioned.
MR. GROBE: Yes. Could we hold off on that?
DR. POWERS: Sure.
MR. GROBE: Because we have another staffing
issue. There is one thing we haven't touched on with Mike
and Sonia that we talked about briefly earlier was how the
SRAs and risk analysts are going to get involved in event
response. We've only had one substantive event since the
new program went into force, and that was at Palisades.
And what we found was that there was a disconnect
between management's expectation of what could be provided
and what we actually had the capability to do.
So why don't you guys talk a little bit about how
Palisades went and what we expect to be able to perform in
the future, how we expect to be able to perform?
MS. BURGESS: With any event, preliminary
information is just that, preliminary, and it seems to
change minute by minute. So with the best information that
we get, based on a senior resident at the site giving us, we
were able to probably within an hour or an hour and a half
give a rough big picture estimate of the situation of the
event, conditional core damage probability.
DR. POWERS: I just have to interject an anecdote.
In the hours following the Chernobyl accident, they called
Moscow to explain they had an accident and the guy on site
says, well, they've had accident here, but things don't look
too bad.
That shows you how good preliminary information
can be.
MR. DAPAS: Pretty gross estimate.
DR. SIEBER: It's all relative.
MR. GROBE: But our residents have a little bit
more flexibility to speak what's on their mind.
MS. BURGESS: So we're able to give -- we have
limited tools with the SAPHIRE model and the GEM model and
obviously our model is not as extensive as the licensee is
being able to model certain components and that, but I think
we are able to provide a rough estimate, for event response
purposes, of whether we need to send a special inspection or
an EIT or an IIT.
I think in a lot of cases, definitely IIT is going
to be self-revealing anyway.
DR. POWERS: You're saying that you've got enough
information that you can provide information to management
to make these kinds of decisions.
MR. DAPAS: Right. Do we need a special
inspection? Are we comfortable that we have the big deal
threshold or do we have time to acquire additional
information and then maybe we need to send another inspector
from another site versus --
DR. POWERS: When you decide, you make a decision
and say I'm going to send a special inspection team to get
to the bottom of this. You give that team a charter.
MR. DAPAS: Correct.
DR. POWERS: And you have enough information to
give a charter.
MR. GROBE: The charter is developed within the
first couple hours.
DR. POWERS: But when they do their best, they've
had their week or maybe a weekend, they never occur at good
times, right? You've had -- and they've brought forth what
they need. Can you write what you would say is a good
risk-informed charter from one of these AITs or IITs?
MS. BURGESS: I believe we can. Just in the past,
before the probabilistic risk insight was used, we also used
deterministic risk insights. And our charters were very
right on the money when we sent out a team and I don't see
any difference now that the probabilistic risk insight is
added.
I think we can do a very capable job of giving a
real good charter to the team.
MR. DAPAS: But I think we would focus on things
like is the licensee evaluating the risk significance, is
the licensee trying to determine extended condition, is the
licensee conducting a root cause, and, if not, we would
challenge the licensee. And, again, that assumes that there
is clearly risk significance associated with this that
prompted us to send the special inspection.
MR. SINGH: I want to ask a question. SRA is a
part of the AIT team most of the time?
MS. BURGESS: Not necessarily. It's dependent.
MR. GROBE: The last time we went an SRA out was
the tornado that hit Davis-Besse. That was a year and a
half ago or so.
MR. PARKER: The flexibility is in the program
that if they think that there is a potential that there is
some uncertainty or some concerns that we have, that they
can --
MR. SINGH: How about, say, if you have an
inspection team inspection, do you have an SRA as part of
the team?
MR. GROBE: We certainly have that flexibility.
But generally, usually, a special team is our lowest level
of response. Generally, that's very targeted on equipment
problems, root cause, things like that.
MR. CALDWELL: But I guess the answer, we haven't
had a special inspection in this new process yet. So we're
telling you what we think.
MR. SINGH: Because the reason I ask, I asked the
question to Region IV when they had a fire at Diablo Canyon
last month, and they had a special inspection and they sent
the SRA up there.
MR. GROBE: That was a significant, complicated
event.
DR. POWERS: One of the things the committee has
to do is advise the Commission on where it should be
spending its research resources and we're wondering if they
are under-investing in developing these tools to be used by
the SRAs.
MR. GROBE: We're clearly resource constrained
right now. Almost all of our agency resources are going
towards the SDPs and as they pointed out, the shutdown
model, low power model, containment model --
MR. DAPAS: Risk-informed PIs is another
initiative that Research has embarked on.
MR. GROBE: The interesting, I get anecdotal
feedback, but I understand that the industry is not
interested in risk-informed PIs. That the amount of money
that it would take to implement it doesn't give them
sufficient payback.
DR. POWERS: What had been proposed up till now, I
agree with industry on that.
DR. APOSTOLAKIS: But if we couple this with the
maintenance rule, will it be much easier to define those
PIs? They already did a lot of it for the maintenance rule.
So there seems to be a distance or gap between the
maintenance rule and risk-informed regulations and using the
PIs. I don't understand why. I mean, what I don't
understand is why didn't the staff at headquarters say, when
they were establishing the oversight process, that the PIs
were plant-specific and the licensees should propose the
thresholds.
They did it with the maintenance rule.
MR. DAPAS: I think the licensee, in many regards,
has weighed in on the thresholds here.
DR. APOSTOLAKIS: But they have their own.
MR. GROBE: Not plant-specific.
DR. APOSTOLAKIS: They have their own.
MR. DAPAS: There was a strong emphasis with the
PIs to minimize the dollar cost of implementation. So they
depended very heavily on indicators.
DR. APOSTOLAKIS: Now they'll pay the price for
the severe criticism that everything that is expected to be
green and they don't mean anything and this and that, and it
seems to me that there was an easier way of approaching it.
DR. WALLIS: Dana was asking about tools and I
think you gave an answer about resources. Tools, to me,
enable you to do more with fewer resources.
MR. GROBE: That's what I was talking about; that
is, the resources are currently focused on other tool
development and our ability to develop all these tools is
resource constrained.
MR. DAPAS: From a regional perspective, I would
offer we are certainly interested in any tools research can
provide us.
DR. WALLIS: It may be we could get some
resources, or someone, to RES to develop things for you,
that's a different kind of resource.
MR. DAPAS: As long as they don't come from the
region.
DR. WALLIS: Yes.
DR. POWERS: That's another question.
Unfortunately, the ACRS has no role to play in that. That's
an NRC management function. But it's one we certainly worry
about, because it doesn't do any good to pay Peter by taking
from Paul.
DR. SIEBER: Well, I think there is one other
point, and that is that recently in the development of a lot
of the criteria involved with license renewal, there was a
notable contribution made by some people from one of the
regions in helping to put together part of that approach. A
lot of us, at least I personally am convinced that the
Commission would do itself a great favor if it would make
greater use of the talent that exists within the regions
and, in particular, those people who are the senior
inspectors, who have real knowledge of how the plants work,
when they put together some of these proposals and ideas.
And so to that extent, we may be doing you the
disfavor of suggesting that you be a greater participant,
but I hopefully would believe that that's, in the long run,
a productive thing rather than counter-productive. I mean,
we have to be frank with you on that.
MR. GROBE: We're one agency, though, and what's
best for the overall safety of the industry is where our
focus is.
DR. SIEBER: Yes.
DR. POWERS: My boss used to say that he was
giving you an opportunity to exercise your management
talent.
MR. CALDWELL: I think you're exactly right that
there are resources in the regions that would help out a lot
of the development of new programs, et cetera, but there
needs to be a shift in resources, because typically the
development is in headquarters.
So in order for that to work effectively, then we
need to shift some resources to the regions so that the
regions have that flexibility to interact or get involved in
the development activities. Because right now, it's the
program office that does all the development and they
resources for that. But we wouldn't disagree that we think
the talent we have in the regions could help that process.
It's just that we are base-loaded right now.
DR. SIEBER: Every time we've had a blood drive in
this organization, the people who have contributed have been
research and the regions. You don't understand.
MR. CALDWELL: I understand. I have to say,
though, that the program offices have taken some pretty
significant cuts and tried to prevent those cuts from the
region. So we have fared reasonably well in the past; in
fact, most recently.
My point is that if we're going to use regional
resources for developmental programs, then you have to
recognize that in the budget.
DR. SIEBER: I agree.
MR. CALDWELL: And take some of the developmental
resources from the program office and put them in the
regions. We are perfectly happy to do that and be involved.
It's just that we have to be careful that we have enough
folks.
DR. WALLIS: You can be very involved in defining
what are the problems, what could be the solutions, what
would help you. You're the customer for something. I don't
see you being quite so involved as a resource in developing
something, but very involved in being articulate and somehow
expressing what it is you need, what the characteristics
have to be of something which comes out of some research
activity.
MR. CALDWELL: A lot of the details of this new
program, though, were developed by regional resources.
MR. GROBE: That's right. That happened under the
old program, so we had some flexibility and we sent a lot of
folks into headquarters.
MR. DAPAS: Task groups, et cetera.
CHAIRMAN BARTON: Since lunch seems to be out the
door, we'll break for lunch from now until 1:15.
[Whereupon, the meeting was recessed, to reconvene
this same day at 1:15 p.m.]. AFTERNOON SESSION
[1:15 p.m.]
CHAIRMAN BARTON: We've got till 3:00. We don't
want to miss anything that you want to tell us you feel is
important, but try to get wrapped up by 3:00.
MR. GROBE: Well, why don't I fly through the
training analysis, then.
CHAIRMAN BARTON: Okay.
MR. GROBE: I mentioned earlier that in the area
of engineering inspections, that we've had to evolve our
expertise and that's because we're doing more design
inspections and we can no longer rely on contract resources.
CHAIRMAN BARTON: Right.
MR. GROBE: In addition, we've got a fairly high
turnover rate. A number of our individuals have left the
jobs with utilities, as well as we had a number of
retirements.
So we've been in a fairly strong recruiting mode
and we've been trying to emphasis recruiting of individuals
with a stronger design expertise.
That's different than the expertise we've had in
the past in the region. We've had some design expertise,
but not a lot.
CHAIRMAN BARTON: Are those people hard to find
now?
MR. GROBE: Absolutely. Absolutely. And we had
some folks go out to the east coast out of Region I to try
to find out if there was anybody interested in joining up.
But basically it's the engineering firms,
utilities and military that are our recruiting pool.
The safety system design inspection is five
engineers for three weeks and, again, if we're going to be
successful in those inspections, they have to be qualified
with design experience, mechanical, electrical and I&C
system engineers.
The Appendix R inspection, the fire protection
inspection is three multi-disciplined engineers, and, again,
they have to have very unique experience. They have to be
experienced in Appendix R inspection capability, and we're
going to talk a little bit about the kind of training that
they --
CHAIRMAN BARTON: Are they ongoing or is that a
one-shot deal?
MR. GROBE: We had inspections early, following
publishing the rule, through the '80s, that were, at that
time, intended to be one-shot inspections. Since then, the
inspections were suspended. Now, under the new program,
we've re-initiated some inspections.
For a while, we were doing inspections out of --
that had the acronym FPFI, fire protection functional
inspections, those were done out of headquarters and they
were not programmatic in nature in the sense that they were
mandated to be done at every plant.
But this is not an FPFI. It's not at that level
of detail. But it does touch on the same elements that a
fire protection functional inspection touched on.
You need somebody with fire protection engineering
capability. We don't have a fire protection engineer, but
we've trained one of our engineers to assess those kinds of
attributes of the licensees' design.
We also need an I&C or an electrical engineer, but
it's unique expertise in evaluating Appendix R types of I&C
issues. And then you need a system operations engineer to
look at how the licensee would implement procedures
post-fire and whether their plans are feasible.
DR. POWERS: Do you have plans to do induced
station blackouts?
MR. GARDNER: Yes. I'm not saying Region III has,
but there are some in the country, that because of the fear
of not being able to contain spurious operations, they go
into a station blackout condition, and that's a concern,
obviously.
MR. CALDWELL: We don't have them.
MR. GARDNER: Not that I'm aware of in Region III,
that's what I said. I'm not sure there are any in Region
III. We'll find out.
MR. GROBE: We should introduce Ron. This is Ron
Gardner. Ron is my electrical engineering branch chief. We
do our fire protection inspections out of the electrical
engineering branch.
DR. POWERS: The induced station blackout is a
problem, it's a recovery.
MR. GARDNER: Well, it puts you into a condition
that you don't want to get into.
MR. GROBE: Just to touch briefly on what we've
been able to accomplish to date, we hired a Ph.D. I&C
engineer who had 12 years of experience designing control
systems for fighter jets, digital control systems for
fighter jets. We're trying to turn him into a nuclear power
plant I&C inspector.
We hired, he's on yet on board, but he's accepted
our offer, an I&C engineer who was one of the co-chair of
the Appendix R BWR owner's group. So extensive Appendix R
experience.
We hired an electrical engineer that had extensive
experience in the industry, as well as prior inspection
experience, and we just brought in a mechanical engineer,
he's a former senior resident inspector, into the mechanical
engineering branch.
The area that we're having trouble is mechanical
design, piping stress analysis, that sort of thing. We're
still looking for that resource and we're still looking for
another electrical engineer. But we've had some success in
this area. They are hard to come by.
I want to talk a little bit about training. Ron?
MR. GARDNER: In the 1980s, when we did the
64-100, I don't know if you remember that number, baseline
inspections, that were actually to make sure licensees were
meeting their required date for implementing 50.48 and
Appendix R, we had degreed fire protection engineers in just
about every region and we augmented our people with NRR
resources and contractors.
