480th ACRS Meeting - March 2, 2001
Official Transcript of Proceedings
NUCLEAR REGULATORY COMMISSION
Title: Advisory Committee on Reactor Safeguards
480th Meeting
Docket Number: (not applicable)
Location: Rockville, Maryland
Date: Friday, March 2, 2001
Work Order No.: NRC-097 Pages 235-325
NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers
1323 Rhode Island Avenue, N.W.
Washington, D.C. 20005
(202) 234-4433. UNITED STATES OF AMERICA
NUCLEAR REGULATORY COMMISSION
+ + + + +
480TH MEETING
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
(ACRS)
+ + + + +
FRIDAY
MARCH 2, 20001
+ + + + +
ROCKVILLE, MARYLAND
+ + + + +
The Advisory Committee met at the Nuclear
Regulatory Commission, Two White Flint North, Room
T2B3, 11545 Rockville Pike, at 8:30 a.m., Dr. George
Apostolakis, Chairman, presiding.
COMMITTEE MEMBERS:
GEORGE APOSTOLAKIS
Chairman
MARIO V. BONACA
Vice Chairman
THOMAS S. KRESS
Member
GRAHAM LEITCH
Member
DANA A. POWERS
Member
ROBERT J. SEALE
Member
WILLIAM J. SHACK
Member . COMMITTEE MEMBERS: (cont.)
JOHN D. SIEBER
Member
ROBERT E. UHRIG
Member
GRAHAM B. WALLIS
Member
F. PETER FORD
Invited Guest
ALSO PRESENT:
KAREN COTTON
GENE CARPENTER
STEVE DOCTOR
BILL BATEMAN
DEBBIE JENSEN
BILLY CROWLEY
LARRY MATTHEWS
VAUGHN WAGONER
I N D E X
AGENDA ITEM PAGE
Briefing Event at V.C. Summer Nuclear Station
Opening Remarks
Chairman George Apostolakis. . . . . . . . 238
Introduction of Participants
William J. Shack . . . . . . . . . . . . . 238
V.C. Summer Background, Karen Cotton . . . . . . 239
Discussion of Technical Review and Future . . . 247
Activities, Gene Carpenter
Materials Reliability Program, Larry Matthews. . 285
Assessment Committee . . . . . . . . . . . . . 301
Inspection Committee . . . . . . . . . . . . . 303
Repair and Mitigation Committee. . . . . . . . 315
Adjournment
P-R-O-C-E-E-D-I-N-G-S
(10:01 a.m.)
CHAIRMAN APOSTOLAKIS: The next item is a
briefing event at V.C. Summer Nuclear Station. Dr.
Shack is the lead member on this.
MEMBER SHACK: I'm sure of the members are
aware that they found a crack in the weld material of
the reactor coolant hot leg piping system at V.C.
Summer.
The crack occurred in alloy 182 weld
metal, which is a high nickel weld metal that they
typically use essentially as kind of a buttering when
you're joining a Ferritic component, in this case a
pressure vessel nozzle, to the stainless steel piping
and essentially it minimizes the mismatch in thermal
expansion and reduces thermal stresses.
We know in the past that PWR primary
coolant piping systems have been very reliable and
we've found very little evidence of cracking in those
systems. They've been approved for leak-before-break,
largely because of that reliable experience, and so
there's some interest here in the general nature of
whether this experience can be generalized to other
systems. And, again we wanted to understand just how
well, for example, inspections can be done in other
plants.
Joining us on the phone today, in addition
to the staff at the front of the table is a
contractor, Dr. Steve Doctor from Pacific Northwest
Laboratory, who's an ultrasonic UT for Dr. Wallis,
expert on crack detection. And I guess Gene, you're
going to start off, Gene Carpenter.
MR. CARPENTER: Karen Cotton will be
discussing this.
MEMBER SHACK: Okay. Ms. Cotton then, you
can start the discussion.
MS. CARPENTER: Okay. First, I'd like to
say good morning. I'm Karen Cotton and I'm a
mechanical engineer and the project manager for V.C.
Summer. Gene Carpenter, from the division of
engineering will talk about the technical review and
future activities regarding the Summer crack.
Larry Matthews of Southern Nuclear, he'll
talk about the MRP. I would like to acknowledge Billy
Crowley sitting here, he's the team leader for the
special inspection team and, as you heard before, we
have Steve Doctor on the phone listening in.
I'm going to discuss the history of the
event and I'll discuss it in three parts. I'll talk
to you about the actual event, I will talk to you
about what the licensee did in response to the event
and I'll also talk about NRC's actions. Then I'll
give a brief synopsis of what was the function of the
special inspection team.
During refueling outage 12, during a
routine walk through, boron deposits were discovered
near the "A" hot leg reactor nozzle. The licensee,
what they did was they continued with their routine
outage activities but they began to investigate where
the boron was coming from. They did a PT inspection
and they discovered a four-inch crack. In the four-
inch crack, they soon found that this was only a
surface indication.
They continued and they did UT and they
did eddy current testing and they found a two-and-a-
half inch crack, and this exited through a weep hole.
Summer designated a team of industry
experts to look at the situation, and the industry
experts they looked at the repair and they looked at
evaluation of the repair. Their focus was to come up
with a root cause analysis and to come up with a
repair method.
MEMBER WALLIS: It surprises me that the
first thing that was mentioned was boron deposits. I
would think there would be all sorts of other
indications of leaks before that.
MS. COTTON: What happened was --
MEMBER WALLIS: A new activity or just
loss of fluid.
MS. COTTON: There were no other
indications of leaking. This leak was very small and
it wasn't detected through our normal leak detection.
MEMBER WALLIS: Doesn't it take a lot of
water to make much boron deposit?
MR. CARPENTER: That is correct, sir.
This is Gene Carpenter. Typically, you have very
small amounts of boron in the reactor system fluid,
and obviously there was literally hundreds of pounds
of water that had to escape before this was detected.
However, as Karen said, it was a very
tight crack and the leak rate was much below 0.1 GPO,
so they never did trigger the tech spec required 1.0
in any unidentified leakage.
MEMBER SHACK: What was the unidentified
leakage sort of in the period leading up to the
incident?
MS. COTTON: It was like 0.6.
MEMBER SHACK: Three-tenths.
MR. CARPENTER: Yes, 0.3 GPO was about the
average over the operating cycle.
MS. COTTON: The team's primary goal was
to ensure that the plant would safely start up. They
looked at all the welds, they looked at the code
requirements, they had to address all the failure
scenarios, even the worst possible case. And they
looked at all the indications and made sure that all
these indications in the other welds were evaluated.
The licensee also developed a
communications plan and this plan ensured good
communications, thorough communications with NRC, with
the other members of the nuclear industry, and also
with the community surrounding the plant.
They also took a further step and they
committed to enhance their leak detection procedures.
They decided that they would examine the B and C welds
during refueling outage 13, and they also committed to
examining all the welds during refueling outage 14.
MEMBER SHACK: When had this weld been
last inspected, or had it even been inspected?
MR. CARPENTER: Yes, it had been last
inspected in 1993 during their ten year ISI.
