491st Meeting - April 11, 2002
Official Transcript of Proceedings
NUCLEAR REGULATORY COMMISSION
Title: Advisory Committee on Reactor Safeguards
Docket Number: (not applicable)
Location: Rockville, Maryland
Date: Thursday, April 11, 2002
Work Order No.: NRC-325 Pages 1-407
NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers
1323 Rhode Island Avenue, N.W.
Washington, D.C. 20005
(202) 234-4433 UNITED STATES OF AMERICA
NUCLEAR REGULATORY COMMISSION
+ + + + +
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS)
+ + + + +
THURSDAY, APRIL 11, 2002
+ + + + +
+ + + + +
The Committee met at the Nuclear
Regulatory Commission, Two White Flint North, Room
T2B3, 11545 Rockville Pike, at 8:30 a.m., Dr. George
E. Apostolakis, Chairman, presiding.
COMMITTEE MEMBERS PRESENT:
GEORGE E. APOSTOLAKIS Chairman
MARIO V. BONACA Vice Chairman
F. PETER FORD Member
THOMAS S. KRESS Member
GRAHAM M. LEITCH Member
DANA A. POWERS Member
VICTOR H. RANSOM Member
STEPHEN L. ROSEN Member
WILLIAM J. SHACK Member
JOHN D. SIEBER Member
ACRS STAFF PRESENT:
JOHN T. LARKINS Executive Director
I N D E X
Opening Remarks by the ACRS Chairman . . . . . . . 4
Final Review of the Turkey Point License
Renewal Application. . . . . . . . . . . . . 7
Advanced Reactor Research Plan . . . . . . . . . 116
CRDM Penetration Cracking and Reactor Pressure
Vessel Head Degradation
Peter Ford . . . . . . . . . . . . . . . . 206
Larry Mathews. . . . . . . . . . . . . . . 207
John Wood. . . . . . . . . . . . . . . . . 257
Ken Byrd . . . . . . . . . . . . . . . . . 278
Mr. Jack Grobe . . . . . . . . . . . . . . 295
Ken Karwoski . . . . . . . . . . . . . . . 333
Westinghouse Owners Group (WOG) and Electric
Power Research Institute (EPR) Initiatives
Related to Risk-Informed Inservice Inspection
Allen Hiser. . . . . . . . . . . . . . . . 337
Andrea Keim. . . . . . . . . . . . . . . . 357
Stephen Dinsmore . . . . . . . . . . . . . 367
Adjourn. . . . . . . . . . . . . . . . . . . . . 407
CHAIRMAN APOSTOLAKIS: The meeting will
now come to order. This is the first day of the 491st
meeting of the Advisory Committee on Reactor
Safeguards. During today's meeting the Committee will
consider the following: Final Review of the Turkey
Point License Renewal Application; Advanced Reactor
Research Plan; CRDM Penetration Cracking and Reactor
Pressure Vessel Head Degradation; Westinghouse Owners
Group (WOG) and Electric Power Research Institute
(EPR) Initiatives Related to Risk-Informed Inservice
Inspection of Piping; and Proposed ACRS Reports.
This meeting is being conducted in
accordance with the provisions of the Federal Advisory
Committee Act. Mr. Howard Larson is the designed
federal official for the initial portion of the
We have received no written comments or
requests for time to make oral statements from members
of the public regarding today's sessions. A
transcript of portions of the meeting is being kept
and it is requested that the speakers use one of the
microphones, identify themselves and speak with
sufficient clarity and volume so that they can be
I will begin with some items of current
interest. First of all, we are welcoming back Mr.
MEMBER LEITCH: Thank you. It's good to
CHAIRMAN APOSTOLAKIS: That's good. I
would like to inform the members that Chairman Meserve
will be here tomorrow at 11 a.m. to welcome our newest
member. And at 1 o'clock tomorrow afternoon we are
all going as a group to have our picture taken
individually because eventually we will get new
MEMBER SHACK: I'll need to dress up for
MEMBER SIEBER: Would that be possible?
MEMBER SHACK: That's the problem.
CHAIRMAN APOSTOLAKIS: You all have this
handout, items of interest. There are five speeches
by the Commissioners at the recent Regulatory
Information Conference. Also, we have summary of the
Reactor Oversight Process Inspecting Findings that
should be of interest and also you will see on page 27
a news item that Westinghouse Electric Company has
submitted an application for design certification of
the AP-1000 design. And Dr. Kress has a tape perhaps
we should all see?
MEMBER KRESS: Yes, I have here in my hot
little hands a copy of a copy of a copy of a copy.
Sandia at work, mostly, that I obtained by nefarious
means and what this is is a tape showing a lot of the
things they did to show the robustness of spent fuel
casks, like running trains into them and dropping them
off of buildings and etcetera. So if anybody is
interested in seeing this and I have it and I guess
Theron can set it up and show it at noon time some
CHAIRMAN APOSTOLAKIS: How long is it?
MEMBER KRESS: It's not very long.
CHAIRMAN APOSTOLAKIS: Okay, so maybe we
can do that at 12:30 or so?
MEMBER POWERS: After members have watched
it and convinced themselves that the casks are
incredibly robust, I'll them what's wrong with the
MEMBER KRESS: Okay, great.
CHAIRMAN APOSTOLAKIS: Okay, so I think we
are now -- do the members have anything else to add by
way of introduction?
Okay, so the first item on the agenda is
the final review of the Turkey Point License Renewal
Dr. Bonaca is our lead member. Dr.
VICE CHAIR BONACA: Yes, good morning. On
March 13, the Subcommittee on License Renewal traveled
to Turkey Point and at that time we visited the site.
We were able to observe on the simulators the ability
of the plant to interconnect the emergency diesel
generators from one unit to the other unit for station
We also heard from the plant about the way
that they addressed closure of the open items. There
were only four open items in the SER for license
renewal. We had an opportunity to observe the site
and note the excellent physical conditions of the
equipment on the site.
In the afternoon on the 13th we met in
Town Hall of Florida City and there we had a public
meeting and we heard from the staff how the open items
had been addressed and closed.
During that meeting we also had some
observation from a member of the public. We also had
in writing some concerns raised by another member of
the public. The two concerns really echoed each
other. One of the concerns that was raised had to do
with voids in the concrete structure of the
containment identified at Turkey Point, both units, in
the early 1980s. We heard from the site personnel on
how the issue had been addressed. We felt reasonably
confident that they had been addressed properly. We
asked questions regarding the generic implications,
how they had been addressed and for those we have
asked the staff to come today and tell us how they
were handled for the other sites.
And so with that in mind, we have a
presentation this morning both from the Turkey Point
people and from the staff and at this point I turn the
meeting to PK Kuo who is here to present us on that.
MR. KUO: Thank you, Dr. Bonaca. Good
morning, members of the Committee. My name is PT Kuo,
the Program Director for the License Renewal and
Involved* Impacts Program. This is my first week on
the job. Chris Grime has moved on to take on new
challenges and we all wish him good luck. I also want
to introduce Mr. Frank Gillespie on my right. Mr.
Gillespie is the Deputy Director for the Division of
Regulatory Improvements Program.
Today, after the Applicant's presentation,
the staff will brief the Committee on the review
results of the Turkey Point license renewal
applications and specifically, the staff will address
in detail the questions raised by Mr. Oncavage in
their letter to the Committee on concrete voids and
the hurricane damages.
We are going to have an assembled panel to
brief the Commission. We also have a technical staff
sitting in the audience ready to answer any of your
With that I will ask whether Mr. Gillespie
has any opening remarks?
MR. GILLESPIE: Yes. Let me just address
the concrete void issue because we may not have done
as much research on it as we would like relative to
everything from the old Oyster Creek problem with
spalling concrete on the outside to the voids that
were identified in the 1980s and going back and saying
did we consider this generically at that time?
The staff is going to be prepared to
address it for Turkey Point where we think it's been
plant specifically resolved and I'm going to tell you
right now we might have an IOU to have to come back as
we were kind of talking about this last night,
prepping for the meeting. We might not have done the
generic research on the other aspects of it quite yet
and we're kind of still in a process. The other thing
is hopefully between the staff and the licensee's
presentation, we will address things like Part 21 on
analysis and decision points that are in Part 21 on is
it significant, is it generic? And the lack of -- and
it's a question of documentation for convenience.
While the letter you got from this individual was, in
fact, an open letter, the Agency did enter it and
Region II is going to be on the phone to try to
address this. They did enter it into the allegation
system. Even though it was an open question it got
put in the allegation system to make sure we followed
up and got with the person and got back to them and
got letters to them and did an inspection.
Unfortunately, that system gives the
appearances because it, in general, was designed to
protect people's identity of being kind of private and
therefore the link to the plant-specific issue and
what was done might not be obvious in public
documentation because of that. So Region II is going
to be on the phone to try to address that to the
degree they can.
We put ourselves in a procedural box when
we put a public issue in a private system.
MR. BLANCHARD: Yes. I realize just for
the benefit for those members who were not in the
meeting, this is all because in their mind there was
an expectation that since this was a potentially
generic issue, maybe the licensee had initiated a Part
21 which speaks of a defect to a significant
component. And Part 21's intent is the one of making
the issue known, available to all plants so that
people can look at their own plant and inform the NRC
that there is an action to be taken on that. And
that's why we raise these kind of issues and we will
hear from Region II how it's handled.
MR. GILLESPIE: So we'll take our best
shot at answering all of the questions, but we may
have a little something. I talked to Goutam here and
depending on how it all comes out when we get all the
facts on the table, we might have an IOU still left.
VICE CHAIR BONACA: Yes, it's important,
however, today that we also separate Turkey Point and
how it was addressed at Turkey Point --
MR. GILLESPIE: Yes --
VICE CHAIR BONACA: From the generic issue
because that may have to be handled actually -- they
should be handled differently. We want to make sure
that there isn't any outstanding issue to the drafting
of a letter of the report at Turkey Point.
MR. GILLESPIE: Yes. And PT told me last
night, he said "I'm the license renewal guy." And he
says, "this is an operating question." I said, "Yeah,
but you're stuck leading the meeting." So --
MR. KUO: And if I may add, we also have
a Region II representative who will be tied up in the
telephone line and to here and to answer any questions
you may have.
VICE CHAIR BONACA: Thank you.
MR. KUO: Thank you.
CHAIRMAN APOSTOLAKIS: Okay, the Applicant
can go ahead.
MR. HALE: Can everybody hear me okay?
Hi, my name is Steve Hale. I'm the Project Manager
for License Renewal for Turkey Point in St. Lucie. I
thank you for the opportunity to talk to you all
today. I know I've met several of you when you came
to the site, as well as the ACRS subcommittee meeting
we had last September.
What I'd like to do today is give you an
overview of the application and then talk specifically
about two of the open items which were a little more
complicated to address than say some of the others and
I'm going to talk about the closure of the nonsafety
related which can affect safety related category of
scoping and the license renewal rule, what we call
Category 2. Then I'll talk about field-erected tanks
and the program that we propose for field-erected
tanks to close that open item.
When we began the license renewal
application effort for Turkey Point, a lot of the
guidance that's in place today was really in draft
form, so we had to drawn on multiple sources. While
we had Part 54, we have a draft standard review plan,
but it was under major revision at the time. We had
a draft GALL report. We tried to address and look at
GALL as part of our overall process, but that was also
in the developing stage for Turkey Point. We had a
draft Reg. Guide, but we had 95-10 which was issued,
I guess the final rev. was in the 1996 time frame
which had undergone somewhat of a demonstration
program, so we utilized the methodology that was in
Additionally, we tried to use the lessons
learned from previous applications, RAIs and RAI
responses which were on-going with Calvert Cliffs and
Oconee at the time. And as generic issues were being
resolved between NRC and NEI, we tried to factor those
also in co-application as they were available and as
they were applicable to Turkey Point.
One of the efforts NEI underwent in 1999
was working with the NRC staff and trying to come up
with a format that we both could agree on so we could
get used to the information being presented in the
same places. This was, I believe, in the 1999 time
frame and essentially, based on the draft SCs that
were issued for Calvert Cliffs and Oconee, plus some
lessons learned through those reviews, we structured,
we came up with a format that both the staff and NEI
agreed to and ANO was really the first to follow that
standard format and then we followed Hatch because of
where they were in the development of their
application, attempted as best they could to address
that format, but based on where they were, they really
had a difficult time in trying to comply with it
And then I think the subsequent
applications that have come down the pike, Dominion's
applications, Duke's other applications as well as
Peach Bottom, followed the standard format. It's
broken down into four chapters. The first chapter
addresses the administrative information requirements
of Part 54. Chapter 2 goes through the methodology we
utilize for scoping and screening and presents that
results. Chapter 3 is where you do your aging
management review and Chapter 4 addresses time-limited
Now I hadn't intended to go through
scoping and screening methodology today. We went
through that in great detail with the subcommittee on
September 25th of last year.
Also, as part of that standard format
there were several appendices. One was the UFSAR
supplement. The second is Appendix B where we have
summaries of our aging management programs presented
in the ten element format addressing staff
requirements on how they want aging management
programs presented. We included an Appendix C and
this was really to address some of the, what we call
generic type RAIs, RAIs regarding positions, regarding
aging effects and that sort of thing. It wasn't
required by the standard format, but this was an
intent on our part to address some of the RAIs we had
seen in previous applications and we felt Appendix C
did a good job of addressing some of those. Appendix
D would include any of the technical specification
changes that would be identified by the overall
process and then as an adjunct or really an attachment
with the application comes the environmental report.
When you look at the scoping criteria in
the rule there's a criteria of safety-related
components that -- and there's three criteria
stipulated for safety-related. Non-safety related
which can affect safety-related, based on our review
of this, we saw two types of non-safety which can
affect safety. One is where the non-safety system has
to function in order not to affect a safety-related
component. And the other is one for potential of
interactions, where the failure of the non-safety
system could potentially affect the function of the
safety-related system. And then category 3 is the
five regulated events: fire protection, PTS, EQ, ATWS
and station blackout.
In the application, you'll find in Section
2.2 a summary of all the systems at the plant and the
ones we had identified as in the scope of license
renewal and we do the same with structures. As you
can see, about half the systems in the plant have at
least some portions that fall within the scope of
license renewal and a little less than or a little
more than a third of the structures at the site.
I have to note that the structures at the
site include anything in the protected area so you
have a lot of the administration buildings and that
sort of thing as why not essentially comes into play
is the power block buildings.
For screening, this is where you really
get down to the nuts and bolts of the components and
structural components that support the functions that
were identified in the system and structure level of
scoping. And going through screening, the first step
you do a component level scoping. Then you look at
whether the component performs its function without
moving parts or change configurations, essentially
what we consider to be passive and/or they're not
subject to replacement based on a qualified life. So
you take each major system or structure in the plant
that falls within the scope of license renewal. You
break it down into its pieces, parts, you determine
which ones support the functions and you establish
which of those components are passive and which ones
are not replaced regularly.
The results of screening are presented in
the six column tables in Chapter 3. One of the
lessons learned that we had with the Oconee and
Calvert Cliffs applications was the fact that it
really makes it good to see the entire IPA on one set
of tables, so you have the scoping and screening
results essentially in the first two columns and then
you have a balance of the aging management reviews, so
rather than including duplicate tables in Chapter 2
and Chapter 3, we simply provide a summary in Chapter
2 and refer to Chapter 3 which lists the scoping and
screening results and then you can see the rest of the
IPA stacked up with each one of those components.
The mechanical sections, again, this is
consistent with the standard format that was
developed. You had a reactor coolant system,
connected systems, emergency safety features,
auxiliary systems and steam and power conversion.
In the structural area, we chose to break
it up between the containment and other structures and
in the electrical and I & C section, it essentially
looks at all the electrical components of the site and
it follows a slightly different process than the
mechanical and civil sections.
We also submitted license renewal boundary
drawings along with our application. Again, the staff
has indicated that really facilitates their review in
the mechanical area and lets them see what the
boundaries were and what equipment was included in
scope based on the actual drawings generated from the
PNIDs at the plant.
Aging Management Reviews are presented in
Chapter 3 and Appendix B because really the Aging
Management Review not only consists of identifying the
aging effects, but demonstrating the aging effects are
adequately managed for the extended period of
To facilitate the review, we grouped the
items in the Aging Management Review the same way
they're grouped in the scoping and screening section
so you had a one to one correlation through the
application. Again, the results are presented in the
six column tables including identifying the aging
program that manages any aging effects that
For nonclass 1 components, again in
Appendix C, some of the technical positions we took
regarding certain types of aging effects are presented
there for non-Class 1 mechanical as well as civil
structural. In the Class 1 area, we develop and
discuss the aging effects specifically in Chapter 3.
One of the things that we felt was
mandatory as part of our review for license renewal
was doing an extensive review of both industry
experience as well as plant-specific experience at
Turkey Point. We reviewed INPO and NRC generic
communications and also our responses and any of those
that really were related to aging we went back and
relooked at those to see if we'd addressed them
In terms of plant-specific history, we
went back and looked at the nonconformance reports and
condition reports, I think all the way back to the
early 1980s. We looked at event response teams.
These are teams we form when we have a significant
event at Turkey Point like a plant trip, those sort of
things. We form teams whose goal is not only to
identify what needs to be done to get the plant
started up, but also root cause and this type of
One of the great source of information we
have, we have a metallurgical lab and all of the
nonconforming conditions or condition reports that
require metallurgical analysis are submitted to the
metallurgical lab for determination of root cause and
the type of aging effects. We also drew on that
population. Those were available, I think, at Turkey
Point we had over 200 metallurgical lab reports so we
used as another major source of information for
And as also part of our process, our
procedures and the way we developed our Aging
Management Review had us go and specifically talk to
the system engineers and the component engineers. My
team was located on the Turkey Point site, so we had
quite a bit of interface with the engineers that deal
with the systems on a day to day basis.
CHAIRMAN APOSTOLAKIS: Now from the
metallurgical laboratory reports, I don't understand
what benefit you had from those. Is it possible that
you would decide to do something by looking at one of
those reports that you had not already done?
MR. HALE: One of the issues that has been
identified as the one -- hey, we don't think aging
effects are occurring, but you need to go in and do
one-time inspections to verify. Pitting is a good
example. But if you go back and you look at
metallurgical and you sort on things like stainless
steel systems with chemistry control, you can look as
whether you've ever had any specific failures related
to pitting or stress corrosion cracking. We use
metallurgical lab reports when they determined that
we've had loss of material due to MIC and we folded
those -- we developed tools for doing aging management
reviews on the non-Class 1 mechanical systems because
those are the ones where you get the wide variety of
materials and environments. And one of the things you
use is hey, the tools the industry may develop may say
that you have to address stress corrosion and cracking
in the system, but if we can go back to the
metallurgical lab reports and say we've never had
stress corrosion cracking in this system and we can
develop a technical basis for it, it provides a good
source of information. Again, on the other hand, the
tools the industry develops may say you don't get MIC
in these kind of systems. Where we have experienced
MIC and we discovered that through our interface with
the metallurgical groups as well as the metallurgical
lab reports. So we're not saying that we just use it
carte blanche. What we're saying we use that
information as additional research in some of the
technical positions we may have taken with regards to
CHAIRMAN APOSTOLAKIS: Okay.
MR. HALE: Any other questions related to
Time Limited Aging Analysis. These were
the major TLAAs at Turkey Point: EQ, class and
balance of plant fatigue, containment tendon
relaxation, reactor vessel irradiation embrittlement.
We had a couple of cases wear/erosion where we had
TLAAs associated with that. Containment liner
fatigue, crane fatigue. Also as part of the rule we
have to do a review of time bound exemptions whether
we had any and our review determined we didn't.
With regard to the UFSAR supplement, we
submitted a markup with the application. In addition
to that we included a new chapter in the SAR which
includes all the AMPs that are committed to for aging
management, as well as a description of every one of
the TLAAs that were identified. Also, in the FSAR
supplements our commitments related to programs are
included. Now additionally, one of the things we did
with the staff, we've updated the SAR supplement to
include all the commitments that were identified as
part of our review of the application. In other
words, with RAIs, responses to RAIs, we included any
additional commitments that came out of that
interchange into a revised SAR supplement that we
issued late last year.
With regards to Appendix B where Aging
Management Programs are located, for each aging effect
requirement management an Aging Management Program is
identified. We presented these programs in the 10
attributes following the guidance issued by the NRC.
We've got three categories of Aging
Management Programs. We have those that are existing,
those that need to be adjusted and those that are
brand new. You see we have pretty equal distribution.
Again, I described Appendix C, non-Class
1 component, Aging Management Review Process,it's not
required by the regulation, but we did submit it to
address some of the previous RAIs we had seen and
other applications. And Appendix D was technical
specification changes. We did not have any for the
Turkey Point license renewal application.
I just wanted to mention the environmental
report because there is an environmental piece. Some
of the unique things about the Turkey Point site, we
have thousands of miles type of cooling canal system
and you see it from satellite photos, in fact. We do
not identify the need of any major refurbishment which
is one of the issues that needs to be addressed in the
We do not use wells at the site. We
essentially, the only water we use from the local
community is domestic water. And the evaluation we
performed against the alternative show that license
renewal is the lowest impact option under the
MEMBER LEITCH: Steve, I have a question.
I'm not sure if this is the right time to bring it up
or not, but the fossil units that are adjacent to the
nuclear units --
MR. HALE: yes.
MEMBER LEITCH: It seemed to me that --
and I'm going on memory of quite a few years back, but
it seemed to me that during Hurricane Andrew there was
some missiles from the fossil unit that damaged a part
of the nuclear unit. I think it was in the fire
protection pump house or something like that.
MR. HALE: What happened was we had a high
tower out in the water treatment plant area and the
high tower fell over on one of our domestic water
tanks. We have two tanks and the domestic water tanks
are also what you credit for your Appendix R, I
believe A-1, whatever, it's our fire protection water
sources. So the tower actually fell over on one of
the tanks and as a result we got into one of the start
up issues we had related post-Hurricane Andrew was
providing the water sources until we could reconstruct
MEMBER LEITCH: I guess my question is in
the 20 years extension period for this license, what
assurance do we have that the fossil units wouldn't be
retired and as many fossil units abandoned in place
and that there might be missiles, if you will, created
as a result of that that could in future storms damage
the nuclear unit?
MR. HALE: Well, for one, the safety-
related equipment is protected from missiles as part
of our design basis. In fact, the safety-related
portions of the plant and even some of the nonsafety-
related portions of the plant survive very well. We
were back on line within a month after Hurricane
There were a lot of missiles during
Hurricane Andrew, independent of whether the fossil
unit was there or not. We had winds in the area of --
the eye passed over Turkey Point and we were in 150 to
160 miles per hour range. The South Florida building
code is about 120 and so trees -- there was a missile
that went through one of the oil tanks, what they call
the day tanks that affected that particular tank.
The nuclear plant fared very well with the
exception of that high tower falling on the fire water
tank and a materials warehouse that was outside of the
protected area. The plant did very well. I think
it's a proof test on the plant so to speak, but one of
the things in terms of interactions that was
identified some years ago and has been evaluated is
the seismic capability of the smokestacks. And they
have been evaluated. In fact, we've included them in
the scope of license renewal for that very reason.
MEMBER LEITCH: Okay, the smokestacks at
the adjacent fossil plant?
MR. HALE: Yes, yes. You'll find them
discussed in the application, in fact.
MEMBER LEITCH: That's good. Thank you.
MEMBER SIEBER: Those stacks aren't very
high though, right?
MR. HALE: About 400 feet. I wouldn't
want to climb to the top of them. There are some
folks who do who have to work on the lights, that sort
Okay now I'd like to go through the
resolution of open items and Dr. Kress, I've tried to
-- you had mentioned the criteria, so I've included
some additional information there. I hope I address
your question that you had.
This is a presentation I went through with
the subcommittee when they were at the site. The open
item is entitled scoping of seismic II over I piping
systems. It really goes beyond that. This is really
interactions between nonsafety and safety-related
system and the potential impact on safety-related.
One of the things I wanted to summarize
was go through the components we included in the scope
of license renewal to start with: (1) any pipe
segment beyond the pressure boundary which is included
in the seismic analysis, we included that pipe segment
in the scope of license renewal because it fit in that
first category which is it's performing a function in
support of the safety system.
We included all piping component supports
for nonsafety-related mechanical systems with the
potential of seismic II over I interactions because
Turkey Point is an older plant. We did this on an
area basis. We basically went through each building
of the plant and any room that contained both
nonsafety and safety-related equipment all the
nonsafety-related supports were in the scope of
license renewal in that area, regardless of whether
the stuff could follow effect or whatever, we just
said this area contains both types, so as a result all
the nonsafety-related supports associated with
ductwork, cable trays, conduit and in addition to that
we included the conduit itself, the cable trays and
other structural components outside of the mechanical
area, in these areas where you had both safety and
In addition to that, we had done a fairly
extensive internal and external flooding analyses so
anything related to that was included in the scope of
license renewal and this basically included curbing.
We have some sump pumps down in the RHR pump rooms and
those sump rooms that were included in the scope of
license renewal as well to accommodate flooding
effects and in addition to that, we included all the
pipe whip restraints, barriers, these type of things
that we credit for jet impingement, effects of spray
and pipe whip.
That's what we included in the scope of
license renewal to start with. After a lot of
dialogue between the staff and ourselves, the issue
that was identified is that the effects of pipe whip,
jet impingement, physical contact, pipes falling on
pipes and leakage due to credible and that's an
important word, credible nonsafety-related pipe
failures, beyond the current assigned break locations
because we've evaluated breaks in certain places, but
we haven't evaluated them across the board, need to be
considered based on the industry operating experience.
In other words, if you'd had failures of
nonsafety-related piping, through operating
experience, and you have a piece of a similar type
piping routed above safety-related equipment, then
that should be something that should be included in
the scope of license renewal and managed from an aging
As a result of this issue, there may be
some additional pipe segments that need to be included
in the scope of license renewal and thus an Aging
Management Review needs to be performed. During our
ACR Subcommittee walk down to the plant, I showed the
ACR Subcommittee an example of the kind of pipe we
were talking about.
What we did as a result of that and all
these rooms where we had both nonsafety and
safety-related equipment we did an evaluation assuming
credible failures based on operating experience of
nonsafety-related piping beyond what's currently in
our current license basis. If there was an
interaction with safety-related components as a result
of this failure, we in turn included that pipe segment
in the scope of license renewal.
To address the criteria --
CHAIRMAN APOSTOLAKIS: Let me understand
this. Something is credible if it has happened?
MR. HALE: In operating experience in the
CHAIRMAN APOSTOLAKIS: Oh, in the industry
MR. HALE: In the industry at large. Not
necessarily -- although a lot of this piping is not in
the scope of license renewal and that sort of thing,
we don't operate with leaks at the site and we manage
that, but the real issue is when you're looking into
the future, without doing specific aging management
say on a piece of pipe, could it fail, such that it
would affect safety-related equipment.
So we used a fairly conservative criteria
in establishing the interaction. Basically, what we
said if we had a nonsafety-related piece of pipe in a
room with electrical equipment and that electrical
equipment is not qualified for outdoor service, then
we said that pipe is in the scope of license renewal.
We didn't do a rigorous evaluation or analysis of
spray and see if the component could accommodate it.
We basically just concluded whether it would actually
affect it or not through analysis, we said that pipe
segment was in the scope of license renewal from a
From the pipe whip, jet impingement and
physical contact and this was basically the high
energy piping out on the turbine building, it really
took walk downs and actual physical observation of the
piping and essentially we took the criteria that if we
could see the pipe and the safety-related equipment,
that piece of pipe was in the scope of license
renewal. It wasn't based on a rigorous analysis, but
we took a very conservative posture on this.
And in this case it was primarily conduit
and cable tray routed out in the turbine building, so
if we had to run a cable tray between two walls and
there was high energy piping in the area, we said that
high energy pipe is in the scope of license renewal.
I don't know if that addresses the
criteria question that you have, but we basically just
took a conservative position on it.
What was the results of all this? We
included a number of pipe segments in five of the
structures that contained safety and non-safety
equipment. We identified the aging effects requiring
management for those pipe segments and for those that
require aging management, we included them in our
chemistry control program, our systems and structures
monitoring program and our Flow-Accelerated Corrosion
Program. And we've already made al those changes in
the program documents. In most cases, they were
already included in the program to start with. We
just had not identified the piece of pipes in the
scope of license renewal.
MEMBER ROSEN: What is the qualifier as
MR. HALE: Well, this is just a broad
statement, you know, you don't use FAC on a non-FAC
system. It was just a broad -- if you locate our
open-item response, I don't know if you all have
copies of that. We highlight specifically what
systems and what programs apply to which pipe
MEMBER ROSEN: But it's not an out -- all
of the above is true except when we decide we don't
MR. HALE: No, no, no, no. The intention
here is not all these programs apply to all the pipe
segments, that's all. FAC applies to only certain
types of systems. Chemistry applies to certain
systems as well as the system structures and
monitoring program. It's in a lot of detail in our
open item response and we've incorporated it on a
component level basis where we identify the specific
programs that are required.
Any more questions on II over I? Now this
is one that I think the industry and the staff are
working towards a resolution such that this will not
become an open item on subsequent applications where
the guidance gets clear, because a lot of it comes to
communications and your ability to understand what the
true issue is and I think once we understood, then it
was easy for us to work through what it was we needed
VICE CHAIR BONACA: Do you think the
guidance now is clear enough?
MR. HALE: I think it's still going to be
a challenge because for -- for older plants. I think
newer plants, we've done an initial scoping review for
St. Lucie. It's not going to be quite the same. The
older plants have some unique design features --
VICE CHAIR BONACA: But the logic is
pretty clear. Older, previous evaluation, II over I
were based on concerns with high energy line break, so
therefore you're looking for those kind of effects,
MR. HALE: Right.
VICE CHAIR BONACA: Whatever. Aging now
introduces potentially some other locations for
failures that are not already covered by previous, so
it seems to me the logic is clear. I mean --
MR. HALE: Right.
VICE CHAIR BONACA: The question is how is
the guidance now because we're be looking for. We
thought that the guidance provided in the SER for
Hatch was quite clear.
MR. HALE: Yeah. Once you understand what
the true issues are, I think that -- again, these
guidance and these generic interchanges we're having
with the staff are a real positive step, I feel, get
some of these down on paper, you know, so we can -- we
don't get into the point where it's an open item.
But the other item I was going to talk
about was related to field-erected tanks. This was an
item where the NRC had identified to us three times
they wanted us to address regarding field-erected
tanks. One, we had not supplied specific acceptance
criteria in the application regarding inspection.
They wanted us to include some additional provisions
in our program that called for additional examinations
if the one-time inspection we had proposed identified
extensive loss of material. And also provide a little
more information regarding why we felt we only needed
to do one-time inspection on these tanks.
With regards to the acceptance criteria
and additional examinations, the acceptance criteria
is any loss of material greater than the tanks
corrosion allowance, okay, will require specific
corrective action in our corrected action program and
as part of that, we'll consider the use of any
additional volumetric or service inspections and
identify as well, whether we need to do follow-up
inspections and that has been incorporated into the
Our basis behind one-time inspection and
I'd like to point out in any of these cases where we
say a one-time inspection is because we're going into
it with the thought that we don't expect to find an
issue. In any of these one-time inspections, if we do
find problems we would be required under our
corrective action program to follow up and establish
future inspections and that sort of thing. So when we
say one-time inspections, we're saying that this is
something where we don't expect to find anything, but
our corrective action process would require us to
follow-on if we had to.
VICE CHAIR BONACA: So if you find
something when you do the one-time inspection, you'll
convert that to a program?
MR. HALE: It depends on the aging effect
and what it may be, but if it's something that looks
like it's going to be a continuing thing that we need
to manage, then certainly we would institute follow-on
inspections, but that would be part of our assessment
and evaluation and what we saw.
Again, the first plan is under the one-
time we don't expect to find significant aging. Our
plant operating experience has revealed no incidents
of degradation of CSTs, RWSTs and DWSTs, other than
some repairs we had to do to the condensate storage
tanks were attributed to one, we had some poor
coatings to start with on the tank and secondly, the
tank was being subjected due to an operational problem
to hotter basically steam fluid was blown into the
tank which was causing some degradation around the top
that it was never really designed for. This is a
field-erected atmospheric tank and it was being
exposed to some higher temperatures.
Secondly, we went into the demineralized
water storage tank recently to install a floating
cover on it to help with oxygen control. We didn't
find any degradation in that tank as part of that
modification we performed.
On top of that, the RWSTs, the CSTs and
the DWSTs, we inspect those. Those are part of our
on-going external inspection program so any problems
with the tank, you would see corrosion that sort of
thing on the outside of the tanks.
When the ACRS Subcommittee was at the
site, we pointed out a couple of the tanks as part of
our walk down we did.
Okay, that's all I had with regards to my
Do you have any other questions?
VICE CHAIR BONACA: Do you have anything
to say about the statements from Mr. Oncavage or are
they going to be at a later time?
MR. HALE: I went back and as part of Mr.
Oncavage's statements I looked at what we did as a
utility, with regards to the discovery, analyzing it,
evaluating any corrective actions. With regards to
the Part 21 issue, our procedures require us to
address defects under Part 21. It's a mandated
requirement. It's in our quality instructions.
One of the things that you have to do
though is to do a significant safety hazards
evaluation to establish whether it is reportable under
Part 21. With regards to this particular event, the
evaluation performed by Bechtel one, determined that
the pressure integrity of the containment was never
compromised and this is documented in the Bechtel
evaluation after discovery of the event --
VICE CHAIR BONACA: The design capability
of the containment?
MR. HALE: Well, two things. One the
pressure integrity, certainly the containment had
undergone integrated leak rate tests as well as the
structural integrity test previously and if you look
where the void was, it was beyond the welded portion
of the pressure battery.
Secondly, in that evaluation that Bechtel
performed they also demonstrated that the structural
integrity of the containment was not affected by the
voids. So for it to be reportable, at least from our
procedures, under Part 21, it would have to represent
a significant safety hazard and based on the fact that
the pressure integrity and the structural integrity
were not affected by the voids, it would not represent
a significant safety hazard.
VICE CHAIR BONACA: What I'm asking about
is the design capability, I'm referring specifically
to what you're committed to in your testing which is
your testing the containment for your design
capability which typically is lower, much lower than
the overall structure -- ultimate capacity.
MR. HALE: Right.
VICE CHAIR BONACA: So the question I had,
I guess, is that evaluation did not address the
ultimate capacity. It addressed the --
MR. HALE: The design capacity.
VICE CHAIR BONACA: Design capacity.
Okay. Just to make that clear. And so because of
that, it is now reported under Part 21?
MR. HALE: Right.
MEMBER LEITCH: I have a question about
your ability to inspect the head as per this recent
NRC inspection, NRC request, I should say. There are
different insulation configurations throughout the
industry which make it more or less difficult for
people to get a good look at the head. What's your
status as far as that response is concerned?
MR. HALE: Turkey Point, we've completed
bare metal inspections on both heads. Unfortunately
Turkey Point was, if you recall back in 1987, we had
a leak that we operated with --
MR. WILLIAMS: Excuse me, Steve?
MR. HALE: Yes.
MR. WILLIAMS: Is that the right slide?
You've got station blackout up there?
MR. HALE: I'm sorry, I apologize.
I'm sorry, I had a slide for the
MEMBER ROSEN: It's going to be
interesting to see you tie the two together.
MR. HALE: All right. As far as the -- as
I was saying, Turkey Point had an event with some
significant leakage in the reactor vessel head area.
In fact, it's what prompted 8805. We had operated, I
believe, about -- I believe it was about six months
with a known leak in the reactor vessel. It was the
As a result of corrective actions related
to that, one our insulation configuration was changed
somewhat to where we had inspection ports.
Additionally, we installed a radiation detector that
actually sniffs the head area and so we can get some
intelligence, you know, when we get high radiation and
containment and can help maybe locate whether --
MEMBER SHACK: N-16?
MR. HALE: Pardon?
MEMBER SHACK: N-16s?
MR. HALE: No, it's just radiation
detector in the head region. It's in an enclosure, so
we actually have a -- it's something we did to tell
us. And we also instituted some very stringent
leakage controls. We require specific evaluation if
leakage reaches .5 GPM and if needed, we'll actually
go in and do containment walk downs.
So the combination of those things,
although it was a negative event, I believe has
created a situation that we're finding and what we did
is we did a bare metal inspection as a result of
bulletin in 2001 related to Inconnel 600 on Unit 3 in
October of 2001 and we also did it in March of 2002.
I would also like to point out we were able to do this
and accommodate it within a normal -- we're doing
refueling outages in a 25-day type time frame and we
were able to accomplish this with that. We used
remote TV cameras. I actually went through the report
evaluation and they addressed each individual nozzles.
We've got videos and pictures, but it was clean.
There was no evidence of leakage and there was no
evidence of boric acid accumulations.
MEMBER SHACK: And you can literally do
100 percent inspection?
MR. HALE: One hundred percent visual.
With remote, television cameras and that sort of
thing. I believe it was Framatone that's developed
the -- but it's very detailed.
MEMBER ROSEN: This radiation monitor you
talk about, is it sampling the environment, the air
and taking it through a filter and putting it in front
of a detector?
MR. HALE: Yeah.
MEMBER ROSEN: Now those filters, are
those looked at?
MR. HALE: Yes. They're replaced
MEMBER ROSEN: What do you find on the
MR. HALE: I'm not sure. You're asking a
question that goes beyond --
MEMBER ROSEN: Well, I ask it because
Davis- Besse found a lot of iron oxide on their
filters and they had a similar system.
I think you ought to be finding that the
filters are clean.
MR. HALE: They replace the paper
periodically because they have to for the monitor
MEMBER ROSEN: They take off the paper to
replace it because they analyze it.
MR. HALE: Yes.
MEMBER ROSEN: But not because it's
plugged up or anything.
MR. HALE: Yeah.
MEMBER ROSEN: But you don't know?
MR. HALE: No.
MEMBER LEITCH: As a result of your
operating with the leakage back in the 1980s whenever
it was, did you find any wastage at that time?
MR. HALE: Not very much, but I think the
number that was quoted, like I said, I'm reaching here
was in the hundreds of pounds, had accumulated on
three of the reactor vessel studs in that stud area,
so there was some wastage on the studs. There was no
real wastage on the head itself, but again, this was
a conoseal leak.
MEMBER LEITCH: I understand.
MR. HALE: And it was, I believe in the --
it was within tech specs, but it was just spraying
over about six months it accumulated boric acid.
MEMBER LEITCH: Okay, thank you.
VICE CHAIR BONACA: Going back a moment to
the issue of the concrete, what did you do? How was
it repaired? What I am trying to understand is what
is the condition of the containment right now for both
units? I understand you repaired what you found. You
did not open every part of the containment so you had
the inspections to identify whether you had other void
MR. HALE: Bechtel essentially did a root
cause on the issue that was discovered. The root
cause determined it was a combination of a difficult
area to get concrete into plus where they had
established a construction joint. The repairs that
were implemented called for -- we were actually
putting in a heavier steel bottom to the equipment
hatch to remove the steam generators, so they removed
that. They poured the appropriate concrete and they
put a thicker piece of metal which was the intent all
along when they had pulled it off and discovered the
void. In terms of generic implications, based on the
root causes that were identified, Bechtel established
based on that root cause that they wouldn't find
similar type of areas like that based on that -- and
so that's documented.
VICE CHAIR BONACA: In other locations of
MR. HALE: Right, right. And that's
documented in there. It was a fairly extensive
evaluation that they performed to demonstrate that.
VICE CHAIR BONACA: So you do have
confidence that there are no voids in your
MEMBER POWERS: It's an incredibly self-
serving finding, isn't it, that everything is okay, we
only made one mistake.
VICE CHAIR BONACA: That's why I'm
interested in hearing about -- what is interesting is
that it happened in the hatch of one of the units and
then they looked at the other one and they found the
same problem right in the location. That's why we
would be raising questions about the generic
implications for other units.
Now so there is a confidence that that was
the only location in that containment that could have
been affected by that and it was this position for the
Turkey Point unit?
MR. HALE: Right.
VICE CHAIR BONACA: Containments.
MR. HALE: And it was also communicated
with -- communicated and inspected by the region.
There was an LER on it. They came in and looked at
the Bechtel evaluation as well as the disposition of
the repairs, so I have confidence. We've also
undergone, I think about seven integrated leak rate
tests on both containments and I have full confidence
in our containments.
MEMBER LEITCH: We had the same problem
with Limerick during construction. I think it was
Limerick I in about 1977 when the forms were removed
from the containment pour and this was above one of
the containment hatches, a large void was found. It's
right above the containment hatch. There was a real
configuration, complex configuration of rebar in that
area, but it was a very significant hole. That was
also a Bechtel job, by the way, and it was a very
significant hole. Were it for the rebar you could
easily put a Volkswagen and maybe a Cadillac into this
CHAIRMAN APOSTOLAKIS: It saved a lot of
MEMBER LEITCH: It saved a lot of
concrete. But of course that was self-evident and it
was all chipped out and replaced.
VICE CHAIR BONACA: Because that was
MEMBER LEITCH: Yes.
VICE CHAIR BONACA: I had another question
regarding another point that Mr. Oncavage raised
MR. HALE: Yes.
VICE CHAIR BONACA: Capability of the site
and he presented the fact that he didn't feel that
Hurricane Andrew was really a category 5 hurricane and
the ability of the plant to withstand a category 5 and
you addressed that issue.
MR. HALE: Yes, yes. In fact, the FSER
highlights are design capability, the two aspects of
a hurricane that you're concerned with is wind and
tidal surge, but with regards to wind design, I think
you'll find any FSER were designed for 225 miles an
hour and all the way up to 300 some miles an hour
without loss of structural integrity. So we are not
concerned from -- wind design is not an issue.
VICE CHAIR BONACA: Tidal surge was the
MR. HALE: When we look at tidal surge, we
are designed -- the plant elevation is at 18 feet. We
can -- we install stop logs as part of our hurricane
preps up to 20 feet and all the safety-related
equipment is located at 22.5 feet.
I had some friends that were affected by
Hurricane Andrew's tidal surge and so I had some
witness accounts of trucks at the top of their garage
as the thing came in and hit their house, but I think
if you look at historical data and that sort of thing,
22.5 feet is plenty to accommodate any tidal surge
that could be expected, even for a category 5
VICE CHAIR BONACA: Thank you.
MEMBER POWERS: Mario, in light of the
Davis-Besse events, have inspections for these one-
time inspections we do for license renewal, have they
come under question?
VICE CHAIR BONACA: I don't think so.
First of all, the components like such as a head are
really under a different kind of inspection program
that clearly is not one-time inspection.
MEMBER POWERS: I mean it's the mindset.
If you go and inspect something expecting not to find
it, you frequently don't find things. And there are
an awful lot of inspections in license renewal with
the predisposition not to find anything. And son of
a gun, they don't.
VICE CHAIR BONACA: Yes, if you look at
the issue or components which are related to the one-
time inspection, I'm not sure that they are the type
where your ability to detect would be so challenged.
For example, it's erosion, certain components or
corrosion and so on and so forth. The presumption is
that if you do the inspection close to the 40-years
life and you do it once and you don't find anything,
then you have -- and the first -- I think there is a
good provision in the license renewal that says you
can roll that inspection into your program until you
find something and it then falls under corrective
action. I think it's a good point you are raising.
I think you have to be sensitive to that as we look at
new license renewal applications in the future and see
what kind of one-time inspection we have, if it is, in
fact, an obvious thing that you would identify those
kinds of degradations easily or if your ability to
detect is being challenged.
MEMBER POWERS: Since we've been talking
about Turkey Point concrete, I've got to tell the
Committee at least one anecdote about the Turkey Point
concrete, but Turkey Point doesn't know. In 1976, the
NRC asked me to look at the effects of interactions
with concrete and they said use prototypic concrete,
so I said well, what's prototypic concrete? I decided
the FSARs probably had prototypic concrete described
in them, so I went to our library attendee and asked
him for an FSAR and they handed me a box of
microfiche, all jumbled together and they said these
are all the FSARs. So I went to sorting them out and
the first one I sorted out so it was reasonably
complete was Turkey Point and Turkey Point's FSAR has
an excellent description of their concrete and I used
that description of the concrete to create the
concrete I was doing and since I wrote it down
everybody else just used that as the specification and
as far as I know every melt concrete experiment that's
ever been sponsored by the NRC has used Turkey Point's
Sand size. I believe your aggregate is
oolite. I had to figure out what oolite was. And I
know more about the Southeastern United States geology
than I ever cared to learn trying to understand what
MR. HALE: Any more questions for me?
VICE CHAIR BONACA: I don't think so.
Thank you for the presentation. It was very
informative. We'll hear from the Staff and the SER.
MR. KUO: Yes. I will call on Mr. Raj
Auluck, the Project Manager for the Turkey Point
license renewal application review and his panel.
MR. CHRISTIANSON: Nuclear Regulatory
Commission, Chris Christianson speaking, may I help
MR. AULUCK: Chris?
MR. CHRISTIANSON: Yes.
MR. AULUCK: Raj Auluck.
MR. CHRISTIANSON: Hi, Raj.
MR. AULUCK: Hi. We are just starting in
a couple of places and I just wanted to make sure you
are on the line.
MR. CHRISTIANSON: Okay.
VICE CHAIR BONACA: Be aware we have about
45 minutes left for the meeting, including
discussions, so I leave it up to you to be --
MR. AULUCK: Okay. Good morning. I am
Raj Auluck, Project Manager for the Turkey Point
license renewal application review. With me around
the table is Jim Medoff. He's from Division of
Engineering and helping us so he'll assist me on a
couple of the slides. Then we have some people from
the technical division, Jim Lazeunick from electrical.
And they will discuss some of the issues which were
especially asked by the Subcommittee during our
meeting last week. Hans Ashar from Mechanical
Engineering and Barry Elliott from Materials.
