Materials and Metallurgy - March 25, 1999
UNITED STATES OF AMERICA
NUCLEAR REGULATORY COMMISSION
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
***
MEETING: MATERIALS AND METALLURGY
***
USNRC, ACRS/ACNW
11545 Rockville Pike, Room T-2B1
Rockville, Maryland
Thursday, March 25, 1999
The subcommittee met pursuant to notice, at 8:30 a.m.
MEMBERS PRESENT:
WILLIAM SHACK, Chairman, ACRS
MARIO H. FONTANA, Member, ACRS
THOMAS S. KRESS, Member, ACRS
ROBERT SEALE, Member, ACRS
. P R O C E E D I N G S
[8:30 a.m]
DR. SHACK: The meeting will now come to order. This is the
second day of the meeting of the ACRS Subcommittee on Materials and
Metallurgy.
I am William Shack, Chairman of the Subcommittee.
The ACRS Members in attendance are:
Robert Seale, Mario Fontana, and Thomas Kress.
The purpose of this portion of the meeting is to meet with
representatives of the NRC staff to discuss NRC activities related to
the status of resolution of the differences between the staff and ASME
regarding the use of the 1994 addenda to the ASME Section III Code for
Class 1, 2 and 3 piping systems; the proposed revision to 10 CFR 50.55a,
"Codes and Standards"; and regulatory research activities associated
with 10 CFR 50.61, the pressurized thermal shock rule. The Subcommittee
will gather information, analyze relevant issues and facts, and
formulate proposed positions and actions as appropriate for deliberation
by the full Committee.
Noel Dudley is the Cognizant ACRS Staff Engineer for this
meeting.
The rules for participation in today's meeting have been
announced as part of the notice of this meeting previously published in
the Federal Register on March 8, 1999.
A transcript of the meeting is being kept and will be made
available as stated in the Federal Register Notice. It is requested
that the speakers first identify themselves and speak with sufficient
clarity and volume so that they can be readily heard.
We have received no written comments or requests for time to
make oral statements from members of the public.
Our opening discussion will be by Ms. Lee on some responses
to GL-92-01, Rev. l.
MS. LEE: My name is Andrea Lee. I am going to be
discussing reactor pressure vessel integrity issues.
[Slides.]
MS. LEE: First I will give a brief background on Generic
Letter 92-01, Supplement 1.
Then I will note the Owners Group activities that have gone
on because of this effort, and the staff's decision to issue requests
for additional information.
Then I will update you on the status of the staff's review
and also the update of the reactor vessel integrity database.
I will also comment on our schedule for closeout of Generic
Letter 92-01.
Then we will switch to a different topic, which is the
status of changes to the American Society of Mechanical Engineers
Appendix G.
The original Generic Letter 92-01 was issued as a result of
Yankee Rowe to avoid the surprises that came out of that review.
The original response to the generic letter and also the
Palisades issues, which I believe most people are familiar with,
provided some additional surprises. So the current program that is
going on now is addressing those issues.
Supplement 1 to the generic letter was issued in May 1995
and it asked that any and all data be collected by licensees so that
they could assess the impact on their integrity evaluations. We also
requested that they submit those reports in detail to the staff.
The purpose was to have an integrated and complete database
to assess all the U.S. commercial nuclear power plants.
Supplement 1 to the generic letter was closed out
administratively in mid-1996, but there was a clear statement in it to
address the ongoing Owners Group activities.
There were several reports submitted to the staff from the
Owners Group. There were two from the Combustion Engineering Owners
Group, one from the Babcock & Wilcox Owners Group, which was a result of
an audit of the Framatome fabrication records, and that audit occurred
in May 1997. There was also a report from the BWR Owners Group.
The staff reviewed the reports and had meetings with the
industry in public to discuss chemistry and surveillance data. Those
meetings occurred in November of 1997 and February of 1998.
From our review, the staff determined that the Owners Group
reports had a lot of information and we felt that they weren't being
reviewed in a timely manner by the licensees. That prompted the staff
to issue RAI packages to PWRs and BWRs. That occurred March through
August of 1998.
DR. SEALE: Could I ask you why the Westinghouse Owners
Group was apparently bulletproof on this inquiry? I noticed you didn't
get any responses from the Westinghouse Owners Group.
MR. WICHMAN: When we are talking about Owners Group, we are
talking about people that fabricate vessels. Westinghouse never
fabricated their vessels. It was either CE or B&W, and in some cases
foreign fabricators.
DR. SEALE: Thank you.
MS. LEE: In terms of the status of our review, all the BWR
reviews have been completed with the exception of Monticello, and that
is currently being worked because there is new data that was submitted
in the pressure-temperature limits report. Most of the PWR reviews are
complete and all are in the final stages right now.
Prairie Island 1 & 2 and Cook 1 have indicated they will be
making submittals by summer of 1999.
In terms of background of the reactor vessel integrity
database, it summarizes the properties of all RPV materials beltline
information for all U.S. commercial nuclear power plants.
It has been switched. The original version was the
classical version. It has been transferred to Microsoft Access version.
The new database diskettes will be posted on the World Wide Web for
downloading once the effort is complete.
There will also be an administrative letter issued to
announce the availability of the new database.
DR. SHACK: I thought this database was going to be merged
with the EPRI database.
MS. LEE: We had at least two meetings on that in the past.
From the feedback we got back, the industry, for various reasons, was
not interested in taking over the database. So it will remain a staff
database.
DR. SHACK: How often do the people review their
pressure-temperature limits? If their vessel embrittles, is that
something that is reviewed on a yearly basis?
MS. LEE: No, it's not yearly.
MR. WICHMAN: It's as needed. It depends on how far their
P-T limits were originally projected to go out to.
DR. SHACK: You give them what, some sort of upper shelf
toughness that they work to?
MR. WICHMAN: Appendix G is their guidance for developing
P-T limits. Staff has -- what it is, 5.2.3?
MS. LEE: MGEB 5.3.2 explains it.
MR. WICHMAN: Surveillance data comes into the picture too
when they pull capsules. All of these things dictate the need for the
revision.
DR. SHACK: Appendix G tells them what they are supposed to
do?
MR. WICHMAN: Appendix G forms the basis.
MS. LEE: In terms of schedule for closeout, because all the
data that came out of this effort is being used to update the database,
the target date for completion of both Generic Letter 92-01 effort and
the closely related effort of updating the RVID is the summer of 1999.
Also, NUREG-1511, Supplement 2 will be issued to document
our closeout.
DR. SHACK: Yesterday we heard that there was only one plant
that was going to have trouble with the PTS limit. How many plants will
have trouble with upper shelf toughness?
MS. LEE: There are none.
MR. WICHMAN: We are revisiting that in license renewal
space.
DR. SHACK: But everybody is okay to 40 years?
MR. WICHMAN: Everybody is okay for 40 years; everybody was
either above 50 foot-pounds or had submitted an equivalent margin
analysis in accordance with Appendix K of the code.
DR. SHACK: How many people had to submit the Appendix K
approach?
MR. WICHMAN: I think basically everybody by owners groups
or topical reports, as I recall, did an equivalent margin analysis or
bounding analysis, but a lot of people were above 50 foot-pounds at the
end of 40 years. Again, it's an item that we revisit in license renewal
space.
MS. LEE: Moving to the last topic, the American Society of
Mechanical Engineers has recently worked to revise ASME Appendix G.
The scope of the change is to use K1c, the lower bound on
static fracture toughness, instead of K1a, which is the lower bound on
static, dynamic and crack arrest, in development of the P-T curves.
An important point to remember is that all the other
conservatisms remain the same in the P-T curves. For example, using the
1/4T flaw or factor of 2 on pressure stress.
The change was initiated for a number of reasons. One
important reason is that there was a significant increase in data since
the curve was originally developed in 1974. It improves overall plant
safety in terms of opening up an operating window and reducing
challenges to LTOP.
The code case that describes these changes, N-626, and the
change to ASME Appendix G was approved by Section XI.
We have no licensee submittals to date, but any submittals
that come in need to be an exemption because they are an alternative to
10 CFR 50 Appendix G.
That's all I have.
DR. SEALE: There is an interesting use of words in here. I
take it that the K1c is a lower toughness level than K1a.
MS. LEE: No.
MR. WICHMAN: It's higher.
DR. SHACK: That's why you get relief.
MR. WICHMAN: Yes. It's relief for the industry.
DR. SEALE: Okay. But it's safety in the sense that you
moved the point on the curve and made it legitimate to move up in your
operational plan.
DR. KRESS: It's actually a relaxation.
DR. SEALE: It's safety because of the operational
difficulty of start-up and that sort of thing. Not absolutely safety in
another sense.
DR. SHACK: We had some discussions yesterday. We do
pressurized thermal shock probabilistically, at least a pseudo-realistic
flaw distribution. Has anybody ever proposed doing this in a similar
manner where you use a realistic flaw distribution?
MR. WICHMAN: No. In fact, this is strictly deterministic
and, if you wish, almost a flaw tolerance evaluation, because you've
have got this 1/4T flaw. You are not relying on ISI, for example. That
could change in the future with advent of Appendix VIII and other
things. I think there is more bang for the buck in this change than
there is in perhaps changing the 1/4T flaw to something less that we
could buy reliably. Then there has been talk of reducing the factor of
2 on pressure.
All of these things would help, I guess. I think there is a
very considerable relief in going to K1c. The French have been doing
this for many years in the development of their pressure-temperature
limits using K1c. It is really more appropriate, because heat-up is
essentially a static process anyway as opposed to a dynamic process,
K1a.
The concern in using K1a was the so-called brittle nugget
and a crack running, and that type of thing. We never found any brittle
nuggets.
DR. SEALE: It is still kind of interesting. You say that
there is probably more bang in making the K1c change than there is in
going to a different flaw distribution. On the other hand, I guess by
the same token you have come to the apparently fairly strong suspicion
that you don't have any brittle nuggets. You should be also getting a
much better idea of what flaw distribution is as a result of some of
these others.
MR. WICHMAN: When I said maybe reduce that 1/4T flaw,
Appendix VIII would go a long way toward having more reliable
inspections, but Appendix VIII has not been mandated by 50.55a yet,
because 50.55a hasn't been updated and out on the street. You are going
to hear more about that a little later.
DR. SEALE: The reason I don't shut up is if there is this
opportunity to make these changes, clearly there is also a cost
associated with it, namely, the additional inspection kinds of things.
MR. WICHMAN: There would not be additional inspections. It
would be people that are qualified to Appendix VIII of the code.
DR. SEALE: Still that would require some additional
resource expenditure.
Has anyone done a trial assessment of what a more realistic
flaw distribution assumption would do?