MS. COTTON: The licensee's activity
included we chartered a special inspection team, we
chartered and formed a communications plan. As part
of the communications plan we did a communications
team, which met on a weekly basis, bi-weekly, we met
twice a week on a weekly basis to handle all issues
dealing with the Summer crack.
We developed a web site that was specific
to just this Summer event. We issued three information
notices, the last one was February 28, was issued
February 28. We received a WCAP from Westinghouse
regarding the integrity of the B and C welds. We did
a safety evaluation regarding this and we completed
and issued the safety evaluation on the 20th of
February.
We also had five public meetings, the last
public meeting was February 15 and that was a public
exit meeting. We chose to have a public exit because
all the meetings were public and we got very good
comments from the public regarding our openness and
our willingness to involve them in this event.
The licensee's root cause analysis was
primary water stress corrosion, and this basically was
due to the susceptible material of alloy 182, coupled
with the repeated welds, or the repeated rewelding and
rework done during construction of the weld.
MEMBER WALLIS: How was the grinding
related to residual stresses? Was it that the
grinding was too gross and rough, or was it something
to do with heat generation, or what was the coupling
between grinding and residual stresses?
MR. CARPENTER: When the weld was
originally installed they had multiple weld repairs.
It took, I believe, something like 40 days to do the
complete weld repair of the Alpha hot leg nozzle weld.
In that time they basically took out the entire weld
and rewelded it in at least once.
MEMBER WALLIS: So grinding wasn't
necessarily a cause of stress at all?
MR. CARPENTER: Well, they did do a lot of
grinding out of welds, of flaws, and then rewelding.
MEMBER WALLIS: But the grinding itself
didn't cause the stresses?
MR. CARPENTER: It added to it, sir.
MEMBER WALLIS: It did?
MS. COTTON: Steve, do you want to talk
about the V-shaped? I have a slide I could put up for
you? Steve? Do you want to talk about the V welds?
I have a slide I could put up for you as for our
discussion about the root cause analysis?
DR. DOCTOR: I think the key point is the
fact that when they put in this bridge path, what this
forced them to do was basically form a double V type
of weld. And the work that EPRI had funded had shown
that the stresses on the inside were much higher with
that type of a weld design as compared a single V type
of weld design.
And so the fact that you've got a grinding
was due to remove the old material, but they used this
bridge path and it forced the design into this double
V type of design.
(Slide change)
MS. COTTON: This basically sums up the
history portion of what actually happened. Now we'll
talk about the special inspection team.
As stated in the history, a special
inspection team was chartered. The focus of the team
was to ensure that the licensee's corrective actions
were appropriate. They looked at and reviewed the
root cause determinations, and they looked at all
corrective actions activities.
The team's activities included, as I said
before, corrective action review, review of the
licensee's records, they observed the welding
processes, NDE activities. They also did on site
metallurgical analyses that reviewed this, of the
spool piece at Westinghouse hot cell labs. So the
team was pretty active and they were on site for
several weeks during this whole incident.
The team's findings were that the root
cause analysis was appropriate and acceptable and that
there were no deviations from weld requirements. From
code requirements.
MEMBER SHACK: Just on this inspection,
doing the weld techniques, do you have to go through
the piping to inspect the weld? Or is this something
that is really done just through the weld metal?
DR. DOCTOR: This is Steve Doctor. When
you perform the inspection according to the Section 11
requirement of ASME code, you're required to inspect
the inner third of the weld plus some adjacent
material on both the pipe side and the nozzle side.
Generally, this is approximately this a half-inch.
So the inspection includes both base
material, structural weld and buttering.
MEMBER SHACK: Steve, I guess I was
interested in whether, for example, if you were in a
plant that had centrifugally cast piping, in order to
inspect this weld you would have to look through the
cast piping.
DR. DOCTOR: That's correct. As a matter
of fact, if you look at the cold legs those are a
nozzle weld with buttering structural weld going to a
cast elbow. And they perform those inspections in a
similar vein. The fortunate thing is that these
inspections are conducted from the inside, so the
amount of cast materials, the very coarse grain cast
materials that you have to penetrate, is relatively
small.
If you perform the inspection from the
outside surface you have extremely long paths through
the coarse grain material, which would then make the
inspection extremely difficult.
MEMBER SHACK: Okay. Thank you.
MS. COTTON: The purpose of my talk was
just to provide a snapshot of the history of what
happened with the event. The actual event history,
the response of the licensee, NRC's actions, and to
give you some detail of what happened with the special
inspection team.
Both the special inspection team and the
staff feel that this event is beyond Summer, and that
there are further generic activities, we should look
into this further. And this will be discussed by Gene
Carpenter. Gene.
MR. CARPENTER: Thank you, Karen. First,
I'm going to go over the technical review that the
staff did of the B and C hot legs, and then I'll be
discussing some of the generic activities that the
staff is following.
The staff performed an independent
evaluation of the licensee's assessment of the B and
C nozzle legs. Now since they had physically removed
the A hot leg nozzle weld and replaced it, there
really wasn't a need to evaluate the cracks in that.
This was done by the licensees submittal,
which was the WCAP-15615 Rev 1, which is a proprietary
document and the non-proprietary version 16616 Rev 0
which is available on the web site for public
inspection.
The WCAP provided the results of the
Westinghouse UT and the eddy current examinations of
the nozzle-to-pipe welds for loops A, B and C. In
those loops, they found in five of the six nozzles,
that there were crack indications -- or I should say
eddy current indications. The C cold leg was the only
one that did not have any indications found.
These indications were evaluated based on
the destructive examination that was done on the A hot
leg nozzle that was removed and was destructively
examined at the Westinghouse hot cell facilities.
And, based on those determinations, the licensee
determined that Summer could be safely operated for
approximately two further cycles before they needed to
do any inspections or possible repairs to the existing
indications in the B and C hot legs.
MEMBER SHACK: Is eddy current an accepted
inspection technique for this kind of consideration?
MR. CARPENTER: No, it is not. They went
beyond that. The UT could not find these eddy current
indications, so basically we are going beyond the code
on this.
MEMBER SHACK: Well, how are they
estimating sizing then for these flaws?
MR. CARPENTER: The eddy current can
determine the lengths. We're doing a 2:1 aspect
ratio, so basically for a one-quarter inch long crack
indication, we're assuming that a depth of one-eighth
inch.
MEMBER POWERS: How did they possibly make
a prediction that they can operate for so long of a
period of time without fixing these things?
MR. CARPENTER: The determination was made
on the basis of the susceptible material. It was made
on the basis of the crack growth rate.
MEMBER POWERS: What is the crack growth
rate in a nozzle in Summer?
MR. CARPENTER: The crack growth rate that
was assumed.
MEMBER POWERS: I'm not interested in
assumptions. I want to know how they knew what the
correct growth rate was.
MR. CARPENTER: When they did the
examination, the destructive examination of the A hot
leg, they went in and basically - if I could flip to
that slide.