The purpose of today's meeting is to
present the staff's review -- Chris, are you there?
MR. CHRISTIANSON: I'm here.
MR. AULUCK: I forgot to introduce Chris
Christianson. He's the Branch Chief, Region 2 and
he'll be helping us respond to some of the questions
you have on the inspections or the allegations.
I will describe the resolution of the open
items and the basis upon which we'll move forward to
make a recommendation to the Commission on this
The application was received 18 months
back, 19 months today exactly. This was the firth
application received by the NRC. Four have already
been approved. This is the first Westinghouse. It is
two-unit site. Each is designed for 2300 megawatt
thermal. The site is shared by two oil and gas fired
generating units in Florida City about 25 minutes from
the Miami, south of Miami.
Unit 3 license expires on July 19, 2012
and for Unit 4, on April 4, 2013. The application is
for two years' extension.
The review schedule originally issued with
an acceptance letter. As you can see, the next --
that line is completing the SERS briefing and
preparing the Commission paper with the recommendation
on middle of next month.
The final SER was issued on February 27th
and final environmental impact statement was issued on
January 15th of this year.
MEMBER LEITCH: I have a question
regarding the length of the extension. I read that
the PTS value is very close to the allowable 300
degrees. It's 297.4. And it's stated that that would
be okay because that was the value, I guess after 48
effective full-power years. Now we're extending this
to 60 years. Is it mathematically impossible to get
a number in excess of 48, full-power years, or is
there some kind of a caveat that says 60, but no more
than 48, effective full-power years?
MR. ELLIOTT: Well, you could get 48
effective full-power years corresponds to 60 years at
80 percent capacity factor. The plant could run
higher than that and therefore it would exceed the --
before it reached 60 years it would exceed 48
effective full-power years. But it's not the 48
effective full-power years. It's the critical factor
here. It's the neutron effluents received by the
vessel and that's the critical factor and that's what
they have to monitor to determine whether or not
they're going to exceed the PTS screening criteria.
As long as they monitor the neutron effluents and they
stay below their projections, they'll stay below the
screening criteria. According to the PTS rule, if
they start exceeding the effluents values they
projected, they're required to do another projection
of where they'll be with respect to the PTS rule. So
it's within the PTS rule there is a flexibility.
MEMBER LEITCH: Okay, so there is no
limitation then at 48 effective full-power --
MR. ELLIOTT: No, it isn't 48 effective
full power. It's the neutron effluents.
MEMBER LEITCH: And even although they go
above -- if they went above 48 effective full-power
years, presumably they'd be crowding that 300 degree
MR. ELLIOTT: They would have to tell us
the impact on the neutron effluents for the vessel and
then from that they would have to project the RPTS
value to determine whether or not they're still below
the screening criteria at end of license to extend the
MR. MEDOFF: Barry, may I add something?
MR. ELLIOTT: Sure.
MR. MEDOFF: I would like to add that in
a reassignment they do exceed the screening criteria,
the rule is written to require the licensee to take
appropriate action including flux reductions and/or
annealing of the reactor vessel. So the rule does
incorporate corrective action should those screening
criteria be exceeded.
MEMBER LEITCH: Okay, thank you.
MR. AULUCK: Continuing, we'll start with
how we reviewed the application. There are two
self-regulatory requirements that govern the review of
any license renewal application. First is Part 54,
the NRC staff conducts the technical review of the
license renewal application to assure public health
and safety requirements. A second is Part 51, then as
the staff completes routine review of the license
renewal application, focusing on the potential impacts
of additional 20 years of plant operation. Now there
are many programs which are routinely monitored and
assessed plant operations, but the license renewal
review focuses only on those which has the potential
detrimental effects of aging and not addressed
routinely by on-going programs.
Part 54 requires the Applicants who
demonstrate how these programs will be effective in
managing the aging process during the extended period.
Now staff's review consisted on reviewing of the
Applicants' scoping and screening methodology, review
of the aging management programs and review of the
time limited aging analysis identified by the
Applicant. These reviews are supplemented by the site
audits and inspections by the NRC staff. There was
one site audit done on this site and two inspections
governing scoping, screening, aging and management
reviews. Scoping and screening methodology review was
done in two parts. And the first one is a desk top
review which is basically initial review of the
application supporting information and second is the
on-site audit with a team of headquarters' staffs and
regional participants in the review of the on-site
documentation, review of the selected engineering
reports, engineering procedures, design documentation
and discussion with engineering staff.
Incidentally, it was during this audit
first done early in the review process which was in
this case November of 2000 when the staff raised the
issue of interaction of nonsafety systems, structures
and components with the safety systems, structures and
components. And then later on this turned out to be
one of the open items in the SER.
We had several discussions with the
Applicant on this issue. Now this Part 54.29
describes the standards which must be met before the
Committee issues a renewed license. We have talked a
little bit already about the first two items on the
slide. The last one relates to hearing and
intervention on the license renewal application.
There was no hearing on this application. There were
two requests filed for -- filed to petition to
intervene and request for hearing. On January 18,
2001, the Atomic Safety and Licensing Board Panel had
a pre-hearing conference in Homestead, Florida to hear
on the petitioner's standing and the admissability of
their preferred contentions. In the order issued on
February 26, the Board ruled that all -- both parties
have standing to intervene. Neither petitioner
proffered admissible contentions, so their
intervention petitions therefore, must be denied.
The Board ruled that these contentions
raise issues that fall beyond the scope of license
renewal and renewal proceedings. And on March 19th,
one of the petitioners -- he appealed the decision to
the Commission. On July 19, 2001, the Commission
issued an order affirming the Board's decision.
We have participated in several industry
groups on license renewal including Westinghouse
Owners Group and for that developed series of generic
reports intended to demonstrate that aging effects
will be properly managed. At the Subcommittee, they
asked us as a staff to make a specific presentation on
these reports and how staff intends to use them.
MR. ELLIOTT: Yes, Barry Elliott,
Materials and Chemical Engineering. The staff has
reviewed all these WCAPs. The first four, in
particular, are license renewal documents in which the
Westinghouse Owners Group has done an aging management
review to determine the aging effects for the
components that are listed in the titles there for the
reports and they listed the aging effects for the
components and the aging management programs that we
used to manage those aging effects. The staff has
written safety evaluations for each one of those and
they've identified license renewal Applicant action
As far as Turkey Point is concerned, the
staff was a little late in its safety evaluation, so
they couldn't reference the actual staff evaluations
in their report, so they wrote how their components
fit the report and it was during the RAI process, the
Applicant addressed the license renewal action items
and the staff reviewed those and found those
Those first four reports were discussed in
detail at the Subcommittee meeting. The fifth item
which is the WCAP-15338 deals with the time limit
aging analysis for underclad cracks, specifically it
has to do with reactor vessel forgings that were
fabricated using a course screen, a head treatment and
fabrication process and where the clad was applied
with high heat input.
This is in BWR and in Westinghouse plants
and we've had two topical reports on this. This is an
extension of a review that the staff did in the 1970s
on this issue and what they've basically done here,
Westinghouse, is extended the review that they did in
the 1970s using 1990's technology and information.
They've updated the analysis for new technology, new
information and also extended it for 60 years.
These are very small flaws on the order of
10 7-inch, the largest in-depth, the largest we've
ever seen is like 3/10ths of an inch. The run in
length from a tenth of an inch to like two inches.
Very difficult to detect with ultrasonics. Therefore,
we're relying on the analysis to assure vessel
The amount of floor growth here from
fatigue is very, very small. In sixty years, it's
less than a tenth of an inch. We don't expect any
growth from stress corrosion or a very small amount of
growth from stress corrosion, cracking. This is borne
out by the recent event this summer where the crack
grew through the weld, reached the ferritic component
and stopped. The allowable flaw size for this is much
larger than the 3/10ths of an inch on the order of one
in three tenths or one in four tenths of an inch. So
there's a large margin here and for that reason
there's no real concern about these cracks for license
VICE CHAIR BONACA: So this WCAP actually
was used to address one of the open items, right, the
MR. ELLIOTT: They're required, licensees
are required to identify time limit aging analysis.
There's criteria in the rule. This would be one of
them and this was used to address that requirement.
VICE CHAIR BONACA: The reason I'm asking
is that the first four were reviewed, but they were
not referenced into the application.
MR. ELLIOTT: Right.
VICE CHAIR BONACA: Although for the fifth
one, the review was completed before the open item was
addressed. So I think it was credited for.
MR. ELLIOTT: The fifth one was credited
MR. AULUCK: That's correct. The
Commission appeal prepared many internal license
renewal documents under the QA program for use in the
preparation of the application and training of their
staff members. The NRC staff reviewed selected
portion of these documents during our site audit and
scoping AMR instructions. According to the Applicant,
they had several discussions with the previous
applicants and reviewed previously issued RAIs and had
other experts look at the application.
In summary, the staff generated about 215
requests for additional information on this
application which was at that time substantially less
than the previous ones of 300 to 400. And as I
understand, the number is going down, which is
expected as the experience, the quality and clarity of
the application is improving.
As part of this review, the staff review
issued four open items in the draft SER in August of
2001. The first one was seismic II over I interaction
of nonseismic safety-related piping because
safety-related structures and components are known as
seismic II over I. This was the same one that was
identified early in the review process, but at that
time the staff was in discussion with the Applicant to
resolve the issue so we asked the FPL to just wait
until the resolution is reached on the application and
the staff position will be issued and then they can
address that issue. So in the meantime the SER time
came, so we issued the SER with open items.
And I think basically, the Applicant has
gone over the criteria of selecting which portion of
the piping was not included in the first time and then
included later on. All I'd like to add here is since
that time, the staff has issued two positions on this
issue. First one is seismic II over I which was a
narrow scope of nonsafety-related piping closely
related to the safety-related piping. The second
position which is broader in scope, it relates to all
nonsafety-related piping and components. I think in
the future, the staff intends to work with industry to
make it an issue to combine the two positions into
The second open item is -- it relates to
the field-erected tanks internal inspection. The
reason it was an open item during the SER stage was it
was a new program and the Applicant had not addressed
all the attributes identified in our process, so we
asked the specific questions in the RAI and the
Applicant said it's applicable to five tanks, two
condensate storage tanks, two refueling water storage
tanks and one shared demineralized water storage tank.
The Applicant responded in late fall and the response
was unacceptable. So this item was considered closed.
The next item relates to Reactor Vessel
Head Alloy 600 penetration program and Jim Medoff, who
was the lead reviewer for this issue when he was in
Division of Engineering, he will speak.
MR. MEDOFF: Good morning, I'm Jim Medoff.
I'm acting as a backup project manager for the Turkey
Point license renewal application.
Prior to my rotation to the License
Renewal Environmental Impacts Program I acted as a
materials engineer for the Materials and Chemical
Engineering Branch. Part of my responsibilities in
that branch included the review of the Reactor Vessel
Head Alloy 600 Penetration Inspection Program.
Basically what I need to say about the
program is that the license renewal application was
submitted prior to the issuance of NRC Bulletin
2001-01 which was the bulletin written on the Oconee
circumferential cracking that they detected in a
number of their penetration nozzles in a couple of
their units. We issued an open item to address
whether the inspection program for the penetration
nozzles was current with bulletin and whether the --
and whether they were going to update the program to
include the bulletin and FPL's responses to the
bulletin and any changes to the inspection program
that might needed to result from the program.
When the Applicant's response to the open
item came in, we not only reviewed that, but we also,
the Applicant referenced the bulletin and we looked at
the bulletin response as well. Our review of the
responses to both the open item and the bulletin
indicate that FPL is committing to continue
participation in the industry-wide program for
inspection of vessel head penetration nozzles and to
update this program as necessary based on industry
experience and any further studies that the MRP or
EPRI might conduct regarding vessel head integrity
Their response to Bulletin 2001-01
provided revised rankings for the plants and indicated
that they were going to do bare-head inspections of
both Unit 3 and Unit 4 vessel heads. FPL has
completed both inspections and has not detected any
visual signs of leakage or boric acids on the vessel
heads for the units. I will say since Davis-Besse has
been brought up that the NRC issued Bulletin 2002-01
to address the Davis-Besse issue and the impact on
vessel head penetrations in pressurized water reactors
in the industry and that FPL has provided its response
to this bulletin. The response further indicates
FPL's commitment to participate in the program and
update the program as necessary based on inspection
The next open item deals with reactor
pressure vessel underclad cracking. I'm not going to
talk in depth on this because Barry has just addressed
what the contents of the WCAP were and the technical
details of the issue of underclad cracking.
What I will say is that when the NRC
issued the safety evaluation on the topical report,
they required two things. One was for three-loop
plants of which the Turkey Point units are three-loop
plants. They wanted the Applicants to indicate
whether the number of design cycles for the transients
assumed in the topical report bounds the number of
cycles for 60 years of operation in terms of -- we're
talking in terms of fatigue analysis for growth of
The second item that the safety evaluation
indicated was that Applicants referencing the topical
reports as being applicable to their facilities would
need to ensure that the TLAA for the valuation of the
underclad cracks was summarily described in the FSAR
supplement for their application.
FPL submitted responses to the RAIs
relative to both of the action items so we decided
that the FPL took appropriate action and closed the
open item up.
MR. AULUCK: As you recall the
Subcommittee meeting, one of the items discussed was
station blackout and staff was asked how we are
addressing that at Turkey Point. At that time we had
stated that the issue is the position has not been
finalized and when it is finalized it will be
addressed like any other -- addressed by plants
previously relicensed. Since then, the staff position
has changed the final position on that station
blackout was issued on April 1 and at that time they
decided since the position has been issued, this must
be addressed by the Applicant on this application
prior to issuing the license, relicense. So we
communicated this issue to the Applicant the following
day and since that time we are having meetings, we had
a public meeting yesterday. We're trying to resolve
the issue from the perspective that certain components
from the off-site power to the plant should be
included as part of the license renewal.
VICE CHAIR BONACA: This is a change from
the discussion we had.
MR. AULUCK: This is a change from the
discussion we had before and that -- so our intent
here is to resolve the issue and still meet the
schedule date of sending the recommendation to the
Commission. What we are thinking is we'll issue --
the FSAR has been issued with all items addressed. It
will go to the printers at the end of this month, but
we are in parallel, we'll be preparing a supplement to
the SER addressing, focusing on the station blackout
issue and our intent is to complete that in the time
VICE CHAIR BONACA: Okay, let me just for
the benefit of the members who were not at the meeting
at Turkey Point, the issue here is that there is a
preferred station blackout recovery path and the
guidance the NRC provided us before the meeting said
essentially that that would include all the equipment
that collects to off-site power. That includes all
the equipment that collects to off-site power. That
includes, for example, the start-up transformers which
the Applicant has not included in the scope of license
And the Applicant made the case that they
did not rely on off-site power for recovery from
station blackout and demonstrated to us that they can
connect one unit to the diesel generators of the other
units and one diesel generator out of four is capable
of carrying all the loads for both units in case of a
station blackout. They also pointed out that the
experience from the Hurricane Andrew that that was, in
fact, providing for them the most reliable source and
they used it for that particular situation.
Our understanding up to now is that, in
fact, that was the way of Turkey Point to address the
license renewal commitments. Now so irrespective of
that, the staff is asking that Turkey Point includes
all the collection to the off-site power?
MR. AULUCK: Yes, that's why --
VICE CHAIR BONACA: This is a change to
SER that we have in front of us?
MR. AULUCK: That's why we're going to
issue a supplement to the SER and we hope to issue
that shortly, address this issue. Jim Lazevnick from
Electrical will speak on this.
MR. LAZEVNICK: Yes, the Turkey Point has
an alternate AC power source as a means of coping with
the station blackout and essentially the point of
disagreement is whether that source is capable of
recovering from a station blackout. In order to
recover from a station blackout, each plant has to
develop a coping duration based on total loss of all
AC power at the plant and the duration for Turkey
Point was determined to be eight hours and they
utilize an alternate AC source to demonstrate that the
plant could cope for that period of eight hours.
These sources may have capability beyond eight hours,
but the staff has not reviewed them to see if they, in
fact, have that capability and the original
requirements of the station blackout rule, the
definition of an alternate AC source did not address
that capability. It spoke of the alternate AC source
being a means to cope with station blackout for the
period of the coping duration.
So based on other requirements in the
station blackout rule, specifically Section 10 CFR
50.63(a)(1), the coping duration itself is based on
four factors and one of those factors is the probable
time needed to recover off-site power at the site.
The four factors that licensees use to
determine the specific required coping duration at
their plant was developed into licensee guidance and
this guidance was included in NRC Regulatory Guide
1.155 and an industry document that the NRC worked on
the industry with which was NUMARC 87-00.
And all the licensees essentially utilizes
a guidance to determine their coping duration,
relative to license renewal and age-related failures,
it's our view that unless we control a portion of that
off-site power system in terms of age-related
failures, the licensee potentially might need a longer
required coping duration if those age-related failures
were not properly controlled and addressed under the
license renewal rule.
Our final position on this has been that
the off-site power circuits between the switchyard and
the safety buses should be included within the scope
of license renewal. We recognize that the off-site
power system actually is a source that the power
source that extends all the way into the transmission
system of the United States. We feel that this
interface, this portion of the circuitry is an
appropriate part to be included within license renewal
because it's the portion of the off-site power circuit
that feeds the plant and essentially has requirements
only in the plant. It has no transmission system type
requirements associated with this portion of the
VICE CHAIR BONACA: So this says we have
SER with one open item.
MR. AULUCK: Out of this stage, right, and
we met with them and there is agreement, close to an
agreement. We have looked at the draft response and
the Applicant believes they can finalize their
response in the next couple of days and we have agreed
to work with the Applicant and issue the supplement as
soon as possible.
VICE CHAIR BONACA: Well, we should hear
from the Applicant what the Applicant thinks. They
made a case for us and they made a demonstration of
what they consider the ultimate power supply and as
far as our review was concerned, we asked questions
specifically about a standard for transformers in
October and the answer was they are not in scope. And
so I would like to hear what's happening there.
MR. HALE: We still do not agree with the
staff position. We had long discussions with the
staff yesterday. We understand what their position
is. We have nothing but confidence in the capability
of our system and I think we demonstrated that for you
at the simulator. But we understand what the staff
position is. We have spent the last two weeks, I
guess week and a half, based on being informed by the
staff what their position was and that they had
finalized it. So we have put together a response,
draft response which they've highlighted the
additional equipment. There's not a lot of equipment
involved based on the boundaries that the staff is
proposing. They're basically calling for the breakers
and the switchyard that feeds the start-up
transformer, the start-up transformer itself and the
feed into the 4160 switchgear.
VICE CHAIR BONACA: Which I'm sure you
MR. HALE: Yes, this equipment is
maintained under the maintenance rule because the
maintenance rule scoping criteria goes beyond our --
is different than license renewal. The maintenance
rule considers things as trip hazards and that sort of
thing. So this equipment is inspected under the
maintenance rule, but base don our interpretation and
our CLE documents which include our safety evaluation
report, on-station blackout which we reviewed in
detail as well as our design basis documents and our
FSAR, we cannot find where we've specifically credited
restoration of off-site power, but we understand the
staff position. We think we're somewhat unique in
that we have fully capable diesels. In fact, we have
over 400 KW, 300 to 400 KW of excess capacity of a
single diesel, so it's their position. That's the way
they've interpreted it. They've issued it formal and
so we've issued a response to address the specific
VICE CHAIR BONACA: So you have already
issued a response?
MR. HALE: A draft response. They are
reviewing it. Once we factor in their comments, we
will issue it formal probably within the next week.
VICE CHAIR BONACA: Any other questions
MR. AULUCK: Continuing, in February of
this year, a public citizen, Mr. Oncavage, sent a
letter to the ACRS identifying four safety concerns.
The first one relates to the effects of wires on
aging, degradation rates and structural integrity of
the containment structures at Turkey Point. At the
Subcommittee, we discussed this issue and you asked
the staff to make a presentation as it may apply to
some generic implications to the other plants.
Before Mr. Hans Ashar will speak on that,
before he starts, I'd like to go a little bit of when
the issue was first raised and what has happened since
The issue was first raised by Mr. Oncavage
at one of our exit meetings. We had gone for
inspection there and at the exit we provided the
results at a public meeting and Mr. Oncavage raised
this issue that he understands there was some voids
formed at Turkey Point containment during 1980s when
during the steam generator replacement process. So at
the meeting, the Region took this, considered this as
an allegation and gave us a tracking number.
And then they asked the Applicant forward
the concern to the Applicant to respond to the NRC.
The Applicant responded with information to the NRC
and on August 10, Region II sent a letter to Mr.
Oncavage summarizing the results of the review.
But then in December 15th, he sent another
letter to the Region stating that he's not satisfied
with the results of the August 10 letter and NRC
should ask FPL to start testing, looking for voids in
Region II informed Mr. Oncavage,
acknowledging the December 15th letter and stating
that they will respond to him after reviewing the
material again. So on April 5th, last week, a formal
response was issued to Mr. Oncavage, summarizing the
review, independent review by the NRC staff and the
inspection reports, other documents. Thus Region II
considers this issue to be closed for Turkey Point.
Now Mr. Hans Ashar will speak on the
VICE CHAIR BONACA: Now I imagine that the
issue was closed for Turkey Point because the two
identified voids were filled and those inspections
were filled in the containment or was it simply some
statement that said we don't expect to find any more?
MR. AULUCK: I think it was review of
other technical documents at the site and there was a
technical member from Region II, went and spent a week
there, earlier this year to review all the reports and
results and discussions with them.
MR. ASHAR: I am Hans Ashar --
MR. GILLESPIE: Excuse me, Mario, if we
could close this out because I know one of your
concerns was documenting the stuff that was done.
Since we have Region II on the phone, if a person went
to the site that means some place there's an
inspection report which documents what he did. Is it
possible to get that inspection report to the
VICE CHAIR BONACA: Chris? Chris, are you
MR. CHRISTIANSON: Hello, this is Chris
Christianson, Deputy Director, Division of Reactor
VICE CHAIR BONACA: Did you hear the
MR. CHRISTIANSON: Is there a possibility
to get a copy of the inspection report? We did not
document this in an inspection report. We documented
this as a memo to file in the allegation folder.
MR. GILLESPIE: Okay, it's still the same
question. Is it possible to get a copy of that,
MR. CHRISTIANSON: Mr. Auluck can forward
it on to the appropriate person.
MR. GILLESPIE: Okay, we'll contact you
off-line, Chris, and we'll get a copy of it and get it
to the right people on the Committee and that might
provide some closure to the issue for Turkey Point and
that might be beneficial.
VICE CHAIR BONACA: Yes, just to
understand what was done to assure the issues of
concerns with additional voids in the containment was
MR. DURAISWAMY: Mario. Raj, you sent
another letter to Oncavage on April 5th?
MR. AULUCK: Yes.
MR. DURAISWAMY: From here? From the head
MR. AULUCK: No, from the Region.
MR. DURAISWAMY: From the Region.
MR. AULUCK: Because Region II considered
the December 5th letter from Mr. Oncavage as the end
of the follow-up allegation.
MR. DURAISWAMY: Yes.
MR. AULUCK: So they tracked it and they
responded to that to him and just closing the loop.
The letter is April 5th from Region II to Mr.
MR. DURAISWAMY: You guys don't have a
copy of that thing?
MR. AULUCK: Those are allegations --
MR. DURAISWAMY: I know what the
MR. LAZEVNICK: I think I have copies of
MR. AULUCK: They can be made available.
MR. GILLESPIE: This is why I say when you
put stuff in the allegation system, it's a very closed
system, even though this individual didn't ask to be
treated that way and so we can deal with it and get
you copies of it.
VICE CHAIR BONACA: But before the
allegation issue, there was a finding, was an open
finding. There was an evaluation being done. There
was a response by Bechtel. There were people that
came in with concrete and poured it to fill those --
I mean there were things that took place and in
addition to that, if anybody had any question, they
would have looked someone else to find are there other
voids. That's -- I would expect there would be some
documentation that says yes, we did the following
steps and then the committee can review it and feel
confident that something was done that we can state
today those containments were taken care of and there
are no voids in containments to the best of our
knowledge within the limitation of detection and so
on. It's not only the file on the allegation, it's
just simply the paper trail that led to the
documentation of the actions taken to deal with the
MR. GILLESPIE: And I'm hoping a memo to
file actually references it reviewed this, reviewed
that and then when you get those things, those things
contain the subject matter and address these actions.
I'm just not sure having not seen the file
how it strings it together, but that's -- the starting
point, I hope would be the memo to file where they
said okay, we reviewed all the existing information
and existing actions taken to date and it appears to
be satisfactory and I hope there's some reference to
what those other documents were so we have a -- we
should have the trail. It's just it's in a system no
one has easy access to. So we'll take back the idea
of working with Region II and copying the paper trail
and trying to get it to you in the very near future
VICE CHAIR BONACA: We asked for those in
Florida City. We asked for -- so that -- and Region
II was there, present during the meeting and when we
asked for this information.
MR. GILLESPIE: Yes, because if this was
followed up in the 1980s and there was an inspection
report from the 1980s, I'm hoping that research was
done that we can just pull it together in this one
memo to file was kind of the cap on top of that
MR. HALE: Dr. Bonaca, this is Steve Hale,
Florida Power and Light, we interfaced with the
regional -- the fellow that came down to do the
investigation. There were LERs on this event. There
was initial LER plus supplements. There was also two
inspection reports which documented the closure of
those two LERs and the individual came to the site,
looked at that information. So I think this memo to
file or whatever should have all the specific
documents, but I can tell you for sure because we were
supporting him and he went in and actually was looking
at the original pours, concrete pours documentation on
the testing that was performed on that concrete, so he
did a very exhaustive investigation, just based on the
interfaced we had with the fellow when he was at the
VICE CHAIR BONACA: Okay, so we'll see for
MR. ASHAR: I am Hans Ashar from
Mechanical, NRR. I had read your transcripts of my
tech. team and concerns expressed by various members
of the SEI subcommittee and based on that, I want to
address only the generic implication at this time as
to what I think about it because we had a very short
time to prepare for any in-depth research, but I'll
try to tell you as much as I can gather from my own
experience as well as other people's input into what
Now first thing, what I want to refer to
is are the worse possible. First thing I want to
emphasize is this, that having voids in concrete
construction, in general, there is commercial
application at nuclear power plant is not an
acceptable way of constructing any structure. It is
not an acceptable matter. People try very hard to
make sure that the concrete that they pour is being
consolidated very well through vibrators and the
construction joints are being formed in such a way
that this kind of voids can be avoided.
I also would like to let you know that it
is possible, it is possible that some of the plants
may have existing concrete voids. Now my own
experience, when I was a specification engineer at
Burns and Roe and I was at Three Mile Island, Unit 2,
and at that time we heard about voids in ring guard at
Three Mile Island, Unit 1 and the United Engineers
Construction was the constructor on that one and their
engineer had found the voids and they took corrective
action after that. So what I would like to emphasize
here is that the way the quality control, quality
assurance works in the industry and it worked at that
time, at least, I know because ACRS had very strict
quality assurance criteria. It had been in force
because people wanted to keep their license and so
there were attempts being made to award this kind of
work being persistence in nuclear power plant
Now somebody might say that that means
that there are no voids in it. I wouldn't say so. I
thin in spite of all the precautions there could be
sometimes back down in some other thing, like a
concrete venting plant, the pumping of the concrete,
the vibratory spin work on the particular areas, voids
might be there in some of the plants. Okay?
Now as I said before, core requirements
require concrete voids -- impact of voids. What could
happen to the containment if there are voids present?
Now in a very narrow way I would say there will be a
reduction in thickness of the thick part of the
sections of concrete.
MEMBER POWERS: Before you go on to the
impact, your slide says voids can occur where
vibrators can't reach.
MR. ASHAR: This is why I explain to you
in much more depth is to what are the factors that can
influence the existence of voids.
MEMBER POWERS: There are many other
causes of voids.
MR. ASHAR: Please?
MEMBER POWERS: There are many other
causes of voids in the concrete.
MR. ASHAR: Yes. Well, in order to avoid
voids in concrete construction, in general, the first
thing to make sure that the construction joints that
they are going to put in are in the right place, so
that you can ensure that the oldest areas, very older
concrete are accessible from the formwork. And the
vibrators can reach into those areas. These are the
items being made all the time. As I told you in my
experience, the voids were in the ring girder of the
containment construction and the ring girder is a very
thick area. It is a liner plate coming down and again
the voids were in the area of the liner plate was
touching the concrete area. But they took out all the
concrete. They rehashed everything. They put new
concrete in there to make sure there are no voids
existing in that particular instance.
The other two you heard about were the
Turkey Point and Limerick. So yeah, voids can occur
in various places and due to various reasons.
MEMBER POWERS: I mean what I'm struggling
with here is for this particular instance, you got an
individual saying there are voids in the concrete.
How do you know there are not voids elsewhere? The
guy that placed -- the architect/engineer went in and
said yeah, there were voids in this concrete and
here's how we explain them. He said it's because the
vibrators didn't get there. That seems very
convenient to me.
MR. ASHAR: Well, it explained to you. I
put one bullet, vibrators can't reach. It is not the
only thing, okay? But the basic thing is to make sure
that the old areas to be concreted out are filled up
with concrete to make sure of that. And then to
consolidate the vibrators to beat the -- now sometimes
it can happen, the water may be a little higher or the
weather might be such that the water can bleed. When
it bleeds what happens the calcium hydroxide from
concrete gets into that area instead of filling of
with full concrete and integrate. Only the water
part, calcium hydroxide stays in that area and it
would look like you filled up the things. As the time
goes by that water starts evaporating and the void
So those things are possibilities. I
would not completely --
MEMBER POWERS: What I'm trying to
understand is the firm went in and they came up with
a hypothesis of why they had a void in Turkey Point.
It was very convenient and it would not be something
that would extend out of places in the containment.
What did the staff do to look and see if there was
alternate explanations for this?
MR. ASHAR: Well, I will ask open forum
for other people to answer to this particular
question. As I said, the construction practice during
that time, the time this plant was being built were
such and the quality assurance requirements were very
stringent because I know from my own experience on
this side of the fence I was not with NRC. I was with
consultants and at that time, as a matter of fact,
after I heard about that void and the cause for those
voids, I wrote my specification for Three Mile Island,
Unit 2 in such a way -- as a matter of fact, it is not
very common for a specification writer to write about
where the constructors would put their construction
But in our case, we did write it. Okay,
because we were concerned about the voids in
construction of Three Mile Island, Unit 2. That's why
-- so people --
MR. GILLESPIE: Dana, let me see if we can
put our package of documentation together. I think
this is getting to the point where it may deserve a
different -- I'm going to suggest a separate meeting.
VICE CHAIR BONACA: The other point I
would like to meet, we are here now, general
considerations here. I think that is on the right
track. The issue is you find a void under the hatch
in concrete. So now you say well, let's see if this
is just one of a chance and you go to the next
containment and you find you have a void in the same
And this seems to be almost like it's a
design feature for this kind of containment, I guess.
It's present in two, let's see how many you've got
where you have a spot. I think you would want to go
beyond. Now typically, you have mechanisms by which
you raise an issue that could be, I thought, would be
Part 21, but Bechtel says oh, it's okay, the
containment is too capable, so it's under Part 21.
I'm sure there was a paper trail by which the issue,
the potential impact of being a generic issue was
evaluated. I mean normally the agency is very
aggressive in pursuing these kind of issues. That's
why we've been looking for how did we address, how do
we get confidence that other containments of Bechtel
design do not have the same voids in the same location
and other containments in general do not have that.
And that is really what we're looking for when we
asked for that information in March down in Florida
City. And we really haven't gotten the information.
MR. GILLESPIE: And I think that's exactly
what we need to pull together. Because now we're all
trying to project what happened in the mid-1980s.
VICE CHAIR BONACA: Yes.
MR. GILLESPIE: And I'm having a tough
time myself remembering what I did last month and
these people weren't there. How well were we
documenting stuff in the mid-1980s? We need to pull
the inspection reports, look at what the people looked
at, look at what the fellow, the inspector from Region
2 that went in and re-reviewed the issue and then ask
the question and look at other records and say now did
we take that? What did we do with it generically? I
just don't know.
I think we're talking about something in
1985 or something like that, maybe and it's 17 years
old at this time. I like to assume the staff did the
right thing. We did pursue things aggressively at
that time. I just don't have the documentation in
front of us. We need to pull it together.
Someone else from engineering --
MR. KUO: Goutam Bagchi, he's going to
make a presentation on related issues.
MR. GILLESPIE: But I would suggest the
opportunity to come back would be also fine with us.
VICE CHAIR BONACA: My proposal will be if
we feel, first of all, this committee will decide
whether or not we feel confident that the issues
themselves for Turkey Point, so we can focus on the
license renewal for that plant. If we feel it is
dealt with properly, then we can say let's concentrate
on that. That will result, probably with separate
letters requesting that we look at the genetic
implications, how they were handled for other units
and that would open the path.
MR. GILLESPIE: Yes, and that would be
fine. I think we can get the Region 2 records pretty
quickly for you for Turkey Point to kind of close that
documentation issue and I'll tell you the truth. I
feel more comfortable coming back to talk about the
generic issue versus trying to do something where
we're potentially kind of patching some things
VICE CHAIR BONACA: I agree with you one
MR. GILLESPIE: Goutam did have some
thoughts of some basic engineering he covered with me
earlier of about why this is a safety issue we can now
look at in an orderly way and not necessarily assume
we didn't look at it 17 years ago, but let's see what
we decided then and what the basis was.
So I'd suggest coming back and let Goutam
finish what he's going to go and we'd be happy to come
MR. ASHAR: If Goutam is going to speak,
then I won't say anything --
VICE CHAIR BONACA: I would like to hear
from the members, is it acceptable with you that we
put this issue here, which is generic, separately and
address it later or would you like through the
MR. BAGCHI: It's a very quick
presentation. I just wanted to share with you some
idea of load sharing, what is it that is unique in the
containment structure of design.
VICE CHAIR BONACA: Okay.
MR. BAGCHI: And I think there is
something unique in the design itself that gives it
the robustness and the ability to withstand the design
And concrete, as you know, takes
compression. It cracks and it doesn't take an tensile
load and it maintains -- the effective purpose of the
concrete is to maintain the reinforcing bars in the
Reinforcement carries all the load.
Post-tensioning tendons keep concrete in compression.
And very high quality, .2 percent ultimate elongation,
ductile liner plates are provided as the leak-tight
Design basis load is internal pressure,
due to the postulated accident load. Containment
structure goes into tension. Concrete cracks due to
tension. Reinforcement bars take all tension loads
and the liner plate maintains the leak tight
integrity. If there is any local void, it deforms
plasticly and then expands and bridges the gap, as we
have experienced in the reactor vessel head at one
At the shell-mat and shell-dome junctions
bending moment puts concrete into compression. But as
you know, this was not the area where the concrete
void was found. The concrete void appears where there
is congestion of reinforcement and special provisions
are sometimes lacking when putting in concrete. And
this is the area of the ring girder near the equipment
So only in those two junctions the
concrete is put into compression. By code
requirement, concrete is under reinforced.
Crushing failure of concrete is prevented
by code provision because the reinforcement has to
yield first. Redistribution of load around any void
provides the necessary strength. Structural Integrity
Test would reveal locations of unacceptable voids by
bulging, spalling or local failure. Every reinforced
concrete structure passed the Structural Integrity
Test satisfactorily the very first time.
There are requirements to make predictions
of deformations and measurements are made,
observations are made, examinations are made
afterwards and they have all been within the predicted
Post-tensioning puts the highest load
during construction. Any weakness in concrete shows
up at this time as we found in the delamination of
dome. It was a weakness in the design. Reinforcement
bars were not provided and later on they learned their
Containment weakened by pervasive voids
will not pass the SIT, the Structural Integrity test.
So my conclusion is that the unique design
of the containment structure, the high quality of
construction, no matter the fact that there were voids
found and these are construction areas those are
imbedded in the code related factors of conservatism
and the allowable stresses and so on, there are going
to be voids and in a very thick structure 4.5 to
5-foot thick walls, you're not going to easily find
the voids. If they were found easily, they will be
taken care of and if there are voids, as I tried to
point out, the load path and the behavior of the
concrete is such that the reliance is not on the
And this is -- the inside I just wanted to
share with you and I feel that the containment
structure is extremely robust as people have seen from
the tests, although in the tests you wouldn't have
expected any voids, but in a scaled condition,
microvoids may well have been there in those third
scale, quarter scale test models. But it's the load
and the design of the structure that provides us with
the assurance that there will be good performance
function, certainly, after the design basis load and
way beyond that.
MEMBER SIEBER: I have a question. I
recall during -- having witnessed a couple of
Structural Integrity Tests of concrete containments
that one of the steps was to find and map the cracks
that appeared. Was that common practice for every
MR. BAGCHI: Absolutely.
MEMBER SIEBER: That would reveal the
presence of the voids because the cracks would appear
around the area of the void as the loads redistribute
themselves. Is that correct or not correct?
MR. BAGCHI: I would like to agree first
and then take away some comfort that I've agreed with
you. If it's 4.5 foot thick wall and if this void is
adjacent to the liner plate, you're not going to see
MEMBER SIEBER: That's right, that's
MR. BAGCHI: This is a conservatism --
MEMBER SIEBER: You will see it on the
inside if it's adjacent to the liner because there
will be a dimple there.
MR. BAGCHI: It has to be a very large
void to do that.
MEMBER SIEBER: Yes, it does.
MR. BAGCHI: yes sir.
MEMBER ROSEN: So the conclusion is small
voids you won't see, but they don't matter because the
loads are being taken by the reinforcement steel and
large voids, if they have occurred, you would see.
MR. BAGCHI: Yes, that's my contention.
MEMBER ROSEN: In the performance of the
MR. BAGCHI: If you allow me to
characterize what kinds of voids, I would not consider
as extremely critical is something in the order of a
MR. KUO: If I might add to it, the large
void, if it is located in critical locations, in other
words, it's a stressed location, void stress location,
you will see during the test, as a result of the test.
MR. BAGCHI: That point about crack, map
cracking, mapping the crack is really intended for
MEMBER SIEBER: That's right.
VICE CHAIR BONACA: Thank you for
MR. GILLESPIE: Mario, now what I'm hoping
is that we'll find that back in the 1980s someone as
smart as Goutam wrote that down as a basis and I don't
know if we will -- we need to look, but that's part of
the reason I think some things didn't happen and how
well did we document things in our actions, we need to
do some investigation.
VICE CHAIR BONACA: Okay.
MEMBER RANSOM: A point of clarification,
in Turkey Point, is it known that there are voids and
do they know how big they are?
VICE CHAIR BONACA: Oh yes. They found
voids, as you know.
MEMBER RANSOM: They have found them?
VICE CHAIR BONACA: Well, they found them,
yeah, sure. That's how the whole issue came up. They
found voids under the equipment hatch when they were
replacing the steam generators. They had to take off
the hatches because they were not large enough. As
they removed them, they found these voids right under
because of the complexity there and the amount of the
rebar that --
MEMBER RANSOM: So those presumably were
remediated when they repaired them.
VICE CHAIR BONACA: Absolutely.
MEMBER RANSOM: This just led to suspicion
that there may be other voids?
VICE CHAIR BONACA: The concern of Mr.
Oncavage was are there other voids in the containments
and so we expected to find that there would be some
documented trail that said yeah, we looked at it or we
tested or we performed some assessment of the type
that we received right now that gives us confidence
that probably there are no voids or there are some
that are not significant to the strength of the
containment. And we haven't found yet this paper
trail. That's what we're looking for.
The other issue is the genetic
implications. If you find this kind of issue in one
location, in one containment and then you go to the
next one and find the same thing as happened there, it
tells us that very likely there is going to be
something similar under the hatch in some other unit
and so one will have to understand the significance of
no remediation of that void and again, that may be
some analysis done of this type that is sufficient,
but we haven't seen any of that, so we're looking for
how the generic implications of the issue were
MEMBER POWERS: I'll point out, Mario,
that there were in construction of the McGuire plant
that they found large voids in the concrete when they
placed, had nothing to do with where they put
vibrators. There are lots of reasons for voids.
VICE CHAIR BONACA: Yes, sure, the timing
of pouring of the concrete, the density, the liquidity
of it, how it flows.
Okay, so are there any more questions?
Your considerations were still related to each other's
presentation we had on the concrete, right?
MR. ASHAR: Pardon me? What's your
question? I didn't get you.
VICE CHAIR BONACA: I'm saying what is the
remaining portion of your presentation?
MR. ASHAR: Yes, I can finish up with a
few lines. Now Goutam very well described this as to
the robustness of containment and how the voids cannot
be that much of a significance in integrity of the
containment at least to resist the design basis
This is exactly what Goutam pointed out in
the initial structural integrity testing, periodic
leak rate testing being performed in the containment.
Containment -- they also conformed intended function
of the structure. Now one other question that I'd
seen being asked was what would be the impact on LERF.
What I would say more succinctly is condition probably
of containment failure. That would be affected if
there is any point in it.
Now my judgment, it's my own judgment on
this particular issue is that there are two model
tests being performed at Sandia. One in 1995 or so on
reinforced concrete model and one in 1999 on viscous
concrete model which was being financed basically by
NUPAC in coordination with the NRC.
On the first test, what I want to point
out is the failure of the model at 137 psig or so, and
at that time the concrete was quite a bit cracked and
heavily cracked, but at that time they did not go all
the way up to the failure of the complete structure.
They stopped when they saw the leakage was too high,
but there was some stiffness left still at that time
and now in the later test in viscous concrete
containment in 1999, they did go a little farther than
just leaking criterion. It was considered the
containment fate, but then they went a little bit more
and they saw that there was few strength left,
stiffness of the concrete to hold the liner in place
and I think they went about 10 psig, more than what
would consider as a failure, not the ability to -- so
that was my judgment that the effects of LERF of the
voids, in general, would not be that significant.
CHAIRMAN APOSTOLAKIS: But the conditional
containment failure probably in NUREG 1150 is
extremely uncertain. I mean it's always between 0 and
MR. ASHAR: Yes.
CHAIRMAN APOSTOLAKIS: I wonder, does it
include the possible presence of voids?
MR. ASHAR: Yes, this is what happens.
Okay, that if the structure were intact completely,
okay, the ideal structure, you find out one fragility
curve occurred for containment probably so there is an
FSAR and ordinate probably to a failure, FSAR used as
pressure as a parameter. Okay, that will give you the
medium design pressure. Point 5 failure could occur.
That was taken in the LERF calculation later on for
structural containment. Now if there is a
degradation, a main degradation is not concrete, but
the liner. In the case of concrete containments,
liner would be the prime candidate for reducing the
effectiveness of containment because it would leak.
So if there is liner degradation of high level, then
you can shift your facility curve in such a way that
it meets with the damage assessment that has been
CHAIRMAN APOSTOLAKIS: My question is if
I look at the -- not the fragility curve, but the
final results of the NUREG-1150, they have very nice
figures with various sequences and then the
conditional containment is computed.
MR. ASHAR: Right.
CHAIRMAN APOSTOLAKIS: And this is a very
uncertain quantity. It goes from 10 to the minus
something, all the way to .9 sometimes or even 5.
MR. ASHAR: Right.
CHAIRMAN APOSTOLAKIS: So that's extremely
uncertain. So I don't know what it means.
MR. ASHAR: But normally the IPEs are
performed with little more preciseness than those --
CHAIRMAN APOSTOLAKIS: You mean the IPE is
no better than your NUREG 1150? I doubt it.
MR. ASHAR: Oh no, no, no. What I'm
saying that the uncertainties which are being in NUREG
1150 considers number of uncertainties. When you
start in plant specific IPE, that means they have
precisely characterizing the sequences and then
putting the -- they also have uncertainties, but not
as much as what we see --
CHAIRMAN APOSTOLAKIS: Yes, but the IPEs
also did not spend as much effort on the level.
MR. ASHAR: I'm not saying I would put a
lot of --
CHAIRMAN APOSTOLAKIS: My question is in
the original 1150 studies, was the possible presence
of voids included? You don't know?
MR. ASHAR: I know that it was not.
CHAIRMAN APOSTOLAKIS: Oh, it was not.
MR. ASHAR: It was not. None of the
damage condition or anything was considered in the
VICE CHAIR BONACA: That's why I made a
distinction between the design pressure that I
believe, this condition is still allowed to meet as a
requirement of the tech specs versus the ultimate
containment. So we don't know and typically we are
looking at penetrations as the weak link or something
of that kind and here you have an unknown.
CHAIRMAN APOSTOLAKIS: Is the effect not
significant because we are so uncertain to begin with
what can happen?
MR. ASHAR: Well, only from the existing
condition. It's not related to the insignificance.
MEMBER FORD: Mario, you also managed to
go about how we felt about this particular issue for
Turkey Point as opposed to generic issues. I feel
really uncomfortable. In all of the rest of the
license renewal examinations we've been asked to
comment upon, we've had detailed documents, ANPs that
we can make good scientific judgments, our own
independent judgments. Here we're hearing engineering
judgment, anecdotes. We've got nothing to go on. So
I don't see how we can make any advice or judgment on
this as an issue.
CHAIRMAN APOSTOLAKIS: Yes, I think this
kind of discussion will take place in the afternoon
VICE CHAIR BONACA: But I would like to --
I know, we know pretty much what we heard already. My
sense is that we should not write a report now. There
are two issues here that need some closure. One is
the station blackout issue. Although we know that the
plant is taking a position, a direction of fulfilling
the requirements, it is important for us as a
committee for us to understand is it a capricious
requirement in addition to what already they are doing
at Turkey Point? Is it essential? I think we need to
reflect on that and review it. Second, we also now
need to look at this paper that will be provided to us
and so my suggestion would be that schedule one hour
meeting at the May meeting and we look at those two
issues and then resolve them at that time. That will
give us at least time next three weeks --
CHAIRMAN APOSTOLAKIS: Well, we have time
this afternoon to discuss the letter. We have already
agreed that there will be some additional information
provided to us with a possible presentation.