MR. WICHMAN: It has been looked at. Reducing the factor of
2 on pressure has been looked at or is being looked at. Also, this has
been looked at for quite sometime. There is a certain amount of relief
involved in this. This was a much less controversial change to make and
I think provides more relief than going to a somewhat smaller flaw, but
I can't quantify that.
DR. SEALE: With all of the interest in reducing burden
presumably by way of identifying and relaxing unnecessary conservatism
-- and that's a judgment call -- we've heard the idea of harvesting the
low hanging fruit, namely, doing the things that are easy, the things
that we do have the database for both in terms of the information
necessary to modify the assumptions in the calculation, and secondly, an
understanding of what the consequences might be in terms of doing that.
My only comment would be that this is probably not the last
time you are going to hear of somebody wanting to come up with the idea
of maybe harvesting some more low hanging fruit. So it seems to me it
might be worthwhile to take a look at these different effects and get an
idea of where --
MR. WICHMAN: It is currently being looked at by Tim
Greasebox's committee in Section XI. Plus I think you heard Mike
Mayfield allude to it the other day as perhaps part of the fallout of
this integrated look at the PTS rule, 50.61. There will probably be
some things that can "improve" the P-T limit, give further relief on P-T
limits.
DR. SEALE: It will change. Whether or not it improves is
something we have to figure out.
MR. WICHMAN: I think the more we know the more we can
reduce some of the conservatisms involved. K1c still provides a very,
very considerable margin. There is a white paper on the subject.
DR. SEALE: I don't want to belabor the point. I thought I
would try to learn a little bit more here.
DR. SHACK: Coming back to this database question, there is
some difference between the EPRI database and your database. Your
database only contains the data that people actually submit as part of
their package and the EPRI database contains a wider range of data. Is
that the difference?
MR. WICHMAN: It's a wider range of data. The data that we
have in our database has actually been used for P-T limits, PTS, et
cetera, et cetera, and it's traceable. I'm not sure that all the
industry data may fall in that category.
DR. SHACK: John Fair is going to update us on the ASME
Working Group.
MR. FAIR: Good morning. I'm John Fair with the Office of
Nuclear Reactor Regulation. I'm going to be discussing the piping
seismic design criteria in the ASME Code.
[Slides.]
MR. FAIR: This is an update of a briefing that you had
about a year ago. So what I will try to do is go briefly over the
background, the discussions on the research study that was conducted by
the NRC, the work done by the ASME Code, and a brief summary on where we
stand right now on the issues.
The topic is the 1994 addenda to the ASME Code. There was a
significant relaxation of the piping seismic design criteria in which
the staff had concerns with the technical adequacy of the data
evaluations that supported this change. So we did not endorse the
change rules at that time.
To go through the background of how these changes occurred,
dynamic testing was performed in the 1980s. In 1991 there was a Special
Task Group assigned to look at the conservatism in the current design
rules on piping. As part of that Special Task Group, an evaluation of
the test data that was performed in the 1980s was done.
NRC staff members were members of this particular Task Group
and we had concerns at that time with the data evaluations that were
going on.
Although this Task Group report did not recommend the
changes to the Code, the work that was done there was used as the basis
for the changes that were adopted in the 1994 addenda to the Code.
As these rules were going through the Code, the NRC members
of the various committees and subcommittees voted negative all the way
through the main committee.
After the rules were finally passed, the NRC sent a letter
to the ASME telling them that we could not endorse the rules at this
time. This was in December of 1994.
We followed that letter up with a letter in May of 1995
which detailed the technical concerns that we had with the rules. This
was a very detailed letter with several attachments, which included the
NRC staff technical concerns, the technical concerns of the researchers
that the NRC had contracted to look at the rules, and technical concerns
of a peer review group that was also part of the research effort.
As a result of that, the ASME formed another group, the
Special Working Group on Seismic Rules, in September of 1995 to evaluate
the NRC's concerns as well as some other concerns they had received
after the rules were published.
The staff and our contractors participated in these meetings
as observers and the staff briefed this subcommittee in June of 1998.
What I will do now is highlight the major areas of technical
concerns that we identified in our May 24th letter.
Recognizing that in the letter itself there are numerous
amounts of specific technical concerns raised that are being addressed
by the Special Working Group, the four main concerns that we highlighted
in our letter were:
There was insufficient test data to support a conclusion
that piping system collapse cannot occur.
The relevance of this is the new seismic rules were a
significant relaxation in terms of stress criteria over the old rules;
the stresses on their own are well above the yield stress of the piping,
and the only technical justification for the rules is that the types of
failures of piping systems during earthquakes are fatigue type failures
and not collapse type failures. Looking at the test data, we believe
there was evidence of collapse type failures occurring in the test data
and that the data itself did not support the conclusion that there is no
piping system collapse possible.
We also had concerns with the evaluation of the test data.
The data evaluations involved extrapolating the data to cases that
weren't tested. We had identified some errors in the data
extrapolations and some technical concerns with some of the data
extrapolations.
There was also a concern that the test data was used to
establish the criteria, and a margin was developed to evaluate against
the test data. We had a concern that there was no real technical basis
presented for this margin value. As the data was being evaluated, at
first there was a higher margin selected as the target value. When the
data evaluations were showing lower margins than originally expected in
the piping systems in the component tests based on some of the
extrapolations, they selected a lower margin of safety for evaluating
the test data again. There really wasn't a technical basis for that
margin of safety that was selected.
DR. KRESS: Is there ever a technical basis for that? From
what I've seen, it's always a judgment call: what margin are you
comfortable with?
MR. FAIR: Let's say the normal code design practice. There
are fairly well established criteria such as material yields, and you
maintain a margin against yield or something like that.
DR. KRESS: The level at which you maintain it is a judgment
call.
MR. FAIR: That's correct, but the value itself is fairly
well established. What we had here was a handful of test data in which
the test data had a considerable amount of data scatter. The margin was
compared to the minimum value of the test data. The question was, did
they do an adequate evaluation of the potential data scatter to
establish this margin?
DR. KRESS: Okay.
MR. FAIR: Also a general area of concern was the fact that
all the testing that was evaluated in terms of a margin were component
tests. As you saw last time, these were cantilever tests and there was
a general presumption that these were conservative compared to system
tests, but there was nothing done to establish a technical basis for
that assertion.
The reason we had a concern is that when you go from a
component test like this to a system test, there are other loads in an
actual piping system such as thermal type loads that could impact the
overall margin.
I will go quickly over this. I think you had a fairly
detailed discussion last time on the research program that was
established with ETEC in 1993.
It included an independent evaluation of the test data.
That was discussed in the previous group.
Independent analytical studies of test margin extrapolations
were performed. These were done by the Cal Tech researchers.
The program also included a peer review group of experts
looking over the evaluations.
The program was completed and the results published in
NUREG-5361.
The review concluded that the technical basis for the rule
changes in the 1994 addenda was not complete.
Following the publication of the NUREG, ETEC continued to
provide some technical support to the Special Working Group that had
been formed to look at these concerns.
As I said before, after the NRC identified its concerns and
other people had identified concerns with the 1994 rule change, the ASME
set up another working group to take a look at the technical concerns.
The Special Working Group has basically relied on the data
evaluations performed by ETEC, because it takes a lot of effort to do
the data evaluations when these is test data, and the members simply
don't have the resources or time to do the detailed evaluations.
The group did vote to adopt the margin definition proposed
by Dr. Kennedy, which I believe he presented at the last meeting. This
was the 1 percent failure probability margin capacity of 2. This was
based on his risk evaluation to keep piping from becoming a predominant
risk factor in the seismic PRAs.
They also voted to investigate a proposal by Dr. Kennedy to
include a minimum value for the Code bending stress index. If you will
recall from the last presentation, when he did a data evaluation of the
EPRI test data, it turned out that the margins for weld joints were much
lower than they were for elbows. He made a proposal that they raise the
stress index on the weld joints in order to get them consistent with the
elbow test data. That is still going on. They are still reviewing that
proposal.
They are also reviewing some fairly recent data provided by
the Japanese with some additional testing and data evaluations.
DR. KRESS: This is the shake table?
MR. FAIR: Shake tests. Similar type of tests to what EPRI
did.
There is also a brand new proposal that was forwarded by Dr.
Matzen of NC State University which involved a finite element evaluation
of some elbows. The preliminary indications are maybe there are some
conservatisms in those elbow stress indices. Therefore, instead of
raising the other indices to get the data consistent with the elbow
indices, they may reduce the elbow indices to get the data consistent
with the other and keep the stress allowance the same as they were
previously.
DR. KRESS: Do you know what code he used?
MR. FAIR: I believe he used ANSIS in his evaluation.
DR. KRESS: That's not a dynamic code, is it? It's a static
code.
MR. FAIR: What is being done is static analysis. This is
strictly looking at static collapse of the elbows.
DR. KRESS: He's not trying to do the dynamics.
MR. FAIR: No.
As I said, there was some recent data testing and
evaluations done by the Japanese. One of the areas they did was some
specimen tests for fatigue ratchet. The fatigue ratchet data show a
significant reduction in fatigue life due to ratcheting at very low
cycle regime.
The significance of this is in evaluating the test data, in
order to extrapolate the data to other conditions not tested, you have
to have some kind of a correlation for fatigue failures. Some of the
previous correlations used the shape of the fatigue curves to do this
correlation. If the fatigue failures are significantly different than
what you would get with the shape of the fatigue curves, you could be
off on the data extrapolations.
When ETEC was doing their program for the data
extrapolations that were presented last time, they did two types of
analyses. They did an analysis of fatigue failures using the shape of
the fatigue curves for the damage calculations.
The other evaluation was to do a maximum displacement type
evaluation. The intent of this was to try to bound the possible effects
of the ratcheting. We still believe that those are reasonable bounds,
the displacement studies that were presented last time.
To summarize the status of these issues, I believe that the
Special Working Group members do concede that collapse is a potential
failure mechanism and that the rules have to cover this potential
failure mechanism.
ETEC has done the rigorous technical evaluation of the
existing test data. The ETEC evaluation has addressed the staff
concerns with the deficiencies in the previous data evaluations.
Dr. Kennedy provided a risk-based approach to establish the
minimum required margin. We think that covers the concern we had with
at least having a foundation for the basis for the margin.
He also had proposed some system redundancy and nonlinear
factors to go along with that which we haven't yet endorsed. These
redundancy and nonlinear factors essentially reduce the margin required
out of the component tests. They have a presumption in them that the
system will be more conservative than the components.
Those data evaluations are still ongoing. These were
judgment values by Dr. Kennedy. There has been no resolution of what
these values could be or might be.
There is also a concern in his proposal on the elbow data,
that they excluded one of the tests, which is Test Number 37. That
happened to be a thin walled stainless steel elbow with a heavy mass on
it and very low frequency system which was tested near to point of
collapse, and the test was stopped before it actually collapsed.