(Slide change)
MR. CARPENTER: This is a representation
of the Alpha hot leg and this is the crack that grew
through wall. The assumption -- basically
Westinghouse went through and evaluated the crack,
found that there were multiple initiation sites and
they grew together. And from that they made a
determination using a fairly extensive formula which,
if you'll pardon me for just one second - 1.4 x 10
the minus 11, K minus 9 to the 1.1
MEMBER POWERS: Show me the experimental
data, applicable to Summer, that validates that
formula.
MR. CARPENTER: I don't have that with me,
sir.
MEMBER POWERS: I mean there can't
possibly be any data that's directly applicable to
this to validate that formula - unless there have
been a lot of cracks in Summer. I mean how do you
justify an analysis like this that says oh, we can
operate for two more cycles based on this magic
formula, that is based on data for some other
situation?
MR. CARPENTER: I will grant you that
there is not a lot of data, and that was one of the
problems that the staff had. And that is one of the
reasons that we did not agree with a crack growth rate
that would allow for two cycles before they did any
further examination.
We looked at the crack growth rate in this
extremely, and bear in mind that this crack growth
rate is assumed, is bounding the limited amount of
data that we do have. So what the staff did was we
took what they licensee and Westinghouse provided to
us.
MEMBER POWERS: Now, when you say you
bounded the data you have, I mean how do you go about
bounding this data? Presumably the crack can't go any
faster than the speed of sound in the metal. I mean
I -- accept that as a bound. What other bound can you
possibly come up with?
MR. CARPENTER: When you take a look at
all the data that is provided and look at the scatter
growth, the licensee had a best fit to that data. The
staff disagreed with that and we increased our
bounding crack growth rate so that it incorporated all
the data.
MEMBER POWERS: How do you know that's
enough? If I took two more data points maybe they
fell outside your bound.
MR. CARPENTER: We don't have two more
data points, sir.
MEMBER POWERS: Yes, but if I had taken -
why do I know that bound?
MR. CARPENTER: I cannot sit here and
guarantee that this is the absolute bounding crack
growth rate. It is much faster than what we have
assumed in the past. And it is the reason that we
were only comfortable with wide operation,
approximately 18 months.
MEMBER POWERS: I'm trying to understand
why you were comfortable with five minutes.
DR. FORD: Gene, can I try and help you?
MR. CARPENTER: Please.
DR. FORD: Is that formula that you just
gave the Peter Scott formula?
MR. CARPENTER: Yes, sir.
DR. FORD: That's based on secondary side
cracks in tubes, I think. Therefore, the argument I
think that's being made here, Dana, is that that could
be a worse case material environment situation. The
follow up question to that, however, Gene, would be if
you used presumed residual stress profile through that
182 crack, would you have predicted what happened on
Leg A, using that formulation? I think that would
answer your question, Dana, or go towards it.
MR. CARPENTER: Perhaps, yes.
MEMBER SHACK: Well probably not because
the residual stresses that were used for the Summer
analysis were the standard sort of piping stresses, or
the standard circumferential weld residual stress
distribution which probably would have arrested the
crack.
MR. BATEMAN: This is Bill Bateman from
NRR. We're into an area here when Gene does not have
the technical expertise. We do have a technical
expertise, however, the individual is not here today,
the one who wrote the safety evaluation and was
involved in asking all these questions.
We can follow up at a later time, if need
be, to answer the questions when we have the
appropriate technical expertise available.
VICE CHAIR BONACA: I just have a
clarification, and I apologize for it. Maybe I missed
something during the presentation. I thought that
there was a claim that the crack that we found in the
A leg was due to the unique welding process used in
the location.
MR. CARPENTER: That is the claim that the
licensee made, yes sir.
VICE CHAIR BONACA: Okay. But now you're
telling me that the inspection showed that the other
nozzles also have cracks?
MR. CARPENTER: Five of the six.
VICE CHAIR BONACA: Okay. So these
nozzles, were they subjected to the same welding
processes as the -
MR. CARPENTER: No.
VICE CHAIR BONACA: No, they were not.
Okay.
MEMBER SHACK: But they do apparently have
smaller cracks. At least you can't see them in the
UT.
MR. CARPENTER: Correct. Okay, going
onward. As I was saying, the staff's review basically
disagreed with the licensee's premise that they could
operate for two cycles, and we said that they could
operate for one cycle before they needed to inspect
again.
Again, we decided to bound the crack
growth rate that they had provided to us, and that's
because there is a limited amount of crack growth
data.
MEMBER WALLIS: Can you tell me how much
does happen in this one cycle? How much crack growth
do you anticipate in this one cycle's okay and two are
not. How much crack growth is happening in the one
cycle which makes it impossible to operate any longer?
MR. CARPENTER: Assuming that there is a
crack indication that is one quarter inch in length,
with an aspect ratio of 1:2, with this crack growth
rate it will not grow three quarters of a way through
wall in one cycle of operation.
MEMBER WALLIS: But how much will it grow?
Halfway through the wall?
MR. CARPENTER: Roughly, sir.
MEMBER WALLIS: So it's growing a lot?
MR. CARPENTER: Yes, sir.
MEMBER WALLIS: So you'd better be careful
about up the bounding so much.
MEMBER POWERS: If the plant has some
misadventure and they shut down, something like that.
Does it change the crack growth rate?
MR. CARPENTER: Define "misadventure,"
sir.
MEMBER POWERS: An unplanned shut down
SCRAM. Something, anything. Does that change the
crack growth?
MEMBER SHACK: It would take an enormous
upset event, you know, to cause mechanical crack
growth rate here so this is really a, you know, a sort
of stress corrosion crack growth rate.
MR. CARPENTER: Albeit rather fast.
MEMBER SHACK: Yes. I mean the 182 crack
growth rates are really as high as you find in any
material that you know we know in stress corrosion
cracking. And, as Peter mentioned, there's a fair
amount of data on alloy 600 in the cold worked state,
there's less data on the 182. When you combine the
two data sets, you have a fair amount of data so that
you have a reasonable confidence when you bound the
whole combined set of data because 182 is something
like cold worked alloy 600. You can argue that
there's an analogy there and the existing data points
to the 182 are basically bound by the data that you
see for the cold worked alloy 600.
So if you combine the whole total data set
there's relatively little for 182 in this PWR
environment. I think you can have a reasonable amount
of confidence that the crack growth rate the staff has
used is bounding.
MEMBER POWERS: I guess I just don't know
where you derive your confidence here. You've got a
stress corrosion cracking phenomenon, it depends on
how much stress you have. That must be unique to this
situation.
MEMBER SHACK: Oh, the crack growth rate
that you have is a function of the stress. That is you
have a bounding curve that it depends on the stress
intensity at that location. So that becomes a
variable that you have to account for in a specific
analysis for a specific circumstance.
MEMBER POWERS: It also depends on having
an aggressive corrosion chemistry. That must surely
be unique to this situation?
MEMBER SHACK: No. That's the one good
thing about dealing with primary water stress
corrosion cracking is that you probably do understand
the chemistry environment, that is the environment
that you've studied the crack growth in is a PWR
primary water environment which is carefully
controlled.