We had a very good team. Unfortunately, since the
1980s, we've lost those fire protection engineers. We lost
one to NRR, one went to actually OI. And then the FPFIs
came back, and I can talk about how we got to where -- some
of that's with Generic Letter 92-18, you might be aware. So
you know how we've gotten there.
In any case, unfortunately, today, with the
baseline program introducing the FPI, we don't have degreed
fire protection engineers. We have inspectors that were
doing the base fire protection inspection and that is a far
degree of difference between that and design of fire
protection systems.
As Jack indicated, we have started training a fire
protection engineer. We had a training session that NRR put
on, two sessions each a week in Brookhaven, you may have
heard of that. We're having a follow-up training session in
the region here in September, one day, unfortunately.
For the first couple of inspections, we're having
a contractor assist. We're doing an inspection right today,
the last day of the inspection is Friday, at Braidwood, fire
protection inspection.
We have two Brookhaven contractors. That's OJT
that we're getting from them. We have NRR technical expert
also on that team that's also giving them some training.
So through a combination of OJT and classroom
training, we are attempting to reach a level that we feel
comfortable with as far as the technical capability of our
people in this area.
As you know, it's very complex, though.
MR. GROBE: For a period of six months, we've
gotten limited contractor resources in the fire protection
area, and for about 18 months in the design area, to put one
contractor on each inspection team. And the goal of that is
to develop some on-the-job training.
In addition, we're doing some internal course work
on heat sink, thermal hydraulics, somebody mentioned heat
transfer earlier, because we have a new inspection we hadn't
done before, it's called heat sink. What it primarily
focuses on is the viability of heat exchangers.
And we're exploring the TTC in other regions,
discipline-specific course work in heat transfer, set-point
methodology, instrument loop uncertainties. We hadn't
focused a lot in the past in these areas, so we're looking
at developing some internal course work in those areas.
CHAIRMAN BARTON: TTC?
MR. GROBE: TTC is the technical training center,
currently in Chattanooga. I would expect most of this is
stuff we're going to do.
MR. SINGH: I have a question. Before you
suspended the inspections back in the '80s, did you ever do
the triennial inspections in fire protection?
MR. GROBE: No. We didn't do one in this region.
MR. SINGH: You did not.
MR. GROBE: No. There were, I think, only three
done in the entire country.
MR. SINGH: No. There were lots of them. I did
all of them in Region IV.
MR. GROBE: Oh, did you?
MR. SINGH: Yes.
CHAIRMAN BARTON: How many did you do in Region
IV?
MR. SINGH: Eight. So nothing was done. Thank
you.
MR. GROBE: Any other questions in engineering?
MR. DAPAS: I just wanted to touch upon, starting
with slide 42, some of the staffing challenges in the
resident inspector program. We've experienced a relatively
high turnover rate and consequent with that is the challenge
to fill vacancies.
You have to post the vacancy, go through the
selection process, and then train the individual, and with
the qualification process, it can be several months before
we have a fully engaged resident inspector replacement once
we've identified the vacancy.
CHAIRMAN BARTON: The primary reason for the high
turnover rate or does it vary?
MR. DAPAS: It varies. It can be promotional
opportunity for the resident inspector that may go on to be
a senior resident inspector or come into the regional
office. It can be -- and that goes for both resident and
inspector and senior resident inspector.
It's a bit more limited for the senior resident
inspector in terms of promotional opportunities, but there
have been a number of residents that have received
promotions, or requests for lateral transfers. We had a
resident inspector that wanted to go back to NRR to be a
project manager and we supported that. He, of course, had
family in that area and that seemed to be a win-win.
And in addition to that, there's attractive salary
offers out there in the industry. Some of these plants that
were in extended shutdowns, like Cook and others, plants
that are merging, there's opportunities for experienced
resident inspectors and you're dealing with signing bonuses,
et cetera, and lucrative salary offers, that's been an
attractive draw.
MR. CALDWELL: Was your question -- were you
trying to get to whether there's dissatisfaction? I don't
think we have -- I mean, there's always going to be some
folks.
CHAIRMAN BARTON: But 12 percent is pretty high
turnover.
MR. CALDWELL: I think most of the folks that left
went for either geographic, promotion, or something that
benefited them, either money or whatever. I don't think we
lost anybody that just --
MR. DAPAS: Or early-out, I don't think so.
MR. CALDWELL: -- didn't like the program anymore.
MR. DAPAS: And that's one of the things we try
and probe, was there some concern with or dissatisfaction
with your working environment or et cetera.
DR. POWERS: But if the inspection program is
going to turn them into automatons and eliminate
discretionary and judgmental aspects of it, are you going to
lose people?
MR. DAPAS: I'd challenge that characterization of
the new program, but --
DR. POWERS: I put the worst spin on it I can
here.
MR. DAPAS: I think we are asking the inspectors
to bring judgment to bear and as I said, in the context of
what revisions do we need to make to the program, I know
that Jim and I have had a lot of discussions, we place a
high value and premium on experienced individuals with
mature judgment and we value that and we're going to
consider that input.
And we -- divisional meetings or one-on-one
discussions with the residents, we go out to the site, we're
continuing to encourage them to flush issues up to branch
chief management, so those can be considered and evaluated,
and not get locked into this, well, the new program doesn't
allow me to do X or Y.
CHAIRMAN BARTON: One of the concerns I have is
they do an SDP and they get frustrated because in the past
it was the findings of violation and now you do it and it's
--
MR. GROBE: It's an issue that we're having to
focus some management attention on, because we've completely
perturbed all of the structures that the staff had to
demonstrate their own --
CHAIRMAN BARTON: Exactly.
MR. GROBE: -- in terms of value. So we're
building what is currently called a significant reactor
finding. We're going to rename it, but we're doing more
internal recognition of inspection issues that add value,
but don't get to a white, yellow or red threshold, add value
because they provide insight to the licensee or provide
insight to us as far as inspection techniques or other
issues that other plants can look at.
So we're trying to find ways to give the staff
anchors for their value, but it is a challenge.
MS. NESTON: Does this 12 percent also include the
rotation out of a particular plant because they've been
there for so long?
MR. DAPAS: I'm not sure on that.
MR. CALDWELL: In the range, it could include
someone who has rotated back to the region, because either
their time was up or we've had individuals who didn't stay
the full seven because they were grandfathered with the
five. They came up to their five and decided they wanted to
do something different and rotated either back to
headquarters or here.
MS. NESTON: And they would be included in that 12
percent.
MR. CALDWELL: They would be included in that.
MR. GROBE: In honesty, we haven't had a lot of
folks that have been -- that have moved because they've
gotten to their time limit. That's the exception, not the
rule.
MR. DAPAS: That's with the extension to the seven
years. But I think, and I view this as a positive, I think
we've had a number of instances where feedback we've
provided to the program office, discussion that we've
generated in the different forums to discuss the new program
has resulted in some change, and we try and build upon that
as positive examples for the inspection staff, where
expressing their views has resulted in revisiting of a given
approach.
So we are encouraging that across the board as we
go into initial implementation. The pilot program, we had
input from really two branches, and now we've got input from
all the branches, and there is a learning curve that they go
through. Some of the feedback we're able to address as a
result of lessons learned from the pilot program and then
there's also additional insights that are communicated that
we discuss and forward to the program office.
So I view that as kind of healthy. There's a long
training period, as I mentioned, for qualification. You
have to attend the BWR, the PWR series, plant-specific
system knowledge, on-the-job training, that's certainly a
large aspect of the resident qualification program, and then
the emergency preparedness responsibilities, understanding
the licensee's emergency response plan, the NRC
responsibilities.
And I caveat this, appropriately. Some PRA
training that the residents receive so that they can
understand the use of the SDP process and how risk impacts
inspection activities, and then they go through a course, an
oral qualification board, where we have various branch
chiefs that sit and ask questions to test knowledge in the
regulatory perspective.
CHAIRMAN BARTON: What happens if they fail the
oral board? Do they get another shot?
MR. DAPAS: We have had a couple individuals, in
my experience in the region, that we felt needed another
qualification board. So there were particular areas where
they had to concentrate and devote some additional study and
then they were successful in their second board. But the
branch chiefs, I think, are fairly successful in not
offering or sponsoring a resident for a qualification board
until they're pretty confident that they've acquired the
requisite knowledge to be successful.
So we've had limited experience where that has
occurred.
MR. SINGH: Do you also have an oral board for the
regional inspectors?
MR. GROBE: Absolutely. Every inspector goes
through an oral board.
MR. DAPAS: And then when we looked at the pool of
experienced resources, that's a bit limited. Obviously, we
draw from the Navy or shipyard or licensee operational
experience.
DR. POWERS: If the Navy keeps working its folks
as hard as they are right now, you'll have a big pool of
people.
MR. DAPAS: We get some applicants that have a lot
of experience in the nuclear power program that the Naval
Reactors runs and that's because they are downsizing. So
they're looking for other opportunities.
But this does require an aggressive recruiting
program, because as I said, the competitive salaries and the
signing bonuses in the industry, the lengthy process we have
to go through for selection, rating panels and interviews,
et cetera. So that can sometimes -- where employee X can
say here, we're offering you a job here.
Sometimes we've been in the process of going
through the selection and we're ready to forward an offer
and individual X has said, well, I just took an offer a
couple weeks ago with company Y. So sometimes we're
confronted with that and we look for ways to streamline
that.
One of the things that we're also looking at is
the entry level program, and that certainly is a resource
investment, but we want people with experience. But, again,
that can be limited, so we look and explore the entry level
program.
MR. CALDWELL: Mark is going to try to hustle up
here so we can get into the fire protection stuff, but I
want to make sure, before he gets out of this, if you have
any questions on this, because it is probably one of the
most important programs we have; not necessarily because the
other aspects, what we do is not important, it's because
these are the folks that are on the site that are there all
the time.
What they do is -- what I saw as the biggest
change in the way the agency worked was that licensees now
expect to have somebody there, so that they don't operate
differently than they would if an NRC presence wasn't there.
I talked to some staff people and they told me
that in the old days, when they knew the inspection team was
coming out, they changed their mode of operation for that
week and then changed back after they left.
So the resident program has provided a routine
presence which keeps folks from operating differently when
we're there.
CHAIRMAN BARTON: It keeps them honest.
MR. CALDWELL: Well, I didn't want to say it that
way, but that's essentially it. I didn't mean to interrupt
you, Marc.
DR. POWERS: Well, there's another thing that you
have to bear in mind, that all of us have to bear in mind,
that there is a very, very crucial role that they play and
this SDP process is their process for screening their
findings and whatnot. So their level of responsibility, to
my mind, has actually gone up in this new procedure and some
of these things I worry about are responsibility and
judgment, notwithstanding I think there are still concerns.
MR. GROBE: The SDP is not limited. The residents
obviously have a role in evaluating their findings, but the
region-based inspectors also use that, that the value to the
inspection program that the residents add is -- I can't
remember the number -- but several hundred hours of their
time is allocated to what we call plant status and that --
it's 650, and that's supposed to be a risk-informed
assessment of what's going on, so that they can engage
themselves in the right activities and also engage the
region-based folks that come out in the right activities
from the risk perspective.
MR. CALDWELL: I cut Marc off and I apologize.
MR. DAPAS: One of the things we talked earlier
about is the impact of the training courses at the technical
training center, when they're offered, but branch chief X
has a vacancy and is successful in filling that, but the
annual PWR course just completed, that individual has to
wait till the next year to pick that up.
CHAIRMAN BARTON: So it's only given once a year.
MR. DAPAS: Right, and I guess that is a function
of the demand that they have when you look across all the
regions and all the offices, that they were only able to
justify one course a year, but sometimes that does have an
impact depending on when your individual reports on board.
And we already talked about absence from the site
for an extended period. If you're attending a seven week
course in Chattanooga, you're going through the
qualification process, that impacts baseline program
execution and site coverage and that requires pretty
involved branch management of the inspection --
CHAIRMAN BARTON: You bring another inspector on
board for that period of time, right?
MR. DAPAS: Right. We were looking at like a
contingency plan. A good example is in an outage. The
licensees are short during outages. There's I forget how
many hours associated with the resident inspection portion
of the outage. Do you recall, Laura?
MS. COLLINS: Eighty.
MR. DAPAS: Eighty hours. Doing that at what
might be a 22-23 day period can be a real challenge if
there's only one inspector on-site and branch chief X might
ask the other branch chiefs can you help me out with sending
someone during this outage period.
And as I mentioned, on-the-job training is a large
part of the program. And the experienced SRIs look at
resident inspector development as a high priority and their
responsibility. It's kind of like I'm training my
replacement coach. So they place a premium on that and I
think we get a lot of value-added.
And the other thing, as I mentioned, we look at
reduced training length when hiring high quality individuals
who can hit the ground running. We have had some interim
certifications in selected areas of the inspection program
because an individual comes on board that has an extensive
operations background.
And then the extensive cross-training. I was just
looking yesterday at the number of residents and senior
residents that have both PWR and BWR training, and so
they're fungible to go to other sites without having to take
the specific series course.
And the other aspect of this cross-pollinization is between
DRS and DRP. We've had a resident inspector go to operator
licensing and a senior resident that reported to operator
licensing, as well as an individual from the engineering
branch going out and being a senior resident.
So there is some cross-pollinization between
divisions which we think is real beneficial.