This is an outlier in terms of margin; it has a very low
margin, much lower than most of the other elbow test data. There still
has not been any adequate explanation of why this particular test was
low compared to all the other tests. So this is still an issue that is
to be resolved.
DR. KRESS: It looks like the concerns expressed by NRC were
pretty well out of it.
MR. FAIR: It depends on who you talk to. We think they
were, but some of the industry thinks that the evaluations going on are
still validating their original judgments.
The evaluation of margins is ongoing for piping systems
right now.
DR. KRESS: Yes. I must admit I was trying to refresh my
memory on the 1 percent probability of the capacity. I don't recall
exactly where that came from.
MR. FAIR: He started off with the margins that are supposed
to be in the seismic PRAs. He selected a value such that you didn't
want piping to become a predominant risk factor in the seismic PRA
studies that had been performed previously. He back-derived a value of
capacity factor from that.
DR. KRESS: Using the probability of a LOCA or initiating
event?
MR. FAIR: He started off with a fixed margin number that
came out of the studies and then, using that number, he back-derived a
capacity factor that you need in piping so that you wouldn't exceed
that. As far as the details of his evaluation, I don't think I will
attempt that.
I was going through the original four concerns that we had
raised. The last one was the margin on the actual systems. That is
still being discussed in the Special Working Group. Issues such as
loads due to seismic anchor motion and concurrent thermal loads have not
been resolved as of yet.
There was one other thing that was beyond the four issues
that I put on my previous slide. ETEC in their final report identified
a concern with a lack of high temperature test data for carbon steel,
that there may be a problem that all the testing has been done at room
temperature. The presumption is the margins would be the same at
temperature as at room temperature, and there is a concern that maybe
for carbon steel this may not be the case. This issue has not yet been
resolved either.
DR. SHACK: So it's basically still work in progress.
MR. FAIR: Still work in progress. I think there are still
a lot of tough issues to be resolved before this is done and finished.
DR. SHACK: Since nobody is designing a reactor, what impact
would this have on operating reactors? Would they go in and start
changing supports?
MR. FAIR: I believe they would probably try to use it to
resolve issues. At all plants you are constantly coming up with
problems or identifying new problems on the piping systems where
somebody finds a missing support or something has not been put in the
right place. They probably would want to use it as evaluation criteria.
They may want to use it to further remove snubbers off existing piping
systems.
I guess the other was with the advanced reactors there was a
request at one time to use this new criteria, which was not accepted at
that time by the staff.
MR. DUDLEY: I have a process question. Do you have a feel
for when the working group will be in a position to make a
recommendation to the review committee? A follow-up question to that
would be, at what point should the ACRS take another look at this?
MR. FAIR: I hesitate to even make a guess. I think their
original intent was to have that done by now, but the technical issues
are very tough. I'm not even going to hazard a guess on that. I don't
think they are close to final resolution, but I think some of the
hierarchy in the ASME think they are close. I will just leave it at
that.
DR. SEALE: The working group is a creature of the ASME, as
I understand it.
MR. FAIR: That is true.
DR. SEALE: They are the people that have their hand on the
throttle. I guess there are kind of competing pressures. Certainly
there is a strong pressure to get it right. There is also a strong
pressure to not say "oops!"
MR. FAIR: Exactly.
DR. SEALE: There is a reason for unwillingness to go either
way on this.
DR. SHACK: Is the Japanese program ongoing? You can only
look at the original data so long and so often.
MR. FAIR: Yes, it is. They are still in the process of
doing data evaluations. They have the data. I don't think they have
released it formally to anybody.
DR. SHACK: Not even the working group, then.
MR. FAIR: They have presented it at the working group, but
I don't think they have actually given the data to anybody.
DR. SEALE: Are the elements of their test programs such
that they do focus on these issues?
MR. FAIR: I think their focus is not necessarily on those
issues.
DR. SEALE: Obviously they have other things they want to
test.
MR. FAIR: An example is the way they did their component
testing was to support the components vertically from deadweight. So
they didn't have this deadweight force that could drive something into
collapse. When they did their testing, they were all neutral loads;
seismic tests were all about a zero mean, so they concluded that there
was no collapse in their tests. The way they set the tests up, we would
have predicted there would be no collapse.
DR. SEALE: Is their test program broader than just nuclear?
Are they looking at concerns in the petrochemical industry and things
like that as well?
MR. FAIR: I really can't answer that. I'm only familiar
with the data they are bringing into the Special Working Group.
DR. SHACK: In the collapse load argument you just made, was
that the Japanese, or was that the way the original tests were done for
the EPRI tests so there was no collapse in those tests?
MR. FAIR: No. The Japanese tests were horizontally excited
with a support in the vertical direction so that you didn't have any
existing deadweight on the system. In the EPRI tests they were vertical
cantilevers with horizontal excitation with no additional deadweight
support. So some of them had fairly significant deadweight stresses,
such as Test 37. That is the one that went closest to collapse. As a
matter of fact, it would have collapsed had they not stopped the run
prematurely.
DR. SHACK: So the Japanese tests won't really give you any
extra insight on that because of the way they have set them up.
MR. FAIR: Not on that particular issue.
Dr. Kennedy did do a statistical evaluation of what he calls
ultimate moments capacities. His evaluation came up very similar to the
EPRI tests in terms of maximum dynamic moments.
DR. SHACK: So there is a consistency there at least.
MR. FAIR: Consistency in the data, yes.
DR. SHACK: Any additional questions?
[No response.]
DR. SHACK: Thank you, John.
Next we have Tom Scarbrough.
MR. SCARBROUGH: Good morning. My name is Tom Scarbrough.
I've talked to you all several time before on motor operated valve
issues and today I get to talk to you about a new rule that we are
working on, in-service testing and in-service inspection.
[Slides.]
MR. SCARBROUGH: What we are going to talk a little bit
about today is the supplement that we have prepared to the proposed
amendment to 50.55a that was issued in December of 1997. There have
been some additional discussions and a decision to go out with a
supplement for public comment.
First of all, 50.55a requires owners to construct, inspect
and test their components in accordance with the ASME Boiler and
Pressure Vessel Code with certain conditions and limitations.
It also requires licensees to update their ISI and IST
programs every ten years. That is the subject of the discussion that we
are going to have some more of.
DR. KRESS: Does it specify the components?
MR. SCARBROUGH: Yes. It gets down to Class 1, 2 and 3. It
does it by that classification.
In December of 1997 the NRC published a proposed amendment
to 50.55a to update it to more recent editions. I will talk a little
more about which editions in just a minute. But basically bring it more
up to date to the current editions of the Boiler and Pressure Vessel
Code, and also the newer ASME Code for operation and maintenance,
including nuclear power plants, of what they call the O&M code, but
still retain the update requirement.
DR. SEALE: What is the difference between bullet 2 and
bullet 3?
MR. SCARBROUGH: Bullet 2 is what is in the rule right now.
That is 50.55a right now, that they have to update every ten years. In
December of 1997 there was a proposed amendment to 50.55a to bring up
the 1989 code, which was endorsed and what is on the books now, to the
1995 code. That is what was proposed in 1997 in that proposed
amendment. Everything else would basically stay the same, but there
were a lot of things of that nature. Basically bring it up to date from
1989 to 1995.
As part of that discussion that went along with the issuance
of that proposed rule was DSI 13. That had to do with how we interact
with the industry in terms of codes and using codes and things of that
nature.
There was a stakeholders meeting in Chicago, and a lot of
the discussion involved how we endorse codes and how we update codes.
One of the items that came up in the discussion was, do we really need
to continue to update IST and ISI programs every ten years to the latest
code, or are they good enough? Can we just continue with what we have?
Based on that discussion and the interaction that took place
regarding that issue, we decided to supplement that 1997 proposed
amendment in just the one limited area of should we eliminate the
10-year update.
We prepared SECY-99-017, which explained this and all the
pros and cons in terms of eliminating the 10-year update, and presented
it to the Commission. We are waiting for the Staff Requirements Memo to
be returned. We have seen the draft, and it says go forward, look at
these issues, come back to us, deal with the burdens and the benefits
more fully in the final rule. We plan to do that. We are on hold right
now waiting for the specific language of the SRM to tell us where we go
from here.
In terms of the current 50.55a, it has been on the books for
quite a long time. As you can see, it is still for Class 1, 2 and 3
components. It requires construction to the 1989 edition of Section
III. Parenthetically we say that the proposed rule would update it to
the 1996 addenda, but there is no update requirement associated with
Section III. So that really is not affected by the supplement we are
talking about.
In terms of inspection, the current regulations, it endorses
the 1989 edition of Section XI for ISI, and for metal containment and
concrete containment it endorses the 1992 edition of Section XI for ISI.
For in-serve testing for Class 1, 2 and 3 pumps and valves,
it references the 1989 edition. Then it cross references. Because the
OM code was just sort of in its infancy when that occurred, a cross
reference to any references to OM documents in Section XI really means
the 1988 addenda to the OM standard.
That is what is on the books right now; that is what is out
there right now that is currently required.
I want to give you a little background of what was in the
rule that was published in 1997 from an ISI and IST point of view.
There are some other Section III areas that are also in that 1997
proposed rule, but I won't get into those because they are not affected
by the supplement.
Basically, what the proposed rule in 1997 did was
incorporate by reference the 1995 edition with the 1996 addenda of
Section XI for ISI requirements for Class 1, 2 and 3 components.
It also had a requirement that PWRs perform volumetric
examinations on their high pressure safety injection systems for weld
examinations.
It also expedited the implementation of Appendix VIII of the
1996 addenda of the code, which has to do with qualification for
performing ultrasonic examinations.
Also, for IST, in-service testing, it incorporated by
reference the 1995 edition with the 1996 addenda of the OM codes. Now
we are switching over from the Boiler and Pressure Vessel Code for IST
into the OM, because the OM code is taking over that area.
We also added a requirement that licensees supplement their
stroke-time testing for motor-operated valves with the design-basis
programs, because we found that stroke-time testing wasn't really
adequate for motor-operated valves.
Also, it permitted implementation of several specific code
cases. There is a new code case called OM 1, which has to do with
improving the performance of diagnostics on motor-operated valves in
lieu of stroke-time testing, and also various portions of OM code
editions that were permitted to be implemented.
Those are basically the areas of ISI and IST that were in
the December 1997 proposed rule.
We are working on the actual Federal Register language of
the supplement now, and that is going through the chain as we await the
final SRM language.
In terms of the supplement, right now what we would propose
is that we would eliminate the requirement for licensees to update their
ISI and IST program every ten years.
