ACRS Meeting on Fire Protection - January 20, 1999
UNITED STATES OF AMERICA
NUCLEAR REGULATORY COMMISSION
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
***
ACRS MEETING ON FIRE PROTECTION
***
U.S. Nuclear Regulatory Commission
11545 Rockville Pike
Room T-2B3
Rockville, Maryland
Wednesday, January 20, 1999
The above-entitled meeting commenced, pursuant to notice, at
8:33 a.m.
MEMBERS PRESENT:
DANA POWERS, Chairperson, ACRS
GEORGE APOSTOLAKIS, Member, ACRS
THOMAS KRESS, Member, ACRS
DON MILLER, Member, ACRS
. P R O C E E D I N G S
[8:33 a.m.]
CHAIRMAN POWERS: Let's bring the meeting to order. This is
a meeting of the ACRS Subcommittee on Fire Protection. I am Dana
Powers, Chairman of the Subcommittee for Fire Protection.
ACRS members in attendance are George Apostolakis, Tom
Kress, and we expect to be joined by Don Miller.
The purpose of the meeting is for the subcommittee to review
the reactor fire protection activities, the insights gained in the fire
protection area from the view of licensee submittals of Individual Plant
Examinations and External Event Reports, result of the Pilot Fire
Protection Functional Inspections, Electrical Circuit Analysis, proposed
NFPA Fire Protection Standard and related matters.
The subcommittee will gather information, analyze relevant
issues and facts and formulate proposed positions and actions as
appropriate for deliberation by the full Committee.
Amarjit Singh is the Cognizant ACRS Staff Engineer for the
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 December 18th, 1998.
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 that they can be readily
heard.
We have received no written comments or requests for time to
make oral statements from the members of the public.
This is going to be an action-packed meeting of the Fire
Protection Subcommittee. It seems that both the Staff and the industry
are hyperactive in this area in recent weeks.
We have invited Dr. Miller, who is the Chairman of the I&C
Committee in our answer to Reddi Kilowatt for the ACRS to assist us in
what is rapidly becoming one of my favorite issues, both now and when I
was 17, the issue of Hot Shorts.
[Laughter.]
CHAIRMAN POWERS: As this issue has unfolded, it looks like
it may be a good deal easier to resolve than I had anticipated. The
industry positions include licensees have effectively implemented the
applicable regulations and the issues raised by the Staff either do not
exist or are outside the regulatory requirements or licensees have made
a good faith effort to implement the applicable requirements but
implementation may not be 100 percent for a variety of reasons including
mixed guidance from the Staff or circuit analysis issues raised by the
Staff are covered by existing regulatory requirements and to the extent
they have not been implemented, additional actions by NRC is warranted,
and my understanding of course is the Staff has divergent views on
these. It sounds like an easy issue.
One of the issues that is emerging in connection with fire
shorts is the issue of risk analysis and I will be particularly
interested in Professor Apostolakis's views on the risk analyses,
because to my mind a lot of the risk analyses that have been presented
seem to be persuading me that Browns Ferry did not occur.
The next issue we will be dealing with is the Fire
Protection Functional Inspections. This is an evolving issue, it seems
to me. It seems to be raising questions about what the Staff is
inspecting and what it should be inspecting in our fire protection
areas.
Of course, one of the highlights of this meeting will be the
NFPA draft standard on fire protection. This is the much-awaited
performance-based fire regulation. It is, however, a far cry from the
one-page regulation that Professor Catton used to argue passionately
for. In fact, I would characterize the NFPA standard as the Prego
Spaghetti Sauce Standard. If you like defense-in-depth, it is in there.
If you like determinism, it's in there. If you like performance, it's
in there. If you like risk, it's in there -- and in fact I think Dr.
Kress will find substantial validation of some of his views on
performance-based regulation embodied in this draft standard.
The topic of focus in tomorrow's session will be the
much-awaited Research Program, and it is the chore before the Committee
to develop its input to the ACRS's report on research based in part on
this material, and I will probably ask Dr. Kress to draw upon his wide
breadth of experience and preside over that session.
With that introduction, and insights on all that we have to
accomplish, I want to turn now to Tad Marsh to give us some additional
insights and perspective on this with his Overview of Fire Protection
Activities within the NRC.
MR. MARSH: Thank you. Good morning. My name is Tad Marsh
and I am Chief of the Plant Systems Branch, and I do want to say I am
looking forward to this meeting. I am looking forward to the
discussions and I am looking forward to synergism and there's a lot of
very good topics and I am looking forward to this action-packed meeting,
I really am.
CHAIRMAN POWERS: Tad, let me just interrupt you. I omitted
discussion of some of the real groundrules for the meeting.
Subcommittee meetings are the one opportunity the ACRS has
to get into depth into an issue and we have an agenda and a schedule,
but I think speakers should concentrate on getting their points across
to us and let Mr. Singh and I worry about the schedule. We are
scheduled to go to 4:30. I had promised the members 5 o'clock and I
guarantee them 6:00, so we have got some flexibility here, and I'd
comment that members of the audience should feel free to interject
points that they think would help clarify our understanding here, but
recognize, however, that I have my back to you, so that if you want an
input, get attention, but feel free to do so because this should be a
very collegial exchange. I believe the Staff feels that way. I
certainly feel that way, so that we can get as much input as we can in
this session.
MR. MARSH: Thank you. Thank you very much. I am glad you
said that because I have said the same thing to our staff, especially in
the context of the NFPA 805, which is a document that is going through
change and is being developed and about which we have mixed views, and
so as we present or as industry presents things, I have encouraged folks
to go ahead and chime in. I am looking very much forward to that
discussion. We were dry-running this yesterday it's going to be good.
But I want to step back for a minute, talk about the big
picture on fire protection and what's been going on because it's been a
very active, very fluid situation in terms of fire protection from the
last time that we spoke.
There have been three forces which have been guiding us and
changing us in fire protection within the last year. One main force is
of course the Agency's emphasis in becoming more risk-informed in
everything that it does. You will hear as we go through the
presentations risk elements in almost each one of these presentations,
and well it should be.
You are also going to hear some of our frustrations and some
of the difficulties associated with going in that direction and that
will become clear to you.
The next force is this new assessment process the Agency is
working on. That has a strong influence on where we go on the fire
protection functional inspections, and it even plays out in some of the
circuit analysis issues which we are going to talk about.
The third force which has been guiding us and moving us in a
direction this year is not a new force but it is one which has been
emphasized by the Commission's interactions with stakeholders, et
cetera, and that is working more closely with the industry in some of
these endeavors, and in all of the efforts you are going to hear about
today there's been work with the industry -- sometimes a great deal of
interaction, sometimes less so, but there have been a lot of
interactions with the Agency.
I'll start by giving you an overview of the fire protection
functional inspection, and that is going to be Pat Madden giving his
presentation. This is I think in my opinion one of the most important
things we have done this year, maybe the most important thing we have
done this year, because we have looked very hard at four plants and the
state of their fire protection programs, and we have come with findings
about how we think things are.
As we expected, there are weaknesses there in programs. Now
the challenge is to decide where to go, what is the future of the FPFIs,
and that is a decision that the Staff will have to make a recommendation
to the Commission about. That decision will be probably some time in
the early spring. We have a paper which is due to the Commission -- it
was intended to be at the end of this month. It's now been delayed
until next month, the February timeframe. We did that because we needed
industry input. We needed industry recommendations on how they see the
FPFI program, what views they may have on how to go in the future, and
how that should and could fit into the final assessment process which
the Agency is working on.
Enforcement issues in the context of FPFIs have become
something which we have focused on. The Agency is using risk insights
in enforcement issues and that has become a challenge, and so as these
FPFI findings are to be processed through the enforcement process, the
risk associated with those findings has become something that we are
striving to determine and it is again an area which is not easy, and you
will hear some discussion about that as we go through the presentations.
CHAIRMAN POWERS: One of the things that comes clear to me,
I think should be clear to me, maybe you can correct me if I am wrong,
is that FPFI is really raising questions about how we organize our
inspections and what should be inspected and what should not be
inspected.
Now you are raising a different issue -- do we have the
tools that will allow us to actually assess risk in what happens to be a
context of enforcement but risk in general --
MR. MARSH: Right.
CHAIRMAN POWERS: -- here in a reliable and believable
fashion.
My sense is that we are a little inconfident in that area,
not because we know that the tools are inadequate but because we know
they have not received the kind of scrutiny we have applied to tools
such as we use for normal power events and things like that.
MR. MARSH: Right. I would agree.
CHAIRMAN POWERS: And I think that is where tomorrow
afternoon's session, where Nathan Siu will discuss both the Research
Program and the IPEEE, but I think we will interrogate him quite
closely, because I think he is knowledgeable in the status of these
tools, what we ought to be doing and where we ought to be going.
One question to present to you, as an overall question for
the meeting, Tad, is are there things that we should be planning to
bring forward to the ACRS either at its February meeting or its March
meeting. It sounds like the March meeting is appropriate for this paper
on the functional fire protection --
MR. MARSH: Right.
CHAIRMAN POWERS: -- inspections, but if there are things
that you think should be brought forward to the ACRS or the members
identify to bring forward, let's flag those immediately.
MR. MARSH: Okay, fine. Can I -- let me hold that question
and we can think about it as the meeting goes on.
CHAIRMAN POWERS: Yes, that is an overall question for
whenever it comes to mind.
MR. MARSH: The FPFI program is -- I don't want to overuse
the term but is "out in front" of a lot of Agency efforts and it is
maybe symbolic of a lot of Agency change. It is a major inspection
effort. We are not doing very many of those any longer. It came from
concerns that were older concerns but they dealt with the adequacy of
fire protection programs and features coming from the Thermo-lag
experience and self-assessments.
Now that we have done those inspections, again we are right
on the cutting edge of how do we factor risk into whether a program
should continue, how do we factor the future into this assessment
process, how do we work with the industry in coming up with a viable
program which is a better use of resources and a better, more
coordinated effort -- and risk itself -- in this program, and
enforcement actions.
I led off with FPFIs as perhaps the most important thing
that we have done this year because it is out in front so much.
CHAIRMAN POWERS: I think your assessment is correct, by the
way, and I'll also point out that as we moved in this area in risk we
were also running into the difficulty that the public has when presented
with risk assessments. They see things that from a lay deterministic
viewpoint seem to be risk-significant --
MR. MARSH: Right.
CHAIRMAN POWERS: -- but from a quantitative point of view,
a systematic analysis, may not be, and discontinuity from perception and
quantitative analysis seems to come forward more clearly in fire
protection than any other area, perhaps because the lay public
understands the threats of fire better than it understands other
threats.
MR. MARSH: I think that is a good insight.
We have stylized ways of analyzing scenarios in accident
analyses in Chapter 15 and in licensing bases. We analyze LOCAs, we
analyze feed line breaks and steam line breaks and we have Standard
Review Plans that tell you how to do that, and we come up with answers
about whether things are protected and whether the DMBRs are exceeded
and whether the PCT limits are met, and those types of stylized,
artificial measures of safety of plants, and they construct procedures
and features and measures to protect against those stylized ways.
Well, fire protection is another stylized licensing bases
type of construct. We have layers of defense-in-depth and we come up
with answers about whether systems and features are adequate to protect
against these challenges, but we lack the opposite approach that we have
in normal accident analyses.
We have pretty good tools to assess risk associated with
LOCA and feed line breaks, and we don't have that counter approach in
the fire area, and from the public's perspective fire is something that
people understand. Fires are things that they see, that they hear
about, and it presents a personal challenge to them, unlike LOCA, feed
line break, steam line -- people can't empathize with those scenarios.
They don't know what that means.
CHAIRMAN POWERS: They don't have LOCAs in their house very
often.
MR. MARSH: But they understand fire.
The second issue which we are going to talk a lot about is
one Steve is going to present, and this circuit analysis and that has
presented a unique challenge too, because there are disagreements with
the industry over what the Staff now believes to be the licensing
requirements for plants, what the regulatory requirements are, as
opposed to what the industry has presented to us as being what they
believe the Staff's licensing requirements were.
Risk plays an element here too. You will hear from Fred
Emerson, NEI, tomorrow I believe, or perhaps later on today industry's
approach in resolving this problem, which has embedded in it a risk
element as well as a stylized deterministic element and which again
encompasses risk. There is a challenge here in understanding and trying
to evaluate the probability and the consequences of some of these
circuit failures, so again adequacy of tools has become an issue.
I am looking forward maybe most to our discussion about 805
because it is so new, and it is so novel, and it present discussions
about how licensing requirements and fire protection requirements
compare to this new document, and it will be a lively discussion.
I mean I think that it is a difference construct of fire
protection requirements that we have heard before, and so it will evoke
all kinds of questions about circuit requirements and baseline
requirements, and what is important, and how licensees could actually
use this. That will be an interesting discussion.
Another thing that is on our plate that I want to make sure
you know, and I think we have talked to Jit about this, is the
Commission meeting that is coming up February 9th, okay. We will be
talking with them about the NFPA 805 document, the Comprehensive Reg.
Guide, which is another important program we have ongoing, Quad Cities,
the results of that evaluation and our assessment of that situation, and
more discussion about the status of the research programs.
Well, I didn't want to take much more time than that. I
wanted to give you an overview, and I hope I have. And the thread that
I would like you to see in these discussions is the three forces that
are involved in each one of these endeavors, okay. And the first force
is the effort for the agency to move in a more risk-informed way in
many, many things that we do.
The other force that is involved is the assessment process,
this new assessment process. We are not going to spend a lot of time
talking about that because we are not experts in it and because it is
evolving, but it is pushing us in a way, especially in the FPFI program,
cornerstones, performance indicators, things of that sort.
And the third force is working with the industry. In all of
these endeavors we had been and been successfully working with the
industry on these programs.
Okay. The first presentation is Ed. He is going to be
talking about the status and development of the proposed Reg. Guide, a
Comprehensive Reg. Guide on fire protection.
DR. POWERS: Before Ed gets started, I will just comment,
Tad. I think that you have identified correctly the forces that are
operating now, but I am willing to bet that as we move further in
applications of risk, and as we move away from the traditional
deterministic examination of fire protection, we will see an emerging
force in the form of public dissatisfaction and legislative
dissatisfaction with the persuasive power of risk as a metric of safety,
and that is going to be a force, not active now, or only perceptible
now, but one that we are going to have to confront directly as we go
forward. Because, as you have accurately pointed out, people do
understand fire.
And they are going to find surprising the kinds of
conclusions that emerge from risk assessments, especially in situations
where you have established an elaborate protocol of defense-in-depth so
that you can tolerate deficiencies in one or more elements of that, or
you can tolerate tradeoffs between the elements of defense-in-depth and
still maintain a bottom line safety. But it is going to be
contra-intuitive to a lot of people, and you are going to find people
grabbing ahold of elements of that, and castigating -- claiming to see
how ridiculous this is. Can't possibly be safe because, as everyone
knows, you know. Fill in the blank, there will be lots of things there.
I think that is going to emerge as an issue, and it is not
going to be one that is going to involve presentations in front of
committees like this. It is going to be that manifests itself within
the legislative branch.
MR. MARSH: I agree.
DR. POWERS: Ed, welcome to the meeting, and you are going
to discuss the Comprehensive Fire Protection Regulatory Guide. You come
here with some courage, because I think it is well known that this
subcommittee was not in favor of this activity.
MR. CONNELL: Well, maybe I can change your mind a little
bit. Good morning, my name is Edward Connell, Senior Fire Protection
Engineer in the Plant Systems Branch at NRR and I have a little
presentation this morning on the Comprehensive Fire Protection
Regulatory Guide.
A little refresher background information, as most are
aware, the existing NRC guidance on fire protection is scattered through
many, many documents, different kinds of documents spanning many
generations, even some of it before the Brown's Ferry fire in 1975, and
I think --
DR. POWERS: None of it pre-dated Jit, right?
MR. CONNELL: No, no. No. And this is just to give you an
idea. We have about 17 Generic Letters, seven Bulletins, an
Administrative Letter, 71 Information Notices, and I am not going to go
through all them, just to kind of make my point that there is a lot of
stuff out there. It takes up five binders on my desk, so --
DR. POWERS: It takes up five binders on my floor.
DR. APOSTOLAKIS: Why?
MR. CONNELL: You are asking why does it take up five
binders?
DR. APOSTOLAKIS: Yes.
MR. CONNELL: Because there's over 125 documents.
DR. APOSTOLAKIS: And why are there over 105 documents?
MR. CONNELL: Because that's how many were issued.
DR. APOSTOLAKIS: And why were they issued?
MR. CONNELL: You are asking me why stuff was issued in 1975
till 1998? I don't know. The staff saw a need to issue guidance,
Information Notices, Generic Letters, internal memoranda, Administrative
Letters, Bulletins, Standard Review Plans, Branch Technical Positions.
The staff saw a need and issued that guidance.
DR. APOSTOLAKIS: I mean what is so special about fire
protection that led to all this? I mean did this happen in other areas?
DR. POWERS: Well, I think what you find is that -- two
things. Yes, fire is special. There is no question about it, fire is
special. And that you find that our learning about fire has been
different than it has been about LOCA and nuclear incidents, that we
thought we had a handle on fire and we have thought we had a handle on
fire, every year we think we have a handle on fire, and we discovery new
things, a new diversity of things. At the same time, there have been
imperatives, and I think one of the presentations that is in your
briefing book accurately reflects the imperatives that come along and
cause crunches to get things out, so it has led to an incrementalism in
documentation and whatnot. And this is -- Ed is discussing an effort to
try to wrap this up into one package.
MR. CONNELL: Well, I think, you know, the objective is to
simplify and consolidate the existing guidance into one place, because,
let's face it, we have guidance that spans over 20 years. And if you
look at the second bullet there, you know, some of the existing
guidance, no surprise over a period of time like that, with over 125
documents, some of the existing guidance conflicts. And I pulled out
three little examples, these are simple examples of where the guidance
conflicts and what the difference would be depending on which guidance
document is being used by the licensee or the staff to review a
particular program.
So I think it is worthwhile to try to do a scrub of that
guidance and simplify it. My goal would be to reduce the five binders
down to one. About 25 percent of what we currently have, as far as
number of pages, and I will be satisfied if we get down to about a
third.
DR. POWERS: Set aggressive goals, Ed.
MR. CONNELL: Well, yeah, sure. And revise them as
necessary as you go along.
DR. APOSTOLAKIS: I still don't understand. Why will you be
able to do that now and you were not able to do it 15 years ago?
MR. CONNELL: Well, there was no attempt to do it. Well,
there was one attempt to issue a Reg. Guide many years ago in the early
'80s. The staff did issue a draft Reg. Guide, Reg. Guide 1.120, that
attempted to do this, and then it fell off the table. But that was many
years ago, George, and not many of the people that are here today were
involved 20 years ago in the project, so I can't tell you why a specific
document was issued 20 years ago. The staff felt there was a need, they
issued the guidance. We are trying to do a scrub of that now, we think
it is time.
DR. APOSTOLAKIS: Is the state of knowledge now different,
that it will allow us to do this?
MR. CONNELL: That is not the concern. The issue is we are
basically, with this project, we are going through the existing
guidance, looking where there's holes, and I will give you some examples
further along in my presentation where we think that the guidance is
either not there or is lacking, but it is not a state of knowledge. It
is not something like, well, we have PRA today, we didn't have it 20
years ago, so that impacts this guidance. That is not the point here.
MR. MARSH: One of the forces that drove us to do this was
an acknowledgement on our part, a concern on the Commission's part and
on the public's part about the clarity of the fire protection
requirements. They have existed over a long span of time. They have
evolved. They are reactionary. They are many volumes, and it is not an
easy subject to explain, okay, because there are so many documents out
there and so many requirements out there.
So the effort here is to try to synthesize and to be more
cogent in what the requirements are. I am not sure that there are other
programs out there that could also benefit from this type of effort,
too. I mean I think EQ is an area that -- environmental qualification
is an area which suffers from evolving guidance, NUREGs, Standard Review
Plans, things that are difficult to explain. And the effort here is to
clarify what the licensing requirements and the agency's guidance is.
DR. APOSTOLAKIS: Well, in one of your earlier slides, as an
example of conflicting or inconsistence guidance, you had the maximum
length of hose per hose station, 75 feet versus 100 feet.
MR. CONNELL: Yes.
DR. APOSTOLAKIS: So what you are doing now is you are going
to clean up the regulations and say the maximum length of hose is X?
MR. CONNELL: Well, we are going to clean up the guidance.
Most of this stuff is not in the regulation. We are going to clean up
the guidance so that we don't have three different, conflicting guidance
for a particular topical area.
DR. APOSTOLAKIS: But you are not actually asking yourself
whether giving guidance on the maximum length makes sense.
MR. CONNELL: Well, we are asking that, too. Yes, we are
asking -- if we decide that maybe we don't need to provide guidance as
far as maximum hose length, we will take that away. That will also
eliminate the conflict.
DR. APOSTOLAKIS: And this will come from the current state
of knowledge, I assume.
MR. CONNELL: Yes.
DR. APOSTOLAKIS: Which is risk-informed and
performance-based?
MR. CONNELL: Well, I don't you are going to come up with
hose length out of a PRA. I don't think that is going to give you that
kind of information.
DR. APOSTOLAKIS: But the PRA perhaps will tell you whether
you need such kind of guidance?
MR. CONNELL: I don't see how the PRA is going to tell you
whether you that need that kind of guidance. Most PRAs don't credit
minimum suppression capability.
DR. APOSTOLAKIS: But will you attempt to identify the needs
form a risk assessment perspective?
MR. CONNELL: We will identify the needs from the feedback
that we get internally and externally within the agency and with the
stakeholders, the industry and the public, identify those topical areas
that they think need to be addressed in a fire protection guidance
document.
DR. APOSTOLAKIS: Okay.
MR. MARSH: I don't think -- the idea, as we had envisioned
it, was to take the existing unclarity and clarify. Not so much to go
forward, you know, and to superimpose or to change the guidance based on
new tools and technology, because we don't know yet how that is going to
play out, the 805 is going to play out, and how the -- that is kind of
moving forward more than this effort was intended to do.
This effort had, in my mind, a clarity, a simplicity to it,
and it was an effort to do the job better, to take what is out there and
to try to look at it collectively and say, gosh, look at how these
things line up, or don't line up. And what do you think the right
answer is here with respect to what is out there, not with respect to
where we need to go in the future.
DR. APOSTOLAKIS: Which means that you would not this
Regulatory Guide risk-informed?
MR. MARSH: You know, I am not sure we have thought about
that part of the formula here, to be truthful.
MR. CONNELL: Well, what we did think about was allowing
risk-informed -- risk information to be used by licensees in the
implementation of the guide. We thought about that. But we were not --
at this stage, we were basically going to scrub existing guidance, add
any -- add guidance for topical areas that were lacking or non-existent,
and I have got some examples of those, and then allow licensees the
opportunity to use risk information in the implementation of their fire
protection program. But the guide is not directly risk-informed, and
the guide is not performance-based. It is basically a scrub of the
existing deterministic guidance.
MR. MARSH: Cleaning house is more or less what it is doing.
DR. MILLER: Is it safe to say that you are going to clean
house, which is a challenge with 125 documents? But if you have a
situation where you have three hose lengths, who is going to decide
which one is right, and what is the criteria you are going to use to
make that decision? Is it going to be risk-informed criteria involved
there? How is that going to play out? Or do we know?
MR. CONNELL: Well, we are going to take feedback from this
committee, from the public, from industry, internally in the staff, and
whatever information we have -- I don't think you can use risk
information to determine the appropriate hose length. It just doesn't
exist, it is an example.
DR. MILLER: Well, I am just using that as an example.
MR. CONNELL: So that one is going to have to be more of a
committee type decision. Well, what does everybody think is the best?
And is there a reason not to do something, or is there a better reason
to do something else? Or we may delete guidance altogether for hose
length.
DR. APOSTOLAKIS: So what is the role then of the NFPA 805
in this?
MR. CONNELL: There is not direct tie into the NFPA process,
into this process. What we will address is the objective -- if you
remember option 2 in SECY-98-58, NFPA 805 was going to be an option for
licensees to choose to change their licensing basis from their current
prescriptive or deterministic license condition to a performance-based,
risk-informed license condition by implementing NFPA 805.
We anticipate that many licensees may not choose to do that.
If licensees choose not to select 805, this Reg. Guide would be the
Comprehensive Fire Protection Guidance document for licensees that elect
to maintain their existing license condition.
DR. APOSTOLAKIS: Okay.
MR. WEST: This is Stephen West. I want to add one thing.
We really are trying to address some practical problems, day-to-day
problems that exist, also through this effort, and that is that you have
a regulation, say Appendix R, and behind that there's -- don't look at
all 125 documents, but there are some significant guidance documents
that back that up, some Generic Letters and Branch Technical Positions,
Standard Review Plan, and there's people out there on both the staff and
industry, licensees that are responsible for maintaining the programs or
inspecting them, and they are just not familiar with this stuff.
I mean they can look at something in Appendix R and say,
well, the regulation says you need to do this, and the licensee may have
done something else. And what they are not familiar with are the
Generic Letters, like 86-10, 81-12, 83-33, that provided guidance on
what we meant by the regulation. And it is a big problem, I mean it is
a real problem out there for out inspectors, just trying to get their
arms around this. And thing, some things change from Generic Letter to
Generic Letter. And it just makes sense to try to distill all that
information and put in one place what the actual guidance is, what the
staff expectations are at this time.
And we learned from the workshop we had on circuit analysis,
I will talk a little bit about that later, that even licensees that are
responsible for the programs are confused by some of the conflicting
guidance, or trying to understand what guidance supersedes what
guidance.
MR. CONNELL: Okay. I'll go on and finish this slide up.
Of course, the Staff proposed this in March of last year in a paper to
the Commission and the industry indicated last May that they supported
this effort. In June of last year the Commission approved the Staff's
proposal to go forward with the development of the Guide.
I've kind of covered some of these already. I'll go quickly
through it. The Comprehensive Fire Protection Guidance Document -- this
will be to satisfy general design criterion three of Part 50,
Structures, Systems and Components Important to Safety -- fire
protection for that is not solely related to safe shutdown.
We want to replace the existing guidance documents, the
list, the 125 plus documents, eliminate the conflicting guidance,
simplify it, define the licensing and design basis for plants related to
fire protection. You're aware that plants licensed before January 1st,
1979 had to comply with certain provisions of Appendix R. Plants
licensed after that date had to comply with the conditions in their fire
protection license condition.
There's been a lot of confusion about the plants that are
basically evaluated against different criteria within the NRC, within
the members of the public, Congressional inquiries, and licensees
themselves seem to sometimes get confused about what really applies to
them. Hopefully our goal will be to kind of clear that up and make it
easily understood by everybody, because we get a lot of inquiries about
how come this plant has to do this and this plant that is right next
door is doing something else, and that doesn't make any sense, and what
are you guys up to? -- so we want to clear that up.
One of the areas that we think is lacking, almost
nonexistent, Staff guidance is on compensatory measures, a very, very
important area and if you look through that list of 125 documents that I
provided you will find next to nothing on guidance on implementing
compensatory measures -- fire watches, temporary protection,
administrative controls, what have you.
MR. MARSH: So George, here is an area where I misspoke.
This is something that is filling a vacancy. This is an area where we
are going beyond something. It is not just a clarifying. This is
providing something that we think is needed.
MR. CONNELL: The next bullet there, and I'll just cover
that briefly, is circuit analysis, what our intent is. You are aware of
the NEI BWR Owners Group effort in this area. Steve is going to make a
presentation in a little while on it. That will be inputed into the Reg
Guide.
Also the Staff believes that there is some need for
self-induced station blackout and self-induced loss of offsite power,
and this is where risk kind of comes into it a little bit, George. We
have some preliminary studies that show plants that do this, present a
higher fire risk than plants that do not, so we want to address that
topic.
Also, you are aware of the effort that the Commission is
undergoing with looking at the 50.59 process and we would like to tailor
that for fire protection as well, what changes can be made, what is a
USQ under fire protection, what needs Staff review and approval, what
does not -- so this is all good stuff that should be addressed.
CHAIRMAN POWERS: Right now though, 50.59 is the flavor of
the day. I mean that is an evolving target there.
MR. CONNELL: That's right.
CHAIRMAN POWERS: That may be a particular challenge, to try
to meld yourself with something that keeps evolving.
DR. MILLER: Quicksand.
CHAIRMAN POWERS: Especially since at best you'll have one
solution today and hopefully in two years another solution.
MR. CONNELL: Well, we can always revise Reg Guides. That
is the nice thing about them. They are a lot easier to revise than
issuing a whole bunch of new, conflicting generic letters or information
notices.
Okay. The status -- we initiated this effort following the
Commission meeting in March. We are getting some technical support from
Pacific Northwest National Lab. We started with them in June.
In November we got a draft outline from PNNL. I also wanted
to mention part of this effort is the development of a database so we
know where the source documents are for the different guidance, so if
somebody wants to know, well, what's everything the Staff said on
penetration steels. Well, they can ask Pat or Steve or myself and we
can spend an hour and pull all these things apart, but we are going to
have a database now that says this generic letter says this, this
information notice says this, this internal memorandum says this -- so
we will have that available, and we think that is also a useful
exercise, knowing where everything came from, what our history is.
In December we made the outline and I believe the Committee
has it, publicly available and we sent it to NEI requesting that they
coordinate the industry review and get back to us on that.
CHAIRMAN POWERS: Did you expect specific feedback from ACRS
on that outline?
MR. CONNELL: Sure. That would be welcome. We'd appreciate
it.
DR. APOSTOLAKIS: Are we writing a letter any time soon?
CHAIRMAN POWERS: I get the feeling that there is likely to
be a letter in connection with the paper that goes to the Commission in
middle February, is that --
MR. CONNELL: Right.
CHAIRMAN POWERS: -- and so I would presume that the
earliest we will be able to write a letter in connection with that will
be at our March meeting.
Now what we are going to remain alert for throughout this
meeting is is there anything that we should be bringing forward to the
February meeting, and stay tuned. We'll find out as things go on.
MR. MARSH: May I make a suggestion that at the end of the
meeting or towards the end we think about everything that has been
talked about and what maybe you spend more time? There's a lot of
effort to go through this Reg Guide outline and in terms of feedback it
may be better to look at FPFI in terms of circuit analysis than to spend
a lot of your time looking at the outline.
CHAIRMAN POWERS: I think you are sensing what are some of
the misgivings that the ACRS had about this activity and that just
seemed premature, because there is so much else coming along.
Now in your defense, the things I see coming along I don't
think are going to persuade a fraction of the licensees to change their
existing programs.
MR. MARSH: Right.
CHAIRMAN POWERS: And so in a sense maybe this going to have
utility no matter what all this other stuff does, but you can understand
why the Committee might have misgivings and say, gee, it would be better
to do this after we had all this other stuff in, and then we could make
decisions about should there be guidance in this area or not.
MR. MARSH: I understand.
MR. CONNELL: Okay. This is my last slide.
What we are going to do is of course we will accept and
review the stakeholder feedback on the draft outline. We have a meeting
scheduled in February, February 25th, to discuss the feedback on the
outline. Then of course we will take that feedback and go back and
draft the actual Reg Guide itself and we'll issue the draft for public
comment. Our schedule is September of this year for that.
MR. MARSH: There is an important step between those two
bullets, okay? That is our process step, and that means go to CRGR --
MR. CONNELL: Right.
MR. MARSH: -- and discuss with CRGR about this document,
how it should be couched, whether it should go out for public comment.
I mean there are major, in my mind, implementation backfit issues
associated with a document of this sort, major issues. Can't take
coalescence of all of our guidance and say here it is -- this is what
you folks should be complying with. There are just major questions of
that sort that will come out I'm sure in the public comment period, but
we have to decide internally before we put it out as well.
CHAIRMAN POWERS: You are clearly going to run into the
difficulty that a guy that got his plant licensed --
MR. MARSH: Absolutely --
CHAIRMAN POWERS: -- at document Number 67 and has his
licensing basis --
MR. MARSH: Absolutely.
CHAIRMAN POWERS: -- that history is going to be in rough
sledding relative to the plant that got licensed at document 124 --
MR. MARSH: Right.
CHAIRMAN POWERS: -- and you are simply going to have
accommodate that.
MR. MARSH: Have to face that.
CHAIRMAN POWERS: Where you have got a big backfit issue.
MR. MARSH: I think we need to face that, okay? This is the
first step of this process -- put it all together, find out what is all
the guidance that is out there, and then at the implementation -- or at
the implementation stage but in the decision-making leading up to
implementation we have to decide what it means, what are the backfit
implications. Is it worth backfitting or not?
First is to develop the guidance but there's a major -- the
future holds major challenges in terms of where to go.
MR. CONNELL: And I did note the implementation and also the
relationship with NFP 805 and different paths that would be pursued.
That is the end of my prepared presentation. I would be
happy to answer any questions.
DR. APOSTOLAKIS: Yes, there are.
This Guide right now doesn't really say very much, does it?
It just says --
MR. CONNELL: Which Guide is that?
DR. APOSTOLAKIS: Well, I mean the outline.
MR. CONNELL: It's an outline. That's right -- just
addressing the topical areas of the existing guidance plus the ones that
the Staff believes need to be expanded upon.
CHAIRMAN POWERS: I think we need to at some time decide as
a subcommittee how we want to address this outline, or do we want to
address it?
DR. APOSTOLAKIS: Yes. I was looking at the Fire Protection
Program Goals and Objectives, which is Chapter 4, there is a lot of
stuff there but I think at the end you will really be doing what you
have been doing all along.
Let me ask a couple of questions. Maybe it is there and I
don't see it.
Defense-in-depth is defined here as prevention of fires,
detecting and suppressing fires, and preservation of plant safe shutdown
functions. Then it goes to Fire Protection Program Performance Goals,
and now we have safety-related structures, systems and components,
post-fire safe shutdown, prevention of radiological release.
Well -- these are also defense-in-depth.
MR. CONNELL: Well, defense-in-depth, if you look in the
definition, Appendix R, is the way it is defined in here.
DR. APOSTOLAKIS: Well, we cannot have different definitions
of defense-in-depth. What is the difference between Appendix R?
CHAIRMAN POWERS: In defense of the fire protection folks,
they are the only ones that do define defense-in-depth.
DR. APOSTOLAKIS: Which is?
CHAIRMAN POWERS: Exactly as stated.
MR. CONNELL: Yes, as stated in the draft outline here.
CHAIRMAN POWERS: Yes. The only place in the regulations
that the words defense-in-depth actually appears in Appendix R.
DR. APOSTOLAKIS: And it is defined.
CHAIRMAN POWERS: And it is defined that way so --
DR. APOSTOLAKIS: So prevention, protection, and
preservation of safe shutdown, is that the definition?
CHAIRMAN POWERS: That is the definition that is used.
DR. APOSTOLAKIS: Well, this is defense-in-depth at a --
well -- which brings up again the larger issue here of whether you
should just consolidate what you have or go beyond it. I don't think we
should have a defense-in-depth definition in different areas. I mean
there has to be some line drawn someplace.
DR. KRESS: It would be nice to have an overriding
definition.
DR. APOSTOLAKIS: Yes.
DR. KRESS: Within -- and derive from that what it means
with respect to specific areas.
I think if you actually had that you might come down to this
definition in fire protection.
MR. CONNELL: Yes, but we are not going to do that in the
fire protection Reg Guide though.
DR. KRESS: Well, you just start where you think what it
means for fire protection, and it may or may not be consistent with some
overriding definition.
DR. APOSTOLAKIS: Well, why don't we say --
DR. KRESS: It would be hard to believe it's inconsistent.
DR. APOSTOLAKIS: When we say prevention, make sure that you
detect them, and preservation of plant shutdown functions, does that
apply to every single area in the plant?
MR. CONNELL: In areas that are important to safety.