If there are any questions.
CHAIRMAN BARTON: Do resident inspectors get
overtime?
MR. GROBE: Yes.
CHAIRMAN BARTON: Are they paid overtime?
MR. GROBE: Absolutely. Let's move on to risk
training. What I'd like to do -- do you folks have any
questions about the SRA training program? Are you familiar
with that?
CHAIRMAN BARTON: No, I'm not familiar with it.
What slide are you on, Sonia?
MS. BURGESS: I'm on 46. There's Region III is no
different from the other regions. There's two SRAs in each
of the regions and there is consideration of an additional
risk trained person and that can take the form of a couple
of different options.
One is using existing inspectors with additional
risk training, so they can do it part-time, and another
person that's dedicated to assist the SRAs in the analysis
of risk.
CHAIRMAN BARTON: Have you been in a position of
trying to assess what your needs are?
MS. BURGESS: Yes.
CHAIRMAN BARTON: You need another warm body or do
you need an assistant?
MS. BURGESS: We have. The SRAs have put their
input in and what we would desire, what we think we would
need. We definitely think we'd need at least one additional
risk person.
CHAIRMAN BARTON: A lot of times people will have
one slot and they say, well, what I'll hire is a new senior
reactor analyst. Point in fact, they've got enough senior
reactor analysts. They need an assistant for them to help
them carry out their jobs, and I'm just wondering if you had
thoughts on that.
MR. CALDWELL: There's no plans to have an
additional SRA slot. As I mentioned earlier, there's a task
force that, in fact, the meeting starts -- the first meeting
is on the 26th, of the four regions and headquarters, to
talk about SRA succession planning, and that really is to
talk about the type of training that you would give one,
two, three, four, five individuals, I'm not going to
prejudge how it comes out, but a number of individuals who
would not be fully SRAs, but would have additional training
that they could support the SRAs and the region in risk
assessments.
Not a short-term thing. I mean, the two SRAs are
going to be just up their necks in work, but it's a
recognition that there needs to be some more expertise in
that area and a recognition that that you need to have
somebody in the pipeline unless an SRA gets promoted or
decides to leave.
DR. POWERS: They better not.
MS. BURGESS: That's a great segue into the next
slides.
MR. GROBE: I was going to say there's a lot of
personnel barriers associated with this, because the SRA
position is a higher graded position than any other staff
position we have in the region.
So there's a lot of issues that come up in the HR
area.
MS. BURGESS: On slide 47 is the SRA training
certification program or process is an 18 to 24-month
program. It's divided into classroom and rotation and I've
listed some of the technical training, the statistics, PRA
training, and then the NRC PRA computer modeling training.
That, in itself, can be up to 27 weeks of training.
CHAIRMAN BARTON: Where do they get the PRA
training?
MS. BURGESS: In headquarters. And most of the
time, much of the training is contracted out. Brookhaven,
INEL. So just the classroom portion of the training is a
significant amount of time.
Rotations, there's nine months of rotation. Mike
and I did five months in NRR in the PRA branch, we did three
months in the Research PRA branch, and then we did one month
at another region to get on-the-job training, with the
assistance of an existing SRA, to see what their job duties
were and how they conducted business in the region, and took
that back to our region.
MR. CALDWELL: And that's one of the areas that
we're going to look at. Now that we have experienced SRAs
in the region, we may not need these extensive rotational
assignments. They'll just spend their time with the SRAs in
the region to get their on-the-job training. But the
classroom training that she was talking about is extensive.
It's not a short-term fix if somebody leaves or
you need additional help. It's something that we're trying,
for the long-term, and come up with a plan that will keep
people in the pipeline and bring up the whole level of the
region's expertise of risk.
DR. POWERS: It also offers the opportunity for
substantial job satisfaction improvements there, the guy
feels like he's going into modern technology.
MR. GROBE: I think we've covered slide 48. Why
don't we go on to 49?
MS. BURGESS: Slide 49 just highlights some of the
training that the regional inspector and the resident
inspector would receive. This first bullet is a two-week
class. It's a combination of the PRA basics plus we have
how integrated the SDP process and now the PRA and the IPE
all integrate into the SDP process. That's a two-week
course.
An then also we've given extensive training on the
SDP process itself. We've had a lot of workshops and with a
lot of examples of issues from other regions and that's
helped the inspectors put some practical use to the SDP.
MR. GROBE: Any questions in the risk training
area? Before we get into the fire protection area, there
were two questions that you asked earlier that we didn't
really get a chance to answer, and I'll just give my
perspective and open it up to Jim and Marc.
One had to do with power up rtes. We don't have a
lot of insight on power up rate, other than the fact that I
could share with you a concern that I have. Jim Dyer
mentioned Quad and Dresden are going to be coming in for
some fairly significant power up rates and you indicated
Duane Arnold is, and that has to do with secondary side
capability and the ability of the operators to operate the
plant in a higher, significantly higher power level, and
whether that's going to impact on initiating event
frequency.
I don't have any more insight to share with you,
other than that's a concern that we have, and I'd throw that
open to Marc and Jim.
MR. DAPAS: The power up rate, I guess from the
resident inspector perspective, I think you would get
involved, the resident inspectors get involved in looking at
if there's any tech spec ramifications. Many times, the
tech spec package that comes out, headquarters is
considering, the residents will be asked to review, to offer
any perspective procedural implications.
So it's just really changes to the tech specs and
procedures that result from the power up rate. I can't
really envision any other area where the residents might be
engaged.
CHAIRMAN BARTON: With a number like that, you're going to
have to make some hardware system changes when you go in
that level.
MR. CALDWELL: Right. Set point changes that have
to be made and they have to be made as they go up.
MR. DAPAS: Which are captured in the tech specs.
CHAIRMAN BARTON: Yes, but they actually have to
make changes in the plants. You have to -- because the trip
set point stays the same, but the 100 percent power, as it's
calculated, changed, and so the trip set points have to be
changed in the instrument and control.
But what Jack mentioned is something that I don't
know that we have any insights into, but some licensees find
that they get the up rate and they just don't have the
capacity we have any insights into they tripped their auto
valves or their turbines aren't set up, at least the way
things are set up, to handle that type of --
CHAIRMAN BARTON: Fermi is a good example of that.
MR. CALDWELL: Right. So those are things that
they have to kind of inch up to and that's what we will be
watching, how they do that, how they control it, and most
licensees, at least today, had done it very slowly and very
deliberate.
DR. SIEBER: These major up rates, though, they're
really talking about a new front end on the turbine and the
things like that.
MR. CALDWELL: Yes.
DR. SIEBER: Which really changes the physical
plant.
MR. DAPAS: But I'm not aware of any prescribed
inspection activity where we would go out and verify that
what the licensee communicated in their licensing submittal
is, in fact, the case in terms of equipment modifications.
MR. GROBE: It would be an opportunity through the
affirmative plant mods inspection and the safety system
design inspection to target some of those areas.
DR. SIEBER: But you know that the stress level on
the plant is going to be higher.
The other you raised earlier was license renewals and we
haven't had any in Region III, but we've seen that train
coming down the tracks, and we assigned a project manager to
stay aware of what's going on in the other regions and
headquarters.
The inspection program for that activity is fairly
significant and while it doesn't have any direct impact on
the baseline program, it doesn't change anything we do in
the context of baseline.
It's resource intensive and as we shared with you
earlier, we don't have a lot of resources.
It's also a fairly unique expertise that's
necessary. There is discussion underway right now, and
maybe, Ron, you can expand on this, too, to capture that
inspection activity out of headquarters or currently it's
out of the regions.
Why don't you talk about what we've done as far as
trying to gain insights in this area?
MR. GARDNER: As Jack indicated, we have a
principal inspector that we've assigned to get with the
other regions who have started down that path, to find out
what they've done, how they did it, what worked, what
didn't, to try to get to the point where when we get our
opportunity, we're not starting at ground zero, that we've
already built on what other people have done and tried to
make improvements.
AS Jack indicated there is some discussion about
who will do what. There's a big portion, as you might
imagine, of environmental qualification questions that come
into life extensions of license renewal.
And I have been part of one of the research
working groups, for years I was on that, on aging of
materials and such. So I have an acute background in that
also.
So I think we have the wherewithal to do the
inspections, the challenge will be finding the resources to
do it.
MR. GROBE: What's the total number of inspection
hours we've seen?
MR. GARDNER: I can't remember.
MR. GROBE: My recollection is on the order of 700
and something.
MR. GARDNER: I thought it was 800, roughly 800.
MR. GROBE: It's a very significant impact,
because it has to be done a very short period of time.
DR. POWERS: We have a statutory responsibility
for all those and we're looking at a major tidal wave coming
in at us and it could literally consume everything we do.
MR. GROBE: I think that captured all the
questions that I had written down earlier. I think what I'd
like to do now is go into the fire protection issues and
turn it over to Ron Gardner. I know that you're going to
have some questions. I suspect you're going to have some
questions.
MR. GARDNER: As Jack indicated, my name is Ron
Gardner. I'm the Chief of the Electrical Engineering branch
in DRS and fire protection falls in my branch.
What I've tried to do is make a presentation that
would address where the new program is going with fire
protection, not only triennial, but also the more day-to-day
review of fire protection and the normal fire protection
things that the regions have been doing over the years.
We didn't stop that. We're just doing it in a
different manner.
The first thing, I guess, on slide 55 that I want
to emphasis is the risk contribution of a fire. It is
significant and if you stop and think, with the fire, you
can have a plant transient, you could have a reactor trip,
you could have a loss of off-site power, you can -- we
talked about self-induced station blackout.
All those require fairly significant reactor
operator actions. You can go beyond that, though, with the
high-low pressure interface problem or a stuck-open PORV, a
spurious operation of an SRV, and you enter a LOCA
condition.
Compound that with a loss of off-site power and
you've got very numerous operator actions. Then with a
fire, you may have smoke, which could inhibit or prevent
operator actions. You have flooding, you have the heat of
the fire.
The fire itself is a very significant area of NRC
historical perspective and it looks like it's going to
continue, that we're going to maintain our focus on this.
There were a number of years where we backed off.
Information Notice 92-18 and the subsequent problems we had
with the implementation of that, that was regarding
motor-operated valves and the potential for spurious
operation and control room fires, to have the valves not
only go to the wrong position, but to be destroyed
mechanically because of the bypassing of the torque
switches.
Also, we had some FPFIs that failed, with
significant findings.
So going on to page 56, I wish there was a silver
bullet where we could say here is the fix, that we could say
the risk of fires has gone away by just doing this one
thing.
No one has been able to find that silver bullet.
So that instead, what we find is the best approach and
licensees have found the best approach is the definition
methodology or mentality. It starts off by preventing
fires, and I'll talk more about that when we talk about what
the resident does and how we try to gauge how licensees are
doing in preventing fires.
Then we have the part of rapidly detecting,
controlling and putting a fire out. Great success, if you
remember the Fermi turbine explosion, it released thousands
and thousands of gallons of oil, EHC fluid, et cetera, and
distributed it all over the plants, with all the water
systems that were ruptured, and the fire was extinguished
and rapidly extinguished, and that could have been a very,
very significant fire and it didn't happen.
So that says that in that case, the rapid
detection, control and extinguishment of the fire worked,
and that involved obviously even the hydrogen system for the
generator.
DR. SIEBER: One of the problems is, though, that
when you have a big fire in the plant that involves
operations, it's the operators who are the fire brigade.
MR. GARDNER: Often, and I'll about it. They have
a lot of manual actions, too, sometimes to mitigate the
fire. One of the things that the licensees are required to
do is for any fire area, is to dedicate or to preserve
enough equipment to safely bring the plant to cold shutdown,
and the performance goals they're trying to make is
reactivity control.
They want to make sure the plant is no longer
critical. They want to have makeup. They want to have
decay heat removal. They want to have enough indication for
the operators to know which manual actions to take or which
actions and EOPs to follow. And a support system.
So that's quite a lot that you have to maintain
regardless of whatever fire you can postulate. To do that,
you have barriers, suppression, safe shutdown procedures,
and you have a number of equipment and systems that are
dedicated just for those operations that have to survive in
the event of a fire in any given postulated fire area.
Unfortunately, there are no performance indicators
existing today, and this is slide 57, to provide insights or
to help us to say that we don't need to do an actual
inspection.
And we haven't given credit for self-assessments.
One of the reasons, and I'm not saying a significant reason,
but one of the reasons was when we were doing the FPFIs,
Prairie Island did a self-assessment. When we did the FPFI,
we gave them credit for it. So our FPFI was focused on
determining the adequacy of their self-assessment.
When we went out there and looked around the
plant, we found a number of issues that their
self-assessment had missed, and they weren't small issues.
I don't have -- if you look at the inspection report, you
could see them.
They were fairly substantive issues, and we were
surprised. I'm not sure if that had a major contribution to
the fact that the NRC wants to at least start down the road
of doing our own inspections, but it probably didn't help
the licensee's cause any, because I know NEI was looking to
see if they could have more credit for self-assessments.
MR. DAPAS: Didn't we also, though, Ron, have some
have some real significant inspection findings in that area
and that has furthered the point that we should
independently verify.
MR. GARDNER: Right. Now, a number of licensees
are doing self-assessments and they are finding significant
issues, and that's to their credit. It's just whether or
not we are comfortable with saying they are to the point now
where they can find the amount of problems that we think are
there still, and that's an unfortunate statement, but that's
true, unfortunately.