We would establish a baseline. This is an area that we
would specifically deal with as part of public comment. What is the
right baseline? That is a question we want to finalize, but right now
the current thought is the 1989 edition for IST as well as ISI.
However, for the containment structure components, metal containment and
concrete containment, we would still continue with the 1992 edition,
which is in the rules right now.
Also we would apply Appendix VIII qualification standards
for ultrasonic examination. That also would be part of the baseline.
This is sort of the baseline that we are proposing as we go
out for public comment; we are specifically going to ask for comment in
this area.
We are going to allow voluntary implementation of later
editions. The current thought process is we continue to review more
recent editions of the code, endorse them in the regulations, and
licensees could voluntarily pick them up. That would then be their code
of record and they would be responsible for following it just as if they
were doing the 1989 code. That would be an option that they could do.
In the future, our thoughts would be that licensees would
sort of follow the same process they are doing now. A new licensee
coming on would pick up the code that was referenced one year prior;
they would sort of get a one-year grace period, and they would pick that
up.
That would make sure that the construction matched the
operation. You wouldn't have something going way back in time to 1989
when you would have the plant constructed in 2010 or something. That
would make sure they match, but they would not be required to
periodically update; once they were assigned that particular edition of
the code, they could continue that for their life.
DR. SEALE: What about a revision to an existing plant that
would require an operating license amendment, a tech spec amendment, or
something like that? Would that be governed by the version of the plant
originally, or would it require the latest version with the one-year?
MR. SCARBROUGH: Under Section III, I think they are allowed
to use the more recent editions for construction, but they are only
mandated to what was under construction. They could use the later ones,
but they wouldn't be required to use them. The baseline would be
established, and the baseline would be the requirement.
DR. SEALE: I'm talking about some fairly significant
modification.
MR. SCARBROUGH: That is something we probably need to deal
with. If there is a major license amendment, does it reopen the issue
of would they be considered a future licensee? I guess that would have
to be dealt with. That's a good question. My first thought would be
they wouldn't have to, but if it was something very significant, would
it be appropriate to pull them up to something more recent?
DR. SEALE: What about license renewal?
MR. SCARBROUGH: They would keep going with this.
Larry.
MR. CAMPBELL: My name is Larry Campbell. I'm in the
Quality Assurance Vendor Inspection, Maintenance and Allegation Branch,
or whatever it is called now.
Under Section XI licensees can replace steam generators and
do major modifications to their piping systems as well. Section XI
currently permits them to use the original construction codes or later
editions of the construction codes provided that later edition is
endorsed by the NRC via 50.55a. So licensees do major mods like
replacing generators and major modifications currently under the
provisions of 50.55a.
MR. SCARBROUGH: So typically they would not be expected to
update.
That is what we are drafting in terms of our proposed
supplement.
DR. SHACK: Eliminating the requirement to update, that is
just a judgment that it's too burdensome?
MR. SCARBROUGH: Yes. That is what is driving this. After
all the years of experience and the changes that we are seeing in the
code more recently, are those changes significant? If you look way back
at 50.55a and the same considerations, there were discussions of a
maturing code. There were things like that talking about having you
update every ten years to pick up more recent improvements in the code.
Over the past few years we haven't seen a lot of significant changes in
the code. There are some changes and some improvements.
DR. SHACK: The requirement for a performance demonstration,
it seems to me, is a major change. The requirement for a performance
demonstration of the ISI. You are just sort of assuming that after that
there won't be any more?
MR. WICHMAN: We are talking about voluntary updates in
terms of Section XI. That would be mandatory.
MR. SCARBROUGH: That is part of the baseline.
MR. WICHMAN: That is the baseline.
DR. SHACK: My problem is with the argument you have seen
this major change in the ISI requirements between 1989 and 1995, and now
you are saying it's mature and you are not going to have to worry about
it anymore.
MR. SCARBROUGH: I think it is set forth in the SECY paper.
Geary Mizuno from OGC.
MR. MIZUNO: Geary Mizuno, Office of General Counsel.
The staff's paper indicated that there would be sort of two
paths for future endorsements. One would be where the NRC staff found
that the code was acceptable for use but did not have any specific
requirement that was necessary to assure adequate protection. In that
case they would simply endorse it for voluntary use. However, if there
were provisions that the NRC staff determined were necessary for
adequate protection, those would be mandated; they would not be
voluntary; we would take action to mandate it.
I would imagine that if you had something similar to PDI
occur five years from now, that would be a provision that would be
mandated by the NRC staff in a future rule.
DR. SHACK: But it would take a future rule rather than the
requirements you have now to do the update on a scheduling basis.
MR. MIZUNO: That's correct. Five years from now, were
there to be a provision similar to PDI that the NRC staff determined to
be necessary for adequate protection, the NRC would have to go through a
rulemaking. It would have to address the backfit rule, demonstrate it
was necessary for adequate protection, provide public comment, and go
through the entire rulemaking process to get incorporation by reference
of this new provision, and the rule would mandate as opposed to open for
voluntary use of this specific provision.
MR. TERAO: Dr. Shank, this is David Terao. Geary is right,
but I think what is being misunderstood is the fact that we will still
periodically update 50.55a. I think that is what Geary Mizuno was
referring to. It will take a rulemaking to backfit provisions later on
that the staff feels are necessary for safety, but we would do it under
our routine 50.55a updates.
When we endorse later codes editions and addenda, we will
review those codes and editions and addenda, and if we feel that there
is a significant enhancement in that code for a particular safety issue,
we would then put into the 50.55a update under the backfit consideration
a requirement that utilities would have to update to that.
DR. SHACK: Every one of these would then be subject to a
backfit analysis.
MR. TERAO: Correct.
MR. SCARBROUGH: That is one reason why we think
establishing the baseline is very important, because after this,
everything is going to be under the backfit 109 rules.
DR. SEALE: That's a heck of a batting average.
MR. SCARBROUGH: We are preparing the Federal Register
Notice with the this new language, and we specifically ask for comment
in these areas, what I call issues.
Potential effect on safety, including reductions in
effectiveness of the ASME code.
Selection of the proper baseline.
The regulatory benefits and burdens not only to licensees,
but also vendors, insurers, states, and the standards organizations. So
there is quite a bit there.
The burden on licensees actually to update. We get various
estimates of how significant the burden is in terms of monetary and time
and resources. So we would like to really focus that and find out how
significant is this burden in the overall scheme of burden that is
applied to them.
The potential effect on the number and detail of licensee
submittals. Are we going to get a lot more relief requests? Are we
going to get less? Is it going to be about the same that we get now, or
is that going to change significantly?
The changes to the range of the code editions that are
applied. If the inspectors are going to have this whole hodgepodge of
parts and pieces of more recent editions of the code than 1989 that they
have to deal with, it is going to be much more confusing to them.
The potential effect on licensing actions. Does this open
the door for more license amendment type of issues to come up? Does it
affect the regulatory guides in terms of how we endorse various codes
and code cases? Or will it affect the risk-informed programs? If there
is not the ten-year update, is that going to undermine some of the
risk-informed initiatives that are out there?
Then we come back to the states again. What is the
potential effect on the states and other organizations that rely on the
code? Is there going to be a mismatch between the NRC and what the
states require.
Application of portions of the code. It's called cherry
picking. Are people going to pick and choose the portions of the code
that they want? We have some language in there to try to minimize that,
to focus people on using the entire code as endorsed by the regulations
as opposed to coming in with individual requests for pieces of the code.
So those are some of the areas. We recognize that all these
areas have to be dealt with and resolved before we go final.
Next we have this ambitious schedule. It is already out of
date. This is the schedule we put in the SECY paper.
We had hoped to have the supplement published for public
comment by April 5. Therefore it should have already been down at the
Federal Register at least a week ago, and it is still sitting on my
desk. We have to finish with the SRM, get that issue resolved, finalize
the concurrences from the Office Director and get the EDO to sign it.
So we still have some more work to do.
We have scheduled a public workshop on May 27. It is going
to be in Two White Flint Auditorium. We have invited NEI and ASME.
They have indicated they are going to participate to talk about all
these issues that we have raised in terms of pros and cons and the
benefits and the burdens. We hope that will be a very informative
workshop.
Our goal originally was to try to finish up, by late summer,
complete consideration of all these comments and put it all together,
and in September have the offices concur and present the final draft in
October to ACRS and CRGR, and then November submit the final rule for
Commission approval, and then publish by the end of the year.
We don't think we are going to make the end of the year now,
but our goal is on the fast track and to get it done at least by early
next year. We will have to see how the schedule goes in terms of public
comment and resolving those.
That is where we are. I will be happy to answer any other
questions. I appreciate the guys coming over and helping me out here.
DR. SHACK: I guess my initial reaction is all the effects
seem to be negative except for reducing the burden on the update. The
ASME code is a consensus document. Changes to the program don't get
made unless it is sort of a consensus of the community that there is a
purpose in changing them anyway. To update this every ten years doesn't
seem to me a horrendous burden. The utilities will have already made
their input to the consensus process.
MR. SCARBROUGH: There is a lot of discussion on how
significant the burden is. There are various views on that, and that is
something we want to focus on. Other people have said that they think
it's a significant burden, and other people have said, just like you
have said, that it is not a significant burden. We really need to nail
that down, because we are getting conflicting points of view. To really
finalize this, we really have to know what the program is. In terms of
the feedback we are getting from the Commission, they really want to
know what are the benefits and the burdens associated with this
proposal.
DR. SHACK: Is there some other benefit that I am missing?
MR. SCARBROUGH: I think burden probably is the main
benefit. They do this now; they update; they come up to the latest
code; all the consensus documents are put together. I think the primary
benefit is a burden reduction.
DR. SHACK: Is this something that the industry has been
pushing, or did we volunteer to do this?
MR. SCARBROUGH: We heard from NEI that this was something
that they were interested in. They thought this was very important to
them.
We were ready to go out with the 1997 rule. It was
finished. The comments had been resolved and it was ready to be final.
There was some input from the industry that this issue on the updating
was very important and needed to be resolved as quickly as possible. So
we held up issuing that to deal with this issue, and they felt it was
important enough to hold up the full rule package to deal with the
supplement issue.
DR. SHACK: It seems to me you will have a significant
burden on NRC, because I would assume that as a matter of fact many of
the utilities would voluntarily implement. I would expect to see a
tremendous patchwork of programs resulting.
MR. SCARBROUGH: Yes, and that is something we are concerned
about in terms of a lot of cherry picking where you take the best parts
and leave off the comprehensive pump test. Or you leave off parts that
are more difficult to perform but give you a lot more information from a
confidence point of view. They might try to just pick and choose.