DR. APOSTOLAKIS: Where does it say that?
[Pause.]
DR. APOSTOLAKIS: It doesn't.
MR. CONNELL: It's in GDC-3. I mean this is an outline. We
haven't written the guidance yet, George. This is just an outline.
MR. WEST: I am not sure what your -- what is your point,
George? What is your question? I mean defense-in-depth is covered by
the existing regulation. We are not changing any of that. You can't
change it through a Reg Guide.
DR. APOSTOLAKIS: It's not just defense-in-depth, Steve. I
like to see a top-down approach. I like to see the top goals, then the
next layer, the next layer and understand how well these things come
together.
MR. CONNELL: Then you'll love NFP 805.
DR. APOSTOLAKIS: Pardon?
MR. CONNELL: You'll love NFP 805 then.
[Laughter.]
CHAIRMAN POWERS: Maybe -- not.
MR. CONNELL: That is the way this is structured.
DR. APOSTOLAKIS: Just by putting things down, that doesn't
mean it is very good, but this is a necessary, not sufficient,
requirement.
What I don't like is a list of objectives that apparently do
not relate to each other. Now they may relate in your minds, because,
you know, you have been living with this for a long time, but there is a
4.1 which defense-in-depth and it is defined prevention, detection, and
preservation. There is a 4.3 fire protection program performance goals.
Now we are talking about safety-related structures, systems and
components, post-fire safe shutdown, and prevention of radiological
release, and in my mind again, if I prevent radiological release, am I
not taking care of safety-related structures, systems and components?
See, that is the top-down. If I make sure that I don't have
radiological release, that means I have already protected the necessary
number of structures, systems and components that will prevent that
release, but no, this is a separate requirement, which means all of them
will be under this. Now -- which comes back to the definition of
safety-related, right? What is safety-related?
The point is that this is not risk-informed in my opinion,
and I don't see why it shouldn't be or at least have some element to it.
Why shouldn't it have some elements of risk-informed information in it?
It may be there but right now if I only read the document
without talking to you, I don't get that impression.
MR. WEST: It wasn't a goal of this project to come up with
a risk-informed fire protection program for nuclear power plants. That
is a separate effort. This wasn't intended to cause any perturbations
in existing programs. This is a very simple task.
DR. KRESS: Yes, I think if it's going to be risk-informed,
George, I think you have to start with the regulations.
MR. CONNELL: Right, and the intent of this Reg Guide was it
was going to be used by licensees electing to maintain their existing
license condition.
DR. KRESS: And you really don't have --
DR. APOSTOLAKIS: But it doesn't --
DR. KRESS: -- but you can, you can work some risk
information into the Reg Guide.
DR. APOSTOLAKIS: It doesn't have to be risk-informed in the
sense that you go to Regulatory Guide 1.174 and look at delta CDF. I
don't mean that. But I mean this top-down presentation and justifying
why you do this and why you do that, there is some element of
risk-information there or risk approach that you are using, but it is
not really risk-informed, but I would like to see some of that because
are we looking at what is important and how do we define important?
MR. MARSH: This is not --
DR. APOSTOLAKIS: What is the definition of importance here?
MR. MARSH: This is not intended to change any regulatory
structure, regulatory format, precept, concept. This is supposed to be
a reconstruction of the existing format and we have got to be careful
not to change.
If there are --
DR. APOSTOLAKIS: I understand that.
MR. MARSH: -- are new guidance being out there, I mean
you're right, we would need to be considering risk information in the
new guidance. I agree with you there, and I certainly don't have a
problem, George, with a format if a top-down makes the existing
regulatory structure more understandable, if it's always been there, and
as a way of presentation it becomes clearer what the structure is by a
construct. I have no problem with that at all.
But to the extent that we are changing the regulatory
requirements or reconstructing the regulatory guidance that has been out
there, we shouldn't be doing that in this effort. This is only --
CHAIRMAN POWERS: I think George is asking this question
maybe -- maybe -- is that you make decisions all the time when you
formulate guidance, even within the context of the existing regulations.
You make decisions of emphasis, detail, breadth of coverage -- things
like that -- and I think George is asking the question ought not that
decision-making process that you make, whether it is reflected in the
guidance or not, but that the decision-making process be based on risk
information to the extent that it is practicable, and I think we have to
say practicable here, as opposed to practical because we again come into
this question is the intuition we have developed from the risk
assessment tools of any use to us or is it open to question, or do we
have to rely on another definition of risk, which is not so much the
quantitative but more the engineering judgment type of risk that has
been used historically in the fire protection area to make those
decisions, and if you do, does that change the method of presentation
here without changing any regulations at all.
I think that is a legitimate question to wrestle with --
MR. MARSH: Okay.
CHAIRMAN POWERS: -- in thinking about bringing this up and
maybe a politic one as well.
MR. MARSH: It is simply that.
MR. CONNELL: And understand the timeframe that we are
talking about here. We are trying to get this thing out, you know, near
the end of the year here, so if there is something that we are going to
have to wait for development of, we won't -- I mean we'll skip it.
CHAIRMAN POWERS: I think the ground-rules you have set for
yourself is development is somebody else's problem.
MR. CONNELL: That's right. I'll let Nathan do that.
CHAIRMAN POWERS: Nice -- I'm sure he will appreciate it.
DR. APOSTOLAKIS: I mean that same question can be rephrased
as follow: To what extent have the fire risk assessments that have been
done in the last 15-20 years influenced this and to what extent has this
new inspection program that the Agency has been developing with their
cornerstones that you mentioned earlier influenced this?
MR. CONNELL: Well, I'll give you some examples, George. I
gave you the SISPO and the self-induced loss of offsite power. That is
directly resulting from the risk assessments. We have no guidance on
that and partially through the risk assessments and partially through
some Staff initiative we discovered that licensee were using this
methodology for their post-fire safe shutdown method, so that is one way
that the risk information is being used in it.
Of course, I expect the circuit analysis resolution is going
to have some risk information tied to it, as will probably the
compensatory measure guidance that we generate, but like I said before,
I don't think that the risk tools can tell me how long the hose needs to
be.
DR. APOSTOLAKIS: My fundamental question is whether it is
any of your business telling them how long it should be.
MR. CONNELL: Right now it is. Right now it is.
DR. APOSTOLAKIS: I don't know. Why?
MR. CONNELL: Because we have that guidance now.
CHAIRMAN POWERS: We have a big prescriptive regulation.
DR. APOSTOLAKIS: The boundary conditions here are confusing
the hell out of me. We have to do this because the regulation that we
wrote says we ought to do it.
MR. MARSH: We are.
DR. KRESS: Right. That's right.
DR. APOSTOLAKIS: Does that make sense?
MR. MARSH: We are -- we are writing a new regulation. I
mean we have two parallel efforts going on here.
One is if you want to keep your existing program under the
existing regulation, this is how you do it. This is how you maintain
it. If you want to go performance-based risk-informed, there has going
to be a vehicle to do that. It's a parallel path. Licensees can pick
either way and we are going to talk more about that this afternoon.
DR. KRESS: This is the parallel path we talked about that
we have to be prepared to accept -- remember in our last letter where we
talked about risk-informed regulations. You have to be prepared to
accept a parallel path, and I think this a real example of it, we have
to be able to accept it.
MR. WEST: There are licensees out there that say we are
happy with our programs, we think they are safe, and we have and the NRC
hasn't or the ACRS hasn't shown us anything that says they are unsafe,
we are going to stick with it.
DR. APOSTOLAKIS: I think that parallel path is a little
dangerous, and I will tell you why.
DR. KRESS: What are the choices?
DR. APOSTOLAKIS: I will tell you why -- because any time
you find something from the PRAs that you haven't thought of,
immediately you bring it into the current regulations, right? So PRAs
used to increase the regulations, and you don't just say, ah, this is an
insight from the PRA, but I will only use it on this path because that
is the risk part. You are also affecting the deterministic part, but if
the PRA says certain things are unimportant, then that goes to the risk
part, not to the deterministic part.
DR. KRESS: You are talking --
DR. APOSTOLAKIS: Now that doesn't make sense to me at all.
DR. KRESS: -- talking here historical past practice. I
think we are entering a new -- I hate the word "paradigm" but we are
entering an area where we are going to change the regulation package to
be risk-informed.
DR. APOSTOLAKIS: And this is not --
DR. KRESS: This is not part of that. This is the other
package where you are --
MR. MARSH: This is not the Part 50 risk-informed change at
all.
DR. MILLER: This challenge is just to codify basically what
is already there.
DR. APOSTOLAKIS: So you promise then that if I look at it
more carefully I will not find additional requirements coming from PRAs?
DR. KRESS: Oh, no. I won't promise that.
MR. MARSH: I won't promise that either.
DR. APOSTOLAKIS: But then if you do that, you will also
promise that I will find some relaxation of requirements because of the
PRAs?
DR. KRESS: Yes --
DR. APOSTOLAKIS: In here.
DR. KRESS: No, no. I don't know what's in there. I
suspect not on either of those, right?
MR. MARSH: It wasn't intended to do that.
CHAIRMAN POWERS: You know, I think it would be unfair to
expect these gentlemen to produce something closing their minds to what
they see coming from other sources. I mean they have an intuition about
risk that is derived from a variety of different sources, each probably
a little bit different, but to say that they are not mindful of the
results of quantitative risk assessment when they produced this outline
would be asking for a discipline that I think would defy anyone's
ability to put together.
Just as the people who wrote the guidance 20 years ago were
mindful of a lot of things, including some element of risk assessment,
as they saw it at that time. But that is not the intention in this
particular document. I think you are just hitting upon things that we
had misgivings about it, thinking that this was a little bit premature.
And it may all be fine, because it does -- some of the things that we
wanted to mature will be maturing in parallel with this and,
undoubtedly, will influence it somewhat.
DR. APOSTOLAKIS: But it seems to me, though, that we have
been talking about a two tier system and we have sort accepted it
uncritically that there will be a two tier system, and now I am finding
that -- and that was one, --that it may not be practical to have a two
tier system.
MR. MARSH: When you say two tier, you mean risk-informed
technique and the classical Appendix R technique?
DR. APOSTOLAKIS: Right. Because the classical Appendix R
technique is affected by this finding.
MR. MARSH: Certainly. Certainly, it is.
DR. APOSTOLAKIS: So now the line is beginning to be
blurred.
MR. MARSH: You can't --
DR. APOSTOLAKIS: What I don't know want to see is it being
one way.
MR. MARSH: There is a precept -- there is a precept that
was associated with the formulation of the regulation in the first place
that had a risk notion to it. It may not have been well defined,
certainly not quantified, but it was a precept. I mean it was there.
DR. APOSTOLAKIS: In fact, one other thing, for example, in
the fire PRA a very important first step is screening the locations
according to certain criteria. Right? In such a screening allowed
here? Or every single location, I have to make sure I don't prevent --
MR. CONNELL: Well, the fire PRAs are focused on prevention
of core damage only. General design criterion 3 is a lot broader than
that. So the PRAs, implementing the PRA does not show that you are in
compliance with general design criterion 3. There are other issues,
fire protection issues at a nuclear power plant other than core damage.
DR. APOSTOLAKIS: Like?
MR. CONNELL: Like radiological release. You have a fire in
rad waste building.
DR. APOSTOLAKIS: Okay. So --
DR. KRESS: That's very --
DR. APOSTOLAKIS: Right. Right.
MR. CONNELL: George, and the PRAs don't address that.
DR. APOSTOLAKIS: So I can have a set of criteria, including
those --
DR. KRESS: That PRA could address.
DR. APOSTOLAKIS: Could address and it does not address to
the pleasure of everybody present here.
DR. POWERS: George, let me ask you --
DR. APOSTOLAKIS: Wait, wait. No, this is important. I can
set a number of criteria, core damage, LERF, measure of radiological
release, anything else I want, and then screen the locations according
to those. And then apply the defense-in-depth idea of preventing fires
there, rather than saying in an absolute manner, up front, I want to
prevent fires from occurring, and so on, which implies everywhere. Now,
isn't that a different approach?
DR. KRESS: I thought that was the approach.
DR. APOSTOLAKIS: I didn't see screening anywhere.
MR. CONNELL: Licensees simply do not have different fire
prevention programs for different areas of their plants based on their
PRA results. They have a site-wide administrative control that limits
combustibles, limits hot work, controls hot work. All those things are
incorporated in their program and it is site-wide. It is not -- they
don't say, well, in this room, I have to do this, in this room I have to
do something else. That doesn't exist, and it is impractical to
implement something like across a plant.
DR. POWERS: George, let me ask you a question about your
concern about one-way streets from PRA. It seems to me that there is a
one-way street, but it is the other way. That is I, as a licensee, do
the fantastic, unbelievably detailed and validated PRA, and I find that,
indeed, there is an excessive amount of conservatism within Appendix R,
I have a mechanism for asking for relief through 1.174.
If, on the other hand, I am NRC and I do the fantastically
accurate, unbelievably validated and verified PRA and find that there is
something missing from Appendix R, that should have been included in
there, but was not, for whatever reason, that the only mechanism I have
to add that is to go through a backfit analysis. So it seems to me that
the one-way street is actually the other way.
DR. APOSTOLAKIS: No, because even the first way it can be
killed by saying our objective here is not just to prevent core melt and
LERF, minimize LERF. So 1.174 really doesn't do much for you.
DR. KRESS: Well, 1.174 sort of includes that, because it
says you will also preserve the defense-in-depth philosophy, and you
will also comply with the current regulatory requirements. And when you
throw those two in, it sort of -- it gets a little broader.
DR. APOSTOLAKIS: Yeah, but we can't say 1.174 says preserve
the defense-in-depth philosophy and then say we, as fire protection
engineers, define it this way. That doesn't make sense.
DR. KRESS: No, you will have to -- it will have to be as
defined as 1.174.
DR. POWERS: No, no, no. In this case, what -- the
defense-in-depth philosophy, we are much further aground here with
respect to defense-in-depth philosophy than we are anywhere else in the
regulations because we have -- this is the one area where we have a
definition. Okay. We may not like the definition, and there are
certainly people sitting at this table that don't even think that fire
protection is an area where defense-in-depth should be invoked. But it
is, and it is there, so there's a very clear standard to compare against
there.
DR. APOSTOLAKIS: See, my problem is, when I read this, and
I will read it again, I get the impression that all these fire risk
assessments that have been for the last 15-20 years really have had no
impact on this.
DR. POWERS: I think that is a fair perception, that the
explicit impact probably is small because they are trying to collect
together all the guidance that has been given over the last 25 years.
DR. APOSTOLAKIS: So, there will be another Regulatory Guide
in two years after Nathan finishes with this one.
DR. POWERS: I think that is exactly what the marching order
is here. Now, I agree with Tom, it is more likely five. But I think it
would be inevitably true --
DR. MILLER: Predict two, it will be five.
DR. POWERS: -- that -- it is clearly inevitable if the NRC
says you can comply with Appendix R or you can comply with NFPA 805.
That is absolutely true that there would be two sets of guidance there,
that is unavoidable. It is likely to be true that in two to five years,
the fire protection staff at NRC would say I need additional guidance
because risk-informed regulation has progressed sufficiently that I need
additional things to be considered that, indeed, one need not have a
specific hose length in a fire protection area anymore because of this
risk-informed guidance, and things like that. I think that is pretty
clear.
MR. WEST: I just want to remind George of one thing. A
couple of years ago, in this room -- I think most of us that are here
today were here then, two years ago -- we discussed with you a proposal
to do exactly what you are suggesting today, get rid of Appendix R,
delete it, eliminate it, and go with the one page, or maybe even one
paragraph fire protection regulation. And you thought that was a good
idea, the ACRS thought that was a good idea, and we went to the
Commission with that. And the Commission agreed. Okay, get rid of
Appendix R and replace it.
And largely because of subsequent protests from industry, we
abandoned that plan and came up with this parallel path plan, and that
is where we are today. I mean I don't think we disagree with you, but
circumstances overcome events and we are -- you know, we are where we
are at today. And I don't think we are going to really seriously
consider going back now and starting over where we were two years ago.
Although, personally, I would like to.
[Laughter.]
MR. MARSH: It's much easier.
DR. POWERS: No, it wouldn't be.
MR. WEST: Do you remember that now? You are being awfully
quiet. You may have been out for coffee that day.
DR. POWERS: Trust me. They are absolutely -- Steve is
absolutely correct. We were in here very excited about the idea of
creating an Appendix S prime that we thought the world was hungering
for.
MR. MARSH: Right.
DR. POWERS: And we got our wheels realigned for us real
quickly.
DR. MILLER: The world wasn't hungry.
DR. POWERS: Well, they were hungry for something else, I
think.
DR. MILLER: I wasn't even here and I remember it.
MR. WEST: You read the letters.
DR. POWERS: Ed, thank you very much. Steve, let's move on
to an issue that has much less controversy associated with it.
DR. MILLER: Oh, is that right?
DR. POWERS: We were kidding.
DR. MILLER: Talking about areas where problems could be
involved, I think it one.
DR. POWERS: Well, as you weren't here for my opening
comments, the difficulty I had on the probabilistic assessments that I
have seen in connection with this issue is that it appears to me many of
them are directed at trying to convince me that Brown's Ferry could not
possibly have occurred. And I think that is something where we have got
to really think carefully about our analytic tools and our experimental
activities in this area.
If they cannot demonstrate the kinds of things that were
observed at Brown's Ferry, if they can't demonstrate the kinds of things
that have been observed in a number of industrial fires, then I think
those analytic tools and those experiments are simply not validated.
And I think we are going to have to think about that carefully when
people make these arguments.
And I think it is going to behoove the staff generally,
perhaps the research program, to start rethinking the kinds of databases
they should be collecting, and instead of focusing exclusively on fire
incident rates, to start focusing on fire consequence events, because we
are having people making these arguments -- I ran this test and it
didn't show any shorts to conductors, it only shorted shorts to grounds.
But, unfortunately, the fire showed shorts of every conceivable variety.
And so the question is not what is the applicable regulation
here, the question is what is the valid text.
DR. APOSTOLAKIS: And why doesn't that thinking apply to the
75 foot hose? Why don't I have to see evidence why it has to be 75 and
not 76?
DR. POWERS: I don't know. I think --
DR. APOSTOLAKIS: Is it because the Committee said it is 75?
DR. POWERS: I think it is fair to ask people why in the
world they came up with particular bits of guidance on 75 foot hoses,
just as three foot telephone extensions.
DR. APOSTOLAKIS: That's right.
DR. POWERS: Both ask why the specific length and why the
guidance. But I also --
DR. APOSTOLAKIS: You know, it seems to me the principle is
the same.
DR. POWERS: I also think it is possible for them to come
back with an answer.
DR. MILLER: It seems like that is a totally different
question than some of the --
DR. APOSTOLAKIS: Anyway, anyway --
DR. POWERS: Yes.
MR. WEST: Well, a lot of that, of course, is just based on
-- it is classical fire protection features.
DR. APOSTOLAKIS: I know. Classical, not in the way the
Greeks use the word.
MR. WEST: I think in this case, yes.
DR. APOSTOLAKIS: You mean traditional.
DR. MILLER: Traditional, right.
MR. WEST: Traditional, classical. But you do see a lot of
deviations from it. I mean it is basically guidance. It is a starting
point and you can deviate. I mean some plants probably do have 76,
maybe 78.
DR. MILLER: Steve, are we going to talk about issues, or
are we going to talk about some fundamental, physical aspects of these
issues, or just go through the issue?
MR. WEST: In this presentation?
DR. MILLER: Right.
MR. WEST: Basically, I think -- I think status, really,
kind of tell you where we are at. And I think some of the other
presentations on the topic today from the Owners Group, from NEI, will
get to more of the issues, although I can try and answer any question
you have.
DR. MILLER: I am not -- you may know, I am not a member of
the Fire Protection Subcommittee, but I was intrigued in looking at the
circuit analysis issues. I wanted to run out and start doing
experiments right away. Not a lot, but it seemed like --
DR. APOSTOLAKIS: Well, then you would have the problem Dr.
Powers raised.
DR. MILLER: Dr. Powers raised a very good point. I can do
all the experiments in the world and somebody would question did I do
the right ones. I understand that.
MR. WEST: Right. Well, I will be happy to try to answer
any question I have. If I am not covering it, and you want to discuss
something, just ask.
DR. POWERS: Well, Steve, I hope that you will be around
after we have had presentations from both Emerson and the BWR Owners
Group, because I think we would like to discuss some of the philosophy
of resolution here.
MR. WEST: Yes.
DR. POWERS: I have certainly articulated a concern that I
have, a big time concern I have, but there are a lot of other issues
here that I would like to both --with and to get a better understanding
for. Because every time I go through your presentations, I come away --
your and Pat's presentation, I say I understand this issue, it is very
simple. Every time I go through Fred Emerson's presentations, I say I
understand this issue, it is very simple. Unfortunately, the two are
orthogonal.
MR. MARSH: Well, those are great presentations.
DR. MILLER: Again, being new, I found it intriguing, when I
read 86-10, at least the parts with circuit analysis, and I kept saying,
where did we -- how did we reach these conclusions that we are going to
say these circuits are always going to be failed until somebody correct
them and so forth? Those kinds of questions. Probably I am not the
first one to ask those questions, or the last.
MR. MARSH: Probably not. Probably not.
DR. POWERS: I hope you noticed that we can blame all of
86-10 on the Nuclear Utility Fire Protection Group.
MR. WEST: Well, Dr. Powers, I think you are right, it would
be better to -- it would be good to have a discussion after you have
heard all the presentations today on the subject.
I am Stephen West, Chief of the Fire Protection Engineering
Section. When Jit asked me to come talk about this, I immediately said
yes, because I am always anxious to come to you guys about things. And
it wasn't until I started late last week thinking about what I was going
to say that I was second-guessing my decision to come here. And after
the last presentation, I am ready to run for the hills, to tell you the
truth.
MR. MARSH: But George isn't here, so go ahead.
MR. WEST: Yeah, I know, I am going to hurry up.
DR. POWERS: It won't do you any good. Anything that he
needs to be reminded of, I will see to it that he gets brought up to
speed.
MR. WEST: We came in June of 1998 and made presentation on
this issue. The staff did, Pat Madden made a presentation and got into
a lot of technical detail, and we went through the issues and possible
resolutions. And I think we also heard from NEI at that time. They
gave some industry views on --
DR. KRESS: Too late, he is back.
MR. WEST: Yeah, I see that.
DR. POWERS: And, again, very welcome.
MR. WEST: And so I wasn't going to try to discuss again all
the details of the technical issues, although we certainly can. But,
basically, the problem is that fire-induced circuit failures may
adversely impact the ability of a plant to achieve and maintain safe
shutdown.
DR. MILLER: Is that -- that's nothing new?
MR. WEST: That's nothing new.
DR. MILLER: The designers 40 years talked about that.
MR. WEST: Well, I think that was -- no, I think 40 years
ago, --
DR. MILLER: Thirty.
MR. WEST: -- they probably had some problems with separate
-- you know, the kind of things we would take for granted today, some
kind of routine separation of redundant functions and divisions.
DR. MILLER: They may have not sufficiently made provisions
for it, as we proved as Brown's Ferry.
MR. WEST: Right. Yeah. Well, the Brown's Ferry fire
definitely revealed the problem, the technical problem, the issue. And
we had thought that we had resolved or addressed that problem through
Appendix R. I mean it was basically -- one of the principle objectives
of Appendix R was to specify fire protection features and methods to
provide reasonable assurance that a fire in the plant would not
adversely impact the ability to achieve and maintain shutdown. So we
thought we had addressed the problem.
We found recently in the last couple of years that some
licensees may not have fully considered the effects of fire on their
safe shutdown capability.
DR. MILLER: Does that mean they didn't adhere to the
guidance that was out there or they didn't understand the guidance?
MR. WEST: It could be a combination of both. I think when
Dr. Powers in his opening remarks on this issue, I think he was reading
from something we wrote --
CHAIRMAN POWERS: Yes.
MR. WEST: Summary of our workshop we had on this issue, and
it kind of characterized the different places that the Staff and the
licensees could be, depending on your perspectives.
My personal feeling is I think there are some licensees out
there that made good faith efforts to comply with the regulation as they
understood it, using the guidance that we had issued, and they may have
failed for whatever reason, but it wasn't because they didn't try.
You said you read the guidance in 86-10 and I have been
living with this stuff for so long I kind of think it's easy to
understand but I am not so sure. I have heard convincing or compelling
arguments that it is not that easy to understand, that it can be
confusing, that saying something in Generic Letter 81-12 and then saying
it again but maybe differently in 86-10 is confusing, so we do have some
of that.
CHAIRMAN POWERS: What you can concede is people can
formulate detailed technical questions whose answer is not transparent
without assisted guidance?
MR. WEST: Yes, I agree. But we do have the licensees where
it appears that they made a good faith effort to comply and they may
have complied in many cases but not all cases. Then we do have others
who I think have embraced a certain interpretation of the regulation
that would not lead to compliance and their are sticking by their guns.
There are some out there, and that is the third bullet, the third
problem -- there are some differing views between the NRC Staff and the
industry on this issue and I am going to talk about those a little bit
more. This is kind of just the problem we are dealing with.
Again, without getting into a lot of the technical
discussion again -- we may hear more of that I think from the other
presentations -- but there are a lot of factors or elements in a circuit
analysis, things that need to be considered --
MR. MARSH: Steve, before you go into some of those details,
I want to inject another what I think is a reality.
I think the NRC may be bag-holders to some of these
disagreements or the situation as well. It may very well be that a
technical reviewer may have approved some particular type of design
feature or compliance method back in those licensing arenas. It may
very well be that there are documents out there that support some of
these industry disagreements with the Staff.
I don't want to color it with you as being purely black and
white.
MR. WEST: Okay.
MR. MARSH: That is, that there are some licensees who
misunderstood. I think that there may be situations out there where we
endorsed some approach which we now think is not appropriate, but to be
honest, we haven't seen a document that supports that. We haven't seen
safety evaluations or letters or REIs back and forth with that that
support that notion, but I would not be at all surprised if that is the
case.
DR. MILLER: What would you think -- just take 86-10 -- ten
people might read that in ten different days to get multiple
interpretations.
MR. MARSH: They may. They certainly may, and what I am
saying is that we are a Staff, we are people, we are engineers, and we
may have had different interpretations at various points in time as
well, and I just want to recognize a reality that I think is out there,
that we may have endorsed this as well, so this has backfit implications
as well is what I am saying.
DR. MILLER: That's your bottom line?
MR. MARSH: Yes, sir.
MR. WEST: Anyway, there's a number, there's a lot of
factors and elements that go into a circuit analysis and just briefly we
kind of look at these circuit analysis as the background of the whole
fire protection program.
I mean the circuits are vulnerable, and it's the circuits
that you need to remain functional in the event of a fire or following a
fire, to operate the equipment you need to shut down the plant and
maintain it in a safe shutdown condition.
That analysis really defines what level of fire protection
you put in place, what kind of features you put in place, where you need
fire barriers, where you need suppression systems, so it is really a
critical element of the whole plant fire protection program and the
foundation of it. The whole thing rests on the circuit analysis.
They need to consider both required and associated circuits.
Where a required circuit would be the circuit that powers the pump --
you need that power cable to remain free of fire damage and functional
so you need to provide some kind of protection for that to ensure that
your pump is going to be operable when it is called upon.
An associated circuit is a circuit that is not required but
it may be associated with the required circuit in some way -- could be
connected to the same boss -- but if it is damaged it could cause you to
lose the required circuit.
DR. MILLER: In that first bullet it seems like that is the
nemesis, this whole problem. That defines the scope of what we are
looking at --
MR. WEST: Right.
DR. MILLER: -- and required is probably fairly easy but
associated I think has lots of ambiguity depending on the plant design.
MR. WEST: That's true and that is one of the specific
issues that the BWR Owners Group for example is working on, coming up
with definitions.
DR. MILLER: Because I could go in and make certain all my
safety-related circuits are appropriately isolated and Thermo-lagged and
buried and anything else --
MR. WEST: Right.
DR. MILLER: In the associated circuits I may still have a
problem.
MR. WEST: Right. Well, even for required circuits there
could be a problem, and I'll talk about that when I get into the issues
we are working with industry on, but I agree with you.
So the analyst is looking at the potential for various
actuations, operations, fire-induced -- you know, fire could cause a
problem like a hot short or a short to ground or an open circuit, could
cause -- you need to look at fuse breaker coordination, circuits that
are in common enclosures, need to look at your transfer and isolation
capability. For example, for a control room fire, you need to leave the
control room. How do you isolate the control room and transfer control
to your alternate shutdown panel or panels?
Look at high/low pressure interfaces. They get kind of
special treatment in fire protection and an issue, MOV -- motor-operated
valve mechanical damage -- and I will talk about that in a second too,
but that is just kind of an example of the considerations that would go
into one of these analyses.
DR. MILLER: Now wouldn't your PRA be important to some of
those analyses?
DR. APOSTOLAKIS: What did you say?
DR. MILLER: The PRA would be better today --
DR. APOSTOLAKIS: No. They are not using that. You are not
allowed to use that.
DR. MILLER: I am saying as an analytical tool only.
MR. WEST: It could be.
CHAIRMAN POWERS: Steve, let's come back to your spurious
actuation.
MR. WEST: Okay.
CHAIRMAN POWERS: And I think it is important to understand
that spurious actuation or signal is in the --
MR. WEST: Right.
CHAIRMAN POWERS: -- in the requirement, and to understand
that you have this actuation and signal, and then there is a cascade of
things that follow from that.
I mean a signal tells something to do something and that in
turn tells something else, and you have to track that whole thing --
MR. WEST: Right.
CHAIRMAN POWERS: -- in doing this. It seems to me that's a
crucial point to understand.
MR. WEST: Right. As you can imagine, there's a lot of
things that could go wrong.
I mean you could have valves operate that shouldn't operate,
pumps start that shouldn't start or stop that should be running or you
could have false indications on instrumentation -- any number of things,
whatever a fire-induced failure could cause to a circuit, power
instrumentation or control.
DR. MILLER: Well, spurious actuations for example could be
when you have, say, shorts, hot shorts. You get sparks. Now in the old
days of analog instrumentation, sparks may not be a problem, but sparks
may be a problem with digital stuff -- I mean just like the old spark
gap transmitter almost.
MR. WEST: Right.
DR. MILLER: I don't want to introduce new problems here. I
started to envision it when I started reading all this stuff.
MR. WEST: Okay. Basically the regulatory considerations --
it is an issue of high regulatory interest. It's an issue that has the
attention of the Commission and industry and the public. It has the
potential to have a significant impact on industry or at least certain
licensees, so there is a lot of interest. That is evidenced by the
involvement of the Owners Group and NEI in this issue, for example.
We believe that it has potential safety and risk
significance --
CHAIRMAN POWERS: I noticed that nearly all of the speakers
and writers on this issue used those words, "safety" and "risk
significance" -- but they don't go on and -- I mean they are obviously
drawing a distinction between safety and risk here, but they don't go on
and tell me what that distinction is. Can you tell me what that
distinction is?
MR. WEST: Probably not. I mean it is -- I was getting
ready to say "although" when you asked your question.
We believe that there -- I mean the safety and risk
significance is obviously going to be configuration-dependent. You
could probably go into a plant and find configurations where a
relatively small fire, for example, could cause significant problems --
maybe a control room fire where a licensee has not done an adequate
analysis of the MOV issue to where a relatively small fire in one or two
cabinets could cause a lot of problems with a number of MOVs that would
cause them to go into a configuration that you don't want and that you
cannot recover from, at least not easily, because the valve becomes
bound.
That could have significant adverse effects on your safe
shutdown capability, your ability to either achieve safe shutdown or to
maintain it, and can introduce significant challenges to the operator
and to the plant.
There are probably other configurations where the
configuration may not meet the regulatory requirements -- you know, the
regulation said you should have a one-hour barrier on this cable and
maybe you don't -- but if you look at it, you may agree that it is not
that significant. Maybe it doesn't matter. It may be a valve the
operator would expect to be available won't be available but it is not
going to be that big a deal. He could easily recognize the problem and
have another way to compensate for that valve being unavailable or going
into an adverse position.
We don't really have our arms around the risk significance
of these various configurations. I think it is the focus of the NEI
effort that Fred is going to talk about -- in fact that is on my next
slide -- trying to get their arms around the risk and
safety-significance of circuit analysis problems or fire-induced faults,
fire-induced failures.
I think also our research program -- Nathan is doing some
work, having contractors at Sandia National Laboratory -- try to give us
some insights into the risk significance of these problems, and I think
they are trying to use Brookhaven also. Are you going to talk about
that later, Nathan? Nathan will talk more about the research effort.
MR. SIU: That will be tomorrow.
MR. WEST: Tomorrow, okay.
DR. MILLER: Is that research effort directed toward the
second and third bullets also -- the second and third sub-bullets under
Bullet 3? Fire-induced faults?
MR. WEST: Nathan can address the question -- I think the
second bullet.
MR. SIU: Sure. My understanding, Steve, is that this slide
talks about the rules for how you do a deterministic analysis. The
research program is looking at the likelihood of specific failure modes,
what are the factors that affect the likelihood, so it would bear on
that when it's done but right now that's not being factored.
MR. WEST: Okay. We have touched on the disagreements
between the NRC Staff and the industry on some of the issues associated
with circuit analysis, and that is what has obviously caused the
complication in coming to a resolution on the issue.
But basically a few years ago the NRC issued an information
notice that provided information to the reactor licensees on certain
potential failure scenarios that some licensees had found, and it is the
situation that I was just talking about for a control room fire where a
fire that affects the control of an MOV that could cause it to open or
close in a way where it bypasses the limit switch and becomes
mechanically bound so the valve becomes inoperable and stuck in open or
closed configuration.
DR. MILLER: That was 92-12?
MR. SINGH: 18.
MR. WEST: 92-18 -- Information Notice 92-18.
The industry through NEI took a position later that
consideration of that type of damage, fire damage, would be a new
requirement. It is not covered by the current regulation -- kind of
getting back to George's thing, well, what if something new comes up?
We have taken the position that, no, it is not a new
requirement. It is covered by the regulation -- because we didn't try
to, we have never tried to define fire damage in terms anything other
than the component that is needed being functional. A valve that goes
into the wrong position and can't be restored is not functional, and if
that causes a problem with your safe shutdown capability, then it is a
problem under the current regulation.
Then the other -- it's kind of related to the other two --
and this gets back to the issue of once you identify the required
circuits it shouldn't be a problem. You know what they are. You can
protect them.
The problem is, the disagreement is on the number of
fire-induced faults that could occur in a circuit, and the number of
spurious actuations that could occur during a fire. Those are really
the -- I think these are really the two issues that we are struggling
with.
DR. MILLER: Are these disagreements based on judgment or
are they based analysis? I didn't find any, and I read through all the
presentations, the workshop and so forth. I didn't come away with the
conclusion that anybody really knew -- could put down numbers to say
"and here's why this faults occurred." I saw a lot of philosophy but
not much --
MR. WEST: It really -- it kind of gets back to, I guess
back to the Browns Ferry fire, which is the real fire experience we have
that led to these regulations in the first place, but I mean during that
fire there were multiple spurious actuations.
DR. MILLER: But if a plant were designed under current
guidance and regulations, would Browns Ferry occur in the same way?
MR. WEST: No.
DR. MILLER: Are we fighting the last war and not the
present war?
MR. WEST: I would say no. I think if Browns Ferry were
designed in accordance with the Appendix R criteria that you shouldn't
have had the problems they had, the multiple spurious actuations.
MR. MARSH: Are you saying -- I am asking if the industry
position is correct, that there should only be -- and if the plants were
designed and operated in accordance with Appendix R, would they have the
same type of Browns Ferry fire?
DR. MILLER: And would you have the same number of
fire-induced faults per circuit that we had in Browns Ferry?