VICE CHAIR BONACA: Yes.
CHAIRMAN APOSTOLAKIS: We're already
VICE CHAIR BONACA: I was attempting to
say in a way that you're right and a means of probably
doing some closure, but I think that for us to jump to
something today is going to make it enough.
CHAIRMAN APOSTOLAKIS: Okay. So I'm
wondering now is there anything else we need to
discuss right now?
VICE CHAIR BONACA: Any other questions
that members would raise?
MEMBER LEITCH: Not related to concrete,
but I have a question about there's a figure in the
environmental report. It depicts a 6-mile radius and
usually when you see these figures they have a 10-mile
radius. I don't know that this relates to emergency
planning, but I'm just wondering --
CHAIRMAN APOSTOLAKIS: Which figure is
MEMBER LEITCH: Page 2.1-3 in the
I'm just wondering is there any
implication? Does Turkey Point have a 10-mile EPZ
like everybody else?
MR. HALE: Yes, we do. Steve Hale,
Florida Power and Light. Yes, we do. That's not
intended for emergency planning.
MEMBER LEITCH: Okay and my other question
is can someone tell me what's the CDF and LERF for
these units and are they different from one another?
MR. AULUCK: We'll have to get back to
MEMBER LEITCH: Okay, I'm just looking for
the CDF and LERF and are units, Unit 3 and 4 different
from one another.
MR. HALE: Unit -- I can't cite the
specific numbers, but we're not an outlier or anything
like that. We have reasonable CDF numbers. I can't
speak to the specific numbers.
MEMBER SHACK: Well, actually, your
numbers reported int eh IPE are highest of anybody,
but the discussion at Florida was that, in fact, that
your updated PRA has numbers that are much lower. So
I think it's close to four times 10-4 in the IPE and
the reported number was like 1 times 10-5, some PRA
person gave this in Florida, but that hasn't been
CHAIRMAN APOSTOLAKIS: So how did it go
from four times 10-4 to 1 times 10-5?
MEMBER SHACK: Divide by 40.
MEMBER ROSEN: This is fairly typically
MEMBER SHACK: The discussion was that he
was making some very conservative assumptions when
they did the IPE.
MEMBER ROSEN: That's the reason. This is
fairly typical, you see it in most PRAs that the very
first ones are quite a bit higher than the more
sophisticated ones that are done over time.
CHAIRMAN APOSTOLAKIS: So that's something
that we have to discuss.
VICE CHAIR BONACA: Any other questions?
MEMBER POWERS: But George, I'll remind
you the number is totally meaningless because it only
considers operational events.
MEMBER ROSEN: Because of what, Dana?
MEMBER POWERS: It only considers
operational events. It doesn't consider shutdown.
MEMBER ROSEN: Plants generally have a
shutdown assessment that considers the risk during
shutdown which is additive to the internal events.
It's not meaningless, it's just part of the question.
CHAIRMAN APOSTOLAKIS: Okay, any other
questions for the presenters?
MR. AULUCK: Do you want us to go over the
other concerns of Mr. Oncavage?
CHAIRMAN APOSTOLAKIS: Well, it's too late
VICE CHAIR BONACA: Let's just cover
MR. MEDOFF: This is Jim Medoff again,
Backup Project Manager for Turkey Point. Basically,
when Mr. Oncavage sent his letter in to you, we did an
independent review of its concerns and basically we
categorized them into voids which we just discussed.
The effect of hurricane windspeeds in storm surges,
unsafe operation of the units. He also went into
concerns about the effect of terrorist attacks on the
safety of the plants and he had a concern about spent
Basically, what we did is we called up the
National Oceanographic and Atmospheric Administration
to discuss the hurricanes. Hurricane Andrew basically
was one of the most severe hurricanes ever to hit the
Atlantic coast. It had wind speeds of 149 to 150
miles per hour which puts it in Category 4, but with
gusts above that which put the gusts into Category 5.
The storm surges for the Hurricane Andrew were of the
order of 17 feet maximum. As Steve Hale has
indicated, the Florida Power Light units, the Turkey
Point units, vital equipment are designed to withstand
storm surges above 22 feet and all of the vital
equipment such as emergency diesel generators, the
reactor vessel, etcetera are put in design category 1
structures and they're designed to withstand
differential pressures created by the hurricane of the
order of 225 psi without any deformation of the --
MEMBER ROSEN: Now you said above 22 feet.
I don't think that's what he said. I thought they
said it was up to 22 feet.
MR. MEDOFF: No, the location of the vital
equipment is at 22 feet or higher.
MEMBER ROSEN: Right.
MR. MEDOFF: The maximum hurricane -- in
our discussions with NOAA, the maximum surge ever
recorded for the Atlantic Coast was 20 feet and that
was for, I think, it was Hurricane Hugo on the North
The maximum storm search for Hurricane
Andrew was 17, so the vital equipment at Turkey Point
are designed at levels currently to withstand the
current storm surges for Category 5 hurricanes.
That's not to say that you might get a
really, really severe hurricane to create a storm
surge above 22 feet, but I think the probability, my
educated guess on that would be the probability would
be low given the data that NOAA had given me in our
discussions with them.
The next one is the effective terrorist
attacks on --
VICE CHAIR BONACA: We know that that's
MR. MEDOFF: And the last concern was the
-- Mr. Oncavage was concerned that they were going to
expand the spent fuel capacity in the spent fuel
building. Typically, they're covered by tech specs if
they even come close. FPL will submit action to
MEMBER POWERS: It strikes me that the way
you have approached storm surges is a bit different
than we usually approach natural phenomena, especially
when you're prognosticating for another 30 years or
Don't we usually say what's the
probability of storm surges of various elevations over
MR. MEDOFF: Not being the expert in that
area, I'm not going to say yes or not, but I would
expect that to be the case.
MEMBER POWERS: Taking particular
incidents since it got to 17 feet, it could get to 20
feet within the last 100 years we've had as high as 20
feet and this is at 23 feet strikes me that you're
very close and I certainly listen to people, not too
intently, that tell me that the weather is such that
hurricanes are going to become more vigorous in the
future. I know that despite the prognostications last
year was a particularly hurricane deficit year, so
maybe their predictions are not too good. But it
strikes me that you need a little more quantified
treatment of this.
MR. AULUCK: I think the design of Turkey
Point can handle Category 5 hurricanes. Steve, do you
want to add?
MR. HALE: Well, one, I think this is
beyond Turkey Point, I mean if the issue is that
historically in establishing your natural phenomenon
and what you address in your SAR, you go back, I
believe 100 years or something like that and then you
establish some conservatism on top of that in the
design of your structures.
We are fully confident in the design of
our structures of accommodating our design basis
hurricanes which had margin well above 100 year storm
that was identified. So I believe that in considering
storms in the future, would be more in the generic
arena than I would a specific Turkey Point issue.
MR. AULUCK: So, in conclusion, we have
completed our review. As I understand we owe you
information on the documentation, how Region 2 closed
the issue on voids. It's available. It's just a
question of getting it to you. The staff
recommendation will include the resolution of the SBO
issue and applicant has met all the requirements
required by 54.29.
VICE CHAIR BONACA: So mean the second
bullet is not correct, of course, at this stage. I
mean there's one open item and we will --
MR. AULUCK: All open items identified in
the SER were resolved. This is a new emerging issue.
VICE CHAIR BONACA: You're right.
MR. AULUCK: It just came last week and
that's why I made a separate bullet in the staff
VICE CHAIR BONACA: Thank you. Any
MR. KUO: And this concludes the staff's
presentation on Turkey Point license renewal
application review and we will take two actions back.
The first one is try to put together the paper trail
on the concrete voids inspection from Region 2. We
will try to get as many copies as we can.
The second action is to check the CDF and
LERF values for the containment.
VICE CHAIR BONACA: There's a third one
which is the station blackout.
MR. KUO: Station blackout. We issue the
staff position on April 2nd on station blackout and
the issue has been there for quite a few months. We
have issued the first station blackout proposed
position back in November of last year. Since then we
have met with NEI and the industry three times and
this position was supported by the NEI and the
VICE CHAIR BONACA: On the other hand, the
staff was present during the walkdown of Turkey Point
and the demonstration of the alternate path and there
was no mention that this requirement would come up, so
I think it's important for us to review it to
understand if the requirement is appropriate.
MR. KUO: Sure, sure.
VICE CHAIR BONACA: Because I was very
convinced by what I saw there and that it was
adequate, so I would like to just --
MR. KUO: I understand.
CHAIRMAN APOSTOLAKIS: All right, thank
MR. HALE: Just for my own benefit, so I
understand these issues. I guess right now the
current schedule for the Turkey Point license shows a
letter from ACRS by -- what is it, April 19th?
MR. AULUCK: The 19th.
MR. HALE: And so what I understand that's
not going to occur?
CHAIRMAN APOSTOLAKIS: It looks like it
MEMBER POWERS: Let's make very clear that
that's somebody else's schedule. That's not our
MR. HALE: Oh, I'm not -- I'm not -- don't
-- just for my own benefit in terms of where we stand
with our license review.
CHAIRMAN APOSTOLAKIS: There is a
probability that it would get it, it went down by a
factor of 40 as a result of today's --
MR. HALE: Is there anything that we can
do? Certainly, we can get our hands on the
information ourselves with regards to the concrete
containment. In fact, I brought quite a bit of
information with me today. If there's some way with
regard to the concrete void issue, we can resolve it
by inspection of the information I have with me.
The second item was with regards to
station blackout. We met for an extended period of
time yesterday with the staff and have come in general
agreement to the approach. We also have that
information available. And certainly, the CDFs for
the plant can be obtained very quickly.
MEMBER KRESS: I propose that the
Subcommittee Chairman sit down with him and go over
that information and see if it's enough to satisfy the
Subcommittee Chairman and then he can report back to
the full Committee.
VICE CHAIR BONACA: There are Subcommittee
member concerns, however, raised right here and I want
to make sure that we satisfy those. I'll be certainly
willing to sit down and review what you have and still
there are a number of issues here, it seems to me that
put the Committee under pressure to come to a
determination when these issues are raised in Florida
City, with the exception of the session blackout. And
so it concerns me in the months, the elapse of time we
haven't been able to find --
CHAIRMAN APOSTOLAKIS: Okay, why don't you
then interact with the licensee and report to us maybe
at 5:30 where we have some time to discuss this?
VICE CHAIR BONACA: I'll do that.
CHAIRMAN APOSTOLAKIS: And see how the
Committee members feel then about writing a letter.
MR. HALE: I would like for Dr. Ford, too,
because he's the one that's voiced concerns with
regards to -- if possible --
CHAIRMAN APOSTOLAKIS: Yes. We can do
these things. But you have to remember, the letter is
from the full Committee.
MR. HALE: I understand. I understand
fully. I just want to make sure that I have brought
information today and anything I can do to facilitate
your review I would like to do that.
CHAIRMAN APOSTOLAKIS: Certainly. Okay,
thank you all very much. We'll recess until 11:30.
(Off the record.)
CHAIRMAN APOSTOLAKIS: We're back in
session. The next topic is Advanced Reactor Research
Dr. Kress is the cognizant member.
MEMBER KRESS: Thank you, Mr. Chairman.
The staff is diligently working on a comprehensive
research plan for advanced reactors. We have a draft,
a proposed draft, copy of it which is incomplete. So
I guess we could consider this kind of an interim
briefing and I guess we're looking for any early
feedback from us that we might be able to give them
either orally now or perhaps in a letter. So with
that minor introduction, I'll turn it over to Farouk.
MR. ELTAWILA: Thank you, Tom. You are
exactly right that this plan right now is in a very
early stage, and as a matter of fact, we have not
received the input from the user office like NRR and
NMSS, so it's a work in progress and we'll continue to
update this plan and we envision that we will be
coming to the ACRS at Subcommittee level in the
different areas of this program. But for the time
being, the staff developed that plan to identify the
issues that will be needed to develop the safety
criteria against which this advanced reactor design
will be judged.
The plan is extremely comprehensive and
includes a lot of information. Some of this
information might already exist through international
research that's conducted somewhere else. it is also
available through the vendors and the old history of
gas-cooled reactors, for example.
So the plan should not be construed as
research activities that the Office of Research is
going to be conducting. As a matter of fact, a lot of
the information that describes in the plant would be
the responsibility of the Applicant of the new reactor
design to try to make the safety case. So we will be
receiving a lot of information from the industry on
But regardless of where the source of
information is going to come from, whether it's coming
from NRC, from international cooperation or from the
vendor or the Applicant himself, NRC will have the
best information available to make its regulatory
MEMBER LEITCH: If it's not intended to
identify research, would it be intended to influence
research by the NRC? Maybe identify is not the right
word. "Would influence" be the right word?
MR. ELTAWILA: Influence research. I
really consider it now as a gap analysis to try to
identify the weakness or the lack of information at
the NRC because we saw it in this advanced reactor,
particularly gas-cooled reactor very recently. So we
might identify an issue that there have been a lot of
research being done somewhere else, so if I call it
research or try to make it to influence research, it
might be the wrong way of characterizing it.
So it's really gap analysis right now and
once we collect more information we are going to
refine that and find out which part of the research
would be provided by the industry, which part will be
provided by NRC.
Having said that, one more issue that the
Office of Research, even though if the utility or if
the vendor provide information research data to
support their safety case, the Office of Research will
be conducting confirmatory research to try to go
beyond the information that's usually traditionally
provided by Applicants like poking into the area of
severe accident source term and the issue that not
traditionally being addressed by Applicant and
MEMBER LEITCH: So the operative word is
"by the NRC"? In other words, you're identifying
research that needs to be done by someone.
MR. ELTAWILA: By someone. And eventually
we'll try to narrow down to the research that will be
done by the NRC.
MR. ELTAWILA: Okay.
MEMBER FORD: Can you put a quantitational
thing on "eventually"? When are these decisions going
to be made?
MR. ELTAWILA: I think this decision -- we
are supposed to go to the Commission in the fall of
this year so we are planning to form inter-office task
groups to look at the information in the research
plan, identify which part of this information would be
provided. The NRC is going to ask the vendor and
Applicant to provide and then decide after that the
balance of that will be performed by NRC and finalized
that in the fall and send it to the Commission, of
course, after coming to you here.
MEMBER FORD: So there will be several
meetings with the ACRS to comment on the various
points along that time line?
MR. ELTAWILA: That's correct, yes.
CHAIRMAN APOSTOLAKIS: By fall?
MEMBER KRESS: Oh yes, we will several by
MR. FLACK: I think what's envisioned is
that we would come back at least once to the Full
Committee before we go to the Commission with the
plan. And then Subcommittees as we feel are necessary
or as the Committee feels necessary.
CHAIRMAN APOSTOLAKIS: Maybe you need a
better title though. When you issue a report that
says "Research Plan" it seems to me most people would
think research to be done by the NRC. Usually, these
are technical issues. They need resolution before you
MR. ELTAWILA: George, I agree with you,
but we are -- are embarking on an area here that we
really don't have too much experience, especially in
gas-cooled reactor. We don't have much experience and
we have, for example, we are having a hard time
getting information from the international community.
So the information might be out there, but we might
still have to do the research because we are unable to
get this information.
CHAIRMAN APOSTOLAKIS: No, I understand,
but I think the title of your report should be
advanced reactor technical issues.
MR. ELTAWILA: Information needs.
CHAIRMAN APOSTOLAKIS: Yes, information
needs, something like that.
MR. ELTAWILA: We can change that.
CHAIRMAN APOSTOLAKIS: Instead of Research
MR. FLACK: Well, the reason why it's a
plan is we're trying to build an infrastructure.
CHAIRMAN APOSTOLAKIS: But you cannot plan
for other people, John.
MR. FLACK: No, no. I understand. That's
when we exercise the plan. The plan is to build the
infrastructure and then part 2 is well, we're getting
a license application that at some later date we're
prepared to support the licensing office in that area.
So we have a plan to try to establish the
infrastructure that will support the plan.
CHAIRMAN APOSTOLAKIS: If you change the
title you will not need a separate color for that
bullet over there.
MR. ELTAWILA: We'll change the title, how
about that? Really, it's not a big issue right now.
CHAIRMAN APOSTOLAKIS: The second bullet
there, you know, why do you feel that you have to say
that? Isn't that sort of understood that the
Applicants are responsible for data?
MR. ELTAWILA: It is -- well,
traditionally, the NRC have been generating the data
for all plans, you know, before the 1990s and things
like that. The NRC generated all the thermal
hydraulic database, all the severe accident and the
fuel. So right now we are entering our strategic
plan, put the burden on the industry for providing the
data that's needed to justify the technical basis for
the licensing of the plant.
So it is important to identify that so
people when they read the plan, they don't think that
we are -- whatever we're going to call it, they are
not going to reach the conclusion that NRC is going to
do this work and then they will sit and not do any of
the work themselves.
MEMBER KRESS: I think that's worth
CHAIRMAN APOSTOLAKIS: But you also have
a sentence in the actual report. I don't know if you
want to come back to it, but where you say it is also
recognized that an Applicant of a new reactor design
has a primary responsibility to demonstrate the safety
case of the proposed design.
MR. ELTAWILA: That's correct.
CHAIRMAN APOSTOLAKIS: And later on, you
use a variation of this as well. It wasn't clear, I
mean somehow it sent a message that we are really not
part of this. We are setting the standards, aren't
we, the criteria and the objectives. It's their
responsibility to demonstrate they comply with the
criteria, but not -- what does it mean to demonstrate
the safety case? Are they going to also set the
MR. ELTAWILA: No, no. I think it's very
difficult to put everything in the first bullets, but
if you go a little bit further in our discussion you
will see that one of our responsibilities is to
develop the data to set the safety limits for this
CHAIRMAN APOSTOLAKIS: Sure.
MR. ELTAWILA: So that will be our
responsibility. It's not going to be Applicant
responsibility or anybody else.
CHAIRMAN APOSTOLAKIS: Okay, but I think
in the report it should be made clearer, because that
was something that struck me when I read it.
MEMBER KRESS: But when it comes to
deciding what data and research that the Applicant
needs to provide to you, do you have some sort of firm
criteria for how to pick out of this comprehensive
document so these are your guys and these are
confirmatory and they're ours. Do you have a way to
decide that or is that just going to be judgment?
MR. ELTAWILA: I think it will be a lot of
things: experience, judgment and our interaction with
the user office about what are the information that
they want independent capability from the staff to be
able to do their job. And our own initiative in the
Office of Research about how to build that additional
infrastructure to be able to ask more intelligent
questions from this Applicant and licensees. So it
will be a combination of the three and the way we have
developed this information and the past will play a
major role in deciding which part will be ours and
which part will be the Applicant's. But in the past,
Applicant tends to focus on the operation of the
plant. They have a safety envelope that they work
within the safety envelope and they will provide the
information to satisfy that need only.
NRC wants to go beyond that and to try to
challenge the system in a different way and we will
generate the information for that.
Although the plan itself is for AP-1000,
IRS and GT-MHR and PBMR, you will see that most of our
discussion will be on high temperature gas-cooled
reactor because that's the area we don't have much
CHAIRMAN APOSTOLAKIS: Do you have
sufficient information on IRIS?
MR. ELTAWILA: Okay, IRIS, let me -- IRIS,
we have very limited interaction with Westinghouse so
it's not really a major part of our activities right
The other points that I want to make is
that we -- Jim Lyons from NRR and I attended a meeting
with Framatome and Framatome is proposing to submit
SWR application. So -- SWR -- honestly, I tried to
look in the vu-graphs to find what -- simplified water
reactor or something like that.
MR. LYONS: This is Jim Lyons from NRR.
It's the SWR 1000. It was designed by Siemens from
Framatome and Siemens are now together. It's a plant
that's being considered to be built in Finland.
They're also looking at coming in. That would be a
BWR design that they're thinking about. They're also
exploring whether or not they'd want to come in with
the EPR which is European Pressurized Water Reactor.
That's another one that they're thinking, they're
considering coming in with for design certification.
CHAIRMAN APOSTOLAKIS: Now the SWR is not
the same as the SBWR?
MR. LYONS: No, it's not. It is a boiling
water reactor. It was --
MR. ELTAWILA: It's almost the same
principle, but it's different. So again, we're going
to change our plant as Jim indicated. They are
coming. They want certification. Next year, they
They are serious about submitting
We're having a meeting with them.
MR. LYONS: We're meeting with them on --
they're going to present these two basic designs and
they're trying to understand the design certification
process and to make a business decision on whether or
not they want to come forward.
MEMBER ROSEN: This raises the whole
question in my mind of how you pick the things that
you need to get researched, however you get them
researched. Because I was astonished in reading your
report that the Generation IV program of the
Department of Energy isn't mentioned until the 111th
page which is the last page.
CHAIRMAN APOSTOLAKIS: Because they
couldn't do it after that.
MEMBER ROSEN: Because they could not do
it after that and still mention it.
And in that program which is a very vital
program with lots of effort going into it, hundreds of
people working on it, many of the concepts that were
just mentioned and lots beyond that are being
considered seriously to be down-selected for
development of a roadmap and some research,
significant amounts of research from the Department of
Energy. I know John Flack who's with you. He's aware
of these things and has attended many of the meetings.
So I would ask you why don't you even
reference Generation IV in this report?
MR. ELTAWILA: That's a good question. We
are keeping informed with what's going on in
Generation IV, but it's a Commission direction. The
Commission directed the staff to work with this
applicant at this time, and that's why we defined the
work that will be needed for these four applications
that we have, even though IRIS is at the very early
So we get guidance from the Commission
about what to work on and what not to work on, and for
advanced -- for the Generation IV to continue to
interact with DOE, we're keeping abreast of what's
going on, and we keep the Commission informed with
what's going on. And once the Commission feels that
the staff should be engaged in this process, I think
the Commission will direct us to be working in this
MEMBER ROSEN: I think perhaps the
committee -- our committee ought to discuss this
CHAIRMAN APOSTOLAKIS: It wouldn't make
any difference, though, Steve. I mean, they are
trying to be as general as they can. I mean, look at
the very -- the penultimate arrow there. The
regulations will be technology neutral. I mean, if
they mention Generation IV on the second page, would
it make any difference to what they're proposing?
MEMBER ROSEN: Well, I think it would make
a great deal of difference.
CHAIRMAN APOSTOLAKIS: Really?
MEMBER ROSEN: Oh, yes.
CHAIRMAN APOSTOLAKIS: They are trying to
be technology neutral.
MEMBER ROSEN: Well, but I do think you
CHAIRMAN APOSTOLAKIS: Yes. Well --
MEMBER ROSEN: -- ever do that.
CHAIRMAN APOSTOLAKIS: Then, they will
have, they say, Regulatory Guides.
MEMBER ROSEN: No.
CHAIRMAN APOSTOLAKIS: So they will not
MEMBER ROSEN: For example, this report
includes -- a third of the report is on the research
to support nuclear materials, NMSS activities. The
Generation IV program will be -- if it continues to
evolve the way it currently is, will include a major
research track on sodium-cooled reactors, but the fuel
cycle of it mostly.
CHAIRMAN APOSTOLAKIS: Yes.
MEMBER ROSEN: With an emphasis on fuel
cycle research. And that's not mentioned at all in
this third -- last third of this 111-page report. And
it would seem to me that it would be a major thrust of
the nation's going-forward activity.
MEMBER KRESS: Well, I think Farouk --
MEMBER ROSEN: So my basic --
MEMBER KRESS: -- I think Farouk
appropriately answered, though. They've got
constraints on what this report is supposed to look
at, and it doesn't include that.
MEMBER ROSEN: Right. And I'd say if
those are the constraints that they were asked -- that
they were working within, because the Commission
directed that, then, well, that's certainly what they
have to do.
MEMBER KRESS: Sure.
MEMBER ROSEN: But we can advise the
Commission that maybe they ought to be thinking about
some broader issues.
MEMBER KRESS: Well, that's -- I think
that would be another issue, another thought.
MEMBER ROSEN: I'm not faulting them. I'm
MR. ELTAWILA: No. I think we encourage
the committee to think about the reality of the budget
situation, and things like that. We have to -- even
that we are encouraging NEI and the industry to come
with identification of what's really their priority.
You know, if it is going to be AP-1000,
PBMR, GT-MHR, we really need to get clear guidance
from the industry about what's important, what's
definitely going to be submitted for certification,
and has a chance of continuing with the application
here for review, because, as you can see from the
report itself, the amount of information that needs to
be gathered is tremendous.
And given the staff limitation and even
contractor availability and test facilities, and
things like that, we need to plan in a much better
structured way than trying to address everything at
the same time.
MEMBER ROSEN: I think there are major
strategic issues that need to be addressed, and that
one of them comes out of what you just said, which is
wait for the applicant to come and then we'll get
ready. I'm not sure that's the only way that research
should get defined, and we can discuss that more in
MEMBER KRESS: Yes. But surely you want
to give priority to things you know are going to come
in for certification, or at least you suspect very
soon. So, you know, you can't -- if you've got a lot
of stuff to do, you're going to focus on the ones that
you need first. And I think that's what they've done.
MEMBER ROSEN: Well, they've done what
they were told to do, which is a good thing to do --
MEMBER KRESS: Yes.
MEMBER ROSEN: -- when you work here.
MR. ELTAWILA: Okay. With the -- I think
George alluded about to the new regulatory structure
that we should be looking at. For example, some
feature of the PBMR is not really covered by current
regulation because -- which is developed for light
So Exelon has proposed a risk-informed
approach towards defining the license basing event to
supplement the current regulatory structure. And we
are planning to build on Option 3, and that's why Mary
is here, build on Option 3, try to provide -- maybe we
need to develop additional supplemental risk metrics
for the other type of reactor, and at a very high
level for what criteria this design should mean that
we can be technology or reactor design neutral.
And then, in the specific Regulatory
Guide, we'll try to see how well they should be
measuring against meeting the acceptance criteria, and
we'll provide that for each type of reactor, a Reg
Guide or a set of Reg Guides to address these
The overall objective of the research plan
is to, as I mentioned earlier, to determine the
critical information that is needed to establish the
safety standard new reactor design is going to
meeting. That's NRC responsibility. Although that we
might get some data from the licensee -- from
applicants, we have the major responsibility of
developing this data.
Again, another issue -- the issue of
uncertainty, we are planning to explore uncertainties
in this design and this information, and that's the
responsibility of NRC.
And, finally, is the issue of developing
independent analysis tool and give the data to assess
CHAIRMAN APOSTOLAKIS: Now, the
uncertainties. You have in mind something,
NUREG-1150? That's the only place where I've seen
large uncertainties handled.
MR. ELTAWILA: I think we will be looking
at something like NUREG-1150.
CHAIRMAN APOSTOLAKIS: With expert opinion
elicitation and doing something about it and --
MR. ELTAWILA: For some of this new design
which we're going to have, much of the experience or
much of the data, that we will have to look into
expert opinion. And you can -- maybe when John
discusses the issues of fuel you'll find some of this
in his discussion. I don't know if you were planning
to discuss it.
Again, because of the -- we are going to
rely a lot on cooperative agreement, although we have
been having difficulty entering into some of these
agreements, but there is work in China and Japan,
European community, and we are looking for cooperation
of the Department of Energy to do some testing in the
I want to conclude my brief presentation
here by saying that we looked at Dr. Powers' trip
report. I think Dana identified very important
technical and policy issues that the Commission needs
to resolve before we can say this type of PBMR in
particular is -- can be certified or not.
CHAIRMAN APOSTOLAKIS: Did you find that
MR. ELTAWILA: So the issues are very
CHAIRMAN APOSTOLAKIS: Did you find that
report clearly written?
MR. ELTAWILA: If you heard Commissioner
McGaffigan say, it's plain language, you know, and he
was looking for something from us to say the same
thing. But, unfortunately, he also admitted that our
concurrence process will not allow me to write
something like Dana Powers writes. So --
CHAIRMAN APOSTOLAKIS: Well, it's not that
-- I'm not sure this committee would think about --
Yes, he certainly speaks with sufficient
clarity and volume.
MR. ELTAWILA: Well, they are very
important issues. We identified these issues and sent
them to Exelon, and we are in the process of gathering
information about it, and we actually use this
information in the development in our research plan.
In addition to Dr. Powers, we received other comments
from Dr. Murley, for example, and all of this
information is factored into our plan.
CHAIRMAN APOSTOLAKIS: Now, why did -- I
sense that you have some problems with international
-- not problems perhaps, but you are not -- it's also
clear how you're going to get information from the
international efforts. Why do you need to understand
the status? I mean, you send somebody there, you
understand it. What's the problem? They are
reluctant to give you information?
MR. ELTAWILA: When you -- there is
reluctance -- I think, for example, the European
community is -- their system of working the everybody
do -- does research, and the shared information --
there is no exchange of money.
So for us to try to get information from
the European community, we'll try to get consensus
from all of the members of the community. And you
know that that's extremely difficult, to enter into an
ongoing program right now to try to get information.
So each country has said yes or no to sharing
information with NRC.
When it comes to China, it is just -- we
have limitations through the State Department and
things like that about what level of interaction we're
going to have with them. Japanese, again, the
organization -- so it's just -- in a nutshell, it's
not that easy.
Yes, we're sending people to go and meet
with them. We've been exchanging e-mail. We meet
with them. And it sounds very promising, and it looks
like we are on the right track, and we are going to
get the information. But, unfortunately, nothing has
materialized up to now. We have not signed a single
agreement with any of these countries. You know,
that's one of the most frustrating parts of this
activity right now.
MEMBER FORD: And do you have a backup
plan should those agreements not take place?
MR. ELTAWILA: Our backup plan is to go to
the Commission and say, "We will have to develop this
data, all of it, ourselves." And which I think that
will be -- will put some of this, like the PBMR
schedule, in jeopardy because some of these data are
very crucial for --
CHAIRMAN APOSTOLAKIS: Do they have any
incentive to cooperate with you? Is there any benefit
MR. ELTAWILA: The benefit is that we
definitely -- we are going to be doing research, and
we'll try to exchange the information. It's just
government-to-government communication and the
exchange of information is not that easy as a lot of
people think it is, you know, including our
Our Commissioner believes that we should
have had all of these agreements signed by now, but
it's just not happening that fast, you know.
CHAIRMAN APOSTOLAKIS: It's still not very
clear to me, but, anyway, let's go on.
MR. ELTAWILA: Okay. With that, I will
ask John to complete the presentation.
MR. FLACK: Okay. My name is John Flack.
I'm the Branch Chief of the Regulatory Effectiveness
and Human Factors Branch, which also has the advanced
I know we're time limited, and Farouk
covered a number of things, so I will briefly -- I
will go quickly through the viewgraphs. And please
slow me down if you need more information.
CHAIRMAN APOSTOLAKIS: Don't worry.
MR. FLACK: The plan was actually created
with a number of --
CHAIRMAN APOSTOLAKIS: Does this committee
have a reputation that it does not ask enough
questions? Because every speaker who comes here
encourages us not to hesitate to interrupt them.
Do we have a record of not interrupting?
MR. ELTAWILA: For the record, I did not
ask you to --
CHAIRMAN APOSTOLAKIS: Is our image so
terrible that --
MEMBER POWERS: We're very shy.
CHAIRMAN APOSTOLAKIS: Okay. John, we
appreciate your --
MR. FLACK: Okay.
CHAIRMAN APOSTOLAKIS: I know it was well
MR. FLACK: Thank you. The plan itself
had been created by -- over 20 authors actually wrote
parts of the plan. Many of them you'll find in the
room today, so what I'm -- I'm offering you an
opportunity, if there's anything technical that you
want -- you've seen in the plan or you hear here
today, we have the people here that --
CHAIRMAN APOSTOLAKIS: Would you please
introduce your colleagues?
MR. FLACK: Oh, I'm sorry. Mr. Rubin to
my left. Stu has been the -- in addition to work in
the fuels issue on the HTTR, he is also the project
manager on the pebble bed reactor.
CHAIRMAN APOSTOLAKIS: Okay.
MR. FLACK: And Joe Muscara to my right
prepared most of the material and the plan on
materials, primarily high temperature materials and
graphite. Don Carlson also works in our group and has
prepared most of the material on the nuclear analysis
part of that, for both material and reactor safety.
CHAIRMAN APOSTOLAKIS: Very good.
MEMBER KRESS: When I read the plan -- by
the way, I like the way it's organized.
MR. FLACK: Oh, good.
MEMBER KRESS: Yes. It makes it very,
very well put together to know what the issue is and
what it -- but when I read it, most of it sounds like
it was written by one person, except when you get to
the materials part that sounds like -- a little
different. But did one person write most of that?
MR. FLACK: No. Actually, well --
MEMBER KRESS: It was put together by a
bunch of people, huh?
MR. FLACK: We tried to establish a
certain format I'll cover in a minute, but I'm trying
to get that information out. But what was important
about the development of the plan is we didn't want it
to be issue driven; in other words, try to figure an
issue and then what research you need to resolve the
What we were really focusing on is the
infrastructure, the ability to ask the right
questions. And so we started -- well, I'll get to it,
but we started from that perspective, what are the
tools, what is the expertise that we're going to need,
rather than try to identify issues.
But, in the end, I do have viewgraphs on
some of the issues we see already -- technical issues
that could bubble up to be safety issues, that could
bubble up to be policy issues -- and we'll go through
that towards the end.
Farouk went over many of the objectives of
the -- the reason why we put together the plan. Some
of these I've just summarized on this viewgraph,
trying to identify the areas, the expertise, having
the plan as a communication tool, so people understand
what we're trying to achieve.
MEMBER ROSEN: But wait a minute. Now,
it's not to build an advanced reactor research
infrastructure. It's really to build an advanced
reactor research infrastructure for three or four
MR. FLACK: That's right. The scope is
there, it's only limited -- the scope of the plan
right now is limited to the four concepts that we have
on the table.
MEMBER KRESS: You should read advanced
reactor as these four concepts.
MR. FLACK: That's right. That's right.
CHAIRMAN APOSTOLAKIS: And also --
MEMBER ROSEN: Which may change tomorrow
if somebody else brings another concept in with an
MR. FLACK: Well, the idea is to see what
we'd need to do. We have an infrastructure in place.
It's what additional work or additional tools above
and beyond what we have already. So with these four
concepts coming in, we already see that we're going to
need new data, additional tools, and at that -- we're
looking at it from that perspective.
If another concept came in, we'll have to
see what tools can be applied to that concept. And if
there needs to be something new developed, then we
would take it from there.
MEMBER ROSEN: But, as you know, there
were something like 19 concept sets in the DOE
Generation IV program, which really meant that there
were something like 75 or 80 concepts that were looked
at overall. So there's lot of concepts out there.
MR. FLACK: Right, right.
MEMBER ROSEN: Some day -- so you need a
program that -- a thinking process that sets you up to
be ready to respond to whoever comes in with whatever
MR. FLACK: Well, you have to have that --
MEMBER KRESS: You can't do that for all
of them. I mean, you just don't have the resources.
MEMBER ROSEN: What I think is the list of
the four has some of the things that we might have to
work on in the next decade, but it certainly doesn't
have all of them.
MEMBER KRESS: Well, it probably
encompasses a good many of them.
MEMBER ROSEN: But it would be clearly a
mistake to believe that because the Commission has
picked those four that that's all that will ever be
brought to the table here and --
CHAIRMAN APOSTOLAKIS: From 4 to 80 is a
MEMBER KRESS: Yes, but I don't think --
to think in terms of which ones of these others might
make it to NRC, and then try to prepare --
MEMBER ROSEN: No, but you don't have to
think about it. You can just simply ask -- go out and
see what people are doing.
MEMBER KRESS: Well, I think their comment
that they try to -- try to make the -- at least the
acceptance criteria in the regulations reactor type
neutral is a good way -- is a good thing to do to
MEMBER ROSEN: It is. I agree with that.
CHAIRMAN APOSTOLAKIS: Now, the overall
objective, is it really to build an advanced reactor
research infrastructure, or is it to build the
infrastructure that would allow you to license
MR. FLACK: Now, there's a distinction
between the infrastructure, one being called
regulatory infrastructure and one called research
infrastructure. What we're talking about, at least
aside from the framework, we're really talking about
CHAIRMAN APOSTOLAKIS: But the objective
ultimately is to support licensing.
MR. FLACK: That's right. Which will get
us through the next phase of this plan that --
CHAIRMAN APOSTOLAKIS: So that's what you
should say, actually, right? I mean, to build an
advanced reactor research infrastructure, why? This
is a regulatory agency here.
MR. FLACK: Well --
CHAIRMAN APOSTOLAKIS: Only to the extent
that it's required for licensing. We've been told by
the Commissioners many times, they have said it in
public, this is a regulatory agency.
MR. FLACK: That's right.
CHAIRMAN APOSTOLAKIS: It's not the
National Science Foundation.
MR. FLACK: That's right.
CHAIRMAN APOSTOLAKIS: So the overall
objective probably needs to be reworded.
MR. FLACK: Yes. And it's driven a lot by
CHAIRMAN APOSTOLAKIS: Of course.
MR. FLACK: In fact, that was my next
viewgraph was to say, where are we going on the second
phase of this plan? If I can jump to that, we can --
CHAIRMAN APOSTOLAKIS: Of course you can.
MR. FLACK: -- talk to that issue a little
The first phase of the plan was really to
get out everything on the table as -- that we know it
today, with no constraints to resources, and so on.
And so we held workshops, we had the preapplication
review to capitalize on, we had talked -- we went
around the world looking at what was out there.
So we're coming to the end of this first
phase, and, actually, with this meeting, which will be
the second phase of this research plan. And the
second phase of this research plan is really what
focuses on that particular issue that you just brought
up, George. It's to set up working groups with the
user offices now that we've seen -- and we gave
everything -- put everything out on the table. What
is it that we really need to do now?
CHAIRMAN APOSTOLAKIS: Yes.
MR. FLACK: Okay? And that's going to be
the next phase, and we see this phase coming to
completion. The next time we come to the committee we
would be more focused on that particular issue of
supporting the process, the regulatory process in the
global sense, and then going to the Commission with
that plan at that time.
And then, the third phase is really to
maintain it a living plan, to pick up new designs as
they come in, see what delta needs to be done, what
new tools we need to develop, and to state engaged in
that Generation IV activity, to see if these things
are materializing to the point where we need to start
getting serious about something.
MEMBER FORD: Now, how does the
prioritization judgment come about? Given the fact
that your resources are undecided, management
resources like collaborative agreements, people,
dollars. That's not a fixed amount right now. So
your prioritization is going to presumably change with
time, isn't that correct?
MR. FLACK: Well, I think Farouk might
want to --
MR. ELTAWILA: No. I think the -- our
budget and resources has been established for the next
three years, you know, that at least to -- our 2003
budget is fixed, and 2004 and 2005 is proposed to the
Commission. And we will try to prioritize within
these budget constraints.
And if we're going to be using the same
PPM process, and we'll be competing with other
operating events that depends on the priority, we'll
be funding this research based on the available
MEMBER FORD: No, I recognize that.
That's how you're going to spend your money on your
people and subcontractors. But what happens if one of
the priorities that -- technical priorities -- work on
graphite, for instance.
MR. ELTAWILA: Okay.
MEMBER FORD: That work has been done in
Britain, for instance. And what happens if the Brits
decide that they don't have to give you that data for
whatever reason? What happens?
MR. ELTAWILA: The first point, that we
are going to be asking the applicants to provide us
for the data to support their case, and then based on
the information we're provided we'll see what
additional information we will be -- we need to
MEMBER FORD: Okay.
MR. ELTAWILA: It is not very easy for a
regulatory agency to try to develop a research
program. It has to be issue-driven, as George
indicated, that we -- everything has to be related to
the licensing process that we are working on.
CHAIRMAN APOSTOLAKIS: I think the overall
objective should be reworded to reflect that. I mean,
I appreciate the phases, but you said overall
MR. ELTAWILA: Okay.
CHAIRMAN APOSTOLAKIS: Ultimately, that's
what you're going to do.
MEMBER KRESS: I think it's implicit in
everything already anyway.
CHAIRMAN APOSTOLAKIS: Another thing I
noticed when I read the report is that you list
everybody's workshops except the ACRS. Was there any
reason? Did you find it useless?
MR. FLACK: No. There's no reason why we
missed that. That was an important oversight. Thank
CHAIRMAN APOSTOLAKIS: Maybe it was not
very useful to you.
MEMBER POWERS: Maybe they just didn't
like our --
CHAIRMAN APOSTOLAKIS: That's I thought,
MEMBER POWERS: Nothing useful emerged
CHAIRMAN APOSTOLAKIS: You list
everybody's workshops, the dates and this and that.
Of course, it will never bias our views, but --
MEMBER ROSEN: You're too sensitive,
CHAIRMAN APOSTOLAKIS: I am not too
sensitive. I'm just sensitive.
MEMBER LEITCH: The second bullet is --
CHAIRMAN APOSTOLAKIS: Commissioner Diaz
was there. He gave the keynote speech. Maybe the
staff doesn't think much of what the Commissioner
MR. FLACK: I think if -- you'll find --
I'm sure I've seen it in there somewhere.
CHAIRMAN APOSTOLAKIS: It is not here.
John, it is not here.
MR. FLACK: It might have got scratched
the last time. I don't know.
MEMBER LEITCH: The second bullet there,
Johns, is there some reason the AP-1000 is not on that
list or --
MR. FLACK: No, that should really be on
there. It was for examples, and I was --
MEMBER LEITCH: It says "for example," and
I was just wondering if it --
MR. FLACK: Yes, they're all HTTRs. I
should have put a light -- yes, a light water reactor
on there. Yes.
MEMBER ROSEN: There's an astonishingly
pervasive gas reactor focus on this, because of the --
MEMBER KRESS: Well, you're almost through
with the AP-1000 preapplication review anyway.
MR. FLACK: Yes. The preapplication is
done, in fact. I think the --
MEMBER KRESS: Is that correct?
MR. FLACK: But most of the gap that we
see is in the high-temperature gas-cooled area, so,
you know -- but we have an infrastructure in place
pretty good for a light water reactor.
Okay. I think we pretty much touched upon
this. The meaning on infrastructure, again, is the
staff expertise, the tools, the facilities, contractor
support, and the scope being the four reactors as we
see it today. And the structure -- and, again, we
built the structure around not the issues themselves
but on the technical areas, which you'll see in a
MEMBER POWERS: John, before you take that
down, let me ask you a question about technical
approach on this. The second item on your list there
is called analytic tools and analysis methods. And
one of the challenges that we repeatedly come up with
when we look at things connected with current reactors
and modest changes to those current reactors, like the
AP-1000, is that many, many, many of our analytic
tools going from simple neutronics through thermal
hydraulics to fission product release had their origin
in an era when the computing capabilities that people
had were widely different than what it is now, and
probably we'll see in the next 10 years even more
Yet your plan doesn't seem to act upon
those things. I mean, it doesn't seem to take that
into account. There is lots of things like, well, we
can take TRACM and put another patch on it, we can
take MELCOR and gerry-rig it to handle something else,
rather than saying, "Hold it. We really have
undergone a computer revolution here." The way we do
computing, the way people do coding now, it's just
very, very different than what it was when our codes
had their origin.
Maybe it's an opportunity for us to bring
our codes up and to recognize that the hardware has
just changed, and what not. But your plan didn't seem
to delve into that kind of an approach.
MR. FLACK: You know, it's an excellent
subject for a subcommittee, I think, to revisit this
particular issue. You're right. We're really
building on things that already have been developed
and seeing where we're going to -- how can we extend
them rather than go back to -- you know, and look and
see is there a better way of doing this. And I think
it's an excellent question. We just -- just built on
what we have.
I know TRACM is improving, of course, has
come quite a way from -- just in the Fortran part of
that. But as far as starting with something new --
and this may be an opportunity to do that for these
gas-cooled reactors, where you may have one code,
because of the nature of the beast, that you don't
have the core melt and the accident progression and
that -- you have a fission product release over time
and temperature and using one code to deal with the
whole spectrum, right out into the environment, might
be a way to go.
MEMBER POWERS: One of the things that it
seems to me that -- you know, in trying to think about
the future, and you put it right up front in your
plan, you say, gee, you know, we're going to move to
a probabilistic risk assessment kind of framework.
And whereas I -- I know for a fact that a lot of our
probabilistic risk assessment tools are kind of
They work pretty well until you get to the
questions of, gee, let's do some of these
deterministic analyses for a bunch of scenarios. And
then we run into a problem that our codes are fairly
archaic. And if somebody wants to run 150 MELCOR
sequences, for instance, you know, you're -- and
that's an enormous number for a probabilistic risk
assessment; 150 is actually a fairly modest number.
You really are buying yourself a pretty
big chore here. So if you -- you know, if you were
looking to say I want to make bigger use of
probabilistic techniques in my licensing process, I
want to have more assessments of them, I want to take
that probabilistic technique deeper into the accident
sequences, rather than just looking at Level 1 I
actually want to go deeper into Level 2, and things
like that, then my phenomenological tools, both
thermal hydraulic and structural techniques and things
like that, have to be better.
You might really come to the conclusion
that you need to invest some in your tools, and that's
regardless of what goes on in DOE land or in the
vendor's land, that you really do need to encourage
the Commission to get you the resources to develop
I mean, I guess my thinking on this is
that, for instance, the thermal hydraulic area you
have some people that are pretty qualified getting
TRACM as a consolidation. And that's going to be
awfully useful for existing reactors, but I bet you
they don't find it very satisfactory for looking at
very innovative kinds of thermal hydraulics things
where the analyses go, I think as you say in the
document, instead of working on time scales of a few
hours you're starting to work on time scales of days
and things like that.
MR. ELTAWILA: John, can I try to address
Dana, you are raising a very good issue.
But I just -- actually, our problem is not really the
speed of the computer, because you continue to enhance
that, and the machine speed itself will make up for
But the biggest problem is trying to
develop a code. You have to have a target that this
code is going to be better than what we have right
now. And we really don't have the data to support
development of models that we'll be able to put in
So going -- embarking on a code
development program, without having the supporting
experimental data, will be just a waste of resources.