MEMBER POWERS: Would this alloy, this
particular weld material and all of its associated
impurities have been exactly reproduced in this test
stage?
MEMBER SHACK: Of course not.
MEMBER POWERS: Well, and then you've got
to convince me that you've bounded it.
MEMBER SHACK: But that's why you have
data on multiple heats of material and, again you
know, when can I say I bounded the data. You know, as
Gene said, I don't think you can say you have an
absolute bound but what you have is an amount of data
on a reasonable number of heats of material under
chemistry conditions that are representative of what
you have here. It doesn't exactly represent it but
you think the population is quite representative and
you have to make the judgment that when you bound that
it's reasonably close to them.
MEMBER POWERS: No, I think all they've
done is they've parted the data points and run a
curve that goes over the top of it. And I don't think
they've done any of this, okay, does the chemistry
span the range of chemistries that I'm likely to
encounter or not.
MEMBER SHACK: Well, the other good thing
of course is that under ASME code conditions, the
range of the material chemistries is not all that
broad. It's a fairly tightly controlled situation,
especially for stainless steels. Again, for ferritic
steels, impurity levels are a good deal higher.
Welds, one of the bad things about welds is the fact
that impurity levels are higher but, again, without an
extensive study I'm not sure that you could say you've
bounded the range but you certainly have a reasonable
population.
MEMBER WALLIS: When this crack gets
bigger you said it could grow as much as halfway
through, does its growth rate slow down or increase as
it gets to such a big crack?
MR. CARPENTER: As the crack grows it will
reduce, it expends the energy so it will tend to slow
down a bit.
MEMBER WALLIS: It will slow down. And
when will be the next inspection? And I guess you're
giving them permission to run for another cycle?
MR. CARPENTER: For 18 months, yes sir.
MEMBER WALLIS: Eighteen months. So
what's going to happen in-between in the 18 months?
No inspection?
MR. CARPENTER: I should alter that just
slightly. We said that they could operate for up to
18 months. They have told us that they're going to be
operating for a short cycle so they will be shutting
down before next summer to inspect.
MEMBER WALLIS: And no one will be looking
for boron stalactites till next summer?
MR. CARPENTER: They will be looking for
evidence of leakage. But it's very difficult to get
into this area during operation.
MEMBER WALLIS: And you're satisfied that
there's a good way of detecting these very small
leaks? They weren't detected before.
MR. CARPENTER: That is correct. And that
is one of the things that I will be talking about in
a moment regarding leakage. Okay.
MEMBER SHACK: But, again, these cracks
will still be relatively short in terms of structural
integrity of the pipe. You know, they'll be a long
way from any kind of large failure, the margin to a
small leak is admittedly much, much smaller than it is
to a large -
MEMBER WALLIS: Well it may be fine, it'll
just be sort of embarrassing if you go in there next
summer and find there's a huge boron stalactite
somewhere.
MR. CARPENTER: But also bear in mind,
sir, that virtually all of these cracks were axial in
nature. As Dr. Shack said, it takes a considerable
amount before you have a concern beyond that.
Some of the ongoing activities that the
staff is engaged in at this time is that we're
reviewing similar cracking in foreign reactors. We
haven't seen anything similar to what happened at
Summer here in the U.S., but the root causes of both
the Summer and the Ringhals cracking in Sweden is
PWSCC. So we are talking with the Swedes about that.
We're also investigating -
MEMBER WALLIS: When you talk about PWSCC,
is it stress corrosion cracking?
MR. CARPENTER: Primary water stress
corrosion cracking. And we're also investigating
reports of other foreign cracking. We haven't yet
been able to verify that there are others that are
identical, or at least similar to what has happened
here at Summer, but we are looking at that.
MEMBER POWERS: When you say identical, I
get the sense that we don't have to be very identical
to be about the same. At least as far as our data
base. So I mean how do you -- I'm trying to
understand the links and bounds of identicality in
this sense.
MR. CARPENTER: Well, I'm not trying to
say that it has to be one for one matching every
point. We're looking at similar welds, we're looking
at similar locations. Trying to find something that
we can lump together.
Some of the other ongoing activities that
the staff is engaged in at this time is, again, we're
looking at the generic implications of the Summer
cracking and the industry activities.
Mr. Matthews of the MRP will be discussing
in a few minutes what the PWR owners groups materials
for a liability program is presently engaged in and I
will leave that to his capable hands.
We're also looking at the implications on
the leak-before-break analyses that have been done for
virtually all the PWRs at this time. We're also
looking at ISI programs, both deterministic and risk-
based, seeing if we need to make any alterations to
those programs.
MEMBER SIEBER: Since you're discussing
ISI, is there any movement to augment with another
technique the UT examination?
MR. CARPENTER: Right now the UT
examination is required by code, and if the code needs
to be altered then -
MEMBER SIEBER: Somebody has to put in a
code case?
MR. CARPENTER: Correct.
MEMBER SIEBER: On the other hand, you can
do the UT exam and satisfy the requirements of the
code, but you could require an additional augmented
inspection using other techniques. Seems to me that
when I look at the pictures, the weld prep for the
examination wasn't very good and it seems also that
probes have a pretty good footprint and maybe this is
a very difficult weld to examine just because of
geometry, notwithstanding the fact that you're
shooting through a pretty thick cross section of
material of varying grain structure and composition.
I take it that since these kinds of cracks
are not reliably always found by UT, that nobody is
making a move to do something better.
MR. CARPENTER: Well, I'm not saying that
yet, sir, and you're leading me by about three slides.
So I'll discuss that in just a moment if I could
defer, okay.
MEMBER SIEBER: All right.
MR. CARPENTER: Again, as we were just
mentioning, the ability of the code required NDE to
detect and size small ID stress corrosion cracks, this
is one of the things that we definitely need to get a
handle on. And the appropriateness of the ASME code
standards allowing flaws approximately 10 percent of
wall thickness that, in the case of Summer, could grow
with such an apparent high crack growth rate.
And Dr. Wallis mentioned a few moments ago
about the effectiveness of the leak detection systems.
These are all things that we are very much following
and trying to get a handle on.
MEMBER WALLIS: Are you thinking of
putting in some supplementary leak detection system in
the places where you might detect something such as
just by the well?
MR. CARPENTER: We have discussed that
with the industry and I believe that -- will you be
discussing that, Larry?
MR. MATTHEWS: We're going to be looking
at it.
MR. CARPENTER: Yes. So this is part of
what we're talking about.
MEMBER WALLIS: If you're going to look at
it, are you going to have anything in place during the
next cycle?
MR. CARPENTER: No, sir.
MEMBER WALLIS: You seem that you want to
put something in there now.
MR. CARPENTER: The lead time for
developing supplemental inspections or supplemental
leakage evaluation --
MEMBER WALLIS: Well it seems to me there
was, I'm trying to remember the pictures I saw, but
the boron stalactites were pretty obvious, right. So
even just a camera would see them, and that's not a
remarkable piece of technology.