MR. MARSH: I think you were saying we think plants should
be designed to accommodate this, and if they are designed they should be
able to accommodate that? Are you asking if the industry position is
correct, that there should only be one?
DR. MILLER: No, I wasn't asking whose position. I was
asking what the positions are derived on -- experiments, facts or just
good -- or judgment? If they are derived on what happened at Browns
Ferry we could now maybe agree it wouldn't happen.
MR. WEST: Well, they were derived basically I think from
the Browns Ferry experience.
CHAIRMAN POWERS: Let me contradict you on that and then you
can correct me if I am wrong.
MR. WEST: Okay.
CHAIRMAN POWERS: I get the impression that we had the
Browns Ferry experience. We say, gee, lots and lots of stuff
happened -- it's hard to sort this all out.
Let's take a reasonably conservative position but not wildly
conservative position and establish an analysis tool that will confront
the kinds of things that we saw at Browns Ferry and that involved this
thou shalt look at all the functional failure states of components in a
system that can be affected by fire. Let us concern one spurious
actuation per fire area, I believe, is what the word actually is is one
per fire area.
MR. WEST: That is from where?
CHAIRMAN POWERS: From the 86-10, isn't it? I think it is
86-10. I could be wrong about that.
It strikes me that we did not set up guidance or regulations
that said let us make sure that we can fight Browns Ferry because I
don't think anybody thought you could, that if you had that number of
things go wrong in your plant in a fire you were in desperate shape and
you set up things to avoid that ever happening, and that the analysis is
conservative but not an attempt to fight -- not to, say, reproduce all
the things that happened in Browns Ferry because now you had separation
requirements and redundancy requirements that they didn't have.
That is my perception.
MR. MADDEN: This is Pat Madden. Dr. Powers, I would agree
with you to some extent, providing that the plant was in compliance with
Appendix R, and what I mean by that, if you had complete separation
between these required circuits and these associated circuits -- for
example, one train of those systems needed for safe shutdown or reactor
protection would have been free of fire damage, throwing the one
spurious signal on top or a spurious actuation on time of that separate
scheme I would agree would have been a conservative approach.
What we are getting involved here is we are looking at the
redundancies and the lack of protection for redundancies and arguing
over could, you know, parallel valves and injection path both go closed
as a result of fire damage to those circuits that go through a common
fire area.
If you take a literal interpretation of the regulation, we
would assume that either that A or B valve, those parallel injection
valves, one of them would always be maintained free of fire damage and
available to you.
That is kind of the fundamental position that we are in as
we are looking at to some degree some of this protection is not in
place. It's people pushing to say, well, now let's look at it from a
risk perspective. What is the probability of both those valves if they
are injection valves and they are supposed to be open and they are open
at the time of the fire and they are open prior to the fire, what is the
probability of both of them going closed as one common fire event
occurs?
That is the gray area that we don't have any answers with.
There's no analytical tools and I am not aware of any experiments with
configuration conditions like a single cable and conduit or there's a
cable in multiple cable tray stack -- there's all kinds of different
combinations that could occur that could affect the way a cable shorts.
So to put it in perspective, if we had complete separation,
to go back to what Dr. Powers said, if you throw on top of that a
conservative one spurious actuation, that's probably conservative, but
that is not the area where we're at right now.
MR. WEST: Well, I guess in summary for the Staff, the
Staff's position, I think what we basically say is that regulation --
was never intended that under the regulation we'd get into doing this
detailed analytical analysis of the circuits and trying to decide how
many and what types of faults need to occur before you had some adverse
conditions, but if it was subject to an adverse condition, you would
just provide protection, so that is where we are coming from, and then
Pat mentioned the second thing about protection of redundancies.
CHAIRMAN POWERS: Steve, you are beginning to have a circuit
failure and we are trying to compensate here.
MR. WEST: And this is the last slide -- again, basically on
status. There's a number of initiatives underway to deal with this
issue and come up with some resolution that would be acceptable to both
the Staff and the industry and that resolves obviously any safety
concern.
We had, the NRC had, a while back a path that we were going
to take where we would basically come up with a resolution and issue it
to industry in the form of a generic letter and our approach has changed
somewhat because of the efforts of the BWR Owners Group and the NEI.
CHAIRMAN POWERS: Steve, I think we will stop you and do
some compensating rework on the circuit here.
MR. WEST: Okay.
MR. MARSH: It may already be known by the Committee but I
want to make sure a little bit of the background on this issue comes up.
These issues about circuits and the circuit analyses, as
Steve said, was not a precept of the regulation. The philosophy of the
regulation is to protect circuits. There's barriers and separation --
not to have to go through a detailed circuit analysis because it is
admittedly difficult, and these problems may not have been prevalent 10
years ago. Why not? Because circuits were protected with barriers.
What happened? Thermo-lag and plants to resolve Thermo-lag
may have been removing barriers, doing new circuit analysis, new safe
shutdown analysis in order to have to remove this barrier, which was a
question and a problem. That is why we are seeing perhaps some of these
issues, some of these philosophical and analytical questions come up.
CHAIRMAN POWERS: That is a good insight. That's useful.
MR. WEST: Okay, so anyway, we had a plan that we talked to
you about in June that would have the final regulatory action at least
from NRR would have been a Generic Letter and then obviously I guess
licensees would have had to react to that, and then there probably would
have been NRC inspection follow-up, but because of a couple of industry
initiatives, we have adjusted our plan to basically work with the Owners
Group and the NEI Task Force -- well, not work with them but to wait for
the results of their efforts and they are going to talk to you this
morning about the details -- and to review those to see that they
resolve the issue.
We have had a number of meetings with both over the past
year, so we are continuing our regulatory oversight through that. We
are still doing inspections. The inspections we do are inspecting
against our interpretation of the existing regulation and guidance.
We had a workshop, an industry workshop, last summer, a
one-day workshop, that was well attended by industry, and we went
through in great detail a lot of these issues, tried to achieve an
understanding of where the hard spots were, what could be done to
resolve them, and we considered that in our decision to back off a
little bit and wait for these industry initiatives.
Basically the BWR Owners Group created an Appendix R
committee, and they have identified this effort of theirs to come up
with a resolution of the issues as one of the -- if not the most
important thing they are working on. They sent a letter to Commissioner
Diaz and pointed out that -- described what they are doing and the
emphasis they were placing on their efforts, and they are basically
developing industry positions on a number of technical issues that are
associated with circuit analysis.
For example, one that was mentioned earlier about what is
the duration of a short or how long would a fire-induced condition be in
effect -- you know, is it 30 seconds or is it until some physical action
is taken to correct it?
I think they have identified 20 or so issues that they need
to develop positions on, industry positions, and hopefully get the NRC
Staff to agree.
DR. MILLER: Is some of the research going on addressing
that particular issue, experimental research, at all?
MR. WEST: You mean --
DR. MILLER: By the Owners Group?
MR. WEST: Not by the Owners Group. I don't think they have
any experiments going on.
DR. MILLER: Some of the research at Sandia, for example,
is --
CHAIRMAN POWERS: There is certainly a discussion at one of
your workshops on the experiential base with short circuits of various
types and to the extent that that is based on experiments or events, I
guess there's some experimental work there.
DR. MILLER: These issues can't be unique to the nuclear
industry.
CHAIRMAN POWERS: No, no, I am sure they are not.
DR. MILLER: You may argue what the consequences may be more
here, but the issue of shorts and how long they are going to last during
a fire --
MR. MARSH: The types of impedances that are involved and
the voltages of the circuits and how they relate to the types of
failures that could be there -- that was an excellent presentation and
that is being used as part of the Owners Group effort as an input into
it.
The only other research that I know of is the Sandia
research that is going on. The only other data that has been
presented --
CHAIRMAN POWERS: You might want to look and see what the
military is doing because I know on board ships that are carrying
nuclear power generation devices that operate under extraordinarily
short time scales that we worry about this a lot.
MR. MARSH: Okay.
DR. MILLER: I think if I were designing a big aircraft
carrier with umpty-ump nuclear plants on it, I would be worried about
these things --
CHAIRMAN POWERS: It is not the nuclear plants that we worry
about. It is those short operational nuclear plants.
[Laughter.]
DR. MILLER: The issue of fire-induced shorts and so forth
just can't be unique to what we are dealing with.
MR. WEST: It's not.
MR. MADDEN: No, it's not unique but DOE did some testing in
this area where they actually configured some trays and tried to monitor
cables and look at duration of shorts and output signal, et cetera, et
cetera, but that is very limited testing that they did because they ran
out of funding, but we do have some limited data with respect to it, and
what they were looking at is specifically trying to backfit Appendix R
to their reactors and some of the considerations that they need to do,
to implement because they didn't have the level of separation as far as
through the power ranges that some of the nuke plants, the operating
nuclear plants, commercial plants have.
MR. WEST: Okay, so anyway, the Owners Group effort is
basically deterministic. They are basically saying we have a regulation
we need to implement. We are going to develop positions that interpret
the regulation the way we think it should be interpreted.
We are going to develop a method using those positions that
a licensee could apply to decide what kind of circuits require
protection, what level of protection, et cetera, to meet the regulation,
and they are going to package that up in a topical report that they are
going to submit and have the Staff review through the normal review
process, have us write an SER that hopefully we accept the method. Then
it would be available for any licensee to use.
There may be a couple of -- you notice it is the BWR Owners
Group. There may be a couple of positions that are unique to BWRs but
we think that most of this effort would be applicable to any reactor,
any commercial reactor, and thee is some coordination with the other
owners' groups. I think maybe Fred will talk about that a little bit
later, because the NEI after the workshop also I guess agree to quit
arguing with us and develop, put together an issue task force.
They have done that and we have had a couple of meetings
with them, and they -- Fred is going to talk to you, but I think they
have characterized their effort as the development of a
safety-significance determination method, something like that, and we
are not -- I don't really want to talk about it because I don't think we
understand it well enough to do it justice.
We did get a letter from NEI just yesterday, I haven't even
read it yet, that describes in more detail exactly what they are doing,
where they are going, and I am sure Fred will talk about that today,
right, Fred?
They are also doing some industry coordination and in the
last one of -- a recent meeting we had, it wasn't on this subject I
don't think -- talking about how these two efforts would work together,
exactly how would the NEI method be used, what is it for, what does it
do, how would it be done.
Somebody who was at that meeting from the Owners Group said
that he thought that they would have this deterministic method and you
could use that probably to do a lot of your circuit analysis and decide
what levels of fire protection you wanted, but then you would have some
left where you weren't so sure you really needed to protect them or
provide that level of protection, and then you would drop into -- use
the NEI method and use risk insights and whatever else is in that method
to maybe do a different type of analysis of those circuits and decide
what you needed to do, so it could be some kind of risk-informed input
into the analysis.
We had a meeting last week with Quad Cities, Commonwealth
Edison, to talk about Quad Cities and as you know they had a lot of fire
protection problems there with their safe shutdown analysis and the
various components of it, including the circuit analysis.
We have been following their activities for a long time, a
year or couple years. They had problems with their risk assessment.
But in that meeting they were telling us kind of an approach they used
to where they did a combination of deterministic analysis and based on
that they made changes to a bunch of circuits, but then they did some
kind of risk-informed, applied some kind of risk-informed method to
other circuits and decided they didn't want to make any changes -- they
could do something less than a physical change, maybe just a
compensatory measure or procedure change.
In that meeting we didn't have time to get into all the
details of what they did, but it was intriguing. We were very
interested. We got all the people, most of the people who were in this
room, all the engineering folks and the risk folks, and we are going to
have a follow-up meeting with Commonwealth Edison, working level
meeting, where we get into really the details of what they have done, so
I found that extremely interesting.
It may be the first real case of a licensee risk-informing
their fire protection -- so that's where we're at.
CHAIRMAN POWERS: If NEI formulates a quantitative risk
assessment method, methodology for screening circuits -- I mean
essentially maybe that is an inverse screening of circuits -- and
presented it to you in some forum -- format that calls upon the
preparation of an SER, are you in a position where you think you could
evaluate a quantitative risk analysis?
MR. WEST: Well, with all of the resources of the NRC, I
guess I would have to say yes.
I think we would be obligated to do whatever it took to
review that. I mean after all we are expecting industry to develop
something and I think we wouldn't be asking them to go forward with that
if we didn't think we were going to be in a position to review it and
either accept it or comment on it or reject it.
We are having periodic public meetings with both of these
groups and where they have an opportunity to tell us what they are
doing. We have obviously an opportunity to react to it, so I think if
there was anything, if they were going down some path that we knew was
going to be unacceptable at the end we would take those opportunities to
suggest course corrections.
We will have to see what we get, but I think through the
Staff and Research and our contractors, we would review it.
CHAIRMAN POWERS: Can I turn to you? What Steve has given
us, I think, is a status report here.
MR. MARSH: Right.
CHAIRMAN POWERS: There are some nitty-gritty issues of
lines of language and their interpretation that arise here. I am sure
you saw it in the reading here that we may get into as we go further in
this discussion.
Is this an area you want to chase down, or you think it will
arise naturally or --
DR. MILLER: I assume that -- I am anxious to see the NEI
presentations.
CHAIRMAN POWERS: Steve, I hope you'll stay around for
continued discussions.
MR. WEST: Yes -- I have got a -- we'll work something out.
MR. MARSH: This is this afternoon. We should be here this
afternoon -- we'll be here.
This is another major effort where there's risk information
that is at issue that is trying to be used to resolve an issue. It's
another one of these issues where it is important to work with the
industry to come up with a resolution pathway. It is to our benefit we
believe and the industry's benefit to work in that manner.
I guess my read on this so far, it's a little hesitant, I am
not sure how this thing fits together yet. I am not sure.
We have had a number of meetings with the industry on this
thing, and we need clarity because as an Agency we are holding back
regulatory action because of what the industry is doing and I need
clarity and understanding of what I am relying on the industry to do so
that we don't take regulatory action.
CHAIRMAN POWERS: What I see is the opportunity for getting
a cascading analysis that has no end in this.
MR. MARSH: Right.
CHAIRMAN POWERS: And I take comfort from some of the
presentations that call attention to the fact that what you want to
analyze is not necessarily how you would like to run the plant but how
you may be forced to run it under the circumstance that you judge to be
bounding.
Now it seems to me what I judge to be bounding is the area
where risk might well be capable of telling us how bounding we need to
be, and I am heartened when I see pronouncements that this is the
highest priority for our risk assessment research programs and things
like that.
MR. MARSH: Right, and it is the BWR Owners Group that has
written the Agency and said this is one of the most important things
they have got on their plate to resolve -- they wrote Commissioner Diaz
and said that, but I share your concern that this could be kind of an
endless loop, an endless circle, and we need to be mindful of that in
our responsibilities to not let that happen.
CHAIRMAN POWERS: That's right.
MR. MARSH: Come to a resolution pathway that is not
risk-based either. It just needs to have a risk implication in it, a
risk-informed approach in it, but not as a "dispense with this because
it has no risk implications" -- they need to be careful of that.
CHAIRMAN POWERS: But again I think that risk does not
provide a vehicle for dispensing with this.
MR. MARSH: Right.
CHAIRMAN POWERS: Simply because circuit problems do arise
in fires.
I am going to declare a break until -- I have been advised
or dictated -- until 11 o'clock.
[Recess.]
CHAIRMAN POWERS: Let's come back into session. We have Pat
Madden to talk about the fire protection functional inspection pilot
programs.
Pat, I understand you are going to focus in your discussion
on a summary of the inspection findings?
MR. MADDEN: That's correct, Dr. Powers.
CHAIRMAN POWERS: I hope that we will have time at the end
to discuss implications of those findings.
MR. MADDEN: Well, the ending is going to be kind of crazy
and bizarre and I have a lot of questions for you, George, and you can
help me out here. There's a lot of questions on the table and I am
going to pose those questions in a broad perspective.
CHAIRMAN POWERS: Good.
MR. MADDEN: What I would like to do is go over the program
objectives. Some of what we call the defense-in-depth principles -- you
have seen this before. Ed Connell talked about them a little bit this
morning.
I am going to talk about the scope of the FPFI Program. The
NRC's memo talked a little bit on a broader sense on the current NRC
inspections program versus what we did in the FPFI and where we found
some holes in the current program versus the FPFI.
We're going to look at the actual pilot inspection effort
and what we considered, what plants were part of the pilot program.
I am going to give you a little bit of a summary of what I
want to call the -- a summary of the findings and to some degree the
insights that we got looking at the program as a whole and looking at
the way NRC does inspection business as a whole.
Then the last two bullets of my presentation, assessing risk
and safety significance of the findings and future plans, is where I
have a lot of questions, and we'll pose some of those questions and
where we are going to go with them and what we are trying to answer and
what we are trying to achieve.
The objectives of the FPFI program was to revalidate or
reevaluate I should say the scope of the inspection program or the fire
protection inspection program, as we know it today, i.e. for example the
core program, take a look at that, take a look at the initial Appendix R
inspections and were they good enough and broad enough, and then there
was a triannual module that was supposed to go back in and revalidate
Appendix R compliance which was not used very much by either the
industry or the NRC Staff. It really wasn't implemented.
The FPFI program was to try to develop a coordinated
approach towards doing comprehensive and complex fire protection
inspections, and by that we kind of used integration of people with
varying knowledge including a risk analyst, a systems analyst,
electrical and I&C type specialist, and a fire protection engineer.
The other aspect was to provide a strong and broad-based and
coherent inspection program, and what we were trying to do is trying to
standardize and trying to come up with a methodology that an inspector
in the field if he wanted to use certain aspects of the FPFI module, he
would have all or a good deal of the guidance in front of him that would
guide him on how to do an assessment of a certain program area in fire
protection.
CHAIRMAN POWERS: Is what you are saying that you tried to
design an inspection program for fire protection here?
MR. MADDEN: Well, it is a little bit more than fire
protection.
CHAIRMAN POWERS: Yes.
MR. MADDEN: Post-fire safe shutdown, circuit analysis. We
also looked at what I want to call auxiliary issues that aren't covered
by the regulations -- for example, event-based fires -- for example, a
turbine fire results in a fire and what impact it would have in a plant,
and also try to look at what we want to say are phenomena that we are
just kind of probing in but it's called a fire-induced transient where
you would actually have a transient caused or impact on a plant, i.e.,
what comes to my mind is spurious opening of SRVs, for example, and what
impact that may have on a plant and how a plant would react. --
CHAIRMAN POWERS: I guess you triggered two questions there.
The first question is where did our core inspection program
for fire come from, and the second one is what caused you to grab hold
of this fire-induced transient?
MR. MADDEN: What caused us to grab a hold of it, you know,
logically speaking we went back a little bit in history, of course we
always revert back to Browns Ferry, but we asked ourselves the
question -- you know, we looked at it classically from an exposure
perspective of fire in a room creating a condition that could impact
your ability to safety shut down, but we don't -- at one point in time
we were trying to look at it from a broader sense to say, well, what
could a fire do to a plant and how could it change its configuration and
could those changes in configuration be managed and identified by plant
operators, and even though it may not affect your "ability to safely
shut down" in a classical sense with respect to Appendix R, could it
affect your ability to safely shut down or cause you to use drastic
measures to shut the plant down as a result of a fire in a space that
you would have not analyzed with respect to post-fire safe shutdown in a
classical way.
CHAIRMAN POWERS: Are you -- see, what you did not say was
you did not say gee, I looked at an awful lot of PRAs and what-not and I
found these transients were frequency dominant or risk dominant or
something like that.
Is what you are saying that you find these risk assessments
to be deficient in that they don't cover that sort of thing?
MR. MADDEN: Looking at the IPEEEs, we couldn't find any
indication where they covered any of that. They only covered it or
looked at it in a classical sense with respect to 100 percent power
operations and shutting the reactor down without any perturbations or
external perturbations that may be caused by fire.
CHAIRMAN POWERS: Okay, so what you are doing is you are
really raising a question that the risk analysts ought to pay some
attention to, whether they are being complete or not.
MR. MADDEN: That's correct, that's correct.
CHAIRMAN POWERS: I think that is very, very important. I
think that it is an issue that comes up all the time, this completeness,
and I suspect if I were going to look for incompleteness it is going to
be at the interfaces between disciplines and here you have the
discipline of fire protection versus PRA during operations.
I think that is an interface that is fraught with a
potential for incompleteness.
DR. APOSTOLAKIS: Of course the issue exists also in the
current regulations.
CHAIRMAN POWERS: Oh, absolutely.
MR. MADDEN: That's correct, and that is why we probed in
that area.
CHAIRMAN POWERS: What I see emerging, and do correct me if
I'm wrong, is coming out of this pilot program a suggestion that perhaps
our core fire protection inspection is looking at certainly an
incomplete set of things, but perhaps an inadequate set of things as
well, and so the question that comes to my mind is where did that core
inspection program come from?
MR. MADDEN: Well, I'll get there.
CHAIRMAN POWERS: Okay.
MR. MADDEN: In this presentation we'll go through --
CHAIRMAN POWERS: Good, good -- I didn't -- I can wait.
MR. MADDEN: -- a little bit of the insights that I brought
forward.
To get back a little bit on track, like I said, we wanted to
provide some clear guidance regarding oversight not only to the NRC but
also to industry if they wanted to use the FPFI module for
self-assessment. We wanted to make sure that they had a good clean set
of rules to do that self-assessment by.
The final aspect that we thought would be of benefit to this
program was to refocus everybody's attention on fire protection and
compliance with not only their licensing basis but design basis in
current regulations.
CHAIRMAN POWERS: But that -- go ahead, George.
DR. APOSTOLAKIS: The most important item or bullet in my
mind is Number 4, focus on the fire protection issues of most
importance. The issue there is of course how do we define importance
and then are we really doing this, are we really focusing the NRC
resources on the most important issues.
MR. MADDEN: Well, we tried to augment this program a little
bit. We used what information we could get out of the IPEEE.
DR. APOSTOLAKIS: So we will get to this?
MR. MADDEN: Yes. I'll just give you an oversight here --
that's where we started the inspection, you know, as far as looking and
scrubbing and looking at areas. That is where we started, okay, but in
the other aspect we looked at we concentrated our focus not only on
those areas but if you had a fire in those areas, how are you going to
shut down those areas and what could be affected?
That is where we focused primarily when we started the
inspection and we might have branched out as a result of us looking at
those primary plant areas.
CHAIRMAN POWERS: Let me comment just a little bit about the
last item on this -- "provide an immediate safety benefit arising from
new industry attention to nuclear power plant fire safety."
I think that if my reading of the comments and presentations
made at the workshops that you have had is that the FPFI Program did
this and that it was universally applauded by both the regulators and
the regulatees -- maybe "applauded" is a bit of a strong term but
recognized that that happened, and I came away saying, gee, that's
distressing, that I would have been much more comfortable had the
industry said this program was unnecessary -- gee, we're very focused as
it is on fire protection issues.
MR. MARSH: I don't think they said that. I think what you
read is -- there is another flavor that goes on with it too. I think
they said there were issues in fire protection that need to be addressed
and you helped us address those issues. I am not sure they used the
word "safety" but they talked about fire protection issues.
They went on to say, and Fred I'm sure is going to talk
about this some more, that they believe that the overall effort was way
too intensive, that it caused a lot of industry resources and too short
a period of time to be focused on these things, and their belief, if I
can speak for a minute, in a disproportionate way -- that is, a lot of
attention on things that may not have been significant and so they are
proposing something different and a different allocation and a different
timeframe for looking at fire protection issues in the future.
But you didn't hear, I didn't hear, nobody heard "you guys
are off-base." There weren't any issues out there -- inappropriate
expenditure of Agency and also industry resources -- because everything
was fine. You didn't hear that. You're right.
I would have felt -- I think we constructed this program
thinking that was the problem that was there. There needed to be some
refocusing of industry attention towards fire protection programs, and
that is what we found, and that was an end-result as well, and not just
in the four plants either, I think in other plants as well.
CHAIRMAN POWERS: I think you are having an impact, a ripple
effect --
MR. MARSH: Right.
CHAIRMAN POWERS: -- beyond the four or six plants,
depending on how you care to count here.
MR. MARSH: In the opening comments this morning I said that
this is perhaps the most important thing that we have done this year --
these FPFI programs -- not just because they examined four but because
of what -- the ripple effect that they have done, the ratcheting, the
leveraging effect.
MR. MADDEN: Trying to get back on track, when we modelled
this or developed this procedure, we looked at basically assuring
ourselves that we still had what we want to call the principles of
defense-in-depth fire protection still being applied at plants, and that
there wasn't any findings or weaknesses. That was our goal, that as we
did these pilots that we would not have any finding of weaknesses and
find strong programs in those programmatic areas.
The areas again are preventing fires, detecting, suppressing
and putting them out quickly would be the second element, and then the
third element of course is designing plant safety systems so that if you
did have a fire that wasn't prevented and of course wasn't detected or
suppressed properly that you would still have the ability to maintain
those essential plant safety functions that you would need to either
shut down or compensate for whatever else was going on in the plant as a
result of the fire and potential perturbations caused by the fire.
Another aspect is that, you know, I hear from time and time
again that Appendix R is not performance-oriented, but it is. There is
one specific goal that we did carry through in this theme of the FPFI
and it is a performance goal and it is almost directly out of Appendix
R. It provides some level of safety margin.
I guess you can call it an interpretation of mine or
whatever but it basically says a fire event should not cause the reactor
coolant process variables to deviate from that which would be predicted
by a Chapter 15 accident analysis for a loss of offsite power.
That is a pretty fine point in the regulation with respect
to what kind of safety margin was being anticipated at that time when
the regulation was being promulgated -- so that was part of the theme of
the FPFI.
CHAIRMAN POWERS: I guess I don't understand quite -- at
first it sounds, gee, that's kind of logical -- let's make sure that a
fire doesn't make things any worse than the loss of onsite power, but I
think that a fire does make things worse because with a loss of onsite
power, I know what things fail and I know the direction they fail in,
but it seems to me that with a fire the problem we are running into is
there are many, many different ways things can fail and some of those
failures because they are at once variable and time-dependent make it a
much more complicated system than a loss of onsite power.
MR. MADDEN: I won't disagree with you, but I think what the
message there was was to try as much as possible to protect normal
systems or systems that you would know that could bound or maintain you
within those predicted levels.
If that statement wasn't in there for example on a PWR, you
could go to a bleed-and-feed type operation and reduce the number or the
scope of equipment that you would have to actually protect. Now is that
safer? Now in some risk circles people can draw a parallel that it is
no riskier to do that type of operation than it is to shut down
normally.
Now do I have a comfort with that? I think I would much
prefer having the operator staying within normal bounds of predicted
systems to shut the plant down than having to jump in deeply into an
EOP.
Okay. Let's talk a little bit about the scope of the
inspection procedure and what we call the primary focus and the
secondary focus. And, like I said before, you know, I kind of alluded
to you that we did use risk-informed, or we tried to use a risk-informed
approach, or gather data from the IPEEE, and we did that a risk analyst
that accompanied on most of the inspections.
We started out by assessing the implementation of the fire
protection program in the risk-sensitive areas of the plant.
And then the other primary focus was to look at
configuration management. You know, how well are these plants
maintaining their license and design basis in the configuration that the
NRC perceived as -- you know, that they have implemented? So that was
the primary focus of the FPFI.
The secondary focus was to look at what I want to call these
fire-induced vulnerabilities and try to come up with a handle on, you
know, what we should do about those, gather information about those and
try to look at those findings in the future to determine if there are
some risk implications associated with those findings. But, as you will
see, we had our hands full with just the what I want to call classical
compliance issues, and we had a very limited opportunity to look at the
secondary --
DR. APOSTOLAKIS: Don't take that away. Why can't we take
the thought in your first bullet and apply it to the Regulatory Guide
whose outline we saw earlier today? In other words, I can have a graded
approach and say that, yes, the idea of defense-in-depth, as it has been
defined in the fire area, preventing fires and so on, will be applied to
the areas that are fire-risk-sensitive, and in other areas, I will do
something less, I will not apply the same criteria. Now, that would
immediately start making the Regulatory Guide risk-informed.
MR. MARSH: That may be how this Reg. Guide gets
implemented, but not in its development. The development of the Reg.
Guide is supposed to just pull together existing regulatory guidance,
regulatory requirements, what is the existing verbiage that is out
there. Now, that how that gets implemented -- if there are issues at
plants where they may not be in compliance.
DR. APOSTOLAKIS: But if the Regulatory Guide says that you
have implement defense-in-depth as defined here, the licensee cannot
come back and say we are doing only in the fire -- in the sensitive
areas.
MR. MARSH: Yes.
DR. APOSTOLAKIS: Can they? They can not. They can? I
thought the language of the guide should allow that, the language is not
there. It very clearly states this is 4.1 objectives.
MR. WEST: No, the guide -- the guide will say that the
methods that are specified in the guide are one acceptable way of
meeting the fire protection regulation. It doesn't mean it is the only
way, and it doesn't mean you have to do everything it says in every
area. Licensees --
DR. APOSTOLAKIS: But I mean since we have it here, Steve,
why don't we say it ourselves?
MR. WEST: It does say that.
DR. APOSTOLAKIS: It does say that?
MR. WEST: Well, you have an outline, you don't have the
full guide. Go, for example, to the -- one of the existing guidance
documents, the Standard Review Plan. It says that the guidance in the
Standard Review Plan is one acceptable means of meeting the regulation.
DR. APOSTOLAKIS: I understand that, it is one acceptable
means, but finding another sometimes is not easy. And if it is already
acceptable to the staff, --
MR. WEST: Well, we have 900 exemptions from the regulation,
for example, out there, other acceptable means of meeting the underlying
purpose of the regulation.
DR. APOSTOLAKIS: And I thought the idea of this Regulatory
Guide was to get away from this huge number of exemptions.
MR. WEST: No. No, you are -- remember, parallel efforts.
Think of the parallel efforts. The 805 will do that.
MR. MADDEN: George, my view of the Reg. Guide and, you
know, my bosses can correct me, but it is a vehicle. You have got to
start out with the deterministic space and then you can branch out as
you gain knowledge or look at certain aspects of fire protection that
lends itself to risk-informed approaches. You can automatically
incorporate those into the Reg. Guide in saying that this is an
acceptable alternative. But you can amend the Reg. Guide, and it is a
lot easier to amend the Reg. Guide than it is to try to change a
regulation or to go out with a Generic Letter, for example. So it
becomes a tool or a vehicle to where we can springboard into some kind
of approach like that.
DR. APOSTOLAKIS: Yeah, but if -- I mean there are methods
and there are objectives. And I see how a licensee can come back and
say, well, we have an alternate method that is not in your Regulatory
Guide and you guys will review it.
MR. MADDEN: That's correct.
DR. APOSTOLAKIS: But they cannot come back and say we have
alternate objectives. And if you state them as objectives and you say
that this section describes the basic objectives of the fire protection
program, and the very first subsection is defense-in-depth --
MR. CONNELL: Yeah, those objectives, George, -- this is Ed
Connell again -- are out of the regulation, so they can't be changed by
the Reg. Guide.
DR. APOSTOLAKIS: In other words, we cannot implement this
then. The licensee cannot go back --
MR. CONNELL: We would have to change the regulation to
change those objectives. Those objectives that you see in that outline
are directly out of the regulation. In order for us to -- a Reg. Guide
cannot change the regulation, we would have to change the Regulatory
Guide in order to do what you are suggesting. And industry has
indicated that they do not want to unilaterally change the regulation to
make everybody go into a risk-informed, performance-based fire
protection environment.
DR. APOSTOLAKIS: Well, first of all, that is not what I am
saying. I don't think that it is all or nothing, you know,
risk-informed, performance-based or what we have now, but I mean some
insights, some flexibility would be warranted. So how is this program
then, that you are describing to us, going to affect the Regulatory
Guide? It will not? That is for later.
MR. MADDEN: Yeah. This program wouldn't -- if the
Regulatory Guide was risk-informed, had risk-informed methods in it, we
would translate the Regulatory Guide or the regulation into an
inspection procedure. And, you know, based on the licensing basis, the
safety evaluation which approved their program and their approach, that
is what we would -- we would amend our inspection procedures to
implement a methodology to go out and inspect that.
DR. APOSTOLAKIS: So back to the FPFI program objectives, to
provide clear guidance to the staff and industry regarding oversight of
reactor fire protection programs.
MR. MADDEN: As we know it today. It is not as we know it
in the future, it is as we know it today.
DR. APOSTOLAKIS: So this is in the future, the words "in
the future" are missing here.
MR. MADDEN: No, it is not. The module is already drafted,
it is already in place. I mean it is as we know the licensing basis of
these plants today. Now, if we go to an 805 approach or some other
approach, we are going to have to make adjustments to our inspection
program appropriately in order to inspect and enforce those new fire
protection requirements.
DR. APOSTOLAKIS: So the picture is this, that we have the
Regulatory Guide which consolidates the regulations, I mean the guidance
of today.
MR. MADDEN: As we know it today.
DR. APOSTOLAKIS: As we know it today. And then we have the
inspection program. And the inspection program is allowed to look at
the risk-significant areas.
MR. MADDEN: We start out with risk-significant, but we use
the deterministic tools to judge compliance in those risk-significant
areas. We focus our resources on those risk-significant areas. We
can't look at the whole plant, we can only take a sample, so we sample
the risk-significant areas.
DR. APOSTOLAKIS: But the licensee is supposed to implement
this Regulatory Guide to all areas.
MR. MARSH: No, not the Reg. Guide. The Reg. --
MR. MADDEN: He implements the regulation. Okay. And the
Reg. Guide is one tool or way he can implement that.
DR. APOSTOLAKIS: But the Regulatory Guide reflects the
regulation.
DR. POWERS: Is it fair to say that no plant will be found
to follow the Reg. Guide absolutely?
MR. MADDEN: That's correct.
DR. POWERS: That all of them will take some exception --
MR. MADDEN: Some exception, right.
DR. POWERS: -- or some shade of meaning or something.
MR. MADDEN: Yes. As a part of their licensing basis
review, those exceptions are identified and the staff will make a
judgment against the Reg. Guide under alternative approach and if -- you
know, the alternative approach, based on their situation, and this is
what I want to call the engineering judgment type performance look on
their situation meets the underlying intent, the staff will find it
acceptable.
DR. POWERS: It's not -- my point is, this is not unusual.
MR. MADDEN: No, it is not unusual.
DR. POWERS: In every single plant, you will find some
difference between the template established by the Reg. Guide and what
is actually done, but in no plant should you find a difference between
what is done and what the regulations say.
MR. MADDEN: Right. Generally --
DR. POWERS: Without an exemption.
MR. CONNELL: Right. This is Ed Connell again. I don't
think that we are smart enough to write a perfect Reg. Guide that can be
applied 100 percent at 103 plants.
MR. MADDEN: I think that's true.
DR. POWERS: So, judicious humility there.
DR. APOSTOLAKIS: So just so I understand what is going on,
we are risk-informing the inspection process, but not the regulations?
MR. MADDEN: No, that is not true. We are using the
insights.
DR. KRESS: They are risk-informing this particular
inspection.
DR. APOSTOLAKIS: Just fire protection.
DR. KRESS: It is a one time thing.
MR. MADDEN: Well, we are hoping it is not a one time thing.
DR. APOSTOLAKIS: He is gaining insights.
MR. MADDEN: We are using the insights from the IPEEE, yes.
And I think if we did a review today, we still use the insights as a
part of the review process for any kind of license amendment, from the
IPEEE. It is just one --
DR. APOSTOLAKIS: So if you do an inspection a year from
now, it will be influenced by this, right, the FPFI?
MR. MARSH: FPFI modules.
MR. MADDEN: That's correct.
MR. MARSH: Sure, you would.
DR. APOSTOLAKIS: So the inspection is risk-informed.
MR. MARSH: Sure.
DR. APOSTOLAKIS: It could be risk-informed, but the
regulation is not.