And we face that issue early, you know, when we are
thinking about either developing a new thermal
hydraulic code versus consolidating the existing code
into a single code.
And we'll get a group of experts, and they
all advise us against developing a code from scratch,
because we're going to end up -- the code is going to
be slow because of the limitation of the model, not
because of the machine.
So unless somebody is willing to invest a
few hundred million dollars in developing the data to
support this fast running code with accurate, better
models, I think going into the development of faster
code is not going to be the best way we put our money
MEMBER ROSEN: I'd like to add that,
although it's probably true, that many of the codes
that we'd be looking at using in licensing reviews are
built on older, previously developed codes. There may
be some pockets where there are new codes being
developed in the current computing environment.
And I would give as an example in the fuel
performance area, the European Commission has a high
temperature reactor fuels task group in place. And
one of the areas that they are doing work in is to
develop fuel performance models today. And those fuel
performance codes will be developed, obviously, in the
current computing environment.
Also, INEEL, working with MIT, I believe,
is developing fuel performance models and codes to
predict fuel failure, etcetera. So there are a few
examples at least where codes are being developed in
MEMBER POWERS: Well, I, of course, have
come to learn that fuel research is irrelevant, so --
MR. ELTAWILA: That's the subject of
MEMBER POWERS: I couldn't resist.
MR. FLACK: We'll move right along on
Basically, to your comment, Tom, on how we
structured the report was around three questions --
why we -- why is it important for us to do this
research, what it is we would actually do, and then
how would we use the results. And we tried to keep
each of the people focused.
MEMBER KRESS: And I thought that was very
good. It was very helpful in reading it.
MR. FLACK: And the research plan
structure, which is -- has been developed, and this
was developed to sort of try to get the completeness
of the work that we're doing. We actually started,
again, not from an issue perspective but from the top
down, and we began -- well, we started by looking at
the arenas that we would be working in as far as
research is concerned. Well, as you can see, most of
it is reactor safety.
We're looking and pressing into these
other arenas to see what work can be done, since most
of the work that we do involves reactor. So there is
some of it discussed as far as nuclear waste and
materials safety, and then, of course, safeguards.
Again, we're pressing that area.
But within the reactor safety arena, we
laid out the work more or less along the lines of the
cornerstones of safety. And bringing that down
further, going from accident -- starting from right to
left, accident progression to initiating events, which
dictates the sort of scenarios we need to look at as
an office on a particular plant design, and then from
there -- which actually sets the stage for the rest,
coming down to look at accident analysis and what area
or what technical work needs to be done in that area.
It's primarily driven by the PRA and those
things that -- that influence the PRA, like human
factors and I&C. And so in these areas PRA was
generally that part of the research under Mark
Cunningham, as you know, Mary Drouin, and Alan Rubin,
and John Ridgely. And on the plant analysis it's
primarily the human factors and I&C, which is Steve
Arndt for I&C and Jay Persinski for human factors.
Moving across from there, from left to
right, the next large area is the reactor systems
analysis, which is primarily in Jack Rosenthal's
branch. And under that being the thermal hydraulics,
the nuclear analysis, and the fission product
MEMBER POWERS: You felt that it was --
that the computational tools you have available to you
for doing probabilistic risk assessment -- the actual
analysis itself, you know, calculating out the
probabilities, that those were in such fine shape that
they deserve no improvement at all?
MR. FLACK: Well, no, I don't think that
would be the case. There's really -- I don't know if
Mary wants to respond to that, but there's really
three areas there in PRA that we see as being --
pushing our needs, and that is initiating event
frequency for the high-temperature gas-cooled
MEMBER POWERS: Yes, but those are data
things. I'm talking about the actual computational
MR. FLACK: Oh, the computational tools?
Do you want to comment on that, Mary?
MEMBER POWERS: The way you go about doing
MS. DROUIN: I agree that there is going
to need to be some research in the development of some
of these tools, particularly in the computational
area. And that's --
CHAIRMAN APOSTOLAKIS: But the report I
think says that SAPHIRE will be used for the PRA.
Isn't that so? That's what the report says.
MR. FLACK: Yes, that's right.
MS. DROUIN: SAPHIRE is a starting base,
absolutely. I mean, I would not like to think we
would just start with a clean piece of paper and not
take a tool that we already have and see where we can
use it, modify it appropriately.
MEMBER POWERS: At least through the
classical Level 1 for normal operating events, the
computational pathway is fairly straightforward, I
MS. DROUIN: Yes.
MEMBER POWERS: And adequately -- the
blocks that you need are adequately there in SAPHIRE,
maybe the computational way it's done.
The issue, it seems to me, that's been
raised so clearly by the eminent Dr. Kress is that
that computational framework may not be adequate if we
were to extend the way we do PRA from an operational
events to include all plant operational states.
I think that's a conclusion that has come
from your own studies in looking at the other
operational events, that the tool you have may not
have all of the computational elements you need to do
all operational states.
MS. DROUIN: I don't disagree.
MEMBER POWERS: And as we know, we trust
you implicitly, because you're one of my heroes,
MS. DROUIN: Absolutely.
MEMBER POWERS: I told you I'd get it on
MS. DROUIN: But, you know, when you get
into -- there's a lot of technical issues,
particularly in the Level 2 when you start looking at
the advanced reactors, and this will have a direct
impact, then, on the calculational tools we use and
where we'll be needing to do some work.
And right now we are in the midst of
trying to -- when you look at the RES plan, you know,
that plan there, when it gets into the PRA part, is
very high level. We are in the midst of trying to put
together a very detailed plan of what we mean by that
three-page plan in the RES-1.
MEMBER POWERS: I'd like to see that.
That would be interesting.
CHAIRMAN APOSTOLAKIS: If I look at
MS. DROUIN: We do plan to come to the
ACRS with it.
CHAIRMAN APOSTOLAKIS: If I look at this
figure, I see the acronym -- actually, it's
initialism, right? PRA? It's an initialism. Down
there on the left.
But it seems to me that, you know, again,
your report shows that the thinking is really that --
if you look at the out within the four boxes, and so
on, you will be looking at the accident sequences all
the way from the initiating event all the way to
offsite protection or somewhere in between, and use
that information in your decision-making processes.
And that's PRA, is it not? So it is a little bit
misleading the way it's shown there.
MR. FLACK: Under "accident analysis," do
CHAIRMAN APOSTOLAKIS: Yes. I mean, it's
pervasive. It's --
MR. FLACK: Yes, that's true, very much
so. There was another figure in the report that shows
these loops of information, how it flows between the
groups, which I don't have with me. But you're right,
there is always this feedback mechanism, both within
the groups and background PRA. In fact, that's the
way the office does work. PTS is an example where you
bring in, you know, the PRA people with the materials
people with the thermal hydraulic folks and --
CHAIRMAN APOSTOLAKIS: Well, the biggest
question, really, here would be, how are you going to
use the PRA? I mean, right now, in the most important
decisions the agency is making PRA is very peripheral.
It doesn't really play any role.
MR. FLACK: In your regulatory decision-
making or the use --
CHAIRMAN APOSTOLAKIS: Yes.
MR. FLACK: -- in the --
CHAIRMAN APOSTOLAKIS: Regulatory, like
license renewal, power uprates, PRA really doesn't do
much there. I mean, it's just, oh, by the way, this
is the number we got from the CDF.
MEMBER KRESS: And even in direct
CHAIRMAN APOSTOLAKIS: And in what?
MEMBER KRESS: Just licensing a plant
doesn't seem to play a role.
CHAIRMAN APOSTOLAKIS: Well, we're not
licensing anybody. That's what --
MEMBER KRESS: Well, we will be.
CHAIRMAN APOSTOLAKIS: Yes, that's what
I'm saying, that this will be --
MEMBER KRESS: Same thing is the license.
CHAIRMAN APOSTOLAKIS: I mean, so that
will be a major challenge, I think, how to use that,
how to actually use it.
MR. FLACK: Yes, we're moving towards the
framework box there, I think.
CHAIRMAN APOSTOLAKIS: You're going to
talk about it separately?
MR. FLACK: If you'd like. Do you want to
talk about it --
CHAIRMAN APOSTOLAKIS: Do you plan to talk
about it? Are you planning --
MR. FLACK: Well, we can talk about it to
a certain extent.
CHAIRMAN APOSTOLAKIS: Well, that, it
seems to me, would be a major challenge.
MR. FLACK: Yes.
CHAIRMAN APOSTOLAKIS: Because the
Regulatory Guide 1.174 doesn't apply here. I mean,
that's for changes in the licensing process.
MR. FLACK: Right. That's right.
CHAIRMAN APOSTOLAKIS: And you don't have
a licensing basis here. So it's really using this as
part of your integrated decision-making process.
MR. FLACK: That's right. It is --
VICE CHAIR BONACA: They show Option 3 as
a foundation for this. Option 3 has a very specific
apportionment of certain goals --
CHAIRMAN APOSTOLAKIS: I understand that.
I understand that.
VICE CHAIR BONACA: -- which are really
measurement for PRA. So there is some structure that
you can put inside here already.
MR. FLACK: Yes. But the point I think is
that we're dealing with plants already built, and
we're applying PRA concepts to those plants in the
sense of changes. And now we're thinking, well, what
are we going to do with respect to regulatory
decision-making on future plants that haven't been
CHAIRMAN APOSTOLAKIS: Right.
MR. FLACK: And that gets us -- I think
pushes us into this framework, what do we need? And
there's really two pieces going on there. One is this
blank sheet of paper starting from a clean approach,
which is -- there is going to be work initiated next
year, and there's work going on in NRR is -- how do we
transition to that?
And Mary can talk about the part about the
research plan, and Jim Lyons could talk about the NRR
approach that's now taking place, from that
perspective. So they're coming together in some form.
Mary, did you want to --
CHAIRMAN APOSTOLAKIS: Well, you are
basing it on Option 3, right?
MS. DROUIN: Well, if you remember, the
Option 3 framework has, you know, three parts to it.
It has -- started with, you know, what we call that
CHAIRMAN APOSTOLAKIS: Right.
MS. DROUIN: You know, a top-down
approach. And then, because it is risk-informed, it
brings in how you bring in defense-in-depth both at
the hierarchical, from the top down and the bottoms
up, and then brings in, how do you bring in your
quantitative guidelines? And ultimately that is
producing the criteria and guidelines that you would
be using to help you in your decision-making process
throughout your licensing.
In terms of your earlier question, you
know, the PRA and the framework and -- it's like
they're all very intricately tied, and one of the ways
that you do use your PRA, you know, would help in your
decision-making also in terms of how much research,
using that word loosely here, that you would need,
because you certainly don't want to pursue an area
that, from your PRA perspective, you don't need it to
support the PRA, and you don't need it for -- it's not
going to help you, and it's not going to contribute
significantly to your risk is what I'm saying.
CHAIRMAN APOSTOLAKIS: Well, the point,
though, is -- I understand what you're saying, Mary.
But this is really something that is an ideal
situation. I can't imagine, for example, the guys who
will be working on the reactor plant analysis and fuel
analysis will be willing to take their criteria and
objectives from the PRA guys. They will just --
MS. DROUIN: As an input.
CHAIRMAN APOSTOLAKIS: That would be one
of the angles to their integrated decision-making
process which would have, I think, other major, major
MS. DROUIN: Yes.
CHAIRMAN APOSTOLAKIS: So the question
will be, you know, to what extent will there be --
will the PRA inputs influence that, or they will say,
no, you know, defense-in-depth and safety margins is
really the name of the game.
MS. DROUIN: But that's where you're -- I
mean, what we're calling it, the framework or the
decision-making criteria comes in and provides you
guidelines on that and how you bring in your defense-
in-depth, your uncertainties, your safety margins, and
your risk insights, and how you blend all of those
together in your decision-making process.
CHAIRMAN APOSTOLAKIS: Which we don't have
right now. We don't have those guidelines right now.
MS. DROUIN: That is what we're going to
CHAIRMAN APOSTOLAKIS: Right.
MS. DROUIN: Where we're starting with
Option 3. Now, you can't just adopt Option 3, because
Option 3 is, how do you make current changes?
CHAIRMAN APOSTOLAKIS: Right.
MS. DROUIN: And so there -- you'd have
other questions that you're going to have to answer,
because we're not just making current changes, you
know, in cases you're starting new.
CHAIRMAN APOSTOLAKIS: Right.
MS. DROUIN: So when you're starting new,
you've got to --
CHAIRMAN APOSTOLAKIS: Well, frankly, I
don't know how you can use PRA in light of Davis-
Besse. That was, I thought, a major blow to the whole
risk cause. I mean, unless we recognize that. I
mean, 10-4 means nothing to me now.
MEMBER ROSEN: In the case of PBMR, and we
believe GT-MHR, they have proposed a licensing
approach, which the staff has reviewed. And I think
we have briefed the committee on the licensing
approach, and it is very much PRA-based, in the sense
that licensing basis events are randomized for
probability and consequences.
And they are put into the framework or
approach that they utilize for operational events,
design basis events, and beyond design basis events.
And I think it would be useful to have a PRA -- the
staff to have its own PRA to kind of review those
applicant placement of those events within that
CHAIRMAN APOSTOLAKIS: But, you know,
about I think three years ago or so, or maybe longer,
there was a major issue that was raised. I think it
was before 1.174 was published. People, especially
from the industry, were complaining that PRA was just
another burden, that we had to do everything, you
know, the regulations said, plus a PRA, to get those
So if we are to use it now, somehow those
other requirements will have to be effective, and
maybe some of them should be eliminated. And I --
this is where I think will be a major problem, how to
do that, because we're going to have, again, the same
philosophical conflict. Okay? And I think the Davis-
Besse incident gives arguments to the structuralist
MEMBER ROSEN: If you're correct, George,
CHAIRMAN APOSTOLAKIS: They're about to
win me over.
MEMBER ROSEN: I think you would be
correct if all 100 plants had that problem.
CHAIRMAN APOSTOLAKIS: Hmm?
MEMBER ROSEN: If all 100 plants had that
problem. We're talking about a plant.
CHAIRMAN APOSTOLAKIS: Yes.
MEMBER ROSEN: One of 100 or so. So --
CHAIRMAN APOSTOLAKIS: I missed that.
MEMBER ROSEN: Well, I'm just responding
to your point that the event -- that Davis-Besse
invalidates all of the probabilistic thinking.
CHAIRMAN APOSTOLAKIS: I didn't say it
invalidates, but it creates serious questions in my
MEMBER POWERS: George, I --
VICE CHAIR BONACA: It goes back to the
proposal. It has a means of filling the gap in the
Code of Federal Regulations. I mean, in that sense,
PRA has been extremely successful. Here we've
attempted to see -- it could play a primary role, in
and of itself, rather than defense-in-depth, and
that's really where concern comes. Okay? Can it be
the first, you know --
CHAIRMAN APOSTOLAKIS: Dana?
MEMBER POWERS: Well, I guess I had two
points. One, just to respond to Steve, all individual
plants have individual peculiarities that can be
To your point, George, as one of the more
ardent of the structuralists on the committee, I'll
tell you that, no, I still think PRA has a -- despite
Davis-Besse, and what not, has a really admirable
place to play within any kind of reactor system. It's
just that it doesn't play in the defense-in-depth
argument from a structural point of view. It plays
very much in the redundancy, and what not, within
I still think it has a strong place to
play there, and I think it will be an even stronger
place to play in the advanced reactors where we can
relieve much more of the ad hoc determinism yet again.
CHAIRMAN APOSTOLAKIS: I think unless the
PRA guys do a better job on model uncertainty it will
not play such a significant role in the process.
MEMBER KRESS: I think you're right,
George. That'll be a key.
CHAIRMAN APOSTOLAKIS: I think the lambda
stuff, the log normal stuff, is nothing. It's the
model uncertainty that drives the decisions.
VICE CHAIR BONACA: I think one thing
that, you know, impresses me more and more as we go
forth is the -- some of the wisdom in 1.174. You
know, the whole concept of integrated decision-making,
etcetera, that comes --
CHAIRMAN APOSTOLAKIS: It's an ideal
document. But show me one case where it was applied.
There isn't a single case where this
beautiful discussion on uncertainty was actually
VICE CHAIR BONACA: That's true. You're
CHAIRMAN APOSTOLAKIS: It's model
uncertainty. That's the name of the game. The
distributions in lambda don't mean anything, and I
don't think we're doing a good job there. I
understand, you know, some of the tradeoffs that Dana
mentioned, sure, they are meaningful, and so on. But
it's really model uncertainty that does the trick.
MEMBER POWERS: Well, I bet we see -- I
certainly hope we see good uses of it in the PTS
MEMBER ROSEN: In the PTS stuff?
MEMBER KRESS: Pressurized thermal shock
CHAIRMAN APOSTOLAKIS: Even there I think
there was more promise than actually done.
MEMBER POWERS: Well, we haven't seen the
final story there. But, I mean, that's -- well,
certainly, you can't criticize a program because
there's more promise than was actually done. I can't
think of any program that that's not the case, so --
CHAIRMAN APOSTOLAKIS: There's no question
about it, that it's a pioneering study.
MEMBER KRESS: Well, Option 3, though, is
still highly focused on light water reactors. It
talks about CDFs and LERFs and sequence frequencies
that are endemic to light water reactors, and it tends
to -- to allocate risk among CDF and LERF and allocate
it among sequences, actually.
And you won't run into a difficulty when
you get to the -- trying to apply Option 3 in that
sense to the gas-cooled reactors, because you don't
have the equivalent number of sequences, you don't
have the same ones, you have a different set of
frequencies that are important, and you don't have a
well-defined CDF or even a well-defined LERF.
And so I think one of the things that
you're going to buck up against is you'll need more
precision in your definition of defense-in-depth for
these reactors. You just can't say anymore that it
means a balance between containment and CDF. You're
going to have to be more precise, and it's going to
have to tie in the uncertainty some way, even though
you could still keep the structuralist view. You're
going to have to tie in to uncertainties in some way.
CHAIRMAN APOSTOLAKIS: Well, that
uncertainty has to be a realistic assessment of
uncertainties, not just the stuff that's easy to do.
MEMBER KRESS: Yes.
MS. DROUIN: If you go back to Farouk's
slide, one of the things that we have identified in
developing, you know, this -- taking the Option 3
framework and, you know, modifying it for advanced
reactors, the primary thing was to look at the
surrogates of CDF and LERF.
CHAIRMAN APOSTOLAKIS: Yes. Yes.
MS. DROUIN: And that's one of the
critical items there, that those may not be
sufficient, and we may need to come up with different,
you know, figures of merit here than just those
surrogates, and come up with some others. So that's
one of the big items that we have ticketed to look at.
CHAIRMAN APOSTOLAKIS: Now, coming back to
this figure -- oh, I'm sorry. I can understand, and
I agree, that this thing, you know, by and large is an
effective -- contributing to an effective regulatory
process. I just don't know that it's efficient. You
say effective and efficient. How do you know it's
MR. FLACK: Well, it's something you
CHAIRMAN APOSTOLAKIS: But how? I mean,
if you ask the guys who were developing all of these
rules in the late '60s/early '70s, I'm sure what they
wanted to do was also be efficient. And here we come
20 years later and say they are not.
VICE CHAIR BONACA: I think if you compare
it to the existing system, I mean, probably the
inclusion of the PRA considerations, the risk
considerations, are making it more effective and --
CHAIRMAN APOSTOLAKIS: I'd like to see
VICE CHAIR BONACA: Well, no, because I
think in some cases you will limit the -- the
necessary burden, okay, that's the only -- I mean, to
the extent --
CHAIRMAN APOSTOLAKIS: Mario, you will be
told it's defense-in-depth, period. Do it. Okay?
It's a new system, we don't know, we don't want to be
surprised again. And I think there's a lot to that
VICE CHAIR BONACA: Well, we have seen
some, you know --
CHAIRMAN APOSTOLAKIS: If in a mature
technology we get things like Davis-Besse --
VICE CHAIR BONACA: Yes, I know.
CHAIRMAN APOSTOLAKIS: You know, I'm just
putting myself in a situation of the poor PRA guy who
says, "Your inspections will fail with probability .2
over a number of years." He's going to be crucified.
My inspectors never fail. Are you kidding? My
inspectors will go there and find it in a minute.
Okay? That's exactly what you're going to get. It's
the same thing you were getting before 1978.
My operators know what to do, and it's
always my -- I don't know why they put that "my" in
I remember. I was in a PRA, and we said,
you know, how about if the operators don't know how
VICE CHAIR BONACA: See, but let me just
CHAIRMAN APOSTOLAKIS: Are you kidding?
They will not know?
VICE CHAIR BONACA: Yes. But I don't
think we can make too much -- in a Davis-Besse event,
we have to learn more. There were a lot of
indications for a long time that something was wrong.
Now, at some point --
CHAIRMAN APOSTOLAKIS: And where is that
in the PRA?
VICE CHAIR BONACA: Well, I'm only saying
that there is a burden on operations to, in fact,
respond to the indications that you have. And in this
case, we may have a case where they did not respond
for years to this indication, that they had plenty of
those. And so I'm saying that you cannot address
everything in your PRA.
CHAIRMAN APOSTOLAKIS: It seems to me that
you will never make progress unless you punish people
for the mistakes they make.
The PRA should be penalized now for that.
MEMBER ROSEN: The PRA should be
CHAIRMAN APOSTOLAKIS: Well, or the PRA
practitioners on the use of the PRA.
MEMBER KRESS: You're just going to change
-- you're going to change the frequency of medium
break LOCAs. That's all you're going to do.
CHAIRMAN APOSTOLAKIS: How about the
efficient, though? How are you going to make sure
MR. FLACK: Well, that was the -- the
question is using these risk insights, which you think
or believe at this point aren't doing what they should
be doing, to utilize those and focusing your resources
on the right things and being efficient by doing that.
I mean, without that, I don't know, it's just
judgment. I mean, I --
CHAIRMAN APOSTOLAKIS: Well, one way to do
that is to really put a lot of meat to what Mary just
said. I mean, if you start from the top and with a
PRA structure you go down and you put objectives, then
you know why you are putting them there. But the
moment you start saying, "No, I'll do it because of
defense-in-depth, then you are deviating from
MR. FLACK: Yes, it could be.
CHAIRMAN APOSTOLAKIS: It may be for a
good reason, but --
VICE CHAIR BONACA: I still believe that
the use of PRA in many areas where you don't have this
kind of grayness is going to really yield much more
CHAIRMAN APOSTOLAKIS: How do you decide
when you have grayness?
VICE CHAIR BONACA: Well, I mean, you
know, an area, you know -- I mean, certainly you have
some indications where you have balance with
information and mitigation that you do not want to
compromise, and you're going to be very committed to
defense-in-depth. There are a lot of decisions,
however, in the design of a plant where, you know, the
inclusion of consideration of probabilities will help
you be more effective and have less of a burden.
MR. FLACK: I think in that role of
knowing what's not important, I mean, we are always
focusing on the PRAs, trying to point out what is
important, which is a good thing. But it also points
out things that are not important, and for certain
reasons, then, justify that.
I mean, you have to have a technical basis
for it. But, I mean, it's a thinking process that
allows you to do that. So, you know, I don't think we
should throw the baby out with the bath water, I mean,
CHAIRMAN APOSTOLAKIS: You're more
optimistic than I am.
VICE CHAIR BONACA: But there was really
practical terms. And in the 15 years or 20 years of
use of PRA in this approach, it has paid off
tremendously for the utilities that use it in those
kinds of decisions where you are not only affecting
defense-in-depth, but you are making intelligent
decisions on imposition of your requirements or
elimination of those.
And we have seen some proposals that have
been approved, and 1.174 -- they were really
acceptable, have not been, you know, undermined by the
experience with Davis-Besse.
VICE CHAIR BONACA: I think there's got to
be some efficiency brought in by that.
MR. FLACK: Moving right along --
MEMBER KRESS: Please continue.
VICE CHAIR BONACA: I'm trying to convince
you that PRA is --
MEMBER KRESS: I can't believe we're
having this discussion. Continue, please.
MR. FLACK: Okay. So this is the process
we use. It's clearly -- it's a matrix approach. We
use the entire office resources as input to the plant.
Now, the next few viewgraphs I go through
and identify the different technical areas. I don't
know if we need to spend much time on that. It's in
the plan. Those are the areas that are being hit.
And that kind of leads us on to what the technical
issues are that we're seeing now. Maybe we can, for
the sake of time, jump to that viewgraph.
MEMBER KRESS: Well, let me ask you a
couple of questions about the technical areas first.
MR. FLACK: Okay.
MEMBER KRESS: You know, you're asking us
for -- whether you think you have the right scope or
you're missing anything or something. I thought it
was very comprehensive. In fact, there's so much in
there I don't know how it could ever get done.
But there were a couple of areas I was
going to ask you about that I really didn't see in
there. And one of them was the issue of licensing by
MR. FLACK: Licensing by?
MEMBER KRESS: Test.
MR. FLACK: Test.
MEMBER KRESS: For PBMR. I didn't see
that discussed in there anywhere, and I was thinking
there might be a section talking about the -- where
would that fit into the regulatory structure at all,
if at all, and is it part of the thinking, or is there
any research need? Like, you know, research in the
sense of how that would affect your decision-making
process, or what licensing by test actually means. I
didn't see anything on that.
MR. FLACK: Well, we have been thinking
about it. I don't know if --
MR. LYONS: This is Jim Lyons from NRR
again. This is one of the areas that we've looked at.
There is certainly the ability within Part 52 to
license a prototype reactor, and then you would -- you
know, and then you would perform tests on that
prototype reactor, and then you could continue on with
using that reactor as a way of developing your I guess
licensing by test.
I don't know if we've really completely
looked at how we would do that. One of the things
that may happen if we do a license by test or a
prototype reactor is that we may put extra features or
have -- you know, request extra features be placed on
that plant to provide us any, you know, assurance that
there wouldn't be any real problems.
But it's part of our process. It's
something that could be done, but I don't think that
we saw any real need in the research area to address
MR. FLACK: Yes, it's a difficult question
to deal with until we actually get a plant in as well.
MEMBER KRESS: Well, along this same line,
one of the issues that is sure to arise with PBMR and
GT-MHR, GA, just in general, is how do you know that
you actually have the fuel quality that's required
when you -- after you load it into the reactor.
And one way to do that is what you do with
light water reactors -- you look at the level of
activity in the primary system, and you infer the
quality of the cladding or the quality of the fuel
from that. And the question I would have is: isn't
there some concept like that being thought of for the
pebble bed modular reactor and the others?
So that during start-up of the operational
phases you can say, "All right. Based on what we see
now, you don't have the fuel quality you said you were
going to have in your licensing basis, so you've got
to do something." Is that part of the plan? Is that
MEMBER ROSEN: It's not in there as
explicitly as you just described it, but it is in
there implicitly. The way I like to refer to it is a
defense-in-depth on fuel performance during operation
and postulated events. And you can think of that
defense-in-depth as building in quality absolutely
correctly every time, and that focuses you on the
manufacturing part of the process, to look at the
process and the product specification, make sure
you're doing it right every time.
MEMBER KRESS: You would look at process
MEMBER ROSEN: And that's in our plan.
MEMBER KRESS: Now we're wanting to look
at product, too.
MEMBER ROSEN: Okay. Then, look at the
products. But before it ever gets put into a reactor
and starts operating, then you get to the next
defense-in-depth place, which is monitoring
operations, and looking at activity and monitoring
The question comes up, though, is that
method qualified? Is that method reliable?
MEMBER KRESS: Yes.
MEMBER ROSEN: Is there data that shows
MEMBER KRESS: That's exactly my question,
yes. Is there something in the plan that will answer
MEMBER ROSEN: Yes. Yes.
MEMBER ROSEN: Well, I think you have some
advantages here, if you're thinking about pebble bed,
that you don't have in light water reactor. You could
do destructive examination on the fuel.
MEMBER ROSEN: That brings me to the third
MEMBER ROSEN: And you could afford it.
MEMBER ROSEN: Yes, that's right.
MEMBER ROSEN: But you couldn't do that in
the light water reactor, say, I'm going to destroy
this assembly and say, therefore, the other 80 are
okay. You know, that wouldn't be -- it wouldn't make
any sense. But if you're talking about thousands of
pebbles, you can statistically sample them and do
destructive evaluation and gain some real confidence
as to the quality of the pebbles.
MEMBER ROSEN: Right. And that's --
MEMBER KRESS: You can't, because they
have to be irradiated. And you're not going -- that's
the problem. You've got to run through the
MEMBER ROSEN: That's the research issue
is how do you identify, from looking at the
destructive evaluation of a non-irradiated pebble, how
an irradiated pebble is going to work.
MEMBER KRESS: Yes. You can't make that
judgment. You have to irradiate them, and that's
where your statistical problem shows up. You just
can't irradiate enough of them to get the right
statistics to qualify the level of failure or pebbles
that you think you have to have.
MEMBER ROSEN: So that's the answer to the
research program, Dr. Kress? I mean, I was suggesting
that there ought to be a research program to get to
that answer. But if you already know it --
MEMBER KRESS: Well, you have to -- you
just can't irradiate enough pellets over the timeframe
to do that. You can't do it.
MEMBER ROSEN: Well, the approach that's
taken when you have billions, literally billions, of
fuel particles in the reactor is to test hundreds of
thousands in a materials test reactor to qualify them,
and then, even if you --
MEMBER KRESS: Yes, to the right
MEMBER ROSEN: To the right conditions,
temperature, fluents, burnup, whatever it is, and even
if you have zero particle failures you don't
extrapolate if you have zero in the billions. There's
a statistic that you can use to project what the
number would be.
MEMBER KRESS: But it's an extremely
MEMBER ROSEN: But the question comes up,
are the test statistics going to hold true in the fuel
that you make later?
MEMBER KRESS: That's right, because
you're only testing one batch.
MEMBER ROSEN: In a sense, that's true.
So you need to show that that's going to continue over
the life of the fuel supply and the life of the plant.
And so you're stuck with, well, how do I then monitor
later on fuel that's coming off the assembly line and
put in the reactor?
MEMBER ROSEN: Well, these are good
MEMBER KRESS: But you're saying that's
implicit in --
MEMBER ROSEN: Yes. And if you look at
the plan, and you look under the fuel performance
piece, you see something called fuel manufacture. And
our plan is to try to understand as best we can what
are the really critical aspects of fuel manufacture to
get quality in the product and also performance in
reactor and in accidents. And there is work going on
internationally to try to understand what it is that
in the process and the product specifications that
will do just that. So we're following that.
And the question comes up, should there be
a regulatory footprint in some sense on that piece as
a way of assuring defense-in-depth? I think there's
a general belief that we ought not to regulate the
product but the performance, which puts you into the
next step, which is looking at operating performance.
If you're going to have --
MEMBER ROSEN: It would be preferable to
-- in my view, to regulate the performance. But in
the case we're talking about, because of the
importance of the product protocols, it seems to me
that the regulatory footprint in the processing of the
fuel is crucial.
MEMBER ROSEN: Yes. And part --
MEMBER KRESS: And I think it's analogous
to digital I&C for controls and --
MEMBER ROSEN: And part of the
preapplication review, a big part of the fuel
performance review, is to look at the tradeoffs of,
where do you put your regulatory imprint. Do you put
it in the manufacturing piece and/or also in operation
and/or testing fuel after it has come out? I mean,
you can put it anywhere you want.
The data I have seen on monitoring
operation and looking at some examples going back to
the German testing program, there are failure modes
that will not be caught by monitoring coolant
activity. They don't --
MEMBER ROSEN: Stu, why do you think it is
only one answer? Why do you think that?
MEMBER ROSEN: I'm not saying there's one.
MEMBER ROSEN: Whatever answer you come up
with now is the answer forever. I don't think so.
MEMBER ROSEN: I'm not saying one. I'm
MEMBER ROSEN: I think the answer is
something you -- in the beginning you do almost all of
what you've talked about, until you begin to get
confidence that you don't need to -- that you do not
need to do pieces of it and can begin subtracting away
MEMBER ROSEN: And we very much believe
that this whole area will be a Commission policy
decision. And what we plan to do in our SECY paper at
the end of this -- not so much the advanced reactor
research plan development process, but the end of the
preapplication review, is to lay out those defense-in-
depth opportunities for catching fuel that may not
perform well in an accident, and talk about the
advantages and the disadvantages in each one, and lay
out our -- those options and lay out our
recommendation, and then the Commission will have to
make a decision.
But I'm not going to say what that final
answer is, but it is, we believe, very much a
Commission policy decision on where that imprint or
multiple imprints need to be.
MEMBER KRESS: Well, while I'm on a roll
here, I want to have one complaint. There's a
statement in the document -- now I don't have mine
with me, so I don't know what page it's on, but it's
to -- the statement says that the -- I won't be able
to find it, because I've got it dog-eared -- that the
evolution of severe accidents and source terms will be
similar to current operating plants.
Now, I just think that's flat-out wrong
for IRIS, and it may be wrong -- I mean, you can't
even relate it to PBMRs. But for IRIS I think it's
flat-out wrong, and I think there's contrary evidence,
especially for high burnup fuel, and IRIS, of course,
is going to go to really high burnups. And I just
don't think you can make that statement.
And I didn't see in the plan, Dana,
anything on research for core degradation and fission
product releases for high burnup fuel of the LWR type.
MEMBER POWERS: It's totally irrelevant,
MEMBER KRESS: I know it is. Yes. So
that's a complaint. That's the one major complaint I
CHAIRMAN APOSTOLAKIS: You have commented
on the whole report now, because I want to do that,
too. You are not just commenting on the --
MEMBER KRESS: Yes, that's right.
CHAIRMAN APOSTOLAKIS: Okay.
MR. ELTAWILA: I agree with you on IRIS.
And as I indicated earlier, we have very limited
interaction with Westinghouse on the design of IRIS.
So we really -- this plan does not really address IRIS
in any extent. So your points are well taken. And
once we -- we are going to keep that plan as a living
document. Once we get information about IRIS, we will
modify to address this plant design.
MEMBER KRESS: Yes, okay. Well, another
question I have is you had a section in there
discussing -- I don't even remember where it was
either -- discussing underground siting.
CHAIRMAN APOSTOLAKIS: Yes, I remember
MEMBER KRESS: It's a good idea, but I
don't think anyone is seriously considering that, are
they? I mean, is that -- that wouldn't be a priority
in my research.
MR. FLACK: Underground is pretty much the
GA design, the GT-MHR --
MEMBER KRESS: Well, that's partly
MR. FLACK: Yes.
MEMBER KRESS: Okay. One other thought.
You talked about, for the PBMR and the pebble -- the
gas-cooled reactors that severe accident issues
include water ingression and air ingression. I'm not
so sure water ingression is a severe accident issue.
I think it's a long-term degradation issue and not a
severe accident issue, so you might want to rethink
that one a little bit.
I guess that's my list of items, George.
CHAIRMAN APOSTOLAKIS: Well, I have a --
I mean, if we are talking about broader issues now, it
looks like -- first of all, you mentioned PIRT some
place. I can't find it now, but I remember. I know
it's a major deficiency on somebody's part not to know
what it is. But I've been on this committee for five
years, and people use the word "PIRT" as if everybody
knew what it was from birth. Is there any place where
I can go and find out what it is? I don't know what
MEMBER KRESS: There's a document called
CSAU that --
CHAIRMAN APOSTOLAKIS: Oh, is that part of
MEMBER KRESS: Yes.
CHAIRMAN APOSTOLAKIS: Can you -- I know
what it is, but I'd like to know how it's done.
MEMBER KRESS: Well, I don't want --
CHAIRMAN APOSTOLAKIS: And I know that the
thermal hydraulicists are ecstatic about it.
MEMBER KRESS: I don't know what the NUREG
CHAIRMAN APOSTOLAKIS: So I'm very
Now, that brings me to another point,
which is related to my question about efficiency and
the use of risk information. It's a matter of style,
of tone, how to write this rather than really
substance. I know what you mean, although the
substance is effective.
I'm willing to bet that what's going to
happen is you're going to have the PRA at the high
level, and then you're going to use a hell of a lot of
defense-in-depth arguments to really preserve most of
the criteria you have now.
And here is the sentence that justifies
that. I'm editing now as I go. However, until
appropriate models can be accurately developed for
these new designs to define and prioritize these
issues, conventional methods will -- may need to be
applied." So this is dismissing now PRA. This gives
you a way out.
I would say -- I would change the tone of
this and say the following. Yes, we've had all sorts
of -- I'm reading from the human factors, but I don't
want to single them out, because I don't think it's
unique to them. Yes, you've been looking at task
analysis, at procedure development, training program
development. Please tell us how important these
things are in the risk environment.
I agree -- you see, now they are putting
the burden on the reliability analysts. Until the HRA
models are accurate, we will continue doing what we're
doing. I'll reverse that. Show me why what you're
doing is important to risk, and then you put a hell of
a lot of pressure on a lot of people to actually
quantify, because if that pressure is not there they
will never quantify, and I say that with a license --
I mean, the power uprates.
The answer was, we have an engineer who
looks at the -- who looks at it. You know, the
available time was 42 minutes, now it's 39, and he
says it's okay. Now, where is the incentive of
quantifying if that's the easy solution? An engineer
looks at it and decides it's okay.
So it seems to me it's a matter of tone
rather than really substance. Ask all these people to
tell you why all these requirements are important from
the risk perspective.
Now, they may come back and say, well,
gee, not everything is important, you know, from --
with respect to CDF, but there are other criteria.
Well, that would be progress in itself, because I do
know there are other criteria that are not
MR. ELTAWILA: If we sound quiet on this
side, it's because Mary keeps saying, "I agree with
you," so I -- we are really --
CHAIRMAN APOSTOLAKIS: She agrees with me
MR. ELTAWILA: No, with you. So we are
agreeing with you, and I think that's a good point.
CHAIRMAN APOSTOLAKIS: I think that if you
say that clearly here, then I think you are well on
your way of having an efficient -- I'm not saying that
it will always work, but at least you are shifting the
MR. ELTAWILA: Okay.
CHAIRMAN APOSTOLAKIS: You have to tell me
why this particular requirement is important from the
risk perspective, whatever "risk" means in this
context. You know, it's not -- nothing is important
with respect to CDF, by the way, unless you demolish
the reactor. There may be other intermediate
objectives that are effective, and at least we will
have them on paper.
Ah, come on, Steve. You know you have to
do big things to see a big change in the CDF.
MEMBER ROSEN: Abolish the reactor?
MEMBER KRESS: Almost.
CHAIRMAN APOSTOLAKIS: Almost.
MR. FLACK: Well, there are sensitive
issues like, for example -- that would be difficult to
quantify. And since you brought up human factors, it
would be like a question of whether an operator is
qualified, what would be the risk from an unqualified
operator? I mean, these are --
CHAIRMAN APOSTOLAKIS: All I'm doing is
I'm shifting the emphasis.
MR. FLACK: No, I understand. I
CHAIRMAN APOSTOLAKIS: See, as long as you
say it's the problem of the HRA analyst, they will
never get anywhere. If you say, "No, it's your
problem, you tell me whether what you're doing here is
risk-significant," then you will see a very different
attitude. I repeat, I don't want to single out the
human factors. I mean, it applies to I&C, and I am
sure it will apply to other things with the new
I&C, too -- I mean, you look at it, there
is a lot of work, and this is -- at the end it says,
"Oh, by the way, we really ought to quantify it, too."
Well, yes, sure.
MEMBER POWERS: John, let me ask you a
question. Since, obviously, we've blown your
presentation completely to hell, we might as well just
continue this trend. Teach you to make viewgraphs, by
We have just had the IPEEE insights
document given to us, and with arguable exceptions we
find two things. One is the estimates of risk that
the licensee has submitted for fire were surprisingly
high comparable to operational risks. And the
techniques that they used to derive those were
And, okay, so you can argue that maybe the
risks are not as high; they were just very
conservative when they went through and did it. On
the other hand, you can take them at face value and
say, "Hey, one of the features of our current crop of
reactors is there are very susceptible to fire and is
an accident initiator." And maybe we don't want that
for advanced reactors.
I mean, it does seem kind of a crude thing
to have a sophisticated, high-technology device like
a nuclear reactor susceptible to fire as an accident
initiator. Why, then, wouldn't you want to put
priority on having good technologies for evaluating
fire and advanced reactors?
MR. FLACK: I guess you looked through the
report for that piece and didn't quite find it there.
Fire is a difficult issue. It's a spatial interaction
type of issue that you need to deal with almost on a
plant-specific level. So it's difficult to understand
what that risk would be until a plant actually comes
in and says, "Here is what I got, and here is where
things are," and then you can study it from that
But I guess, again, this comes back to the
code issue, whether or not our codes --
MEMBER POWERS: I'm looking at -- I mean,
I'm taking your lead in saying you're trying to create
an infrastructure here, a capability --
MR. FLACK: Right. Exactly.
MEMBER POWERS: -- and so I'm asking,
isn't this a capability that you want to have?
MR. FLACK: I would -- the answer is, of
course. I mean, it's certainly an important risk
contributor we see in these plants. How they play out
in advanced plants, passive designs, is yet to be seen
in what we'll have -- how we'll approach that problem.
Again, it's a difficult issue to deal with
without seeing a plant. But no, it's certainly
external events. Seismic and fire are two that's part
MR. RUBIN: Can I just -- John? This is
Alan Rubin from the PRA Branch and also the IPEEE
External Event Program. As part of the advanced
reactor research plan, we do include external events
in the PRA -- different operational states as well as
external events, fire, and seismic. So we --
MEMBER POWERS: We don't doubt that you
include them. I'm really asking a question on the
quality of tool that you have available to include
them. For instance, a noted member of this panel, an
exemplary member of this panel, devised a code some
time in the past, and he recount for you the details
of it, called COMBURN, and we universally find COMBURN
gets used beyond its stated limits of applicability,
because there's nothing else available.
And the problem I see that you have is
just what John outlined for you. If you're going to
analyze fire, you're going to have to do it on a
plant-specific basis. If you wait for a plant to come
along in order to do a fire analysis, then there isn't
time to develop a better tool, because you're under
the gun and people are yelling at you to do it faster,
better, cheaper, and things like that.
And so COMBURN lives forever. And though
I know the author of COMBURN is an exemplary
individual, a noted phenomenologist in this world, I
don't think even he thinks that it deserves to live
MS. DROUIN: Dana, let me just also
interject something. We have a huge research
initiative going on in the area of fire that would
support this effort. I mean, that's looking into
things -- you know, the models. I think they've been
in front of the ACRS.
MEMBER POWERS: I get confused, Mary, over
the strategy in preparing the report. It's all well
and good that you have a research effort going on
there, but shouldn't you lay it down here to say, "And
we need that research effort"? I mean, this wasn't a
litany of things that you're supposed to do. It's the
things that are supposed to be done.
MR. FLACK: No, that's a good comment.
MS. DROUIN: I mean, the whole intent was
to take advantage of what was going on in that
program, and, yes, we probably shouldn't have been so
silent on it.
MEMBER KRESS: I think we have reached the
end of the allotted time for this subcommittee
meeting. I would like to, you know -- lest you go
away thinking we were too negative, I think -- I think
you're on the right track with this thing, and you did
a magnificent job of identifying the -- what the needs
are and the gaps that might exist. And it's a
comprehensive, well-written document.
So I think you're on the right track, and,
you know, we got some specific comments. I don't know
if those were sufficient for feedback or should we
have a letter or not. Probably --
MR. FLACK: No, we weren't looking for a
letter at this point.
MEMBER KRESS: Okay. Well, the other
question I wanted to ask is: when should we think
about having you back again on this same issue? July
meeting, is that too soon, or is that too late, or
what do you think?
MR. FLACK: Are we talking about
subcommittee or full committee?
MEMBER KRESS: Well, probably need a
subcommittee and a full committee, too.
MR. FLACK: On this subject.
MEMBER KRESS: Yes. When do you think it
would be worth thinking about another meeting? That's
my question, I guess.
MR. ELTAWILA: We are ready any time you
want, Tom, so just set the schedule according to your
-- the availability of you and other members of the
CHAIRMAN APOSTOLAKIS: There has to be
MR. ELTAWILA: So I think we will have to
start scheduling all of these meetings between now and
to end by August, to be able to finalize the plan to
go to the Commission. So if --
MEMBER KRESS: That's why I was thinking
if it was in July we --
MR. ELTAWILA: -- every month you want a
meeting, we will be supporting that.
MEMBER KRESS: Well, thanks. I guess
we're going to talk about -- yes, go ahead. One more
MEMBER ROSEN: I want to say one thing.
I associate myself with all of the comments of the
eminent Dr. Kress, but I am still concerned about the
scope. So take that away.
MR. FLACK: We gotcha.
CHAIRMAN APOSTOLAKIS: And next time,
John, just come with two viewgraphs. It doesn't
It just doesn't matter.
Okay. Thank you, gentlemen.
MEMBER KRESS: Thank you very much.
CHAIRMAN APOSTOLAKIS: This was a useful
discussion, and we will recess now. How much time do
you guys want? Do you want a full hour? Okay. Shall
we be back at 1:50? 45 minutes? 1:50, okay.
(Whereupon, at 1:08 p.m., the proceedings
in the foregoing matter went off the
record for a lunch break.)
CHAIRMAN APOSTOLAKIS: Next item, "CRDM
Penetration Cracking and Reactor Pressure Vessel Head
Degradation." Dr. Ford, please lead us through this
MEMBER FORD: On April 9, presentations
were made to the Materials and Metallurgy and the
Plant Operations Subcommittees on the 2001-1 and 2002-
1 bulletins relating to cracking of CRDM housings and
the degradation of CRDM housings. Obviously there's
a tremendous amount of work going on on those two
issues by both the industry and the staff. And on
April 9, we heard preliminary information especially
on that from Davis-Besse related to the root cause and
generic implications of the degradation.