We tried to put some language in the proposed rule to try to
indicate that that is not the preferred approach; that is going to be a
much more difficult path to follow. If they pick up the complete later
endorsed editions of the code as endorsed with the limits and
conditions, then that would be without prior approval. They could do
that without coming in, but if they only pick and choose, then they
would have to come in through a relief request process, and that is much
more burdensome for them as well. We are going to try to discourage
that as much as we can, but we know there is going to be a tendency to
do that.
Does anybody else have any comments?
DR. SEALE: I can understand your concern.
DR. SHACK: If we are going to do inspections, it seems to
me one would like to do effective inspections. Otherwise you are
wasting people's money, time and exposure.
MR. COZINS: This is Gary Cozins from NEI. NEI has been
mentioned several times in this discussion. It might be worthwhile
putting some perspective on it.
Part of the logic of the thought on approaching this is our
plants are safe today. Every time we go through a 10-year update it is
an expensive process. I've been told by many licensees it's on the
order of about $1 million per plant for update. I don't know if anybody
else has numbers that would confirm or deny that, but that is my
understanding. So it is a fairly significant resource.
The point is today we are safe. Do we need to implement all
these new requirements that may be in a new update to continue to be
safe? Is there something that we don't have presently? So there is
this balance. We have defined what is safe.
If there is something that is necessary for adequate public
health and safety, then the staff has the obligation to impose that upon
licensees. That process still is in the works. The question is, we
have gone from the safe level. Do we need to go to a different level
that may or may not be safe?
Then we talk about the differences that may exist presently
between licensees within a licensee's plant itself on the different
editions. That is the present status today. That is not a change. No
plant out there has a singular edition applicable to everything. They
have come in for changes in the code where it has been beneficial to the
plant and gotten exemptions from the staff, or modifications, to
implement changes that have occurred in the code that have not been yet
endorsed by the staff. So right now we have a patchwork. I don't
believe that will change in the future any more than it has to date.
It is a little bit of a hassle, but we do need to keep good
records that will assure us that we understand what requirements go to
which code.
DR. SHACK: It is one thing to have construction
requirements from one edition to the code and inspection requirements.
It would seem to me that would patchwork a lot more.
MR. COZINS: We have come a long ways in time. The question
is, are we safe today? That's the real question. And do we need to
change to new technology to continue to be safe? That is a question
that continues to evolve, and I believe the staff will continue to think
about that. I know you have done a lot of thinking on that with regards
to Appendix VIII.
There are other requirements that a plant may need to be
safe to operate today that they don't necessarily need to change to
continue to be safe to operate. There is the discussion. I think that
is the crux of the matter as they examine this.
I have not seen the SECY yet, so I don't really know the
exact words that will be in there. We do look forward to this workshop
in May. I believe the list you put up there is the ten odd questions.
This is very similar to what they did in the ESI. Teams take over, and
I understand that was an effective way to get all stakeholders an
opportunity to have input, and I think staff will probably have a good
basis to make its final decision on that point.
MR. NORRIS: My name is Wallace Norris from the Office of
Research. I would just offer another data point. I'm on a couple of
ASME committees. Talking to ISI coordinators, some of the numbers I've
heard thrown around for 10-year updates, it is between $150,000 to
$300,000. As far as computerization of their programs, they have done a
lot of that now.
DR. SHACK: It seems to me the maturity of the programs
would make the updates self-limiting in the sense that if the ASME feels
the programs are mature, there will not be significant changes in the
code. We all know that ASME code changes are not made lightly.
MR. TERAO: Actually, with respect to ISI Section XI in the
past ten years, or since the 1989 code, there have not been any new
requirements added to Section XI. They are mainly reductions in
requirements. From that perspective, the staff feels it is no longer
necessary to require these codes. If utilities would like to
voluntarily adopt these codes, later editions, addenda, they may do so
provided they are endorsed in 50.55a. I believe Section XI has had no
enhancements in their requirements in the last ten years.
MR. WICHMAN: There is another enhancement, and that is in
the area of flaw evaluation and analysis, Appendix K for the low upper
shelf; improvements to flaw evaluation appendices; Appendix G. All of
these things have been upgraded tremendously during the past several
years. So that's an enhancement.
MR. CAMPBELL: Another enhancement, which is somewhat
forgotten about, is Section XI not only covers ISI; it covers repairs
and replacements, and there are some unique repair procedures dealing
with everything from steam generator tubes to other repair procedures
that are contained in Section XI that you will not find in the
construction code. So there are rules for repairs, replacements and
modifications contained within Section XI.
DR. SHACK: Does the 10-year update include all the Section
XI? When we say ISI and IST, that is really shorthand for Section XI.
MR. WICHMAN: Yes, it does.
MR. TERAO: I'm sorry. I have to disagree. 50.55a does not
require that the industry adopt all of Section XI. It only requires
in-service testing of pumps and valves and in-service inspection,
including hydrostatic tests. Areas such as repair and replacements are
not included under 50.55a in the current regulation.
We are trying to consult with our counsel here.
MR. CAMPBELL: This is Larry Campbell. Before I joined the
NRC I used to prepare ISI and IST submittals. I know the code has
changed. In this nice submittal I prepared in 1988 before I joined the
NRC we put in there that that submittal for repairs and replacements
could conform to this edition and addenda to the code. I've been away
from this for ten years, so I don't know what the current submittals
say.
DR. SHACK: I am assuming that the licensees can voluntarily
choose to do that. The question is whether they have to do it.
MR. TERAO: That is correct. We are talking about
regulation covering it versus a licensing commitment. Certainly there
are licensing commitments that they would adopt all of Section XI, but
as far as what 50.55a requires, it is only in-service testing of pumps
and valves and in-service inspections, including hydrostatic tests.
MR. CAMPBELL: And 50.55a also addresses codes for
construction, depending on when the construction permit was issued.
MR. MILLMAN: My name is Gilbert Millman. I'm in the Office
of Research. I'm the NRC representative on the ASME Board on Nuclear
Codes and Standards, the ASME Boiler and Pressure Vessel Main Committee,
and the ASME Subcommittee on Nuclear In-service Inspection, which is the
committee responsible for Section XI.
This is really a very complicated issue. When Section XI
writes its rules, it is always making the assumption that those rules
for any specific edition or addenda will be implemented in their
entirety. The code itself does have provisions for using portions of
later editions and addenda, but the fundamental assumption is that the
base code would be a specific code edition or addenda.
It is also assumed by all many hundreds of people
participating on the Section XI committees that the work they are doing
is useful; that when they participate for a week to write new code, that
they are taking into account the experiences of 100-odd operating
plants.
When people say that they are reducing requirements, that is
really not an accurate interpretation of what is going on. I am as
critical as anyone else on the code. I probably have more negatives on
valid actions than probably anyone that has ever participated in the
code process, but I do recognize where the other participants of the
code are coming from.
There is a great deal of experience in using Section XI. We
have been using Section XI since 1971. That is 28 years of experience.
It was written in a very conservative manner, as were other standards
and other portions of the Boiler and Pressure Vessel Code. We have got
a great deal of experience. There are certainly examinations that have
not proven to be fruitful and can be backed off, and when we back off an
examination, it is not just eliminating an examination; it really is
reducing occupational exposure in certain areas.
When we say the burden on a utility is $200,000, or
whatever, to update to their new 10-year code, we are not taking into
account in that calculation the fact that when they don't do something
that has not proven to be fruitful in previous examinations, that is a
major reduction in burden that is not accounted for. It is a reduction
in occupational exposure, which is also very significant.
Baselining to, say, the 1989 edition is totally contrary to
any other code that is used in other areas, like fire protection and
electrical, where it is assumed that the later editions and addenda of
the code make useful contributions to safety, which is an overwhelming
term, and it includes occupational exposure as well.
I believe, and I believe other members of the staff believe,
that this is simply a mistake; it is a mistake to remove the 10-year
update.
DR. SEALE: Thank you very much.
DR. SHACK: Thank you very much.
[Recess.]
DR. SHACK: We are back in session. Mike Mayfield is going
to talk to about PTS.
MR. MAYFIELD: Thank you.
[Slides.]
MR. MAYFIELD: Yesterday afternoon, when Ed Hackett was in,
he mentioned several times all the great things we are going to tell you
about this morning. If he oversold your expectations, we will call him
back down to explain it.
What I am going to talk to you about this morning is a
little of the background information. I suspect all of you know, but I
wanted to start there.
The goal for this activity.
Why we think there is a success path.
What our approach to it is going to be.
And some milestones and schedule for the program.
PTS is an overcooling transient that is typically taken as a
rapid cooldown of the vessel, either with concurrent high system
pressure or with repressurization. There have been operational
transients. Perhaps the most famous of them was the Rancho Seco
overcooling transient.
In 1985 the NRC finally promulgated a PTS rule, 10 CFR
50.61. It established some screening criteria for axial welds and plate
and then a separate one for circumferential welds.
If the pressure vessel is projected to exceed the screening
criteria, the licensee is first to implement a flux reduction program.
If that doesn't get them to stay below the PTS screening criteria, then
they would have to do an analysis and submit that three years in advance
of exceeding the screening criteria to demonstrate that the vessel could
be operated safely beyond the screening criteria.
In the 1982 to 1986 time frame Research funded 3
plant-specific studies to look at pressurized thermal shock, what would
have to go into the plant-specific analysis. That led to the
promulgation of RG 1.154 that lays out the plant-specific analysis
methodology. It talks about it being a format and content guide, but it
really does lay out what would have to be in a plant-specific analysis.
The staff had always intended that the PTS rule and the
regulatory guide would be something that would lay out a process for
getting beyond the PTS screening criteria. The notion is you would come
in in advance of exceeding the screening criteria. You would deal with
plant-specific considerations and then look at what changes might be
useful in the plant, for example, heating, safety injection, tank
temperature, so that you could mitigate the cooldown in such a way that
you could continue to operate the plant.
The staff considered the screening criteria as trip wires as
opposed to hard limits. When the Yankee Rowe plant came in, the
embrittlement estimates had them some 30 degrees or more over the PTS
screening criteria. The whole analysis sort of went in the drink,
because we were now trying to use the reg guide in a situation that it
really wasn't designed to deal with. They were already over the
screening criteria and we were playing catch-up.
The net result of all of that is that a lot of people in the
industry, and I suspect some on the staff, had started viewing the
screening criteria as hard limits. Licensees go to some considerable
lengths to avoid those limits, and there is also an impact then as
people make license renewal decisions: Oh dear, am I going to have
pressure vessel problems? Do I have a PTS problem? Could I ever have a
PTS problem if the staff made some further adjustments, for example, to
RG 1.99?
As we looked at this, it appeared to us that there had been
a fair bit of heartburn associated with the rule and the guide. So we
thought we probably ought to take that on and revisit it.
I want to talk a little bit about what is in the guide and
the current technology base.