DR. MILLER: The regulation is not, no.
DR. POWERS: The regulation is what it is right now.
DR. APOSTOLAKIS: Which is not risk-informed.
DR. POWERS: I think that is wrong.
DR. APOSTOLAKIS: What is wrong?
DR. POWERS: To say that the regulations are not
risk-informed.
DR. APOSTOLAKIS: They don't use PRA.
DR. POWERS: What we now have are tools to do more
quantitative, more systematic assessments of risk than we had in the
past, but I think that every one of the authors of every one of the
regulations, save perhaps for a few of the legal ones, felt that they
were operating from a risk-informed perspective.
DR. APOSTOLAKIS: I understand that point, but what I am
saying is that we use the word risk-informed the way we have been using
the last two years, namely, using the quantitative results of what is
known as PRA, so let's not argue that point. Of course, there was
risk-information embedded in the regulations before. So what Pat is
describing to us here is an inspection program that will use these
results. In that sense, it is risk-informed. But the basic regulation
does not.
DR. POWERS: Yeah, it didn't exist at the time they were
written.
DR. APOSTOLAKIS: That is correct.
MR. MARSH: Just a point of clarification --
DR. APOSTOLAKIS: Does that make sense? Okay. Okay for
today.
MR. MARSH: Pat is telling you how we have done four FPFIs.
Okay. Pat is not telling you where we are going to go. Okay.
DR. APOSTOLAKIS: I understand. I understand. But I
suspect there is a strong correlation with where we are going to go and
the findings here.
DR. POWERS: I think, in fairness -- it is my perception --
DR. APOSTOLAKIS: I am not trying to be unfair. To
understand --
DR. POWERS: Let me say it is my perception that the FPI
pilot findings in the conclusions of that are opening up the debate on
inspections.
MR. MARSH: Absolutely.
DR. POWERS: That is my impression.
MR. MARSH: Absolutely.
DR. APOSTOLAKIS: Opening up which debate?
DR. POWERS: Debate on the inspection program.
DR. KRESS: It may need to be more --
MR. MADDEN: Yeah, it is opening up a unique debate and we
will get some of those questions on the table here.
This is to go back to your question, Dr. Powers, on the
baseline of the core, and I don't know what it is going to be called in
the new context of the cornerstone. It may be called complementary
inspection, or I don't know exactly what the new buzz phrase term is.
But we do have a module, it is called 64704, which is predominantly done
routinely by the regions at all the plants.
And in that center section you can see, these are the items
that are looked at repetitively. And when I show you the results of the
FPFI, you will see that to some degree there is either weaknesses in
here with respect to the core, that we weren't looking at the right
thing, or we weren't being critical enough, or the items, like
administrative controls, for the most part. We did find some findings
in the administrative control section, but you could probably say that
everybody looks at that, everybody looks at how you control transients,
so, therefore, plants do it quite well.
I looked at the quality assurance program the core did, but
it only looked at in only looked at it in the sense of auditing fire
brigades and some what I want to call routine programs and never really
evaluated the engineering or the design issues.
Generally speaking, for, I think it is an 18 month period,
fire protection is either under the baseline or core, is focussed on
either a resident or a regional inspector and about 32 hours of
inspection effort.
In the initial phases of the late '80s -- or early '80s, I
should say, we conducted a one time inspection, it was Appendix R
implementation. It was done under 64100, and that was a headquarters
regional initiative, one time inspection, one week, basically, five
people, I believe, so five people for basically three days on-site
inspection time, equating to about 128 hours to inspect the actual
implementation, safe shutdown analysis, et cetera, et cetera, associated
with Appendix R.
There is another inspection procedure that is on the books
that could be used as a regional initiative but has sparsely been
implemented, and that was intended to look at long-term compliance of
Appendix R. So, to my knowledge, I think Region IV is the only one that
did any of those type of inspections, and they were done quite some time
ago.
Understanding this, understanding the thermo-lag dilemma,
understanding the self-assessment report that was one on the staff,
there is a lot of issues in fire protection, so that raised -- or that
is why we developed the FPFI program. Like I said, the principal focus
of the FPFI was to look at fire protection features, post-fire safe
shutdown design and compliance, the fire protection licensing basis and
look at the program elements, you know, of the defense-in-depth and how
they integrated and how they were actually implemented in an integrated
fashion at a plant.
What was being done in the past, everything was being looked
at fragmented and, you know, the core looked at administrative controls
and fire brigade, for example, and that was its primary focus. You had
a one time look at Appendix R and its implementation. Nowhere in
between was anybody really looking at how adequate 5059s are, how
adequate configuration management was, how adequate the actual
robustness of the Appendix R safe shutdown analysis was.
So the FPI program basically equated to 400 hours on-site
and there is -- and the initial inspections, I believe, three of were
led by headquarters with a fire protection engineer, like I said, an EE
with INC and electrical engineering background, a reactor system
engineer or analyst, a PRA specialist and a regional fire inspector.
We did -- I showed -- I had that last bullet on the bottom
of that slide, but I want to couch it with the actual pilot program.
The last concept, we did try a different concept, and that was looking
at some self-assessments where we cut the scope back to look at how well
the licensee performed its own self-assessment using the FPFI
techniques, and, so, therefore, we cut back on the scope of our
inspection and pushed it back to the 128 hour concept similarly used
under the 64100 module.
The FPFI pilot program, we looked at -- we did three full
scope inspections, two BWRs and one PWR, and then we did two reviews of
self-assessments, one BWR and one PWR, and then we used the FPFI
techniques on a confirmatory action validation inspection, which was a
BWR, and I think can guess to some degree which plant that was.
And I'll have to show -- yeah, that shows up pretty good.
This is kind of like what I want to call the scattering of the findings.
We had approximately 140 findings out of the inspection and, you know,
if you want to equate it into -- we had -- if you start around the
circle, we had 10 percent -- 10 percent of the findings were associated
with implementation procedures for safe shutdown; 6 percent of the
findings were dealing with what we want to call alternative shutdown
implementation; 6 percent -- or 5 percent, I mean, 5 percent of the ASD
implementation; 5 percent in emergency lighting; 2 percent in
communications; and those -- those four chunks right there go all hand
in hand.
I mean you have to good communications in order to do the
implementation. You have to have adequate emergency lighting to do
shutdowns and move through the plant. And you have to have adequate
procedures in order to do the proper implementation. And what we mean
by ASD implementation, some pieces of equipment or steps in the
procedure didn't match the safe shutdown analysis and, therefore, were
not picked up in the field as a required action, so there were some left
out actions from the procedures.
Then we get into some of the classical fire protection where
we had 5 percent in fire prevention, and here are some surprising
things. Twelve percent in fixed suppression, and that's -- what we did
there is looked at the design basis and how the suppressions systems, as
installed in the plant, actually met the code or record with respect to
like the NFPA codes and the sprinkler code for installation, sprinklers,
halon code for the installation of the halon system. So we did find
some surprising findings in the way these systems were installed in
plants.
DR. POWERS: My experience with those kinds of findings is
colored heavily by experience within DOE facilities, where in some cases
the discrepancies were catastrophic. Detection systems that simply
didn't detect, and couldn't possibly have ever detected even if they
were fully operational. Suppression systems that had been -- that are
on the drawings that are in the safety analysis report, but, in fact,
had been blanked off.
What -- can you give us some feeling for the -- you know,
well, it said put a suppression sprinkler nozzle every 12 feet and they
were 11.6 feet or something like that?
MR. MADDEN: Well, not to that extreme, but what I would --
how I would couch it is that, you know, we use a reliability factor, a
factor that these things are 98 percent effective against -- you know,
we use industry, nationwide industry data, and I don't think that is a
fair portrayal in the nuclear industry, because, you know, if I were to
argue on one side, I could say, well, you know, the general industry is
probably more conservative than the nuclear industry. You know, the
nuclear industry took liberties with the rules. For example, they put
in the sprinklers, but then they went and ran things underneath the
sprinklers such as duct work, et cetera, et cetera, but they never went
back and looked at the sprinklers, or they misapplied head.
You know, for example, the UL listing and the way they were
tested for a certain head, it had to be installed in a certain
configuration, and this and that, you know, and they had specific
criteria based on the testing in order to get optimum performance in
order to meet the minimum code requirements. Well, they misapplied that
type of -- they never looked at that data, but they misapplied the head.
You know, they installed it and they used a configuration that should
have been installed.
So, you know, with that aspect, you know, that is what I
talk about a little bit about configuration management, making sure that
your system keeps abreast of plant changes, et cetera, et cetera.
And the same thing would hold true for the detection
systems. You know, there is a fundamental argument about high ceilings
beam pockets the way detectors are tested. Detectors are usually tested
on -- for optimum performance, and that would be for general industry,
it would be on 15 foot high ceilings with no beam pockets. So now we
have 30 foot or 40 foot ceilings beam pockets and we expect the detector
to respond in the same manner it would have been when it was tested by
UL and it is -- you know, to meet the standard.
DR. POWERS: Well, they just can't possibly do it.
MR. MADDEN: Right. So, you know, but yet, you know, this
goes back to risk. You know, what data do we use to -- and what kind of
judgments do we make to say that we have optimum detection in these
plants and optimum suppression?
DR. POWERS: That particular kind of thing is one that is
going to very, very difficult for the risk assessment analyst to come up
with, because --
MR. MADDEN: That's correct.
DR. POWERS: -- he is going to sit down and say, okay, you
have got a detector here, and he says, well, how good is that? And he
is going to go look at the UL test data, because there is not going to
be any test data for --
MR. MADDEN: I guess the point I am making is that making
the nuclear industry, to some degree, has to take a conservative
approach to install additional detectors in lieu of trying to stick with
what I want to call the traditional methods of the NFPA standards, for
example, or the traditional methods of -- you know, with respect to
sprinklers, when in doubt, install a sprinkler head under an
obstruction, you know, because plants change and configurations change,
et cetera, et cetera.
So, to go on, we had 10 percent of the findings were in
passive barrier design. Some of that dealt with penetration seals.
Some of it dealt with thermo-lag barriers and the implementation of
upgrades of the thermo-lag barriers. We had 8 percent of findings dealt
-- or were associated with the IPEEE and some of the assumptions that
were done with the IPEEE. Five percent was in that area that I said we
didn't spend a lot of time, the evaluation of plant transients and -- or
4 percent, excuse me. And 22 percent of the findings were associated
with SSA, safe shutdown analysis and its implementation.
MR. MARSH: Pat, can you take a minute and talk about, or
maybe you are going to talk about it a little later on, some of these
fire brigade issues that were so surprising to us?
MR. MADDEN: Yeah, I can when I get to the slide. I mean
the next few slides are real quick overviews as far as where we found
them and how many plants they were associated with.
DR. POWERS: Now, the trouble with those slides is, again,
they don't communicate to me what in your written words you have said
you need to do there. They don't communicate to me the significance of
the findings.
MR. MARSH: Right. Right.
DR. POWERS: And when you get to the bottom, to the last of
those slides, when I have seen them before, I don't know and I haven't
looked ahead yet in your package, your written word, you say I can't
understand this until I get a plant-wide view.
MR. MADDEN: That's right.
DR. POWERS: And I don't have that yet.
MR. MADDEN: And you won't have it until I tie it all up at
the end. I mean it's -- you have got to -- and that's why I kind of
stress, you have got to keep in mind what I want to call
defense-in-depth, you know, and where we are going to draw the line here
is you are going to see that by doing the FPFIs, we identified
weaknesses, or I should say findings in all the levels of
defense-in-depth with respect to fire protection, George, only that.
Okay.
So -- and to say, you know, what is the punch line, is this
-- how risk-significant is this, how gross is this, how -- I guess what
I want to say is, in a classical sense, it didn't meet the expectations
of current guidance and regulations.
So in the administrative end point of it, we had three out
of six plants where they had control combustible problems, or
weaknesses, and they may have been isolated. But the point I guess I am
making here is that they did have a process, a program and yet that
program has -- or you can circumvent the program, and it can go on for
some period of time. In one plant, the uncontrolled combustibles had
been there for some period of time.
DR. POWERS: And that is simply one of those issues that
seems to have a life of its own. I am certainly familiar with a DOE
plant where we cited them for transient, uncontrolled transient
combustibles every single inspection. They were removed and replaced
every single inspection, and our inspections were substantially more
frequent, they were every three months. It was always a different set
of combustibles, but it was always there.
DR. MILLER: Absolutely.
MR. MADDEN: Compensatory measures, two out of the six
plants, we had some views or findings associated with how they implement
compensatory measures. And by that, I mean when they do an operability
assessment, do they look at the whole problem, or do they just go for
the quick fix compensatory measure? But if it is like a safe shutdown
function that is out of service, do they only put a fire watch, an
hourly fire watch in the area, or do they take the step further to say,
well, if this safe shutdown train is out of service, what kind of things
could we do to get around this problem, operationally-wise, and alert
the operators?
The other one out of two, and there was only two PWRs we
looked at, but one of them had problems with their oil collection
system.
The next area, the second element of defense-in-depth and
that is detection and suppression. I think over on the lefthand side of
the slide it kind of couches it right -- Designer installation did not
meet industry standard, or code or record. In absence of an engineering
evaluation to evaluate those deviations, some of them you can simply
dismiss, but some of them, you know, would have warranted either
additional detectors or additional sprinklers.
So we had one area, or one -- five out of six plants had
problems with detection. Five out of six had sprinkler concerns. We
had a gaseous suppression system and we looked at the acceptance, the
initial acceptance tests of the gaseous suppression. There were two out
of six plants that we looked at, two of them had -- we had concerns over
the initial acceptance of the test -- testing. It actually had failures
and lack of ability to hold a concentration, but they did an engineering
evaluation to determine that that was acceptable.
One plant out of the six we had questions about the
suitability of the agent for the hazard and the design concentrations of
the gaseous suppression. For example, cables can be considered if -- if
a cable fire progresses to a deep-seated fire, the agent was not applied
for that concentration to control a deep-seated fire.
MR. MARSH: Let me interject a thought here as we go through
these slides. These slides could also be read as not giving any sense
of significance, okay, any either risk significance or of any safety
significance, or compliance significance or anything, these are just a
frequency charts. These -- it says, on how many plants did we have
findings of this sort?
DR. POWERS: The trouble with that I have is that some
people will take it as an indicator of significance.
MR. MARSH: We need to be careful about how these -- what
these connote.
DR. POWERS: Yeah, I think that -- I think it is a real
problem because there are people that will leap upon this as proof --
MR. MADDEN: Right.
MR. MARSH: Right.
DR. POWERS: -- of this or that contention.
MR. MADDEN: Right. You know, and I -- yeah, I do want to
caution. I mean these are findings -- I mean, you can go back to the
details of the report and look at the findings. Now, would I classify
any singular finding as being drastically risk-significant? Probably
not, because I have all these other layers of defense-in-depth that I
would rely on. But if you start, you know, looking at the whole picture
and why we are having so many findings, and where we are -- why we are
where we are today, then that raises other questions.
MR. MARSH: It raises questions in my mind about -- how come
five out of the six plants that we looked have detection issues?
MR. MADDEN: Right.
MR. MARSH: Is it something that we haven't said clearly, or
that the guides aren't clear, or is there -- what is the synthesis of
this information? Not significance yet, just the synthesis of it. If
you see a large fraction, it should raise questions in your mind, but
that is the only thing that it should do, is just raise a question in
your mind.
DR. POWERS: Well, with respect to detection, it is my
experience that actually what I find surprising is it is not six out of
six, because, quite frankly, you have an awful lot of detectors in a
plant, and to say that every single one of them is located in exactly
the optimal position and is at any time absolutely functional is asking
a bit much.
MR. MARSH: Yeah, see, there is a lot of talking that needs
to go along with this.
MR. MADDEN: That's right.
MR. MARSH: Because we didn't inspect in that vein. We
didn't go look for a 100 percent.
MR. MADDEN: So, like if you said detection, you know, my
response would be, well, what is the significance of that? Well, the
detector may not react for 10 minutes, okay, when, you know, it should
probably react in a minute. But it will react sooner or later, I mean
it will react. I mean it is just that you don't know -- does it meet
the criteria of prompt detection? Is 10 minutes prompt? You know, it
depends on who defines it.
MR. MARSH: But there's another dimension.
DR. POWERS: Those were even subtle issues.
MR. MADDEN: Right.
MR. MARSH: There's another dimension that goes -- must
accompany this chart. That's, you know, a vertical dimension. How
significant are these five out of the six? It needs discussion, and I
worry sometimes about these things being misused.
DR. POWERS: Well, I think this discussion of significance
is a point raise by some of the findees.
MR. MARSH: Sure.
DR. POWERS: And at some point we are going to have to put
that third dimension on this.
MR. MARSH: Absolutely. In a lot of ways. You need to --
in terms of a recommendation to the Commission about whether this
program, what did it find and is it safety worth -- what is its safety
worth, you have to do it in that context. You have to do in the
enforcement context. You evaluate these findings and what to do about
them in enforcement space. There has got to be a significance element
associated with that. So there is that synthesis going on.
DR. POWERS: We have a comment.
MR. WARREN: Yeah. Glen Warren of the BWR Owners Group.
And Pat, just for clarification of me, too, and understanding the chart
that you have here, when you are talking frequency, and in the example
of hose stations, a plant could have hundreds of hose stations in the
plant, and a finding of one hose station in that plant would make this
show up as one of the six, that this is not four of 6, like at six hose
stations, we looked at four of them and had a problem?
MR. MADDEN: No, no. No, no, no. It's plants, we are
talking plants. Frequency of plants.
MR. WARREN: Thank you.
MR. MADDEN: For example, one plant had an exemption which
required, as part of the basis of the exemption, that it would have hose
stations available to suppress a fire in the control room. Well, when
we went out and measured the actual -- or the distance, and they
committed to meet the NFPA standard, when we went out and started taking
a look at -- I shouldn't say we, the licensee started to take a look at
it because they knew the FPFI was coming along, they determined that
those hose stations would not have reached, or the hose wasn't long
enough to reach the control room, or weren't strategically placed to
accomplish what they stated in their exemption request and stated in
their licensing basis. So that's the type of examples that, you know,
we were finding.
Now, why? I don't know. That should have been something
that should have been evaluated back in the early '80s.
MR. MARSH: See, now, you can -- you misled by a comment
like, gosh, there's 300 hose stations, they only found one problem, so,
therefore, it adds up to be four rather than three. That finding has
some significance, whereas, a finding someplace out in the turbine
building where there's many hose stations may not.
MR. MADDEN: Right.
MR. MARSH: So there is a synthesis that is going on about
these numbers that is important.
DR. POWERS: Yeah. If the hose won't reach the place it is
going to and there is no other hose to do that, you have got a problem.
MR. MARSH: Right.
DR. POWERS: The fact that the hose has a prescribed length,
but that --
MR. MARSH: It didn't quite make it.
DR. POWERS: Make it. But it is still adequate to meet its
function, then --
MR. MARSH: Then it is nickel, dime.
MR. MADDEN: That is a different story. But, you know, some
of the arguments that you will hear on the other side -- Well, that
would have been no big deal. I mean if we had an actual fire, we would
have just sent somebody to go get another piece of hose.
DR. POWERS: Well, I think I would be less susceptible to
that kind of an argument.
MR. MADDEN: Yeah. I mean I was. I wasn't very sympathetic
to that argument, you know, but that's --
DR. POWERS: Yes. It is very easy to claim that somebody
will take a remedial action when you do not recognize that when one of
these events occurs, there are a few distractions.
MR. MADDEN: Yes. Well, the other aspect is, you know, I am
a firm believer that the design of the plant ought to take care of
itself and you shouldn't have to rely on all these remedial actions to
place the plant in a safe condition.
Meaning of suppression. We looked at the fire brigade, and
I am not going to say we looked at them very hard, I mean I don't think
we were that critical on them, but we had findings with respect to
equipment and the adequacy of equipment for the hazards that are on the
plant, you know, on the plant site at three out of the four plants.
Drill performance. We don't -- we didn't feel that the
utilities were, you know, putting enough emphasis on what I want to call
realism of the drill scenarios and expectations of what they wanted to
accomplish through these drills. For example, we had people using --
during the drill, they didn't use their breathing apparatus, okay. We
had other examples where they didn't put on their turn-out pants and
boots. We had other examples where they pulled the hose off the hose
rack and they left it all in a heap and they didn't unfeather it or,
otherwise, if they have been charged, it would have been kinked and the
nozzlemen wouldn't have been able to use it right.
We had another technique that was -- we had a situation
where the hose flat-out didn't reach the seat of the fire whatsoever. I
mean that was in the drill. We had another situation where they used
cellular phones for communications, and the cellular phones went belly
up. I mean it just -- you know, there was all kinds of things that went
on and, you know.
You know, the expectations and being critical of, you know,
a defense -- what I want to call layer of defense-in-depth was not -- we
didn't see it, the criticality.
Qualifications of fire brigade members. We think
fire-fighting is pretty stressful, and we found at one plant -- our
guidance documents require medical examinations. We saw at one plant
where they decided that annual medical examinations were too frequent
and they decided to extend that.
Pre-plans and strategies. Three out of four plants we had
some questions about their pre-plans and strategies. A lot of these
plants are older plants and they don't have smoke removal capability
and, you know, any type of lubricating or hydrocarbon fire, oil, puts
off a significant amount of smoke -- or any type of cable fire, the
cable jackets put off a significant amount of smoke. And we really had
some questions about how do you manage smoke in a plant, and they all
pointed to the pre-plans. And when you went to the pre-plans, there
would be just a one line statement, you know, vent smoke as needed, or,
you know, use your portable fans.
So our questions were, portable fans, you know, what is your
smoke management plan? Do you put the smoke in an area where the
redundant train isn't? Do you put the smoke in an areas where sensitive
electrical equipment that you may be relying on isn't? How do you
protect your operators from carrying out the manual actions that they
may need -- that may be associated with that fire in that fire area?
So we asked all, you know, all those questions. And so we
looked at, you know, the in-depth look at -- you know, if you needed a
fire brigade and it was a serious fire, how you would cope with some of
the what I want to call the real situations that they may be faced with.
MR. MARSH: I think this was a surprise, right, Pat? The
breadth and depth of the findings in the fire brigade?
MR. MADDEN: Yeah. For a long period of time, and even in
the IPEEEs, there's, you know, always a reliance on we have a plant fire
brigade. They will always say -- you know, and they are well qualified
and well trained. And it was kind of a surprise. And maybe our
expectations are being held a little higher than theirs, but we think
that if they are one of those barriers of defense, they ought to be
pretty good.
DR. POWERS: I think I agree with you that we have always
presumed that fire brigades were supermen, quite frankly, supermen.
MR. MARSH: Especially at a nuclear plant. We thought,
gosh, they have got to have good training, good equipment.
DR. POWERS: And I don't know that we have, or would ever
have anything that said, okay, if your fire brigade is deficient,
medical examinations, training, whatnot, equipment, that how does it
function in the bounding fires that we know, that we anticipate? It
seems like a very difficult area.
MR. MARSH: We haven't focused on it either.
MR. MADDEN: Yeah, I mean if you were to, quote-unquote, use
the term realistic or credible fire that they all been called out on, to
function, you know, and I can only think of a couple of cases where I
would say that the fire brigade has been absolutely unsuccessful, and
those, you know. But, generally speaking, you know, most fires are
handled with a fire extinguisher.
DR. POWERS: Yes.
MR. MADDEN: Granted. Okay. And I guess if this was a true
performance-based world, that is all we would be training on.
MR. MARSH: Dr. Powers, I have a question for you, sir.
DR. POWERS: I didn't have sex with the woman, if that's
what --
MR. MARSH: I didn't ask that question.
DR. KRESS: The question is why didn't you?
MR. MARSH: We have a group of slides -- I don't know where
to go with that.
DR. POWERS: Don't go anywhere with that.
MR. MARSH: State of the Union.
DR. POWERS: The President suggests it is a lot more trouble
than it is worth.
MR. MARSH: We have a group of slides following this one
which give you a frequency connotation, and the discussion that Pat is
going through is important because it gives you the breadth and depth of
the inspection that went behind these frequencies. It gives you some
sense of what was found, not just a number, but a sense of what was
found. Okay.
Would you like us to continue this way, to go through these
and give you the discussion or --
DR. POWERS: I think it is fair to say that the members have
these slides already.
MR. MARSH: Okay.
DR. POWERS: And have probably looked at them, and they
probably have just exactly the same -- all the same questions that we
have been dancing around.
MR. MARSH: Okay.
DR. POWERS: And so --
MR. MARSH: And so where do you go? Okay.
MR. MADDEN: Well, I would to jump to 17.
DR. POWERS: Right.
MR. MADDEN: Okay. The first bullet is that we have had a
real fragmented approach as far as looking at fire protection as a
whole. In the implementation of the fire protection plan, every plant
has a plan. And what we found during the FPFI, that Appendix R is
generally not considered a part of that plan, and that plan generally
focuses on administrative controls, fire brigade, surveillance of
testing -- surveillance and testing of detection and suppression. But
when you get into actual engineering evaluation or the analysis
associated with safe shutdown, that is always outside or very rarely
integrated into the plant.
So, the integration, we couldn't really draw a conclusion
that the site plan actually ever integrated or validated the SSA. The
safe shutdown has always been viewed as engineering and not an integral
part of operations. So that was one insight we got out of the program.
The other insight was what I call internal self-assessment.
And, true, the NRC has put out guidance on QA audits for fire
protection. And some of that guidance is relatively old. And that
guidance never really asks the licensee to assess or evaluate its
compliance or long-term compliance with Appendix R. So out of the
audits that we looked at, there wasn't an application of QA or a
principle focus on QA audits on compliance with Appendix R, the analysis
of Appendix R, and the implementation of Appendix R.
Pre-FPFI inspection procedures, they never verified -- and
this goes hand in hand with the one above, that safe shutdown was
actually being thoroughly audited by any implementation, was thoroughly
audited by licensee's programs and, therefore, some of the issues that
were identified were only identified as a result of reanalysis done by
the licensee, not based on audit. And some plants started to go that
way as they got closer to being chosen for an FPFI.
Another thing that we noticed, that the administrative
controls or the fire prevention aspects are overly emphasized by core
inspections. Basically, the core inspections done by the NRC are
primarily focused on the administrative or the operations side of the
plant and not so much on engineering and design, and maintenance of
engineering and design.
The next bullet is -- basically what I said is that the
design basis of suppression systems, detection systems, components used
to protect your safe shutdown features was never really evaluated,
reviewed or inspected in-depth with respect to its design basis prior to
the FPFI program.
MR. MARSH: These are weaknesses that Pat and the teams
discovered through doing the FPFIs, and in looking retrospectively at
the core and other past inspection efforts, okay, because that was part
of the charter. Part of the charter was, hey, how -- what is our
inspection core? What are these procedures, are they any good? Okay.
So that is part of what we are finding here, that there were weaknesses
in past inspection efforts.
DR. POWERS: And I think this is really the key insight
here. And it raises the key question, okay, now what?
MR. MARSH: Right.
DR. POWERS: And this is where I think the risk insights
come in. One can now conceive of going through and taking these things
and doing a top-down -- what should I be inspecting? What really is
important to do?
MR. MARSH: Right.
DR. POWERS: Are the administrative controls so important to
my fire prevention activities, one leg of the defense-in-depth that they
have to receive some emphasis, or should there be a more uniform
emphasis to achieve the risk reduction that I am really looking for
here? I think that is what Professor Apostolakis would argue for in
bringing insight to this.
MR. MARSH: Right.
DR. POWERS: This slide is the stuff that I really got
excited about when I saw it in a previous presentation. It continues
on, though, you have got lots of insights.
MR. MARSH: It does.
MR. MADDEN: Yeah, it does continue, but they are all kind
of interconnected. I mean they really are interconnected to some
degree.
DR. POWERS: But they all ask the same question -- what
should the inspection program really be? And I think Professor
Apostolakis would argue passionately for a top-down risk-informed
adjudication on that.
DR. APOSTOLAKIS: I don't know about passionately.
DR. POWERS: I have never seen you dispassionate, George.
MR. MADDEN: Some of the other aspects were actually
weaknesses in the program with respect to procedures and how to inspect
some of these deterministic analyses. And that could hold true if we
were to look at risk-informed analyses, you know, we would still have to
have the same procedures or aspects incorporated in the inspection
procedures.
One thing that we did notice, that the core procedure does
not -- and I have said this before, does not look at any of the Appendix
R program elements. So the NRC really does not visit Appendix R in any
type of depth, other than specific issues coming up here and there.
Another thing that we had problems wrestling with is 5059s,
adequacy of 5059s and what is considered a USQ, et cetera, et cetera, in
fire protection space.
Post-fire safe shutdown operability. How do plants control
operability? Could an element in the system that is required for
post-fire safe shutdown only go out of service and, you know, is a
compensatory measure adequate for that?
Configuration management and reportability. When is a plant
outside its design basis with respect to fire protection? Those are
issues that were not fully addressed but were looked at and inspected
under the FPFI.
Now, to say do we have any answers, we have some views, but
I don't think we have any answers.
Another thing that we did find, and we did, even in the FPFI
program, to some degree, we took a stab at this, but we believe that
additional inspection guidance is needed to fully address the adequacy
of risk analyses. And by that, I mean the IPEEEs. I mean one plant has
a pilot fire -- which may be no better than this piece of paper, and
another one does a bounding fire modeling technique, which, you know,
looks at creating a challenge or finding out when the challenge would
occur to their plant.
DR. POWERS: You said that you needed examine the IPEEEs.
In the IPE space, we really don't do that. The IPEs exist, there is an
underlying PRA for most of them. People have reviewed the IPEs against
a set of criteria that are pertinent to the original request, but not
pertinent to the question of, can I use the IPE for analysis of a
regulatory question? They just haven't looked at them for that purpose.
MR. MADDEN: Well, the thing that I guess I am driving at
with this statement is that when we go to looking at risk with respect
to these findings, let's say all these were compliance issues, and I
know how to deal with those compliance issues, and we are going to get
to this in the next slide, but when I go to deal with risk, if I only
have a snapshot in time -- first of all, I have got to have some kind of
IPEEE that I have confidence in, to start evaluating each individual
finding or findings grouped together -- I need to know what impact that
has on that IPEEE. And right now, you know, with, you know, divergence
and assumptions methodology, and what our perception is about the
results, I mean the only thing that we really use them for is relative
ranking of sensitivity of rooms, and understanding what that relative
ranking is.
Now, to judge the findings, you know, is it only those rooms
that we do enforcement against? Or is it -- you know, are there other
elements outside those boxes that we do enforcement against?
DR. POWERS: I guess my point is that I would not expect the
IPEEE to be a very reliable source for risk information, ipso facto.
DR. MILLER: Right.
DR. POWERS: That -- it was done under a set of ground rules
that exist, they are fixed. They probably didn't recognize this
potential use of it, and so don't ask it to do that. Now, you still
have this problem. How do I risk rank my findings? Because people tell
you you have got to have some risk ranking here -- how do we go about
doing it? That really is the problem.
MR. MARSH: And it is something we are struggling with
today.
MR. MADDEN: Right.
MR. MARSH: It's very true, very relevant.
MR. MADDEN: Some of the point is that when we ask the
utilities to do it, they revert back to the IPEEE.
DR. POWERS: And that -- if they did, then I think it is
fair game to go look at their IPEEE at that point. But I don't think it
is fair to go until they say, yeah, verily, you can use this.
MR. MARSH: We clearly need some method to come up with some
judgments, qualitative, quantitative, some tool to assess the risk
significance of some of these findings, some type of rack up, some type
of ranking, and we are struggling with that.
Now, as part of the assessment process, there is efforts to
build a new baseline inspection program, baseline inspection across the
regulatory board, in all programmatic areas, looking at performance
indicators and things of that sort. There is a group of individuals
that are involved in that that are trying to do the same thing, assess
the risk significance of inspection findings, and there are some tools
that are being developed, not mature yet, but there's thinking along
this way, because that is an element which must exist in this inspection
program.
We have not asked ourselves this question before, that I
know of in other -- any inspection effort. What is the rack up of
findings in risk space of relative -- these relative findings? We
assess like a point estimate of a finding, which is a dangerous thing to
do, to just come up with a point.
DR. POWERS: Well, it is an unfair thing to do.
MR. MARSH: It is an unfair thing to do.
DR. POWERS: Because you have deliberately set out to have a
fire protection structure that is tolerant of failures.
MR. MARSH: Right.
DR. POWERS: And I think you have laid down an interesting
challenge to the risk community here with your question, and I realize
that you are in the hot spot, you have got to have an answer
immediately. I would really enjoy hearing what you think you need to
have, not maybe perhaps this year, but let's say what you would like
your son, when he takes your job, to have for doing that question.
MR. MARSH: Well, he is in computer information systems, so
he is --
DR. POWERS: Much smarter than you, I understand. But I
think that's the question we have got to ask. As we go to look for the
right tools, we also have to address your immediate question. And I --
you know, I am sitting here saying, gee, how do they -- how do you do
that? How do you get the risk analysis on this from the staff's point
of view when you don't have something like a NUREG-1150 for a
representative cross section of plants dealing with fire protection
issues? I mean you have one or two that came out of NUREG-1150, but it
is not this robust, --
MR. MARSH: No.
DR. POWERS: -- analyzed device and subdevice that gives you
some insights on the risk significance of these items.
MR. MARSH: Yes.
DR. POWERS: You have this IPEEE per plant that is
presumably a good shot at things, but an unreviewed good shot.
MR. MARSH: Right.
DR. POWERS: And you have got a tough one here.
MR. MARSH: And there is danger because we are making
decisions about futures of programs which may hinge on this, this part
of the assessment, and all the frustrations that you are recognizing, we
are there as well.
MR. MADDEN: One other aspect of the IPEEE, you know, its
foundation was compliance with the regulation.
Anyway, to go on to trying to assess risk and safety
significance of the findings, this is where I told you I was going to
ask you questions. I am turning the table now.
DR. POWERS: Ah, but, see, we don't have to answer.
MR. MADDEN: I know, I know. I know, that's the bad part
about it.
Purely from the inspection findings now, you know, I am not
going to say that the findings identified gross values of any one of
these elements, but it did indicate that there are program concerns that
could, you know, be involved with one element or multiple elements of
defense-in-depth, and that is an indication which was, to some degree,
somewhat surprising.
The safety significance of individual findings is kind of
dependent on the overall program effectiveness and should be judged in a
plant-wide context. And I guess what I mean by that is that, you know,
if they had a single finding and one area or one element of
defense-in-depth showed an inherent weakness, but the other elements
were very strong and robust, then, of course, the safety significance of
that finding would probably be relatively insignificant.
But when you start to look at I have a few findings here in
this area, a few findings in this area, and few findings in this area,
now, in a plant-wide context, what does that mean? And it starts to get
a little grayer.
DR. POWERS: What we know is that risk assessment portrays
itself as a particularly suitable vehicle for making judgments of
exactly that kind.
MR. MADDEN: That's correct.
DR. POWERS: That's what it portrays itself. Now, the
question is, is it now adequate to do that kind of fine judgment? We
frequently find in operational PRAs that we ask regulatory questions
that are too microscopic to be explicitly addressed by the PRA. Now,
does that mean those questions are not risk-significant? I think not.
But -- and so how do we interface them? I mean that we had that
particular problem in 5059, that the questions being asked are below the
resolution of the PRA and it is -- how do you make a risk-informed
judgment on a question that the PRA can't see?
MR. MADDEN: That, you know, that may fall in suit here
also. I mean --
DR. POWERS: There's a real potential here.
MR. MADDEN: Yeah, there is a potential here. And it
depends on, you know, where that threshold bar is, you want to place. I
mean if you want to place it fairly hard -- or high, I mean you can
probably filter a significant number of findings.