Today, we're going to hear an update on
these issues, and it's primarily for information. The
staff have not requested a letter from us. Future
meetings with the subcommittees and the full ACRS are
scheduled somewhere in the near future for which there
will be a letter, presumably, requested. Jack, you
didn't have any comments?
MEMBER SIEBER: No.
MEMBER FORD: I'd like to move on then.
We're going to take it in order, from the industry
perspective, given by Larry Mathews, and then we'll
move on to the Davis-Besse, and then finishing off
with the presentation by the staff. So Larry is the
Chairman of the MRP Program and from Southern Nuclear.
MEMBER SIEBER: What's MRP?
MEMBER SHACK: The first test.
CHAIRMAN APOSTOLAKIS: What's MRP?
MEMBER FORD: Materials Reliability
Program, sponsored by EPRI.
MR. MATHEWS: Like Dr. Ford said, I'm
Chairman -- is this on? I'm Chairman of the Alloy 600
Issues Task Group of the Materials and Reliability
Program. I work for Southern Nuclear, in case you
care, or at least they pay me. I don't do much for
MEMBER POWERS: An extraordinarily honest
MR. MATHEWS: Not to imply I don't work.
I just don't --
These are kind of four topics I'd like to
run through fairly quickly here today and provide a
summary on: The Alloy 600 82/182 strategic plan that
we have developed, an update on where we stand on
crack growth rate issues, some brief words on the risk
assessment and the probablistic fracture mechanics
that we're doing for the reactor vessel head
penetrations and then, basically, how we are
responding to the Davis-Besse issue at this point.
This is basically an outline of the
strategic plan that the MRP has put together to
address the Alloy 600 and the 81/182 issues. The plan
has a problem staying on the goal and mission of
trying to manage the issue, how we're going to go
about it, what the roles of our various stakeholders
are. And then we have a strategy right now, which are
the five areas you see here.
Basically, on the -- are you looking for
MR. MATHEWS: Are you looking for the
PARTICIPANT: No, no, no.
MR. MATHEWS: Oh, okay, okay. On the butt
welds, the basically strategy we've laid out is we're
going to rely primarily on the ASME Section 11, the
guidance for inspections and the frequency, but we're
driving and we're trying to drive improvements into
technology for doing those inspections. And,
primarily, Appendix 8 has to be implemented by next
fall, and at that point, all the inspections will be
done by qualified inspectors.
One of the things we will have to be
looking at potentially in more detail is the
frequency, is it appropriate, et cetera? But that's
where we are right now is we believe Section 11,
coupled with Appendix 8, will be the appropriate way
to do it. There is a potential issue with the pass
rates and the qualifications of the inspectors, and
we're trying to address that right now.
There's other areas up here, excuse me.
The head penetrations in the near term, we put
finalizing a safety assessment, but the real thing
we're doing here is putting together mockups to drive
the technology for doing volumetric inspections and to
demonstrate those inspections. We're having mockups
built that will be used in blind tests this summer for
vendors that will be qualifying to do volumetric or
under-the-head inspections next fall. There's also a
mockup that was built that was available for people to
use early and then another one for the spring outages.
In the area of the longer term, what we're
doing to do is get out inspection guidelines on what
people ought to be doing, as far as inspecting their
head penetrations. And then we want to work with the
NRC and ASME to make sure this is, you know, all in
conjunction with what's the right thing to do as far
as inspecting the heads.
All the other locations, we're working
with the owners' groups to see what's already been
done. We don't want to duplicate anything for all the
other Alloy 600 locations. And where there are holes
in what they've accomplished, we know they've done a
lot of work, where there's holes in what they've
accomplished, we'll work with those owners' groups and
vendors to figure out where's the right place to
develop those guidelines and get those programs
And, ultimately, the goal is to get out a
management guideline for all the locations that would
either provide information on how to manage it for
your plant or direct you to where it would be
One of the first things we want to work on
is the inspection plant. We have draft inspection
plant out now. This is something we need to get with
the staff and make sure we're all in agreement on
what's the right thing to do in the inspection. But
it basically marches toward -- as the plant gets older
and it has more time at temperature on the vessel
head, the inspection should become more rigorous, if
you will, going from a visual to ultimately,
potentially all the way down to you must do a
volumetric on some frequency. We haven't finalized
that. That's in the final stages at this point.
In the area of crack growth rate for Alloy
600, what we're trying to do is figure out what's the
right crack growth rate people ought to be using when
they're trying to do evaluations of cracks in the
Alloy 600, initially looking at the base metal. We've
created an expert panel. That expert panel has met
several times, and they've screened databases
available in the world. They're trying to refine
their approach. It's been consolidated, but
apparently, recently, we were very close to publishing
the report, but then one of the labs said, "Well, we
want to take another look at our own data."
And then while that's going on, Davis-
Besse occurs, and so especially with respect to what
the annulus environment might be and the impact of the
annulus environment, the experts said, "Well, we know
what we said," and I'll tell you what that was in a
second, "but before we publish we want to take another
look at that and make sure we still believe it." And
so they're meeting next week. It's a sid bar meeting
to a meeting going on in France to look at that issue.
CHAIRMAN APOSTOLAKIS: So when you say
"curve," what are the axes? I mean one must be the
MR. MATHEWS: Growth rate and stress
CHAIRMAN APOSTOLAKIS: Stress intensity.
Now, isn't there any uncertainty in those curves? I
mean are you displaying --
MR. MATHEWS: Oh, yes, quite a bit.
CHAIRMAN APOSTOLAKIS: And you are
MR. MATHEWS: Pardon?
CHAIRMAN APOSTOLAKIS: You are displaying
it or are you just showing one curve?
MR. MATHEWS: What we're proposing is a
couple of different approaches.
MEMBER FORD: Well, before you -- are you
going to continue answering that specific question?
MR. MATHEWS: Yes. Go ahead. What were
you going to say?
MEMBER FORD: Well, answer that question,
because I want to come back to that.
MR. MATHEWS: Okay. What we've done is
we've taken the whole database and we've come up with
a curve that we feel can be used for the deterministic
evaluation of the crack growth rate for real flaws.
And, basically, any flaws that you're trying to
evaluate to leave in surface, the main ones that have
been evaluated are flaws that are either ID axial
flaws or if they are on the OD, they're below the
weld. Anything above the weld it has to be a leakage
path, and we can't leave that in service, so we
wouldn't be evaluating real flaws above the weld.
We do want to evaluate hypothetical flaws,
for instance, all in the circ direction to determine
if it flows into the safety, how long have we got and
that sort of thing. And so above-the-weld flaws
they've recommended a factor of two to account for the
chemistry in the environment, but that's one of the
things that the guys are going to take a look at next
week in France, will make sure that Davis-Besse
doesn't really throw a monkey wrench in.
CHAIRMAN APOSTOLAKIS: But are on the
issue of uncertainty now? You said it can be used for
MR. MATHEWS: Right. And the curve that
we're proposing is for deterministic evaluation is
like the one that would fit the 75th percentile of all
the heats and material in the database.
CHAIRMAN APOSTOLAKIS: Oh. So you're --
MEMBER FORD: I think this is an ongoing
argument within the industry for quite some time, and
you've got a big scattered database, experimental.
How much of that scatter is due to experimental
control? Is much of it due to heat variations, for
instance, in the materials in that database? And we
have requested that at the next meeting that that
database will be shown to the committees and how that
has been analyzed. So that will directly answer your
CHAIRMAN APOSTOLAKIS: Because it would
seem to me to be an ideal place for a family of
curves, would it not?
MEMBER FORD: For a --
CHAIRMAN APOSTOLAKIS: A family of curves
rather than one curve.
MEMBER SHACK: People recognize there is
a distribution. Just for deterministic evaluation
you'd like to have --
MR. MATHEWS: No, but if you knew exactly
-- if you knew exactly.
CHAIRMAN APOSTOLAKIS: No. CGR data for
base material feeds directly into the PRA.
MR. MATHEWS: Well, that's not how we feed
it into the probablistic approach, though. Instead of
feeding it into the probablistic approach as a single
curve, we put the whole database and all the scatter
of the database to be sampled in the probablistic
approach. The whole scatter for the whole database is
put into the probablistic analysis.
CHAIRMAN APOSTOLAKIS: I'd like to see
MEMBER FORD: That is one of the things
we've been asking that we do all see the database so
we can understand the reasoning behind these words.
MR. MATHEWS: Yes. And some of the staff
is saying but we haven't shown them the ACRS. And
part of the reason is it's in a state of flux right
CHAIRMAN APOSTOLAKIS: So you're going to
do this in a subcommittee meeting?
MEMBER FORD: We'll do it in the
subcommittee and present it at the full committee,
MR. MATHEWS: And hopefully we can do that
at the next meeting.
MEMBER FORD: Correct.
MR. MATHEWS: I think we'll be much closer
and we can do that.
MEMBER FORD: Could you go back to your
MR. MATHEWS: Sure.
MEMBER FORD: The implications of the
Davis-Besse, your last bullet, is that in terms of the
question as to what the environment is in the
MR. MATHEWS: Yes. That's what -- I
believe that's what the experts would want to take a
look at. They had made some assumptions, some MULTEQ
calculations and some other discussions amongst the
experts about what are the possible environments that
could be in there in the annulus region, and then what
effect would that have on the crack growth rate? And
they came up with what they felt was a conservative
multiplier, a factor of two.
Given the situation at Davis-Besse,
thought, they said, "Well, I don't know that it's
going to change, but let's take a look at it and see
if there's anything coming out of the Davis-Besse
situation that would say that environment that we
predicted is inappropriate to use for a
circumferential crack growth.
MEMBER FORD: And, again, that information
will be discussed, presumably, at the next meeting,
this specific information.
MR. MATHEWS: We hope to have our report
published well in advance of that meeting, and we can
come talk about it.
CHAIRMAN APOSTOLAKIS: Next meeting.
MEMBER FORD: Well, in the near future,
maybe one, two months time.
CHAIRMAN APOSTOLAKIS: Subcommittee
MEMBER FORD: Correct.
MR. MATHEWS: Also, the expert panel they
met very recently to look at the weld metal Alloy
82/182 and what we know about the crack growth rates
in the weld metal. And they will be coming back to
the MRP with recommendations on where there's holes in
that database, and there are likely to be some because
it's a limited database and where testing may be
There's also a research effort that's
being undertaken right now by EPRI, and it's a DOE
part of the NEPO Program to look at some crack growth
rates in weld metal. And there may be some additional
base metal crack growth rate in there, I'm not sure.
And we will certainly be willing to continue to update
you as we get more data, maybe provide you some.
In the area of the risk assessment work,
the approach is to predict the probability of leakage
based on the industry experience and where we've seen
links and modeling that in a Weibull model, Weibull
statistics model. Then compute, after a leak
develops, the probability of a nozzle ejection,
looking at or considering the initiation and growth of
a circumferential flaw above the J-groove weld. We
can factor into that inspection and the probability
that a leak might be detected prior to growing to an
CHAIRMAN APOSTOLAKIS: How would you do
MR. MATHEWS: I left that slide out. What
you do is as the model progresses through the time,
it's a statistical model but it progresses through
time, and at given points in there, depending on the
inspection frequency that you put in, you can put in
a probability of detection. And if you -- and you do
a sample on that. And if you find the probability
that it is detected on that particular sample, you
take it out of the database for an ejection.
And if you don't, it goes on down to maybe
the next level of inspection or the next whatever.
You just the run the statistics, and if you put a
probability of detection of 80 or 90 percent in there
and you're doing inspection at a certain point in
time, then 80 or 90 percent of any flaws that might be
in existence there would be taken out of the database
or if they're not --
CHAIRMAN APOSTOLAKIS: Would that be
consistent with the Davis-Besse experience? An 80, 90
percent probability of detecting?
MR. MATHEWS: Today, I would say, yes,
probably. I'm not sure what the POD, probability of
detection, that we're going to put in there. That's
just the way it's modeled, and we'll have to decide.
We haven't settled down on exactly what kinds of
inspections or when they would be into the model to
figure out the risk. But, you know, before Oconee the
world was different than it was after Oconee, so
people look at things a whole lot different.
CHAIRMAN APOSTOLAKIS: See, what worries
me is that I don't know how many times the world is
going to change.
MR. MATHEWS: Oh, yes. I know what you
CHAIRMAN APOSTOLAKIS: I mean it
shouldn't. It should change any more for the current
generation reactors. That's my problem.
MR. MATHEWS: Knowledge isn't perfect, I
CHAIRMAN APOSTOLAKIS: Boy, you can say
MR. MATHEWS: Yes. Anything else?
Finally, what we do is we grow the flaw to the
critical flaw size on a statistical basis from Monte
Carlo sampling, and some of them grow to critical flaw
and some of them don't. And then they take the
fractions that do and that's the probability there.
Couple that with the probability of a
conditional -- I'm sorry -- yes, with the conditional
core damage probability from a small break or medium
break LOCA, and you have the core damage frequency.
What we're going to do is assess the potential impact
on the conditional core damage probability of the
collateral damage. We think it's going to be minimal
that might occur from an ejection.
CHAIRMAN APOSTOLAKIS: Is it clear to
everyone why nozzle ejection is the issue here?
MEMBER SHACK: That's what causes your
MR. MATHEWS: Yes.
CHAIRMAN APOSTOLAKIS: Oh, that's --
MR. MATHEWS: In almost all -- you know,
if you look at all the times that plants run most of
the time, almost all the time these plants are up at
power and all the control rods are essentially all the
CHAIRMAN APOSTOLAKIS: So what's the
MR. MATHEWS: The inside of a nozzle is
about two and five-eighths inches, I believe.
MEMBER SHACK: But when the whole thing
comes out, it's like four inches.
MR. MATHEWS: Yes.
CHAIRMAN APOSTOLAKIS: Oh, okay. So then
MR. MATHEWS: Well, you've still got to
get through the part that's left. If you have a circ
flaw above the well, then you've got a segment that's
left from the well down that's not ejected and the
inside diameter of that is two and something inches,
and if it's a control rod location, it will still have
a shaft in it unless that gets pulled on out too.
CHAIRMAN APOSTOLAKIS: How will you go to
the condition core damage probability? I mean you
would just consider the new probability of a medium
LOCA? The probability of nozzle ejection would be --
MR. MATHEWS: Well, the CCDP is the
conditional core damage probability.
CHAIRMAN APOSTOLAKIS: Right.
MR. MATHEWS: Given that you have a
medium-break LOCA, the plant risk assessments already
have looked at what is the probability that you have
core damage, given that you have a medium-break LOCA.
And that goes through all the possible failures of
your ECCS systems and all of that.
CHAIRMAN APOSTOLAKIS: Would you consider
dependencies between the initiating event and some of
the other events?
MR. MATHEWS: Yes.
CHAIRMAN APOSTOLAKIS: In particular
SCRAM? Would SCRAM be affected?
MR. MATHEWS: Yes. And that's what we
would look at as would there be collateral damage from
the ejection of a control rod nozzle that could make
that conditional core damage probability of a medium-
break LOCA higher than if it was on a pipe somewhere.
We'll look at that, and if it would make that
conditional core damage probability, given the LOCA
here as opposed to on a pipe higher, then that effect
would be factored into the risk assessment. We think
that effect's going to be minimal and we've gotten
some preliminary work from the vendor, but we need to
CHAIRMAN APOSTOLAKIS: So you are also
looking at small-break LOCA, I see. All right.
MR. MATHEWS: From a risk standpoint, yes.
We're not doing a deterministic blowdown of a small-
break LOCA type thing, it's more of a risk analysis.
CHAIRMAN APOSTOLAKIS: Okay. You're going
to have to have experts again telling you what's going
to happen if you have a nozzle ejection.
MR. MATHEWS: Yes. And the vendors know
CHAIRMAN APOSTOLAKIS: And how it will
affect the SCRAM system.
MR. MATHEWS: -- what's up there, and
we're asking them to provide us input on that, and
they've given us some preliminary stuff, and we need
to follow-up on that and figure out how to factor that
input back into the risk assessment.
CHAIRMAN APOSTOLAKIS: So when will this
MR. MATHEWS: We were hoping to be through
this month, but everything's kind of taken a --
everybody's busy on Davis-Besse issues right now.
CHAIRMAN APOSTOLAKIS: Okay.
MR. MATHEWS: Some of the key elements of
the probablistic fracture mechanics analysis, which is
the major part of the risk assessment, is the
simulation of the leakage as a function of time and a
Monte Carlo model. That's based on our time and
temperature model using the fracture for the stress
intensity factors, for the various types of flaws that
would be in there as the flaws grow. The entire
database for the structure crack growth rate database
and the statistics, all of those statistics would be
fed into for the sampling and then the effects of the
inspection and the inspection reliability.
We have some very preliminary results for
a tight temperature plant, and I do stress
preliminary. First cut thereafter after you've an
inspection, the probability of nozzle ejection within
the first or so is less than times ten to the minus
three after you've done inspection. And then the
conditional core damage probability, the worst one we
could find on the high temperature plants was five
times ten to the minute three. Multiplying those two
together you get a core damage frequency in the range
of five times ten to the minus six.
CHAIRMAN APOSTOLAKIS: What is the main
reason why the probabilities are so low?
MR. MATHEWS: The main reason the
probability of an ejection is so low after you've done
an inspection is that you've found your leaks and
repaired them. But in a few cases, when you do the
statistical Monte Carlo approach, you can have some
very high crack growth rates on some of this sampling.
And those that grow very, very rapidly a few of them
may grow all the way to the ejection in the sampling
process, but it's a very, very few of them within one
cycle or before you come back to do another
CHAIRMAN APOSTOLAKIS: So you're assuming
that when the size reaches a certain level, then
there's a very high probability that they will be
caught by inspection and somebody will act on it.
MR. MATHEWS: Yes. Given today's
environment and what everybody knows about what they
need to be looking for, yes.
CHAIRMAN APOSTOLAKIS: Today's environment
MR. MATHEWS: After Oconee. I mean Oconee
showed that you could have a leaking penetration that
didn't have a lot of boric acid coming out down the
side of your vessel. And so now people are keyed into
you have to look for popcorn instead of big piles.
CHAIRMAN APOSTOLAKIS: And CCDP, why is it
MR. MATHEWS: Because a small-break LOCA
CHAIRMAN APOSTOLAKIS: No, a medium LOCA.
MR. MATHEWS: Okay. I'm not sure of the
exact square inches on the small and medium LOCA, but
we have lots of safety systems that are designed to
handle the LOCA and to keep the core from being
damaged. And the way you get damaged typically on a
risk assessment analysis on the LOCAs is something
fails, and there's probability and statistics put in
on a failure probabilities of your various safety
systems, and as you do that sampling on all the
systems and their probabilities, it comes out with a
fairly low probability for that size break that you're
going to have core damage.
CHAIRMAN APOSTOLAKIS: But how much credit
are you taking for scrap?
MR. MATHEWS: I'd have to go look at the
PRAs. I'm not sure if we -- I know in the design
basis axis on LOCAs I'm not sure we take any credit
CHAIRMAN APOSTOLAKIS: You're not sure of
MR. MATHEWS: I'm not sure they take any
credit on the design basis analysis, but on the risk
assessment I think we do take credit for SCRAM.
CHAIRMAN APOSTOLAKIS: The question is how
much because I don't know that we really know what's
going to happen if you have a medium-break LOCA at
MR. MATHEWS: Well, that's what we're
counting on the collateral damage assessment to tell
us: Does it have an impact on the conditional core
CHAIRMAN APOSTOLAKIS: Oh, so the
collateral damage is not part of these numbers?
MR. MATHEWS: Right. But like I say, the
conditional assessment we have from the vendors is
that it will have very minimal impact, if any, on the
conditional core damage probability. A break at the
top of the vessel is better than one that's at the
bottom, and the CCDP is for all breaks. But --
MEMBER ROSEN: A break at the top of the
vessel is better than one at the bottom but not for an
event when you want the control rods drives to
CHAIRMAN APOSTOLAKIS: That's right.
MEMBER ROSEN: Because the control rod
drives on a PWR are at the top.
CHAIRMAN APOSTOLAKIS: They're at the top.
MR. MATHEWS: That's right. And that's
what we have to see and have to assess in this
collateral damage is is there something that could
happen that would prevent a SCRAM or a significant
portion of the rods from not going in? Severing the
cables is great.
MEMBER SIEBER: It's designed to have one
rod stuck up.
MR. MATHEWS: At least one.
MEMBER SIEBER: And still get enough
MEMBER ROSEN: From a reactivity
MEMBER SIEBER: But if you damage the
adjacent rods somehow so that they don't, then the
probability of core damage goes up.
CHAIRMAN APOSTOLAKIS: That's exactly what
we're exploring here.
MEMBER SIEBER: Wiping out 60 of them, I
think, is pretty improbable.
MEMBER ROSEN: What we're worried about is
the steam environment, the jet environment and all of
that that will be up there in very aggressive to the
operation of the drives and the rest of the equipment
MR. MATHEWS: Well, most anything that's
going to -- the real concern, if there is one, from a
collateral damage, is if you could something that
would prevent the rods from moving physically.
MEMBER ROSEN: That's right.
MR. MATHEWS: Severing the cables, no
problem, they're going in. It's the --
CHAIRMAN APOSTOLAKIS: Physical, yes.
MR. MATHEWS: If you bend the tube or
something like that, that's the condition --
MEMBER ROSEN: If you have a plate right
above this, you know, above the point where you have
the break, and you create a high pressure environment
between the plate and the top of the head and what if
that plate cocks or something like that? I mean you
can imagine --
MR. MATHEWS: The insulation plate.
MEMBER ROSEN: Yes.
MR. MATHEWS: Yes. Those are pretty low.
MEMBER SIEBER: But what's the point if it
MR. MATHEWS: And that's what -- we have
to look at the --
MEMBER SHACK: We're not done.
MR. MATHEWS: We're not done yet, but, you
know, I think I heard yesterday and it's, at least to
my way of thinking about it, the first thing that's
going to happen is the voids are going to shut the
CHAIRMAN APOSTOLAKIS: The point is that
the five ten to the minus six number does not include
considerations of this type.
MR. MATHEWS: Right.
CHAIRMAN APOSTOLAKIS: Okay.
MR. MATHEWS: It includes an initial
estimate that it's going to be a very minimal impact
on that number, but we still have to go back and tie
all that together. We're not through yet.
MEMBER FORD: The first time that such an
analysis was given, to the staff that is, was during
the Duke presentations relating to Oconee, and my
question now is have there been any subsequent
discussions between you and the staff on this whole
MR. MATHEWS: We've had some fairly
detailed meetings with the staff on how we are
modeling primarily the probablistic fractured
mechanics part. We haven't really gone in in much
detail on the rest of the risk assessment. I think
we've laid this level of detail out and discussed it
with the staff. But on the probablistic fracture
mechanics and how we're modeling the crack and the
crack growth rate, we've met with Ed Hackett and the
research folks and their contractors and had a couple
of rounds of questions about how we're doing it versus
how they're doing it and trying to reach resolution on
some of those issues.
MEMBER POWERS: Suppose that after all
that they said, "Gee, you're just doing great. The
crack growth rates are great, everything's great."
How do you know the results are right?
MR. MATHEWS: Well, from the probability
of leakage is -- well, it's based on the experience in
the field, and we continue to get experience in the
field, and that is adjustable to match the experience
in the field. We're trying to be somewhat
conservative in this, and although it is a statistical
MEMBER POWERS: How do you know you're
MR. MATHEWS: There are a number of
details of how we're modeling the probability fracture
mechanics work that are -- like immediately upon a
crack going to a leak, we assume that it's instantly
like -- I think it's 20 or 30 degrees around branch of
the flaw, and it's going to take some time to initiate
a circumferential flaw, but we assume it happens
instantly. That's one thing.
CHAIRMAN APOSTOLAKIS: Would assuming the
presence of the degradation around this nozzle,
similar to that of Davis-Besse, be a conservative
thing to do and what numbers would you get?
MR. MATHEWS: It might be a conservative
thing to do, and we could model it. And I guess the
next slide is --
CHAIRMAN APOSTOLAKIS: You don't know what
number you're going to get, though, do you? Because
it's not just the normal rejection.
MR. MATHEWS: No, I don't know.
CHAIRMAN APOSTOLAKIS: You may have
MR. MATHEWS: There is the potential there
that if you got a nozzle that was in a situation like
Davis-Besse where there is a wastage cavity next to
it, if the cavity comes all the way around so that you
lose a back wall on the opposite side from where the
cert flaw is growing, it might have an impact on how
fast the crack grows. And we can model that and do
some studies on that, and we probably will do that,
where we remove the nozzle, the constraint from the
nozzle on the opposite side from the cert flaw.
CHAIRMAN APOSTOLAKIS: Well, that would be
an interesting case to see, a sensitivity case.
MR. MATHEWS: Yes. And it's not that hard
to do. There's gap elements on that side of the
nozzle that we just set them to a gap instead of an
interference and then see what happens to the nozzle
leaning over as a function of the crack growing.
Really, the way we've modeled it, it would only have
impact after the flaw hits 180 degrees in through
wall. If it's part through wall, we don't even model
that restraint; that's ignored. So, basically, we're
modeling it without that restraint already.
CHAIRMAN APOSTOLAKIS: So if you were
doing this analysis before Oconee, what number would
you get? You said earlier, "in today's environment."
So in yesterday's environment, what number would you
get, five ten to the minus nine or five ten to the
MR. MATHEWS: Well, we probably would
CHAIRMAN APOSTOLAKIS: Huh?
MR. MATHEWS: Yes. It probably would have
been in that --
CHAIRMAN APOSTOLAKIS: So all Oconee did
was raise the number from ten to the minus nine to ten
to the minus six? No? What? That's what they said.
MR. MATHEWS: I didn't do it before
Oconee, so I don't know what the number would have
been if we hadn't -- where it comes in is the
probability of the ejection.
CHAIRMAN APOSTOLAKIS: Yes.
MR. MATHEWS: Which starts from the
probability of a leak. We would have thought that
prior to Oconee in those flaws that have been recently
discovered, we would have felt that the probability of
developing a leaking penetration on a USPW head was
lower than it really was.
MEMBER FORD: I think the answer to both
your questions, to a certain extent, is, again, I
don't think you can -- the proof of the pudding, of
course, is observation versus theory, and we haven't
had any raw dejections, thank goodness. But you can
do it what's the probability of a number of through
wall -- through circumferential wall cracks that have
been observed. And that's essentially the approach
that Oconee did, or Duke did for Oconee, to compare
these predictions against the number of
circumferential cracks that they saw. Now,
admittedly, it's not going the whole way, you're
absolutely correct, but it is going -- they're doing
a check of observation versus theory.
MEMBER POWERS: What I guess -- I mean
you've certainly interpreted my question correctly,
and what I'm really struggling to find we apply this
probablistic fracture mechanics in a lot of regimes
now. This seems to be the first one where we don't
get answers like ten to the minus 45, which I thought
was a constant --
-- in probablistic fracture mechanics.
But I never -- I mean I'm sufficiently unfamiliar with
the technology that no one ever shows me that it
actually gives you good answers for any circumstance
that isn't fairly well-contrived laboratory
circumstance. And so I'm wondering as the geometry
has become more complicated, and here they're about as
complicated as comes quickly to mind, do we really
have data for any circumstances, I mean it doesn't
have to be a reactor vessel, but how about an
internally pressurized vessel of some sort where we
can show that indeed the probablistic fracture
mechanics has got all the physics in it so that if we
do what the speaker has said, we parameterize the
model conservatively, we should get a conservative
MEMBER FORD: Do you want to answer that?
MR. MATHEWS: I'm not a probablistic
fracture mechanics guy.
MEMBER POWERS: Well, that speaks well of
MEMBER FORD: I don't know -- quickly, off
the top of my head, I don't know --
CHAIRMAN APOSTOLAKIS: Are there any cases
where probablistic fracture mechanics gave
probabilities on the order of 0.2, 0.3 value? Or is
it an inherent thing of the methodology?
MEMBER POWERS: Ten to the minus 45 is a
really common number, I know that.
MEMBER SHACK: Just to come back, George,
you know, one of the things one observes is the way
things depend on diameters, your famous Thomas
correlation that you PRA guys love, you know, that
comes out of the fracture mechanics. The low
probabilities, of course, are for a large diameter
pipe where, again, for the crack to grow all the way
around the pipe, you have to grow a crack that's many,
many inches long. So, obviously, that's going to take
a lot longer than it does to, say, grow a crack around
a four-inch pipe. I mean the physical -- you still
have to grow 330 degrees, it's just the 330 degrees on
a four-inch pipe is a whole lot less metal than 330
degrees on a 24-inch pipe.
Now, it's very difficult, of course, to
get one-to-one comparisons, because we just don't have
a whole lot of data, but when you go back to the
database, you get probabilities of failure that aren't
all that -- you know, they're in the ballpark of what
you're computing for your probablistic fracture
mechanics; it's not a one to one.
We have experimental confirmation of the
ingredients; that is, you know, crack growth rate is
measured independently. It's not in a probablistic
fracture mechanics test. The biggest thing that you
have are the loads on the pipe where we know the
pressure loads very well. PR over T really work. The
residual stresses you can measure independently. So
you can measure those independent ingredients, and
MEMBER POWERS: But I never see anybody
put the whole thing together and say, "Okay. Here are
a bunch of data on this thing, and this thing works."
MEMBER SHACK: When you come out with the
probability of large diameter pipe failure of ten to
the minus nine, you're not going to find data.
MEMBER POWERS: Well, give me a small
MR. HACKETT: If I could add, this is Ed
Hackett from the staff, we briefed the Committee, I
guess, numerous times now on the pressurized thermal
shock reevaluation program. I think that's where the
staff and the industry have done the best job of
applying this type of methodology. And in fact that
has been benchmarked to international reference
experiments, and in several cases has done quite well.
In think in the case of Professor
Apostolakis' comment, I'm not aware of any that have
come up that high. We see these failures for vessels,
and, again, thankfully, as Dr. Ford was mentioning,
are in the range of E minus six or less when we're
looking at reactor pressure vessels, different
application than what Larry's talking about here
MEMBER SHACK: But even there, Ed, when
you benchmark that, you benchmark the fracture
mechanics, "Yes, I failed a vessel with a crack so
MR. HACKETT: That's correct.
MEMBER SHACK: Just to say that the
probability of the vessel failure is ten to the minus
eight, you're not going to get a whole lot of
statistics to --
MR. STROSNIDER: This is Jack Strosnider.
I'd like to make a few comments on this too and maybe
to defend the credibility of probablistic fracture
mechanics somewhat. First of all, I think, you know,
when you talk about benchmarking this, as Ed pointed
out, thankfully we don't have an empirical database on
pressure vessel failures or CD control rod drive
mechanism failures, for that matter. So it is rather
difficult to get that sort of benchmarking.
However, I think when you look at the
probablistic fracture mechanics, you can get results
that are reasonable depending upon the conditions that
are being considered. And I think the ten to the
minus 42nd number that was brought up a couple times,
I think you're referring back to some of the PWR work
on vessel inspection. And in fact that number, it
turned out, was the number that was generated when you
assumed design basis conditions were satisfied. In
fact, when you go through the full risk assessment
that was done and what we ultimately ended up with, we
came up with more like ten to the minus six to the ten
to the minus seven numbers when we took into account
beyond design basis events. The conditional -- or the
vessel failure probability, given those events, was
somewhat higher. It certainly wasn't those low
But the other comment I'd make is that the
analysis, methodology exists. We know how to put
models together, we know how to identify random
variables, we know how to model those, how to do Monte
Carlo simulations. There's some challenges looking at
dependence between the variables. But the biggest
challenge, and frankly I would say this is true in all
our PRA modeling, is coming up with the distributions
that represent those random variables.
For example, in this case, where one of
the first things you had to look at was the initiating
frequency, when does a crack initiate one of these?
There's very little data available until we started
getting results from the inspections that were done
and could try to construct a distribution. So the
biggest challenge that we have when we go into this
sort of analysis is being able to define those random
variables, the distributions for them, with some level
of confidence. And usually you have to go out and do
some work, inspections or whatever to get the
information to do that.
CHAIRMAN APOSTOLAKIS: But speaking of
that, though --
MEMBER POWERS: Jack, you make huge
amounts of -- when you do these probablistic fracture
mechanics analysis, you're making huge simplifications
in the way you describe the metal and the way you
describe the crack, things like that. And I guess
what I'm struggling with is how do you know you got
them all. All the physics and all these
approximations really are good ones to make. I mean
some of your approximations are made because you know
how to solve the mathematics.
MR. STROSNIDER: Well, again, I would come
back to if you look at all these models have an
underlying deterministic model associated with them.
If you look at the ability to predict crack growth
rates as a function of stress intensity values, if you
look at the ability to predict failure using either
limit load or linear elastic correction mechanics,
they work pretty well if you have a really well-
controlled situation. And it comes back again to
defining the distributions that are associated with
those in real life. And I agree, that's a challenge.
MEMBER POWERS: Well, every time I look
for things that you predict well, you predict well
those things that have been used to derive the
physics, you know, nice, simple specimens, simple
geometries. Now, you're applying them in really
complicated geometries. There doesn't seem to be any
database that I'm aware of, and I can't say that I've
looked exhaustively, that says, okay, I've done my
laboratory specimens, now I'm going to do this
complicated thing that I don't understand very well
and see if I can get it about right. Is there such a
MR. HACKETT: I guess the one -- this is
Ed Hackett again -- I guess the one I could point out,
Dr. Powers, is the one -- it's a complicated acronym.
They called it fracture assessment of large-scale
international reference experiments; it's the FALSIRE
project. And then there have been follow-on series,
and this is an international collaborative effort,
where they have gone from the small specimen
geometries where things are nice and fairly simple to
predict, to trying to predict what actually happens in
a vessel. The Germans have blown up scale model
vessels, we have at Oak Ridge.
MEMBER POWERS: Yes. Now you're hitting
exactly what I want to see.
MR. HACKETT: And we have in fact --
MEMBER SHACK: Plus an enormous number of
pipes at Battelle.
MR. HACKETT: Absolutely. The most recent
one, thinking of the follow-on activity, the NESC 1
spinning cylinder experiment in the United Kingdom.
In fact, the folks at Oak Ridge, using their
probablistic model, the FAVOR code, which is what
we're using in the PTS Program right now, predicted
the propagation of an embedded flaw in that vessel
almost dead on in terms of initiation and arrest.
CHAIRMAN APOSTOLAKIS: Don't take the
MEMBER POWERS: But if somebody can point
that out -- point it out to me or come present it or
something like that, it adds a lot more credibility to
some of these categories.
MR. HACKETT: Probably in the context of
the PTS project we'll do that.
MEMBER POWERS: That would be great. You
know, if we could take a half an hour and just go
through that, that would be great.
MEMBER FORD: Could I suggest, Larry, that
-- this will be --
CHAIRMAN APOSTOLAKIS: What does it mean
the probability is less than ten to the minus three?
Have you done an uncertainty analysis? How uncertain
is that? How high can the ten to the minus three be?
MR. MATHEWS: I don't have that right now.
CHAIRMAN APOSTOLAKIS: But you will?
MR. MATHEWS: I'm not sure we were going
to do a full-blown uncertainty analysis.
CHAIRMAN APOSTOLAKIS: Well, then what are
you doing? I mean there are so many questions about
all this. To give one number, what does it mean? If
the ten to the minus three can be ten to the minus
one, I don't know what conclusion I can draw from
this. I mean all kinds of doubts have been raised,
and it seems to me doing an uncertainty analysis means
exactly, precisely to address these doubts and
comments. There's something about the five ten to the
minus six that bothers me, okay? That it was five ten
to the minus nine and now it's ten to the minus six,
that's all we learned. I just don't believe that.
And the other thing I want to finish is
that there is a certain pleasure in listening to Mr.
Strosnider defend the probablistic method. Usually
he's a skeptic. Today, he was on the other side.
MEMBER SHACK: It's probablistic fracture
mechanics he's defending.
CHAIRMAN APOSTOLAKIS: I don't care what
you put after probablistic.
It was nice to hear him talk that way.
MEMBER POWERS: But, George, there is a
MEMBER SHACK: One's a science.
MEMBER FORD: If I could just --
CHAIRMAN APOSTOLAKIS: Go ahead, Dr. Ford.
MEMBER FORD: -- move along here. In
defense of the MRP, a lot of this is dependent on
having a reasonable database for crack growth rates
upon which that is dependent. Now I'm told that we're
close to it. The next meeting we will see that
database, and then we will see the follow-on to your
CHAIRMAN APOSTOLAKIS: Great.
MEMBER FORD: -- on that particular
Could I ask you to finish in five minutes,
Larry? I realize that I've now cut you down to your
MR. MATHEWS: I will. In response to the
Davis-Besse issue, we've had lots of interaction with
the staff, but even before the bulletin came out we
conducted, as an MRP, a survey, and it was based on
some -- basically assumptions about what the possible
causes at Davis-Besse were before the root cause or
even the preliminary root cause was out. And there
were three possibilities that we tried to consider in
our survey, and that was leakage from above, leakage
from a crack in a nozzle or a combination of the two.
And then we'll be -- the ongoing Davis-Besse work will
We did that survey, we came up with four
questions basically aimed at how confident are you
that you don't have wastage on your head? And we
received responses from all the PWRs in the country.
We wound up categorizing the responses into four
categories plus another group that didn't quite fit,
and they range from -- you know, category one was they
got the best knowledge, they're darn certain, they've
gone and looked, they don't have any wastage.
Category four, it was more like they were able to do
from a historical view of leakage, et cetera, to feel
confident. And then there was a category, other, that
they had leakage and perhaps had not fully cleaned it
up or there was some other reason they didn't fit into
one of the other categories. And we categorized all
these plants, gave the names of the plants to the
staff, and I believe they've actually used our tables
to help guide a little bit how they're contacting
plants as far as what their intentions are.
This is our ranking of the units that we
put together a while back. If you look at it, the red
triangles are the leaks, and most of those leaks are
to the left of the graph, which is kind of where -- if
the model's worth anything, that's where they'll be.
A couple outliers, we do have one plant that had some
cracks that was a little bit further out. Those
cracks were nowhere near as severe as the cracks at
these plants that have had leaks, so maybe we're
picking up the precursor here. That's something we
have to look at.
All the blue diamonds have done
inspections and haven't had leaks or the open blue
diamonds are doing inspections this spring, yet to do
a few plants in the fall and a few more next year.
We'll have done inspections per bulletin 2001-01.
Here's the table we sent to the staff.
Turns out most of the plants, as far as the wastage on
the head, feel a good degree of confidence that they
don't have any significance wastage on the head. Some
of these plants have even done inspections since then.
Cook 1 I know plans an inspection very soon. Wolf
Creek, I believe, has done an inspection, and I think
Palo Verde just finished their inspection. So most of
these plants are moving into greater degrees of
confidence that they really don't have an issue with
wastage at this point in time.
MEMBER FORD: You should point out that,
Larry, that that's on the basis of your survey, not on
the basis to the replies of 2002-1.
MR. MATHEWS: Absolutely. This was all
put together -- it was probably right at about the
time the bulletin was coming out or maybe shortly
thereafter, but it was based on the response to our
questions, not the responses to the bulletin.
A couple of points about that. All the
plants that are less than ten effective full-power
years on our histogram will have been inspected by the
end of this spring outage season. That includes the
highest ranked 20 units in the country. And they
should have a reasonable assurance that they don't
have any significant corrosion on top of their head
because of those inspections. And of the plants that
were less than 30 EFPY, 34 out of 45 will have
inspected by this spring. We're showing five in the
fall and six in the spring of 2003. There's a little
bit of confusion right now. We're not off more than
one or two plants, I don't believe, but we've got to
settle that out, straighten that out.
This is something that we wanted to say,
that of the 34 leaking nozzles and penetrations that
have been discovered to date, all of them displayed
visible evidence of leakage or corrosion on top of the
head, leakage primarily. A total of 203 nozzles have
been inspected at those -- let's see, is it nine
plants where leaks have been discovered? And NDE has
confirmed through-wall leaks or cracks -- I mean
through-wall defects in all 34 of the nozzles that
showed leakage. NDE did not detect through-wall
defects in any of the others, and there have been,
this says, four plants without evidence of leakage,
and I'm sure by now it's much more than four plants
have inspected the nozzles without any defects found.
MEMBER SHACK: It would interesting on
your chart, you know, where you've got the one with
cracks that you found by NDE, to also see where the
guys that inspected by NDE and found no cracks were on
MR. MATHEWS: Yes. Up until when I put
that together there weren't a lot. There was Cook 2
and maybe a couple of others that had done volumetric,
that didn't have a prior indication of a leak that
they were going and confirming. But we're getting
more and more of the plants now that are doing
volumetric inspections. I think Palo Verde just
completed a volumetric inspections, and I don't even
have them marked as having done that. But we will
update the chart and try and figure out how many
colors we could put on it. But we'll do that.
Recent experience of the -- except for the
Davis-Besse issue, in the other 31 leaking
penetrations, there's no evidence of any significant
corrosion or wastage. There has been a hint at a
couple of other nozzles that there was a little bit
here and there on top of the head or whatever but no
significant evidence. And also on the plants that
have repaired their nozzles that were leaking, most of
those repairs have been performed using the Framatome
repair technology where the nozzle is bored out and
then rewelded up inside the head to the low alloy
steel. And if there were significant wastage there,
it would have been evident. They have to go PT that
surface before they weld to it, and if there's a big
gap, they can't even get it to weld. So out of all
those other nozzles, there hasn't been any significant
wastage like the one big cavity at Davis-Besse.
CHAIRMAN APOSTOLAKIS: So what do I learn
from that? What's the conclusion from that?
MR. MATHEWS: Well, the conclusion is that
something's different about Davis-Besse, the waste,
the big cavity like they had compared to the rest of
the industry. And they're going to talk about it --
CHAIRMAN APOSTOLAKIS: And the rest of the
industry also had wastage there for the number of
years that Davis-Besse had it?
MR. MATHEWS: Well, that may be the key,
and in fact it may be the difference between this one
nozzle and the rest of them is the amount of time that
the nozzle leaked. And Davis-Besse will discuss that
when they get up here. That may in fact be the key is
how long was the leakage allowed to go on without
being detected? But do I know that that's absolutely
the reason? I don't know that, not right now. Okay.
I've only got two more. Ongoing
activities, we're reviewing or have reviewed the
Davis-Besse initial root cause, and we will review the
final root cause for generic implications of that and
use that information to get back into MRPs
recommendations as far as inspection to the plants.
And we're also taking a look back at the Owners' Group
work that was done back in the early '90s. They did
some work on head wastage, and we want to take a look
at that and see does this really change any of that?
CHAIRMAN APOSTOLAKIS: Are you done?
MR. MATHEWS: Yes. I'll quit.
MEMBER FORD: Questions?
CHAIRMAN APOSTOLAKIS: Yes. I mean I'm
amazed that you say you are not planning to do an
uncertainty analysis. Uncertainty analysis is not an
academic exercise. You keep telling me that there are
all these experts that are looking at the huge scatter
of data and so on, and then at the end we're not going
to do an uncertainty analysis.
MR. MATHEWS: Well, we're definitely going
to do all kinds of --
CHAIRMAN APOSTOLAKIS: I'm amazed.
MR. MATHEWS: We're going to do all kinds
of sensitivity studies and look at the various
parameters that go into the model and determine --
CHAIRMAN APOSTOLAKIS: Sensitivity
studies, are you going to do them two at a time, three
at a time, variables, playing all sorts of games to
really gain insights? I mean to vary one variable at
a time doesn't really do much for me.
MR. MATHEWS: Well, the nature of the
Monte Carlo is you do them all at once.
CHAIRMAN APOSTOLAKIS: And that's a
MR. MATHEWS: No. You do -- well, yes.
You put all of the uncertainty of all of the databases
and all of that, it goes in there at one time and you
do a Monte Carlo sample --
CHAIRMAN APOSTOLAKIS: Well, that's not
sensitivity, that's uncertainty analysis.
MR. MATHEWS: Right. But doing the
sensitivity we'll go in and we'll change some of those
parameters and distributions.
CHAIRMAN APOSTOLAKIS: But you said you
were not planning to do that. That's why I'm amazed.
If you were planning to do it, I wouldn't be amazed.
MR. MATHEWS: The term, "uncertainty
analysis," caught me off -- we are going to do
sensitivity studies to look at what the sensitivity of
the analysis is to the various --
CHAIRMAN APOSTOLAKIS: Well, that's a way
of doing it. That's a mechanics review.
MR. MATHEWS: Yes.
MEMBER FORD: Could I, just in terms of
time management, call this one to a close but
recognizing that there are questions along these
lines, and when you come back within the next two
months be prepared to answer them.
MR. MATHEWS: Yes.
MEMBER FORD: Mr. Chairman, am I allowed
to go five, ten minutes over?
CHAIRMAN APOSTOLAKIS: Well, if the Vice
Chairman went over 45 minutes, I don't see why the
members can't go over five minutes.
MEMBER FORD: Okay.
CHAIRMAN APOSTOLAKIS: There's no schedule
today anyway, so keep going.
MEMBER ROSEN: Let's establish some sort
of quantitative mechanism or a curve here, we can
begin to --
MEMBER POWERS: Could I ask a question?
CHAIRMAN APOSTOLAKIS: Yes, sir.
MEMBER POWERS: Something perplexes me a
little bit here. The speakers indicated the time that
the nozzle was allowed to leak, I guess is the word,
and Davis-Besse may have been key. And he said leak
without being detected. Okay? And then we have
inspections of the other things, which presumably have
some probability of detection so that some of those
declared not to have any cracks may in fact have
cracks and may in fact be leaking but we just don't
detect it. What are we doing about that?
MEMBER FORD: A related question to that
is we are assuming that when you see a nozzle, the
popcorn on the top of the nozzle, that is the
sufficient evidence that you've got a crack
underneath. That's something that we've questioned.