Now, as I was indicating, it's not just a
triennial or design inspection. We have a constant focus on
fire protection that's brought about by the residents.
On a quarterly basis, residents tour six to 12
areas of the plant, and they're looking at the classical
fire protection features. They're making sure that the
licensee doesn't have extensive combustibles or ignition
sources for those combustibles in the plant.
The licensee has requirements for storage of
combustibles, et cetera, they're looking at that. They're
making sure that the material condition of the fire
protection systems is up to par, that they're not degrading.
Operational lineup, say, for a C02 or a halon system,
they're making sure it's properly lined up, so if there is
an automatic initiation, it would function.
They look at operational effectiveness of the
equipment and of the licensee's fire brigade and fire
barriers.
CHAIRMAN BARTON: Are those quarterly inspections
what you require to be done during an outage?
MR. GARDNER: Required to be done during an
outage.
CHAIRMAN BARTON: Yes.
MR. GARDNER: I'm not sure that the procedures
differentiates between an outage and a non-outage condition.
CHAIRMAN BARTON: The only reason I bring it up,
because in outage, you've got an opportunity to bring in a
lot more fire loading combustibles.
MR. GARDNER: That's why the residents are out
doing this, because they're there during the outages and
not. Usually, the region stays away from an engineering
type inspection during an outage, the residents are there
anyway.
MR. SINGH: This question came up last week when
we were got in the NEI conference on fire protection. They
don't want it during the outage, because there's too many
combustibles, too many --
MR. GROBE: They don't want us to do an
inspection?
CHAIRMAN BARTON: Yes, that's why.
MR. SINGH: They emphasized the point that they do
not want the NRC doing inspections in the outage.
DR. APOSTOLAKIS: Do you have any IPEEEs yet?
MR. GARDNER: Any I what?
DR. APOSTOLAKIS: IPEEEs.
MR. GARDNER: We have the Generic Letter 88-20,
Supplement 4, IPEEEs that the licensees have been providing.
DR. APOSTOLAKIS: So you have their IPEEEs.
MR. GARDNER: Yes, we do. If you recall, in fact,
several years ago, Quad Cities released their
5E-to-the-minus three that really stirred up the region to
take action on that.
DR. APOSTOLAKIS: I wonder whether you can
prioritize these fire areas that you're inspecting according
to their --
MR. GARDNER: And I'll get into that in a minute,
if I could, because that's one of the things we do as part
of our triennial and it's also done by the resident
inspectors when they are looking in their areas.
MR. GROBE: Step back for just a second. That's
one of the reasons that we have to spend more time preparing
for these inspections, because everything we do has to be
risk-focused. So something as simple as selecting which
plant fire areas to look at would involve some consideration
of the risk significance of fire areas.
DR. APOSTOLAKIS: That's a one-time job, though.
After you've done it, you have it for that time, correct?
Unless something dramatic changes.
MR. GROBE: That's correct.
DR. APOSTOLAKIS: So it's an initial investment in
a new process.
MR. GARDNER: No. What we find, and I'll go into
that. It's changing. It's not a static. It's a dynamic
number --
DR. APOSTOLAKIS: But the critical locations,
unless you really change the plant, it's where the cages
come together.
MR. GARDNER: Evidently there's other things. We
have found that it's changing, and I'm going to that in a
minute.
DR. APOSTOLAKIS: Okay.
DR. POWERS: There's a little problem in using the
IPEEEs as the basis for prioritization.
MR. GARDNER: Right.
DR. POWERS: Because there are some crucial
assumptions that some licensees have made in screening
things out, I mean, things that just don't appear in the IPE
have gotten screened out because though there's a high
combustible loading, you can say, well, there's no ignition
source. I can screen this area out.
Well, that's all well and good. What happens when
an ignition source gets introduced?
MR. GARDNER: And I hope to get to that point,
too. That's one of the subtle aspects of the new program
versus the old, in that when we postulate a fire that can
affect safe shutdown equipment, we have to be able to
demonstrate how the combustibles, whether they be cables,
scaffolding, whatever, how it can ignite, what is the
ignition source, and then how you can get the fire to
migrate from one part of the fire area to another.
In the past, we used to assume it just happened.
We just say you have to assume it happens. Now we have to
develop a scenario to show reasonably that it will, in fact,
because of the heat plume and of the effects of that plume,
it will transverse the fire area.
So if I don't get into that further, if you need
more when I go through it, let me know.
DR. POWERS: There are other subtleties in there,
as well, because a lot of the IPEs have been done saying,
well, the fire barrier penetration seals are 100 percent
guaranteed absolutely effective. And I don't know of
anything that's that guaranteed.
It's just one of these problems. You just can't
look at an IPEEE and say, well, this is truth, it's truth if
one person saw it.
MR. DAPAS: Ron, do the inspectors look at the IPE
to understand the assumptions before they go out?
MR. GARDNER: Yes, and that's what I'm going to
get into in just a couple slides.
On page 59, if we can go to that one, we shift
from what the residents are doing on a monthly and a
quarterly basis to an annual inspection.
It's always important to understand how the
licensees fire brigade can perform. It may come down that
they are the last of the defense-in-depth for a given fire
area. So we hold them to a high standard.
DR. POWERS: We usually just assume that
defense-in-depth. In the good old deterministic days of
Appendix R, we just assumed that fires aren't out until the
fire brigade goes in to put it out.
MR. GARDNER: That may be true. I don't recall
that.
DR. POWERS: Automatic suppression systems were
assumed only to control fire and to actually put it out, you
had to have somebody walk in there and put it out.
MR. GARDNER: At 3G, it gives credit for
separation and if you don't have separation, for suppression
and detection.
DR. POWERS: It's just suppression. It's not
putting it out. The fire's not over until somebody actually
goes in there and declares it out.
MR. GARDNER: From a design approach, it gives
credit for suppression.
DR. POWERS: Under the new program, we weigh
suppression. We have a fire mitigation frequency, I don't
know if you're familiar with the new SDF, significance
determination process for fire protection, and there is a
formula for SMF which includes fire barriers, ignition
frequency, and automatic suppression, manual suppression,
and CC, which is common cause.
So that is figured in to the equation. Again, on
the resident inspection portion, on an annual basis, they
check certain aspects of the fire brigade. What they would
probably do is not ask for the fire brigade, but find one
that is routinely scheduled and observe it.
The triennial inspections do not demand that a
licensee do a fire brigade just for the triennial. We would
get information from the residents about whether their
perception of the fire brigade's adequacy was, as well
reading what licensees are finding and documenting their own
critiques of their fire brigade drills.
Now, on page 60, I shift to the triennial team
inspection. This is not a classical fire protection,
looking for combustibles. It is more focused on design.
And in the preparation aspect, we talk or
communicate, get with the SRAs, the regional SRAs; if they
are tied up, we get with headquarters SRAs, and we get the
risk rankings for different fire areas, and we have found
that the IPEEE can give you some numbers.
We go out to the site and we find that those
numbers may have changed. That just happened at Braidwood.
The numbers changed. I don't have all the reasons as to why
it happened, I just know that it did happen.
We also look at the transient sequences. All of
this is done in conjunction with the SRAs to assist us in
saying which of these fire areas would probably be the best
for our inspection to focus on.
One of the things we may stay away from, by the
way, is the control room. The control room is so analyzed
and has so many people in it that some of the other rooms,
sometimes we think would be more bang for the buck, so to
speak, to look at than the control room, which a licensee
automatically assumes they're going to evacuate anyway.
But that is a case by case basis, we'd have to
look again and look at the rankings.
We have a very important two to three day full
team information gathering visit. That's where the full
team goes to the plant. They walk down the fire protection
systems, safe shutdown systems. They look at the P&IDs.
They determine what might go wrong. They say that the
licensee is relying on HPSI for makeup and they may look and
say, okay, let's see if we can find a valve that, if it were
to close, would isolate HPSI from the water supply it needs.
And then they would check that cable or that valve
to see if it's been protected or not. They would look at
spurious operations, et cetera.
So that first two or three days is a very
important aspect of our inspection. Obviously, we look at
risk rankings, we look at things like that.
Then we come back into the region for a week, the
whole team does, take that information that they gleaned
from that two to three day bag trip, we call it, and
determine their inspection plan.
They've finalized the areas they're going to
inspect. They determine some cables, some areas of question
they're going to focus on, and they get just about ready to
go out there and start the inspection as if they had a very
limited time, which they do, by the way.
CHAIRMAN BARTON: Wouldn't an inspector go look in
the corrective action system to see how many outstanding
items there are against fire protection system, deficiencies
that haven't been corrected or are backlogged?
MR. GARDNER: We don't go into the licensee's
corrective action program in detail. We have a small
percentage of our inspection that looks at that.
What we try not to do is mind the licensee's
corrective action program. We try to do an independent
assessment of the licensee's fire protection program.
DR. APOSTOLAKIS: On 60, it says that you select
three to five plant areas important risk for inspection.
MR. GARDNER: Right.
DR. APOSTOLAKIS: Then on 58, you said that you
are inspecting six to 12 fire areas on a quarterly basis.
MR. GARDNER: On page 58, I was talking about the
resident inspections. That's covered on a monthly or
quarterly basis.
MR. GROBE: And that's just looking at classical
fire protection, combustibles, controlled ignition sources.
DR. APOSTOLAKIS: But the question is why can't
these six to 12 plant fire areas be ranked according to risk
so you focus on the risk significant areas?
MR. GARDNER: We do under the triennial design
inspection. We pick the most risk significant --
MR. GROBE: Laura, did you guys, when you did this
module, did you use IPEEE insights to focus risk?
MS. COLLINS: We did.
CHAIRMAN BARTON: So even the six to 12 areas are
among the --
MR. GARDNER: Yes, sir. They are also risk-based
or risk-informed. Excuse me.
DR. APOSTOLAKIS: The areas are risk-based. They
come from the PRA.
MR. DAPAS: The inspections risk-inform, though,
when they're selected in the areas.
DR. APOSTOLAKIS: That's right.
MR. GARDNER: The triennial inspection shifts from
the classical fire protection to a design focused
inspection.
DR. APOSTOLAKIS: Are these areas, though, you
take them from the licensee's risk assessment.
MR. GARDNER: We look at the IPEEE, we talk to the
SRAs and we get the licensee's assessment of the relative
risk.
DR. APOSTOLAKIS: So you may decide there are additional
that require a tool, even though the licensee may have not
found them to be a not very significant safety.
MR. GARDNER: That could happen. I'm not saying
it's going to happen, but it could happen, certainly, if we
found a basis for it. The resident inspector may have a
reason for us to go to a particular fire area based on what
they've been seeing.
DR. APOSTOLAKIS: See, that's where the standards
w discussed earlier this morning become very important,
because many licensees have used screening methodologies and
unless you really look carefully at the assumptions that
they have made, you may have missed important five areas.
The IFPI-805 is going to solve that, right?
That's why ASME and ANS are not looking at fires. It's an
IFPI that will do it. That means there's something fishy.
You have to understand means this. Go ahead.
DR. POWERS: IFPI's expertise in fire risk
assessment, just the personnel on the committee, it's just
very, very limited. It's like one guy that really knows a
lot about fire risk assessment. He may be the only guy in
the country who a lot about fire risk assessment.
So to say that we will have a standard that means
that you can look at a five analysis and have some
confidence that you don't have to go plowing into the
assumptions. I think that's overly optimistic.
DR. APOSTOLAKIS: So they should have given to the
ANS then.
DR. POWERS: We haven't see any product from ANS
at all.
DR. APOSTOLAKIS: Yes. They are more experienced
fire analysts there.
When are we going to review this?
MR. SINGH: August 28.
DR. POWERS: That's when the committee meeting is.
DR. APOSTOLAKIS: Do I have it? You gave it to
me.
MR. SINGH: Yes, so you do have it. I have a
question. Did you have a chance to provide a comment on the
NFP-805?
MR. GARDNER: I did. I believe I did. It was
some time ago, I believe, and I think I remember --
MR. SINGH: Let me ask you another question. When
I was at the conference last week, they discussed this
NFP-805.
Did you realize that they have taken out the high
pressure enthalpies from the core and also the -- it's
really watered down.
DR. APOSTOLAKIS: The agency is going to endorse
it for sure.
MR. GARDNER: Isn't it true that 805 will not be
required to be endorsed? Is NFPA-805 going to be required
to be endorsed or is it going to be --
MR. SINGH: It's not required, but they are
forcing the NRC to look at it.
MR. GARDNER: But licensees will have an option as
to whether they choose to enforcement.
DR. POWERS: And I suspect the number of licensees
that will pick it up is going to be zip.
MR. GARDNER: That's my point.
DR. APOSTOLAKIS: I don't know about that. If
it's nice and doesn't get into too much detail and it's a
national standard, I think the licensees are going to push
for it.
DR. POWERS: It makes Appendix R look like a
cavalier off-the-cuff document. It's like doing Appendix R
with a risk assessment.
DR. APOSTOLAKIS: That's tragic.
MR. GARDNER: Okay. Slide 61, the triennial team
inspection has about 200 hours direct inspection and Region
III is doing it in two weeks, other regions are doing it in
a one week time period. And Region III is an outlier.
We think that two weeks gives us more time to
develop our inspection questions and to have the licensee
give us the answers in a more deliberate fashion, so that we
feel like we've accomplished what we need to accomplish.
DR. POWERS: One of the issues that came up at the
fire protection forum, and if you're not attending those, I
would really encourage you to attend. They are great
meetings that are put on by NEI, but they have lots and lots
of information coming in about lots of things.