DR. POWERS: One of the problems I could foresee in using
PRA in this area has to do with the objective function. You have -- a
PRA will give you information on CDF and a PRA will give you information
on LERF. That is only half of the objectives of your fire protection
program.
DR. APOSTOLAKIS: What is the other half again?
DR. POWERS: The other half is going to be life safety for
occupants, whatnot, and plant damage control.
DR. APOSTOLAKIS: You can do a PRA for those.
DR. POWERS: I think you can, but you are going to have to
-- I think you can do a PRA for any objective function you would care to
define, but the objective function is going to be different than the
ones that get used for -- the operational guys have a different set.
They only have to go against CDF and LERF. You have got an extra one.
If you choose to preserve that extra one, and I think right now there is
no evidence we shouldn't, then you have got to add it in.
MR. MADDEN: My reaction to this is that I could look at
these findings and I could probably judge these findings on the merit if
I was just looking at what I want to call a fire protection, risk
assessment mitigation model. You know, if I could just -- you know, if
I could evaluate against the change of my success, you know, just a
delta in the success of -- with this element out of the picture, what is
the delta in my success in the mitigation model? That, I could probably
understand, you know. And that, I could probably make relatively fast
judgments about the potential safety significance of these findings.
DR. KRESS: The importance function.
MR. MADDEN: Right, exactly.
DR. KRESS: I think there is a lot to be said about that.
DR. POWERS: I think it is fair to say that when you read
the reports on Prairie Island and St. Lucie and things like that, that
an experienced fire protection engineer can go through and say this one
is important, this is one is -- this one is a gotcha, but I don't care,
and things like that. Literally, that fast. I think that is fair to
say.
MR. MADDEN: Yeah, it is fair to say. But I think what you
need to do is also step back. Like if you looked at the St. Lucie
report, I mean there was such a divergence in all the findings, to come
out to the conclusion that none of it, from a programmatic standpoint,
was safety significant, I don't know how we get there from there. I
mean if you look at the whole big picture. And that is why I have that
second kind of sub-bullet there, that you have got to look at it from
the big picture, also, you just can't focus on the individual findings.
DR. APOSTOLAKIS: The fire community out there has done some
work that uses PRA kind of methods to find the egress time. Is there
another area where the word egress is used or is it just in the fire
protection? I mean is that used anywhere else?
DR. KRESS: Lots of places.
DR. APOSTOLAKIS: I have never seen it, only in the fire.
It is like the urn, it is in probability --
MR. MADDEN: If your ship is sinking, I think you would want
to know the egress to the lifeboats.
DR. APOSTOLAKIS: Yeah, that's right. Where they estimate
the time that it would take for people to get out, and that's meaning
PRA really, although not as sophisticated in the other PRAs, in other
words, the tools for these kinds of things, for these analyses, exist to
some extent, not -- they are not really advanced. But now, you know,
they are talking about -- I am sure you know that -- performance-based
fire regulations in Australia and in other places, where this is the
main theme, actually, estimating the time it takes people to get out,
and then the time, you know, to kill them, and whichever is greater. So
we will not be starting from scratch. I mean this is just a comment. I
mean if we want to address the other half.
DR. POWERS: Well, I think we have a confidence that focus
on any of these tasks could lead to resolution. The question is -- what
focus, when and how much?
MR. MADDEN: See, I think if you -- you know, this -- if you
take it all the way to core damage, you know, in an analyst's space, it
takes away a lot of flexibility that an inspector has in the field to
make rapid judgments. And that's why I said if you -- and George is
touching on this -- is that if you say in the mitigation aspect, or
trying to figure out just what my probability or likelihood of success
is with all these different elements out of this little model out of the
picture, that can give, you know, relatively quick, you know, how these
items interact and what their importance are, too, if you decide to go
that next step. You may not.
I mean you may say that that is good enough for fire
protection. You know, that we have got to protect the plant personnel,
we have got to protect life safety. We have got to, you know, preserve
the shutdown ability and the operator's ability to move through the
plant. We have got to have the ability of the fire brigade to take
prompt action. We have got to have smoke control measures. You know,
if that was all integrated into a mitigation model, it would give you
probably a better picture of how success we are really going to be.
DR. POWERS: Okay. You are saying a little different kind
of analysis than classic CDF analysis.
MR. MADDEN: Right. Exactly. It is a little bit more
simplistic. I mean --
DR. POWERS: I think it is -- I think it would end up being
just every bit as sophisticated.
MR. MADDEN: Well, maybe you are right.
DR. POWERS: But I think it is an intriguing idea because it
avoids you going to the step that I think just leads to defocusing.
MR. MADDEN: Right. Exactly.
DR. POWERS: And that might be very advantageous to do.
DR. KRESS: The problem I have with it is the same problem I
have with performance-based regulations in general, is how do you know
when you have done enough? What acceptance criteria are you going to
use to say this is good enough here? And that is always a problem with
performance-based.
DR. POWERS: Could you imagine doing it, Tom, where you make
occasion tie points to the CDF analysis?
DR. KRESS: Or to other measures.
DR. POWERS: Other -- something else.
DR. KRESS: I don't like -- you know, I don't like sticking
CDF. But if you had occasional tie points, enough that it made it
relatively coherent, I think --
MR. MARSH: So that you could interpolate.
DR. KRESS: -- the more you go, the better off you are.
But, you know, I don't even think it has to be -- I think performance
type of regulations can stand on their own as long as you put together a
set of acceptance criteria for them. Now, they don't have to be PRA
based, or they don't have to be CDF based. They could be like any other
acceptance criteria, it is what people are willing to accept.
DR. APOSTOLAKIS: But they would be rational.
DR. KRESS: They would be rational. And they would be
arrived at by a process that incorporates things like decision criteria,
loss functions and stuff like that. But it is how you arrive at these
acceptance criteria that has always bothered me in the performance.
DR. APOSTOLAKIS: I want to ask a question. If you had gone
-- if one of your plants in this program had been Quad Cities, --
MR. MADDEN: It was.
DR. APOSTOLAKIS: Did you find that thing yourself?
MR. MADDEN: Yeah.
DR. APOSTOLAKIS: That the IPEEE found? Did you know about
it before?
MR. MADDEN: Well, what the IPEEE found, I think we found a
little bit beyond the IPEEE. Let me kind of explain.
DR. APOSTOLAKIS: When did you go, though? When did you do
this, before or after the IPEEE?
MR. MADDEN: After the IPEEE. Yes. But I mean we found
things that were not even considered by the IPEEE.
DR. POWERS: I understand that the IPEEE finding is
relatively narrow considering --
MR. MADDEN: Were they important? Yes.
DR. POWERS: In comparison to the raft of things that have
been found at Quad Cities now.
MR. MADDEN: Were they important? Yes. We found conditions
where a diesel generator could have been loaded on faulted buses. I
mean that wasn't picked up by the IPEEE. We found conditions where
smoke management was never even considered and operators would have had
to go through rooms full of smoke. That wasn't picked up by the IPEEE.
I mean there was all kinds of conditions that we picked up.
I mean I am sure you read that report and that's pretty --
okay.
DR. APOSTOLAKIS: So where do we go from here?
MR. MADDEN: Okay. I guess what I am talking here is what I
think -- if I -- you said something about if I wanted a tool, what kind
of tool would I need? And it is basically the methodology that I kind
of regurgitated to you, but I mean it would have to be a tool that would
not only, number one, assess compliance, but look at the risk in the
changes, basically, on what I want to say in the safety margins of the
individual findings, and that is what I mean by -- you know, our rule is
written around mitigation of consequence, you know, starting out with,
you know, the event is one. And I had a tool that could just tell me
the change in the mitigation of success, that would give me a good
indicator of what worth that individual item or individual finding had.
DR. APOSTOLAKIS: So you are talking about compliance now
with the existing regulations?
MR. MADDEN: Absolutely. Right. I mean I have to, I mean I
can't talk about future because I don't know where it is going to be.
DR. APOSTOLAKIS: So why do you care about risk?
MR. MADDEN: Why do I care about risk? Because it would
come into play on how strong of a viewpoint I take with respect to
getting it fixed, enforcing it, and what kind of generic implications it
may have. I mean if I am going to push -- if it is low risk, I am not
certainly not going to push the licensee to immediately take action, to
shut his plant down to do this, to do that, you know, so I am not going
to do that. I mean if it is dealing with improper fire-fighting gear or
whatever and it is low risk, sure, it is a finding, it may be a
non-cited violation, it may be he agrees to fix the problem and I go
away happy as can be.
But if it is high risk, I may have to get other individuals
involved such as management, et cetera, et cetera, and we may have to
start another dialogue with that utility and ask them why they think
they are justified to continue to operate.
So I think risk has to play a little bit of a picture in
here. But I think the real thing is that we really need to understand
what the change in the safety margin is over the perceived regulation to
where they are at that given point.
And then I think another thing that we need to do is have a
methodology for assessing what I want to call the big picture, the
programmatic aspects of what we perceive as a weak program, and that
means across the board, and what impact that may have on the plant as an
overall.
DR. POWERS: I guess I don't see the distinction between the
two methodologies.
MR. MADDEN: Well, there maybe isn't. And my mindset
up-front was looking at individuals, but if you had a robust enough
tool, you could look at the whole big picture with one tool.
DR. POWERS: What you want to do is look and say I have got
a bunch of individual findings, they are not in each -- they are not in
just one leg of the defense-in-depth.
MR. MADDEN: Right.
DR. POWERS: I have findings in every one of the legs of
defense-in-depth.
DR. MADDEN: Right.
DR. POWERS: I am confident that if I have all three legs of
defense-in-depth operational, I am in good shape. If I have two out of
three, I think I am still in good shape. If I have one out of three, I
am maybe a little less confident. But the question is, when I have
findings in all three --
MR. MADDEN: Weaknesses in all three.
DR. POWERS: Weakness. Do I ever get the combination of
events that just follows that weak link all the way through?
MR. MADDEN: That's correct.
DR. POWERS: That's what you want to have.
MR. MADDEN: That's correct.
DR. POWERS: And that seems to me tells you about the entire
program.
MR. MADDEN: That's correct.
DR. POWERS: And I think you have certainly raised an
interesting approach to that that looks a lot more tractable to me than
the conventional PRA type of approach.
DR. APOSTOLAKIS: Well, actually, what I find very
interesting is that we are willing to make the inspection process
risk-informed, but not the regulations.
MR. MADDEN: Well, let me do this first, George, and then we
will work on the regulations.
DR. APOSTOLAKIS: I'm sorry. Go ahead.
MR. MADDEN: I said let me do this first and maybe we will
work backwards and do the regulations.
DR. APOSTOLAKIS: This is kind of selfish, Pat. You are
trying to save the agency money but not industry.
DR. POWERS: I don't understand, George, I thought with
1.174 --
DR. APOSTOLAKIS: I'm sorry.
DR. POWERS: I thought with 1.174 we announced that they
could make these regulations risk-informed.
DR. APOSTOLAKIS: Well, up until this morning, I was under
the impression we are talking about risk-informed, performance-based
regulation. Now, we are talking about traditional regulation,
risk-informed inspections.
Now, Pat may be right, maybe you want to understand this
first before you go to the first one, which is what -- which is not what
we did with 1.174.
DR. POWERS: To the detriment of 1.174.
DR. APOSTOLAKIS: This is a revolutionary change in my
thinking.
DR. POWERS: Well, understand that I think --
DR. APOSTOLAKIS: No, the change has not happened yet, but
something I had to --
DR. POWERS: Understand that --
MR. MARSH: I am going to -- I have a challenge. My
challenge is for myself, to understand the depth of your concern about
this Reg. Guide.
DR. APOSTOLAKIS: I didn't say it was a concern. Oh, this
guide?
MR. MARSH: The Reg. Guide, yeah.
DR. APOSTOLAKIS: Yeah, I don't like the way it is put
together.
MR. MARSH: Yeah, because I am not understanding. And let's
not do it -- but my -- the Reg. Guide started, and its intent was to
address an existing, real life problem out there now, to put into one
place all this stuff. Okay. I think that is a great thing to do. I
mean I think that is a wonderful, noble and necessary thing to do.
Now, what you do with that, where you go with that is the
next more difficult step. And if this becomes a parallel track to go
with 805, whether this becomes backfit to operating plants, that is
where you may inject risk information into that, that type of the
process, but --
DR. APOSTOLAKIS: But there are elements of the thinking
that has evolved over the years from PRA without necessarily being
strictly PRA that, like this top-down approach and stating your
objectives. Pretty much what the inspection program does, the other
inspection program with the cornerstones. That could have been used
here.
All right. I realize that there is a need for consolidating
the existing regulations, but you can inject a little bit of your
thinking in there without really going all the way to PRA. It is not
all or nothing.
The other thing that worries me -- see, I am not raising one
of my concerns, because I haven't really studied it. I suspect that
when the PRA tells you to do more, you are very happy to embrace it.
MR. MADDEN: It's backfit.
DR. APOSTOLAKIS: But when the PRA says relax on all the
requirements, you are not too willing to do that, not just you.
MR. MARSH: Why do you say that?
DR. APOSTOLAKIS: Why do I say that? Because all the
evidence points that way, and I use Bates theorem every day.
DR. POWERS: I am going to interrupt this discussion.
DR. APOSTOLAKIS: We have a major problem identifying
regulations that were eliminated because of a PRA.
DR. POWERS: I am going to --
DR. APOSTOLAKIS: It is very easy to create long list of
things that were added to the regulations because of the PRA. And then,
of course, you go to the industry and they say, PRA, what's that? I
don't want to know about it.
DR. POWERS: I am going to --
DR. APOSTOLAKIS: I'm sorry. Okay.
DR. POWERS: -- interrupt this discussion, not because I
disagree --
DR. APOSTOLAKIS: Or agree.
DR. POWERS: -- or am fascinated with it, but because I have
to be fair to the subsequent speakers this afternoon. But we have -- we
do have time on our agenda to pursue exactly this issue, and I think the
Committee needs to wrestle with how it is going to respond to your
Regulatory Guide. I mean we responded at one time and said we didn't
think it was a useful expenditure of resources.
MR. MADDEN: Right.
DR. POWERS: But, obviously, we were in gross error over
that judgment.
MR. MARSH: The Commission thought it was a really good
thing to do.
DR. POWERS: So now we have had our wheels realigned, I
think we would like to get -- we need to decide how we want to pour oil
in the water, and we need to answer your question -- what is our depth
of feeling? But, on the other hand, I want to be fair to the other
speakers, so I am going to defer it till a little later this afternoon
and whatnot.
MR. MADDEN: Okay. Just the line and final slide here, and
the big question is, where do we go from here? You know, how does the
FPFI program fit into the NRC cornerstone inspection process, or
philosophy? Big picture. Do we rely on audits, audits of licensees'
self-assessment, for example? You know, do we turn the FPFI program
over to the licensees and say, hey, do them? Come up with a plan,
blah-blah-blah, and we will come in and audit them.
Do we do a mixture of both? You know, is it just a FPFI
audit, you know, at a certain set of plants -- you know, maybe one or
two a year, maybe three a year -- and, also, do audits of licensees'
self-assessments?
One thing I think we do have to do, and this is not really a
question, but I think we have to revisit what is going to be in the
complementary inspection program with respect to the cornerstone effort?
I mean there are -- I have seen elements in there, but I don't believe
they are the right elements.
And then how do we monitor licensees' fire protection
programs with respect to the effectiveness of their programs in the
cornerstone program? By that, I mean what do we gauge it against? What
is the benchmark? Who monitors it? How do we get the information and
how do we know that the information is reliable with respect to
monitoring?
So, with that, I close. I thank you for your attention.
DR. APOSTOLAKIS: By the way, I can see the need for
consolidating certain things. I don't necessarily have to like it. I
mean there are certain things that need to be done that could have been
done a different way, and I am not --
DR. MILLER: Isn't the real question the drafting? In the
process of consolidating, you are recommending they build in risk
insights.
MR. MARSH: That's right. That's what I hear.
DR. MILLER: The danger of doing that is you fail in doing
both very effectively. The other way is consolidating, then building
the risk insights later. That's --
DR. APOSTOLAKIS: See, I wouldn't mind that, if it weren't
1999, and the reactor safety study was published in 1975. To hear that
Nathan will give me research results five years from now to start doing
-- using PRA, depresses me, because that will be 2004. Okay. That will
be 30 years after the draft WASH-1400 was published. Thirty full years.
DR. POWERS: The proof that the PRA analysts had been very
effective -- ineffective in communicating their message.
I am going to recess until 1:30. If any of the subsequent
speakers are going to have difficulty, please bring it to my attention.
[Whereupon, at 12:50 p.m., the meeting was recessed, to
reconvene at 1:35 p.m., this same day.]. A F T E R N O O N S E S S I O N
[1:35 p.m.]
DR. POWERS: Let's come to order. I have already come to
order. Fred, you are going to discuss with us two topics, I guess. You
are going to talk on my favorite, hot shorts -- boy, that is good for
after lunch -- and the FPFI. So, I guess, the floor is yours.
MR. EMERSON: Actually, four topics.
DR. POWERS: Four. Oh, sorry, Fred.
MR. EMERSON: It remains to be seen whether I get through
them all in 60 minutes, but I will give it a shot. The four topics that
I wanted to cover today, first of all, I will try to address some of the
issues that the staff addressed this morning and then some that they
haven't addressed yet. I would like to start with discussing the fire
protection inspections and assessments, which is what Pat wound up with
this morning, then standards and guidance development, including NFPA
805, fire-induced circuit failures and a new area for us, risk-informed
fire protection applications. I am going to move fairly fast because of
the quantity of information.
First, in the area of fire protection inspections and
assessments. I am not sure to what extent your Committee has been
exposed to SECY-99-007. I am not going to try to go through it in any
detail, but I would like to provide some background information for what
we are proposing as a follow-on to the FPFI program.
In the overall proposals in the SECY that is currently out,
there is both industry roles and NRC roles, industry and assessment, and
NRC and baseline inspections and other activities. The new assessment
process, as I said, is embodied right now in the SECY-99-007, and it
integrates licensee audits and self-assessments into regulatory
oversight. Just as quick sketch of the process, I am not going to try
to again go through this except to indicate that it shows both the NRC
and the licensees -- I'm sorry -- the NRC and the licensees as having
distinct, but complementary roles in this new assessment process.
I have labeled this an NEI approach. Actually, it is not,
it is what is currently embodied in the SECY document. It involves
several performance levels, which I will put up in just a minute. It
involves thresholds for increased NRC attention, a threshold for overt
regulatory action to assure public health and safety, and these
thresholds are intended to be a blend of regulatory requirements and
risk insights.
There are -- if you look on the subsequent pages, these next
two slides are shown in full sizes, so you don't have to get our your
magnifying glass to read the type. As I say, there are four performance
bands, and I am just going to cover this again very quickly as a
backdrop to the recommendations which we have very recently made to the
staff.
The first area is a Licensee Response Band, which will
involve licensees monitoring performance indicators where they exist,
and NRC engaging in a core baseline inspection program where the
performance indicators do not exist. Below the top band is called the
Increased Regulatory Response Band, and that is an area where the NRC,
when performance begins to drop off, either as measured by performance
indicators or as core inspection results, NRC and licensee become more
active and increased attention is needed for those performance
indicators.
The third area is the Regulatory Response Band. That is
where some significant action is needed. That is a yellow color. And
then the bottom one is Unacceptable Performance where plants won't
normally be allowed to operate. This slide represents the performance
for any specific performance indicator or core inspection module. There
are those at NEI who are much more conversant with this than I am, so I
am not going to try to address the process in more detail than this.
DR. KRESS: There would be a number of these for each
performance indicator. Supposedly, each performance indicator has some
way to quantity so that you can have these --
MR. EMERSON: Yes. As I understand it, performance
indicators either have been or are being developed, and in areas where
performance indicators do not exist, there is a core baseline inspection
module that the NRC would use to evaluate performance.
DR. APOSTOLAKIS: So what is the difference again between
increased regulatory response and required regulatory response?
MR. EMERSON: Increased regulatory response means the
performance indicators or the core inspection results are dropping off a
little bit. It is time to take a closer look at that indicator and the
reasons why it is not going the way it should. As you drop lower in
performance, either you have more indicators going south, or you have
one going south repeatedly and still more increased attention is devoted
to it.
If I can put up the next slide, it might make things a
little clearer.
DR. KRESS: Let me ask you a question on this slide.
MR. EMERSON: Sure.
DR. KRESS: The slide implies to me that the measurement of
these performance indicators is continuous, because there is a
continuous line. Would they be, in fact, though, continuous, or would
they be something that would the result of an inspection or
self-assessment that would be at specific points in time, specific
periods?
MR. EMERSON: I think to the extent that the licensees
monitor their own programs, there would be a continuous level by the
licensees of monitoring performance against these indicators. I don't
know that they would do it on an everyday basis, but they would do it
frequently enough to keep in tune with --
DR. KRESS: Continuous doesn't have to mean that.
MR. EMERSON: Right. I am not familiar enough with the core
baseline inspection programs to know how the NRC, how frequent those
would be.
Okay. This is what is intended, and, again, this is out --
almost entirely out of SECY-99-007. This involves actions by the NRC
and licensees with respect to what do you do when either one performance
indicator moves into a certain band or more than one moves into a
certain band, and it reflects increasing levels of attention by both the
licensee and the NRC as the performance indicators decline, or as core
inspection results indicate a declining performance.
Now, the way fire protection is currently wrapped into this
-- and, staff, feel free to correct me if I misstate this -- as I
understand it, right now, it is intended to be a complementary baseline
inspection program because no performance indicators have been worked
out for fire protection. It is intended to address the two cornerstones
of initiating events and mitigating systems. The types of things that
the baseline inspections are intended by -- as written in the SECY
document currently, to address are review of ignition sources, control
of combustibles, fire protection system and equipment, licensee controls
to minimize the probability of a fire, and the availability and
reliability of equipment to mitigate the effects. That is, as I
understand it, how it is stated in the current SECY, current version of
the SECY document.
What I would like to get into next is what we propose to
augment this, if you will, to replace the fire protection functional
inspection program. In the broadest sense, our view of the future of
the FPFI is that all assessment and inspection guidance for fire
protection ought to be consistent with and integrated into the changes
in NRC oversight and assessment that are included in the SECY document.
We believe that performance indicators should be used
wherever possible to monitor and assess fire protection and safe
shutdown programs. And in cases where performance indicators cannot be
developed that are satisfactory, or readily acceptable, then the core
inspection would be -- modules would be developed to carry on core
inspections to be able to identify and trend licensee performance in
those areas.
In general, the licensee would have the principal
responsibility as long as his performance stayed in the highest
performance band, the green band. NRC would continue to conduct
baseline inspection, but no significant amounts of activity would be
needed. And, as we pointed out earlier, there would be increasing
licensee and NRC activity as performance declined into the lower
performance bands, that is in the broadest sense.
What I would like to cover are our recommendations for
short-term activities and long-term activities for fire protection to be
integrated better into this process. I have separate recommendations
for NRC actions, NEI actions and individual licensee actions.
In the short-term, we had understood from conversations with
the staff that they were developing the current FPFI inspection guidance
into discrete inspection modules, and we encourage that process to
continue, with the addition of developing performance indicators
wherever that is possible for each of those inspection modules, and I
will indicate what we would like to see done with those modules shortly.
That is in the short-term, and by short-term, I mean in the next six to
12 months. The letter that we just sent to the NRC, and advised ACRS
and provided copies of, provides a more detailed schedule for these
recommendations.
What we propose as short-term activities for NEI, the first
activity would be to form an issue task force, which is a group of
working level experts in a given area who can pool their knowledge and
expertise for whatever specific task NEI needs to have done. This task
force's responsibilities would be, first, to develop some near-term
inspection guidance, or assessment guidance for licensees.
Now, this short-term activity is intended, while we are in
the process of implementing the long-term activity, to see whether there
are any short-term needs, or any holes that licensees need to fill while
we are getting the long-term process rolling. It is intended to address
an NRC concern that we might not want to wait, if findings in the FPFIs
had proved significant, to -- for all licensees to address it, and to
address the question of consistency and comprehensiveness of licensee
self-assessments. So we would be developing guidance and requesting
licensees to look at their own programs if they had not performed recent
inspection or -- I'm sorry -- assessment activities.
Then we would ask the task force, and this would be one of
the important roles, is to put their minds to drafting risk-informed,
performance indicators to provide input to the staff. This -- if the
staff is developing inspection modules with performance indicators, we
feel that industry expertise should be factored into the staff's
preparation of these modules, and we would devote some significant
efforts to suggesting performance indicators for staff consideration.
DR. POWERS: When you say risk-informed -- sorry.
DR. APOSTOLAKIS: The risk-informed indicators, I think we
are going to have a problem defining those for an event like a fire. I
can see an indicator dealing with perhaps the frequency of minor fires,
but anything beyond that, I don't know what performance means anymore.
I mean I have difficulty seeing the performance indicators that will be
a good set that will cover a lot of things. I mean you necessarily, you
have to rely on programmatic requirements. It is not like trains of
safety systems that are being tested frequently and you can produce an
unavailability figure and so on.
With fires, I mean it is such -- especially the fires of
interest are such rare events that I don't know how many indicators we
can have.
MR. EMERSON: I am not going to try to pre-judge what the
indicators would look like, but I would agree that they would probably
need to address perhaps at least the elements of defense-in-depth. You
might want to measure your ability to prevent fires from starting,
control of combustibles, your ability to detect and suppress, fire
brigade performance perhaps. Those might be areas where performance
indicators -- no single indicator is going to do the job for the whole
program, but, taken collectively, I believe that we can come up with
enough to make the effort worthwhile.
Now, as I said, I can't pre-judge what they will look like,
and, as you pointed out, this probably is not going to be an easy
process, but we definitely think it is worthwhile trying it.
DR. KRESS: It looks particularly difficult to put that on
the four region map you have, because, you know, like in detection, that
is can you or can't you, it is not four regions there.
DR. APOSTOLAKIS: And I don't know to what extent the
detectors are tested. Maybe Pat can enlighten us in a little bit.
MR. EMERSON: Well, there is in some test frequency, but,
recently, there has been trends to extent that testing frequency by the
utilities. I think it was on a once per like every refueling cycle,
they would go and then actually do some testing of the circuitry, maybe
not specifically the detectors.
DR. KRESS: Those, in that sense -- or maybe the frequency
at which it doesn't work.
DR. APOSTOLAKIS: Yeah, but when it comes, for example,
suppression systems.
MR. EMERSON: You don't touch, generally touch those, no.
DR. APOSTOLAKIS: The ability to shut down.
MR. EMERSON: Safe shutdown systems, you mean.
DR. KRESS: Check and see if the water available and there's
valves that are reliable.
DR. APOSTOLAKIS: But that is not performance.
DR. KRESS: Oh, performance is anything you want it to
cover.
DR. APOSTOLAKIS: Oh, I don't know.
DR. KRESS: That's another problem I have.
DR. APOSTOLAKIS: See, that's the problem I have, and, in
fact, I copied the definition. It says the extent to which something
meets its intended purpose. Now, I don't know that if you have water
that is a performance by an objective criterion.
DR. KRESS: It's the extent to which it could meet it, and
then you back off of that when you can --
MR. EMERSON: George, I would share your view that it will
be a challenging exercise. But, given the level of expertise in this
area both within the staff and the industry, I think that if these
people can't come up with useful performance indicators, the inspectors
and the users of them, then I will agree then that it is not a possible
job.
DR. APOSTOLAKIS: And this is in the short-term, so that
makes it even more challenging. But, anyway, I am just expressing a
thought here.
MR. EMERSON: I understand. And those opinions have been
expressed by the staff and by my own peers in the industry, too.
DR. POWERS: When you say risk-informed on your performance
indicators, you are using that in the sense of engineering judgment of
risk?
MR. EMERSON: I don't know that I would even try to
characterize how risk information would be factored into them just yet.
I just -- we just haven't thought about them enough.
The next task would be to recommend steps for integration
fire protection into the overall assessment process. Those are the
questions that Pat asked on his last slide this morning. How are we
going to factor all of this into the new process? And those questions
have yet to answered and we would, hopefully, have the opportunity to
work with the staff and make recommendations on how that could be done.
Our task force would also request plants with no recent
self-assessments, and by recent, I mean pretty much since the inception
of the FPFI program in the last year-and-a-half, to take a look at
themselves critically and see whether an assessment is needed in the
short-term to address holes. The main thing is for them to be aware,
thoroughly aware of the health of their own program.
DR. POWERS: As I understand the status of self-assessments,
and my understanding is imperfect at best, only just barely half the
plants have done a self-assessment on both the fire protection and the
safe shutdown programs.
MR. EMERSON: That's what our survey showed, it was
approximately that. Now, more had done one and more had done the other.
And, certainly, other plants that have not done what I would call formal
self-assessments have had other types of either inspections or
assessments that they might credit at least partially for that.
DR. POWERS: So it is a situation of you would ask people
that had done self-assessment of their fire program, surely, but had not
done the safe shutdown, now to go ahead and do the safe shutdown?
MR. EMERSON: I am not necessarily asking to go out and do a
full blown self-assessment. What I first would be asking them to do is
to take a look and see whether, based on their -- based on what is going
on in the industry, based on the FPFI results, based on their own
knowledge of their own programs, whether it is time to take a look, and
we would provide them some guidelines for how to do that.
The whole idea of this is to get people to be looking at
their own programs in the short-term, not asking them to go out and do
expensive and time-consuming self-assessments unless they see a need for
that. You know, they need to be the masters of their own destinies.
DR. POWERS: Sure. And a guy that has spent a lot of money
keeping his program up to date doesn't need to do a self-assessment.
MR. EMERSON: Right. That is also true. We would ask the
licensees to use the NEI guidelines to determine whether a
self-assessment is needed in the short-term, and, although it is not
reflected in there, if they do determine that one is needed, we would
ask them to complete it probably by the end of the first quarter of the
year 2000. As I understand it, formal self-assessments are not easily
scheduled and accommodated within a plant's cycle, so expecting them to
do within a space of three months might be very unrealistic.
Moving into the area of long-term recommendations, for NRC
and the industry both, they would be basically implementing the
performance indicators and the core inspection programs that were
developed in the short term. We would suggest that the core inspections
use the modules that staff was intending to develop from the FPFI
guidance, and we would suggest to the licensees that they use these
modules for regulatory -- with those addressing regulatory requirements
for their self-assessments when they determine that a formal
self-assessment is needed.
Some of the feedback that I've gotten from licensees, I
won't call this a universal reaction, but a number of them have said
well, you know, the FPFI was a good, thorough approach to reviewing a
program, and it would make sense to use the guidance in a
self-assessment as long as we didn't have to do it in the space of a
month, either as part of an FPFI or as part of a self-assessment
preparing for an FPFI.
Now we would -- the licensees would conduct routine
performance monitoring and/or NRC would conduct baseline inspections in
the green band depending on the existence or lack of performance
indicators. We would propose that licensees conduct what I would call a
focused self-assessment if a single indicator starts going south into
the white band. By focused I mean you focus on the area where the
indicator is going south.
If performance declines on a broader scale, we would propose
a general self-assessment using more than one NRC module for -- with NRC
oversight in the yellow band. And this in general is consistent with
the -- we believe with the guidance in the SECY-99-007. Where FPFI, a
large team inspection such as the FPFIs would play a role is if
performance continued to decline further into the red band, obviously
there are some major breakdowns in the licensee's fire-protection
program, and an FPFI or an inspection similar to the FPFI pilots might
be warranted because something more drastic is needed to turn things
around.
Continuing, NRC and licensees would continue other
inspections and assessments that were not superseded by the new process,
and I can't sit down and give you a full list of what those other
inspections might be, but we're not precluding other appropriate
activities. We would expect appropriate credit for results of those
other inspections or assessments against the self-assessments and core
inspections that the NRC would be doing on an ongoing basis.
I'd like to spend a couple of slides addressing some
questions that the NRC has asked us in previous meetings or concerns
that I've heard expressed privately as to why don't we just maintain the
FPFI program in parallel with this new process. So I'd like to address
four specific concerns that I've heard.
First of all, the first concern would be the FPFI is needed
to adequately evaluate fire protection programs. Pat explained very
well this morning what some of the deficiencies in earlier inspection
programs were, and the reasons why FPFI had beefed up the quality of
inspections. Now we believe that integration of fire protection into
the proposed assessment process improvements will achieve that same goal
with greater cost benefit, especially if we continue to use the FPFI
guidance in discrete inspection modules by both the NRC and licensees.
A second concern was what assurance do we have the licensees
will conduct adequate self-assessments, and I think we heard some
references this morning and some concerns that licensee self-assessments
may not have been adequate. We believe that this proposal provides for
more timely and comprehensive assessments and inspections than
continuing routine functional inspections, because the licensee is
looking at himself with a set of known criteria, and the NRC is using a
set of core inspection criteria that are based on the FPFIs. I really
believe that dual look by the licensee and the NRC would fill the same
need without the need for the full-team inspections unless again as I
pointed out earlier performance really declines in a drastic way.
The third concern has to do with the significance of the
FPFI findings. You know, Pat again talked eloquently this morning about
the measurement of significance for the FPFIs, and pointed out clearly
what the status of the staff efforts are to assess significance. Really
the purpose of this overall program is to focus on safety-significant
issues, and if we construct the fire-protection elements of that
properly, we I think will meet a great many of the needs for focusing
inspections in the right area. I think if we set it up properly and
implement it properly, the goal is certainly to identify
safety-significant areas, and the program should accomplish that goal.
Lastly, it was the question just raised by George about
performance indicators, and as I said before, it'll be challenging, but
we think the expertise is there to maximize chances of success, and we
think that the effort is needed. We at least need to give it a try.
Now that concludes my discussion of the assessment and the
FPFI follow-on program.
DR. POWERS: Fred, one of the original objectives of the
FPFI program, one that probably has been subverted -- or subordinated,
not subverted -- was reviewing the status of Thermo-Lag, and that has a
certain political visibility that the NRC has to respond to. In your
view, does this combination of licensee self-assessments and NRC
auditing or oversight, are these self-assessments going to give them the
ability to respond to political interest in the Thermo-Lag issue?
MR. EMERSON: Well, the political --
DR. POWERS: The slow pace which is being resolved at some
plants.
MR. EMERSON: Well, the political discussion is basically
over one element of defense in depth, and that's the adequacy of fire
barriers. And if we have constructed the inspection modules properly,
we'll be in an objective way continually reviewing on a frequent basis
the viability of the fire barriers you have at your site, whether it's
Thermo-Lag or some other material.
As far as addressing the political aspects, I can't speak to
that, but I think it should be adequate to determine any
safety-significance aspects of the fire-barrier issues.
DR. POWERS: As you probably understand, the question they
have to answer every once in a while, episodically, it seems, is why is
it taking the NRC so long to get the plants to resolve this barrier
Thermo-Lag issue.
MR. EMERSON: Um-hum.
DR. POWERS: And I'm wondering if a response that went to
the effect that we know we have multiple lines of defense here in
degradation and one of those is not deemed risk-significant to some
level that's going to be an effective response to that particular
audience.
A rhetorical question.
MR. EMERSON: I understand the question. I don't have an
answer for it.
DR. POWERS: No one else does, either. So let's move on.
MR. MADDEN: Dr. Powers, just one thing. I mean, the
Thermo-Lag question is just a little bit more complex than just
responding to looking at the adequacy of a barrier, for example. I
mean, there have been changes in shutdown methodology and approaches as
a result of Thermo-Lag and rerouting of cables, et cetera, et cetera.
So it's a little bit more complex than just that.
MR. EMERSON: I didn't mean to oversimplify it, Pat.
Okay. If there are no other questions, I'll move on to the
area of fire-protection standards and guidance development.