MEMBER SIEBER: Well, the boron that was
visible was not directly at the crack in the pipe, it
was underneath the boot and, you know, and that's
where the air flow ventilation for that section of
piping comes from. So it had to appear where the boot
was not tight.
MR. CARPENTER: When you're talking about
all this, you know, it came out through here so it is
not readily accessible during power operations.
MEMBER SIEBER: And it might not come out
there.
MR. CARPENTER: Correct.
MEMBER WALLIS: But it's too rough an
environment for some sort of video surveillance?
MR. CARPENTER: I don't know the answer to
that, sir.
MEMBER SIEBER: It could be done.
MR. CARPENTER: Some of the further
activities that the staff is working on right now is
that we're proposing confirmatory research into the
primary water stress corrosion cracking issue, and
that will include some of the NDE and the ISI issues
that we've discussed so far.
MEMBER WALLIS: What are you trying to
confirm?
MR. CARPENTER: Confirmatory research
meaning is our capability of other inspection toolings
to go in and find and size these indications. For
instance --
MEMBER WALLIS: This is a kind of, I know
it's a misnomer, but research is to figure out
something new not confirm something.
MEMBER SIEBER: But they're not
anticipating a problem. They have the problem.
MR. CARPENTER: Determination of a
bounding crack growth rate and just how the residual
stresses play into that. Development of
susceptibility model, and because we know that the
welds that were, for instance, the welds at Summer
were field fabricated, you also have some that were
shop fabricated at other PWRs. You have different
materials being used at different PWRs. So the
susceptibility model is going to have to take a look
at multiple factors.
The assessment of possible repair and
mitigation methods that the industry may come up with
and overall following of industry activities.
MEMBER SIEBER: Now, in all Westinghouse
plants with stainless steel piping do they use the 182
weld model?
MR. CARPENTER: No, and all of the
Westinghouse plants from my understanding is that they
have something like virtually every weld was slightly
different. So it's going to add considerably to the
complexity here of all this.
DR. FORD: Gene, could I make a comment?
You mentioned at the beginning of this is the generic
activities. When you're looking at the industry
experience, are you confining yourself to pressurized
water reactors? I'm thinking specifically of the vast
amount of boiling water reactors which are into
hydrogen water chemistry with a lot of 182 welds.
MR. CARPENTER: Yes. Yes, sir I have been.
At this time we are looking specifically at Ps, we may
expand into Bs. But this appears to be a PWSCC
concern right at this moment. If we need to, we will
expand the scope beyond that.
MEMBER SIEBER: I mean he is running at
hot leg temperatures that are a good deal higher than
your BWR.
DR. FORD: Yes I recognize that but we
know what the activation enthalpies are for the
cracking in these systems, so you can make some sort
of comparison.
VICE CHAIR BONACA: I had a question
regarding all previous inspections, including the year
2000. They found no indications. Now, what kind of
inspections were they? They were not using eddy
current, of course.
MR. CARPENTER: If you don't mind I'll
defer to Steve Doctor on that one. Steve?
DR. DOCTOR: Yes, I couldn't hear the
complete question. Would it be possible to have you
repeat it, please?
VICE CHAIR BONACA: Yes. My question is
all previous inspections of these nozzles showed no
indications, including the 2000 inspection. And I was
wondering what type of inspection that was, I mean
what kind of technique do you use?
DR. DOCTOR: They basically employ the
same techniques that they employed back in 1993. The
biggest improvement on the ultrasonic side was that
they employed an improved transducer sled that allowed
each transducer to independently gimble to do a better
job of tracking the surface and thereby providing
better coupling.
This was a significant improvement. It
didn't accommodate all the conditions that in fact are
associated with the ID conditions of these particular
wells. The size of the footprint of the transducer
and the housing that the transducer goes in is quite
large and, as a consequence, it has some difficulty
accommodating the root, the counterboard, and there is
a difference in the diameter between the nozzle and
the pipe.
And, as a consequence, it did create some
problems. I believe that obviously one of the things
that industry is going to be looking at in the future
on how to ensure a better ability to track the surface
and thereby improve the quality of the UT inspections.
And for this particular inspection, the
staff at Summers and Weston, you know, agreed to use
the eddy current. An eddy current inspection is
primarily a surface inspection type of technique. It
has not generally been used for this kind of
application.
However, it is very sensitive to any kind
of surface breaking flaws and, as you saw from I think
the result from the Alpha leg hot outlet nozzle
dissimilar amount of weld, the eddy current was very
effective at detecting a number of cracks that were
verified through the destructive testing. And it
should be noted that there are indications in four of
the five other dissimilar amount of welds that we have
found with the eddy current, they have similar
characteristics to the cracks that were in the Alpha
leg, but at this point there are still indications.
They have not been, you know, verified by any other
means.
There's a possibility that some of those
in fact may not be cracked because some of the
indications that were found in the Alpha hot leg were
not verified through destructive testing. So there is
some uncertainty there and in the analysis they've
taken the approach to assume that, in fact, all of the
indications are assumed to be cracks, although that's
not proven at this point.
VICE CHAIR BONACA: So the previous
inspection used the ultrasonic testing. And some
forms of eddy current, if I understand it.
MR. CARPENTER: No.
VICE CHAIR BONACA: No. No eddy current.
Okay so it was ultrasonic testing. And that is the
standard testing that is being done by the industry,
right?
MR. CARPENTER: Correct. Code required.
VICE CHAIR BONACA: Thank you.
MR. CARPENTER: Okay. That brings us to
where the industry is right at this time. The PWRs
have proposed an industry initiative to respond to the
cracking issue that was found at Summer. And, as we
have discussed with the ACRS before, the staff has an
industry initiative process that we can utilize for
this, in which case an issue occurs, the industry and
the staff meets on this. The industry proposes to
follow this as an industry initiative, and the staff
either forgoes any generic communications or generic
letter per se, to tell them what needs to be done in
lieu of the industry coming in and actually telling us
what they're going to do following this.
Now at this time we have met with the
materials reliability program on this twice now. And
they have proposed to respond to the issue and, again,
Mr. Matthews will be discussing this in a couple of
moments.
VICE CHAIR BONACA: I just want to get
back to it. It seems to me that there has to be some
past experience of V&V on the ultrasonic testing that
is adequate or is not adequate. We were left here
with a statement that says that we have eddy current
indications of cracks in the other nozzles which were
not identified by UT. We're not sure yet that they're
cracks, they may be something else. So we're trying
to understand, in fact to validate these observations
here.
And so my question, again, is do we have
V&V of ultrasonic testing identifying these kind of
cracks?
MR. CARPENTER: The Alpha hot leg had both
eddy current testing done on it and ultrasonic
examination. It was then cut out, the weld was cut
out, and was destructively examined. As Steve Doctor
mentioned before, some of the indications that were
found by eddy current were not found in the
destructive examination. Some of the indications that
were found by destructive examination were not found
by UT. So we're still struggling with that, sir.
VICE CHAIR BONACA: Okay.
MEMBER SHACK: Let me ask it in a
different way. The inspectors that do the
inspections, do they go through a performance
demonstration on stress corrosion cracks?