One of the questions they had, when you take this
bag visit, people have been through this, said, gee, it
works a lot better if the whole team comes for the bag
visit, not just a few guys.
Is that what you're planning to do?
MR. GARDNER: Yes, sir. In fact, we had the first
plat, which was Braidwood, they questioned us as to why we
had more than a team leader coming. They thought that just
the team leaders only should show up and for the reasons I
spoke to earlier, it's of great benefit for the whole team
to be there, and that's what we plan to do.
DR. POWERS: And I think that's the experience in
industry. It makes life a lot easier for them, and actually
NRC got some pretty high praise for the people running these
things, saying that they had -- they get a letter that says
assemble the entire universe of documentation on fire
protection, that the team leaders have been very effective
in whittling that down to what actually was needed and used.
So NRC got some real strokes from the licensees on
that, triennial inspections.
MR. GARDNER: Going on. We look at the fire area
boundary design. Some plants have been forced, because of
the vintage of the plant, to use huge areas. Quad Cities,
originally, based on their design, used practically a whole
turbine building as one fire area.
They and most licensees, through further review,
are trying to narrow the scope of the fire areas to make it
more user friendly, so to speak, for themselves and for the
inspectors.
MR. DAPAS: That's because Quad Cities had to use
bounding assumptions, because they didn't know the cable
routing configuration.
MR. GARDNER: Yes, and also because unfortunately,
when the first plants were built in the '60s, they didn't
understand that it may be better to have more concrete walls
than fewer, that those concrete walls could, in fact, be
natural fire barriers. Brown's Ferry hadn't occurred yet,
in other words.
Safe shutdown system selection adequacy. We see
if the system they chose to have for makeup or for heat
dissipation is functional during the fire or after the fire,
et cetera.
System separation evaluation, we look at the 3G2
aspects. Any questions about those, I can enumerate on
them. There's three basic ones.
When you're doing the inspection, you do a fire
suppression -- slide 62 -- fire suppression.
DR. APOSTOLAKIS: What happened to 61? I have a
question.
MR. GARDNER: Yes, sir.
DR. APOSTOLAKIS: The separation, as I recall from Appendix
R, it says that trace carrying cables or redundant trains
should be separated by at least 20 feet.
MR. GARDNER: There are three criteria, 20 feet is
one,. No intervening combustibles, and automatic suppression
and detection, if you use that method.
MR. GROBE: That's one exam criteria. Plus
suppression and detection, plus no intervening combustibles.
The 3G2A says --
DR. WALLIS: So this is 20 feet in the horizontal
direction.
MS. BURGESS: Right.
MR. DAPAS: Right.
DR. APOSTOLAKIS: But in a PRA context, though, if
they are 20 feet apart, that will, of course, inhibit spread
of fire from one tray to the other, but there is a fire in
the room and they're near the ceiling. Does it matter if
it's 20 feet or 30?
MR. GROBE: That's why it requires -- the 20 feet
is permitted, but only with suppression and detection. So
you've got a sprinkler system to knock down the heat, you've
detection to bring the operators in promptly or the fire
brigade.
DR. APOSTOLAKIS: But these are all the
defense-in-depth measures. But the separation criteria
means nothing to identification, because you have a layer
that tries --
MR. GROBE: It's somewhat of a compromise. There
is a three-hour barrier or 20 feet horizontal with
suppression and detection and no intervening combustibles.
The staff concluded that those were approximately
equivalent in protective capability.
DR. APOSTOLAKIS: But if I have the suppression
capability, then why do I need the 20 feet? Why is that
important if I have --
MR. GROBE: Defense-in-depth. Probability of
failure.
DR. SIEBER: If they're right up next to each
other, suppression isn't going to help you.
DR. APOSTOLAKIS: I think they had in mind only
propagation from one tray to the other. The fact that you
will have a layer of gases that are hot.
MR. GARDNER: Well, if you have the 20 feet of
separation, you don't have intervening combustibles, and you
have detection and suppression, we don't affect that the
fire will affect both redundant trains and we will give you
credit and say you are successful, you have protected
adequately.
DR. APOSTOLAKIS: If there is a fire somewhere
else in the room generating hot gases, then both the trays
will be --
MR. SINGH: No, George.
DR. APOSTOLAKIS: No?
MR. SINGH: If the fire is in the corner, you
still meet the 20 feet criteria.
DR. APOSTOLAKIS: If I have the trays 20 feet
apart, near the ceiling.
MR. SINGH: Right.
DR. APOSTOLAKIS: And there is a fire in the
corner. Very quickly, if you have enough combustibles,
you're going to have a hot gas layer there.
MR. DAPAS: You have a sprinkler system.
MR. SINGH: You have a sprinkler system and you
have a detection system.
DR. APOSTOLAKIS: So then why isn't the sprinkler
system relevant if the separation is only ten feet? See, we
selectively use it when it's --
MR. DAPAS: We can only conjecture what was in the
thought process. Some of us were around when that happened.
DR. APOSTOLAKIS: I think that you do not
anticipate the hot gas layer from a third fire, that what
they had in mind was spreading from one to the other, in
which case all these measures make sense.
MR. DAPAS: We could only conjecture what was in
their thought process.
DR. APOSTOLAKIS: There is one fire in the corner.
You don't need a second fire. It is too hot. The reason
I'm saying this is because the first time it was pointed out
was after the first fire PRA was done and people said, yes,
that is correct.
MR. GARDNER: Again, though, if you're going to
use 20-foot, you can't have intervening combustibles. If
you get into a diesel generator room, you're probably going
to have to use a three-hour or a one-hour fire barrier.
So, you can't just blindly pick 20-foot. It
depends on whether or not there's a chance that a fire that
could occur as you were postulating in the middle. Then
both drains go, but if that can happen, don't try to use the
20-foot. Use another one. Okay? That's where we'd be
looking.
One of the things -- on the first slide -- the
first point on slide 62 is the fire suppression damage
assessment.
This is the part where, when we come into a fire
area that we've picked and we do the what-if scenario, what
could go wrong, in other words, how likely is it, and then
what are the consequences of it, that's the basis of our
inspection.
Licensees would have protected, let's say, through
20-foot separation, three-hour fire barrier, whatever. We
don't find a problem with the barrier and we don't find a
problem with the 20-foot, our rule indicates it's 21-foot,
whatever.
We still don't stop, because what we find is that
-- let's say, again, the licensee for a fire in that area is
relying on a charging pump.
They have reliance on the BCT to be the initial
source of water.
DR. WALLIS: How do you use this ruler when the
conduits aren't parallel?
MR. GARDNER: We can take a average plane, a
vertical plane, and walk that off. We can do it. We look
at the valves from the DC-2 -- in fact, we've got this
question at Braidwood.
The licensee had a cable for one of the valves on
the BCT that ran through the fire zone and was unprotected,
and it had been overlooked.
So, that's the kind of things we look at.
Sometimes the licensee has manual actions in a
fire area, and they have -- in their procedure, the operator
will come in and operate the valve manually.
At Braidwood, we found they were going into a room
that was going to be 178 degrees. Our question was is this
going to be a good idea?
They said water packs, and we said, well, it looks
like he has to be there for cold shutdown. That's 72 hours.
You know, most water packs will start boiling, if you're not
too careful, after so many hours at 170-some degrees. It
won't be boiling, but they'll be darn hot.
So, we have issues like that. That's the kind of
thing we do through every fire area we pick, even when the
barriers look pristine.
DR. POWERS: The step at which you have to assess
the level of degradation of these is a step I've never
understood very well.
MR. GARDNER: What level of degradation?
DR. POWERS: Okay. When you come in and you look
at either manual fire capability or the fire suppression and
detection capability, you have to make some sort of an
assessment on the level of degradation -- high, medium, or
low.
MR. GARDNER: Right.
DR. POWERS: And that's the step I've never
understood.
What constitutes high and what constitutes low?
MR. GARDNER: It is somewhat subjective. I'm not
sure it is completely objective.
Let's say you found the BCT valve and now you say
I have a potential fire area degradation; I want to run it
through a SDP screening.
Phase one, which would be just a cursory, is there
a potential for any significance, you whip right through and
say yes.
You go into a phase two and you have to calculate
the fire mitigation frequency, which uses, then -- which
requires you to have first an ignition frequency for
whatever combustibles are in that room, it looks at the
barriers, and if there is degradation of the barriers,
starting with the fire barriers, you do a moderate or --
what's the term? -- highly degraded, I think, and those have
numbers that adjust the risk.
That's somewhat subjective.
DR. POWERS: Yeah. I mean the numbers that are in
there, that you actually plug into the formula -- I even
actually found out where they came from, and they come out
fine, but you have to make the subjective judgement on these
things, what's the level of degradation here, and that was
the step I never understood, and I have a set of notes from
the BNL course to see if I could understand better just that
exact issue.
MR. GARDNER: I went to the BNL course, and I
don't think the notes will help you.
What will happen is this -- whatever method -- and
we usually are fairly conservative -- you go to, you will
come out with, let's say, a white issue. That doesn't end
the process. That's when you start refining the level two
evaluation. You'll get the SRAs. The licensee will get
their own SRAs in there.
You will elaborate to the licensee what
assumptions you used to come to a white conclusion. One of
them would be that you're assuming significant or high-level
degradation to the fire barrier or the manual suppression,
whatever it may be that you're doing in that part of the
calculation, and the licensee would obviously come back and
say they think it's moderate, and the difference between
moderate and significant can make you from a green to a
white, as you know.
DR. APOSTOLAKIS: But shouldn't the ultimate
criteria, though, be, really, the relative speed with which
a fire is expected to spread, how quickly you can stop that.
That really should be the ultimate criteria.
MR. GARDNER: That's a part of it. It's much more
complicated than that.
DR. APOSTOLAKIS: Like what else?
MR. GARDNER: Well, ignition frequency -- okay.
First of all, you have to postulate --
DR. APOSTOLAKIS: Oh, you mean when you deal with
--
MR. GARDNER: -- the plume and that there is a
potential for --
DR. APOSTOLAKIS: But suppression deals with a
fire that's already there.
MR. GARDNER: Yes.
DR. APOSTOLAKIS: So, Dr. Powers asks how do you
decide that degradation is significant. What I'm saying is
the criterion really should be can you arrest the growth of
the fire before it does damage.
DR. POWERS: That's not the way the thing is set
up, George.
DR. APOSTOLAKIS: I know it's not, because it was
not done using risk assessment.
DR. POWERS: Yeah, it was. It was done using your
wonderful fire technique.
DR. APOSTOLAKIS: No. No. We very clearly have
an equality there. The time to damage has to be less --
greater.
MR. GARDNER: I think if you're familiar with the
fire protection SDP process, you can see that they have
tried to make --
DR. APOSTOLAKIS: It's very hard to do.
MR. GARDNER: -- a mathematical estimate of the
significance, and I think the fire protection is less
subjective than the internal events. It makes it more
difficult and it makes the people that use it have to be
more sophisticated in their capability to understand risk
and how to use it, but it's not perfect, and we're going to
use it, and just like with the other one, we'll probably be
revising it before long.
Continuing on with operator recovery action, when
the fire has been somewhat put out, there's still smoke
removal, de-watering.
At FERMI, we had six or seven hundred thousands of
gallons of water to -- because of surface contamination --
to decontaminate, and you'd be surprised at the public
outcry when you tell them you're going to put it through
filters and send it out to the lake.
DR. POWERS: I'm not going to be surprised.
MR. GARDNER: That's quite tricky.
Control re-unitization -- you try to re-establish
your power systems that you've lost, get all your systems
back now, instead of the ones that got you to safe shutdown,
and return to service.
We also do a manual fire-fighting capability
assessment just to assist us with the SDP if it becomes an
issue.
As parts of the design aspect we're looking at --
and that's slide 63 and 64 -- we're looking at electronic
circuit analyses common enclosure, high-impedance faults,
spurious circuits.
If you want to discuss a high-impedance fault,
it's an arcing fault.
Any of those things I could talk to you about in
specifics, but in general, just for the purpose of what we
do, is we're looking -- as electrical engineers, we're
looking at common enclosure, associated circuit faults.
We're looking at common power supply. This goes
into breaker coordination, fuse coordination.
A high-impedance fault is not your classical
volted fault. It's not the one where you're estimating the
contributions of your inductive motors. As they start
stopping, they will actually feed faults, and when you're
doing a normal fault analysis, you have to get all your
contributions.
In this case, you're just doing a -- assuming that
the fault is what they call a arcing fault, and that
actually can be of more problem than a volted fault.
DR. APOSTOLAKIS: How can you have a spurious
signal from an open circuit? Can you give me an example?
MR. GARDNER: If you have a circuit that's
supposed to be open and you have a dual ground -- first
ground on one side of the contact that's open and then you
ground the other side, you now create a bypass around that
closed -- an open circuit.
DR. POWERS: The Europeans, in testing their new
modern cable insulation, found out that open circuits became
closed circuits, because there was some copper oxide in the
material that got reduced by the boric acid or borate that
they put into it, and open circuits all became closed. I
mean it was a conduction pathway.
MR. GARDNER: Sure.
DR. POWERS: And so, needless to say, they've kind
of redesigned that new super insulation.
MR. SIEBER: Why are high-impedance faults more
significant sometimes than volted vaults?