I think everyone's familiar with the provisions of the
Commission decision directing the staff to work closely with the NFPA to
develop Standard 805 on an expedited basis and to ensure a proper
risk-informed performance-based approach, that that type of approach is
maintained.
The NFPA Technical Committee of Nuclear Facilities, and I'm
not going to try to get too much into that, because you'll hear a
separate talk from the NFPA representative later, released a draft for
proposals or, in layman's terms, comments. The distinction was lost to
me until fairly recently. These proposals or comments are due February
19, and there are several cycles of comment still ahead of us which I
think Mr. Bielen will elaborate on further, perhaps.
DR. APOSTOLAKIS: Are we going to comment, Dana, on 805?
DR. POWERS: That's one of the questions the committee has
to address. If we choose to comment, we'll have to bring it forward in
front of the full committee, unless we choose to comment separately and
individually.
MR. EMERSON: I'd like to offer our current views on the
NFPA 805, and these are understandably at a fairly high level. Just
based on my personal observations of the committee's activities over the
last six months to a year, there has been a great deal of effort, and I
can't stress that strongly enough, a great deal of effort by the
committee members -- Ed Connell is one of them -- to develop a
high-quality draft. Recognizing that this foray into performance-based
risk-informed standards is a fairly new area, there have been some
extremely -- extreme amount of work has gone into developing this
product to the extent that you see it so far. In our view it is a
high-quality draft, but it is still a draft, and it needs to go through
the rest of the process to achieve its full potential as a useful
standard for possible adoption in the regulatory arena.
So we would urge that the process continue using the NFPA's
approach --
DR. APOSTOLAKIS: Let me understand something you just said.
It's not clear to me what the role of the standard is in the whole
structure of things. We are publishing the regulatory guide, right?
MR. MARSH: Yes, sir.
DR. APOSTOLAKIS: And I have perhaps the --
MR. MARSH: FPFI.
DR. APOSTOLAKIS: And then the standard gets published. How
does that change the way we do business?
MR. MARSH: Well, this is all part of a Commission-approved
process.
DR. APOSTOLAKIS: I understand that.
MR. MARSH: We asked the Commission, we recommended to the
Commission option 2, which option 2 said let's not go forward with our
own rulemaking to develop our own risk-informed performance-based
fire-protection reg guide. Wait for the NFPA. And if that standard
meets our -- we would recommend approval, then we'll come back to the
Commission and recommend approval of that standard.
DR. APOSTOLAKIS: So that would be the risk-informed here.
MR. MARSH: Yes, sir; that's the risk-informed avenue.
DR. APOSTOLAKIS: And you will stay during the discussion
with the NFPA?
MR. EMERSON: No, I'll let the NFPA representative, Mr.
Bielen, address --
DR. APOSTOLAKIS: But you will be in the room?
MR. EMERSON: Yes, I'll be here.
DR. MILLER: So the plan is to, if it meets the needs, to
endorse this standard as a reg guide down the line.
MR. MARSH: It would be endorsed through a rulemaking
effort, okay? You'd have to modify the rules, the regulation that would
be either 50.48 or Appendix R, however it gets constructed, as an
alternative to the existing regulatory structure. The option was to the
Commission we would allow -- this is a voluntary standard. If it's
endorsed, it would be a voluntary way of meeting 50.48. It would not be
a requirement for all licensees to comply with the NFPA standard. It
would be voluntary. And it would have to be done through a rulemaking
effort, changing the regulations to endorse the standard.
DR. MILLER: And this standard is being developed by --
MR. MARSH: NFPA.
DR. MILLER: And that -- the standards development group
comprises a selected group. Does that include staff --
MR. MARSH: I'd like them to tell you all about that group.
DR. MILLER: Okay.
MR. MARSH: There's a lot of important insights that people
have brought to that committee.
DR. MILLER: We'll hear more about it later.
MR. MARSH: Yes, sir.
DR. APOSTOLAKIS: You will hear a lot.
MR. EMERSON: We expect because of the effort that's going
in to risk-inform the standard that it could be an important factor in
improving the use of risk information for fire protection. I expect
there will be a number of insights captured and ways to use risk that
will end up benefiting the industry.
DR. APOSTOLAKIS: So these are your personal views. NEI has
not reviewed it.
Does NEI participate in the development of the standard?
MR. EMERSON: NEI has a representative on the committee, as
do other industry organizations.
DR. POWERS: Do you have any sense of the enthusiasm that
there might exist within the industry for this guide?
MR. EMERSON: I think there is a lot of interest, and I
think there are concerns. And that's the reason why I'm recommending
that we allow the process to continue to completion, because those
concerns need to be addressed in making sure we have the best possible
standard.
DR. APOSTOLAKIS: I'll bet you the concerns have to do with
the word "voluntary."
DR. POWERS: No, I suspect the concerns have to do with the
word "burden."
DR. APOSTOLAKIS: What?
DR. POWERS: "Burden."
DR. APOSTOLAKIS: Well, that goes together with --
DR. POWERS: Go ahead, Fred.
DR. APOSTOLAKIS: It will be as voluntary as 1.174 is, don't
you think?
DR. POWERS: No, I think this will be a completely voluntary
thing. I think this is just an alternative to using Appendix R.
MR. MARSH: 1.174 is a technique one can use, right? This
would not be a -- this is a regulatory requirement. NFPA 805, if it's
endorsed, and if a licensee chooses to go that route, forms the
regulatory oversight, it's a new licensing basis, you know.
MR. EMERSON: One of our beliefs on maximizing the
usefulness of the standard to the industry would be to assure as much
flexibility in its application as possible. We think that we should be
allowed to use it not just in an -- say an all-or-nothing proposition,
you either choose it as an alternative to the existing standard or you
can't use any portion of it at all for any reason, and we believe that
flexibility in its use, to take some of the insights that have gone in
through the efforts of the committee should be made available to
licensees, even those licensees, even those who choose to remain within
the existing licensing basis to improve the application of risk insights
to their compliance with that licensing basis.
DR. POWERS: Well, you're raising a big issue here, which
has been a concern to some members of the committee. Of course, if the
industry can use this standard flexibly, then the staff should be able
to do the same thing. So using PRA insights, for example, to increase
regulatory requirements in the deterministic side should be as
acceptable to the industry as it would be to use this guide to support
exemption requests.
And then the big question is do you really have two tiers?
Can you really pick and choose if you have two tiers and say well, gee,
you know, tier 1, this looks good to me, I'm going to do it, but the
other part I don't like, I'll switch to the other tier. Then we are
creating an even bigger mess than we have now.
So -- and that's something that's bigger than fire
protection, in my opinion. We wrote in one of our letters that there
will be two tiers of regulation, and in the stakeholders' meeting they
picked it up and said yes, yes, we'll have two tiers. But I'm not sure
that everybody understands what having two tiers means, and I don't
think it means that you will pick and choose and create your own tier.
You know, either you follow tier 1 or you follow tier 2, because there
may be other requirements that are in tier 1 that are related to the one
that you like, or you don't like. So that is something that perhaps NEI
also wants to think about.
MR. EMERSON: We certainly have thought about it.
DR. POWERS: I think when you look at 805 the thing that
comes through almost immediately is gee, can a guy come pick and choose
in here? I mean, it just struck -- by page 3 I was asking that
question. And I think it's one that we ought to pose to the committee,
have they thought about it. Because I think they have thought about an
integrated structure which doesn't lend itself well to picking and
choosing, and then what Fred's saying is gee, for maximum flexibility,
there are -- I believe your words were ideas and thinking that went
behind some of these things that could be applicable. And that's
interesting. That may be the grist for reg guides, and that may be
grist for regulations. And it would be interesting to interrogate on
that point.
DR. APOSTOLAKIS: Well, I really think our committee should
look at this issue of the two tiers, you know, at a higher level, not
just in the context of fire protection and maybe say something. Because
evidently now it's acquiring momentum, we're going to have two tiers,
and I'm not sure if people have really digested the implications or
possible problems.
DR. POWERS: When the ACRS as a whole has addressed that
issue, I guess right now they have a mixed record. In one case they
recommended picking and choosing, and in one case they rejected picking
and choosing, and in both cases they were overridden.
DR. KRESS: In the source term reassessment as a committee
we recommended picking and choosing.
DR. POWERS: I guess we were endorsed in that.
MR. MARSH: Right. That approach is being endorsed. That
is allowed.
DR. APOSTOLAKIS: But in the last letter I think we just
observed.
DR. POWERS: Yes. What we wanted the Commission to be aware
of is that they were going to proliferate regulations and proliferate
difficulty in having a trained work force to inspect and enforce those.
And it's very difficult --
MR. MARSH: Assessing compliance with the structure,
whatever that structure may be. And even in proposing changes to the
structure in an amendment space, what is being deviated from.
DR. POWERS: We have enough difficulties with the
pre-'79-post-'79 split, the BTA and Appendix R split, and all the
exemptions. This is a group of people that may not be wild about
picking and choosing.
MR. EMERSON: I would agree that the picking and choosing
process is not on the surface something that you would want to sign up
for, but I guess where I'm going is that a licensee who has a great deal
invested in his current licensing basis and doesn't want to make the
extreme effort to transfer to an entirely new, different licensing basis
should nonetheless be able to reap the benefits of risk information in
his compliance with the current licensing basis, and if there are
insights that the NFPA process has offered that would be useful in that
area, if there are places where it would make sense to apply those
insights, then I think we should explore ways to do that.
DR. APOSTOLAKIS: Now wouldn't or shouldn't the regulatory
guide in preparation now have words to that effect? I mean, I
understand its main purpose is to consolidate the existing regulations,
but since we are going through the process, there should be some
allowance for risk information along the lines you just set.
MR. EMERSON: Well, the staff this morning talked about the
extent to which you put risk information into the reg guide versus
whether you use risk information to implement the reg guide, and I think
that's going to have to come out as the reg guide advances farther from
its current outline form.
DR. POWERS: It seems to me that just Reg Guide 1.174
provides the adequate opportunity for the occasional use of risk
information to modify your licensing basis in light of risk information.
You know, if you're not going to switch completely over but you find one
element of your program under Appendix R or the BTA, that would benefit
from risk information; 1.174 looks like it's a perfectly legitimate way
to go about taking advantage of it.
DR. KRESS: I thought that too, Dana. Then I got to
thinking that 1.174 in total requires a section that says you will meet
the deterministic requirements of the rules. And then when you come up
against here you've got one set of rules versus another set. What does
that particular provision 1.174 mean, and how do you decide whether or
not you've met that stricture? I'm not sure how you --
DR. POWERS: Well, it seems to me that that stricture exists
that you will meet the requirements of the rules except the rule that
you're trying to get changed. I mean, it's not a catch-22.
DR. KRESS: Well, that is catch-22 I think.
DR. POWERS: Well, it could be, but we were told explicitly
that it is not.
DR. KRESS: It meets all the rules except the one you're
trying to change.
DR. POWERS: Yes, I mean --
DR. KRESS: So we'll just systematically change all the
rules -- I think that is a catch-22 at 1.174. When you're dealing with
rule changes --
DR. POWERS: We worried very much about that when we
discussed 1.174 when it was being produced, and it became -- not that we
thought it was the intent. We were just worried about the explicit
language. And I think there's clarification in that, that it's meeting
all the other rules, not the one that you're actually addressing.
DR. KRESS: 1.174 is generally related to plant changes
themselves, specific changes to the procedures or the plant
configuration or something, and the question is does the resulting new
configuration meet all the rules. It normally doesn't talk about
changing rules because licensees don't change the rules. They get
exemptions from them or something. So that's why I think it tends to be
a catch-22.
DR. POWERS: I mean, here you would not really be changing
rules either. You'd be implementing in a different fashion.
MR. WEST: This is Steve West. This is a question I think
we're going to have to deal with later. I'm not sure what it has to do
specifically with the development of NFPA 805 at this stage. I mean, I
think the objective of the committee is to develop a standard for a
performance-based, risk-informed fire protection program at nuclear
powerplants, and I trust that the committee members aren't doing that
with also the -- in the back of their minds trying to work out how they
could fit this piece or that piece into an existing program without
adopting the standard in whole.
You know, right now under 5012, which governs the exemption
process, a licensee can request an exemption and it's basically up to
them to submit a technical basis that supports the exemption. And I
guess if they were to say the standard is in place, if they could use
techniques that are specified in the standard or other information in
the standard, we'd have to decide at that point if it was technically
justified.
Now if all of a sudden the standard's out and some licensee
comes in with 100 exemptions, but they're not going to adopt the
standard, I mean, that raises a question, right? Because you haven't
heard much about the standard yet, but to implement the standard is
going to require some effort on the part of the licensees.
DR. POWERS: No kidding.
MR. WEST: For example, there's a sitewide risk assessment,
and that's one of the bases that gives us comfort in having this
performance-based risk-informed standard in the first place. There's
this overlay of the sitewide risk assessment, which we'll hear more
about. I don't want to get into that. But if all of a sudden licensees
see this great -- a lot of flexibility and they can come in with a lot
of exemptions, you'd have to wonder if they're trying to come up with a
cute way or having a risk-informed, performance-based, fire protection
program without the sitewide risk assessment, for example.
DR. POWERS: I'm pretty sure I know which one of those is
the least pained route, and it is not the exemption route.
MR. WEST: Yes.
DR. POWERS: I think we're going to have to move on.
MR. EMERSON: Last in this topic, there was a good
discussion this morning of the comprehensive reg guide, and my only
comment here is that we are looking at it very carefully, and will
provide comments at the time that Ed Connell indicated this morning.
DR. MILLER: Do you have a comment on the objective?
MR. EMERSON: I'm sorry, on the --
DR. MILLER: This reg guide. Do you have a comment on the
objective of the reg guide?
MR. EMERSON: As far as consolidating guidance? I think Ed
characterized it correctly this morning, that industry supports the
consolidation of guidance.
Okay. Next is another area that we spent some time talking
about before, and that's the area of fire-induced circuit failure
resolution. And what I propose to talk about here is basically work in
progress by the NEI task force as to what our currently envisioned
approach looks like, and certainly we would be receptive to feedback
from your committee, as well as from anyone else.
The approach that we're looking at at this point and working
to put more meat on involves three basic elements. One is looking at
the cable failure modes and mechanisms if you get a fire large enough to
cause cable damage, how will it fail, to try to hopefully add some
knowledge to what currently exists as to how cables will fail in the
event of fires of certain sizes.
DR. MILLER: Is there an expectation for a data base out
there on that, or are you going to generate one in that area?
MR. EMERSON: Well, we're certainly looking at the existing
information that's out there, and we would consider augmenting that with
more research if the need arises. I'm not going to try to address what
the Office of Research is doing in that area right now, and we haven't
made any decisions on what more research might be necessary, but
hopefully we could compile some additional data from existing literature
to the extent that it applies to this area.
Again, the first area involves cable failure modes and
mechanisms. The second area involves risk information. And when I say
risk information, I'm not talking about the likelihood of cable failure
given a damaging fire, it's the likelihood that you will either get a
fire large enough to cause cable damage, the likelihood that you can
recover from the damage that's involved in a fire, or the availability
of alternate systems to help you in the event that your systems are
damaged.
And the third area of course is other considerations. Since
this is a risk-informed method and not a risk-based method, we would
have to consider the usual deterministic considerations. We would have
to consider defense in depth. We would have to consider safety margins
and other things that were touched on by the staff this morning.
I think I just talked about this slide without putting it
up.
DR. MILLER: As I look at that process, it sounds quite
good, except I expressed a concern, is there a data base that will
provide you to get off bullet 1, so if you don't get off bullet 1, you
won't go to bullet 2, right?
MR. EMERSON: Well, I expect we will be able to apply bullet
2 insights. I think the best possible method would address all three
areas. If I can get into subsequent slides, I want to get into each of
these areas in a little more detail.
Our intent first would be to try to look at the data base or
the information we can accumulate and the types of failure mechanisms we
can postulate as a result of our review to screen out potential failures
involving our initial goal interactions between two different cables in
the same tray. Then next we would consider what we could do with
interactions between conductors within the same cable.
This would be again based on an improved understanding of
cable failure modes based on review of test results, new testing if that
was appropriate, some sort of analysis. We haven't pinned down exactly
what is going to serve as the basis for this yet. We are developing
what needs to be done to make this method workable within our committee.
DR. POWERS: It seems to me that there is a real challenge
in knowing whether the test that you identify or what analyses that you
have on cable failure modes have really improved the understanding. I
have certainly seen presentations in which individuals have done a test,
and they have said see, this particular kind of failure is very
unlikely, 12 out of 12 times I tried it, I didn't get this particular
kind of failure. That kind of failure was precisely the kind of failure
that has been observed in accidents, Browns Ferry not being the only
one.
This whole area of what is an improvement in understanding
and what is an inapplicable data point seems to be a challenge to me. I
certainly don't know the answer of what is an adequately appropriate
test. Is there someone trying to figure that out, what is a good data
base on cable failure modes and what goes into it and what is excluded
because it's just not applicable?
MR. EMERSON: I think Nathan would like to say something.
DR. POWERS: He gets his opportunity tomorrow morning.
MR. SIU: I was waiting to hear what Fred was going to say.
Again, tomorrow I think we'll talk about at least how we're
approaching the problem, and I think you're right, if you try to do this
in a purely statistical fashion saying gee, I've got zero failures out
of 12 or I got one failure out of 12, you're really susceptible to this
issue of the applicability of that kind of test to the situation you've
got at hand, and what you really need is some sort of model-based
approach where you understand what are the parameters that affect that
likelihood and then use the test results to quantify the model
parameters.
We are right at the beginning of that stage trying to
identify what are those parameters and what are the sources of data that
we might be able to use to quantify those parameters. I don't know what
industry is doing along these lines.
MR. EMERSON: I would say that's a pretty good
characterization of what we're doing too. We've been reviewing the test
reports that we can find that have results applicable to this activity.
We've been talking to individuals who are knowledgeable in cable failure
modes and mechanisms, and given fires, given their experience in the
industry, both the nuclear industry and outside the industry, we're just
at the stage now where we're trying to pull some of those insights
together and decide what else we need to do to improve the knowledge
base to the point where we can make some statements that we the industry
can accept and the staff can accept as a screening criterion for this,
and I wouldn't pretend that we could come up with this lightly or in the
near future.
You know, this is where we intend to go, but the process of
fleshing out our intent with solid information that the staff and
industry can accept is going to take a little longer.
DR. POWERS: You and I both attended an international
conference in which we saw lots of people standing up and saying we're
going to create a new fire-incident data base.
MR. EMERSON: Um-hum.
DR. POWERS: It's going to be international in scope. And
we did not hear a lot of people standing up saying we're going to create
data bases on effects of fire like cable failure modes and things like
that. It is my suspicion that fire incident data is not transferable
from one country to another, but fire effects, like cable failure modes,
is transferable from country to country. Should we be pressing these
various people creating data bases to, if not abandon their incident
data base, at least augment it with failure mode data bases?
MR. EMERSON: Well, I wouldn't try to say whether we should
scrap the efforts to improve the incident data bases internationally,
but certainly we should make the effort to broaden the data bases to
individuals in other countries, and as an example of that, we heard at
our last information forum a presentation from a senior engineer at
Electricite De France about cable testing that they have done. It
wasn't with this in mind, but the fact is that other countries are
interested in cable failure modes, and there may be some ways that we
can take advantage of work that they're doing and vice versa.
As far as developing a data base, certainly the process of
compiling information and test results, it would have to go into some
sort of a data base. I just don't know what that would look like.
Certainly you could not make a decision on screening out
something based on a cable failure mechanism in isolation from
consideration of things like defense in depth and safety margins when we
would propose that any decision based on this tool to screen out
potential interactions at this step would be done through consideration
of defense in depth, so that you didn't just say well, we think it's
unlikely, boom, it's gone. We'd have to think about it as to what the
consequences of screening that out were.
Okay, the second method has to do with -- the first was
intended to be a generic method. The second step is intended to be a
plant-specific method. One member of the staff pointed out at one of
our meetings, and rightly so, that it's difficult to create a generic
criterion for screening based on risk, because plants have
plant-specific risk elements. So we would propose that the second
element, the application of risk, be done on a plant-specific basis to
address failures or potential failures that had not been screened out
with the first one.
We would -- the types of analysis that would be necessary to
apply this, first we would ask the proposed user to determine either the
frequency of a damaging fire or the probability of conditions supporting
a damaging fire, determine whether that reached some threshold value.
We would ask them to determine the probability of recovery of equipment
if it were damaged through either operator action or some other type of
recovery, and determine the availability of redundant equipment if it
could not be recovered.
We would look at the PSA elements that went into each of
those three determinations and develop an appropriate criterion with
which to compare that, and again in parallel consider the impact on
defense in depth and safety margins.
As a result of this second step, should it come about as we
foresee, you would have screened out more potential failures based on
plant-specific determinations.
The rest of the method has to do with, you know, if you were
unsuccessful in screening out potential interactions at this stage, you
probably have a pretty good idea that they were safety-significant
enough that you should do something about it. Before you do that, you
might want to analyze it using more traditional analysis methods, the
kind that are used currently by licensees. And then lastly, once you'd
finished that step, to take appropriate action to address any remaining
potential failures that you could not write off as being nonsignificant.
And obviously you need to document what you've done for future
reference.
I don't want to spend a lot of time on this. This covers
some of the elements that you would consider in the --
DR. MILLER: Do you have a kind of expectation for a time
frame in --
MR. EMERSON: In the development of this method?
DR. MILLER: For developing this method and knowing that the
data base is certainly a sizable unknown in this whole issue.
MR. EMERSON: Yes. I have some slides at the end which kind
of explain what our time frame is for developing the pieces which
support this, if you can wait.
As I said before, our primary goal is to -- we feel we
can -- if we could address interactions between cables, that would help
us rule out a lot of interactions. If the likelihood of interactions
between cables is low, we could screen out a large fraction of potential
failures.
DR. MILLER: That would go a long way in answering a lot of
problems we have right now independent of the rest of your process,
wouldn't it?
MR. EMERSON: Those on our committee who do this sort of
analysis say that would help a lot. We always face the likelihood that
perhaps this is an unrealistic goal, but this is where we're going at
this point.
Obviously we'd like to consider failures within single
multiconductor cables, that that would be a second-order goal.
In the area of the risk information, I'm not going to try to
go through this in any detail. In looking at the frequency of fires or
the probability of conditions supporting damaging fires, these are the
types of parameters, PSA parameters that we would consider --
initiation, exposure, growth, detection and manual suppression, and
automatic suppression are the principal parameters that would allow you
to calculate a probability of achieving a damaging fire.
The other two elements involve the probability of recovery
of equipment, as I said, and the availability of redundant equipment,
and these are the parameters that you would be able to calculate those
values from.
With respect to the timetable, our current timetable is to
in the area of the generic circuit failure mechanistic method we're
developing a method of -- a draft, a matrix of failure combinations by
the end of this month, and by the end of February we're going to have
drafted a program plan for achieving our goals. It will reflect the
information we've gathered to date, the test results and so forth. It
will address what further work we need to do to gather more information
or conduct testing or address further analysis requirements, and it
would factor in the work that the Office of Research is doing to make
sure that we're not doing things that are entirely redundant to their
activities. So this program plan is intended to say how we are going to
get this tool developed and what we have to do to get there and lay out
in more detail how we will get there.
In the area of the plant-specific risk tool, we have drafted
an outline for the three specific risk methods and their use, and we
intend to turn that into the program plan that I mentioned for the
earlier tool, again by the end of February.
Factoring in the deterministic methods including the defense
in depth, safety margin, and traditional deterministic methods, we are
going to put that piece of it into the overall program plan, again by
the end of February. So our initial timetable for developing how we are
going to turn this plan into something of a reality is due by the end of
next month.
We expect to run some sort of a proof of concept or a pilot,
and we currently envision that this would take place during the March
and April time frame.
As Steve has pointed out this morning, we have been
interacting with the staff on at least a monthly basis, I think, since
we actually started our work, and we've briefed the staff within very
short periods of time after our committee meetings. And so they are
pretty much in tune with what we are doing and where we plan to go.
DR. MILLER: What will be the purpose of the pilot?
MR. EMERSON: The purpose of the pilot would be to see
whether this method could work at an actual site, whether a plant could
take this method and see does it work or not, given how we expect it to
work.
DR. POWERS: Fred, we were anticipating having another
subcommittee meeting in Region I in June. I wonder, would that be an
appropriate time to come discuss what you've done here?
MR. EMERSON: We'd always be available to share with you our
progress.
DR. POWERS: It would be in Region I offices.
MR. EMERSON: Should be no problem.
DR. POWERS: Okay. We might just stay in touch as you
progress along. Sometimes schedules slip. Not NEIs, but sometimes our
schedule slips. And if it's convenient and what not, that might be a
good time to do it.
MR. EMERSON: That's fine.
To wrap this up, I have no illusions that this plan is the
same thing as an actual method. You know, what we intend to do is all
well and good, but it's what we end up with that will be whether we have
achieved our goals or not, and we're going to be exerting some
significant efforts to try to get this to be a workable method. And we
look forward to further discussions with the staff on making this a
method the staff can embrace as well as the industry.
The last area has to do with -- the last area I want to
cover, and I only have one slide for this, is the area of risk-informed
fire applications. We were asked to review draft NUREG-1521, which
covered a number of such potential applications for risk information and
performance bases to the fire-protection area, and we think it's
important, as I indicated earlier when we were talking about the draft
standard, to be able to apply risk information to existing
fire-protection programs under current licensing bases to gain the
advantages of that information as perhaps a more objective way for
addressing either plant-specific or generic issues.
And what we propose here is that we agree with the staff
that this is a worthwhile goal to promote the increased use of risk
information and to work out agreed-upon actual applications that staff
can have some confidence that the industry knows how to use properly and
the industry can have some confidence that the staff will accept them if
they are properly done. But anyway, predeveloped applications, if you
will. And we would propose to start that.
DR. POWERS: We had discussions earlier this morning on what
it meant to do fire-risk assessments and the quality of our tools, and I
would hope that that would be an area of a great deal of cooperative
agreement, because it's clear that we're all suffering from a lack of
tools, the industry and the staff, for exactly the same reason, where we
can find nonconformances and noncompliances, but it's very difficult
right now to do anything that looks like a justifiable or well-supported
assessment of the risk-significance of those. And as we move our
inspection and enforcement policy more toward risk-informed, it's in
everyone's best interest to have acceptable tools that are broadly
supported within the technical community to make these judgments about
risk and what the significance of these findings are.
MR. EMERSON: I would agree that having those tools is
needed, and having ways to use those tools in agreed-upon ways is also
needed.
DR. POWERS: That's right.
Do the Members have any additional questions?
Fred, always useful to get your insights, and we'll stay in
contact perhaps for our June fire-protection meeting, what not.
I have made an adjustment to the schedule to accommodate
some travel plans, and we propose now to move on to the presentation on
NFPA.
MR. MARSH: From the industry? From the --
DR. POWERS: From the group.
MR. MARSH: NFPA?
DR. POWERS: Yes.
MR. BIELEN: Good afternoon.
DR. POWERS: As you progress through your presentation, if
there is a point at about midway that it would be appropriate to take a
break, you might indicate.
MR. BIELEN: Okay. Will do.
I just wanted to give you a little bit of a brief overview
of where we are with the development of NFPA 805, just go over the
agenda.
The purpose of this presentation really is to give really a
progress report, and I would want to differentiate that between a
progress report versus a defense. We're not going to really be
defending the document as far as what we did or some of the details.
Some of the details actually I'm not going to be able to answer, because
some of the experts that we have, for instance, on PRA are not here
today. I have some committee members here in the audience to help
support the presentation if we need, but as far as some of the details
go of the report, we may not be able to answer as we're going through.
So this is really to be just a progress report on where we are with the
document, where we are in the cycle, and how much further along we have
to go.
I think it would be useful to also review the NFPA
standards-making process. I've done this before on some other
presentations previously, but I think it would be worthwhile to go over
again a little bit about the NFPA process and how that works so that you
get a better feel for what a consensus-base document actually is, the
way we present it.
Also I'd like to go over where we are in the cycle, again to
show you how far along we are, where we are at this point, and what the
steps are in the future that we have to still go through. And then I do
want to go over a brief overview of NFPA 805 and I call it Draft 6.3,
which was released to the public back at the end of November -- so we've
gone through quite a few number of drafts -- and then finish up with the
conclusions.
The NFPA standards-making process -- and again this is just
a real brief overview, and I left out a lot of the details and some of
the minor points in between -- but essentially the NFPA Standards
Council, which is a 12-member committee that oversees all the standards
activity, approves a new project, as in this case with NFPA 805. An
NFPA technical committee is then formed. In this particular case the
Technical Committee on Nuclear Facilities was an existing committee with
some other documents, NFPA 801, 802, 803, and 804. So in this
particular case, an existing technical committee took on the
responsibility of developing this new project, NFPA 805.
The Technical Committee would then have to on new projects
develop a draft document to be released to the public. And the reason
why we have to have the draft document released to the public is because
we have a proposal closing date or a call for proposals, and we have to
have a document that the public has to see in order to submit proposals
on.
So this is a consensus-based or a consensus process, so we
have several different steps as far as proposals and comments, and I
heard earlier how there was some confusion and differences as to what
they are, but again I won't get bogged down with the details of the
differences between it. But basically it's a two-step process where
there's a proposal period and a comment period. So there's a deadline
for call for proposals on the draft document. The committee --
DR. MILLER: Excuse me. What's the draft document?
MR. BIELEN: The draft document is NFPA 805, and I --
DR. MILLER: The draft document you're talking about in
bullet 3 is what we have in our hand right now.
MR. BIELEN: Right, Draft 6.3?
DR. MILLER: Right.
MR. BIELEN: Yes, I believe the committee has that. Okay.
DR. MILLER: I don't know if it's 6.3; we have a draft.
MR. BIELEN: Right, a draft of NFPA 805. Yes, it would be
6.3, which is one that was released to the public.
There's a deadline on proposals, and I'll go over the cycle
on the next series of overheads that I have. But essentially then the
committee would have to meet and review all the proposals that we have
both from the public and any ones that are generated by the committee
and take an action on them. And there's --
DR. MILLER: Proposals meaning in this case comments?
MR. BIELEN: We can call them comments, yes. In the NFPA
process we call them proposals, but they're essentially comments on the
draft.
DR. MILLER: So the Technical Committee basically as a
consensus committee spent the better part of a year, I suppose, coming
up with that.
MR. BIELEN: About two years, yes.
DR. MILLER: Or two years. How large a committee was that?
I could count the members, but --
MR. BIELEN: Yes, there's -- let me just see here. I do
have something here. Actually not for the committee. This is just some
information as far as the number of people. We have about 51 people
altogether, including technical committee members, both principals and
alternates, and then there's several volunteer people. And just to give
you a feel for the types of people that we have, I could just go over
some of the different areas and the names and the industries.
As far as the industry users or the user groups we have Stan
Davis from Pennsylvania Power & Light, Frank Garrett from Arizona Public
Service, who represents NEI on the committee, Dennis Henneke from
Southern California Edison, Chris Ksobiech from Wisconsin Electric,
Cliff Sinopoli, who is with us today, from Baltimore Gas & Electric,
representing EEI. Ray Tell from Los Alamos National Laboratories. Mike
Vitacco from Westinghouse Savannah River. Lee Warnick, who is with us
today from Virginia Power. Then we have a series of special experts.
Rick Alpert from Triad Fire Protection Engineering, Mario Antonetti from
Gage-Babcock, Tom O'Connor from EPM, Arie Go from Bechtel Corporation,
Liz Kleinsorg from Duke Engineering. Don Kohn, Dave Notley, Dan
Sheridan, Black & Veatch; Bill Sullivan, Contingency Management.
And I guess maybe it will take a little bit longer --
DR. MILLER: I didn't really want you to read all the
members. I was just trying to -- I meant through the process. I'm
trying to relate it to the process I've been through, so --
MR. BIELEN: Okay. You've been on the NFPA process before?
DR. MILLER: No, I've been through a different --
MR. BIELEN: Different process.
DR. MILLER: Standard process where we start out with a
reasonably small committee that develops a draft, and that goes to a
larger committee.
MR. BIELEN: Yes, we've had task groups that went over
different chapters, and then the task groups would get together with the
main committee and go over the whole document.
DR. MILLER: Page 50, I assume that was the one that went
over the overall document, and then different groups developed different
parts.
MR. BIELEN: Right. Correct.
DR. POWERS: In light of the public interest that
fire-protection issues have generated, I'm surprised that there isn't a
member that would -- what we refer to as the public when we talk about
nuclear regulations.
MR. BIELEN: Yes, we have a category for consumer, but very
few consumers or the general public join into the technical committees,
whether it's this Nuclear Facilities Committee or any of the other ones.
It is a category that we have in the process, and unfortunately the
consumers don't have a lot of funding I think it is to attend a lot of
the committee meetings, so they don't apply and don't attend. But also
the meetings are open committee meetings, so any member of the public
can attend any of the committee meetings.
DR. POWERS: Okay.
MR. BIELEN: So they don't have to formally be a member of
the committee, they can attend. And actually we do get several guests,
but they're usually from the industry or insurance or something along
those lines.
DR. MILLER: Did you invite specific people from the public
to participate from the very beginning or just let it know it's open?
MR. BIELEN: The latter. We pretty much publish when we're
having our meetings and where they're going to be, and it's open for
anyone to attend. So we didn't specifically invite members from
anywhere.
DR. POWERS: Yes, I think that is one of my inherent
objections to the entire consensus process --
DR. MILLER: Well, the process I was involved in -- in fact
I represented the public for four or five years on it -- they also
invited somebody specifically who participated from a quote "antinuclear
group," who did participate, as well as I did. We too represented the
public. Of course we had to pay our own way, so that's a big issue.
This is not an inexpensive process.
MR. BIELEN: Correct, and all the members of the committees
are volunteers, and they pay their own way. So this isn't funded
through NFPA in any way.
DR. MILLER: So at this point the document we have has been
through the consensus process to the point where all issues, all
objections or issues brought up by every member of the committee have
been addressed independently?
MR. BIELEN: No, actually --
DR. MILLER: That process at that point now?
MR. BIELEN: We're still in the process right here where
it's call for proposals. The proposal closing date in some of the other
overheads I have is February 19 of this year, '99.
DR. MILLER: My question was, have all the issues brought up
by the Technical Committee been addressed at this point or not?
MR. BIELEN: Well, the committee is addressing the issues as
we go along and always making modifications to it, so we're not done
with addressing technical issues yet.
DR. MILLER: So internally the consensus process is not
done, as well as externally.
MR. BIELEN: Correct. Yes, internally we've still got a
ways to go, and I'll show you when I go through the cycle, and actually
we're still -- we haven't even reached this point yet where -- well,
we're in the call for proposals. We haven't reached the deadline yet.
The committee's gone through several editions of drafts, but we still
have the proposal, the ROP meeting, and what we call the ROC meeting to
finalize it. So we still have several steps left in the process to go.
Once the committee has acted on the proposals, it's
published in a report called "The Report on Proposals." At that point
the public again gets a chance to look at the proposals and how the
committee acted on it, and then they get a chance to submit public
comments on the Report on Proposals or the ROP, and the process more or
less repeats itself where the committee has to act on the comments. The
comments get published in a report called "The Report on Comments" or
the ROC --
DR. MILLER: That basically addresses in some way every
comment, right?
MR. BIELEN: Correct. I think what is important to remember
also is the official actions are taken by the committee on the proposals
and then the comments, so if there are concerns or issues that are
raised by either committee members, the public or anybody where it says
you didn't address this issue and you ought to look at that issue,
unfortunately because of the way our process is set up that doesn't mean
that that issue is necessarily going to be addressed.
The way it is going to get addressed is if that person or
organization of someone submits a public proposal or a public comment.