MR. CARPENTER: The PDI program --
Performance Demonstration Initiative -- as I
understand it, I'm not the expert on this. Steve, did
you want to respond to this?
DR. DOCTOR: Yes. Right now this is a PWR
issue and all the people that are really trained for
stress corrosion cracking do inspections on BWRs.
There is a requirement for a supplement independent --
regarding dissimilar amount of welds. That has not
been implemented as of yet. It's in the process of
being developed with regard to the PDI program and
it's something like I think about 18 months off until
that will be fully implemented, and then all
inspectors will have to go through that.
And, of course, the timeliness of the V.C.
Summer event is that now we've identified failure
mechanism and so the type of flaws that have to be
included in that demonstration are PWSCC.
MEMBER SHACK: Thank you.
MR. CARPENTER: Okay. Going back to where
we are with the MRP. Again, the staff has met with
the industry on this at least twice now. We've had
multiple telephone calls with them following up,
discussing the agenda items that have been in these
public meetings. We will have another public meeting
with the MRP in three weeks time to discuss the
assessments that they are going to be providing to the
staff. We have also gone down to one of the vendor
sites, Framatone specifically, to take a look at the
mock up that they have been making use of to look at
the welds for four plants that we'll be inspecting
this spring outage. And further technical and
management meetings are planned to discuss what is
going on.
And now we get to the slide that Dr.
Sieber was leading me to.
(Slide change)
MR. CARPENTER: Some of the staff
expectations of the generic activities. What we are
hoping to come out of this with. The MRP assessment
of the generic susceptibilities; they have promised
this to us by the end of March and that is what we
will be discussing in three weeks time.
The NDE methodologies and the toolings
that the industry is going to be using to do their
examinations. The staff has told the industry that
they should be making use of the best practices and
capabilities to address potential weaknesses that seem
to have come out of this examinations at Summer.
If potential code cases are necessary to
address some of the things that we have discussed
already, the staff will be looking at those in an
expedited manner.
We also need to get a better handle on the
implications for the ISI programs and also for leak-
before-break. And long term assessment of the alloy
82/182 applications, we're going to be discussing with
the industry to take a look at that.
And we'll also be looking at the review of
their repair and mitigation methods that they will be
proposing to us.
And that concludes my discussion for this
morning.
MEMBER WALLIS: I'm just wondering, the
expectations, are results expected or activities?
MR. CARPENTER: We're hoping that there
will be activities that will lead to results, yes.
MEMBER WALLIS: Well that's the thing, I
see a lot of activity and I just wonder about the
results.
MR. CARPENTER: We are at the very
beginning of this, sir, and it's too early to --
MEMBER WALLIS: That's what concerns me a
bit, yes. You may discuss with industry for a long
time without achieving anything.
MR. CARPENTER: We have made our
expectations very clear that this is something that
needs to be expedited. It's not going to be a five or
ten year practice before something occurs. That we
need to have something sooner. And, again, under the
industry initiative process, if the staff determines
that the industry is not being as proactive as we
would like, we always have the option of going out
with generic communications of some sort.
MEMBER WALLIS: Well at least by next
summer you'll have some data points?
MR. CARPENTER: We certainly hope so, sir.
MEMBER LEITCH: The licensee seems to make
quite a bit out of the uniqueness of this weld in its
original instruction. But I guess from hearing your
presentation, it sounds as though you are not
accepting that idea, that you feel there's something
more generic going on here. Is that a correct
assumption?
MR. CARPENTER: Well, initially, sir, we
were in agreement that there were, as the licensee
correctly points out, extensive repairs done on this,
especially as opposed to the other five welds. That
there were mitigations there that could have caused
this one to be of concern but not the other five
welds. And then there were indications found in four
of the other five. And also there was indications
found at Ringhals in Sweden, which is a similar plant.
MEMBER LEITCH: Indications but not
cracks, right?
MR. CARPENTER: Not through wall cracks,
no, sir.
MEMBER SHACK: No, but Ringhals is
confirmed to be a crack.
MR. CARPENTER: But not through wall.
MEMBER: Yes, but it is a crack.
VICE CHAIR BONACA: Ringhals identified it
through eddy current?
MR. CARPENTER: They did do eddy current
examinations there also, yes. That leads us to be a
little less accepting of the uniqueness suggestion.
MEMBER LEITCH: Yes, do we know if
there's anything unique about which plant is at
Ringhals weld?
MR. CARPENTER: Well, it's a double V
weld. It's fairly similar to what we were discussing
earlier. But, again, we really need to get a better
handle on all this information.
MEMBER WALLIS: So the Ringhals crack is
still growing is it? Or has it been fixed?
MR. CARPENTER: Debbie, do you remember
what they said?
MS. JENSEN: They did some repairs.
Debbie Jensen from the Office of Research. They did
some repairs to the Ringhals crack, but we're going to
meet with them next month and have face to face
conversations and exchange of technical information
with the similarities and the differences between the
two plants in this particular issue with the pipe
crack.
MEMBER WALLIS: Well did some repairs, do
you mean they cut out this area and rewelded it or
something?
MS. JENSEN: From what I understand, yes,
they did some grinding and they replaced with addition
weld metal and they took out some samples to do some
testing.
MEMBER LEITCH: I have a question about
the enhanced leak detection procedures that were
mentioned. Are we going to get some more about that
later, or is now an appropriate time to ask that
question? I'd like to know specifically how these
leak detection procedures were enhanced.
MR. CARPENTER: Well, that's one of the
things that we have asked the industry to go and look
into, to see what they can develop as far as enhanced
leak detection capabilities. Right now we're not
ready to discuss what could be used.
MEMBER LEITCH: Ms. Cotton in her
presentation said that one of the licensing
commitments was to enhance their leak detection
procedures. Am I to understand that the plant will go
back in service with -- that is that that enhancement
is future, that the plant will go back in service with
the same leak detection procedures?
MR. CROWLEY: What they plan to do is they
plan to do noble gas sampling and analysis to provide
additional verification of the RCS integrity. The
other thing they plan to do is they're going to --
MEMBER LEITCH: You say they plan to do
that, but will they be doing that when the plant gets
back in service?
MR. CROWLEY: Yes. Yes, that'll be when
the plant goes back in service. The other thing, the
calculation of RCS water inventory balance, they plan
to do that on a more frequent basis than they've done
in the past to try to determine if they have
additional leakage.
They're going to add a main control board
enunciator to alarm at 0.75 GPM such that the
operators will be alerted prior to reaching tech spec
limit.
MEMBER LEITCH: 75 GPM?
MR. CROWLEY: .75 GPM. And then they're
going to, of course this is not what the plants
operating -- the inspection they do when they come
down next time, they're going to have an enhanced
boric acid inspection. They've had boric acid
inspections every time a plant comes down, but they're
also going to enhance that program also.
MEMBER WALLIS: What happens when you have
this crack and there's a leak? Do you get a jet of
suppurated steam coming out of it or what?
MEMBER SHACK: Sure.
MEMBER WALLIS: And this has properties
like momentum?
MR. CARPENTER: Yes.