MR. GARDNER: If you can visualize the fact that
you have a distribution panel -- let's say it's feeding
125-volt DC and you're feeding, let's say, three loads that
are part of your safe shutdown, and then you have four or
five loads that aren't, but unfortunately, those four or
five loads run through the fire area, and we will postulate
that you will have multiple high-impedance faults on each
one of those loads that runs through that fire area.
Each one of them could be an arcing fault, which
means the current of that fault will be slightly less than
its breaker.
So, the combination of all of those currents can
equal the tripping of the supply breaker to the whole
distribution panel, which cuts off the power to the one you
needed to suppress the fire or to deal with the fire.
MR. GROBE: We have about 30 minutes left. We're
still in fire protection, and then we had a discussion of
on-line maintenance.
Is your preference to stay with fire protection?
DR. POWERS: I would like to.
MR. GROBE: Okay. And if we have a few minutes --
Laura, I'm kind of cutting you off, but -- Laura and Mike.
If we have a few minutes, we'll talk about on-line
maintenance; if not, then we'll just conclude with fire
protection. And we'll skip the break.
MR. GARDNER: Any other questions about hot
shorts, open circuits, high-impedance faults, common
enclosure?
DR. POWERS: Well, you'll never get a resolution
on that between the NRC and the licensees.
MR. GROBE: Well, you're not going to get it from
us.
DR. POWERS: I understand. I'm asking for
prognostication, not resolution here.
MR. GARDNER: I think you're talking about the
classical question that's confronting us about whether a
licensee has to assume multiple hot shorts versus a single.
That issue we wrote a TIA on, which is a task
interface agreement, and we have not seen the definitive
answer yet.
There have been meetings between the staff and NEI
and the owners groups.
I believe, in talking to the staff, they're pretty
sure that our position is going to be the position, but I'm
sure if I talk to NEI, they'll probably tell me the
opposite.
DR. POWERS: Are your licensees in this particular
regional happy with that, or are they resisting?
MR. GARDNER: No, but Braidwood -- Commonwealth
Edison is one of the licensees, and they were the basis for
our task interface agreement. They emphatically said one.
MR. GROBE: Put some time-frames on it, Ron. The
TIA was based on Dresden, wasn't it, and that was about four
years ago?
MR. GARDNER: Yeah, four years ago, I'd say, we
wrote that, right.
MR. CALDWELL: I think we were the first region to
really address the issue.
MR. GARDNER: It might have been, yeah.
MR. SINGH: Hey, Ron? Does Perry have that same
similar problem?
MR. GARDNER: Who's that?
MR. SINGH: Perry?
MR. GARDNER: As far as their position?
MR. SINGH: No, I mean do they comply with their
hot short issue?
MR. GARDNER: When you're talking about hot
shorts, you mean do they assume multiple hot shorts?
MR. SINGH: Yes.
MR. GARDNER: I'm not sure. We're getting ready
to go to Perry, and one of the next two inspections will be
Perry, and we'll find that out.
I didn't keep a catalog of who does what. We're
going to pick them up on the FPI, and hopefully that will
give NRR an opportunity to come to one position or the other
when we find it during these inspections.
MR. CALDWELL: I think we scheduled our fire
protection inspections to target those plants where we
thought we would probably have the most question in terms of
their approach, if I recall correctly.
MR. GARDNER: We did Braidwood partially for that
reason. That was the first one. Perry is number three, and
we're going to be looking at that.
Actually, we also picked Quad-Cities in December,
because Quad-Cities will complete, we hope, all of the
modifications necessary to establish full compliance with
Appendix R by November, which would make our December
inspection like just in time, and if you have any questions
on Quad --
DR. APOSTOLAKIS: Was there a high number?
MR. GARDNER: Yes.
DR. APOSTOLAKIS: The result of wrong analysis,
very conservative analysis, or are they actually doing
anything about it?
MR. PARKER: It depends on who you ask.
DR. APOSTOLAKIS: See, that's why I'm asking.
MR. PARKER: The licensee pointed out that there
were some over-conservatisms in their analysis. So, they
had to make some bounding assumptions.
So, that was part of it, and then they did
implement some compensatory actions and were making
modifications, because they did agree that their plant had a
high fire risk vulnerability, but they claimed the 5 times
to 10 to the minus 3 was really over-stating the full
as-found condition, if you will.
MR. GROBE: You have to appreciate that the
refined analysis with significant improvement is still 5 10
to the minus 5. It's not low-risk, but it's equivalent to
their --
DR. APOSTOLAKIS: Just from fire.
MR. GROBE: Yeah, just from fires.
We have two more topics. One's the SDP, which I
sense a lot of familiarity with. The other is -- we've put
together some slides on Quad, if you guys are particularly
interested in Quad.
DR. POWERS: I think we can get Quad from another
route.
MR. GROBE: Okay.
MR. GARDNER: Okay.
I can finish the last two slides, then.
Sixty-five is where I was headed.
The next, baseline use of risk information at the
baseline fire protection inspection -- and as I tried to
state earlier that both the triennial and the resident
inspections are using risk information to guide where they
look and how significantly and deeply they look when they
pick those areas; also, that the fire protection
significance determination process is in its own
compartmentalized document, and it's IMC-0609, Appendix F,
and that's a good document to have available if you're going
to be following fire protection issues.
DR. POWERS: At least in the version they gave us,
there's an egregious typographical error in Appendix F.
When you go through the calculations, you come up with --
depending on how you read the typographical error, either
with astronomical numbers for any plant or minuscule numbers
for any plant.
MR. GARDNER: We had tried it a few weeks ago at
Brookhaven, and we didn't find any errors like that, so
maybe the version we had was a later version.
Slide 66.
We would expect that the resident inspector, with
their understanding of the fire protection issues and the
complexity of the SDP, would only be involved in phase one
screening.
If it looks like it had to go further, they would
engage the region and the SRAs.
The inspection team, however, will do a phase one
and a phase two, and if, in fact, we find that the phase two
is heading us towards other than green, we would continue to
do that, and that would be a more protracted evolution, with
inputs from the licensee and more refinement with the SRA in
helping us to look at our assumptions and seeing if we were
overly conservative.
That was all I had prepared.
Jack indicated I have some material on Quad, but
you indicated you didn't need that.
So, any questions you have on this material, I'd
be glad to discuss.
MR. SIEBER: I think your presentation was very
good.
MR. GARDNER: Thank you.
MR. GROBE: You can tell, this is about as excited
as Ron gets, but this and the SSDI inspection we feel are
very meaningful inspection efforts. You can really find
stuff with this kind of inspection, and we're excited about
both of those inspection efforts, very detailed,
design-oriented, intrusive-type inspection.
If there's a problem, we could find it with this
type of inspection.
MR. PARKER: I hope all our inspections are
meaningful, though.
MR. GROBE: Yeah, but these are new tools that we
didn't have before.
DR. APOSTOLAKIS: There are no performance
indicators. They are planning to --
MR. GARDNER: No, sir. I think we haven't --
we're not smart enough to figure out which ones would be
relevant.
DR. POWERS: Great men have tried.
MR. GARDNER: That's right.
DR. APOSTOLAKIS: How about fires, the number of
fires?
DR. POWERS: It just turns out to be meaningless.
MR. SIEBER: They're mostly wastebasket fires.
MR. GROBE: And they're fairly frequent. You'll
have a couple of fires a year.
DR. SEALE: Any good performance indicator is
something that is not so rare that, in itself, it's a
catastrophic event. So, you want something that happens
every once in a while as a performance indicator.
DR. POWERS: Yeah, but wastebasket fires just
aren't going to do anything.
DR. SEALE: I agree with you. I'm saying the
frequency is not the problem. It's the wastebasket.
MR. GARDNER: I think we're also concerned,
though, that a low number might lull you into a false sense
of security.
So, there's some danger on taking any number and
saying that is going to make your determination as to
whether you're there or not as far as defense-in-depth.
DR. POWERS: With NFPA, when they tried to do it,
they ended up putting in this incredible core of Appendix R,
essentially, kinds of inspections and deterministic
activities, because there was no way to say, okay, if
they're doing all this, this indicator will indicate that.
MR. GROBE: I think you could develop an indicator
that could result in your ability to cut back in the
classical fire protection inspection area, but this and the
SSDI are very design-oriented, and I can't think of any
performance indicator that could result in you giving
justification to cut back in this area, because this is
focusing not just on ignition sources or initiating events,
those kinds of things. I think we could develop an
indicator in those areas. It's focusing on did your
engineers do a good job designing it, in a very complex
design.
DR. POWERS: And are your people maintaining it
and subverting it inadvertently?
MR. CALDWELL: Right. In actuality, the
performance indicator is the results of the inspections over
a period of time.
DR. POWERS: Yeah, that may be it.
MR. PARKER: When we met in Region II to discuss
inspection resources and how we were going to implement the
new program and what is the appropriate estimated number of
hours, there was discussion about the frequency of these
inspections, and I think there was the recognition across
the regions that the safety system design inspection and the
fire protection inspection were -- the two inspections where
probably the most risk-significant findings will emanate,
and as a result, do you want to continue with that intrusive
inspection, versus looking at performance indicators, and
so, there was that discussion.
DR. POWERS: One question, in thinking about
smoke, are you staying aware of these difficulties people
are having with their assumptions on how well-sealed their
control rooms are?
MR. GARDNER: You mean to keep the smoke out of
the control room?
DR. POWERS: Yeah, leakage rates.
MR. CALDWELL: The control room habitability has
been a problem as long as I've been in this agency.
DR. POWERS: We're seeing occasions of enormous
discrepancies between what's assumed in the FSAR and what
the actual tracer gas types of mixing are. I mean they're
just not even close. I mean it wasn't even a good guess.
And it's really because the FSAR is writing about what
somebody drew up on a piece of paper.
MR. GROBE: That in-leakage is when the door is
closed. If you have an event, that door is going to be
opening and closing on a regular basis.
DR. POWERS: That's another question that comes up
on the leakage test, is there's a lot of other things
happening. The HVAC system gets manipulated around and
changed, may be off, and whether the test actually relates
to the environment during an accident, but over and above
that, even with the test and the conditions you have, we're
seeing huge discrepancies.
MR. SIEBER: Well, the duct work is like a furnace
duct in your house, and it deteriorates, too. They use
those Pittsburgh seams to hold them all together.
MR. GARDNER: Well, there's also an over-reliance
on IEEE-383, I think, cable fire tests, to say that that's
the end-all to say I won't catch fire.
In reality, all that does is raise the ignition
temperature, but once it's ignited, it burns faster and
hotter than a non-IEEE-383 cable.
DR. POWERS: I've heard that.
MR. GARDNER: It's true.
DR. POWERS: I have not seen the data, but that's
definitely what I've heard.
MR. GARDNER: Yeah.
DR. POWERS: But on the other hand, we also find
that aging cables are less combustible.
DR. SEALE: They've already evaporated.
DR. POWERS: It's actually a cross-linking thing
and you get rid of the plasticizers, which are the real
flammable part.
MR. GARDNER: It's the oxygen scavenging from the
neutrons, yeah.
MR. GROBE: Any other questions?
Laura, you're on.
MS. COLLINS: We can be brief. We don't have that
many slides. I'll answer whatever questions you have.
DR. POWERS: If you haven't learned by now, the
ACRS has an infinite supply of questions.
MS. COLLINS: I'm going to talk on the topic of
risk associated with on-line maintenance, and we have a
procedure in the new baseline inspection program that's
carried out by the resident inspectors, and it's actually
7111.13, titled "Maintenance Risk Assessments and Emergent
Work."
Part of that inspection, we would sample between
five to eight maintenance activities per quarter, and that's
dependent on a unit size, and I'll say right up front that
this is a lot more emphasis on reviewing these types of
assessments than we had under the old core program.
The concept is to evaluate the effectiveness of
the licensee's risk assessment and control of the
maintenance activities.
That's the objective, and this was really
developed because we knew (a)(4) was coming, (a)(4), the
requirement of the maintenance rule, which, really, under
(a)(3), we previously said they should do a risk assessment,
and they were for the most part, but we didn't have -- it
wasn't really a requirement, so now it's becoming a
requirement.
Since we knew it was coming, we put it in a
baseline inspection program and we've been doing it kind of
ever since then, but I will say, because of that, and
because (a)(4) isn't fully in effect, we really anticipate
more changes to this procedure.
We've had two throughout the pilot program. The
guidance is changing. My understanding is that NRR is even
going to come out to the region and do a temporary
instruction, go out to the licensee's facility and really
see what they're doing and what we should be looking at, to
help us, I think, define what a finding is going to be in
this area.
On the next slide, I've just written down the
inspection objectives from the inspection procedure.
We looked at planned work. We also look at
emergent work, and then the last bullet is verifying that
the licensee has adequately identified and resolved problems
in this area, and that's just a standard thing we have in
all of our inspectable areas.
If they come up with some kind of problem in this
area, we can select that and go in and see what they do
about it to fix it.
MR. BONACA: Some of the emphasis in -- you know,
in the rule is manage risk. Any consideration to limit the
risk? That's a question which is somewhat open, because in
absence of criteria and in absence of tools to quantify the
risk, I mean it seems to me like there is some option there.
We were shown yesterday that, you know, increasing
risk from a baseline of about a factor up to 10 is not
considered high enough increasing risk that you have to go
to management for approval. It's a judgement. It depends
on how low your baseline is.
So, any sense on how this is being implemented at
the sites?
MS. COLLINS: Well, we can go on to the next
slide, where I start to talk about our inspection
techniques.
MR. BONACA: Okay.