The integrated pressurized thermal shock studies, IPTS,
involves 3 PWR plants, one from each of the vendors: Oconee for B&W;
Calvert Cliffs for CE; and H.B. Robinson for the Westinghouse designs.
We list here the 3 NUREGs and their publication dates.
When you go back and look at those reports, you find some
general classes of transients that cause pressurized thermal shock. You
see small-break LOCAs; you see steam line breaks; you see steam
generator overfeed; reactor trip with a stuck-open secondary.
So basically you are talking about a primary site break that
leads to a rapid cooldown; the repressurization comes; if the operator
can isolate that break, safety injection picks back up and you drive the
system pressure back up.
The other kind of transients you see are secondary side
events where you get a rapid cooldown of the secondary and then an
associated cooldown of the primary.
This top bullet is just going back to the notion that the
dominant PTS transients are the ones where we have repressurization
fairly late in the transient. So the vessel has gotten cold; you have
established some fairly high thermal stresses across the wall; and then
you repressurize the vessel.
Qualitatively, the initial thermal stresses get you a crack
initiation, and then the pressure stress that comes with the
repressurization will cause the crack to drive through the wall. That
is sort of the phenomenology that you see.
Sensitivity analyses that we did show that the flaw-related
variables are the most significant, and that is driven largely by the
level of conservatism that was in the choices for the flaw variables.
So what did we do to make the flaws look so bad? At the
point in time we did the IPTS studies we took what were treated as best
estimates on the flaw distribution, but they were known to be
conservative. For example, all the flaws were placed on the inner
surface of the vessel. We know that to be the most conservative.
Hackett talked about that a little bit yesterday.
The flaw size and flaw density distributions were assumed to
follow this Marshall distribution. I will talk a little bit more about
that as we go. That distribution was based on expert judgment based on
early vintage NDE data where they tended to gate out the first inch of
material. So you lost that inner inch of material. You don't really
know what is there from those early studies.
It also lumped nuclear and non-nuclear vessel data. So
you've got this real mixture of information that went into the Marshall
distribution. We knew then and we can now confirm that those data were
not appropriate for this application.
There were some other factors that were found to be
important. The embrittlement variables, both the shift and the
reference temperature as well as the unirradiated reference temperature,
were found to be sort of the second most important variable.
The fracture toughness characterization, the K1c curve, was
sort of the third most important.
Finally, the thermal-hydraulic transient data came in.
Then there is a list of other variables of lesser
significance. But the flaw-related variables dominated the whole
business.
DR. FONTANA: Is that in order of importance?
MR. MAYFIELD: Yes. It starts with the flaws and then
pretty much runs down that list.
Most of what I am going to tell you this morning are
generalizations that come from those three studies and some other work
we have done. It gets to be somewhat plant specific. So you could see
some swapping around. The flaw variables dominate in all of these.
That is one that we know for a fact. The relative importance of some of
these others moves around a bit vessel to vessel.
What are we trying to do with this?
Our overall goal is to promulgate a risk-informed revision
to 50.61. That, of course, involves both Research to develop a
technical basis and then NRR to promulgate the rules.
The Research goals are, first, to develop the technical
basis for this fundamental change. That would build on a lot of the
work we have been doing over the last several years with some real
emphasis on the stuff in the last five or so years.
Secondly, to approach this project as a full participatory
project bringing input in from the industry and other technical experts.
We want to make sure that everybody that has something useful to say has
a place at the table.
Finally, we want to achieve a common understanding among the
stakeholders about the proposed revisions, that they are practical, that
they are technically credible, they are cost-effective to implement, and
that they are scrutable.
DR. SHACK: You've got every ACRS buzzword in the world
there.
MR. MAYFIELD: We have been reviewing the transcripts.
[Laughter.]
MR. MAYFIELD: Those are important issues. It is some
feedback we have heard not just from the committee, but also from the
industry and others: What are you people off doing? We don't know what
you mean. We don't know why you said that.
Then when we revisit some of the SECY papers, some of the
statements of consideration, occasionally the staff is left saying, what
does that mean?
So what we want to try to do here, at least our goal, is to
five years from now or ten years from now people that look at the
documents will understand how we got where we are and the basis for it.
It was a noble goal. Why do we think we would even get
there?
Research results have made some considerable improvement in
the understanding of these critical variables. We now believe that we
can develop a technically defensible approach to dealing with the flaw
variables.
Ed talked yesterday about the investigations of the PVRUF
vessel and the flaw distributions being put together. This is
cumulative distribution of flaws as a function of flaw depth, the lower
curve being the original Marshall distribution, the upper curve being
the information from the PVRUF vessel; those flaws with an inner tip
between the cladding and about 1/8 of the vessel thickness.
What you see is a much larger proportion of small flaws than
what you have with the Marshall distribution. The range of interest in
flaw sizes for pressurized thermal shock is sort of in the 1/2 to 1
inch. If you get real deep, you don't care, because the thermal
stresses don't do much. If you are real shallow, you don't care,
because the thermal stresses don't do much.
What you find is that the much larger fraction of the flaws
are coming outside of the size range of interest with the PVRUF
distribution.
The other thing that you would do and that we are working on
is to figure out how to distribute the flaws through thickness. We have
that information from the PVRUF vessel. So we are looking at how do you
generalize that vessel-specific information into a flaw distribution
both in terms of size and location that would be suitable for these
probabilistic analyses.
DR. SHACK: How many feet of PVRUF are we talking about?
How many feet of weld are we looking at here?
MR. MAYFIELD: Shah, do you remember?
MR. MALIK: Twenty meters of weld.
MR. MAYFIELD: We also have acquired not quite that much. I
think there were about 11 feet at the Midland weld. We also have the
material from the Shoreham vessel. We got some material from River
Bend. I think that is all of them. So we are going to continue the
work on the flaw distribution.
We felt the PVRUF was representative. PVRUF was a
Combustion Engineering fabricated vessel and it was one of the last two
or three they made. We acquired it because the plant that was destined
for was canceled. Oak Ridge bought it and had it shipped, which was
quite a story in and of itself. Insurance, what happens if the barge
sinks, and that sort of thing got interesting.
DR. KRESS: They brought it up the Tennessee River to
Chattanooga.
MR. MAYFIELD: Yes. It is quite an undertaking just to get
these big vessels shipped. If you never had the opportunity to see any
of these, the way they handle these vessels is really remarkable.
At any rate, we acquired that vessel and ran out of things
to do with it. So we starting doing the nondestructive examination.
It's a late vintage vessel. So there was some concern about, well, what
about the earlier vintage vessels? Would we expect to see any real
difference in the flaw distribution?
Statistically we see some difference between the Midland
data and the PVRUF data, Midland being a B&W fabricated vessel, much
earlier vintage vessel. Some difference, but there is a lot more data
off PVRUF than there was off Midland. We are still reconciling some of
these pieces of information.
The scatter band is not that large. It is not so large that
you have to throw your hands up and say, well, distributions have to be
vessel specific. That is not where we are, but we have got some work to
do yet in terms of what would be the distribution when actually doing
this analysis.
DR. SHACK: That cumulative distribution you showed, is that
based on the nondestructive examination, the NDE results for the flaw
sizes that you had?
MR. MAYFIELD: It's based on the SAFT UT results. They are
off now doing some destructive examination to confirm that. It looks
pretty good. We have done a lot of work with SAFT UT on sizing
accuracy. So we felt reasonably comfortable going in. This is not the
sort of inspection you would see in the field. This was people up close
and personal with that vessel where they could scan and go back over
regions and acquire the data. If they needed to, they could reacquire
the data. We are pretty comfortable with that.
DR. SEALE: Mike, one wonders how risk averse one should be.
It's risky to get another vessel to look at it, in a sense.
DR. SHACK: Risky to the budget, for sure.
DR. SEALE: That's the other part of it. There are several
vessels that ought to be coming available from various plants. Is there
any thought of trying to get another candidate or to look at these in
more detail?
MR. MAYFIELD: We wanted to stay with the ones we have
simply because they are all unirradiated and we can get up close and
personal with them. As you start getting on the more highly irradiated
vessels the quality of the UT results -- I want to be careful to not
imply that what we are doing in any way relates to what would be done
during a traditional ISI. Consequently, our confidence in the sizing is
much higher than you could reasonably expect with people working off the
end of an ISI mast with all the complications that go with handling an
irradiated vessel.
Is there any interest? Risk averse maybe is the right
consideration there. We have been interested in getting as much
material as we can simply because we tend to reduce the uncertainties at
the end of the analysis.
We did some analysis. I will show you a plot in a second.
We are seeing about an order of magnitude in probability of thru-wall
cracking for this example case. We know that there are additional
conservatisms in the analysis.
The flaw size and density for the plate material. We really
haven't done much on that and we were staying with some conservative
assumptions.
We are still dealing with flaw density in welds and the
number of surface-breaking flaws you would actually have to consider.
If you inspect 20 meters of weld and you find no surface-breaking
defects, statistically you can't say there aren't any. There are ways
of estimating how many surface-breaking defects you should consider in
that material.
We need to impose an up-to-date fluence map, because we just
assumed that peak fluence occurs uniformly around the vessel.
And you would need accurate information on the chemistry so
that we could influence the reference null ductility values.
Finally, there is reason to believe the PTS screening
criteria can be increased significantly by simply reducing some of the
excessive conservatism in the analysis.
This is the result of one of the example calculations. This
is from the IPTS study for the Calvert Cliffs plant. I want to
emphasize that this does not represent Calvert Cliffs as it sits today.
This is simply the vessel that we were working on at that point in time.
The upper curve represents essentially what you would get
out of RG 1.154 as it sits today. The lower curve is what you get out
of these example calculations. So you pick up about an order of
magnitude reduction at 32 EFPY between what you would get out of the reg
guide calculation and where we can get without having to work at it very
hard.
This is basically just the flaw distribution information and
making some distribution of the flaws through thickness. By the time we
impose the fluence map, we sharpen up the embrittlement estimates, and
we do some of these other things, we think it's conceivable you could
get another order out of this analysis.
DR. SHACK: I don't need another order.
MR. MAYFIELD: Let me come back to that. In principle you
don't need another order, but what we don't want to do is revisit this
in another ten years; we want to try and do it right today and get this
as sharp as you can justify it.
The approach we are taking will be basically a two-track
approach.
The bulk of the work will be determining the appropriate
screening criteria.
The other piece of it is to look at the numerous other
aspects of the existing rule that need to be revisited: the
embrittlement correlations that Ed Hackett talked about a bit yesterday;
what margins need to be included in this analysis. The idea isn't to
embed conservatism at each stage but rather to do the analysis using
your best technical input and then put a known margin at the end of it.