Could you have a crack down below the J-weld and not
see the popcorn at the top?
MEMBER POWERS: Well, I think the answer
to that is yes.
MEMBER FORD: Well --
MEMBER SIEBER: It's not through-wall or
plugged. Either way you won't get --
MEMBER FORD: Well, plugged over the
surface. We've asked that question, and that's under
The other question is to whether from
human error you don't see it.
MEMBER SIEBER: Right.
MEMBER FORD: That one has not been
addressed apart from in the Duke presentation on
Oconee the human error was addressed of not seeing it.
But recognize this is still a fairly recent
phenomenon, if you like.
MEMBER POWERS: Well, I mean isn't it the
conclusion that you come out of this as, "Gee, our
methods of inspection are inadequate."
MEMBER FORD: This is something you may
have from the staff, because this might be a policy
CHAIRMAN APOSTOLAKIS: I'm not sure it's
the methods. Ultimately goes to the safety culture.
MEMBER FORD: But that question about --
CHAIRMAN APOSTOLAKIS: It didn't say -- it
doesn't say here that they didn't know because, it's
just they didn't pay attention.
MEMBER FORD: This question of management
of this whole situation by inspectors --
CHAIRMAN APOSTOLAKIS: This gentleman
wants to say something; he's been trying for a while.
MR. MATHEWS: I was just going to say that
the human error -- this is Larry Mathews, I was just
up there. The human error part could be easily
factored into the inspection on a probablistic
fracture mechanics as a probability of detection.
CHAIRMAN APOSTOLAKIS: It could be easily
placed there. Now what value you use is not going to
MR. MATHEWS: Oh, yes. We have to figure
CHAIRMAN APOSTOLAKIS: That's the whole
MEMBER SHACK: Sensitivity studies.
CHAIRMAN APOSTOLAKIS: Oh, you do
sensitivity, excuse me.
MEMBER FORD: The answer to your question
may well come up in the staff's presentation. Could
I ask the representatives from Davis-Besse to come up.
Normally half an hour but make sure you have enough
time to present the stuff on the risk assessment
aspect. John Wood and Ken Byrd from Davis-Besse.
MR. WOOD: Good afternoon. My name is
John Wood. I'm the Vice President of Engineering
Services for First Energy Nuclear Operating Company.
In our agenda today, I'll be discussing the
information that we presented to the subcommittees on
Tuesday. And then at the end of that, we'll have, at
the subcommittees' suggestion, a discussion of the
safety significance assessment that was given to the
staff early this week.
I'd like to just cover a couple points on
background for Davis-Besse in that if you'll note in
the middle there we have 15.8 effective full-power
years at that Unit. Toward the bottom, hot leg
temperature is a little bit hotter than other Babcock
& Wilcox plants at 605 degrees up. That's about three
or four degrees higher based on our core delta T. And
we have 69 nozzles at our Unit. Sixty-one of those
have control rod drive assemblies, seven are spare and
one is used for a head vent that goes to our steam
This is a depiction on the next page of
our reactor pressure vessel head configuration. The
insulation is shown across horizontally here. You'll
note that the dose above the insulation in the area of
the flanges is about one-half a rem per hour. And
beneath the head the dose is approximately three rem
per hour. And those are the fields that we have to
engage as a head sits on the head stand.
In our next picture, or actually two
pictures, what we have shown on this slide is the
reactor vessel head sitting on the head stand in the
left-hand picture with a couple gentlemen working up
above. The picture on the right has been cut open
this outage in order to access at the flange level.
That area is 20-some feet below where those gentlemen
on the left are standing, so typically people would be
working in and around the flanges using 20-foot-long
The next diagram depicts a typical B&W
control rod drive nozzle. It is shown in its
position. There's a shrink fit of about one-half to
one and a half mils that enters into the low alloy
carbon steel. You can see there the shell cladding
and the J-groove weld. Now, when I talk in a little
bit about cracks, the cracks that we have depicted
actually are on the OD of the tube on the wetted side,
or ID, of the main reactor vessel head. And then
through-cracks would go up past the weld into this
annular space here.
We went through details Tuesday with the
subcommittees in regard to the UT examinations that we
performed at Davis-Besse. This picture depicts the
below or underhead UT examination tool. It has been
demonstrated, using EPRI capability, to detect actual
and circumferential flaws. It is delivered with a
robotics system and an automated data acquisition
system. This was used on all 69 nozzles at Davis-
Besse, and then those nozzles produced indications of
flaws were also inspected the top-down UT examination
tool, and that has ten transducers in order to
characterize the flaws.
MEMBER POWERS: Would you give me an idea
how long it took to inspect 69 --
MR. WOOD: That inspection period for
Davis-Besse was approximately 96 hours. And that is
Our UT examination results, and these,
again, were detected with the underhead and then
confirmed top-down, are shown on the next page.
You'll see that there's six nozzles listed here. The
first five had cracks indicated, the first three were
the through-wall cracks. You can see Nozzle 1 had
nine actual tracks, two went through-wall, and nozzle
Number 2 had eight actual cracks, one circumferential
flaw. And that circumferential flaw was approximately
30 degrees, a little bit more than an inch in length,
1.2 inches in length, and was about 50 percent
through-wall for the nozzle. I should mention also
the nozzle is approximately 0.63 inches thick.
Number 3, of course, the one that has the
cavity associated with it, had two through-wall leaks
and there were cracks on Nozzle 5 and 47. Number 46
did not have a crack indicated; however, there's an
investigation with a backwall signal on 46.
CHAIRMAN APOSTOLAKIS: These examinations
were done when?
MR. WOOD: These were done approximately
in early March, the first week in March. Actually,
the last part of February, early March.
CHAIRMAN APOSTOLAKIS: After the problem
MR. WOOD: That's -- no. This led to the
finding of the problem.
CHAIRMAN APOSTOLAKIS: Oh, this led to the
MR. WOOD: That's correct. This was the
100 percent UT examination of the nozzles at Davis-
Besse was done in conjunction with our answering of
2001-01 in our extension from the end of the year to
CHAIRMAN APOSTOLAKIS: But were
examinations like this done routinely and on a
MR. WOOD: No. At the time, we had the
most extensive examination of the head using
CHAIRMAN APOSTOLAKIS: So that was the
first time you did this?
MR. WOOD: That's correct.
MEMBER POWERS: These were surprises to
MR. WOOD: It was not entirely surprising
that we had axial cracking. Based upon the
information of 2001 and the information that we were
getting from the industry, we expected to find some
cracking. We did not expect to find through-wall
necessarily and certainly didn't expect to find the
cavity that we found on Nozzle 3.
MEMBER POWERS: I'm sure that was a -- but
I'm just asking about the --
MR. WOOD: Right. In fact, our plans
included fixing up to four nozzles in our base plan
for this refueling outage.
This diagram lays out the nozzles that
were found with cracks. Those are indicated in both
the red and the green. I will note that the five
nozzles in the center of the head are all from the
same heat, and I'll talk about that later. Those are
the only five nozzles from that heat at our Plant.
You can see Nozzle 2, which had the circumferential
crack, was located in this quadrant, and there was a
very small amount of wastage in this area of Nozzle 2
that I'll talk about in a little bit as well.
I guess that's the next slide. As we were
going through the repair process for the nozzles, we
did note, as it's shown here, as we machined up, as
Larry discussed the repair process used by Framatome,
you machine up and then the intent is then to weld
onto the carbon steel. We did find a small cavity in
that area. Its dimensions are approximated on this
sketch. We have since removed that nozzle for further
clarification. It is essentially as depicted here.
It goes about a quarter to three-eighths maximum
depth, as indicated in the reactor vessel head.
MEMBER POWERS: You mentioned that the
afflicted nozzles came from a particular heat, and the
reason you know that is because of your Appendix B
MR. WOOD: That is part of the MRP process
that we have been working on and also the response of
2001 and the Babcock & Wilcox Owner Group efforts,
knowing what the heat numbers are for the various
nozzles in all the plants.
The primary reason we're here today is the
Nozzle 3 cavity. This is depicted in this drawing, or
this picture. I will remind you that this circular
hole where the nozzle was located is approximately
four inches across. You can see there is some wastage
on the right-hand side at the surface level, and this
is the stainless steel cladding evident at this
location. This is our number one nozzle, so this
would be the dead center of the head, and flow
downhill in that direction.
The next page is more of a display of some
of the numbers that we have determined using various
tooling. It does not show the surface wastage that is
off to the right. You can see there's a difference in
color here. This is to represent a nose or an
overhang, and there is additional erosion at -- or
corrosion that goes on underneath that zone.
You'll also notice that there is a
proposed 13-inch circular cut line indicated here. In
order to better capture this area, we're going to cut
that out in one piece using an abrasive water jet, and
that will then be retained for further evaluation as
we go forward. That abrasive water jet will also
leave us a very smooth finish that we can then prepare
a final fit up of the forged disc that we discussed in
concept yesterday with the NRC staff. The exact
location of that cutout will be determined to optimize
all things involved.
After we found the cavity area around
Nozzle 3, we chartered a root cause initial
investigation team using First Energy personnel to
lead the effort. Those individuals were not from the
Davis-Besse staff. We did include members from the
Davis-Besse staff on the team, as well as augmented it
with industry experts from Framatome, Dominion
Engineering and EPRI, as listed here.
The team came up with a probable timeline
using best engineering judgment in looking at the
evidence that we had from the period of time in
question. What you see here is a summary of that
probable timeline. It shows that the crack
potentially propagated through-wall in the '94 to '96
time frame, and thus went basically unaddressed for a
period of two to three operating cycles.
CHAIRMAN APOSTOLAKIS: Now, that's where
I have a question. What does that mean? Were you
aware that there were cracks?
MR. WOOD: No, we were not aware that
Davis-Besse had cracks at that time.
CHAIRMAN APOSTOLAKIS: So when you say
unaddressed, what do you mean by unaddressed?
MR. WOOD: Unaddressed means that the leak
was allowed to be active without awareness for that
period of time.
CHAIRMAN APOSTOLAKIS: Did you have any
indications there was a leak?
MR. WOOD: In a retrogressive look,
certainly there were missed opportunities, and I
believe the staff will relate those as well. And as
I go through some of the contributing causes, there
were reasons that the staff used to perhaps not center
on those clues that a leak was occurring on the nozzle
Now, I'll talk --
CHAIRMAN APOSTOLAKIS: All the rules and
regulations were followed. You were not in violation
MR. WOOD: I don't think I'm in a position
at this point to say that there was nothing that was
violated. Certainly, there were people with very good
intentions that were doing the things they thought
were right. As we look back, things did not go
according to the desires and the expectations that
should have been in place.
CHAIRMAN APOSTOLAKIS: And that was, in
your opinion, more a matter of judgment, which perhaps
was poor in this case?
MR. WOOD: Certainly, poor judgment.
CHAIRMAN APOSTOLAKIS: Okay.
MEMBER LEITCH: What gives rise to the
probability that the crack initiated about three years
before it went through-wall? Is that based on some
crack growth rate?
MR. WOOD: That's based on the same crack
growth rate that you would have heard from the MRP
individual -- Larry.
MEMBER LEITCH: Then I guess one could
assume that since we see no crack in Nozzle, what is
MR. WOOD: Number four.
MEMBER LEITCH: That we have a certain
degree of confidence that it would not go through-wall
within one cycle of operation.
MR. WOOD: That's correct. But that's
based on probabilities and not certainty.
MEMBER LEITCH: Yes. Because Nozzle 4
seems like it's crying out to crack, right? I mean
MR. WOOD: Well, and there have been
numerous people, including myself, who have asked over
and over and been told again and again that Number 4
does not have cracks.
MEMBER SIEBER: Yet.
MR. WOOD: Yet. And that's an important
yet, and that's true with all the nozzles that are in
MEMBER LEITCH: Okay. Thank you.
MR. WOOD: Now, the probable cause here is
really of the failure mechanism, that being the
cracking. And since we were in the repair process
prior to finding the cavity -- as I have mentioned
earlier, the repair effort requires us to grind up the
nozzle from below to above the J-groove weld, and so
the cracks themselves were taken out as a result of
doing that. So that's why it's listed as probable
cause because we don't have material to identify it as
a factual root cause. But every indication --
MEMBER SHACK: Nobody tried to map the
cracks as they were grinding them either.
MR. WOOD: That's correct. We did have UT
data that we showed the subcommittees Tuesday that
mapped them out in the general sense but not to
progress and grind in PT, as an example.
With what we know that is happening in the
industry on Alloy 600 and the control rod drive nozzle
issue, we feel confident that it is primary water
stress corrosion cracking that resulted in the crack
initiating propagation and then allowed leakage to the
reactor vessel low-alloy steel head.
MEMBER FORD: If I could ask a question.
It's fairly obvious that the initiating event was
primary water stress corrosion cracking rising to a
liquid of some sort in the annulus. But the key
question is why did that environment give erosion or
corrosion of the low-alloy steel in your condition but
did not in many of the others, like Oconee? And
that's the root cause question that needs to be
MR. WOOD: Correct. And the root cause of
the cavity being there is this next page.
MEMBER FORD: Okay.
MR. WOOD: And that is our Boric Acid
Corrosion Control and In-Service Inspection programs
did not allow us to see that leakage at an earlier
time. Now, this is, again, looking backwards at the
data that we had at hand, but we feel that the leak
had existed through-wall for two to perhaps three
operating cycles and thus did not allow us to identify
CHAIRMAN APOSTOLAKIS: I'm confused by the
words on this slide.
MR. WOOD: Okay.
CHAIRMAN APOSTOLAKIS: "The Boric Acid
Corrosion Control and In-Service Inspection programs
and the program implementation resulted in the Plant
not identifying the through-wall crack." What does
that mean? That the program resulted in you not
MEMBER SHACK: The failure to implement
the Boric Acid Control Program.
MR. WOOD: Right.
CHAIRMAN APOSTOLAKIS: Oh.
MR. WOOD: The Program neither robust
enough nor was it implemented sufficiently in its form
to detect the crack. So had it been, let's say, more
robust and more rigorous applications, that would have
been one approach. Even apart from that, had it just
been implemented appropriately or properly, it would
have been the other case.
CHAIRMAN APOSTOLAKIS: So you are blaming
both the Program and the implementation, at this point
MR. WOOD: That's correct.
MEMBER SIEBER: Now, I have a question.
You, actually, when you asked for your extension from
the bulletin schedule for inspections, you relied on
videotapes, as I understood it, to say that leakage
was not there?
MR. WOOD: Yes. And what I think is being
asked, as we went through the effort on 2001-01 to
extend our outage from the end of the year, as was
requested from the staff, until the time of February
16, we did an evaluation of the information we had in
hand and knowing that there was some boric acid in the
vicinity, the thought of the staff was that that boric
acid had come down from the flanges from above and the
mindset, for whatever reason, was focused on circ
cracking and not on the potential wastage issue that
we eventually found.
MEMBER SIEBER: Did anybody from the NRC
staff see those videotapes before the extension was
MR. WOOD: I cannot answer that question
MR. BATEMAN: Yes, I can answer that
question. We spent about three hours looking at
videotapes from the 1996 inspection, the 1998
inspection and the 2000 inspection. And there were
substantial amounts of boric acid on the head at that
MEMBER SIEBER: Did you, like the
Licensee, assume that it came from the joint in the
housing up above?
MR. BATEMAN: We did not have that
discussion at that point in time.
MEMBER SIEBER: Okay. Thank you.
MR. BATEMAN: By the way, Bill Bateman
from the staff.
CHAIRMAN APOSTOLAKIS: So let me
understand the second bullet here, "Plant returning to
power with boron on the RPD head after outages." So
Plant personnel knew that there was boron on the RPD
head after outage?
MR. WOOD: There were individuals at the
Plant that knew there was boron on that head, that's
MEMBER SIEBER: And, apparently, the staff
did too prior to granting the extension.
CHAIRMAN APOSTOLAKIS: They thought it was
coming from the flanges.
MR. BATEMAN: This is Bill Bateman from
the staff again. I want to make it clear that the
videos that we looked at were videos inside the shroud
area around the mechanisms, not outside where the weep
holes -- I think you saw the picture yesterday --
where the weep holes actually -- it dripped down from
the holes onto the -- near the bolt circle on the
head. We did not look at -- we did not see those
particular pictures. We were inside that shrouded
area of the videos that we looked at.
MEMBER SIEBER: This was through those
MR. BATEMAN: Right.
MEMBER SIEBER: Camera on a stick?
MR. BATEMAN: Right. Yes. Those are the
videos we looked at.
MEMBER SIEBER: Okay.
CHAIRMAN APOSTOLAKIS: Okay. You said,
Jack, that they knew there was boron there and they
assumed it came from the flanges. So what, didn't
they still need to clean it up? I mean whether you
clean it up depends on where it's coming from?
MEMBER SIEBER: I would have thought so at
the time, but I'm not sure that everybody makes their
-- up until today, makes their reactor vessel head
squeaky clean each time they do an inspection.
CHAIRMAN APOSTOLAKIS: But there's a
difference between each time and not doing three or
MEMBER SIEBER: That's true.
MEMBER POWERS: By the way, George, I just
remind you of a point that was made at the beginning
of the presentation. This is -- doing things on the
vessel head that aren't absolutely required is a
highly costly thing, not only in time but because of
the radiation dose that you incur to your workers. So
if you don't think you have to do it, you're probably
not going to do it.
CHAIRMAN APOSTOLAKIS: So the question is
when do you decide that you have to do it?
MEMBER POWERS: That's right.
CHAIRMAN APOSTOLAKIS: Now, maybe you have
already explained it, what is 12RFO?
MR. WOOD: Twelfth refueling outage.
We're currently in our 13th refueling outage.
CHAIRMAN APOSTOLAKIS: Okay. Thank you.
MR. WOOD: Okay. And as we have just been
discussing, the environmental conditions which
contribute to this is the cramped conditions of the
design. And by that I mean there's about two inches
of clearance between the top of the head and the
insulation. As was mentioned, we have 18 weep holes
near the bottom that provide us some access. And we,
therefore, did not take appropriate compensatory
measures as a result of these cramped conditions to
allow ourselves to find that leakage.
Another contributing cause was the fact
that in the late '80s, early '90s, there was much
leakage of the CRDM and flanges above the insulation,
which allowed some boron to pass through to the head
and participated in the mindset of the staff at the
Now, I did mention the fact that we had a
material heat that was unique for five nozzles, four
of which had cracking, three of which had through-wall
cracking. And all three of those nozzles that had
through-wall were from this heat listed. We're aware
that that heat is used at two other B&W plants. One
plant has all but one of their nozzles from that heat;
another B&W plant has one nozzle from that heat. The
one that has the majority has been well-inspected and
has thus contributed to a database that suggests that
20 percent of this particular heat of nozzles has
cracked or has had evidence of cracking thus far.
We spent some time Tuesday talking about
crack length versus leakage. I don't intend to go
into a long conversation on that, but I did want to
mention that our unidentified leak rate at the Plant
during the period of time in question was
approximately 0.1 to 0.2 gallons per minute. So that
is well below the tech spec limit of one gallon per
minute. And you can see the fact that the longer
crack lengths have more damaging corrosion resulting
from them. Whether that's just evidence that it is
interesting at this point or it is matter of fact, we
don't know for certain.
MEMBER POWERS: Could you give me some
idea of what the width of the cracks is?
MR. WOOD: The width of the crack, I don't
have that information. I don't know if anyone from
the staff does in the back there.
MEMBER POWERS: Real tiny, as big as my
MR. WOOD: Very tiny, and we're talking in
the orders of a thousandths of a gallon per minute up
to the 0.2, 0.8 region. And so --
MEMBER POWERS: That's what I was looking
MR. WOOD: Okay. As a result of our
meeting Tuesday and getting together with --
CHAIRMAN APOSTOLAKIS: Before we go on, if
I were to take with me the top two causes why this
situation developed, what are they? Something must
have gone wrong someplace, so what are the top two
causes, so I remember? I read a lot of stuff and they
say a lot of things, the timelines and this and that,
but if you ask me what was the number one and number
two contributing causes, I have difficulty figuring
those out. So can you summarize them for us?
MR. WOOD: Well, I think number one was
the Boric Acid Control Program and the application of
CHAIRMAN APOSTOLAKIS: Okay.
MR. WOOD: I guess almost everything else
pales by comparison.
CHAIRMAN APOSTOLAKIS: Okay.
MEMBER KRESS: I would have listed the
potential for having a bad heat. There are cracks
MR. WOOD: Granted however in this
business we're accustomed to dealing with things that
may be first of a kind or second of a kind or
whatever. So we wouldn't want to use the fact that we
had a bad heat as the indicator of the cavities, the
indicator of the crack.
MEMBER KRESS: You still have to deal with
MR. WOOD: Correct.
MEMBER SIEBER: There may be an issue of
standards involved too on the part of the inspection
personnel and decision makers.
MR. WOOD: Yes. Those standards of course
will go to the very top. That's where standards come
CHAIRMAN APOSTOLAKIS: I'm sorry. What
standards are these? I missed it.
MEMBER SIEBER: The kind of standards one
would expect from a professional organization that
operates a nuclear power plant.
CHAIRMAN APOSTOLAKIS: Isn't that what
some other people call safety culture?
MEMBER SIEBER: That's a piece of safety
CHAIRMAN APOSTOLAKIS: Yes. It can be all
MEMBER SIEBER: Questioning added to high
CHAIRMAN APOSTOLAKIS: Yes. Okay.
MEMBER SIEBER: Vigilance.
MEMBER ROSEN: The application of the
corrective action systems.
CHAIRMAN APOSTOLAKIS: Okay. Thank you.
MR. WOOD: Okay. Then as a result of our
meeting on Tuesday, Peter Ford asked that we would
include safety significant assessment. So we have Ken
Byrd who will present that.
MR. BYRD: Okay. My presentation will be
a very brief summary of the results of a safety
significance assessment that was provided to the staff
earlier this week. For this assessment, we considered
a range of breaks from very small to the size
described on the top of this page 23.
So that for the maximum size, we assumed
the failure of the exposed cladding area which is
approximately 25 square inches. In addition, we
assumed that the whole was 50 percent larger than the
exposed cladding area for about 38 square inches.
We also assumed that CRDM Number 3 would
eject. So our total area was approximately 50 square
inches or 0.35 square feet. We're looking at a range
from very small up to 0.35 square feet. For our
analysis, we evaluated three critical functions.
MEMBER ROSEN: Now before you get off that
in terms of assumptions. You've obviously made the
assumption although it's not shown here that nothing
else was damaged. There was no additional damage.
MR. BYRD: No, sir. I'm going to talk
about that next when I look at these next three
MEMBER ROSEN: Okay.
MR. BYRD: I'll get to that. We looked at
three critical functions when we did this analysis. We
looked at the ability to have core cooling, to
maintain shut down margin, and finally containment
We do not have a Davis-Besse ACE, an
analysis for a LOCA at this specific location.
However our LOCA analysis covers a spectrum of LOCAs
from 0.01 square feet up to 14.2 square feet.
Setting aside at the moment collateral
damage, this particular LOCA is equivalent to a hot
leg LOCA with respect to core cooling. In that
respect we would get injection flow going through the
core for both core cooling and for boron precipitation
control. Therefore with respect to core cooling, we
were bounded by our existing LOCA analysis.
Let's go on to my second bullet here which
relates to shut down margin. I think this is where we
get into the concern about the issue of collateral
damage that might occur to adjacent control rod drive
mechanisms. Consequently we had Framatome ANP do an
evaluation of the potential for damage to adjacent
control rod drive mechanisms.
The Framatome Analysis looked at several
different mechanisms. They looked at jet loadings.
They looked at pressure loadings. They looked at
loose debris which might mechanically jam an adjacent
control rod drive mechanism.
The results of their analysis was that it
was unlikely that an adjacent control rod drive
mechanism would be affected. Not withstanding that
result, we went ahead and had them do a further
analysis to look at the impact of all of the control
rod drive mechanisms. We actually looked at five
control rod drive mechanisms surrounding the affected
Failing to insert is a result of
collateral damage. In addition to that, we added one
additional control rod which would be a random control
rod failing to insert with the highest shut down
margin for that control rod. With those six control
rods failing to insert as a result of this accident,
we were able to have both immediate and long term shut
MEMBER ROSEN: Is that for the conditions
that the Davis-Besse found themselves in at the end of
the day on February 16 or whenever it was that you
shut down? Was that a more general conclusion for any
time during the cycle?
MEMBER FORD: Before you answer, Ken,
could you just let the Committee know if the staff
have not reviewed this analysis yet?
MR. BYRD: No.
MEMBER ROSEN: So let me repeat my
question. Is that result that you had plenty of shut
down margin even with those six rods not reinserting?
Was that a general result for if this had happened at
any time during the cycle or a specific result that
applies only to that day, the day you shut down?
MR. BYRD: It was really intended to apply
only to that day. But the analysis was done using the
beginning of life for cycle 14 which was actually a
more conservative time period.
MEMBER ROSEN: Okay.
MEMBER SIEBER: But is the break size you
had, the larger the break the better able you would be
to get reactivity reduction because of the insertion
of highly borated water?
MR. BYRD: Yes, sir. That would be true.
CHAIRMAN APOSTOLAKIS: The rod ejection
effect is instantaneous, but you're at full power. So
you have some full power conditions.
MEMBER SIEBER: Right.
MR. BYRD: Right.
CHAIRMAN APOSTOLAKIS: So that reduces the
concern with the rod ejection.
MR. BYRD: Okay. If I could go on to the
third condition that we considered. We also
considered containment integrity. The issues we were
concerned with here were two issues.
One was the control rod ejection, actually
impacting on our containment. The other issue would
be the mass and energy release from the particular
With respect to the first of these issues
at Davis-Besse, we have missile shields above the
control rod drive mechanisms which would prevent an
ejected control rod from impacting a containment.
With respect to the second issue, mass and energy
release, this particular LOCA is bounded by much
larger LOCAs which have been analyzed. So we did not
see any significant issues with respect to containment
MEMBER POWERS: Let me ask a question that
you may not have the answer to. If you have blow out
in that particular location, do you put an unusually
large amount of mass into your sumps that could clog
some pumps and things like that?
MR. WOOD: No. That area would not be
directly driven towards the sumps. That would be
within the refueling canal. Then you saw the service
structure arrangement around it. So there's not a lot
of direct accessibility out of that into the sump area
which is quite a ways away from that.
MEMBER SIEBER: The refueling canal is
empty during operation.
MR. WOOD: That's correct.
MEMBER SIEBER: You use a diaphragm
between the vessel flange and the edge of the canal.
MR. WOOD: No. There would be an opening
in that area.
MEMBER SIEBER: During operation.
MR. WOOD: During operation.
MEMBER SIEBER: That's the flow path to
MR. WOOD: Right.
MEMBER SIEBER: Okay. So there is a
MR. WOOD: The sump itself is up on a
different level beneath the head. But would initially
MEMBER SIEBER: Okay.
MEMBER POWERS: So it's a fairly contorted
path that something would have to follow to get to
MR. WOOD: That's correct.
MEMBER SIEBER: It would have to go
MEMBER POWERS: It wouldn't be so uphill.
MEMBER ROSEN: The insulation that's above
the head in that region is reflective insulation.
There's no silicacious insulation.
MR. WOOD: That's correct.
MEMBER ROSEN: That's all metal in pipe
MR. WOOD: Right.
MEMBER POWERS: That didn't help you much.
MEMBER KRESS: It's gets really pushed
around a lot.
MEMBER ROSEN: Well it does actually.
MR. WOOD: However all that insulation
would have been inside of the service structure.
MEMBER ROSEN: The three GSI-199 is the
most damaging kind of material. It is the kind of
material that can plug the screens. Typically it's
the silicacious sand-like material that --
MEMBER POWERS: No.
MEMBER ROSEN: Plans toxin fibrous
material and end up building the building up across
MEMBER POWERS: Fibrous material is of
course very bad. But we've seen experiments showing
that you can shred this stuff up. That shredded
material is not too good either.
MEMBER ROSEN: It may be. But I think if
you read GSI-199, the most recent staff stuff that
came out of, which lab? I'm trying to remember which
lab. I think that report indicates that the worst
material comes out of Los Alamos and the University of
New Mexico. So I'm reasonably familiar with it.
MEMBER FORD: If I could interrupt, could
we just get this one through? Again I'm looking at
MR. BYRD: Okay. Going on to the next
page. As a further effort to address the safety
significance of this condition, we had a stress
analysis of the as-found head condition performed.
This stress analysis is a three-dimensional finite
element, stress analysis of the wasted -- and the
reactor pressure vessel head.
We had a failure criterion set at the
maximum strain of 11 percent through the thickness of
the clad. We had the results verified by an
independent analysis. We had this both performed by
Framatome ANP and Structural Integrity Associates.
The results were that the degraded cavity
would maintain its integrity in excess of twice the
transient loads. The results for the two analyses
were fairly consistent.
MEMBER SHACK: What's the rational for the
MR. BYRD: This particular analysis is an
input to my safety assessment. I think I have an
expert here from Framatome who could probably address
that better than I can.
CHAIRMAN APOSTOLAKIS: Please identify
MR. FYFITCH: I'm Steve Fyfitch from
Framatome. The rational here is that's actually a
conservative value that they used for the analysis.
The 11 percent comes from an Oak Ridge report that we
have access to that looks at 308 in stainless steel
The 11 percent is where necking starts to
occur in the tensile test. We assumed that 11 percent
was the failure strain. So it's in fact a very
conservative because once the uniform elongation
starts to disappear, it actually goes out and total
elongation about 30 percent.
MR. HACKETT: Bill, this is Ed Hackett
from the staff. A follow up to that would be we're
doing confirmatory analyses too as you know for the
criterion failure strain. That number probably needs
to be adjusted, Vom Mises or Treca for the multi-axial
state of stress that would exist in the head.
So probably the real number should be less
than 11 percent. I don't know what the number should
be. As Steve pointed out, that number is from uni-
axial tension test. So what you have is at least a
bi-axial state of stress in the head. That will come
down somewhat. We're looking into that right now.
MR. HERMANN: Ed, I think in the models
the tensile stresses that were taken were compared to
Vom Mises output in the models.
MR. HACKETT: The 11 percent already
reflects a Vom Mises or Treca adjustment.
MR. HERMANN: Yes. It's just a comparison
of what came out of the tensile stress versus that's
not what was in the model. It was just a comparison
of that. A unilateral strains.
MR. HACKETT: Okay. Thanks.
MEMBER FORD: For the Recorder, that was
MR. HERMANN: Bob Herman from Structural
MR. BYRD: Now going to my last page. The
results of this analysis on the previous page
indicated that the expected failure pressure was well
in excess of the pressure for any postulated
transients. It's also well in excess of the pressure
for any transients that have actually been experienced
However to estimate a risk of the as-found
condition, we looked at the probability of a failure
occurring at less than this estimated pressure based
on our stress analysis. The results of this indicated
that there are core damage frequency we estimated to
be in the range of 1 times 10 to the minus 5th per
year. The larger the release frequency was
approximately of 1 times 10 to the minus 8th per year.
Our public health risk was approximately 0.56 person
rem per year.
CHAIRMAN APOSTOLAKIS: Are these Deltas
given these conditions?
MR. BYRD: Yes, sir. These are Deltas.
CHAIRMAN APOSTOLAKIS: So what is your
MR. BYRD: My baseline currently for
internal events is 1.2 times 10 to the minus 5th per
MEMBER ROSEN: Ten to the minus what?
MR. BYRD: Fifth per year.
CHAIRMAN APOSTOLAKIS: So your doubling.
MR. BYRD: Approximately doubling our
internal event baseline.
MEMBER SHACK: Now as I'm corroding away
at two inches a year, how many weeks do I have to wait
until this thing goes?
MR. BYRD: We have that analysis currently
in progress. We're expecting an answer to that
relatively soon. We have an analysis that will give
us the size at which point we would have a failure at
a normal pressure. As far as how long it would take
to get to it, I think that's a little bit more
CHAIRMAN APOSTOLAKIS: So this is given
that I have the amount of degradation that was
observed, the core damage frequency would be 10 to the
MEMBER KRESS: The maximum it could be is
conditional. What's the conditional core damage
CHAIRMAN APOSTOLAKIS: Well it is
MEMBER KRESS: Given that you have the
CHAIRMAN APOSTOLAKIS: Oh, the hole.
MR. BYRD: If we had a LOCA?
MEMBER KRESS: Yes.
MR. BYRD: That would be a conditional
core damage probability. In the calculation of this
core damage frequency, we evaluated the conditional
core damage probability from a range all the way to
very small up to the 0.36. The largest was at about
0.1 square feet. That was 2.9 times 10 to the minus
CHAIRMAN APOSTOLAKIS: You said 0.36?
MR. BYRD: The hole size with the maximum
core damage probability.
CHAIRMAN APOSTOLAKIS: So you estimated
the probability of this LOCA to be the order of 7 10
to the minus 3.
MR. BYRD: I'm sorry.
CHAIRMAN APOSTOLAKIS: What's the
frequency of this LOCA?
MR. BYRD: I guess it might be easiest if
I could just take a minute here and walk through the
process because I think I have a few questions.
Essentially what we did was we understood that at the
pressure we calculated we weren't supposed to get a
failure. So we looked at ways that this would fail at
There's a couple of things that came to
our mind. One was a sizemic event. The other being
overpressure transients that didn't actually get to
With respect to the sizemic event, we have
recently completed a sizemic PRA. We looked at that.
Based on the results of that a sizemic event of
sufficient magnitude to cause this damage in Northwest
Ohio the frequency is very small. So that was a very
The other thing that we looked at though
was overpressure transients. We recognized that this
number that we had from the stress analyses is a
calculated number. It's dependent on a number of
things such as the analysis, the actual condition of
the clad, and the material strength.
So we employed a process that is outlined
in NUREG 2300, the PRA Procedures Guide and NUREG 5603
and 5604. This is a process we've used for doing our
interfacing system LOCA type of evaluations in our
PRA. It's also similar to what we use in our sizemic
analysis and in our external event tornado analysis.
To do that you actually assume a median
failure capacity which we took to be the number we got
from the stress analysis. Then we had to develop a
logarithmic standard deviation. To do that we went to
the new rigs and looked at the various different
tabulated standard deviations for materials, for
temperatures and different kinds of configurations.
We took one that basically bounded the
results we've seen in there. This is a way of
approximating the probability that the failure might
occur earlier. Based on that we were able to
calculate the probabilities of failures at pressures
of about 5600. We were able to come up with
probabilities of 3 times 10 to the minus 3rd to 7
times 10 to the minus 3rd depending on the pressure.
So that gave us a probability of failure
at a given pressure. Then we had to determine since
we weren't trying to calculate a frequency, we had to
calculate a frequency which over pressure transients
would occur at the plant. To do that we went back
through our plant history all the way back to 1979 and
looked at all of our overpressure transients.
We actually calculated frequencies for
various different categories in terms of the extent to
which they overpressurized the plant. Then we were
able to calculate a frequency of that we would get a
transient that would actually cause a LOCA. That
number was in the order of 4 times 10 to the minus 3rd
which is about to give you a feeling two orders of
magnitude higher than our normal medium LOCA number.
CHAIRMAN APOSTOLAKIS: Does the number of
10 to the minus 5 include as part of the conditions
the possibility of the six rods not going in?
MR. BYRD: Based on our deterministic
analysis, we had evaluated that even if the six rods
did not go in, we would have sufficient shut down
margins. So we did not specifically include that.
CHAIRMAN APOSTOLAKIS: All right.
MEMBER FORD: Okay. If I could jump in
here. I'm watching the time here, George, unless you
want to extend into your other time.
CHAIRMAN APOSTOLAKIS: No. That's unfair.
I shouldn't extend it if I want to ask questions
MEMBER FORD: That's right.
CHAIRMAN APOSTOLAKIS: Let's move on.
MEMBER FORD: Thank you very much indeed.
I appreciate your comments. Let's call on Jack Grobe.
You're now going to hear two presentations by the
CHAIRMAN APOSTOLAKIS: Should we take a
break? We've been going forever. Do the members want
to take a short break?
MEMBER KRESS: Yes.
MEMBER SIEBER: That would be good.
CHAIRMAN APOSTOLAKIS: Okay. We're
recessing until 3:50 p.m. Off the record.
(Whereupon, the foregoing matter went off
the record at 3:40 p.m. and went back on
the record at 3:50 p.m.)
CHAIRMAN APOSTOLAKIS: On the record.
Back in session.
MR. GROBE: My name is Jack Grobe. As was
mentioned, there's three presentations this afternoon
from the staff. I'm going to present the results of
a recent inspection that was completed about a week
ago. We exited on that inspection last Friday. Allen
Hiser will then present the status of Bulletin 2001-
01. Ken Karwoski will present the current status of
the bulletin responses for Bulletin 2002-01.
Being from Region III, I'm the Director of
Reactive Safety. I don't get to see you folks very
often. I appreciate the opportunity to be here.
Quite frankly I'm quite embarrassed to be here. As I
go through this you'll see why.
This wastage occurred over a period of
years. Our staff did not identify it. Certainly the
Davis-Besse caused it and had many opportunities to
identify it. We'll get into that a little bit.
I was going to cover three topics. The
first and third I think we've addressed pretty
extensively with the staff's presentation from Davis-
Besse. There are just a couple of issues that I'll
touch on in that area.
As was mentioned there were five cracked
nozzles, three were through wall. I'm going to get
into a little bit of the description of the cavity,
just some of the information that I think was
important but not presented yet. You've already
understood what happened at nozzle 2.
This is just a little bit different
rendering. This is an artist's rendering of the
cavity. They spoke of the nose. There was
substantial undercut in the cavity.
In addition to that, there were some UT
measurements were taken from beneath the cladding.
There was an unusual result. They were taken on one
inch centers. There were indications that for an
extended distance outside of the visible cavity on the
order of maybe two and sometimes more inches, there
appeared to be a gap on the other side of the
It's not clear what that is. When the
licensee cuts out the cavity, they'll be able to
investigate that more clearly. It's not clear whether
that's a reflection. Whether it's actually a
separation, it's just not clear.
If you look at the physical character of
the cavity, there's an uneven area quite a bit bigger
than the cavity that appears to be as a minimum de-
bonded between the stainless steel and the --
VICE CHAIR BONACA: Could you show us the
location there? Is it possible to see the location?
MR. GROBE: I don't have a slide that
shows the layout of that. A plan view as it were. I
don't have that. I apologize.
MR. HISER: Yes. I guess just to try to
provide a little bit of an answer this is Allen Hiser
from NRR. It's around nozzle 11. It's just not clear
at this point how far --
VICE CHAIR BONACA: Okay. Down there on
MR. GROBE: Well, it actually goes
laterally across the cavity as well as downhill. It
appears to go the whole way to nozzle 11 and maybe
somewhat around nozzle 11. Like I said it's at least
in some cases two or more inches beyond the visible
aspect of the cavity.
VICE CHAIR BONACA: The reason I'm asking
the question is that in the repair, they've already
defined the size of the plug.
MR. GROBE: Right.
VICE CHAIR BONACA: Does that mean the
plug may have to be larger than what they are planning
MR. GROBE: Or there may be repairs
necessary. One of the first things that they are
going to do after they cut out the 13 inch diameter,
their current plan, is they're going to do
diapenetrate testing of the surface to try to identify
whether or not there's additional damage to that
VICE CHAIR BONACA: Okay. I understand.
MR. GROBE: This is a view of the cavity.
I think you can see in the lower section of the cavity
there's a shiny area. That's where it was machined
prior to the penetration to pitching as it were. The
tube has been removed. You can see the walls of the
cavity are fairly smooth. They slope in.
You saw this drawing in the last
presentation. There's nothing more to report on this
except a characterization of the wastage area is a
little bit incorrect. It comes out a little bit more
now that we have impressions in the lower area. Then
it tails off to be a little bit thinner.
So it appears that there may be more than
one mechanism. It may not just be corrosion. There
may be some other things as well.
I want to get into missed opportunities.
I'm going to cover three areas. They are the
containment air coolers, the containment radiation
monitor filters and also the Boric Acid Corrosion
Dr. Apostolakis, you asked what are the
two main causes. The easy cause is to blame the Boric
Acid Corrosion Program implementation. The entire
operation of these facilities depends on human beings
whether it's people doing designs, operators of the
control panels, human beings make mistakes.
Implementation of this program was not
well implemented. That's by engineers. But the
results of the program implementation were known to a
number of people as well as a number of other
I believe that the most important cause
here is a complete failure of the Corrective Action
Program. You'll see that as I go through my
Just a little bit of system knowledge that
you may not have that's important to this. There's a
ventilation that the system intakes as suction on this
volume here. Discharge is near the top of containment
above the D-rings.
The area below the insulation is connected
to the area above the insulation through small gaps
around the nozzles and things of that nature. So
there is a communication of the ventilation system
between these two areas.
There are a series of almost 20 five by
seven inch what are called "mouse holes" or "weep
holes" that are right down here at the edge of the
vessel. (Indicating.) So they are for air coming in
through that direction. It's critical to understand
that the discharge from these areas at the top of
containment just to see what happened in the
containment air coolers and radiation monitors.
MEMBER SIEBER: The way out of that bottom
plate and the mirror insulation is such that since the
air flow is up, they don't have conoseals, but in
those joints the leakage is probably not going to go
down. Some of it does.
MR. GROBE: The leakage will likely be
MEMBER SIEBER: That's right.
MR. GROBE: It will be steaming
horizontally. It will spray against other surfaced
and evaporate. Then the vapor will be taken up
through the ventilation system.
There's been sufficient leakage at times
during the past ten years that has actually leaked
down along the penetrations, through the floor of this
service structure and through the insulation and
gotten onto the top of the head.
MEMBER SIEBER: My recollection is that
it's pretty windy in that area.
MR. GROBE: I haven't been there.
MEMBER SHACK: That is a plate though
MR. GROBE: Yes.
MEMBER SHACK: There was some picture
there yesterday that gave me the impression of a
gridwork that you attached the insulation to rather
than a plate.
MR. GROBE: I think it's a framework. Is
MR. MCLAUGHLIN: It's angle iron.
CHAIRMAN APOSTOLAKIS: Identify yourself
MR. MCLAUGHLIN: This is Mark McLaughin
from Davis-Besse. There is actual angle iron that
goes across the service structure. That's what the
insulation is laid on top of.
MEMBER ROSEN: So you would not expect
there be a large Delta P that would arise across that
structure if there was a substantial steam leak below
at the top of the head. Is that correct?
MR. MCLAUGHLIN: That would be correct.
The other thing that's not shown on there is there's
insulation. See on the outside of the flange, that's
were the reactor vessel hold-down bolts are. There's
another layer of insulation that's L-shaped that's
outside of that which covers up the bolt holes. So
that would even further restrict air flow in that area
MEMBER ROSEN: What I was getting as was
I was postulating that if you had a big leak right at
that point of steam at the top of the head that
somehow that insulation in that structure would
somehow cock and cause some stresses. I'm trying to
get the sense of whether you think that's possible.
I think you're saying is this the gridwork that came
with the Delta P that could create some kind of
cocking of that structure.
MR. GROBE: No. I think there's a fairly
tight clearance around each penetration hole. This is
a sheet material. Clearly the floor of the service
structure is sheet material.
I would expect if you're discharging 2,200
pounds into this area that you're going to get a very
substantial differential pressure between these two
areas. You would see some deflection in these plates
which may result in some movement of the penetration
I don't remember who asked the question.
But they were very interesting and complex questions.
These are also restrained near the top for sizemic
purposes. I think you'd really have to get into how
much would those bowl and what are the clearances
inside before you could say how many rods would be
MEMBER ROSEN: Now you made me worry
again. I was almost to the point where I was done
worrying. I was the one who postulated this
originally. Now I'm back to work. That's exactly
what I was worried about. Because of the yards Delta
P across some of this, there would be enough
distortion caused by flexing of something that you
could have some sort of common cause failure.
CHAIRMAN APOSTOLAKIS: More about six
MEMBER ROSEN: Yes.
MEMBER SIEBER: Well, the mirror
insulation is in blocks. Right?
MR. MCLAUGHLIN: I'm sorry. I didn't hear
MEMBER SIEBER: The mirror insulation is
in blocks. Right? It's a puzzle that you put
MR. MCLAUGHLIN: The way the mirror
insulation was manufactured is if you look at it
there's a flange right up above the insulation.
MEMBER SIEBER: Right.
MR. MCLAUGHLIN: The mirror insulation is
really in long strips, I'll say. Each strip has a
cut-out area for half of a nozzle along an entire row
though. So what they did is they slid it in on its
side. Then they laid it on top of the angle. So the
insulation is installed with long strips.
MEMBER SHACK: It's like around recessed
lighting in your basement.
MR. MCLAUGHLIN: Exactly. If you cut it
around if you have recessed lighting in your basement
and you cut half of one of your ceiling tiles, that's
how it would look. So that's how it's installed. I
would think that if you had enough of a force you
might move one strip. However there is sufficient
room between the insulation and the nozzles that it
should move up. I would think it would tend to flip
out of the way.
MEMBER SIEBER: Now is there or is there
not a plate involved here someplace?
MR. MCLAUGHLIN: There is no plate.
MR. GROBE: What's the construction of
this, Mark, the floor of the service structure?
MR. MCLAUGHLIN: That's just showing the
circle. There's no plate inside there. The only
thing that you have is the angle iron that supports
MEMBER SIEBER: The insulation is sitting
in there loose.
MR. MCLAUGHLIN: That's correct.
MEMBER SIEBER: Does that help you?
MEMBER ROSEN: A little bit. I'd actually
like a more detailed drawing so I could conclude.