Then it has to be reviewed by the committee. So in the consensus
process there is a formalized process of submitting proposals and
comments and the committee has to address those, and that is really the
direction that we have to go as far as making any modifications to this
draft or after the ROP period.
In the consensus process the NFPA membership then gets a
chance to vote on the document as with all documents.
DR. MILLER: That is assuming they have already commented,
they just vote on it yea or nay --
MR. BIELEN: The membership? No. That is not until, and
you will see in the next couple of overheads, until May of 2000. That
is at the end of the process.
DR. MILLER: At that step in the process you say the
membership. Is that a consensus process or do they just vote yes or no?
MR. BIELEN: Oh, okay, yes. At that time it is a go/no go
or yes/no, as to whether or not the document should go forward or get
released. The final step is the Standards Council actually issues a
document based on the Report on Proposals and the balloting results,
based on the balloting results of the Report on Comments, and then based
on any floor action at the fall meeting.
They take all that into consideration and then they either
issue or don't issue the new document.
DR. MILLER: So this doesn't go to ANSI though?
MR. BIELEN: It then goes to ANSI -- right, and then it
becomes ANSI-approved.
DR. MILLER: I should have looked one ahead -- you probably
had that on your next overhead.
MR. BIELEN: I didn't about ANSI but I do about some of the
other questions.
The cycle, just to give you an idea of how long it takes for
a document to get released and some of the steps that are involved and
where we are, we are what we call an annual 2000 cycle right now, and
you will see later on that that means the document gets published in the
Year 2000.
Let me jump to Bullet Number 3 here first, and let me just
back up. As I mentioned before, February 19th of this year, the public
proposal closing date, that is the deadline for the proposals, to back
up -- and this is actually a mistake -- it should be November 25th,
1998 -- a draft document that you have in front of you has been released
to the public for review with a call for proposals with that February
19th deadline for proposals, so we are still in the proposal stage, any
submissions from the public.
The committee is going to be meeting next week down in
Orlando, the 26th through the 28th, to act on any internal committee
proposals that we generated, which is kind of an extra step that we
added into the process. We didn't want to wait just for the public
proposals. The committee also wanted to continually work on the
document and we wanted to formalize that a little bit so we developed an
internal committee proposal process.
DR. MILLER: So by that time all of the committee members
should have submitted all their proposals?
MR. BIELEN: For the internal deadline that is correct and
then if they miss the internal deadline, they still have until February
14th for the formal submission of proposals. Actually to this point
right now we have 50 public proposals, all from Ed Connell, out of
150 -- actually, we have 100 internal committee ones and 50 public and
the 50 public are Ed Connell's so at this meeting next week we will be
acting on all 150 of those proposals.
DR. MILLER: So you have none outside of the committee or
NRC?
MR. BIELEN: That's correct for public proposals. The
internal ones are all committee members, but as far as the external
proposals from the public right now they are all from a committee
member, correct.
DR. MILLER: But you would expect, I assume, by February
19th that you would have some from industry and others outside of --
MR. BIELEN: Right, I would assume that we would be getting
some from industry and outside groups.
Historically what's happened is people don't submit
proposals until the last week, and then they realize there is a deadline
and you get 50 or 75 percent of them in the last couple days of the
closing date. I can tell you from experience that is what happens all
the time, so, yes, I expect to have some from industry and whoever else.
DR. APOSTOLAKIS: So if we write a letter, then, Dana, that
would be a proposal?
CHAIRMAN POWERS: They might well treat it as a proposal.
DR. MILLER: Yes --
MR. BIELEN: There is actually a formal proposal form to be
filled out, but if you were to submit a letter I think we would probably
accept it as a proposal.
One thing we do want in the proposal form is we require
specific wording for what you want changed and substantiation for why
you want to change it.
For instance, we get bogged down with proposals or comments
that say you didn't address this issue -- you should address it. Okay.
But how do you want us to address it is the question.
DR. APOSTOLAKIS: I don't think that the ACRS can write a
letter to you. If we write a letter we do it with the EDO. I don't
know of any instance where we wrote a letter to somebody other than the
EDO or the Chairman of the Commission.
CHAIRMAN POWERS: I don't know of any reason why we
couldn't.
DR. APOSTOLAKIS: Or the Staff can forward our letter to the
EDO --
CHAIRMAN POWERS: They could do that too.
MR. BIELEN: Yes, you can have all your correspondence -- we
don't care how it comes through us, whether directly from you or from
the NRC Staff, and in this situation we are blessed with NRC Staff
members. We have got Nathan, Pat and Ed, so any one of those people can
deliver your concerns to us through a proposal.
DR. APOSTOLAKIS: Because I think we have to write to the
EDO. I don't know of any case where we wrote to anybody else except
Congress.
CHAIRMAN POWERS: I known of no restriction on who we write
to. I know it is usually to write to the Staff or the Commission.
DR. MILLER: It might be -- well --
DR. APOSTOLAKIS: It might be what?
DR. MILLER: I was thinking since we are going to be dealing
more with the standard, we might develop a policy on how we deal with
issues of this type, but I think that is an issue we should discuss
later.
I agree with you, George, that based on past practice we
have never written a letter to somebody besides the EDO or the Vice
President --
DR. APOSTOLAKIS: -- the Vice President --
DR. MILLER: -- or the Chairman of the Commission.
MR. BIELEN: And I would strongly recommend if you have
comments, concerns, issues, whatever, that you do provide us with input.
DR. APOSTOLAKIS: So at this February meeting it would be
very timely if we wanted to write something?
MR. BIELEN: Yes. There is a meeting we are having next
week, but then we are also having another one in March -- March 16th
through the 18th -- to complete the ROP meeting, so we have until that
meeting actually if you wanted to submit anything.
MR. MARSH: George, it is also timely in the sense of the
Commission meeting on February 9th where we are going to talk about the
NFPA standard.
DR. APOSTOLAKIS: The letter though will probably not be
ready. I mean the committee will write it if it decides to write it
during the meeting, but then it takes awhile for the Staff to process
it.
MR. MARSH: I am just thinking if the Commission would want
to know what your views would be on the standard thus far, that would be
helpful. I understand your time constraints though.
DR. APOSTOLAKIS: No, I think the Staff can read the draft
after the committee has approved it. It's just that it takes some
editing after that and so on. Anyway, that is something that is not
relevant at this point.
CHAIRMAN POWERS: We have an opportunity to resolve that. I
think it is more important that we get to the point that we can
formulate a position the full Committee can adopt, and I am hoping that
we do so. The mechanics of it I think we can sort out.
DR. APOSTOLAKIS: Well, Mr. Connell will talk about the
contents of it tomorrow, right?
MR. BIELEN: Right.
DR. APOSTOLAKIS: So he will get the subcommittee's views --
individual members' views, I'm sorry.
CHAIRMAN POWERS: That's okay -- the subcommittee's views
are fine.
I mean this subcommittee does not operate on a basis of the
full Committee.
MR. BIELEN: Okay, so as I mentioned, February 19th for the
public proposal closing date, and then the committee is going to have a
second ROP meeting, the final one, March 16th through the 18th, in
Albuquerque to review the remaining public comment -- proposals and take
action on those.
The report then gets published and there is a public comment
closing date of October 8th, 1999, so there's your second shot at it
once we submit proposal, the committee takes action on it. You are able
to see what the actions are, hence to submit another round of comments
by the October 8th deadline.
DR. APOSTOLAKIS: And then there will be a third chance?
MR. BIELEN: Nope.
DR. APOSTOLAKIS: Oh, no, no, no -- for the NRC Staff has to
accept it.
MR. BIELEN: Okay.
CHAIRMAN POWERS: Lots of opportunities here.
MR. BIELEN: The committee would have to meet by December
27th of 1999 to have their ROC, Report on Comment, meeting to take
action on any of those public comments.
DR. APOSTOLAKIS: Do you plan to meet the last couple days
there --
MR. BIELEN: I'm not -- around Christmas? No. I think we
will probably meet somewhere closer to the end of October into the
beginning of November and stay away from the holidays.
It then goes to the NFPA Annual Meeting of May 14th through
the 18th in the Year 2000 where the NFPA membership votes on it and that
will be in Denver, Colorado.
DR. APOSTOLAKIS: So the whole membership votes on the
standards?
MR. BIELEN: Yes. The NFPA -- when I say the NFPA
membership, it is whoever shows up at the Annual Meeting.
DR. APOSTOLAKIS: Usually how many show up?
MR. BIELEN: In the Annual Meetings they are typically a
little bit bigger -- I would say between 3,000 to 5,000 people show up.
Not all of them ar NFPA members. There might be 1500 to 2,000 people --
1,500, around there, that will be actually in the hall that are voting
members that will vote on the document.
DR. MILLER: What is the size of the NFPA as an association?
How many members total?
MR. BIELEN: We have got approximately 70,000 members --
just under, like 68,000 -- somewheres around there. Cliff?
MR. SINOPOLI: I am Cliff Sinopoli, Baltimore Gas &
Electric. I think one important thing too from a public standpoint,
also from the subcommittee's standpoint is once you put in a proposal or
a comment, you may not be satisfied with the response that you get from
the committee, you have the opportunity actually two more spots -- you
have the opportunity to get up in front of the entire membership to make
your case and you also have the opportunity, if I am not mistaken, Rich,
to go in front of the Standards Council -- say you lose on the floor of
the Annual Meeting, you can still go to the Standards Council provided
you have made these comments at each of the proper closing period, so
the public as well as this committee -- anyone has an opportunity to get
satisfaction up to the level of the Standards Council.
MR. BIELEN: Right. There's a whole number of steps
involved. If you submitted a proposal and if you submitted a comment,
and you didn't like the action that the committee took on either one,
you then have that option to get up in front of the membership at the
Annual Meeting and to try to get a floor amendment and convince them
that they should accept your proposal or accept your comment if the
committee rejected it, for instance, and then if you are still not
satisfied with the action there, you can then take your issue to the
NFPA Standards Council and file an appeal and argue your case with the
Standards Council, so there are several steps involved.
DR. MILLER: How often does that happen on an NFPA standard,
that somebody comes to the membership and makes an appeal?
MR. BIELEN: All the time.
DR. MILLER: All the time? How often does the membership
agree with the person making the appeal?
MR. BIELEN: I'd say it's about 50/50, depending on how
strong of an argument and they'll argue issues for hours on the floor
and at that point it is kind of a 50/50. It is a tossup as to which way
it is going to go.
When it then gets to the Council, that is where the odds
start to go against you. The Council doesn't generally like to go
against what the committee and what the membership does on the floor,
although they will if they have to in certain cases.
DR. MILLER: So at the membership meeting, of course
obviously that includes a lot of people who hardly know what a nuclear
plant is.
MR. BIELEN: Right, but they are --
DR. MILLER: They know the issue of --
MR. BIELEN: -- are fire protection people.
DR. MILLER: You get the input of fire protection
independent of a nuclear plant.
MR. BIELEN: Right -- and then at some point, at the July
Standards Council Meeting, if everything has gone well with the ROP, the
ROC and the floor action and any appeals, they then issue the document
in July and it gets published around the September-October 2000
timeframe and it is then available to the public at that point.
Kind of a brief overview of the document and actually there
is a flow chart that I will be going over at the end here, but the
document is comprised of several chapters -- Chapter One being the
introduction; Two is methodology -- telling you how to apply the
standard; Chapter Three is fire protection elements of defense-in-depth;
Chapter Four is performance requirements; Chapter 5 is fire protection
during decommissioning and permanent shutdown; Appendix A is just
explanatory material that goes along with the requirements in the
document; Appendix B is still being worked on by a task group --
DR. APOSTOLAKIS: So you don't have it?
MR. BIELEN: You have some material in Appendix B right now,
but it is very general, very brief.
DR. APOSTOLAKIS: Definitions?
MR. BIELEN: As a matter of fact, we put that in just as a
place-holder in order for the committee to do something. If we didn't
have anything in the draft, the public doesn't have a chance to submit
any proposals on that so what is in Appendix B right now more or less is
just a place-holder.
The task group met January 8th and I believe they are
meeting this Friday or beginning of next week before the main meeting,
but they are meeting again very shortly to still go over some of that
material, so by the time we meet next week or in March, we should have
something that is a little bit more detailed, but right now, correct,
what we have is probably not what you are looking for. It is very brief
and very general.
Then Appendix C is just references.
CHAIRMAN POWERS: I guess you are not here to defend so
much, but some of your committee members are here, so I'll let them
interject if they choose to.
In the course of this morning's discussion, questions were
raised and critiques given of the existing tools for doing risk analyses
in connection with fires, and concerns were expressed that those tools
may not be adequate to the tasks presented to them.
Those tasks are not wildly different at all from the kinds
of tasks NFPA 805 would have risk analyses do. Why aren't those tools
then inadequate and incomplete for doing the NFPA? In other words, why
is Appendix B ever going to be an inadequate document?
MR. BIELEN: I saw Nathan out of the corner of my eye going
up there. Do you have any comments?
MR. SIU: Yes. I guess Rich hasn't had a chance yet to walk
you through the flow diagram. I assume you saw it.
The concept is that risk will come in in a couple of
different ways. One is more or less as a vulnerability search. At the
end of going through all the plant fire areas, all the fire protection
features, and having complied with the performance requirements, you are
still taking a risk shot at the end or taking the risk picture to see
where you are and to identify if there are any additional things that
need to be taken care of.
That kind of use of the risk assessment is very much like an
IPEEE vulnerability search and I guess the feeling is based on the
experiences from the IPEEE reviews we certainly know a number of areas
where we can say what appears to be the best way to approach an issue,
even if it is not necessarily truth. It is our best shot at identifying
vulnerabilities.
The second use of the risk assessment tools is actually at
the performance level -- let's say in lieu of doing some fire modelling,
one chooses to use a fire risk assessment to show that a particular
feature is not needed above and beyond what is required in the base fire
protection program. Again, that is something that I think Rich will
walk the whole committee through.
In that situation I think that the burden on the fire risk
assessment is greater. We probably at this moment can't specify what
are the precise tools that we should be using to address some of the
issues there, but we can certainly say what are some of the concerns and
things that have to be addressed, and again the document I believe makes
numerous references to this phrase of how the analysis has to be
acceptable to the authority having jurisdiction, so there is going to be
some give-and-take in the context of how a standard has been written for
that particular issue.
So the bottom line is we probably won't be able to specify
in great detail how the analysis should be performed just as ASME is not
specifying in great detail how, let's say, each RA should be performed
and internal, that's PRA standard, but at least we can provide some of
the functional requirements that we think need to be addressed in a
better assessment.
CHAIRMAN POWERS: The document makes reference to something
called qualitative PSA. Would you tell me what that is?
MR. SIU: As Rich said, there are discussions ongoing within
committee. Honestly, there are some differences of opinion right now as
to how we should be using risk in this.
DR. APOSTOLAKIS: We will discuss in more detail the
standard I think when Ed comes to the floor, right?
MR. SIU: Yes, although that particular --
DR. APOSTOLAKIS: I would like to pursue Dana's comment and
paraphrase it a little bit.
CHAIRMAN POWERS: I was being gentle, George.
DR. APOSTOLAKIS: No, the previous comment.
CHAIRMAN POWERS: I was being gentle then, too.
DR. APOSTOLAKIS: Dana said that questions have been raised
about the tools of PRA and so on, and you gave an answer. Questions
have also been raised, and if they have not I will be very happy to
raise them, regarding the completeness of the deterministic methods.
I would like to see a discussion in the standard, the same
number of lines as the lines you will use to discuss the incompleteness
of the PRA, discussing the incompleteness of the deterministic methods.
MR. SIU: Well, I think the standard is going to make -- is
going to describe how the analysis should be done. I don't know the
extent to which it is going to talk about the limitations in any of
these methods.
DR. APOSTOLAKIS: Well, I can assure you that there will be
something from the ACRS that will raise the issue. If not from the
whole Committee, there will be definitely individual members' comments.
[Laughter.]
DR. APOSTOLAKIS: I want to see exactly the same number of
lines devoted to the incompleteness issue of PRA and I will count them,
yes, because I don't think deterministic methods are complete and the
false assumption that this is perfection, that the new kid on the block
has to prove herself I don't buy -- that PRA has to be complete and has
to have nearly perfect methods, so I would like to see the same
criticism and evaluation raised against the so-called traditional
methods.
MR. SIU: Which you will see, I suspect, and I don't know
how far the fire modelling part of Appendix B has gone -- by the way,
Appendix B covers fire modelling as well as the fire risk assessment.
I think you will see some fairly high level requirements for
the analysis that don't get into what are the specific limitations of a
particular approach. It will say that the approach must be -- and then
name a quality -- like it must be --
DR. APOSTOLAKIS: CFAST?
MR. SIU: No, but I suspect even as we get further along, we
are not going to be specifying, for example, you must use CFAST in this
situation. It is not to that level of specificity, nor will it talk
about if you use CFAST in this situation be careful for this because
this is a problem. It just doesn't get to that level of detail.
DR. APOSTOLAKIS: Right. Right, but --
MR. SIU: So what it does say is that you may choose to
employ various methods to show that your functional requirements have
been met -- your fire protection requirements, and these are the ways
you can show that.
DR. APOSTOLAKIS: I understand that. All I am saying is you
should give equal time to traditional methods and -- for example, there
is some discussion on completeness someplace, which I can't find it now
and maybe it is inappropriate to raise it right now, but uncertainty
analysis --
CHAIRMAN POWERS: But the issue of completeness or
uncertainty certainly arises and is cited explicitly in the discussion
of uncertainties, but though people cite uncertainties in the document,
which is a very good feature of the document, one of its few good
features, that it doesn't tell me what to do. It doesn't tell me
whether I am to use the mean of my estimates to compare against the
standard mode, the median --
MR. SIU: What you are seeing -- there was actually material
that was in an earlier draft that covered exactly that --
CHAIRMAN POWERS: Well, we didn't see another draft.
MR. SIU: I understand. I am just explaining how I think
you got where you are. I understand that that material was supposed to
have been moved to an appendix, and I don't remember if it is Appendix A
or Appendix B and it might have gotten left out.
MR. MARSH: B.
MR. SIU: B -- it was supposed to be in there and if it
isn't in there, there is every intent to make sure that it is in there.
DR. APOSTOLAKIS: Are we -- okay. When are we going to
discuss the details? Not now, right?
CHAIRMAN POWERS: What is the course of presentation now?
MR. BIELEN: Just to go over a general approach on now the
standard is to be used, go over the flow chart, and then to wrap it up,
but our intention at this particular meetings wasn't, as I said, to
defend the document, to get into details.
If we need to have a more detailed presentation on the
specifics of the document in the future we can do that, and get the
right people here -- not that we don't have the right people here now,
but we can get the people who are PRA experts, for instance, here if we
need to have those questions answered, but the thing to remember is it
is still in a state of flux because this is only a draft.
We still have the proposals to go through. We still have
the comments to go through and it probably wouldn't be appropriate to --
well -- I guess you could, at some point you might want to have a
detailed discussion so you can maybe want to submit comments or
something during the appropriate time.
But I was going to say that since it's still in a state of
flux, I'm not sure what the right time is, because it's always going to
be changing.
DR. POWERS: It's always in a state of flux.
MR. MARSH: But it exists now, and it's out for some
comment, and I think the staff has comments about it. You're going to
hear some details tomorrow. I want to be cautious that we are not in
the position of defending it either. We're going to be explaining it to
you as we understand it with its -- and our concerns too.
DR. APOSTOLAKIS: Okay. So Ed then will be speaking as an
NRC staff member, not as a member of the committee.
MR. MARSH: As a member -- that's right. He's not going to
be --
DR. APOSTOLAKIS: Okay.
MR. MARSH: Defending it; he's explaining it to you.
DR. POWERS: Sometime I have to understand why this becomes
what I've characterized as the Prego spaghetti sauce of all standards.
This has something for everybody in it, and as a result, it's not clear
to me that it takes a direction that the NRC wants to go or that it
could ever be approved as a regulation. Built this way, it's just not
clear to me that it ever makes the grade as a regulation.
MR. BIELEN: Do you have specific recommendations on how you
might improve the document so that it might, which is what we'd be
looking for in the proposal period and the comment period, so we
could -- you know, if we know what your concerns are, you know,
specifics, let us know so that the committees can address it.
DR. APOSTOLAKIS: So maybe tomorrow we can hear some of
our --
DR. POWERS: Yes.
MR. MARSH: We have concerns about it too. And I want to
say something that Fred said earlier, there's a laudable effort that's
gone into this standard. It isn't perfect. It's got concerns that we
have too, but there's been a great deal of work that's gone into it,
great number of hours, very diligent folks who have been working hard at
it --
DR. POWERS: I think the people who have worked on NRC's
core capabilities will attest to you that the number of hours spent on a
document carries little weight with this committee.
MR. MARSH: Okay.
DR. APOSTOLAKIS: It's a performance-based --
[Laughter.]
DR. POWERS: I'm going to declare a recess for 15 minutes at
this point.
[Recess.]
DR. POWERS: I'd like to resume the meeting, and we are now
on to the NFPA overview.
I don't know why I have a mental block about sticking the
805 in there. It's not like the only thing that NFPA is doing.
MR. BIELEN: I think I'm going to skip over to the flow
chart, because a lot of the other overheads I have with the general
approach basically say the same thing except in words, and we maybe
follow it a little bit better, maybe not, in the flow chart. But the --
look at the flow chart. This is -- it's supposed to tell us basically
how the NFPA 805 is to be used, and I think this might generate a few
questions here.
DR. POWERS: Well, it does, and it does just from the very
top. As I have read and understood the document, which I'll be the
first to admit I have not pored over it in critical detail, but there is
what you call on the slide the minimum fire protection program, and then
one comes down and there are choices in what additional requirements you
choose to impose upon yourself. And those are what you call either the
deterministic or the performance. And then wrapped around this whole
thing is a risk evaluation, okay, which will allow you to go through and
reassess a deterministic versus performance additional requirements, but
does not include this first initial very deterministic what you call up
here the minimum fire protection.
Why was this particular structure selected?
MR. BIELEN: I don't have an answer for why it was selected.
It was kind of a structure that was generated by committee, so I don't
know if there's one person who can have a real good answer as far as why
this particular structure was --
DR. POWERS: Well, see, my question about is the risk -- you
asked the question is risk acceptable, and the answer is no, but the
only things I can change are in here. The problem may be up in my
minimum acceptable program. Or I can come down and say is risk
acceptable? I may find myself down in the minuscule noise of risk, and
the reason I'm there is because of this minimum program may be far more
than what I need.
I mean, it seems to me that if I'm going to use risk to
adjudicate on what I need to do, I cannot box myself into the one-way
street Professor Apostolakis frequently mentions, which is I can -- I
only use risk here to augment the requirements that I choose to impose
on myself.
MR. WARNICK: I'm Lee Warnick with Virginia Power. I'll
make a feeble attempt at answering that question, but as far as the
establishment of the minimum fire protection program, the intent there
is that there are certain things, certain elements of a fire protection
program which are going to be basic to every nuclear plant regardless of
the risk or the degree of risk in any specific area.
The best example is the fire brigade. Every plant has to
have a fire brigade. It's the committee's opinion that every plant has
to have a fire brigade regardless of your calculation of risk in any
specific area of the plant. And so it was intended to establish those
type of requirements that fell into that box as a separate category of
the standard. And then you move on beyond that to establishing specific
requirements based on the hazard in any particular area of the plant,
based either deterministically or risk-based.
DR. POWERS: I think you've chosen a particularly
appropriate example in the fire brigade. Your minimum program said you
all have things that deal with prevention of fire, suppression and
detection of fires, and mitigation of damage from fires. Had that made
up the elements of your minimum program, I think I probably wouldn't
have objected. I can probably find people that would object, but I
wouldn't.
When you come to the fire brigade, that is a means to
achieve an end. It is not the end in itself. I mean, the objective of
a fire-protection program is not to assure employment of
fire-suppression personnel. I can perfectly well imagine a plant who
says my costs come from people, and having five fire-protection people
on each one of my shifts, what, a typical plant has five shifts, is a
big expense to me. I can get rid of that expense by enhancing my
automatic suppression systems. I can have a super whammy dammy
fire-suppression systems that are just the cat's meow, and I don't have
to have all those people. That should be an option open to him.
MR. WARNICK: One other comment as far as the administrative
controls are another category within that base program, certainly that
would be common. As far as the fire brigade, all I can say is again
it's the sense of the committee that the brigade should fall under that
minimum requirement.
DR. APOSTOLAKIS: Now -- are you done, Dana?
DR. POWERS: No, but I will pause to take a breath.
DR. APOSTOLAKIS: Is the committee that's putting this
together familiar with the work that the NRC is doing on inspections,
the cornerstones and all that data that has been developed? You
probably cannot get the document itself, but the viewgraphs at least for
the presentations that show the objectives, the goals of the inspection
program. These are very good, in fact, and as I started reading this
document, chapter 1, page 3, it has a statement of the goals. And these
are not quite consistent with the cornerstones that that document has.
So the first question is whether this agency -- I mean, this
is NFPA, but eventually we want to adopt this, we can have different
goals in different documents.
Second, it talks about radioactive release goal and talks
about a release that will result in adverse effects to the public, plant
personnel, or the environment. Now is protecting the environment part
of the regulations?
DR. POWERS: Yes, it is.
DR. APOSTOLAKIS: Where?
DR. KRESS: It's down in the regulations --
DR. POWERS: GDC 4?
DR. KRESS: That have dose acceptance requirements in them.
DR. POWERS: Well, I'm --
DR. APOSTOLAKIS: That's the environment? I thought it was
for people?
DR. KRESS: It's the same. You can view that as protecting
the environment, because the doses are so low that they have no adverse
effect.
DR. POWERS: I am informed that appeal to GDC 4 is
appropriate for the environment.
DR. APOSTOLAKIS: Which says?
DR. SINGH: That's a general area.
DR. POWERS: I can look up what the general design criterion
is for you.
It's Appendix A.
DR. APOSTOLAKIS: But anyway the goals are -- and then the
objectives -- I think they should be consistent with the other
activities in the Agency, but also within the document there are
inconsistencies. For example, it's not clear to me how one decides
which equipment to protect from the goals. We really need a top-down
approach that goes from the top goals that are stated here, protect the
environment, to show schematically how you decide which equipment to
protect.
The way it is now, and maybe I'm wrong, the way I read it is
that the determination of which equipment to protect is something else
that somebody else decides again as an additional requirement, and that
goes back now to the current deterministic approach where, you know,
safety-related equipment are supposed to be protected. But how these
equipment relate to the nuclear safety goal of providing reasonable
assurance that the fire will not result in nuclear fuel damage or will
prevent radioactive releases is not clear.
I think the top-down approach, the power of the top-down
approach has been underestimated by a lot of people, and this program
that the NRC has to put together an inspection program does a very good
job using the top-down approach, starts with the top goals, goes down to
the cornerstones, and then goes all the way down using fault-tree type
logic to what they want to inspect, what the performance indicators will
be, what is it that's not covered by the performance indicators,
therefore there has to be a programmatic inspection and so on. And I
think that's great, and I think we should try -- now I don't know to
what extent your committee can become familiar with that before the
staff publishes that report, but since you have NRC staff on your
committee, they can certainly read it.
But that's something that struck me as needing help here, to
structure this whole thing so we start with top-level goals, and then
when I go to page 35 I will be able to understand why there is an
attempt to protect a certain area or piece of equipment. That's missing
now.
MR. SINOPOLI: This is Cliff Sinopoli with Baltimore Gas &
Electric. I think we've tried to do that, and apparently
unsuccessfully. I know the format that we've used, the NFPA has come up
with a primer. I think there's a number 3 now. And that's what we
tried to follow. And I think -- I mean, I don't think the committee
would have any objection to trying to make it as consistent with the NRC
approach as possible. Again, though, provided that we can try to
maintain the NFPA framework to the best, we need to marry those two.
DR. APOSTOLAKIS: Okay. Now the other thing that really
confused me is the distinction between deterministic requirements and
performance-based requirements. So if I go to page 29 that talks about
water supplies, part of the deterministic requirements is the
fire-protection water supply shall not be less than two separate
300,000-gallon supplies.
Okay, makes sense, deterministic.
Then I go to performance-based. An adequate fire protection
water supply shall be provided in all areas of the plant as determined
to be necessary by the FAHA. So that's performance? Having an adequate
fire-protection water supply is a performance-based requirement? I'm
utterly confused. It sounds to me like a deterministic requirement.
And then the second performance-based requirement, an
untreated raw water source may be used as a fire-protection water supply
when both quantity and quality can be shown to be acceptable. Is that
what we mean by performance, Dr. Kress?
DR. KRESS: You certainly don't want to ask Dr. Kress what
he --
[Laughter.]
Can't we do this in private?
DR. POWERS: I would say -- I'm surprised, Tom, because I
would have thought you would have said yes, that's exactly what a
performance-based requirement is --
DR. KRESS: Yes, yes, among other things.
DR. POWERS: Performance-based rules either lead you to
determinism or risk --
DR. KRESS: You said it very well.
DR. POWERS: And there is no middle ground for you. Except
here's proof that a middle ground exists. It's both.
DR. APOSTOLAKIS: Another performance-based requirement. If
fire pumps are powered by electric motors, the electrical supply shall
be independent such that a fire involving one electrical supply and a
simultaneous failure of any one fire pump shall not prevent delivery of
the required fire flow.
I would not call that a performance-based requirement.
Where is the performance? That's a design requirement.
MR. SIU: Yes. Without commenting on the last one, George,
I think there is a difference -- I think there's a difference in
interpretation of what performance-based means between how NFPA
approaches it and how that term has been used here.
DR. APOSTOLAKIS: And I looked it up.
MR. SIU: Yes.
DR. APOSTOLAKIS: The Random House dictionary.
MR. SIU: Now wait, wait, wait, wait, wait.
[Laughter.]
DR. KRESS: That doesn't work.
DR. APOSTOLAKIS: That doesn't work?
MR. SIU: This is the NFPA --
DR. APOSTOLAKIS: It worked beautifully.
DR. KRESS: I don't think the words mean anything.
DR. APOSTOLAKIS: They have --
DR. KRESS: They define words. They don't define the fine
concepts.
DR. APOSTOLAKIS: A lot of them refer to the dramatic arts.
We understand that. Then they have one that comes the closest to what I
think performance is. The manner in which or the efficiency with which
something reacts or fulfills its intended purpose.
That's actually pretty good. That's a performance. But to
say that an adequate fire-protection water supply shall be provided,
that's not performance, it's design.
MR. SIU: Rich, correct me if I'm wrong, but my
understanding of how performance-based requirements are being developed
for nonnuclear applications in the fire-protection industry is basically
they're talking about the use of codes or models to predict something
like temperature, like smoke, and so forth. It's not a measurement of
something that you will observe down the road. Hopefully you'll never
have that fire that leads obviously to that.
DR. APOSTOLAKIS: And I agree with that.
MR. SIU: Okay.
DR. APOSTOLAKIS: If you predict the behavior of smoke --
MR. SIU: Um-hum.
DR. APOSTOLAKIS: And you show, for example, that it remains
under certain conditions below a value, a pressure --
MR. SIU: Yes.
DR. APOSTOLAKIS: That's performance, yes, because you're
dealing with the manner in which the smoke fulfills its intended
purpose. In other words, being below -- but if you say I want a water
supply, I don't see any performance. I mean, you design it and you have
it, you build it.
MR. SIU: Well, but you have to have an adequate supply.
DR. APOSTOLAKIS: So when you inspect it then you make sure
that the water level is --
MR. SIU: Or your model, whatever -- the design calculations
that you use to show that it's adequate.
MR. SINOPOLI: I think in the past we've always been told
you're going to have 300,000 gallons. That's what you've always been
told in nuclear energy. What we're saying here is we're not telling you
anymore how much you have to have. What we're saying is your fire
hazards analysis finds those areas where you see that you need to have
some pressure or whatever, you're going to fight a fire that occurs,
whether it's manual fire suppression, what your flow to the hoses are
going to be, to the sprinkler system, whatever. And you design your
supply based on that requirement for the water, as opposed to the past
where it said here's -- you have to have 500,000 gallons or 300,000
gallons or whatever it was.
DR. APOSTOLAKIS: So performance then means that you don't
specify numbers.
DR. POWERS: Well, what it is is, I think the question is
not whether they specify numbers but whether they specify function.
That is, why would not one write a performance-based regulation that
says you can put out a fire before it does any damage. I mean, maybe
you'd write it somewhat differently.
MR. SINOPOLI: I think that's the intent of the --
DR. POWERS: And why would you care whether the guy put out
the fire with water or with CO2, with appeals to the Great Water Demons?
I mean, you really wouldn't care, as long as he can do that and show
that he can do that.
DR. APOSTOLAKIS: Well, that would be performance.
DR. POWERS: That would be performance.
DR. APOSTOLAKIS: Putting it out before it does damage is
performance. In other words, in my mind, performance has an element of
time in it. Something is happening in time and stays below or above a
certain level. Then I say yes, it performed well. But to say that the
electrical supplies of the pumps must be independent of the fire is not
performance. It's design, isn't it?
DR. POWERS: I mean, it seems to me that that's part of --
that particular requirement sounds like it's straight out of a standard
review plan.
DR. APOSTOLAKIS: It's design, though.
DR. POWERS: Yes. I mean, it's not performance.
DR. KRESS: If one uses the top-down approach, say the
top-down performance objective is to be able to put out the fire before
it does any damage, one has a lot of trouble figuring out what that
actually means. One can say well, in a performance world, we know if we
can do this, this, this, and this, like supply enough water, like have
enough detection, have enough sprays and so forth, we know that if we
have enough of these, we have satisfied this objective. So let's put
our performance measures down here on things we can look at. In that
sense it is performance.
DR. APOSTOLAKIS: That's a prescriptive approach.
DR. KRESS: Not necessarily.
DR. APOSTOLAKIS: So the current regulations are
performance-based.
DR. KRESS: Not necessarily.
DR. APOSTOLAKIS: Well, that's what they tell you to do.
DR. KRESS: No, I maintain that they know --
DR. APOSTOLAKIS: Twenty field separation --
DR. KRESS: That if you look at performance-based
regulations in the current ones that they're almost --
DR. APOSTOLAKIS: Isn't Appendix R performance based?
DR. KRESS: Based on different forms, though.
DR. APOSTOLAKIS: Appendix R is performance-based. It says
that you have to have a minimum of 20 feet empty space separation,
right? How is that different from saying that the electrical supplies
to the pumps must be independent of the fire or that you have to have an
adequate fire-protection water supply? I go there and I measure it. I
go there and I look at the water.
Now, if I had a model that would predict the time to
suppress the fire, then that would be my performance measure, and the
other stuff down here I wouldn't care about. But I don't have that.
Therefore, I abandon the performance-based approach and I go to the
prescriptive approach, which says what you said. I suspect that if I
have enough water and trained people, they're going to put it out in
time. But that's not performance, that's prescriptive, because of lack
of model to calculate that time, to estimate that time.
DR. KRESS: Well, it gets to be performance if you don't
actually start describing distances and length of hoses and say you must
have enough separation or you must have enough water to let the
conditions determine how much that is or let the licensee's calculations
and plant configuration determine --
DR. APOSTOLAKIS: I think if I compare the similar
requirements that run the deterministic and performance-based, there is
a requirement that's called deterministic that says each pump and its
driver and control shall be located in a room separated from the
remaining fire pumps and from the rest of the plant by fire barriers
with a minimum rating of three hours. So they specify the rating.
In the performance-based part, it just says they should be
independent from the fire and the simultaneous failure of any other fire
pump. In other words, what is not specified is that there should be
fire barriers with a minimum rating of three hours.
Now is that the distinction between deterministic and
performance-based, simply that you don't specify the fire rating of the
barrier? That's news to me.