VICE CHAIR BONACA: It'll cut like a
knife.
MEMBER POWERS: You can't get away from it
can you.
MEMBER WALLIS: But it's invisible isn't
it? It's invisible. It's a jet of some --
MR. CARPENTER: Yes.
MEMBER WALLIS: But it impacts on things
and it carries boron with it. The boron is in the
steam in some form, droplets or something?
MR. CARPENTER: Yes.
MEMBER POWERS: Vapor.
MEMBER SHACK: It's dissolved in it.
MEMBER WALLIS: Dissolved and then it
comes out when it condenses on a cold surface
somewhere.
MR. CROWLEY: Well the pipe is insulated,
of course, so it has to -- whatever vapor that comes
out has to get through the insulation.
MEMBER WALLIS: So what happens in the
insulation? It deposits or is the insulation sort of
blown off, does a hole get made in the insulation?
VICE CHAIR BONACA: Cut right through it.
MR. CROWLEY: Through the same, goes
through the same.
MEMBER SIEBER: Well, most plants have
mirror insulation and it travels all over the place.
MEMBER WALLIS: So this steam gets all
lost in the insulation somewhere?
MR. CARPENTER: Yes.
MEMBER SIEBER: I was under the impression
that most PWRs had some kind of noble gas detection as
part of their containment radiation monitoring.
MEMBER WALLIS: Well let's go on. That
insulation gets hot when you put steam through it and
you get a hot patch on it when the steam comes in
there, so if you had thermal couplers on the
insulation they would get hotter if you had a steam
leak. There seem to be so many things that could be
done using pretty robust technology to detect some
change that would be affected by a steam leak.
MR. CARPENTER: And these are things that
we have asked the industry to go in and investigate
and come back and talk to us about.
MEMBER WALLIS: But if you had to do it
next week I would think that someone could actually
come up with something. I just wonder why -- it
seems to be a slow process, this asking and coming
back with things. The agency doesn't seem able to
respond quickly to the idea say let's put thermal
couplers, or whatever it is, around something so we
know what's going on.
It may take months to make a decision. By
then the cycle's over anyway. Am I describing things
right? It just takes forever to make -- not forever,
but it takes so long to make decisions that it's
unlikely that any detection system will be in place
before the end of the cycle.
MR. CARPENTER: Well, there is detection
systems in place at Summer. As to what needs to be
done at other plants --
MEMBER WALLIS: Well what's the new --
there's a new detection system in Summer?
MR. CARPENTER: Yes --
MR. CROWLEY: Just the improvements that
we --
MEMBER WALLIS: Just the ones that you
mentioned. But they are still -- gross balances for
the plant. They're not focused on the area of
concern. There's nothing installed around the welds
or anything like that.
MEMBER SHACK: Local leak detection is
harder than you think because, as Jack mentioned, you
know, water and steam have a way of moving around a
lot.
MEMBER WALLIS: Well it's very true in
your house, you get a leak in the bathroom and it
appears in the living room.
VICE CHAIR BONACA: I just wanted to ask
you a question about, this is not the first time that
PWR nozzle cracks have been identified, right? It is
not the first time.
MR. CARPENTER: I believe it is, sir.
MEMBER SHACK: Ringhals is the first,.
VICE CHAIR BONACA: Is it? I thought
there have been some events.
MR. CARPENTER: But this is the first
through wall crack.
VICE CHAIR BONACA: Oh through wall, yes
I understand. But cracks which were not through wall,
I thought there had been some instances.
MR. CARPENTER: Not that I'm aware of but
I will get back to you on that.
VICE CHAIR BONACA: I guess I'm going after
the issue of, you know, this seems to throw in full
doubt the effectiveness of ultrasonic testing as an
inspection means, and I thought that there had been
some significant validations of the technique.
MR. CARPENTER: The ultrasonic
examinations of dissimilar metal welds is a little bit
more of a challenge, so that is something as Dr.
Doctor mentioned a little bit ago, the PDI initiative
is looking at that and they have approximately 18
months to come up with a solution to that.
MEMBER SHACK: I mean non-destructive
examination, you know they're sort of trained to look
for certain things and at this point PWSCC wasn't
really considered to be a major problem for PWR.
VICE CHAIR BONACA: I understand. In 2000
they had no indications. Then eddy current comes and
says there are indications. There are indications to
the point where now we're putting restrictions on how
long they can run. It begs the question of what do
you do about all the other PWRs for which you have
inspections using ultrasonic testing. And so that's
why I'm asking those questions. I had more confidence
in that testing than I'm getting out with now.
MR. CARPENTER: And these are questions
that the staff are asking ourselves, yes.
If there are no further questions we'll
turn this over to Mr. Matthews of MRP.
MR. MATTHEWS: My name is Larry Matthews,
I work for Southern Nuclear Operating Company on the
managing inspection and testing services group. I'm
also chairman of the Alloy 600 issues task group of
the materials reliability program, and I'm going to
give you some information about where the industry is
and where we're headed on this issue.
First off, a little brief history of how
we got to where we are. You've heard about the crack
and what's been done at the plant at V.C. Summer. The
MRP Alloy 600 issues task group took the lead on this.
The event occurred in October of 2000, the initial
root cause was available early December and the issues
implementation group, or issues and integration group,
we can't ever decide what IIG stands for but it's the
parent of the ITG, recommended in mid-December that
the MRP take on as activity the resolution of generic
issues relative to the V.C. Summer event.
We received executive approval from some
utility execs here in early January to begin
activities. We developed an organization and we
worked out a fairly detailed plan and budget but it's
evolving as we go and as we learn more.
The issues task group met in January 19,
after the V.C. Summer public meeting on the 18th, to
address the key focus areas and we organized into
three committees: an assessment committee, inspection
committee and a repair and mitigation committee and
I'll be going into what the activities of those
committees are.
We met with the staff on January 25, at
which point we outlined the approach that we were
planning on taking with respect to this issue, and
solicited feedback from the staff at that point in
time as to whether they saw things additional we
needed to be doing.
The feedback was basically they felt we
were on the right path, saying the right kind of
words. Of course, the proof is in the pudding -- can
we deliver what we say.
On February 1, two of the committees had
their initial meetings, inspection committee and
assessment committees both met in Charlotte. They
further refined their plans and schedules and budgets.
On 2/16 there was an MRP/NRC executive management
meeting. This is typically a meeting we've been
having on an annual basis where MRP executives were
meeting with the NRC management.
MEMBER WALLIS: Has anybody done any work
yet?
MR. MATTHEWS: Yes.
MEMBER WALLIS: No, I mean you have all
these meetings in the management and budgets, has
anyone done any engineering yet?
MR. MATTHEWS: Yes, yes. I'm going to get
to that.
This issue was one of the topics that was
discussed at the meeting along with all the other MRP
activities, and just this week we've scheduled another
technical meeting with NRC staff and I'll go into what
we're going to discuss in that meeting.
The industry plan includes a short term
assessment in which we want to demonstrate that the
continued operation of alloy 82/182 welds is
acceptable. We're trying to get that to the staff by
late March. The NSSS vendors are at work right now
performing the analyses and working on this
assessment.