There are issues such as credibility collapse for the
surveillance data that need to be addressed. There are several of these
points that need to be dealt with.
As I mentioned, we see this as a fully participative process
that would involve the public and industry at each stage. This isn't
something where the staff goes behind closed doors, does some stuff, and
then comes back out and pronounces what the answer is. Rather, people
would be involved all along.
We are hoping that would include this committee. I think we
are scheduled with the full Committee in April. The notion would be to
inform the committee, seek your comments, if any, as we go along, and
then provide additional status briefings during the course of the
project.
What we don't want to do is get to the end and then have
somebody identify a fundamental flaw that should have been picked up in
the early stages.
DR. KRESS: No pun intended.
MR. MAYFIELD: Yes.
This little flow chart tries to lay out -- it is probably
difficult to read. I have to get my glasses fixed.
DR. SHACK: It's not a whole lot better on the page either.
[Laughter.]
MR. MAYFIELD: I started to bring the 11 by 17's and the
just didn't get them printed.
We would start up at the top using some public meetings to
identify the major issues that need to be addressed. We have those
meetings scheduled, at least the first cut at them. One we have already
had; the second one we would have April 20th and 21st here in
Washington.
We would be collecting the output from those two meetings,
assimilating that, and that would lead to some specific analyses and
additional meetings, so that we could come up with the list of questions
that need to be resolved.
DR. KRESS: Who shows up at a public meeting like that?
MR. MAYFIELD: By and large the industry.
DR. KRESS: People who know about this thing.
MR. MAYFIELD: We notice the meetings. Occasionally we will
get some folks not directly associated with the industry. We will see
vendors; we will see some of the A&E show up.
DR. KRESS: But these are technically minded people.
MR. MAYFIELD: They tend to be technical people that show up
at these meetings.
Right after the Yankee Rowe experience, when we first
started doing these kinds of public meetings, we had a couple of the UCS
folks show up; we had a couple of reporters show up. For reasons of
their own, they apparently decided their resources were best put
elsewhere. We see more involvement from a broader range as you get
towards the end of some of these things, but when you are dealing with
the detailed technical meetings, by and large it's the engineering
sciences folks that show up.
So we have come up with our list of questions. This other
path would be to deal with these other issues in the rule, the
embrittlement correlations, credibility criteria.
DR. KRESS: Are you going to talk about the square on the
right?
MR. MAYFIELD: That's why I started over here, so I can come
back to this one.
These are the more mundane parts of what we want to change
on the rule. Although they may be more difficult to deal with, they are
the more routine considerations.
Over here is to reevaluate the technical basis for that 5
times 10 to the minus 6 target. That one seems to capture everybody's
attention. That is an area that Mark Cunningham and his guys will lead.
The notion is that we would base that criteria on some of
the other risk-informed initiatives. I don't want to leave you the
impression we have decided on this, but as an example, work backwards
from either the CDF or the LERF criteria in RG 1.174 to come up with
what would be an appropriate vessel failure criteria.
DR. KRESS: This is one set of sequences among many. Are
you going to distribute the LERF among some sequences? Is this the
idea, Mark?
MR. CUNNINGHAM: Mark Cunningham from Research. Yes, that
is exactly the type of issue we will have to talk about, coming from
1.174 or coming from the regulatory analysis guidelines for backfits,
which are interrelated at this point; talk about what does it mean for a
relaxation like this for a subset, if you will, of whole plant risk.
DR. KRESS: That was my issue.
MR. CUNNINGHAM: I envision a number of interesting meetings
with the committee talking about just that subject.
DR. KRESS: We look forward to it.
MR. MAYFIELD: Mark will be in front of the committee for
those.
DR. FONTANA: A related question on the curve that you
showed for Calvert Cliffs. This frequency of vessel failure, does that
assume the whole spectrum of initiating events that can get you into
this kind of a failure?
MR. MAYFIELD: For this the original study did. They went
through a very detailed system evaluation and bin events. For this
particular calculation, the small break LOCA gets you 90 percent of so,
and they just ratio it up. The original IPTS study and what we would do
in this overall study, we will go back and look at the full range.
DR. FONTANA: You were discussing the acceptance. That
heavy line is an acceptance line?
MR. MAYFIELD: That's the 5 times 10 to the minus 6 that we
have today. Bill said, well, you've done enough. If, for example, this
came down tomorrow like 10 to the minus 6, now I'd like to pick up some
more. You want to sharpen this analysis and make sure we are making the
best use we can of all of the available information.
DR. KRESS: You are reevaluating this solid line, too.
MR. MAYFIELD: We are reevaluating the solid line as well as
what would go into these others.
DR. FONTANA: I hate to bring up ancient history, but back
in WASH-1400 weren't they guessing around 10 to the minus 7 for vessel
failure?
MR. CUNNINGHAM: Yes. I don't think people recognized in
WASH-1400 PTS was not on the table at all. It was by and large a
traditional, spontaneous, if you will, rupture of the vessel.
MR. MAYFIELD: This box will be a key input to everything
else we are going to do in terms of reevaluating screening criteria. We
are anticipating that will involve a fair bit of public discussion.
In one of the schedules that we had put out for this program
we had that activity completing much later in the program. One of the
comments we got from the industry meeting we had a week or two ago was
that that was coming too late. We agreed. We have moved that up
significantly.
This is also something that potentially will continue during
the whole course of this program, but we needed to focus on that
activity much earlier. So we have moved that up much further in the
schedule.
DR. SEALE: It kind of washes over into these other things
quite a bit.
MR. MAYFIELD: Absolutely. This is something I anticipate
there will be continuing dialogue on, but you need to frame it up front
and make sure you understand the context of that answer and then look at
the rest of the program.
DR. KRESS: That is an interesting box over there. That is
where defense in depth has a head-on collision with risk criteria.
MR. MAYFIELD: The other part of the approach is in the
middle.
One of the things that we have been emphasizing was a
mistake that was made in some of the earlier studies, a mistake that we
have made a couple of different times, to not involve the statisticians
up front.
We know we want to do a detailed sensitivity and uncertainty
analysis. We don't want to wait until the mechanical engineers and the
metallurgists are done having fun and then bring those folks in to make
sure that we can put together the sensitivity and uncertainty analysis.
Put them in up front; make sure we understand the type of analysis we
need to do and the level of rigor we need to do, so that at the end we
can have a sensitivity and uncertainty analysis that is believable.
DR. KRESS: I would recommend you draw some sort of line to
that box and the other box. I think uncertainty is going to be
important in your acceptance criteria.
MR. CUNNINGHAM: Yes.
MR. MAYFIELD: Coming out of listing the questions, making
sure that we have an approach framed in a way that will be credible at
the end of the day, we will look at three pieces, identify and bin the
events, meaning going back and revisiting existing PRAs for the plants.
Our hope is to build on the IPTS 3-plant study. However, Palisades has
volunteered to make themselves the fourth plant in the study and to be
totally participative. That was a nice addition.
One of the first things that Mark and his guys have to do is
to look at the existing PRAs, contrast those that were used in the
original IPTS study, and see what additional work we need to do to bring
that list of events up to reflect the current plant. It may be a lot of
work; it may not be very much work. It just depends on how the systems
have changed over time.
DR. KRESS: Going to your box on sensitivity and
uncertainty, since you already know that it is the flaw parameters that
are most sensitive, more than likely that is where you would need to --
MR. CUNNINGHAM: Probably the most uncertain.
DR. KRESS: Most uncertain also. I don't think you can rely
on the data spread to get an uncertainty in this case. Your only
recourse is probably going to be expert opinion. It sounds to me like
you are back to Marshall.
MR. MAYFIELD: Except we've got a lot better information to
feed that panel.
DR. KRESS: Your experts can be fed with better stuff, and
you can sharpen up this expert opinion.
MR. MAYFIELD: I think that is exactly the case. We had to
start somewhere. Making the next step based on the PVRUF information
just to see if we were even close, we felt that, yes, we were close.
That next step is how do you generalize that vessel-specific information
into something that can be used in this kind of program? Do you do that
with expert elicitation? Exactly how do you get there?
DR. KRESS: I think you are going to have to.
DR. SEALE: The Marshall distribution was terribly
subjective. I think everybody knows that it was terribly subjective.
DR. KRESS: I think Dana would have been appalled at the
expert elicitation that went into that.
DR. SEALE: They loaded the deck because they had a loaded
agenda.
MR. MAYFIELD: Yes. The other thing is that the people from
HSE came in and told us ten years ago that the Marshall distribution
wasn't credible. There wasn't anything else on the table. We really
didn't have anything else to use, but even the people that generated
that distribution have been disavowing the thing for a long time.
This is the stuff we have done on PVRUF and the continuing
work on that activity, some of the really major steps forward in getting
a look at what really can be in a reactor pressure vessel.
DR. KRESS: I have to say it looks like a good approach,
Mike. I think you are going to have problems in a couple of the boxes.
These problems are tough, but you guys can do it.
DR. SEALE: They look like the right boxes.
DR. KRESS: They got the right boxes. That's for sure.
DR. SHACK: Let me ask you a question you will get to answer
next week. That box on the right is clearly NRC turf. What about the
rest of this? Why isn't industry doing it?
MR. MAYFIELD: Industry is certainly involved significantly.
Bob Hardies is here with us again today. The industry is involved. The
reason the NRC takes this on is it is NRC's regulation.
DR. KRESS: I think that is a reasonable criteria.
DR. SHACK: We'll see how it flies next week.
MR. MAYFIELD: That same question can be posed on virtually
anything we do. The criteria we have looked at is that it is in fact
NRC's regulation. There is in fact a Commission initiative to reduce
unnecessary burden. There is also a Commission initiative to
risk-inform Part 50. We believe that this activity is consistent with
both of those initiatives. We can reduce excessive conservatism and it
is consistent with risk-informing Part 50.
DR. SEALE: And if we are going to have the gall to continue
to license plants on a full-term basis or the next 20-year term basis
instead of in a progressive way that a lot of the countries and the rest
of the world do, then we need these data in a mature fashion so that we
don't ask ourselves 50 years instead of 60 years out is the whole story
there. That is clearly an NRC imperative at this time.
MR. MAYFIELD: This is something where we are the ones that
ultimately get faced with making the yes-or-no decision on continued
operation of the plant. The industry has a vested interest in looking
at license renewal and other major plant considerations about whether
they want to continue operation. That is why there is interest on both
sides; both sides are participating in this activity; both sides are
putting money on the table in terms of resources that are being brought
to bear on this program.
DR. KRESS: When you come out with a new screening criteria,
will it be like the old one in the sense that it will be a flag, or is
it actually going to be a limit?