MR. GROBE: Okay. Thank you. The tubes
and fins of the containment air coolers obviously are
cooler than atmosphere. Anything that's in the
atmosphere they'll condense water out of the air as
they're cooling the air. Contaminants in the air and
moisture in the air will plate out on the fins and
The containment air coolers need to be
cleaned occasionally depending on leakage inside
containment. They were cleaned in 1992. Prior to
some substantial leakage, there was equipment that
needed corrective maintenance in the 1998 time frame,
late '98/early '99 which resulted in unidentified
leakage in containment going from about one-tenth of
a gallon per minute to about 0.8 gallons per minute.
During that time frame it was necessary to clean the
containment air coolers 17 times.
A mid-cycle outage was taken in April 1999
to repair that equipment. Unidentified leakage only
went down to about 0.3 gallons per minute after that
outage. It remained higher than it had been prior to
Also during this time frame after the mid-
cycle outage, the containment air coolers had to be
cleaned twice in late '99 and seven times throughout
2000 and 2001. During that time frame, the engineers
reported that the character of the material on the
containment air coolers had changed.
Previously it might appear as a spray
painting, a very white dusty material on the fins and
the tubes. During this time frame it took on a
different color. It was dark brown. The Davis-Besse
staff assumed that the change in color was due to
corrosion of low alloy steel components in the air
MEMBER ROSEN: Did anybody do any
measurement of the activity of that deposit?
MR. GROBE: No. I don't believe so. When
you say "activity" you mean specific activity, radio
MEMBER ROSEN: Yes.
MR. GROBE: I'm not aware of that. I'm
not sure if the Davis-Besse folks here are aware of
that either. I did not ask that question.
Okay. The radiation monitor filters.
There were routine preventive maintenance to change
the filters on the airborne radio activity monitors
inside containment every 31 days. Prior to the '99
time frame, that was sufficient to maintain that
Beginning in May '99, this is after the
mid-cycle outage, the frequency of filter changes
increased. Between May and August of '99, it went
from about once a month as a preventive activity to
every other day. In July '99, the engineer
responsible for this equipment requested to have the
material analyzed on the filter.
The filter itself had previously never
appeared reddish-brown in color. That was the
character of the filter in this time frame. It was
analyzed in July '99. The analysis came back that the
filter was clogged with boric acid and iron oxide that
was produced in a steam environment, not surface
The facility staff looked for a leak that
might cause this. They were unable to find one. They
assumed that the leak was from flange leakage. You
can't observe the flanges during operation.
In August '99, they installed banks of
HEPA filters with high volume fans to try to reduce
the frequency change for the radiation monitor
filters. That was successful. It reduced it to about
every other week.
In July '01, the frequency gradually began
to increase again. This is after refuel outage in
2000. It continued to increase to every other day.
In October '01, the staff reported that the filters
were abnormally dark brown.
MEMBER KRESS: Are these little filters?
MR. GROBE: I haven't seen them. What's
the physical size of these filters? I don't think we
have anybody here that's seen them. They're in-line
filters in the air sampling system so I don't expect
them to be very big.
MEMBER KRESS: They're small I would
MR. GROBE: Yes. I've talked about the
containment air coolers and the rad monitor filters.
Nothing associated with the air coolers was reported
in the Corrective Action System.
The rad monitor filters was captured in
the Corrective Action System. But the Corrective
Action was inadequate to identify the source of the
material. In fact some of the actions taken
potentially insulation of the HEPA filters masked any
ability to detect whether it was increasing on the
I want to talk next about the Boric Acid
Corrosion Control Program. I think you're aware that
this is an NRC required program. Through our Quality
Assurance Regulations, it's clearly a procedure
affecting the safety of the plant. So it's required
to be implemented.
In 1998, we issued a bulleting that
required licensees to describe their program for
monitoring boric acid. It's an extremely sensitive
but not on-line of course way of detecting leakage.
Just a little analogy here. One drop per second will
leave about 15 pounds of boric acid in a year. So
it's an extremely sensitive indicator of leakage.
Ongoing nozzle flange leakage. The
engineer responsible for maintaining the quality of
the flanges was provided a period of time each outage
to repair nozzle leakage, flange leakage. During some
outages there was a little flange leakage. All of
them were repaired.
During some outages there was more
extensive nozzle leakage. The engineer would
prioritize those nozzles as far as how badly they were
leaking and get as many of them repaired as he could
before it was time to restart the unit. Nozzles were
left in service leaking.
In 1990, the Davis-Besse staff identified
that it was necessary to have a modification to the
skirt beneath the service structure. The mouse holes
or the weep holes at the bottom of that skirt were not
sufficient to do adequate inspections and cleaning of
the vessel head. That modification would involve a
number of large diameter openings around the parameter
of the skirt, much higher in that skirt structure.
That modification was approved for
implementation in the early '90s. I think it was '94
or '95. It was scheduled in successive outages and
deferred out of each of the successive outages. So
the fact that the licensee was unable to do thorough
inspections and cleanings of the head was of their own
Reactor vessel head boric acid deposits
were not removed at the end of each outage. It was
believed throughout that period of time that boric
acid deposits on the head were not significantly
hazardous. Moisture would be driven out of the boric
acid and the remaining crystals would not be
In the '96 outage, the boric acid that was
left on the head was characterized as "patches of
white loose consistency material." What could be
gotten was cleaned up with mechanical means vacuuming.
In '98, the boric acid was characterized
as "fist-size clumps and a thin layer of generally
brown boric acid around the center penetrations."
Again, most of the boric acid was removed by just
In the year 2000, the boric acid was
characterized as "accumulating over the head." There
was a thick layer of boric acid in the center of the
head. I'm going to put a slide up now. This is from
the 2000 Bulletin and as Bill Bateman mentioned a few
minutes ago, the staff did not have the opportunity to
see the condition of this part of the vessel head.
The Boric Acid Control Program clearly
indicates that if there are indications of red or
brown coloring, that's an indication of corrosion. It
should be pursued.
In 2000, this material was approximately
one to two inches deep. It had flowed out the weep
holes. In fact, the material inside the weep holes
was high enough to cover the weep holes. The material
had to be removed with crowbars. Eventually a water
wash was used to dissolve some of the material. But
a substantial amount of material was left on the head.
This was documented in the Corrective
Action Program as was the boric acid on the head
throughout this period of time. The close-out of the
Corrective Action Program document, the Condition
Report, actually they call them "peacocks" at Davis-
Besse at this time, was listed as "head was cleaned
MEMBER ROSEN: I'm sure that you're going
to take a close look at the corrosion effects of all
this leakage on those bolt circles.
MR. GROBE: Yes. We issued a confirmatory
action letter that requires a review of the entire
primary reactor coolant system. Not only the head and
the bolts on top of the head, but throughout the
entire system including the bottom head and other
Clearly there were indications of reactor
head corrosion. They were not recognized as
indications of corrosion and not evaluated.
The licensee described the preliminary
root cause, outside diameter, primary water stress
corrosion, cracking cavity caused by boric acid
corrosion. Significant corrosion began at least four
years ago. It's pretty difficult to argue with any of
There's a lot of issues that are clearly
not addressed yet at least in documents that we've
seen. They haven't submitted their corrective action
document to us yet.
There's very interesting chemistry I'm
learning from this opportunity. Boric acid crystals
begin to react with air at a temperature far below the
temperature of the head and begin to form boric oxide.
In addition to that the melting temperature is only
slightly higher then the temperature at which that
So you could have had a very interesting
combination of boric acid, boric oxide, and liquid
boric acid flowing down the head. It's not clear what
role that chemistry played in that cap over the top of
the head and corrosion that might have initiated from
the head down.
The role of head temperature throughout
the operating cycle, outage times, start up times, it
appears that there were times that boric acid was
pooled in the bottom of this cavity. That's certainly
an opportunity during shut down times when the head is
at ambient temperatures. It's not clear what role
that may have played in the corrosion process.
The rate at which the cracks progressed
and the corrosion progressed is not clear. I don't
see a reason to believe that the corrosion progressed
at a uniform rate through the years. So those issues
are not answered. Clearly the correlation between
Davis-Besse and the rest of the industry hasn't been
So there's a lot of outstanding questions
that I'm hoping are answered to a large extent in the
licensees root cause assessment. That completes the
information. I apologize for being quick.
MEMBER FORD: Jack, who has the action to
provide that data.
MR. GROBE: I'm sorry.
MEMBER FORD: Who has the action to
provide that data.
MR. GROBE: The licensee is required to
provide us the root cause. It's not clear to me that
those questions can be answered without research. The
grinding operation on the nozzle in penetration 3
started. The nozzle twisted a little bit and tilted
a little bit.
At that point the licensee did extensive
cleaning operations on the top of the head to discover
the cavity. All of that material is gone. Had we
been able to take samples of that material, it would
help. The licensee at that point had no reason
preserve that material because they didn't understand
what was going on. Maybe that's reason enough to
In addition, of course all the cracks were
machined out. So we have no information on the
cracks. It's not clear to me that we're going to have
sufficient data from the licensee's analysis to answer
all these questions. Likewise it's not clear to me
that we need all those answers necessarily to approve
an appropriate repair to the head.
Those answers are important for going
forward as far as Davis-Besse and the rest of the
industry. So there's a lot of things that play here.
I anticipate there may be some research, Hackett's
ears are perking up, that will come out of this.
MEMBER FORD: That comes down to the
question of the timing of which this research goes to
get to an identifiable goal. Bearing in mind that
it's assumed that there are no other observations of
such magnitude in the existing fleet. Until we have
that data we don't know. Tomorrow it may start,
unless we know the chemistry, physical dimension
MR. GROBE: It may be that the right
answer is to do volumetric examinations of these areas
every outage. I don't know what the right answer to
MEMBER FORD: Okay.
MR. GROBE: Then you never get into this
situation. At least not from these cracks.
MEMBER POWERS: This is the part that I
don't quite understand, Peter. In the inspections of
heads that we're doing elsewhere, are we looking for
boric acid corrosion of the mild steel pressure
MEMBER FORD: Inside the annulus?
MEMBER POWERS: Yes.
MEMBER FORD: Not as far as I know. Not
unless they're doing 100 percent UT. They're not.
MR. STROSNIDER: This is Jack Strosnider.
I just wanted to make two comments on the discussion.
First of all with regard to the research, NRR has
requested the Office of Research to start doing some
work in this area including looking at what
information is already available. Also looking at the
feasibility of mock-ups. We've also had some
additional discussions with the industry I believe
with regard to doing that kind of work.
With regard to what the inspections are
expected to look at, I think that's a subject of the
next presentations. In particular Bulletin 2002-01.
When you hear the presentation, you'll see that's
exactly the issue that we're trying to get to in that
CHAIRMAN APOSTOLAKIS: If I look at this
incident from the New Reactor Oversite Process. Is
MR. GROBE: The licensee's analysis puts
it at the white, yellow order. We haven't even begun
to review that. That's the next inspection that will
begin in the next week or so, both to look at the
regulatory implications of the findings of the AIT as
well as the risk analysis.
CHAIRMAN APOSTOLAKIS: But are you using
the action matrix right now? No.
MR. GROBE: The AIT, the Augmented
Inspection is an event response. Now we'll go into
the follow up inspections and apply the Significance
CHAIRMAN APOSTOLAKIS: Okay.
MR. GROBE: It's an interesting
CHAIRMAN APOSTOLAKIS: Yes. We've been
hearing a lot about the utility personnel there and so
on. How about the resident inspectors?
MR. GROBE: That's an excellent question.
As part of the follow up activities, I'm required to
recommend to appropriate offices actions to take.
CHAIRMAN APOSTOLAKIS: Were they aware of
any of this?
MR. GROBE: No. The residents were not
aware. Our inspection program does not require
inspections in these areas. The in-service inspection
program primarily focuses on piping and welds in the
BWRs, BWR internals, as well as steam generators.
Reactor vessel heads was not included as part of our
CHAIRMAN APOSTOLAKIS: They were aware of
the fact that the 1990 modifications to improve the
reactor vessel heads had not been installed.
MR. GROBE: No.
CHAIRMAN APOSTOLAKIS: They were not aware
MR. GROBE: No. I don't know how many
modifications every year that Davis-Besse has. But I
would expect that it's certainly in the dozens and
maybe many more than that. Corrective maintenance
activities would be in the thousands. So the chance
that a resident inspector may choose to pick one of
these activities to look at is fairly small.
CHAIRMAN APOSTOLAKIS: Now the Corrective
Action Program is one of the cross-cutting issues. Is
MR. GROBE: That's absolutely true.
CHAIRMAN APOSTOLAKIS: So what? We're not
doing anything about it. It's an old issue between us
and the staff. The staff claims that even if you have
a defective Correction Action Program, then you will
see the consequences of that. That's what happened
MR. GROBE: I think that's what we have
MEMBER ROSEN: I think that's what you
said, Jack, is that you're doing a Significance
MR. GROBE: Right.
MEMBER ROSEN: What comes out of that is
what's off the action matrix.
MR. GROBE: Exactly. Also to answer your
question, we're going to have to look at our
inspection program and how we implement it to make
sure that we're addressing appropriate inspection
CHAIRMAN APOSTOLAKIS: The question is
whether you should stick to this point of view that if
there are problems with the Corrective Action Program
let them be until something happens or you should try
to devise some ways of evaluating the quality of the
Corrective Action Program before things happen.
MEMBER ROSEN: I don't think your premise
is correct. I don't think that they do. I'm not
talking about Davis-Besse, any place without a serious
event. If the inspection, resident inspectors and the
NRC find that the Corrective Action System is somehow
not working as it should, then that becomes an issue.
CHAIRMAN APOSTOLAKIS: They're not
looking, Steve. They're not looking.
MEMBER ROSEN: I think they are.
CHAIRMAN APOSTOLAKIS: No. It becomes a
MEMBER SIEBER: There's a module for that.
CHAIRMAN APOSTOLAKIS: There's a what?
MEMBER LEITCH: It's 4500. Isn't it?
MEMBER ROSEN: I think it's a major focus
of the inspection program now.
MR. GROBE: There's three areas where we
look at the Corrective Action System. There's an
inspection that's now conducted every other year which
is a team inspection. It's a large inspection. It
covers several weeks.
CHAIRMAN APOSTOLAKIS: Of what?
MR. GROBE: It's of the Corrective Action
System itself. A wide variety of condition reports
are chosen on a risk informed basis to examine the
effectiveness of the Corrective Action System.
There's also a series of interviews of staff across
the facility to get a sense for their safety focus as
In addition to that a certain percentage,
I believe it's 10 percent of the hours of every
inspection whether it's a radiation safety inspection,
security and safeguards, maintenance, surveillance
testing, or whatever it may be, is intended to spend
in the Corrective Action area looking at Corrective
Actions for deficiencies identified in that specific
area. In addition to that now we're implementing
sampling of about ten more minor events.
Events that wouldn't get to the level of
a special inspection where you send a team out to the
region. More minor daily events that by following our
nose, catch our fancy. We spend a little bit drilling
more on that specific event into how it happened. So
there are three ways we look at the Corrective Action
It's very difficult to apply the
Significance Determination Process to Corrective
Action violations. The Corrective Action Program if
it's a violation of not fixing things correctly, it
will most likely found the issue before it became
significant from a risk perspective. But didn't fix
it properly. So by definition that would be a low-
There's still quite a bit of dialogue
among myself and my peers about whether or not it's
appropriate to apply a risk-based, risk-driven
Significance Determination Process to a Corrective
Action Programmatic deficiency. Or whether there
should be some programmatic Significance Determination
Process developed that's more deterministic.
MEMBER ROSEN: So given all that, what was
the staff's conclusion about the Corrective Action
Program at Davis-Besse prior to this event?
MR. GROBE: The staff's view is that the
Corrective Action Program is well implemented at
Davis-Besse. That's what's very troubling. It's
something that I'm going to be getting to the bottom
of over the next several weeks, maybe months.
The extent of the behavior that created
this problem is multiple people weren't following the
Corrective Action Program. For example, engineers
were not speaking laterally. The rad monitor engineer
wasn't talking to the containment air cooler engineer,
who wasn't talking to the head engineer.
There were several decisions that were
made which included supervision and management that
don't appear to have been good decisions. Some
examples are the delay of the modification,
installation of HEPA filters in containment, the
decision to not continue to pursue the source of iron
oxide in the '99 time frame, quite frankly the
decision to restart after the 2000 refueling outage.
So there's just a plethora of issues that
we need to continue to follow up on. Why those
decision making processes, communication processes,
supervision deficiencies didn't manifest themselves in
other areas, that's another question we have to ask
ourselves and try to find the answer to. But they
didn't. I'm fairly comfortable with our inspection
CHAIRMAN APOSTOLAKIS: Okay. They didn't.
But we, the NRC, have no way of finding out that they
did not because we were not looking for that. Is that
correct? We were not looking for the existence of
communication channels between this group of engineers
and that group of engineers because that's a safety
issue. We're not supposed to look at that. Is that
MR. GROBE: Whenever you identify, it's
what I refer to hardware and software. Most problems
have fixes in two sides. They have a hardware fix.
For example in this case potentially drilling out a
hole in the head, installing a plug, welding it in.
They also have a software fix. It's a human
performance problem or a communications problem or a
CHAIRMAN APOSTOLAKIS: Right.
MR. GROBE: We look at all of those issues
when we look at fixing a deficiency in the facility.
If it's our violation, we follow up on it. The 10
percent of each inspection procedure is spent doing
that. We pick about a half a dozen less significant
events per year. We drill down in each one of those
to make sure that the root cause is identified and
fixed. Every two years we spend a significant period
CHAIRMAN APOSTOLAKIS: I think I'm getting
a different picture from you of what our inspections
do. Then you guys would develop the ROP.
MR. GROBE: Well, I can tell you that you
get a picture of what we're doing in Region III. I
believe it's the same as the other regions.
CHAIRMAN APOSTOLAKIS: Yes.
MR. GROBE: I apologize.
MEMBER POWERS: In fairness, you explained
this when we visited you. All of the regions have
explained this. They do this baring down on the less
significant issues and things like that. It's one of
the values of our visit to the regions.
CHAIRMAN APOSTOLAKIS: I know. Sure.
Another thing that you said that I find very
interesting is you said that you are not sure of the
Significance Determination Process as it is structured
now. That makes sense for things like the Corrective
Action Program. Put another way, should we evaluate
everything on the basis of CDF and LERF? That's
really what you are saying.
MR. GROBE: Exactly.
CHAIRMAN APOSTOLAKIS: I don't think we
MR. GROBE: I agree.
CHAIRMAN APOSTOLAKIS: You agree with me.
MR. GROBE: When you look at the Design
Control Program for example if our inspectors go in
and we spend a week and we find 20 calculational areas
which are not minor oversights like a transposition of
numbers or something like that --
MEMBER ROSEN: This is at Davis-Besse.
MR. GROBE: No. This isn't Davis-Besse.
This is philosophical.
MEMBER ROSEN: I apologize. I won't
CHAIRMAN APOSTOLAKIS: That's fine.
Philosophy is good. Keep going.
MR. GROBE: If you find 20 calculational
areas where the calculational area had a precursor of
not understanding the engineering a mis-application or
a mis-assumption or something of that nature but each
one of them came out as to not render the equipment
inoperable, currently the Significance Determination
Process would classify those as either minor or green.
They would be non-cited violations.
When in fact that's a clear precursor that
there's a problem with the competency of the engineers
as well as the competency of the engineering
supervisors. So there are areas and these are the
things that we're still working out in implementation
of the ROP.
I think the Corrective Action Program is
likewise. It needs something less than less rigorous
analytically than a risk analysis to evaluate the
significance. I certainly appreciate this podium to
express these views. I don't get it very often.
CHAIRMAN APOSTOLAKIS: It can be a risk-
like analysis but not using core damage frequency is
the end stake. Something before that.
MEMBER ROSEN: It sounds to me like what
you're suggesting is the Reactor Oversite Process
ought to be risk-informed not risk-based.
MR. GROBE: That's exactly right. In some
areas it can be risk-based, but overall it should be
CHAIRMAN APOSTOLAKIS: Nothing we do is
MEMBER ROSEN: Well, if you're writing
something that's agreeing because it's number that
you've calculated is way down there, that's risk-based
CHAIRMAN APOSTOLAKIS: No, but that's a
MEMBER ROSEN: What Jack is arguing for is
a true risk-informed regiment which is in my view the
right answer. It's always I think the wrong answer to
use a risk-based regiment.
CHAIRMAN APOSTOLAKIS: No, but the point
is should you be using core damage frequency to make
all these determinations. I think that's a
VICE CHAIR BONACA: For example one
concern that you have raised and I brought out at
least personally was the fact that the Significant
Determination Process doesn't take into consideration
CHAIRMAN APOSTOLAKIS: That's true.
VICE CHAIR BONACA: And yet it is
something that traditionally we have looked very hard
at the plans as indicators of problems with the
Corrective Action Program. You fix something, you say
you fixed it and it's not fixed again and again.
That's a major indicator. Yet the Significance
Determination Program doesn't deal with that.
CHAIRMAN APOSTOLAKIS: Also the example
with the calculations is a very good point.
VICE CHAIR BONACA: Yes.
CHAIRMAN APOSTOLAKIS: Because you have 10
wrong calculations spread over time. Each one would
probably become a "green." But if you find a common
cause behind them then I don't know what you are going
MR. GROBE: I think we still have growth
in the area of how to apply our risk tools. A good
example of that in the maintenance area was at Quad
City several years ago. They were incorrectly
maintaining their motor operated valves. They were
repetitively failing. But at each failure they didn't
have redundant equipment in a failed state or out of
Consequently there was essentially no risk
significance to each individual failure but there were
17 valves that failed over a period of two years. It
was because the maintenance activity was inadequate
and the Corrective Action Program wasn't identifying
it. So that's a situation I think that goes to right
to both these issues.
CHAIRMAN APOSTOLAKIS: Exactly.
MR. GROBE: We need to continue to mature
in how we are using our risk tools.
CHAIRMAN APOSTOLAKIS: Very good. It has
been really very useful.
MR. JOHNSON: George, this is my chance.
Over here at the table. George.
CHAIRMAN APOSTOLAKIS: Oh, you again. I
thought you weren't in the room, Mike.
MR. JOHNSON: I was hoping not to say
anything here. But I couldn't not say anything. I do
want to point out that we have had continuing dialogue
with ACRS on cross-cutting issues. I couldn't sit
there and remind us that the goal of the ROP was never
to make sure that we didn't have issues. There is
never a guarantee in the ROP that would say that we
would not have issues and then you would find and look
back and say hey you know what. There were some
cross-cutting issues that if the licensee had taken
care of we wouldn't have gotten here.
In fact what the philosophy of the ROP is
is that if in fact there are problems in cross-cutting
areas that those will be reflected in performance
issues like perhaps this performance issue that we're
talking about in time for us to take action before the
performance is unacceptable. So that's the premise of
the ROA. I wanted to be very clear about that.
The other thing is that I wanted to be
sure that we remember that the commission has given us
some specific direction with respect to treatment of
cross-cutting issues. The direction from the
commission was before the agency takes action on a
cross-cutting issue we need to make sure that it is an
issue that has reflected itself in terms of
performance that it has crossed some threshold.
So the commission has been very clear with
us with respect to our previous process of looking at
issues that have continued to aggregate if you will.
Aggregation was a feature of the previous process and
has steered us away from aggregation towards where we
are in the ROP.
I'm sorry, George. I just couldn't sit
there and not say that.
CHAIRMAN APOSTOLAKIS: Are you still the
head of that?
MR. JOHNSON: No, I am not.
MEMBER FORD: George, I have one question
from the public. Then I'd like to get back on to the
CHAIRMAN APOSTOLAKIS: Sure. We can never
MEMBER FORD: That's true.
MR. GUNTER: Paul Gunter, Nuclear
Information Resource Service. Just a quick question.
Jack, could you inform me if the 1990 modification
that Davis-Besse didn't undertake was that part of
compliance with generic letter 8805? I mean 8805 had
a specific piece about increasing accessibility for
inspection. I'm wondering in what context did the
1990 modification come about. Did Davis-Besse just
volunteer it or was this part of 8805?
MR. GROBE: That's Paul Gunter by the way
for the records. Paul, 8805 didn't require any sort
of modifications. It simply required the licensee to
have a program in place that addressed certain
attributes of boric acid corrosion management and to
describe that program to us. The modification that
was identified in 1990 was proactive in a sense that
the Davis-Besse staff identified for themselves that
this would be a benefit to them. There wasn't any
requirement to implement a modification of any sort.
As a matter of fact of the B&W pressurized
water reactors most of them have implemented such a
modification. Some have not. So it's simply a matter
of what a licensee views is necessary for their own
The disturbing issue at Davis-Besse is
that over the years their staff had identified that
one of their inabilities to effectively inspect and
clean the head what influenced that inability was the
fact that they had limited access through these mouse
holes or weep holes. That reemphasized the need for
implementation of the modification. I think I've
answered your question.
MEMBER FORD: I'd like to move on if I
may. Ken, do you want to swap your presentations?
You deal with 2002-01 and finish off with 2001-01.
It's a suggestion.
MR. KARWOSKI: That's fine. For
continuity purposes, I'll be discussing Bulletin 2002-
01 which was issued in response to the findings of
Davis-Besse. Just to recap, the NRC is taking a
number of generic actions as a result of the findings
at Davis-Besse. I'll be discussing some of those.
I'll also be discussing some of the results that we
have to date as a result of reviewing responses to the
bulletin and talking to licensees.
Just to go through it quickly because I
know we are behind schedule. The first slide just
recaps what we knew about the findings at Davis-Besse
at the time. We knew that they had boric acid on the
top of their head and we knew that they had leaking
With that information and the knowledge
that there was a cavity, we contacted the industry and
asked them three questions. Those three questions are
listed on this slide. Basically we asked them for
plants that had just recently completed their
inspections in response to Bulletin 01-01 which had to
do with circumferential cracking of the nozzles. Were
the techniques used during that inspection capable of
detecting the type of wastage that was observed at
The other thing we asked them is to
provide a justification for continued operation for
the plants that had not performed those inspections at
that point. We also asked them for a risk assessment.
The industry conducted a survey and Larry
Matthews of MRP described that survey. They
categorized their results. While the industry was
performing that survey and about the time we received
those results, the NRC issued Bulletin 2002-01 on
March 18. We had several reporting requirements in
that bulletin and I've listed those on this slide.
Within 15 days of the date of the
bulletin, we asked licensees to provide a summary of
the reactor vessel head inspection and maintenance
programs. We asked them to evaluate those programs
for the ability to detect degradation such as what was
observed at Davis-Besse. We asked them to identify
conditions that may lead to degradation such that was
observed at Davis-Besse. We also asked for their
plans for their next inspection outage and then the
justification for continued operation.
We also asked that within 60 days that
they provide a more comprehensive evaluation of their
Boric Acid Corrosion Prevention Program. We also
asked the results of their next inspection to be
provided within 30 days of the completion of that
With respect with where we stand today,
the staff as a result of the MRP survey, we took the
plants that were listed in the other category that
were on the slides of Larry Matthews that presented
including Beaver Valley, Calaverdi, Wolf Creek, Watts
Park. We've contacted all those licensees because of
possible concerns because the other category is a
category where the results of the inspection were
questionable and we felt we needed to understand a
little better why they were categorized that. Some of
those plants have subsequently performed inspections.
We are still pursuing additional information from one
of those plants.
We are also contacting licensees that are
currently in outages to obtain the results of their
results of their inspections and also to discuss their
plans for the inspection recognizing that the bulletin
went on the 18th and the responses weren't due back
until the first week of April. We wanted to make sure
that we understood the licensees inspection scopes and
we wanted to make sure that the results of inspection
whether or not we wanted to evaluate those results to
determine whether or not we needed to take additional
regulatory actions. Those phone calls are still on-
As a result of those phone calls, we have
not identified any other plant with similar
conditions. In most cases, I have characterized the
results as there is small debris on the top of the
vessel head. That debris could be a result of
maintenance activities and be metal shavings or pieces
of metal or small pieces of boric acid crystals as a
result of previous leaks but nothing to the extent as
what was observed at Davis-Besse.
We are reviewing the responses to the
bulletin. We have completed initial categorization.
We are proceeding on those reviews now. That's
basically where we stand with respect to the
activities of this bulletin.
MEMBER FORD: Thank you, Ken. Questions?
MR. HISER: I'd like to describe that the
status of review of Bulletin 2001-01 looking back that
was on circumferential cracking of vessel head
VICE CHAIR BONACA: Could I ask a
question? I'm puzzled. It will be a quick question.
When they looked at the Davis-Besse, they looked from
the bottom. Then they did the inspection and
identified cracking I guess through UT inspection in
the sense. So that means they never looked from the
top because of the super structure (PH) I guess it
MR. HISER: As a part of the 2001-01
inspections for the prior bulletin, they looked using
ultrasonics to determine whether or not they had any
circumferential cracks. As a part of their overall
activities, they intended to do a visual inspection of
the head as well. The sequence of events was such
that they completed their ultrasonic inspections and
then begun repairs before they did their visual
VICE CHAIR BONACA: I just wanted to make
sure for the other plants in genera that there is
always a plan to inspect visually from the top.
MR. HISER: For many plants that's true.
For some plants the insulation configuration is such
that the insulation is directly on the head. Then
there are cases that it really isn't feasible to do a
visual exam of the head's surface.
VICE CHAIR BONACA: So would you find the
same problem if you -- Do you see where I'm going?
MR. KARWOSKI: There are a number of
plants whose insulation is either glued or cannot be
removed for the head easily. One of the recent plants
that shut like that is Genet. They had a well
documented history of prior leaks. They also did a
visible inspection of the surface of the insulation.
In areas where it was stained they cut up
pieces and looked down to the bare metal. They also
did additional examinations in areas where there was
a known prior history of leaks. In the case of Genet
specifically they did UT thickness measurements from
the bottom of the head near the center nozzle. They
also did some UT in the periphery around the shroud
ring as result of a prior leak in that area.
So there are other actions that plants who
have nonremovable insulation can take. Certainly if
they have never had a leak there is a possibility that
leakage would come down from the top.
VICE CHAIR BONACA: But you would expect
provisions however that they would take so if there is
a faradic erosion over time taking place in the
ferritic steel would be identified.
MR. KARWOSKI: Yes. I was just addressing
the corrosion from the top of the head.
VICE CHAIR BONACA: I understand. I have
just been wondering though since in some cases you
cannot have a visual from the top, how do you assure
that if you have an event of this type it's going to
be identified in all cases? That still puzzles me.
MR. BATEMAN: Just a point of
clarification. Bill Bateman from the staff. When Ken
says leaks, he's referring to flakes from above from
the phalanges at the conoseals that would run down and
land on the header and the insulation.
MR. HISER: One of the things that the
industry talked about on Tuesday was interpretation of
the ultrasonic data above the weld and the inference
fit zone and the ability of that to characterize
whether they have metal behind the nozzle or not.
That's one approach that the industry is taking.
VICE CHAIR BONACA: But they're addressing
MR. HISER: Right. Here's what I would
like to do today is to just provide a brief summary of
the inspection results and how that fits within the
context of the susceptibility ranking approach and
then provide some observations and forward looking on
where we are headed with this.
The table illustrated here provides the
inspection results for all the high susceptibility
plants along with two moderate susceptibility plants,
Crystal River 3 and Millstone 2 that did identify
cracked nozzles. In general, plants have tried to use
a qualified visual exam if they are able to do that.
Again the qualified visual means that you are able to
inspect the inner section of the nozzle with the head
so that you can split to that bare metal to see if
there are any boric acid deposits. Also you have done
a plant specific analysis to demonstrate that any
leaks in the annulus between the nozzle and the base
metal would provide a deposit on the head that would
be available for detection. In some cases in
Millstone 2 and Davis-Besse, they also did a 100
percent ultrasonic inspection because they were not
capable of doing a visual exam with the as-found
Now for the plants that have identified
leaking or cracked nozzles, any positive findings from
the qualified visual exam were followed up with
ultrasonic techiques in order to characterize the type
of degradation or is it actual flaws or a
circumferential flaw whether it was through wall or
not. A number of nozzles have been repaired. I guess
two things to point out is from the susceptibility
rankings, we do have two plants in the moderate
susceptibility bin that have found cracked or leaking
nozzles. One of those Crystal River 3 is actually the
first plant in the moderate susceptibility range.
They did identify a circumferential crack in the one
nozzle. Millstone 2 identified three nozzles with
crack from the ultrasonic test. None of those were
thrown wall and none of them appeared to provide any
Some discussion of Oconee 3. That was the
first plant that identified circumferential cracking.
That was identified in February of last year during a
midcycle maintenance outage. A refueling outage in
past November did identify additional degradation with
the seven nozzles having cracks or leakage. One of
those nozzles did have a circumferential crack.
So I guess some of the points to be made
here is at this point all of the high susceptibility
plants with the inspection of Davis-Besse have been
inspected. We have continued to find cracked nozzles
and also some circumferential cracking. Looking at
this within the context of the susceptibility ranking,
plants are within zero to five EFPY of Oconee 3 were
classified as high susceptibility. As you can see
many of these have identified cracked nozzles. In two
cases they have not from recent inspections this is
the Crystal River --
MEMBER SHACK: Those are really leaking
nozzles. Right? They did visuals.
VICE CHAIR BONACA: That's right.
MR. HISER: In some cases. In at least
one plant all of the nozzles that were found to be
cracked did not have definitive indications of leakage
on the head, did not have definitive conclusions of
MEMBER SHACK: No, the two that we have
down there in the high zone that say no cracking.
Those had some visuals on them.
MR. HISER: That's correct. Yes.
MEMBER SHACK: So the no leaks is the true
MR. HISER: No leaks. Yes. That is
correct. The highest ranked plant that has leakage is
Crystal River at this point. Again Millstone 2
identified cracking because they did an ultrasonic
exam. Probably if they had done a visual exam they
probably would have been a blue square. We would have
said they have no cracking. As you can see there
clearly are a lot of plants that still will be doing
inspections either later this spring, next fall or
even next spring because of the cycle of outages.
MEMBER FORD: Allen, did I hear that
correctly that particular plant a visual inspection is
not sufficient to determine that you have no cracking?
Is that what you said?
MR. HISER: In this case the cracking that
was identified as the maximum extent was about 40
MEMBER FORD: Oh. So there it wasn't a
MR. HISER: Right. It was not a through-
VICE CHAIR BONACA: Some of the confusion
is that you are using the expression "cracking." You
should use the expression "leaking" because that
really is what you are monitoring with the exception
of that plant there, Millstone 2. I would suspect
that all of them are somewhat cracked.
MR. HISER: They may be. That's correct.
We'll improve the indications on this chart.
MEMBER SHACK: No. Matthews' chart says
it has four plants with volumetric inspection that had
VICE CHAIR BONACA: I thought there were
two. There were two on that table. Only two plants
with UT. Millstone 2 and Davis-Besse.
MEMBER SIEBER: But there were others who
MR. HISER: Yes. The plants that are
shown in the table are predominantly those that are
less than five EFPY. Some of these other plants
probably also did ultrasonic inspections. They should
be indicated a little bit differently. That's
I guess the one point we wanted to make is
that although all of the leakage is down in the low
EFPY area we have seen cracking here. Ultimately it
is going to get to the point that cracking extends
throughout the histogram. At this point in time the
history does justify I think the susceptibility
ranking model that we have.
MEMBER POWERS: I guess that's not
apparent to me. You have appointed 15 EFPY. It seems
to say that this ranking is not correct.
MR. HISER: From the standpoint of
circumferential cracking in nozzles, the plant had no
circumferential cracks. It had three nozzles with
about 40 percent through-wall.
VICE CHAIR BONACA: And no leakage.
MEMBER POWERS: If I wait until 12 EFPY it
has two wall cracks.
MEMBER FORD: I think an explanation,
Dana, is that this model is based purely on time and
temperature. It misses out the fact there is
differences in stress and especially differences in
heat. Therefore you are going to expect a scatter
around those values. So it doesn't surprise me at all
that you have at least one plant who when you look at
the distribution of those plants that have seen
VICE CHAIR BONACA: If that plant had
performed a visual --
MEMBER POWERS: Well, I think what this is
telling you is that this ranking is just not adequate.
MEMBER FORD: You're always going to
scatter around those points. You are absolutely
VICE CHAIR BONACA: If that plant had
performed visuals like the other reds it would not
have been red but it would have been green.
MEMBER POWERS: That also says that visual
inspection is not adequate.
MR. STROSNIDER: This is Jack Strosnider.
I'd just like to make a comment on this discussion.
As was pointed out with these susceptibility models
there are parameters that aren't taken into account
here such as residual stresses, materials, et cetera.
We wouldn't expect this to be exact.
I think the one thing I want to caution is
when we say it's not exact. When we ask the question
is it adequate from a regulatory perspective, I want
to point out that even the largest circumferential
crack found in these plants had substantial margin to
Is it adequate in terms of protecting
against the circumferential crack that's going to lead
to failure? That's what we're concluding that yes the
inspections are happening soon enough to give us that
It's not going to predict this plant is
going to be at exactly this time or this plant will be
exactly before that plant. But when you look at the
results of the inspections, we believe it's adequate
to provide confidence that the cracks will be caught
in time to preclude any failures.
I guess the one other thing that I'd point
out is then you ask the next question. What about the
Davis-Besse experience and the fact that a leak lead
to the sort of thing that we saw at Davis-Besse?
That's the point of the bulletin that Ken talked
For people who have already done these
inspections, one of the things that they have to
respond to is tell us why that inspection was good
enough to tell you that you didn't have any
degradation occurring in the head. So I think you
need to look at both the bulletins and what they're
MEMBER KRESS: Yes. But there's going to
be an unfinished part of that. They're going to come
back and say we're sorry we couldn't have found the
Davis-Besse thing without inspection. Then you'll
have to come back with now what.
MR. STROSNIDER: Yes. If we see a
responsible Bulletin 02-01 which says that we can't
tell you a licensee that can't provide the argument as
to why they don't have degradation occurring in the
head, we need to have more discussions with them.
MEMBER KRESS: They'll have some
arguments. But you'll have to use judgement as to
whether they're good enough. I think what you'll find
out is they really can't tell you. Then you have the
decision to make. What are you going to do? I think
you ought to be thinking about that.
MR. STROSNIDER: We are.
MEMBER KRESS: Okay.
MR. STROSNIDER: If we get a response to
Bulletin 02-01 which doesn't provide confidence that
the type of degradation saw at Davis-Besse is not
occurring, then we will have to follow up on that.
That's the point of our argument.
MEMBER POWERS: Jack, let's come back on
this regulatory adequacy. You have this, I think it's
Crystal River up there at 15. Is that right?
MR. HISER: That's Millstone 2.
MEMBER POWERS: That's Millstone 2. I'm
sorry. You say it's okay because this things going
through a wall. Isn't that an accident? If I look at
the next plant down, couldn't it be that it has
MR. STROSNIDER: Which one?
MEMBER POWERS: One of them.
MR. BATEMAN: Right now we're managing
this issue through leakage. If we look at that plant,
do a visual inspection and we see popcorn there then
we know there's leakage. The licensee fixes it. They
don't restart until they've fixed all their leaks.
Right now the way we're managing this issue is through
MEMBER POWERS: Right now this curve is
used to tell you the urgency with which they're doing
MR. HISER: Actually I should have set the
stage on this. The bulletin had two main purposes.
First of all is to identify any plants that had a
safety issue such as the cracks that were identified
at Oconee. So far we've found no plants that have a
safety issue with large circumferential cracks.
The other is to provide us with data in a
graded approach that would help us to determine what
the long term management, i.e. inspection methods need
to be to assure that we don't get any large
circumferential cracks. Within that context, the
susceptibility ranking is supported by the data that
we have at hand.
MEMBER KRESS: I don't think you should
overlook the blue squares, Dana. They tell you a lot
MEMBER POWERS: You have blue squares down
here at three.
MEMBER KRESS: I know. You would expect --
MEMBER POWERS: They don't tell me
anything except that the curve is not adequate.
MEMBER KRESS: You expect some overlap at
that level down there.
MEMBER POWERS: It looks to me like the
density is about the same. I would argue that the
blue squares are about uniform across that grid.
MEMBER FORD: You don't think that the
ratio of cracking to no cracking changes as you go
from the left hand side to the right hand side.
MEMBER POWERS: It doesn't look to me like
MEMBER FORD: There's no red squares up in
the right side.
MEMBER POWERS: But you haven't looked.
MEMBER KRESS: I'm presuming that you've
looked at the blue squares.
MEMBER POWERS: First of all I have two
blue squares in the first block. I have four in the
next block. I have three in the next block. I have
three in the block. Two in the next block.
MEMBER KRESS: That's just an indication
of which ones you looked at.
VICE CHAIR BONACA: But let's change the
name to leaking because really the cracking is just
misleading. Those two boxes on the left between zero
and five may be --
MEMBER POWERS: That's what I disagree
VICE CHAIR BONACA: May be 90 percent
through right now. They show however no cracking. No
that's not true. No leaking. They haven't seen any
leakage. But they may be so close to all extent
they're in the same bunch.
MEMBER POWERS: I think I agree with you.
VICE CHAIR BONACA: What will you shift
the criteria? Do you call the other one up there no
cracking? That means no leaking actually. You have
seen no leaking in less than two. But you know that
there is cracking.
I can make the same statement about any of
those. I probably could go at 20 years and find some
at 20 years that have cracking but no leaking.
MEMBER KRESS: But I would be awfully
surprised to see that many blue squares if indeed
you're supposition is right. Some of them are that
close to being --
VICE CHAIR BONACA: I was talking about
the one between zero and five, those two.
MEMBER KRESS: Well, those two might very
VICE CHAIR BONACA: They may be very
MEMBER KRESS: But that just validates the
curve if that's the case.
MEMBER POWERS: It may also be true that
the two up around 15 are within 95 percent of through
MEMBER KRESS: But I would be very
MEMBER POWERS: You see if I didn't have
the red dot, I might be surprised. But now I have the
red dot. Why am I going to be surprised? You know
MEMBER KRESS: The red dot is the one
thing that raises a flag.
VICE CHAIR BONACA: That's apples and
MEMBER KRESS: If I had two red dots, I'd
be more concerned.
VICE CHAIR BONACA: But you don't have
CHAIRMAN APOSTOLAKIS: So this is the one
MEMBER LEITCH: Another important variable
and it becomes a limitation I imagine of how much you
can plot, is the inspection method.
CHAIRMAN APOSTOLAKIS: Good.
MEMBER POWERS: The one uncontested
conclusion I get out of this is visual inspection
looking for evidence of leakage is --
MEMBER FORD: This is going to come up in
further discussions because this is relating to the
policy of how you manage these.
MR. HISER: Okay. I believe initially
this whole two hour meeting was going to be on
Bulletin 2001-01. That overtook us. So we're trying
to squeeze two hours into about five minutes.
MEMBER FORD: If I could just interrupt
because this is a serious point. Dana, this will come
up for discussion in the near future to discuss that
policy with regards to how we're going to manage this.
MEMBER POWERS: Good.
MR. HISER: This says conclusions. But
really these should probably be observations and
status. I guess what I really want to focus on is the
implications of Davis-Besse to the future inspection
needs for CRDM nozzles is yet to be determined. Once
the Bulletin 2002-01 review activities are completed
and the root causes end then we will have a better
understanding of that.
In addition the bulletin addressed the
next refueling outage for plants after August 2001.
In some cases plants a year from now will be up to
their second inspection. In all honesty, the
bulleting really doesn't apply in that case. What we
hope to do is have some inspection guidance in hand by
that time so that plants will be able to implement
that next spring.
I believe that the Committee was provided
with a copy of our draft action plan that will be used
to resolve the VHP nozzle cracking issue. Again that
was drafted before the Davis-Besse findings. We have
chosen at this point not to modify it because things
are in such a state of flux. Clearly that will be
revised as the implications of Davis-Besse become
MEMBER FORD: That's both underlining I
think, Allen, that parts of the actual experiments and
analyses in that action plan are already being done by
the MRP. So you say it's a draft. It is in fact.
The actions are already going on.
MR. HISER: Yes. That's correct. That's
what we had planned to talk about today.
MR. STROSNIDER: This is Jack Strosnider.
I'd like to just add one comment here if I could to
emphasize something that Allen touched on. I don't
know if this will go fully to addressing Dana's
concern. Hopefully it might help.
Again the bulletin was just a one time at
their next outage, that's all it addressed. We
recognize that we need a longer term program to manage
this. I think that's where the work is ongoing.
The Sub-Committee heard on Tuesday and the
Committee today heard something very important from
the MRP that I just wanted to go back and highlight.
That was that the MRP has reached a conclusion that
just visual inspections to look for leakage is not an
appropriate long term method for managing this type of
degradation which has very important implications with
regard to the type of inspections that would be done.
Basically it draws you to doing volumetric
examinations and finding cracks before they ever
develop into any kind of leak at all. Hearing that
from the MRP and that's an issue that we were looking
to have some resolution on I think we'll be working
with them to look at a longer term program that
follows that philosophy. We're waiting to see their
proposal on that subject.
Recognize that, yes, there is a longer
term follow up that has to happen here with regard to
managing this problem because it will show up at other
plants. This distribution is marching forward in
time. It will have to be managed.
MEMBER FORD: I'll pass it back to you.
CHAIRMAN APOSTOLAKIS: Well, thank you
very much. I guess we'll take another break now.
Then we'll go with the last item on the agenda. We'll
take 15 minutes, until 5:20 p.m. Off the record.
(Whereupon, the foregoing matter went off
the record at 5:07 p.m. and went back on
the record at 5:21 p.m.)
CHAIRMAN APOSTOLAKIS: On the record.
We're back in session. Risk-informed inservice
inspection, break exclusion, region piping, that's
what it says here.
MEMBER SHACK: Just to remind everybody
that we've been through this notion of risk-informed
inspection for piping which seemed like a good idea at
the time. Again it was a notion. Now we've learned
about where pipes fail and about the consequences of
failing. In fact we could do better inspections by
looking mostly at regions where we expected to find
degradation of piping and looked hardest at the piping
who's failure had the most severe consequence.