DR. KRESS: I don't -- you know, I once made the comment
that there's very little difference, and that in fact, when you go to
performance-based regulation, all you're doing is proliferating all the
bad things that you already have, and just relax --
DR. APOSTOLAKIS: Well, yes. Yes.
DR. KRESS: And that's a prime example of it. You're just
doing the same thing over again in a different way.
DR. APOSTOLAKIS: Well, if you stretch the definition of
performance.
DR. KRESS: Yes, which you almost have to, because the only
reason you have that kind of performance-based rule is because you can't
do a risk-based one. If you could do a risk-based one, it would all
make sense.
DR. POWERS: I'm going to ask that we let the speaker move
on. I think the point, if not elaborated upon, has been made.
DR. KRESS: Let me ask another question about his viewgraph.
The triangle where it says is risk acceptable, is that fire risk or is
that plant risk?
DR. APOSTOLAKIS: Yes, that's what I mean, fire.
DR. KRESS: Fire risk?
DR. APOSTOLAKIS: It says somewhere in the standard, that
the licensee should specify the fraction of the total risk --
DR. KRESS: Due to fire.
DR. APOSTOLAKIS: Due to fire.
DR. KRESS: And that's what was in --
DR. APOSTOLAKIS: Which I don't understand how they're going
to do and why they should --
DR. KRESS: That's why I asked --
DR. POWERS: And why they should have to.
DR. APOSTOLAKIS: Huh?
DR. POWERS: And why they should have to.
DR. APOSTOLAKIS: And why they should have to.
DR. KRESS: That's why I asked.
MR. BIELEN: Well, I don't know if I have the expert -- I
don't have the expertise to debate why we did things, why we didn't, and
all the details of how we did it, except I can try to explain what's in
there, which you can see already, and probably have gone through it
already.
But again I do encourage you to submit proposals so that the
committee can hear your concerns.
DR. APOSTOLAKIS: And in terms of communicating to the
public -- and I'm a member of the public -- when I see deterministic
approach, no further analysis required, a performance-based-approach
engineer, I'm turned off. Because that again shows deterministic
approach is nice and good and no further analysis required. The other
stuff requires additional work. I don't know whose idea it was to put
it up there, but it seems to me when you take the deterministic
approach, don't you have to do any engineering analysis someplace?
Sure.
MR. BIELEN: Well, the NFPA process is a dual-track
approach -- I don't know if you remember that terminology that we used
before -- where you can go down a what we normally call prescriptive,
but I guess in this industry we call it deterministic, versus
performance-based. So that's why there's kind of that split there with
deterministic and performance-based. And I hear you as far as your
concerns go that -- about the engineering analysis for deterministic.
DR. APOSTOLAKIS: Yes, I mean, this no further analysis
required, I mean, there's a flag out there that says hey, do it this
way.
MR. BIELEN: Well, there is further analysis, because you've
got to go down to doing a sitewide risk evaluation.
DR. APOSTOLAKIS: In either way, though. I mean, whether
you do it deterministically or performance-based in that box you still
have to do the sitewide risk evaluation, which is a good thing.
And I noticed, by the way, that the title of this does not
have the words "risk informed."
MR. BIELEN: Right.
DR. APOSTOLAKIS: It's only performance-based.
MR. BIELEN: Right. But the risk is kind of built in there
when you look at if the performance-based approach, risk-based methods
are encompassed in there, and then also in the sitewide risk evaluation
and is risk acceptable, that's where we use risk methods.
DR. APOSTOLAKIS: So why not then put it in the title,
performance-based, comma -- no, is it comma? Risk-informed.
MR. BIELEN: Yes.
DR. APOSTOLAKIS: Yes, is an official position on that.
DR. KRESS: Comma, yes.
DR. APOSTOLAKIS: Comma, yes.
MR. BIELEN: I hope to see a proposal from you stating that.
[Laughter.]
If we kind of move down through here, as Dana started with
the first box with the minimum fire-protection program, that's correct,
chapter 3 kind of is your base fire-protection program, here's what you
have to have as far as a minimum goes. And then there's some additional
requirements.
DR. APOSTOLAKIS: Now, the minimum, is there a rational way
of determining the minimum? You mentioned, for example, the fire
brigade, and I would say yes, that makes perfect sense to me. Is the
requirement that the hose be 75 feet part of the minimum?
MR. SINOPOLI: No, that's -- what we considered the minimum
was primarily those things such as your fire prevention. For example,
combustible control, hot work permits, things that we've learned over
the years every construction site should have if you're doing hot work.
The basic sort of what I call the first leg of the defense in depth.
And, I mean, that's just something that we've learned over the years.
DR. APOSTOLAKIS: So it's a fairly high level.
DR. MILLER: That's just independent of whether it's a
nuclear plant or just any facility.
DR. APOSTOLAKIS: Well, is that true?
DR. MILLER: Well, yes, I suppose it --
DR. APOSTOLAKIS: You don't have a fire brigade in a
building, do you, an office building?
DR. MILLER: Any industrial facility where fire may be an
issue.
MR. WARNICK: Well, and in fact I think the fire brigade is
actually in the next --
DR. APOSTOLAKIS: Do all chemical process facilities have
fire brigades? Probably not.
MR. WEST: No, they probably do.
DR. APOSTOLAKIS: They do?
MR. SINOPOLI: Some level of trained people. Even the
fossil plants typically have some level of trained people. They may not
be under the fire brigade. There's different categories -- or like we
use interior structural fire fighters, you know, the fire extinguisher
person.
DR. APOSTOLAKIS: So the minimum program then is the
accepted collective wisdom of the fire protection brigade.
MR. SINOPOLI: Remember, this is a committee, because you
have insurance people on there who have certain aspects, because they
want the warehouse protected, you know, they see valuable assets there
as well, you know, the staff we knew.
And the other problem with the five-man fire brigade, for
example, why did you pick five? And we struggled with that a little bit
too. But right now with OSHA is you're going to have a fire brigade of
the type we have, you have to have two in, two out, which means if two
guys go in with a hose, you have to have two guys back out to back them
up, with one person left for a brigade leader. So unfortunately that
number is going to be pretty much automatic. We picked out a minimum of
five just because of other requirements. You can't get away from OSHA
requirements, for example.
DR. APOSTOLAKIS: So is it fair to assume then that you're
selecting performance goals and objectives because this is a nuclear
plant? That's where the fact that it's a nuclear plant --
MR. SINOPOLI: I'm still up in the first box.
DR. APOSTOLAKIS: No, but the first box, as Professor Miller
said, would apply to any industrial facility. This is the minimum
requirements for any industrial facility. Then the additional
requirements and goals and so on in the second box come because this is
a nuclear plant, so you want to protect the fuel and so on.
MR. SINOPOLI: In most cases I think there's probably a
couple exceptions. First of all, life safety. Any facility is going to
have to have certain life safety requirements.
DR. APOSTOLAKIS: Right.
MR. SINOPOLI: Okay. And the only thing we did a little bit
differently here is that typically, you know, life safety requirement
situation, you want people to get out. Here we recognize there's going
to be some individuals who may have to maintain their positions or make
entries into certain areas, so we adopted that in here. And also,
business interruption. Most facilities have some sort of, whether it's
a fossil plant, nuclear plant, chemical plant, they have some sort of
contingency plans or some sort of level of risk they're willing to
accept, you know, a million dollars', two million dollars' worth of loss
or a certain amount of downtime. And that was put in here as well,
which I don't think is strictly nuclear. But obviously the first --
DR. POWERS: At the risk of provoking my colleague even
further, or maybe the hope, hydrants with individual hose gate valves
shall be installed approximately every 250 feet apart on the yard main
stream.
Provided at intervals of no more than 350 meters along the
yard line mainstream. These are among those minimal requirements.
These are very prescriptive things where, another one, weekly valve
inspections, why one week?
DR. APOSTOLAKIS: Exactly.
DR. POWERS: Eight days is unacceptable, six is too often.
I mean --
DR. APOSTOLAKIS: And that is something that really can be
analyzed using risk analysis.
DR. POWERS: Yes. Yes. Yes, these are ones that can be
answered.
DR. APOSTOLAKIS: We have a whole risk-informed standard now
-- I mean Regulatory Guide.
DR. MILLER: Weekly was probably established years ago by
the --
DR. APOSTOLAKIS: Yeah, that's what --
DR. POWERS: Yes, but that is not an excuse for including it
here.
DR. APOSTOLAKIS: See, that is the whole point with the
redundant trains of safety systems, that years ago somebody said once a
month, it became the Bible.
DR. POWERS: That's fine.
DR. APOSTOLAKIS: And now with risk-informed, one expects
one can change that, like some of the diesels have gone up from seven
days to 21 and so on. So even the minimum requirements then can be
questioned, that is what you are --
DR. POWERS: No, I think they can't. I do think we need to
move along, though.
DR. APOSTOLAKIS: Sure.
MR. BIELEN: Okay. Moving off of box one.
DR. APOSTOLAKIS: I think we covered that. We really don't
want to be discuss that anymore.
MR. BIELEN: Does everyone have a general feel for how the
standard is working better now?
[Laughter.]
DR. APOSTOLAKIS: Is there anything else you would like to
tell us?
MR. BIELEN: We can probably move on to the next one.
DR. POWERS: You can be secure that the Subcommittee members
have read the document.
MR. BIELEN: Okay. A lot of the other overheads I have are
kind of just word explanations for the flow chart, which I think we have
gone over, and I will just go up to the conclusions.
I would like to state that the project is on schedule.
Remember, it is the annual 2000 cycle. We have a public proposal
closing date of February 19th. We are going to be meeting -- have to
ROP meetings, so the project is going on schedule according to the NFPA
process.
The Committee still has to review the ROP and ROC, or still
has an ROP and ROC period to make changes to address concerns. I just
want to emphasize that we are still at the very beginning stages of the
process. We just have the draft out. We still have the ROP and the ROC
period to go through.
So, again, I know there seemed to be a lot of questions and
concerns, and, like I said, we are still early on in the process, and I
would like to get, as specific as we can, some input from the Committee
as far as what we need to address and how to address it, which really
addresses point number 3, is all public input proposals and comments
must be addressed by the Committee.
So just by saying I have got a problem here, I have got a
problem there, you should address this, it doesn't mean it is going to
get done. What we really need to do is put this in as a formal process,
a formal proposal or a formal comment if we are that stage, and then the
Committee has to address in some way. So that is what I would like to
finish up is, any input we do have, do try to get it in on a public
proposal by the February 19th deadline for the Committee to review.
DR. POWERS: I will comment that there has been information
discussions among members of the Subcommittee where we would like to put
together a set of comments. We are struggling now as to how to bring it
-- to follow our own process, which means we have to -- we can propose
this to the full Committee and they are the ones that actually write the
ACRS positions. And we have to figure out a mechanism for doing that.
We think we have hit upon one. And we will be -- we find ourselves in
camera for the next two-and-a-half weeks or something like that, so
there should be adequate time for us to work out the details of that.
DR. APOSTOLAKIS: By the way, what is A 2000?
MR. BIELEN: Annual 2000.
DR. APOSTOLAKIS: What does that mean?
MR. BIELEN: The NFPA has two cycles, an annual meeting and
a fall meeting. So this document is up for the annual meeting of 2000,
which is May of 2000.
DR. APOSTOLAKIS: Okay.
DR. POWERS: Let me ask you a question about the
philosophies behind these consensus standards a little bit. No, it is
not -- I am not going to go into that. I promised you I wouldn't go
into that. I see we have your -- among the people developing these
standards, those involved in the regulatory process that are
knowledgeable about regulatory law, we have those who the licensees and
are the victims of regulatory law, and we have another class of people
who are the insurers who have a knowledge about standards and
qualitative perceptions of risk.
DR. MILLER: And they also vendors of equipment.
DR. POWERS: And vendors perhaps.
DR. MILLER: I don't know if this one has it.
DR. POWERS: But it seems to me that, in looking to take a
standard into the de jour, that you have a problem, that you are
carrying along baggage that is imposed -- you have the possibility to
carry along baggage that is imposed for financial rather than safety
reasons. Is that a concern we should have here in looking to some day
take this standard and endorse it as a part of -- as an option within
the NRC regulations?
MR. BIELEN: Yeah, that concern comes up with actually any
standard because of the make-up of the Committee.
DR. POWERS: It is not yours in particular, it is anybody's,
yes.
MR. BIELEN: Or any of the NFPA ones. But because of the
way we are set up, we have to have two-thirds agreement by the Committee
in order for it to go forward. So we try to have balanced Committees by
having no more than one-third of any particular interest group on the
Committee, so the Committees basically can't be dominated by just the
users, or just the enforcers, or just the manufacturers, or just the
whoevers. So you really have to have two-thirds agreement by the
Committee in order for it to go forward, so that is how we have
addressed that concern, by limiting the membership, balancing the
Committee, and then having the two-thirds agreement.
DR. POWERS: My concern is not balance, I mean I, as an
insurer, said, gee, I want a standard out there that I know if the guys
comply with that, that I -- they have minimal risk of fire damage that I
have to pay for, and there are a lot of things I want to have. And so I
work very hard and persuasively and get those in there. But when we
come from a legal basis that says, no, we don't worry about that sort of
thing, what we worry about is protecting the public health and safety,
we are imposing things, especially when we have prescriptive elements.
We are imposing things upon the licensees that are none of
our business to be imposing. We have no legal basis to impose those, if
one were to look at it in microscopic detail, and that is what concerns
me a little bit about taking this and moving it into -- as an
alternative to the regulations. It is less concern to me if we say this
is a Reg. Guide, this is a way to meet the regulations, but to say it is
an alternative to the existing regulations, I think you run into a
problem there. It is an expression that has no answer, it is just a
question.
DR. MILLER: Well, there is certainly precedence for that
happening, though.
DR. POWERS: Steve has a question.
MR. WEST: I have a question? I have an answer.
MR. BIELEN: Steve has an answer.
DR. POWERS: Good.
MR. WEST: Well, I think that is really a question that is
more appropriate to ask the staff, that we will need to address through
the rulemaking process. I think, generally, the NFPA as an
organization, in their charter mission, they are concerned about
property, protection of property damage.
DR. POWERS: Oh, yes.
MR. WEST: I don't think not considering that in the
standard is really an option for them. Life safety, too, it is kind of
part of their foundation, is to --
DR. POWERS: It was not meant as a criticism, it was meant
as a question.
MR. WEST: Right. I think through the -- if the staff finds
-- and this is a question that has been raised by people in industry
also. We were at an NEI forum earlier, or late last year, and a lot of
the participants there were questioning having property protection
criteria and life safety criteria in the standard and what it meant in
the regulatory space.
But I think through the rulemaking process, we don't have to
necessarily endorse the standard, the entire standard, we can endorse
parts of it, or endorse it in part, or only those parts that we care
about for regulatory purposes. But there may be other authorities
having jurisdiction out there that may be interested in the property
protection or the life safety aspects and they may impose them through
their own regulatory vehicles.
DR. POWERS: Good. Sounds points all.
DR. APOSTOLAKIS: Why is the NFPA doing this? Who started
the process? The industry or --
MR. BIELEN: Well, the NFPA has been involved with fire
protection in nuclear facilities, I am trying to think of -- I think in
the early '70s when we first started developing standards for the
nuclear facilities. 805 kind of developed as a need, from what we have
been hearing, through Committee members, through industry, but,
basically, it was NRC, I think, from what I remember from the early
discussion, was looking for a rulemaking and performance-based,
risk-informed and the Committee said we can do that.
MR. MARSH: I think we wrote you a letter saying we needed
this to be a component in the NFPA 805 in order to gain Commission
approval.
DR. APOSTOLAKIS: But I remember a letter to the Commission
from the NFPA saying that we hear that you guys are about to do
something and we are already doing this so why don't you join forces.
So that sounds like --
MR. WEST: You need to get to a microphone.
George, there was also, there was a public law that said
that federal agencies should, to the extent they could, rely on the
consensus standard process instead of rulemaking and we did --
DR. APOSTOLAKIS: I understand. But there was this letter,
though, from the NFPA to the Chairman --
MR. WEST: Right. When we --
DR. APOSTOLAKIS: -- saying that we are already doing this
and why don't you guys join. So it doesn't sound like the NRC was the
one that initiated it.
MR. WEST: Well, there was a negotiation. There was the NRC
was looking at options for creating a performance-based, risk-informed
rule or Reg. Guide for use by the industry. And in parallel with that,
the NFPA was thinking of creating NFPA 805. But my sense, my memory is
that they had not specifically -- and they came and talked to you --
MR. MARSH: Right.
MR. WEST: -- and you guys raised this question, too. They
did not specifically plan to include risk in the standard. So the
letter, the letter from NRR back to NFPA was, yes, we are interested in
utilizing the NFPA consensus standard process for this new regulation,
but if we are going to -- if you are going to do that, you have to
understand what we need, and you are going to have to agree to include
in your process our needs.
DR. APOSTOLAKIS: Okay.
MR. WEST: And then that's when we introduced the risk
aspects to the standard.
DR. POWERS: I think we have -- have we covered your
presentation?
MR. BIELEN: Believe me, I am done.
[Laughter.]
MR. BIELEN: Well done.
DR. APOSTOLAKIS: Very good.
DR. POWERS: Well, let me assure you that we very much
appreciate your coming here. You may not -- you may never come back
again, but I hope you do.
DR. MILLER: Every year, we want you to come back.
DR. POWERS: Thank you very much.
MR. BIELEN: Okay. Thank you.
MR. WEST: Dr. Powers, what is going to be fun is that Ed
Connell has missed a lot of this, and he is going to be giving his
presentation tomorrow, so.
DR. APOSTOLAKIS: Yes, I am not really sure, though, there
is any purpose to that presentation. Oh, no, he is presenting us --
MR. WEST: He is going to make a --
DR. APOSTOLAKIS: Your use of the 805.
MR. WEST: Right.
DR. APOSTOLAKIS: That should be interesting. That should
be interesting.
MR. WEST: Without the benefit of these comments, though.
DR. APOSTOLAKIS: Well, that is his problem. The room was
open to all.
MR. WEST: I said it should be fun.
DR. POWERS: It will not be the first time we have set Ed
up.
DR. APOSTOLAKIS: So is this the end of the day, Dr. Powers?
DR. POWERS: No. We are going to go back to my favorite
subject, which is hot shorts.
DR. APOSTOLAKIS: And why are we going back, did I miss
something?.
DR. POWERS: Yes. You have missed, or not had the privilege
of a hearing from the BWR Owners Group, who is a major player in this
whole area of the hot shorts and the hot shorts resolution effort.
And I will have to say that, to introduce this, that I am
gratified to see the spirit of resolution that so permeates some of
these activities now in this area, rather than arguing simply over
language and the like. With that, the floor is yours.
MR. BACANSKAS: Good afternoon, my name is Vince Bacanskas.
I am not Greek, I am Lithuanian.
DR. POWERS: Well, you are very foolish, you should have to
claimed to be Greek to begin.
MR. BACANSKAS: And there are two words missing from my
presentation, one is "risk" and the other is "performance", and that is
by intent. I am the Appendix R Committee Co-Chairman. We have two
utility chairmen on this group and let me get right into it.
To give you a little background, we formed this Subcommittee
in 1997 in response to, first, some issues that came out of the early
FPFIs, which the first two were BWRs, and some issues specific to BWRs
came out. I also wanted to mention to you that this was charted as a
generic committee within the Owners Group. What that means is that all
BWRs are full participants and have full voting rights within this
Committee. You have had other groups speak to you who are cafeteria
committees, or special interest committees such as the Mark I Owners, or
some of the smaller groups. This is all BWRs involved.
What we are trying to do is -- and you heard earlier this
morning from Ed, we have, I think, seven volumes of documents, guidance
documents, on implementation of safe shutdown analysis. We are trying
to gather all of that, read through it, understand it, and put it into a
format where it can be used as more or less a cookbook approach, a
generic approach, for performing a safe shutdown analysis.
So you are aware, we have a very explicit charter from our
executives that we will not do this in a vacuum. We have participants
on both our Steering Committee and our Subcommittee who are active
members of the Fire Protection Working Group at NEI, who are members of
the NEI Circuit Failures Issues Task Force. As Fred generates his, I
believe two new tasks forces, we will probably have membership there as
well, or members of our Committee, who are also serving on his.
And we speak frequently with some of the industry
representatives of the NFPA 805 Committee. And I want to state that
very clearly, we have been discussing a directed vote member for the
Owners Group on the 805 Committee. We are supposed to be in final
discussions on that, but at this point we are coordinating and
discussing issues with members from utilities of the Committee. So we
have no direct relationship with the NFPA 805 Committee.
We have had three outstanding meetings with the NRC staff,
which I would like to thank them publicly for today. At each of these
meetings we have had open discussion. We have brought items to the
table that we felt were our understandings of issues. The staff went
through in detail with us their history on that specific issue. We have
looked at solutions to provide a high level of safety, and have taken
the NRC's input and gone back and looked at our work again. And I think
we have opened their eyes to some things that -- or perspectives that
they did not see, and how we arrived at those perspectives.
As Steve said earlier this morning, what we are attempting
to do is to provide a licensing topical report, a GE -- I guess it would
be a NEDO document, or an NEDC document, to the staff with the explicit
methodology and positions for review by the staff, and request of a
safety evaluation report and approval of our methodology.
DR. APOSTOLAKIS: So, your methodology -- I mean let's say
that 805 is cleaned up and it is accepted. Your methodology will go
beyond that and adapt it to BWRs, or you might end up saying, gee, we
like 805, so this is it.
MR. BACANSKAS: I think that as we get into it a little
further along, you will see that what the Owners Group is doing at this
point would parallel more the Reg. Guide effort than anything else. We
all have different licensing bases, different SERs, different
idiosyncracies in our plant design, and we are trying to establish a
common methodology that envelopes as much of the existing, as it is
today, with the appropriate basis stated, and the source documentation
from the regulations, so that we can walk through and understand how did
we determine that this is adequate and safe.
DR. APOSTOLAKIS: Okay.
MR. BACANSKAS: We have developed an outline of the
document, which I will run through quickly for you, and identify those
areas where we will take what we call positions. And I guess to explain
a position, it is more or less an interpretation of the existing
guidance and how it would be applied.
We have gone through, as I said, seven volumes of regulatory
information, SECY letters, Commission briefs, Information Notices,
Generic Letters, the entire gamut, and have reviewed SERs, reviewed and
catalogued all of this information, and we are trying to put it together
in a cohesive story so that someone ten years from now can understand
how we got where we are, which seems to be one of the biggest problems
that exist today, that if we look back -- how did they do it ten years
ago? -- we are not sure how we got where we are today.
DR. POWERS: Since that kind of objective seems to parallel
very closely one of the objectives the staff has, is there a point where
you think you -- you anticipate cross-checking your two conclusions? I
would not say reconciling your conclusions, because it would not
surprise me at all if you were at odds here and there. But will there
be a cross-checking?
MR. BACANSKAS: We are doing through our meetings. I don't
know that they are holding up the draft Reg. Guide and saying this is
what we have written. But we are bringing forth, as we meet with them,
the detailed written positions, making presentations to the staff, and
then discussing them. And it is nice to know they haven't thrown us out
yet.
DR. POWERS: I think they probably salivate every time you
come because you probably make their job a little easier.
MR. BACANSKAS: Well, I think it is -- as I said earlier, it
has been a very, very informative process, and I think a very positive
process for both industry and the staff, because we are free to discuss
these things in a non- -- even though it is a regulatory environment, it
is a non-regulatory environment because there is no specific license
action on the table.
DR. POWERS: I notice that the workshop in July, at least
associated with the BWR Owners Group, there were a set of very specific
questions on spurious operations. Some interesting questions,
situations, that asked for what the NRC position was. Did you get an
answer to those questions?
MR. BACANSKAS: On the July workshop? From the Owners
Group, --
DR. POWERS: I just assumed it was from the Owners Group.
MR. BACANSKAS: No, we did discuss some of this in detail at
our December 2nd and 3rd meeting. We are -- I believe Steve left with
an action that he was going to give us certain bounds on what they have
seen, and that we would continue to develop our position and discuss it
with them further.
MR. WEST: Does that answer your question? I mean during
the workshop, we had question and answer sessions set up and we did take
questions from the floor, and we responded to them.
DR. POWERS: Well, I will just --
MR. WEST: I don't know, I don't remember anything from the
Owners.
DR. POWERS: I will read the start of a question, maybe it
will trigger somebody's memory on that. Assume that the spurious start
of an ECS -- ECCS pump, such as the RHR pump A and a boiling water
reactor is a concern. Further assume the spurious start of the pump
requires three short circuits to occur on separate conductors. The
separation analysis show the control cables associated with these
conductors may be affected due to a common fire. Is the starting the
pump considered to be a valid spurious operation that needs to be
addressed?
There is a lot more to the question, but it ends up -- What
is the NRC position on this?
MR. WEST: Yeah, we answered that one.
DR. POWERS: Okay. What was the answer?
MR. WEST: Yes.
DR. POWERS: Okay.
MR. WEST: And I mean that is kind of also an example of the
level -- some of the level of discussion we were having in the meetings
with the Owners Group. We get into some specific examples like that,
looking at wiring diagrams and schematics, and postulating different
events and what could happen. It is the same type of -- parts of those
meetings are the same type of discussion.
MR. BACANSKAS: And Steve reminded me to mention that. I
didn't put a specific note. But one of the things we are including in
the document, and have brought to our discussion are schematics, so that
we can sit down and talk in detail, this circuit with this
configuration, and these cables in this area, because sometimes, to get
a clear path to resolution, you have to go to that level of detail.
DR. POWERS: Especially in this area, I don't see how you
can avoid it.
MR. BACANSKAS: Correct.
DR. POWERS: Because the way the concern and the resolution
comes clear to you is by looking at what the circuit actually does and
where it is.
MR. BACANSKAS: Absolutely.
DR. MILLER: Are you starting with some of the assumptions
that are in the Generic Letter 86-10, I think it is? Are those the
starting point for some of this?
MR. BACANSKAS: I think that is an accurate statement.
DR. MILLER: In other words, you are not looking at physical
mechanisms of board circuits and stuff like that?
MR. BACANSKAS: No, we -- for instance, in question, I
guess, question and answer 531 talks about hot shorts, shorts to ground
and open circuits. We are looking at those types of faults occur. Now,
we are trying -- they occur as a result of the fire.
DR. MILLER: Okay.
MR. BACANSKAS: What we are trying to do is look at, is
there a means to bounds for a specific event and configuration? How
many of those would be credible? That also walks into position on the
hot short duration that you mentioned earlier, and then there's other
sideline issues such as, if I have a fire of significant magnitude that
it would burn up all of the cable in the tray, or burn up the insulation
on the cable in the tray such that these faults could occur, is it
credible then to assume that it would also damage the circuits such that
they could not continue to carry the type of current that would be
necessary to energize a pump motor, a valve, you know? And we are also
looking at, how far does that go.
DR. MILLER: So you are in a sense trying to bound, you
point out the --
MR. BACANSKAS: We are trying to bound the deterministic
analysis.
DR. MILLER: The answer in that particular Generic Letter.
MR. BACANSKAS: Yes. Also, I wanted to mention that we have
obtained abstracts and full reports on approximately, I guess it is
around 95 NUREG contractor reports on testing of cable related to fire.
Steve Knowlen from Sandia assisted us in that area. We have reviewed
all of that for pertinent data with respect to the electrical fault
conditions. And, unfortunately, what we have found to date is there is
not the database that you were alluding to earlier out there and
available.
DR. MILLER: Which is not a surprise.
MR. BACANSKAS: No, it is not.
DR. POWERS: Well, you know --
MR. BACANSKAS: There is a lot of information with respect
to how does the insulation system respond, but not to the electrical
performance characteristics. There's a fairly detailed report on the
effect of aging and the loss of antioxidants in the insulation system
and the impact on its propagation of flame with an ignition source. But
the cable was not energized during the testing, it is not in the
standard configurations, and so there is no real value from the
electrical circuit analysis standpoint.
DR. POWERS: Are you looking -- is anyone looking to see
what other people have? Certainly, you would think that maybe other
industries have it, or the insurance industry has, or perhaps the
military has it, especially, the Navy.
MR. BACANSKAS: I think I heard the same presentations you
did from the one individual. NEI is investigating that and speaking
with that individual, and we are trying not to do the same thing in five
different places. So we are looking to NEI to lead that effort.
DR. POWERS: Okay. Good.
MR. BECANSKAS: I think I've already stated this. Steve
mentioned earlier this morning that there were 20-odd positions that
we're looking at developing and after additional committee meetings and
discussions we found that, well, some of these really aren't position
statements. They are definitions and they are terms that are generally
undefined in the existing guidance documents and also in the
regulations, so we would like to attach a very specific definition so
that wherever that word appears you get the same meaning from the word.
We felt that we had to start with definitions so we are all
starting on a level playing field, as far as understanding.
We have reviewed several of these position statements with
the Staff, as I said, and our basis for them, and they have provided
feedback. We are discussing them and we continue to discuss them and we
have an Owners Group meeting planned the first quarter to work on more
of our positions and hopefully not long after to meet again with the
Staff to go through those.
The schedule -- we provided a full schedule for submittal of
our document to the Staff and originally we had looked at either late
first quarter or early second quarter of this year for completion of our
document with some of the input we are looking for from NEI we are
probably going to submit later in the second quarter of this year.
However, we still see this as a very high priority issue.
We have a lot of resources applied and we want to continue to move and
complete the activity in a reasonable timeframe.
The rest of my presentation I think is just a breakdown of
the document and the sections, and again our entire purpose with this
activity is to take all the existing guidance and put together a
methodology that can be used for support of the existing information
that we have and should design modifications be done in a plant to
factor in those changes through use of this methodology so that we are
all working to the same way of doing business.
We have had discussions with the other owners groups about
their participation in this activity and we have received interest from
them. I think that the BWR Owners Group is set up a little differently
as far as how they function and start committees and issues, which has
brought us to the front end of this issue and through NEI, through many
of our own utilities -- mine, for instance, we have Combustion
Engineering units, B&W units, and GE units -- we also have a fire
protection peer group that is supporting this activity internally, so we
are disseminating this information.
Unless somebody wants me to go through the specific details
of any of these sections, I can stop right here and hopefully answer any
questions you might have.
CHAIRMAN POWERS: I guess the first question that comes to
mind flows immediately from what you have just said -- disseminating the
information.
Do you anticipate that there are going to be wild
differences between what you -- this cookbook that you have come up with
at least with respect to approach on circuit analysis for a BWR or any
other plant? I admit the specific circuits are going to be a bit
different, the acronyms for the equipment are a bit different, the
functions of the equipment tend to be about the same?
MR. BECANSKAS: I think the simple answer to that is no.
There are some unique functions in a BWR that will be addressed, however
we expect that 80 to 90 percent of the overall document could be applied
to any operating reactor.
CHAIRMAN POWERS: And the next question that comes to mind
is that if indeed late in the second quarter you are able to submit a
document, I hope you would have a chance to come by and talk to us about
what you came up with in that document?
MR. BECANSKAS: We have anticipated all along that when we
had the details of the document or had a draft document that had been
approved through our executives, far enough along through our process we
would bring it here as well as the Staff.
CHAIRMAN POWERS: I think that would be very edifying for
us. We will promise that the words "risk" and "performance" will not be
allowed in the discussions.
MR. BECANSKAS: It seems to go quicker if you leave those
words out.
[Laughter.]
DR. MILLER: It would be interesting when you feel it's
available for the public to share your definitions and your positions.
I realize you wouldn't want to do that now because they're certainly in
flux, but that would be very illuminating right at this point.
MR. BECANSKAS: I don't think there would be any
disagreement with that, and we are still fighting with one or two of the
definitions because of all the various documents and how it is couched,
and you end up with a definition that is eight pages long. That doesn't
buy you a whole lot.
DR. MILLER: You need to shorten that a page or two.
MR. BECANSKAS: A little bit -- get it down to 10 and we'll
say we'll use Number 6. Thank you, gentlemen.
CHAIRMAN POWERS: Thank you very much.
This was -- I guess that I am enthused about this activity
just because it seems like it's a damn good QA check for reactivity that
you are taking under with the Reg Guide activity.
MR. MARSH: Well, I am enthused in that we have moved off of
this --
CHAIRMAN POWERS: Oh, yes. Yes. This is going to get us to
resolution.
MR. MARSH: We may not agree 100 percent with how this thing
ends up, but it is a lot better process to get to that point than
arguing over "this document is wrong" --
CHAIRMAN POWERS: These particular set of words and --
MR. MARSH: By using deterministic, risk insights and test
information, I think there's hope to end up at really the right place.
CHAIRMAN POWERS: Yes. I agree with you 100 percent on
that. This is going to take us to some sort of a resolution --
MR. MARSH: Absolutely.
CHAIRMAN POWERS: -- that everybody can live with.
MR. MARSH: Absolutely.
CHAIRMAN POWERS: That brings us to the close of the formal
presentations for the day. We will have a presentation tomorrow and --
quit chuckling -- Ed's got a hard job.
MR. MARSH: I am not going to tell him anything that was
said.
CHAIRMAN POWERS: Oh, you are a dirty dog.
MR. MARSH: He missed the day. That's his problem.
DR. MILLER: You are going to make him clueless?
MR. MARSH: That's right.
CHAIRMAN POWERS: I think it will be an important
presentation but I think it is more important for the subcommittee to
get its act together and figure out how we can comment effectively on
what we have heard, and we will do so.
We are also going to discuss the Research Program and the
IPEEE and that of course is a much-anticipated discussion by an erudite
and authoritative figure in the field --
DR. APOSTOLAKIS: You are not speaking for the subcommittee,
are you?
CHAIRMAN POWERS: You will quickly find that I will be
recusing myself and turning the chairmanship over to Mr. Kress.
So now let me close this session by asking if there are any
comments that the members would like to make at this point, recognizing
that we will have comments again at the end of tomorrow's session, but
George, aside from the fact that I think we all agree that we do have a
problem of PRA being one-way street to enhanced regulation that has to
be guarded against, anything else you would like to say?
DR. APOSTOLAKIS: Well, I think I have said enough. I'll
summarize tomorrow, I suppose.
CHAIRMAN POWERS: Dr. Kress?
DR. KRESS: I think I have said it.
CHAIRMAN POWERS: Your definition of performance-based
regulation still rings true in my mind.
Our distinguished guest from the I&C Committee?
DR. MILLER: Well, I find that this was a very illuminating
day and I am anxious to get more involved in the circuits aspect of
this. I learned a lot today and what I learned most is there's a lot
more to go here with this. The circuits aspects are kind of challenging
and kind of interesting.
DR. KRESS: We realize you I&C guys are the whole trouble.
DR. MILLER: Yes? I thought we were the solution.
CHAIRMAN POWERS: If the I&C guys would just use cables
about this big around, a lot of our problems would go away.
DR. MILLER: Fire resistant cable -- that's what you need.
What if we have to use all fiber optics cables and replace
all the electrical cables?
DR. APOSTOLAKIS: Who else is a member of this subcommittee?
CHAIRMAN POWERS: Frequently Mr. Barton attends -- because
we frequently find ourselves involved with plant operations and so often
he attends.
DR. APOSTOLAKIS: I think one of the utility people should
be a member, a standing member. Mario?
CHAIRMAN POWERS: Barton just frequently attends and I will
remind the members also that we are still anticipating a subcommittee
meeting in June in Region I.
DR. MILLER: On this particular issue or just --
DR. APOSTOLAKIS: You can change the dates -- June -- seven
months.
CHAIRMAN POWERS: It is a much more wide-ranging issue than
that.
With that I will recess the meeting until tomorrow after
thanking all the speakers and participants. I found it fun today.
[Whereupon, at 5:00 p.m., the meeing was recessed, to
reconvene at 8:30 a.m., Thursday, January 21, 1999.]
Page Last Reviewed/Updated Tuesday, July 12, 2016