DR. WALLIS: I was going to ask you -- I see
you're evaluating effectiveness several times and you're
looking at adequacy. Is there a lot of judgement involved
in this?
MS. COLLINS: Absolutely.
DR. WALLIS: It's all judgement.
MS. COLLINS: It's all judgement at this point,
and we're looking forward to new guidance and new
information from NRR, as I said a minute ago, to what would
be a finding in this area.
Even right now, we have preliminary information in
our inspection procedure that I understand is from the NEI
guidance which we're endorsing with our reg guide, and
there's different levels with increase in CDF and increase
in CDP, and I had an inspector call me recently because I
was in a pilot program and say, well, I'm here at this plant
and they don't calculate increase in CDP, they only do CDF,
what do we do about that?
I don't know what we do about it at this point.
You know, we're going to -- those are the kinds of
questions and some of the feedback, I think, that we've been
giving throughout the pilot program to the program office,
that not only do we need guidance for licensees, but we need
the guidance for the inspectors to say what is really an
issue in this area?
DR. APOSTOLAKIS: Now, the NRC staff developed
this upper bound on the CCDP of 5 10 to the minus 7, I
believe. Why can't we use that here?
I mean instead of having a licensee say, well,
gee, I'm really managing risk, because under exceptional
circumstances, all I'm doing is raising the CDF by a factor
of 3 and I'm already very low, but in the context of, what
was it, allowed outage times, they came up with this number
of 5 10 to the minus 7, which means about three hours you
have a CDF of some value.
Can that be -- you know, lacking anything else,
why can't that be a starting point for evaluating or
verifying how the licensees manage the risk?
MR. PARKER: There are some thresholds in some of
the documents. The problem I think Laura is pointing out is
there's no requirement.
So, if the licensee were to exceed those and the
residents and the SRAs or challenge the utility, what do we
do with that and how do we address that?
DR. APOSTOLAKIS: The 5 10 to the minus 7 is one
of the Region V risk-informed regulatory guides. It may not
be a requirement here.
DR. POWERS: It's an allowed outage time.
DR. APOSTOLAKIS: Yeah. Well, it's an increase,
an increase in CCDP.
MR. PARKER: But I think Laura's point is that
this task group is looking at the maintenance rule,
implementing procedures associated with (a)(4) here. We
would assess that, you know, what is a finding.
If we identify that the licensee did a CCDP and
determined it was greater than 5 times 10 to the minus 7, in
what context do we put that on the table, what's our
assessment of that, etcetera.
DR. APOSTOLAKIS: I'm not saying this is the
answer. I'm saying at least there is a starting point there
where somebody thought about it and came up with a footnote
that is really very nice. We don't know what to do, but
let's assume this.
MR. BONACA: Yeah, because -- in part, also, is
because -- I mean the risk increases associated with how
many components you're taking out of service and what kind
it is.
Now, especially for those power plants that are on
24-month cycles, they have plenty of time over two years to
do maintenance on-line without taking multiple components
out of service.
So, what does it mean, this managing risk? I mean
does it mean that since I can go up to whatever I want, I
can take five components out of service simultaneously.
There is a balancing act there that I don't think
has been properly defined, and that's why I was asking those
questions.
DR. APOSTOLAKIS: Maybe that will be the next
round of refinement. We haven't really had a chance to
think about these things.
MR. CALDWELL: First of all, at least there's a
recognition that they have to put something in place to do
an assessment of it.
I guess I'm a little removed from the inspection
program, but what Laura is saying -- we don't have the
criteria or guidelines yet to do an assessment of it. But
at least we're requiring them to do an assessment.
As we get smarter, those licensees that -- they
actually know what is good and what's bad. Those licenses
that -- because we don't have the tools yet or the whip or
whatever, the lever -- that want to push the envelope will
be the ones that we catch as we get smarter and come up with
our criteria.
Those that are good and smart and know how to do
this -- they'll already have set themselves a limit that
will be within where we end up.
DR. APOSTOLAKIS: I think the staff, though, at
headquarters should think a little bit about this, because
this is very important.
Now, yesterday, as Dr. Bonaca said, we were shown
some spikes in the core damage frequency, but I don't recall
any discussion of the duration.
MR. BONACA: There was no duration.
DR. APOSTOLAKIS: There was no duration. It was
just the core damage frequency went up, and then they said
themselves, regions -- you know, we told them to change
their names, but they call them now very high risk, high
risk, and so on.
But they were prepared to go up by a factor of 10.
Now, you might say, well, gee, they're already
starting at 5 10 to the minus -- no, 1.5 10 to the minus --
so, why can't they go to 10 to the minus 4 or a little
higher?
MR. BONACA: And they implied that they could
higher if they get management approval.
DR. APOSTOLAKIS: I guess the issue you're raising
is, even if the CDF goes up by some number, it's still not
clear that adequate protection is still preserved.
MR. BONACA: Absolutely.
DR. APOSTOLAKIS: I mean that's even higher.
MR. BONACA: The other issue is, even if you stay
within a certain limit, wouldn't just limiting the number of
components you're taking out of service mean good
management?
I mean there is the other issue that it doesn't
say that you have the liberty to go wherever you want, as
long as you don't meet a certain number. There is another
way to do it, which is to only limit the number of
components you're bringing out of service.
DR. APOSTOLAKIS: But then again, you are going
back to the deterministic way. I would be reluctant to do
that. I would like to explore the CDF and CCDP first.
Instead of calculating probabilities of minimal cut-sets,
count the number of events in there. So, let's be
consistent in our evolutions.
I think we should explore the CDF and CCDP issues,
see how far we can go with those, and if necessary, then
we'll go back and limit it more.
MR. CALDWELL: For those licensees that have real
strong management, that are interfacing with the plants on a
day-to-day basis, they're no different than we are, and
they're old school, too, deterministic approach.
For those licensees, they'll probably do that.
The manager is going to say I don't want the diesel -- I
don't want these six components being taken out at the same
time, I don't care what it says, that doesn't feel good to
me, and you know, until we have a better approach, we're
going to have to rely a lot on licensee management in order
to keep their plant safe.
MS. COLLINS: Some of them are pretty developed.
I mean they already have these kinds of limits. The limit
I've seen in the guidance that's coming out -- CDF -- it
says something like 10 to the minus 3 should not normally be
entered.
Well, the procedure I'm familiar with is not even
close to that. So, they're already way far away from that.
The other thing that I think is kind of
self-limiting is resources, taking these systems out of
components. Oftentimes, they have LCOs that -- they don't
have enough resources to take all this stuff out, equipment,
so I think it's naturally limited that way.
DR. APOSTOLAKIS: Let me ask you a question. In
your view, should the criteria be bounds on CCDP or CDF?
MR. PARKER: Our procedure has both in it.
MS. COLLINS: Yeah.
DR. APOSTOLAKIS: Very good.
MR. PARKER: It has a threshold of the ICCDP of
less than 10 to the minus 5 and ICCF less than 10 to the
minus 3.
So, it's asking the inspectors to look at that if
they exceed either of those thresholds, because some
utilities, like Laura pointed out, are using CDP, some are
using CDF.
But you want to -- CDP, I believe, would be
looking at the duration, and you want to factor that in
there.
DR. APOSTOLAKIS: If you say that you have an
ICCDP of 10 to the minus 5, that's almost two orders of
magnitude greater than what the NRC staff had proposed.
Now, you are NRC staff, too. The other staff.
Yeah, we have to really work on those things and
make sure that we have some consistency.
MR. PARKER: That's instantaneous, too.
DR. POWERS: When you look at these plants, do you
find them taking out multiple systems at the same time?
MS. COLLINS: We do find that there are multiple
systems or multiple components at the same time.
DR. APOSTOLAKIS: What's multiple?
MS. COLLINS: There could be two or three, but --
DR. APOSTOLAKIS: Two or three systems?
MS. COLLINS: Yeah.
MR. PARKER: Some plants may have divisional
outages and take out all their divisional equipment or any
maintenance on a particular division at a time.
MR. DAPAS: Train outages. They'll take out maybe
RHR and the charging pump, let's say, associated with the
same train.
DR. APOSTOLAKIS: But that doesn't defeat the
whole system, does it?
MR. DAPAS: Sure.
MR. GROBE: Sure. They'll do maintenance on
several systems on the same train.
MR. CALDWELL: Multiple systems within a given
train.
DR. APOSTOLAKIS: Is it fair to say, Laura, that
there is a need for guidance in all three bullets?
MS. COLLINS: Oh, yes, absolutely, and we know
that there are major changes coming to this. I mean we
already know that.
Of all the procedures that we have, this is
probably the one that is sort of newest to the resident
inspectors and where they need additional guidance, and I
think that's a well-known fact.
DR. APOSTOLAKIS: Okay. Thank you.
MS. COLLINS: When we go to slide 70, though, and
talk about inspection techniques, kind of the way -- what we
do -- we would probably select a planned work week, a week
or so in advance, or if it's emergent, you know, we don't
have that time, and we focus on that work that does involve
the risk-important systems and components.
We also tend to focus, I think, on unique
activities or first-time evolutions, and then we take that
safety assessment, we try to understand what the assumptions
are, we talk about the licensee's PRA staff, and their
operations staff, and the next week, perhaps when the work
is going on, we evaluate the plan and the safety assessment
against, really, the conduct of the work to make sure that
it's consistent, and this also applies to shutdown risk
assessments, where configuration of the plant is changing,
and we try to know up front what the assumptions are, this
has got to be back in service before we take this out.
Those are the kinds of things we would go out and check.
MR. CALDWELL: Laura said something about we'd
focus in on first-time evolutions.
I can tell you that once on-line risk started, the
majority of the transients or events that were caused were
because they transitioned from an activity they did while
shut down to an activity while they were operating and
didn't fully evaluate how they were going to get there, and
they either didn't tag out a component correctly or they
operated a piece of equipment the wrong way or whatever that
resulted in a transient.
So, it is a good area to focus on as they're
moving to on-line risk.
DR. APOSTOLAKIS: Let's say you have a plant
that's a 18-month cycle. If I look at a random -- at the
plant at a random time during that 18-month period, is there
a high probability that some on-line maintenance is going
on?
MS. COLLINS: Yes, every week.
DR. APOSTOLAKIS: Every week.
MS. COLLINS: Yes.
DR. APOSTOLAKIS: So, I wonder, then, whether the
-- what so-called baseline CDF is meaningful anymore. We
should revise it to take into account this plant's on-line
maintenance.
MR. DAPAS: Supposedly, the SDP accounts for that.
DR. APOSTOLAKIS: No, no, no, the baseline, the
PRA itself.
MR. DAPAS: When you look at, if a component is
out of service, what's the additional contribution to the
baseline CDF, and there's some assumed amount of
out-of-service time associated with that.
DR. APOSTOLAKIS: What I'm saying is you don't
have a baseline. If your baseline is moved to the point
where you have something --
MR. SIEBER: You already have assumed a certain
amount of outage time per component.
DR. APOSTOLAKIS: Not with on-line maintenance.
MR. GROBE: With on-line maintenance, if you look
at the fault tree, there's some component for equipment out
of service time, which can be on-line, can be shutdown.
DR. POWERS: What George is saying, I think, is
that that's gotten kind of averaged over the entire year,
and in truth, it's peaked, it's spiked, and so, now he's
moving from spike to spike with maybe a little trough in
between or something like that.
DR. APOSTOLAKIS: What we used to call baseline
CDF perhaps is not baseline anymore.
MR. DAPAS: It may not truly capture the risk
posture of the plant at the time a piece of equipment is
taken out of service.
DR. APOSTOLAKIS: This is a very interesting
thing.
MR. BONACA: If they showed that they were
integrating that value, as I've seen other plants do, to
assure that you stay within the assumed unavailability in
the IPE. So, I mean there is a self-controlling mode.
DR. APOSTOLAKIS: No, but what they showed us was
that there was a line that said this is 1.5 10 to the minus
5, our baseline, and here we had a spike because we did
this, then we had another spike because we did something
else.
Now, Laura is telling me that actually they should
have spiked every week.
MR. PARKER: There are typically spikes every
week, but I think you're right, they generally --
DR. APOSTOLAKIS: If you have a lot of spikes,
then --
MR. PARKER: It has to balance out, because
they're looking at the availability and the un-availability,
and that all should be modeled appropriately within the
scope of the PRA.
I understand what you're saying as far as the
spikes, and we need to look at it in a different context.
MS. COLLINS: The other part of the maintenance
rule is sort of their annual assessment where they're
supposed to be looking at that, and we also go in -- and the
concept of balancing the unavailability and reliability,
which I guess we assume that, if they meet their performance
criteria for those systems and components, that they've
achieved that goal.
So, that's under a different inspection procedure
that's done by Division of Reactor Safety. They do that
once a year.
DR. APOSTOLAKIS: Okay. I got the answer.
MS. COLLINS: Okay.
Page 70, the last bullet, I say consult with
senior reactor analysts. If we have some kind of an issue
-- and I say we haven't really decided what a finding in
this area is -- the SDP doesn't apply to these findings.
So, my understanding is that there is a SDP for these kinds
of issues under development in NRR.
To date, we've just been using our best judgement
and the judgement of the SRAs.
DR. POWERS: Your understanding is our
understanding, and you've apparently seen just as much as we
have.
MS. COLLINS: Okay. But I guess the good thing
is, throughout the pilot program, we've seen pretty good
programs with risk assessments, and we haven't identified
what we believe to be any significant issues.