When we approved that it was basically for
piping that was covered by the ordinary Section 11
plants. The augmented inspection regions were not
covered under that one. Now the industry is proposing
to extend that to regions who are augmented and
inspections were required.
One of those is the break exclusion region
where in fact you're supposed to do 100 percent
inspection of the welds. There's a proposal then to
risk-inform that. The staff is going to tell us about
their assessment of that proposal.
MS. KEIM: Okay. I'm Andrea Keim. I'm
going to be handing off this presentation later to
Steve Dinsmore. We have a few other support staff
here to help us answer any questions. Again we're
here to talk about the risk-informed inservice
inspection of an augmented inspection program covering
break exclusion region piping.
A little bit of the background of the PRA
implementation plan included the following guidance
that was developed for devising risk-informed decision
making. There were some general guidance developed
and four application specific guidance in four areas.
They covered technical specifications, inservice
testing, graded quality assurance and inservice
inspection. So far mostly the inservice inspection
has been the most useful for industry.
MEMBER ROSEN: A point of order. I think
our hand out is every other page. At least mine is.
No, there's two on each page. I'm sorry. Human
MS. KEIM: A little bit more on the
regulatory project covering risk-informed inservice
inspection. Again we've developed a regulatory guide
that was issued in September 1998 and a standard
review plan. We've also reviewed topical reports from
Westinghouse Owners Group and an EPRI topical report
covering inservice inspection. Again that covered
ASME code piping from code class 1 and 2.
These were issued back in '98 and '99.
Now what we're looking to do is extend that to a
different augmented inspection.
First I wanted to go also and show the
status of risk-informed ISI reviews. We're proposed
to receive 99 plants wishing to implement a risk-
informed ISI inspection program. We've received 46
through December 2001. We anticipate getting another
42 in 2002. We anticipate an additional 11 post-2002.
The 37 of these submittals that we've
already received used the EPRI methodology. The 13
have used the WOG methodology.
CHAIRMAN APOSTOLAKIS: What's the
difference between the second bullet and the third
MS. KEIM: Not much.
MEMBER KRESS: A few months.
CHAIRMAN APOSTOLAKIS: Major bullet.
MS. KEIM: Yes.
CHAIRMAN APOSTOLAKIS: Number of plants
expected to implement RI-ISI is 99. Number of plants
that have submitted, what is that?
MS. KEIM: That's what we have received so
far to date. So we have 50 applications so far.
CHAIRMAN APOSTOLAKIS: So it's the 46
through 2001 plus a few --
MS. KEIM: A few that we have gotten this
CHAIRMAN APOSTOLAKIS: Okay.
MS. KEIM: We've approved 46 of these
plants. All the ones through 2001.
CHAIRMAN APOSTOLAKIS: I don't understand.
Why do you have to approve them since they are
following methodologies that you have approved?
MS. KEIM: Because these cover ASME code
piping class 1 and 2 which require a submittal for a
CHAIRMAN APOSTOLAKIS: Okay. Even though
they follow an accepted methodology.
MS. KEIM: Yes.
MR. BATEMAN: It's never quite so simple
that they follow an accepted methodology. Each
licensee always has their own little differences they
want to take from the accepted methodology.
CHAIRMAN APOSTOLAKIS: So you have number
of plants that have submitted is 50 or approved.
MS. KEIM: So we have 50 that are
submitted. Our current activities are covering the
Westinghouse Owners Group and EPRI submittals that are
extending this risk-informed ISI methodology to the
augmented inspection of break exclusion region piping.
MEMBER KRESS: Could you give me a little
idea of what break exclusion is about?
MS. KEIM: We're going to get to that.
MEMBER KRESS: Okay.
MS. KEIM: That is coming. Where that's
defined and where those requirements came about.
Primarily our today's presentation will focus on the
EPRI methodology and the EPRI submittal because that
one is farther along in the review process.
A little bit more background on the
objective of ISI, inservice inspection. That's to
identify degraded conditions that are precursors to
pipe failures. I think we're all familiar with that.
For normal ISI, it's referenced in 10 CFR 50.55(a)(g).
That's the requirement that still requires them to
still submit a relief request for the code class
piping. That again references ASME code for the
Now to what everybody's interested in.
The break exclusion region came around from reviews of
general design criteria, number 4 which requires that
structures, systems and components important to safety
be designed to accommodate the effects of a postulated
accidents and include appropriate protection against
the dynamic and environmental effects of postulated
pipe ruptures. The staff has issued a number of
documents that provide criteria for implementing the
above requirements. That covers the Standard Review
Plan chapter 3.6.2 which also includes a staff
technical position MEB 3-1.
The Standard Review Chapter states that
breaks and cracks need not be postulated in break
exclusion region piping provided they meet certain
design and inspection criteria. So from this they
designed these pipes with the different criteria.
They also are required to inspect 100 percent of the
piping welds in these regions.
CHAIRMAN APOSTOLAKIS: I must say it's not
clear to me what a break exclusion region is. What is
MS. KEIM: Well actually it's piping that
is in the vicinity of the containment which is from
the inside isolation valve to the external isolation
CHAIRMAN APOSTOLAKIS: Okay.
MEMBER KRESS: That's piping that you guys
want them to design and inspect so that you can
exclude the possibility that it won't break.
MS. KEIM: Right.
MEMBER ROSEN: That's what exclusion
really means. It doesn't have anything to do with
excluding from the welds or from the inspection.
MEMBER KRESS: Yes. Okay.
MEMBER ROSEN: It has to do with excluding
breaks from the process.
MEMBER KRESS: There are important regions
of piping that you just don't want to break. You want
to be sure.
MS. KEIM: Right.
MEMBER SIEBER: So you have to do 100
percent of every weld.
CHAIRMAN APOSTOLAKIS: This is the only
place where 100 percent inspection takes place.
MEMBER SIEBER: I think that sampling in
CHAIRMAN APOSTOLAKIS: Everywhere else
MS. KEIM: Yes.
MEMBER ROSEN: The code typically requires
I think 25 percent.
MS. KEIM: Yes. For class 1.
CHAIRMAN APOSTOLAKIS: What is MEB?
MS. KEIM: MEB is another acronym that we
use to identify different branches. MEB is the
Mechanical Engineering Branch.
CHAIRMAN APOSTOLAKIS: Oh, okay.
MS. KEIM: That's included in the Standard
Review Plan which is attached into the Chapter 3.6.2.
MEMBER SIEBER: I think the nickname for
the break exclusion region piping is superpipe
because it gets inspected so much.
MS. KEIM: Also because it has additional
MEMBER SIEBER: Right.
CHAIRMAN APOSTOLAKIS: Okay. So now I
understand what a BER is. What is the first sub-
bullet? "Pipe breaks not postulated in BER if
criteria is satisfied including augmented IDI of
piping welds." What does that mean?
MS. KEIM: I think some of that we're
going to cover a little bit later.
CHAIRMAN APOSTOLAKIS: What do you mean
MR. DINSMORE: This is Steve Dinsmore from
MEMBER SIEBER: You don't have to consider
CHAIRMAN APOSTOLAKIS: Oh, if the criteria
is satisfied --
MEMBER SIEBER: You don't have to
postulate a pipe break.
CHAIRMAN APOSTOLAKIS: You do the safety
MEMBER SIEBER: Right.
MR. ALI: This is Syed Ali from the staff.
Maybe I can clarify just a little bit. I think one of
the big differences between the BER and the non-BER is
in the regions breaks had to be postulated and
hardware had to be installed for the effects of those
breaks such as pipe replacing, check shields.
This region which is generally between the
inside and the outside containment isolation valve is
so congested that the staff came up with the criteria
that you don't have to postulate breaks. Therefore
you don't have to install all that hardware provided
a number of conditions can be met.
One of those conditions was 100 percent
inspection. Other conditions were stress below a
certain level, you critique below a certain level.
CHAIRMAN APOSTOLAKIS: Okay. So I guess
if you had written "pipe breaks need not be
postulated" then it would be clearer.
MR. ALI: Right.
CHAIRMAN APOSTOLAKIS: Okay. This is an
interesting situation that you just described because
it goes against the defense in depth philosophy. Does
it not? It says you are shifting everything to
prevention. They say no longer areas. You also do
something to mitigate, to contain the possibility.
But here you just convince yourself that the break
will not happen.
MR. ALI: There are a number of conditions
that have to be satisfied.
MEMBER POWERS: George, you're promptly
committing the cardinal sin of defense in depth. That
is applying it to every damn sub-system in the whole
CHAIRMAN APOSTOLAKIS: That's a cardinal
MEMBER POWERS: Yes.
CHAIRMAN APOSTOLAKIS: So big.
MEMBER POWERS: Yes.
CHAIRMAN APOSTOLAKIS: Jesus. I'm
beginning to become a rationalist again. All right.
That's clear now.
MS. KEIM: So now what the proposal is --
CHAIRMAN APOSTOLAKIS: Well excuse me.
But it doesn't tell me anywhere that the defense in
depth stops at some point. If I read all the
documents, that's a philosophy.
MEMBER POWERS: If you read the exemplary
paper by Sorenson, Powers and Apostolakis, it would
outline this for you.
CHAIRMAN APOSTOLAKIS: That was probably
the part that Apostolakis did right. Okay. Sorry,
Andrea, it's late.
MS. KEIM: That's okay. So what the
proposal is --
CHAIRMAN APOSTOLAKIS: You're doing fine
MS. KEIM: Risk-informed methodology to
select piping elements and welds to be inspected in
lieu of the 100 percent examination. With that I'm
going to hand it over now to Steve Dinsmore.
MR. DINSMORE: Hi. I'm Steve Dinsmore
from the PRA branch. I've been involved in this risk-
informed ISI since pretty much day one or since the
beginning of time, whichever is longer.
CHAIRMAN APOSTOLAKIS: That's where time
MR. DINSMORE: Just to give you a brief
overview that can avoid some confusion later. What we
have is this temporary ISI TR, the original TR. It's
about 200 pages. It has a whole description of a
methodology. It's been approved to use. Except it
was explicitly excluded for use in the break exclusion
Now we have this second topic. This is
what we call the EPRI BER TR. Not topical essentially
identifies tweaks to the original methodology. If
they used them, they can take the original
methodology, tweak it and apply it to the break
This slide is a quick overview of the
different steps in the original methodology and how
they're changed to let the BER program be included.
The first one is scope definition. It's easy. It
used to be excluded. Now we include it.
The consequence evaluation. The BER TR
includes a fairly well defined criteria which should
be used to determine the consequences of ruptures in
these regions. So that's probably the major
Degradation mechanism evaluation. There's
no change. Piping segment definition. There's no
change. Risk categorization. There's no change.
Selection of welds. There's no change.
Risk impact assessment. Essentially what
CHAIRMAN APOSTOLAKIS: Let me understand
that. When you say "no change" to what?
MR. DINSMORE: To the original
CHAIRMAN APOSTOLAKIS: Okay. Not to what
you used to do to the break exclusion area.
MR. DINSMORE: Right. This is to the
CHAIRMAN APOSTOLAKIS: This is to the
MR. DINSMORE: This is to the methodology.
CHAIRMAN APOSTOLAKIS: The methodology.
MEMBER ROSEN: The existing approved
methodology to the 46 plants.
CHAIRMAN APOSTOLAKIS: Now it makes sense.
But did you explain to us what they propose to do to
the exclusion region?
MR. DINSMORE: The tweaks are described
here. This is a quick overview.
CHAIRMAN APOSTOLAKIS: Okay.
MR. DINSMORE: The risk impact assessment.
We had to figure out how to apply the risk criteria
that we'd been using to this region and to the plant
in total. There's also a slide on that.
Monitoring feedback. There's no change to
that. The implementation is another one of the bigger
changes. A lot of these BER programs are only
referenced in the FSAR. You could use 50.59 to make
changes that are referenced in the FSAR.
CHAIRMAN APOSTOLAKIS: What does that mean
implementation if you use 50.59?
MR. DINSMORE: If you do a 50.59
evaluation, you can determine whether you need to make
a submittal for prior review or not. Sometimes they
are in other places, but those plants have their own
If it's only referenced in the FSAR, you
should be able to apply your 50.59 evaluation, use
this methodology and then apply the evaluation. Then
you won't have to come in with a submittal. You can
just make a change.
CHAIRMAN APOSTOLAKIS: How would you apply
50.59 to piping in the exclusion region? Have you
thought of the questions that you're effecting
MR. DINSMORE: Actually the seventh
question is are you --
CHAIRMAN APOSTOLAKIS: I thought the first
question of 50.59 was what you are about to do could
effect initiating events.
MR. DINSMORE: We have our 50.59 person
here specifically for that.
CHAIRMAN APOSTOLAKIS: Okay.
MS. MCKENNA: This is Eileen McKenna from
the NRC Staff. I think you're going to get to it a
little later in the presentation. I think part of the
point that was trying to be made here is that this
part of the program, the BER, is not in 50.55(a). So
you don't have to follow a 50.55(a) review and
Then you look at what is the approval
process if there is one that might apply to this. To
the extent that it's in the FSAR, then it would be
50.59 that would apply to it.
What we're talking about as you'll see a
little bit later is we're really looking at the
methodology by which you select your inspection
locations as changing from the 100 percent inspection
to the risk-informed approach. Then using a
methodology that has been approved through the topical
process. Then you would go through Criteria A which
is the method of evaluation criteria in 50.59.
CHAIRMAN APOSTOLAKIS: But I suspect that
all of this will fail to pass the Criteria 50.59.
Would it not? So you would actually have to come to
MS. MCKENNA: We're approaching it from
looking at it as being the method for determining the
CHAIRMAN APOSTOLAKIS: Right.
MS. MCKENNA: We're looking at it as being
Criteria A method of evaluation. The criteria that's
established is that if you're changing from the method
that you had in your FSAR to another method that has
been approved by the NRC for the intended application,
that is a change that can be done under 50.59.
MR. DINSMORE: You don't have to answer
the other seven questions.
MS. MCKENNA: Right. If it's methodology.
CHAIRMAN APOSTOLAKIS: It's only
methodology here? You say you are reducing the number
MEMBER SHACK: You're changing the method
that you're selecting the inspection.
MR. DINSMORE: Right.
MS. MCKENNA: It has that effect, yes.
MEMBER SIEBER: But that's already been
approved by the staff as a generic methodology. So it
doesn't result in an unreviewed safety question.
CHAIRMAN APOSTOLAKIS: No. But it has
been approved for regional solid of the exclusion
MR. DINSMORE: We're in the process. If
we issue this SE, it will approve it for use
specifically in this region. The SE even says that.
CHAIRMAN APOSTOLAKIS: Let me understand
this. Before this, we were inspecting at how many
MR. DINSMORE: At 100 percent.
CHAIRMAN APOSTOLAKIS: At 100 percent.
Now it's going to be in a smaller number.
MR. DINSMORE: Yes.
CHAIRMAN APOSTOLAKIS: You consider that
a change in method. Is that an unresolved question?
MR. DINSMORE: No. We're reviewing it as
a change in methodology.
CHAIRMAN APOSTOLAKIS: That's what I'm
saying. Why is that so? It doesn't sound to me like
it's a change in method. It's a change in results.
You are inspecting less.
MEMBER ROSEN: I think it's a change in
method that results in a change in results. It's a
change in the methodology.
CHAIRMAN APOSTOLAKIS: Which results
though in a real change which may effect initiating
MR. DINSMORE: But all methodology changes
could result in a real change.
CHAIRMAN APOSTOLAKIS: All?
MR. DINSMORE: I think so.
MEMBER SHACK: The assessment will find
that it doesn't significantly increase your risk.
MEMBER SIEBER: The generic assessment.
MEMBER SHACK: If you follow the
MR. DINSMORE: Yes.
MEMBER ROSEN: George, you're having a bad
MR. ALI: This is Syed Ali from the staff
again. The original EPRI methodology is specifically
excluded from its scope the application to this
region. So what they are doing now is coming with an
addendum to that methodology that says their
methodology can be applied to this region also.
We are reviewing that addendum. If we
approve the addendum then we would have approved the
original methodology but now being applied to this
region also. There are some slight tweaks to the
methodology changes. But it's basically the same
MR. DINSMORE: I think the idea is first
put out this NEI 97.06 that if you use this approved
methodology or an approved methodology for the purpose
it was approved for, you don't have to address those
other questions. The NRC has accepted that as
guidance for using 50.59.
MEMBER KRESS: These pipes penetrate the
containment generally. There's isolation valves on
either side of the containment. If the pipe breaks on
the other side of containment, you've automatically
violated your containment.
MEMBER SIEBER: Not if the valves work.
MEMBER KRESS: Well, the valves are
generally open. You have to close them. Right?
MEMBER SIEBER: Well, they close generally
MEMBER KRESS: What I'm trying to
reconcile is that 1.174 and by extension to the
inservice inspection part of 1.174 there's a
stipulation that you don't violate the defense in
depth principle. It seems to me like this is a
defense in depth consideration. I don't know whether
it violates it or not. It appears to violate it to
me, but I'm not sure.
CHAIRMAN APOSTOLAKIS: No. The 1.174 says
the defense in depth philosophy.
MEMBER KRESS: Well, that's a philosophy.
CHAIRMAN APOSTOLAKIS: So that's a way out
MR. DINSMORE: Well, we include the
spatial effects of the failure of this piping in the
evaluation. Exactly what you gentlemen are talking
about is why we have a much more well defined spatial
effects evaluation process in the TR instead of
leaving it somewhat up to the licensees to develop and
document how they want to address spatial effects.
In this case, we've taken the extra step.
We've put in a good bit more description and criteria
about how they're supposed to do that analysis. But
if the results of the analysis are acceptable
according to all the other criteria that we have, then
MEMBER LEITCH: It seems to me that if you
get past this first issue of the questionable
definition of methodology and you applied the other
seven questions, it would fail. Would it not?
Clearly it would fail.
CHAIRMAN APOSTOLAKIS: Yes. Clearly fail.
MEMBER LEITCH: So if the whole arguement
is hinged on the definition of methodology then you're
not going to get to the others.
CHAIRMAN APOSTOLAKIS: Exactly.
MR. DINSMORE: It might not fail so bad
though because we did look at the questions a bit.
MEMBER SIEBER: My way of looking at it,
and you can correct me because it's a simple way of
looking at it is that if it fails, that means it is an
unreviewed safety question. Then you have to go to
the staff to get approval.
MR. DINSMORE: Right.
MEMBER SIEBER: But they've already
approved when they write this SER the methodology. So
it's no longer an unreviewed safety question. I think
that's what that means. So you don't end up having to
go down that chain of questions to legitimately apply
the methodology because the staff has already approved
the methodology. Is that a way to look at it?
CHAIRMAN APOSTOLAKIS: How does that
compare with the earlier information that Andrea gave
us about the number of plants submitting risk-informed
ISIs and being reviewed by the staff?
MR. DINSMORE: But that's a totally
CHAIRMAN APOSTOLAKIS: You are reviewing
the process that you have.
MR. DINSMORE: If you want to get a relief
from applying, that's going to be Section 11
inspections, you have to come in to the staff and
MEMBER SIEBER: An exemption. Right?
MR. DINSMORE: It's a relief request.
CHAIRMAN APOSTOLAKIS: So that doesn't
MEMBER SIEBER: From 50.55(a).
MR. DINSMORE: Yes.
MEMBER SIEBER: Right.
MR. ALI: Again, it's Syed Ali. I just
want to add something on that also. In the original
program, they were specifically going below the
inspections that are required by ASME 11. So they had
to come in for a relief. Here in this region there's
ASME piping and there's non-ASME piping.
For ASME piping that is in this region,
they would have to maintain at least the ASME 11
inspections in order to apply 50.59 and not come for
a relief. If they go below the ASME 11 then it will
go into the same kind of a treatment as the rest of
the plant. They will have to come in with a relief
request. So the floor is still the ASME 11 in this
region for the 50.59 process to be applicable.
MEMBER LEITCH: The actual floor is about
a 10 percent inspection.
MR. ALI: Well, it's 25 percent for ASME
class 1 and about 7 and a half for ASME class 2.
That's the ASME level in the floor.
CHAIRMAN APOSTOLAKIS: Well, I guess if
it's clear to all the members, we can go ahead.
MEMBER LEITCH: Just one more question.
Is that 25 percent per 10 year interval?
MR. ALI: The 25 percent per each 10 year
MEMBER LEITCH: Thank you.
MR. DINSMORE: Okay. Now we move to the
consequences. We'll explain a little bit again the
difference between BER piping and non-BER piping. The
non-BER piping had pipe failure postulated during the
design and evaluated using these SRP guidelines. The
mitigative hardware was added as needed. I guess we
already talked about this a lot.
In the BER piping, the pipe failures were
not postulated and the mitigative devices were not
constructed. So essentially when we did the original
risk-informed ISI we were looking at the non-BER
piping because that's the only place they were
changing inspections. We were more or less crediting
this SRP analysis out there. They had done this SRP
analysis one time already. So these guys can do their
PRA realistic analysis on top of that.
Now inside the BER piping, we don't have
that fall back. It's just whatever is there. That's
the reason in the EPRI BER TR, we essentially said you
can use the SRP guidelines or criteria or somewhat
more conservative. They can use somewhat more
conservative because it's not as sensitive. What the
result is, is that the segment goes into higher
medium. The result of that is they do 10 percent or
25 percent of inspection.
It's not that they have to build in all
this equipment. So I think the two pilots were
somewhat conservative because it didn't hurt them that
much to be conservative.
MEMBER LEITCH: Once again I just want to
make sure I understand this. Under the BER piping,
the reason that pipe failures were not postulated is
because this particular piping was very conservatively
designed and because we were going to do 100 percent
MR. DINSMORE: Right.
MEMBER LEITCH: Not because it's not
important. In fact it's to the contrary. It's very
CHAIRMAN APOSTOLAKIS: Yes. I think that
was the reason.
MEMBER LEITCH: These are high energy pipe
MEMBER SIEBER: Some are, some aren't.
MR. DINSMORE: We're working on it.
MEMBER LEITCH: It's main stage. It's
feedwater. Isn't it?
MEMBER SIEBER: Sure.
MR. SULLIVAN: This is Ted Sullivan. I'd
like to add a little perspective. I think Dr. Kress
really hit upon it earlier. You couldn't postulate a
break in these areas. If you postulated a break for
example in a boiler and coupled with it the single
failure of the isolation valve --
MEMBER KRESS: Or leaking at that.
MR. SULLIVAN: You violate containment.
So it's really an outgrowth of that.
MEMBER LEITCH: All the more reason for
inspection though as I say. I agreed you couldn't
postulate a break. But I just don't understand the
logic of this. If you couldn't postulate a break,
it's not because it's not a problem. It's a big
problem. So all the more reason to inspect.
MR. SULLIVAN: I don't disagree with you.
There are some representatives of industry here if
they want to add to what I'm saying, industry's view
was that these are fairly high radiation areas. They
really have not been finding anything to speak of or
much to speak of from doing these inspections.
They've done thousands and thousands of
weld inspections. The performance of this piping is
very good. So what they proposed and we've been
reviewing is a concept of focusing inspections
basically for cause. Where is the degradation
expected to have some potential to occur? Let's
inspect in those regions and couple that with regions
where the consequences would be high rather than
forcing the licensees to continue to do 100 percent in
a lot of area where they really can't even identify a
potential degradation mechanism.
CHAIRMAN APOSTOLAKIS: It's a performance
based initiative. Because they haven't found anything
in many inspections, they say why should we keep doing
MR. DINSMORE: Why should we keep doing
CHAIRMAN APOSTOLAKIS: Yes.
MR. DINSMORE: I think that's right.
MEMBER KRESS: That's a different
arguement than we've been hearing.
CHAIRMAN APOSTOLAKIS: It's a very
MEMBER KRESS: It's a more persuasive
CHAIRMAN APOSTOLAKIS: In fact, it's much
more persuasive, yes. This is not risk-informed
stuff. This is performance based.
MEMBER POWERS: In fact, it has to be a
risk-uninformed thing. I mean, WASH 1400, NUREG 1150
all tell us if you want to get yourself in real
trouble you have a bypass accident.
MEMBER KRESS: That's exactly right.
CHAIRMAN APOSTOLAKIS: Yes.
MEMBER POWERS: So if you bust these
pipes, you have a bypass accident. Anything that
degrades your confidence in these, would have to be a
risk-uninformed activity, inverse of risk-informed.
CHAIRMAN APOSTOLAKIS: You would never
pass 50.59. You just don't.
MS. KEIM: We have someone from industry
that would like to speak.
MEMBER KRESS: You might if you postulate
that the inspections aren't doing you any good because
they never found anything.
CHAIRMAN APOSTOLAKIS: No. The
inspections are always doing something good. They
never found anything. That's strong evidence that the
uncertainty has been reviewed significantly. Right?
MR. DINSMORE: Yes, sir.
MR. BALKEY: This is Ken Balkey from
Westinghouse. I'm working with our team on the
Westinghouse Owners Group methodology. They fall as
the same procedure in the EPRI method as well.
To add to Ted Sullivan's comments, when we
did the risk-informed ISI work from the original
topicals a few years ago, we learned a lot. That ASME
code had 25 percent and 10 percent. There was a
history of how they came up with that. It just says
there's a history is why there's 100 percent here.
To do these exams, it's not simply just go
out. They are in congested areas and high radiation
areas. There are only so many examiners to go around
as well too.
When we did the risk-informed ISI process
with either method to do the Section 11 exams, we feel
that we've done a real service. Even though we're
doing a smaller population, we are in the process of
moving the exams to the areas of active degradation.
Therefore making very good use of the utility's
resources in doing those examinations.
We knew about this area when we did the
original program. We even had a lot of discussion
with the NRC of could we include this, even in the
original topical three or four years ago.
The staff felt and industry agreed that we
have to take one step at a time here. It was enough
of an issue to get through the ASME Section 11 exams
and working through a regulatory process with the
relief as Andrea said in terms of utilities making
submittals and getting approval for a relief request.
The industry now said we should be able to
take the same knowledge we just gained from that
program, and apply it to the high energy line break
exclusion region. We're not taking exams down to
zero. I think we're trying to support what Dr. Kress
said. Do you really 100 percent to give you assurance
that the integrity is good within this piping?
If it was easy to do, we wouldn't be here.
They are difficult exams to do. So we're saying can
we do a smaller population and still get the same
level of assurance in this region like was done in the
same piping for the Section 11 program. All the
questions in terms of if it breaks, would it take out
other areas or what it's effect is from a PRA, we
still have to look at that. There are areas where we
will not remove examinations because the PRA indicates
them a consequence. You really still need to do a
number of exams in that area.
In summary, what we are trying to do is
really take what we learned on the original
application and now extending it to this for the 100
percent. It does free up the resources to really get
at some other degradation issues we're dealing with in
MEMBER KRESS: Let me ask you a question.
MR. BALKEY: Sure.
MEMBER KRESS: When you say 25 percent of
piping instead of 100 percent, let's just pick a
MR. BALKEY: Okay.
MEMBER KRESS: Does that mean you
eventually inspect all the piping? You would only
spread it out in time a little more.
MR. BALKEY: That's a good question. The
original concept for the 25 percent came from 30 years
ago. You do 25 percent in the first 10 years, 25
percent in the second and so forth. So over the life
of the plant, you do 100 percent.
But guess what? As plants operated, folks
said we did the first 25 percent and we really should
go back and take a look to see if anything changed.
If you go another 25, going back to a location you
just did 10 years ago and you get a different signal
from your ultrasonic, you know degradation is under
way. So you're better off getting to a smaller
population and really monitoring the degradation
closer than trying to do it all one at a time.
MEMBER KRESS: You could do a combination
of those two.
MR. BALKEY: Right. In this application,
the intent would be you'd have a smaller population.
But they are the areas that you would expect
degradation and of course areas of high consequence.
You would go back to those areas each ten year
CHAIRMAN APOSTOLAKIS: So you are always
inspecting the same 25 percent?
MR. BALKEY: Yes. Or whatever the percent
ends up being in this region. Yes. You would go back
to the same. But the program also as part of its
update if you find something whether it's in the
Section 11 program or if it's in a break exclusion
region, you may have to expand your sample. Not may,
There's a sampling scheme that if you find
something in that outage, you have another population
that sees it now somewhere else you weren't
inspecting. If you find something there, then you're
doing 100 percent of your area. So the process allows
you to get to 100 percent if you start finding
degradation in the sample that you're doing.
MEMBER LEITCH: How big an issue is ease
of inspection in determining which 25 percent?
MR. BALKEY: I would actually ask one of
my colleagues here who is an examiner at his plant.
Dave, do you want to speak to the difficulty in
getting to some of the locations.
MEMBER LEITCH: I know some of the
locations are very difficult. My question was really
how do pick your 25 percent.
CHAIRMAN APOSTOLAKIS: Do you pick them
MR. BALKEY: Right now Dave has to do 100
percent of the exams at his plant.
MEMBER LEITCH: I know some of them are
really hard. What I'm saying is when you determine
your 25 percent sample view, do you eliminate the real
MR. BALKEY: No. I can give you an
example. Turkey Point is one of the plants that's
been submitted not for break exclusion but in the
original Section 11. We looked at their risk-informed
ISI. We indicated in their surge line for their
operational experience. They had to do 100 percent of
the surge line.
That was a very difficult finding because
they had to go back and spec underneath the
pressurizer. It's a very high radiation. But we said
you have to examine it because of the information you
had. We would use the same philosophy. The same
philosophy would apply here.
Just because it's hard to get to is not
the reason you would drop it out. If you find it's an
area of degradation and your PRAs telling you that
it's really important if it fails, unfortunately
you're going to have to go in and make the effort to
do the examination.
MEMBER KRESS: What is the risk criterion?
How do you establish whether the one pipe section is
more risky than another one? Is it because of
equipment that may be around it?
MR. BALKEY: Yes.
MEMBER KRESS: Is it the size of the pipe
or the flow rates or a combination?
MR. BALKEY: It's a combination of the
temperatures and pressures. That's part of what
Stephen was talking about and the consequence
evaluation on this slide here. One has to go in and
look a lot more carefully. You look at your pipe whip
for jet impingement effects and also flooding effects
on the electrical equipment if there's anything that
happens to be nearby.
MEMBER KRESS: That's how you decide the
MR. BALKEY: Yes. That's part of the
MEMBER ROSEN: The functions of the piping
MR. BALKEY: As well as the functions of
the piping. We usually break it in to a direct
consequence to address the functions. Then the
indirect effects are the pipe whip and jet impingement
of pipes whipping and taking out other equipment
nearby. That has to be done as part of the process.
MEMBER KRESS: Thank you.
MR. DINSMORE: Okay. I'm not quite sure
this is resounded. We do use some risk information in
the process. So that they don't have to come in with
a submittal, you have to keep that in the back of your
mind, the quality of the PRA needs to be the same
acceptable quality as for risk informed ISIs since
it's pretty much the same process.
MEMBER SHACK: Can he do this without
having a risk-informed ISI program for his Section 11
MR. DINSMORE: They can apply this to the
BER region without doing a risk-informed ISI.
MEMBER SIEBER: Right.
MR. DINSMORE: Within the BER region then
as Syed was saying earlier --
MEMBER SHACK: Could you do it with 50.59?
MR. DINSMORE: Yes. But you couldn't
change the ASME Section 11 inspections if there are
any in this BER region. You could only change the BER
MEMBER ROSEN: Do you expect anybody to
actually do that, someone who hasn't done the basic
MR. DINSMORE: I have Pat O'Regon back
there nodding. He's from industry. So I have a
feeling he knows.
MR. O'REGON: I'm Pat O'Regon from EPRI.
The answer is yes. There are several plants that
would like to implement BER only. In particular a
couple of BWRs will be implementing BWR VHP 75 on the
stainless steel piping and risk-informed BER on the
carbon steel piping.
MEMBER POWERS: How would the quality of
your PRA affect the conclusion that seems to be robust
trough all PRAs that containment bypass accidents are
very hazardous accidents?
MR. DINSMORE: Well, they would assign a
pretty high conditional core damage probability or a
conditional large early release probability to those
segments which would contribute to those sequences.
Then it would be up to whatever degradation mechanisms
are in those segments.
If there's no degradation mechanism and a
very low failure probability then those segments would
be lower risk. If there's some degradation mechanism
and a high probability, there would be a higher risk.
MEMBER LEITCH: Do we have any idea how
much man-rem per plant per year is attributed to the
execution of this program as it now stands? In other
words, what's the man-rem saving per plant per year
estimated to be?
MR. DINSMORE: Maybe industry would know.
I don't. I guess not. No.
MEMBER ROSEN: Another way to look at that
same question is what's the percentage reduction in
the program that would come out of this. How big an
effect is it on the remaining overall program? Can
you give us any feel for that?
MR. DINSMORE: The EPRI TR says that if
you get below 10 percent, you need to provide a good
explanation of the design features in your plant which
supports finding that you have to inspect less than 10
percent of the welds in this region.
MEMBER ROSEN: That's not exactly the
question. That's not the answer to the question that
I thought I asked.
The question is let's say before you have
a start at this you were inspecting 1,000 welds in the
10 year period. Then you go to risk-informed ISI.
Now you're only inspecting 350 welds. You knocked out
two-thirds of them which I think is the number I
So you're down to 350 welds in the 10 year
period. Now can go to break exclusion piping and
knock that out. Now you're inspecting not 350 but
only 175 or 300? I'm trying to get a feel for the
MR. DINSMORE: This is one of the pilots
that we didn't review by the way we just looked at it.
If you had 135 welds, one of them went down to 20 for
example. So that's about 11 percent. The other one
went down to 3 percent.
MEMBER ROSEN: Wait a minute. You said
135 and you went to 20.
MR. DINSMORE: Yes.
MEMBER ROSEN: That's a reduction of
almost 90 percent. Right?
MR. DINSMORE: That's because we're
starting with 100 percent. You see if you start with
MEMBER ROSEN: Out of 135 welds you're
total example was the BER scope.
MS. KEIM: Yes.
MR. DINSMORE: Right. You inspect them
all to start with. In the ASME class 1, you were
going from 25 percent down. Here you're going from
100 percent down.
MEMBER ROSEN: So basically it's a very
large reduction in the BER scope.
MR. DINSMORE: It can be.
MEMBER KRESS: When you do the risk
assessment to calculate the change in LERF for
example, can you check it along with the absolute
LERF? If you have more than one unit on the side, are
you going to add the LERFs together?
MR. DINSMORE: We don't have process to
deal with that. If you had more than one unit on the
site I think what happens is if you add the two
together, the relative increase would be the same. We
don't really apply these criteria.
MEMBER KRESS: No. You have an absolute
LERF then you have a Delta LERF. The Delta LERF stays
the same. If you do it to one unit only, the Delta
LERF is for the unit. But the LERF is a LERF for the
site. It ought to be the sum of all the plants that
are on the site. That's a glitch or a short coming of
1.174 that I've been trying to get fixed. That's why
I ask the question every time.
MR. DINSMORE: We haven't fixed it in this
CHAIRMAN APOSTOLAKIS: A straightforward
answer. You'll wait until 1.174 is fixed first I
MR. DINSMORE: Right.
CHAIRMAN APOSTOLAKIS: Okay. Let's move
on. Go to 11.
MR. DINSMORE: This is 11.
CHAIRMAN APOSTOLAKIS: This is 11?
MR. DINSMORE: I have a different
CHAIRMAN APOSTOLAKIS: So what number do
you have for this one?
MR. DINSMORE: I have 11 for the other
one. We took one out. We put one together.
CHAIRMAN APOSTOLAKIS: We discussed this.
MR. DINSMORE: Yes. We discussed this in
the beginning. We can just maybe even skip it.
CHAIRMAN APOSTOLAKIS: Yes.
MEMBER KRESS: This is the final
conclusion you have.
MR. DINSMORE: Right.
CHAIRMAN APOSTOLAKIS: Now let me
understand the first bullet. As I recall Regulatory
Guide 1.174 as we said earlier today has a beautiful
discussion of uncertainties incompleteness, models.
Are you guys doing any of that?
MR. DINSMORE: Those are included mostly
in the system level guidelines. We don't allow them
to for example take a bad weld in a dangerous system
and start inspecting that. They get a big plus risk
from that and use that to stop inspection many welds
in other systems. We don't believe that the numbers
support those type of large shuffling of risk.
CHAIRMAN APOSTOLAKIS: When you say the
basic acceptable quality of the PRA is the same as the
risk-informed ISI, so you have already approved 46.
MR. DINSMORE: Right.
CHAIRMAN APOSTOLAKIS: These are 46
submittals. You are now reviewing four.
MR. DINSMORE: There are five. We got one
CHAIRMAN APOSTOLAKIS: Five. Okay. So
you are really busy then. When you reviewed the 46,
did you look at issues like model uncertainty and
incompleteness? My impression is that nobody's doing
uncertainty analysis anymore.
MR. DINSMORE: What we required for the
risk-informed ISI is that the licensee go back and
look at all the negative comments made by the research
review and the peer review process, the BWRG. They
evaluate all these comments and make sure that either
they don't affect the results of the ISI analysis or
that they incorporate somehow the comment into the
CHAIRMAN APOSTOLAKIS: But what if the PRA
has not done an uncertainty analysis at all? We were
told last month that asking for uncertainty analysis
means killing the program because nobody does it. So
I don't know how you conform with Regulatory Guide
1.174 if you don't do that.
MR. DINSMORE: Well, I think 1.174 says
that if you do a reasonably conservative analysis or
if you do something that you think is a bounding
analysis, you can address uncertainty in that way.
CHAIRMAN APOSTOLAKIS: I thought 1.174
really looked at all these uncertainties. How do you
know something is conservative if you don't understand
the uncertainties? Don't you have to understand what
is uncertain first before you say now what I'm doing
MR. DINSMORE: It's also that the
uncertainties in the pipe failure probabilities are
probably much larger than in the PRA.
CHAIRMAN APOSTOLAKIS: That's also true.
So how are these uncertainties handled?
MR. DINSMORE: We handle them by having
different criteria. Again this risk level criteria,
we don't allow them to move risk around between
systems very much. The risk level criteria is you
can't get more than a 10 to the minus 7th increase in
So it's a factor of 10 below the plant
level criteria. It's regardless if you only have
three systems. Then the plant level is going to be 3
times 10 to the minus 7th and not 1 times 10 to the
We've tried to deal with uncertainty by
putting in this backstop of what you can move and what
you can't move. We've actually done it in the BER
program as well. We've taken the BER program by
itself. They have to apply the same criteria to the
In other words, every system within the
BER program they cannot increase the CDF by more than
10 to the minus 7th per year. For the total BER
program although it's not really useful, they couldn't
increase the CDF by 10 to the minus 6th. Then if they
put it together with the risk-informed ISI, they have
to apply those criteria to the total change as well.
So there's a couple of steps in the
criteria. That's the main --
CHAIRMAN APOSTOLAKIS: What you're saying
is that they don't need to do the uncertainty analysis
because the criteria we have established have allowed
for the uncertainties that you may have which is a new
interpretation of 1.174.
MR. DINSMORE: We used it in the basic
CHAIRMAN APOSTOLAKIS: I understand that
you have used it. Okay. Let's go on.
MEMBER ROSEN: I have a question about
those few licensees that might come in and just want
the BER program. Would they have to come and get
approval or could they completely avoid any review,
just do 50.59 and off they go?
MR. DINSMORE: If they don't change the
ASME Section 11 or any other licensing basis, they
could. Yes. They would not have to come in. They
could just do it. They have to put it in their yearly
report that they've done it.
MEMBER ROSEN: So the staff would never
get a chance to talk to them about their PRA and how
good it is or any of those things.
MR. DINSMORE: No. But they're required
to do the same analysis which we've been requiring
them to do for risk-informed ISI which is to take all
the comments and everything and document it. The
documentation requirements to be maintained onsite are
the same if they just do the BER as they are if they
do a risk-informed ISI. It's just that they don't
send us anything.
MEMBER ROSEN: That part troubles me quite
a bit. At least in the basic risk-informed ISI
program licensees came in with the EPRI method. The
staff reviewed what they wanted to do, looked at their
PRA and their peer review and had some handle on it.
With the small number of licensees I'm told who would
never have to go through that process, could use 50.59
and change the break exclusion region piping sample
size without any staff at all of anything except after
MR. DINSMORE: We do very limited reviews
of the PRA. Really all we ask for is who said what
bad things about your PRA and what did you do about
them. We look at what they do. They usually give a
reason. If somebody said you had a bad human error,
they say we applied these new methodologies and so on.
We've occasionally gone back and said
that's not enough, please give us more. But that's
not often. These guys if they just do the BER,
they're still going to have to do the same process.
If we go out and eventually audit one of these guys
and they didn't do it or they didn't document it, then
I'm not sure what we'll do. But we'll do something.
MEMBER LEITCH: I'm still a little bit
confused with the approval of this proposal. What
determines whether it's 25 percent or 10 percent?
MR. DINSMORE: Well, 25 percent of the
welds in high safety significant segments have to be
inspected. The 10 percent of the welds in medium
safety significant segments have to be inspected.
That's a hold over from the old methodology.
MEMBER LEITCH: So the determination is
based on whether it's high or medium safety.
MR. DINSMORE: Right.
MEMBER LEITCH: There are no low safety
significant systems in this set, I guess.
MR. DINSMORE: There are. You do not have
to inspect those.
MEMBER LEITCH: Are they inspected now?
MR. DINSMORE: On the BER everything is
MEMBER LEITCH: So there are some where
there are low safety significant that you would go
from 100 percent inspection to zero inspection. Is
that what I understood you to say?
MR. DINSMORE: That's correct.
CHAIRMAN APOSTOLAKIS: I'm missing
something here. Has anybody objected to that? Why
are they reluctant to do that when we talk about
option 2? The low risk significant SSC still impose
some requirements. They are unwilling to lump them
with non-risk significant. Yet for pipes it seems
that they're willing to go to zero.
MR. DINSMORE: Well we did a bounding
MR. O'REGON: Pat O'Regon from EPRI again.
We looked at three plants, two sites out of the BER
application. We did find some low safety significant
locations. But they were as a result of the utility
conservatively applying the BER rules. They extended
piping beyond where they would have had to if they
held strictly to the SRP requirements.
So that's why they fell as low safety
significant. They weren't big pipes that created big
holes in containments. As Steve mentioned, the high,
medium or low are from the EPRI TR ISI, the base case
methodology where we rank things as high, medium or
low. We just kept that consistent when we extended it
to the BER programs.
CHAIRMAN APOSTOLAKIS: All right.
MR. DINSMORE: The methodology is
consistent with the EPRI Topical Report. The
inconsistencies are the things we've explained to you.
The changes to BER program as described in the FSAR
may be made under 10 CFR 50.59. Inspections within
the BER program to change that come from other
regulatory requirements need to be changed according
to how you change the other regulatory requirements.
MEMBER SHACK: Anything else?
CHAIRMAN APOSTOLAKIS: No letter. Right?
No request for a letter.
MEMBER SHACK: There's no request for a
CHAIRMAN APOSTOLAKIS: So there will never
be a letter.
MEMBER SHACK: Not unless we decide one.
They're not requesting one. We can discuss whether we
want to send one.
CHAIRMAN APOSTOLAKIS: Okay. Anymore
questions to the lady and the gentleman?
MEMBER POWERS: Well, there's another
point to be made. That is it is true enough that
bypass accidents are risk dominant. But bypass
accidents initiated by failure of this particular
piping don't show up in the PRA at all. They never
MEMBER SHACK: There is one difference
though. When we did the original in service risk-
informed, you could make the argument that you were in
fact approving safety. Obviously you might have been
looking at fewer welds. But you were looking at the
more important welds. So you could make an argument
that your Delta CDF could have gone down. In this
case, it might be a small change but it has to go.
MR. DINSMORE: That's part of the reasons
that we applied the criteria specifically to the BER
as well. That was the best way we could think of to
deal with that.
MEMBER POWERS: But you still have this
MEMBER SHACK: Right.
CHAIRMAN APOSTOLAKIS: That's really a
MEMBER SHACK: That's incorporated in the
argument that you're going to apply all that good
performance to assign most of this stuff to a low
probability of failure. You don't want to give them
double credit for that. They're going to take that
credit already. Again, it's a very small change in
LERF for perhaps ALARA reasons.
CHAIRMAN APOSTOLAKIS: Isn't there a table
that the regional methodology has when they have the
risk significant of a piece of piping? Then they have
a susceptibility. That's where the performance comes.
MEMBER SHACK: That table still applies.
CHAIRMAN APOSTOLAKIS: The performance
MEMBER SHACK: Yes.
CHAIRMAN APOSTOLAKIS: Is this for
everything or at Westinghouse?
MEMBER SHACK: Yes. It's everything.
MR. DINSMORE: I wouldn't bring
Westinghouse to EPRI SE.
CHAIRMAN APOSTOLAKIS: No. I mean, they
have something similar I think.
MR. DINSMORE: They have something
similar, yes. But you can see here if it's a really
high consequence in this methodology, it would end up
in a medium box even with no degradation mechanisms.
CHAIRMAN APOSTOLAKIS: Medium means?
MR. DINSMORE: The 10 percent.
CHAIRMAN APOSTOLAKIS: My concern is
bigger than what you're doing. I think that the
implementation of Regulatory Guide 1.174 has drifted
away from what the guideline is saying. It has a lot
to do with you. Are there anymore questions for Steve
and Andrea? Well, thank you very much.
MR. DINSMORE: Thank you.
CHAIRMAN APOSTOLAKIS: I would ask the
members to stay here for a few more minutes. Maybe we
can discuss things among ourselves.
Shall we take a five minute break? Eight
minutes. We don't need transcription anymore. Thank
you. Off the record.
(Whereupon, the above-entitled matter
concluded at 6:21 p.m.
Page Last Reviewed/Updated Monday, July 18, 2016