MR. MAYFIELD: It is interesting you raise that. That is
something we talked about just a little bit. There is a presumption in
some of these lower boxes that the rule is going to fundamentally look
about like it looks today; just the numbers would change.
One of the things that Jack Strosnider and I have talked
about, and I think Mark and I a little bit, is, should the nature of the
rule change? Jack has commented a couple of times that 50.61 was one of
the very first efforts by the staff to be risk informed and performance
based.
By today's criteria, you just scoff at that; it's laughable.
But the fact is the screening criteria were approached from a
probabilistic risk assessment. Certainly not by the standards that Mark
and his colleagues would apply today, but in the late 1970s or early
1980s that was the kind of thinking that went into it. It is
performance based in terms of there is a trip wire screening criteria
that would lead you to do something else rather than just go shut the
plant down.
The idea is laughable by today's standards, but it was an
approach that dealt with some of those concepts very early on. We would
like to try and get there. The formula in the rule where you have got
the screening criteria and then you go do something else, does that
still hold water? That is one of the things we want to talk about as we
go.
Again, there is a presumption in some of these lower boxes
that that is how it is going to go, but we are not locked into that.
DR. KRESS: If you follow the attributes that define
performance based, you would go that way. You want your performance
criteria to be something that you can live with.
MR. MAYFIELD: That is the notion. Again, you will see a
lot of boxes down through here talking about public meetings to gain
shared understanding. That doesn't necessarily imply agreement, but
everybody needs to understand where we are and how the staff is going
about making its decisions.
There is a final box over here, a meeting to resolve any
outstanding issues. This would be sort of the last look at what the
staff is going to consider as it develops the proposed technical basis
that would be provided to the RES and NRR office directors as a proposal
for how you could go about revising the PTS rule.
Going back to your question of what would it look like and
can it be performance based, a specific regulatory approach is what we
would be proposing to the office directors and then along the way
dealing with the notion of how do you make it performance based and what
should it really look like.
Finally, we started laying out a schedule for this. Most of
what is listed here is the public meetings. We figure those are the big
targets where we need to have specific things done. So we listed those
as the major milestones.
The notion here is that we would be done with the activity
and would propose the revisions to the rule to the office directors by
the end of 2001. That provides NRR roughly two years to promulgate a
rule change. That would support Palisades as the first plant that has a
PTS problem, and they run into the screening criteria in December of
2003.
If nothing else is done with that plant, they have PTS
screening problems December of 2003. This schedule is consistent with
getting a rule change promulgated in time to support the staff's
consideration as well as licensee consideration.
DR. SEALE: Have you thought about how many times you are
going to want to meet with the ACRS on this thing in the next six months
to a year?
MR. MAYFIELD: I'm guessing at least once more. It is going
to depend on how quickly this starts to go together.
DR. SEALE: You have 10/99. There is a lot of input, I
would think, that the committee would have into that right-hand box over
there, and 10/99 is not that far away.
MR. MAYFIELD: It is not that far away. I guess I do need
to defer to Mark on this. I don't know if he has thought much about how
many times we would be back with the committee. Certainly at 10/99 we
would be back.
MR. CUNNINGHAM: I envision this discussion, particularly
the right-hand box, being of the style like we did as we developed
1.174.
DR. SEALE: In other words, fairly interactive.
DR. KRESS: We could help you decide the approach.
MR. CUNNINGHAM: Absolutely. I'm sure we would like to do
that.
DR. KRESS: I'm sure we would like to do that.
MR. CUNNINGHAM: The ACRS was very much involved in the
original decision on why 5 times 10 to the minus 6 was the right number
to use. I would envision several subcommittee meetings just on that
issue this spring and summer.
DR. KRESS: I would suggest we have meetings on the box on
the uncertainty too, because that is where you are likely to hear some
criticism from some members of the committee.
MR. MAYFIELD: Just don't invite them.
[Laughter.]
DR. KRESS: I thought about that.
DR. SEALE: You are doing us a service too. We have been
noodling this question around for a while now, as you are perfectly
aware, but we have been awfully general, unfortunately, or fortunately,
however you want to look at it. The time has come to grab the nettle,
if you will, and this is certainly an important area to do that in. I
think this is happening at the right time as far as ACRS is concerned.
DR. KRESS: We have on our agenda how to do deal with
defense in depth in a risk-informed regulatory world. This is a
beautiful example right here.
DR. SEALE: This is it.
DR. KRESS: This is defense in depth and risk informed right
together. So it fits right in.
DR. SEALE: We'll come looking for you if you haven't come
here.
DR. FONTANA: How many man-years of effort does it take to
do something like this in this schedule you have got here?
MR. MAYFIELD: We have done those estimates and I've just
gone brain dead.
Shah, do you remember?
MR. MALIK: Two to three man-years
MR. MAYFIELD: So it's about a man-year per year. Maybe
it's a couple of man-years per year.
DR. KRESS: Are you guys going to do the reevaluation in
house, the IPTS?
MR. MAYFIELD: It will be a mixture of in-house and
contractor support. Part of it will be in house. We simply don't have
the contractor support resources to keep running out and doing
everything with contractors.
DR. SHACK: I don't see any box that says improve FAVOR
here. You think the fracture mechanics code is basically in a state
that it will do this job?
MR. MAYFIELD: There will almost certainly be some level of
tinkering. By tinkering, I mean making specific changes to address
specific questions. We are not having to reinvent the FAVOR code. We
are pretty well satisfied with that code. It has had a lot of look from
a number of reviewers.
DR. SHACK: You don't need to incorporate biaxial effects?
MR. MAYFIELD: Some of that is already available. That is
one of the considerations that Shah and his group will be looking at,
exactly how you want to do that, exactly what changes need to be made.
Out of this box and some of these, there will be specific
questions listed and specific approaches on how we are going to address
those questions that will be coming down. As we need to make changes to
FAVOR, they would be made on the input side to that lower box. It is
our intent to use FAVOR essentially as it sits today.
DR. KRESS: That middle box called thermal hydraulics, will
Theofanous be part of that, do you think?
MR. MAYFIELD: We have had some conversations with Professor
Theofanous. It is my understanding Farouk Eltawila, who unfortunately
couldn't be with us this morning, he and his staff will be leading up
that piece of it. The last understanding I had was that we would
anticipate using Professor Theofanous and others as consultants to the
activity but not as the leader of the activity.
DR. SEALE: This expert elicitation thing that Tom referred
to as the input over here on the right-hand box is something that is
going to be interesting to see.
MR. MAYFIELD: That one will be interesting. The other one
that we think is going to be interesting is to look at uncertainty
bounds on the thermal-hydraulics calculations. We know that we cannot
do a brute force Monte Carlo approach to calculate uncertainty bounds on
the thermal-hydraulics for each of the transients. That is just not a
practical matter to undertake. How you then go about estimating the
uncertainty bounds in a credible manner, that is one that Farouk and his
guys are talking about.
DR. SEALE: Graham will love that.
DR. KRESS: Yes.
DR. SHACK: Something for everybody in this.
MR. MAYFIELD: That was one of the reasons we wanted to meet
both with this subcommittee and then with the full committee early on,
so that you see what we are off doing; you can identify where you would
like to be more involved or less involved, rather than just wait until
we got to the end of the day and then ask you to try and review several
years worth of work.
DR. KRESS: We certainly like all your boxes. There are
three of them in particular. The thermal-hydraulic box, the uncertainty
box, and the RG 1.174 box are ones we would be particularly interested
in.
DR. SHACK: Is there a danger you are going to have a
disconnect here on your revision to the embrittlement rule? This is
going to come ahead of the embrittlement.
MR. MAYFIELD: One of the things we will do here is to
include those new correlations and include that as part of the
sensitivity and uncertainty analysis. We know there are going to be a
number of things. Hackett talked yesterday a good bit about the master
curve and that approach versus the K1c curve. That is something that
Bob Hardies and I have talked about a little bit.
The notion would be accept the fact that there are two
different notions on how to deal with these things: Accept the fact
that there is some uncertainty in the embrittlement correlation, and
deal with that rigorously. Rather than try and dance your way around it,
here it is, and here is what kind of impact that has.
DR. SHACK: Then you just build it into the uncertainty
analysis.
MR. MAYFIELD: Build it in and deal with it explicitly.
I think we mentioned yesterday the notion about taking on
the P-T limits as a second question. I think you can see that a lot of
the elements of this approach could equally apply to the P-T limits. So
when we put the question on the table with the industry at our last
meeting, the notion was to get people to think about what might go into
a revision to the P-T limit approach and then keep that in mind as we
take on these activities. There may be some tweak we can make here and
there. It would be a little more than we need to do just for this
purpose but that would satisfy that end goal.
DR. SEALE: Sharpen up the learning curve.
MR. MAYFIELD: The best of all possible worlds is at the end
of this we are proposing not just changes to 50.61 but changes to
Appendix G as well. That would be a wonderful end to this. I am not
overly optimistic that we are going to get there, but if keep those
things in mind, then we are going to stand a better chance.
DR. KRESS: According to what we heard yesterday, the
changes to Appendix G would be worth more to industry.
MR. MAYFIELD: I think in terms of operating costs, that is
true. PTS tends to be a plant killer issue.
DR. KRESS: That's really a killer. You're right.
MR. MAYFIELD: Whereas Appendix G is a routine operation
issue.
DR. KRESS: They are different.
MR. MAYFIELD: That's why we want to make sure we are
getting both of them.
That was all I had, Mr. Chairman.
MR. DUDLEY: Do you have additional needs for experiments?
MR. MAYFIELD: Not so much in the fracture area. We think
that is pretty well in hand. Farouk has some interest in trying to
experimentally assess flow stagnation. That is in the
thermal-hydraulics part of this analysis, the notion about whether if
you really stagnate flow it tends to drive the cooldown severity. If
you continue circulation, the cooldowns are significantly mitigated.
DR. KRESS: That could be a real significant uncertainty.
MR. MAYFIELD: Farouk has the notion of trying to build in
some limited experimentation in one of the flows they have used
previously. Those are budgeting considerations, how much of that can we
really do and can it be done in a timely manner. We are still
struggling with budget formulation, as I think you all know. There is
an interest in that kind of experimentation. We believe the rest of it
we are in pretty good shape. It's a fair bit of work to get to the end
product, but we think we have all the fundamental pieces in place.
Anything else?
[No response.]
MR. MAYFIELD: Thank you.
DR. SHACK: Thank you very much.
DR. FONTANA: Let me add that we have seen some
presentations where we wished that they were planned as well as this is.
MR. MAYFIELD: We hope you are saying nice things to us six
months from now.
DR. SEALE: There is no question that you are tackling the
problems that count.
[Whereupon at 11:15 a.m. the meeting was concluded.]
Page Last Reviewed/Updated Tuesday, July 12, 2016