DR. POWERS: That does seem to be what we see.
For these planned outages, they're doing good work, they're
doing real good work, and there's an economic incentive,
because people that do well-planned, well-thought-out work
have short outages, costs less money to get more done.
The difficulty is what about unplanned and what
you call emergent events, and how well is that going to be
done, and I don't have a handle on that.
MS. COLLINS: I think in our experience we've seen
it done pretty well, but we don't know what's coming.
DR. POWERS: Your experience is extremely
important to me, because you have an experience that I
don't, and so, I take your word very sensitively.
MR. DAPAS: There is a spectrum of performance
depending on the licensee.
DR. POWERS: I'm sure that's true, but I mean if
the general feeling is, hey, they're doing a pretty good job
here, then I'm going to worry a lot less about it than oh,
my god, can I tell you some horror stories.
MS. COLLINS: There are a couple of areas that I
think are of interest to us, and that is how the licensee
might evaluate initiating event frequencies or
probabilities, which is kind of what I'm seeing in the
guidance.
Other than weather-related, impending weather kind
of problems, I don't necessarily see a lot of that, and I'm
not sure how that will be done. So that's another area that
I think we'll explore.
On page 71, inspection observations, again, I said
--
DR. APOSTOLAKIS: Yeah, the second bullet -- would
you elaborate a little bit? We don't have to go through all
of them.
MS. COLLINS: Right. We have seen where the
duration of the maintenance exceeds the planned duration,
but if it doesn't exceed an LCO, there isn't much
involvement we have other than a comment.
DR. APOSTOLAKIS: But this is common? Is this a
common occurrence?
MS. COLLINS: No, I wouldn't say it's common.
MR. BARTON: It happens occasionally, yes.
MS. COLLINS: But we've seen it.
MR. BARTON: Because you have a system outage
scheduled for 36 hours and it ends up 42 for some kind of
problem, and that happens not too infrequently.
MR. DAPAS: And I just wanted to comment -- this
has brought to bear an issue where the procedure would ask
us to assess is the actual time to execute the maintenance
greater than planned, okay?
You're asked to look at that as part of the
inspection procedure. Then what do you do with that,
because does that really translate to an increase in risk,
and they're within the LCO time and they may be within the
time assumed as part of your baseline CDF.
What do you do with that, and that's one of the
questions that we've been wrestling with with the program
office.
MR. BARTON: I think you understand why it is it
happens, and if it's the same cause that always happens,
then you've got an issue.
MR. DAPAS: You're right, but again, the result of
that has to be some increase in CDF that crosses some
threshold where you can land that issue with the licensee
and engage them, versus an observation per se.
DR. POWERS: It's like drunk driving convictions.
The penalties are very severe in New Mexico for the second
one, but since they always excuse the first one, nobody ever
has a second one.
MR. PARKER: We've seen that happen on occasion,
and the SRAs have gotten involved on a few of the issues
where the licensee's risk assessment assumed, let's say, 36
hours on a 72-hour, and they had some bounding analysis, and
now, because of parts availability or some additional
concern, they might have went up to the 72-hour, and so,
we've asked the residents, that this is a good opportunity
to challenge the utility on their risk assessment and their
bounding analysis and go back to risk assessment and see if
the licensee is comfortable with the new numbers, where it's
taken them.
DR. POWERS: It's also a good vehicle for asking
them about the uncertainty in their analyses, what kinds of
things did they think about that might change their numbers?
DR. WALLIS: If this is a best estimate, then half
the time the duration will exceed the plan, roughly
speaking.
DR. POWERS: Based on the reports I see, I think
most maintenance is less than the plan.
DR. WALLIS: Less than the planned time?
DR. POWERS: Yes.
DR. WALLIS: So, it isn't so bad that a few take
longer.
MR. DAPAS: Getting back to Mr. Bonaca's point, I
would offer that a licensee that is managing the risk would
say, okay, if we run into a problem, then here is the risk
if it takes 72 hours versus 20, and that's a sufficient
increase, now we want to doubly insure we've got parts and
we've, you know, done mock-up training or what have you to
ensure the actual time is bounded.
DR. WALLIS: But surely all you're really
interested in is the average over all the maintenance you
do, and the fact that some may take longer and be a bit more
risky doesn't matter, as long as it's compensated for
throughout the year or whatever by the others that take less
time.
MR. CALDWELL: We probably have a little more time
than we anticipated. O'Hare is closed right now. So, we're
calling on your particular flights to find out what that
actually means.
My secretary is going to call and check and see if
it means they've been canceled, delayed, or whatever, and
then we'll let you know.
MR. GROBE: We can give you a nice list of
restaurants.
MR. CALDWELL: Flying out of O'Hare and into
National, which is what we do when we go to headquarters,
your chances of one of the two of them getting there is 100
percent.
DR. POWERS: Now, I know why the risk analysts
here are so busy. They're calculating the probability the
boss is going to get back.
MR. GROBE: Other questions on 71?
[No response.]
MR. GROBE: Mike, do you want to go into a little
bit of what you're doing?
MR. PARKER: Yeah. I just wanted to take a little
time and go through some of the observations that Sonia and
I had during our SRA site visits.
We went out to all the sites over the last --
probably -- I think it was six months to a year ago, and we
went to each one of the sites together as a team and tried
to get a pretty good idea of what tools the licensee has,
how they're using them, and how they're integrating into the
organization.
So, it was more of an observation visit to
introduce ourselves, to go through the new inspection
program, and how we're going to -- how we would like to deal
with them on risk issues, but some of the things we found --
on-line risk assessments -- most of the utilities were using
a probabilistic risk assessment such as Safety Monitor,
EOOS, or Sentinel.
There were a few outliers out there that are still
using deterministic. In other words, they're still using a
matrix or procedure to look at things, and it's more of a
defense-in-depth-type approach, and some of them also have
pre-solved cut-sets that they're using on some of their
on-line monitors, but it looks like quite a few of them,
including Commonwealth, is moving to some very good systems.
They're going to Sentinel at the Commonwealth facilities.
So, most of the utilities are using risk programs,
and there's, I think, one or two outliers right now in our
region that are still using matrix procedures.
Shutdown risk assessments -- the majority or all
of them at this time are deterministic. Several of them are
matrix procedures with defense-in-depth, and I'd say the
majority at this time are using an ORAM-type program, outage
risk assessment matrix, and that's defense-in-depth.
We have seen a couple plants that are in the
region -- I mentioned Perry as an example -- that are
looking at developing shutdown models right now. So, that's
going to be very interesting seeing a full-blown shutdown
model and how they're using that and integrating it into the
organization and into outage planning.
So, it will be a very good tool, but they're
completing that. They expect to use it the next outage,
which is in February, and they're hoping to use it for some
of their pre-planning activities right now, and they're
going to tie it back in to -- they need to do some
conversion and put it into Safety Monitor.
So, that will be one of our first plants in our
region.
I know several plants out west are using shutdown
models.
So, that will be very interesting.
As far as risk assessments, most of the utilities,
I think, are doing some very good risk assessments.
Generally that's involved with the work week managers and
not the PRA organizations.
Generally what we've seen is the PRA organization
or the corporate staff develop the tools and put them in
place and then it's turned over to the line organization to
look at normal work activities, and it's not until they've
determined that they have a risk-significant configuration
that they may have the PRA organization get involved and
deal with the issue and look at the acceptability or
challenge the model.
MR. BARTON: Well, don't they -- if they have any
changes at all to that planned maintenance, don't they
bounce that back off their PRA groups?
MR. PARKER: Right.
MR. BARTON: Okay.
MR. PARKER: But some of the organizations will
have the line organization where they'll put it in the
schedule and then run the program, and as long as the
program is, let's say, a green baseline, they won't get
involved.
So, to address George's question as far as what
happens if they have a higher risk, do they try to balance
that, some of the plants do, other plants will have like a
12-week rolling average or rolling schedule, where they have
certain equipment that comes out periodically, and they will
try to stick with that equipment at that timeframe and to
complete that 12-week cycle. So, they've looked at certain
combinations of equipment that they would like to take out
at the same time.
So, they'll try not to manipulate that equipment
and put it into a following week.
As far as integrating risk assessments, I think
Laura mentioned that, in general, the information we're
familiar with is licensees are doing a pretty good job at
integrating their emergent work with the pre-planned, and
we've seen a lot of occasions where the licensee has
pre-planned activities, some equipment to identify
degradation. They'll put off or defer or cancel some of
their pre-planned activities so as not to incur that
additional risk, and so, we've seen some good indications of
that, which makes us feel pretty comfortable.
Maintenance rule (a)(4) -- as Laura mentioned,
that's not out yet. I think that's supposed to take effect
in November.
There are some direction coming out NRR right now.
There's two visits planned for Region III.
There's a visit, I think, in the next few weeks that's
tentatively set up to go to Braidwood, and with the region's
assistance, they want the SRAs, the regional inspector, and
then headquarter involvement just to see how Braidwood does
activities. I think Braidwood was picked because they
indicated they think they have a pretty good program in
place. So, that's a one-day visit.
And then the other activity that's being planned
is more of a comprehensive V&V inspection, and that's
planned for Clinton, and that would be more than likely --
and I'm somewhat speculating, but I think it's to actually
have the draft TI and see how the utility does things. So,
it would be somewhat of a pilot or just maybe go through the
exercise and see how our procedures develop.
MR. DUNLOP: I just got off a phone call a little
while ago about the (a)(4) rule. NRR is not really going to
prepare a TI. What they're going to do is -- in the
verification -- is re-validate the new Attachment 13.
So, during the first survey visit to Braidwood,
they'll figure out what kind of -- and at all the other
regions -- they're doing five surveys -- they'll go out,
look at the different types of assessments that the
licensees are doing, come up with a new or revised
inspection program procedure, and then, during the four
verifications -- ours is at Clinton -- verify that the new
procedure will work and it will be acceptable.
That's one change that we just found out, that I
had just found out today.
MR. PARKER: That's Andy Dunlop. He's with the
maintenance rule in Region III.
One of the challenges I think we're going to have
in the maintenance rule -- and like we said, we don't know
where it's all going, but we have some thresholds, there's
some thresholds in some of our reg guides and other
guidance, but I don't know how we're going to deal with the
fact if the utilities exceed those thresholds and how they
balance that and what tool do we have to encourage the
licensee to reduce that overall risk, and so, those are
questions that we have outstanding and we'll be involved
with the development of these activities.
As far as risk assessments, I think, since the new
inspection program, there's been significant implementation
of the licensee evaluating risk.
In the past, as far as events, we've challenged
the utilities, and we didn't see that they were truly
assessing it.
So, I think the new revised oversight program has
really forced the utility to look at some of those emergent
work activities and the impact it has or transients, and
we're seeing significant involvement on the part of the
utility to assess that, and Sonia and I are actively
involved in looking at the impact particular transients have
on the plant, overall risk, and communicating that in our
morning meetings and other avenues that we have.
We've also seen the utility and we've been
strongly encouraging the utility to address the risk
significance in LERs.
An LER asks the licensee to talk about safety
significance of the event of interest, and we're seeing the
utilities taking an opportunity to address what they
characterize as the overall risk significance of the
activity as part of the safety significance.
So, I think that helps the region and anybody
that's following that particular activity to put it in
perspective. It gives the licensee their first shot, and
then certainly the residents and the SRAs are evaluating the
risk significance of LERs.
The last thing is -- Sonia has already talked
about how we're involved with the SDP process in the phase
two.
Is there any questions?
That's all I have.
MR. BARTON: I want to thank you all for a real
informative session and thank you for the work you've put
into it.
I think, of the visits we've made, this has
probably been one of the best if not the best, from my
perspective. I don't know how the other members feel, but
it's been very informative and a good dialogue and we
learned a lot.
DR. POWERS: Yeah, I'd say that the meeting far
exceeded expectations. I think it was an extremely good
discussion among colleagues in these areas, and we got some
things for us to go puzzle about.
I reiterate my belief that the wealth of
experience that needs to be injected into this process,
especially as we look to the next year of refining some of
it, because you guys are really finding the rough edges, and
I don't blame the people that put these new systems
together.
They had millions of things to take into account,
and they did a wonderful job doing as much as they did, and
they knew they weren't going to get all the rough edges, and
so, now, it's a process of making sure we find out about all
those rough edges and do things, and what we just heard
about on this maintenance rule business is something I
hadn't anticipated.
We've clearly got to think about that a lot in the
coming weeks.
So, it's starting to make me think. This is
difficult, but I really appreciate it, and we had a
fantastic visit out at Davis Bessie. They really pulled the
stops out for us.
MR. CALDWELL: I thoroughly enjoyed this. I
learned quite a bit.
I wanted to compliment the staff, those folks that
are here. They did an excellent job, I thought, and Marc
and Jack, and I certainly appreciate that, as I understand
you did.
MR. BARTON: I think that's what was better. In
our past visits, we've heard from the management of the
region, and I think what was great today is we really heard
from the people that are out there involved in the process
and doing the work and having the interface with the
licensees.
MR. SINGH: I just want to thank you, especially
to Bruce Burgess, for his hospitality here. He has been
really helpful, and he has worked since last October to
arrange all this.
So, I really appreciate his help.
MR. CALDWELL: I think I ought to tell Bruce I
appreciate it, too, because I jerked him around a bunch
today, and it came out relatively smooth.
MR. BARTON: On that note, the meeting is
adjourned.
[Whereupon, at 3:08 p.m., the meeting was
adjourned.]