We had a goal of getting interim
inspection guidance.
MEMBER POWERS: You say here the continued
operation with alloy 82/182 welds, that's just the
weld not a flawed weld that you're dealing with?
MR. MATTHEWS: We want to -- well, what
we're going to show is the margins that are available
in there to cracking and even if it does crack, the
margins that are available to a rupture of the pipe.
We're going to try and prove here that it's not really
a safety issue, it's a leak issue, it's an operational
issue, we have to be very concerned about it, it's
very expensive to have this kind of leak. But we want
proof and show that it's not a safety issue.
MEMBER POWERS: What you want to show, I
think, is that if you have a flaw in that weld, that
it will not propagate rapidly to create a pipe
rupture.
MR. MATTHEWS: Exactly. Analyses to that
effect were certainly part of the report that
Westinghouse put together for V.C. Summer. And we
will build on those analyses for the whole industry.
MEMBER POWERS: But you're not going to
have any more data than they did.
MR. MATTHEWS: No, not at this point in
time. I mean there's no more data that we can get our
hands on right now. We've got to go create some or
find out what else is out there.
MEMBER POWERS: Well when you think about
data on stress corrosion cracking, you think about
things like residual stresses, you think about
chemistry. Do we have now data that are taken in
irradiated water of the type we have in --
MR. MATTHEWS: I don't think we have data
in irradiated water, but we do have data that was
taken with several heats of alloy 182 weld metal and
there was created, samples cut from it, several
samples were put into a PWR environment in an
autoclave and tested to crack --
MEMBER POWERS: When you say environment
you're speaking of the pressure temperature
environment not the radiation environment?
MR. MATTHEWS: Not the radiation but these
things are very, very low radiation where these welds
are. These welds are not in the belt line region,
they're above the core.
MEMBER POWERS: Well, I mean I just can't
help but ask, you have a lot of radiolysis product,
water radiolysis products in these and they tend to be
fairly aggressive chemicals as far as oxidation and
reduction reactions. Do they not affect the chemistry
in these?
MR. MATTHEWS: I guess I don't know the
answer to that but the tests that we've done are
trying to stimulate the PWR primary water as best they
can, given that we're not doing it with a reactor.
MEMBER POWERS: Well my question is is
temperature adequate or do you have to simulate the
ozonides and peroxides and things like that because of
water radiolysis?
MR. MATTHEWS: I guess I don't know the
answer to that. We do run these plants with a
hydrogen over pressure and it tends to scavenge those
things pretty quickly I would hope.
MEMBER WALLIS: Even short life -- and
this is a hot leg, this stuff has been irradiated and
everything else a very short time before it comes to
this spot.
MR. MATTHEWS: Yes.
MEMBER WALLIS: So there could be some
very transient type products which are in there.
MR. MATTHEWS: Yes, there could and I
guess we haven't looked at that as an industry and
perhaps we need to.
VICE CHAIR BONACA: I don't want to
belabor it but it seems to me that the gentleman said
they're trying to see if in fact this eddy current
data is credible.
MR. MATTHEWS: The eddy current data?
Yes.
VICE CHAIR BONACA: Now, assuming that you
could prove that the eddy current indications were not
correct, that would support your claim that this is a
unique issue to do with that particular weld in that
particular A leg, and all this would be gone. So why
won't you focus immediately on the issue of the
validity of eddy current as a means of inspecting
these cracks?
MR. MATTHEWS: Well, we have some
information, as I understand it, from the Ringhals
test. They did UT and eddy current, and over there
the eddy current was not the save all, in fact it
missed flaws that the UT picked up.
VICE CHAIR BONACA: Okay. The combination
of the two seems to be an effective means you mean?
MR. MATTHEWS: Perhaps. But the eddy
current here was --
VICE CHAIR BONACA: So you can't discount
the eddy current indication, that's what you're saying
right now.
MR. MATTHEWS: Yes. It's not a proven
technology for going in and detecting and sizing
flaws.
MEMBER SIEBER: And it focuses more on
surface indications.
MR. MATTHEWS: Yes, that's right. Another
thing the plan included was to get out some interim
inspection guidance for the near term outage plants,
those plants that are coming down this spring. That
was completed yesterday I believe. The letter was
signed out to the industry.
The plan also includes a longer term
assessment of all the alloy 82 and 182 welds in the
plants, in the PWR primary systems. We'll be looking,
reviewing and improving inspection technology where
it's appropriate. And we'll also be reviewing repair
and mitigation methods, if necessary, working to
develop some improvements in those.
MEMBER WALLIS: Let me ask about UT.
Isn't this a developing technology in the medical
field that's highly developing, a lot more
intelligence is used for it and they can see things
they couldn't see before and it's improving very
rapidly. Is this sort of a fossilized technology, or
are improved UT methods coming out regularly?
MR. MATTHEWS: I think the industry is
constantly looking to try and improve their technology
for detecting --
MEMBER WALLIS: Is it happening?
MR. MATTHEWS: Yes, there's phased arrayed
technologies that are coming out and have not been
applied at this point to these welds, but that has
been applied in the industry for turbine blade
examination and things like that. There's new
technology being looked at by the EPRI and NDE center
right now for much smaller --
MEMBER WALLIS: Is it difficult to get
approval for new technology because of the regulatory
process?
MR. MATTHEWS: I think the code process
would be the more difficult thing to get it through
but, at the same time, if there's a better way to do
things I can think we can push it through.
DR. FORD: Can I just come back to the
very first bullet there, the short term assessment.
What are the criteria for that? What are the criteria
that your short term assessment is correct?
MR. MATTHEWS: I'm going to -- oh, I'm
going to give a lot more detail of what we're going to
do there.
DR. FORD: Okay. But there will be data?
There will be stress corrosion data to back it up?
MR. MATTHEWS: There will be what data we
have available will all be factored in to putting
together the short term assessment.
DR. FORD: Okay.
MR. MATTHEWS: And while we've already
started work on much of this, we expect there is an
approval process for the -- the senior reps we
anticipate them approving what we're laying out in our
plan on March 9. But we've already started work with
funds that were already available.
Basically, these are the three committees
under my Alloy 600 ITG and we're part of the MRP and
the MRP is looked to the NEI as the regulatory
interface with the NRC. That's not to say we don't
have technical discussions. We do. When there's
technical issues to discuss with the staff, we'll
discuss them directly.
MEMBER WALLIS: Excuse me. Who does the
work? Do you contract with somebody?
MR. MATTHEWS: Most of the work would be
contracted to vendors or consultants or done in house
at EPRI. Most of the technical work, a lot of the
guidance and overseeing of all that work is done by
these committees. And these people are knowledgeable
people in the industry in these areas too, on the
committees.
These are just the chairmen of the three
committees that we've set up. The chairman of the
assessment committee is Vaughn Wagoner from CPNL. The
chairman of the inspection committee is Tom Alley from
Duke. And the chairman of the repair and mitigation
committee is Gary Moffatt from the V.C. Summer plant.
One more detail about the committee
activities. The first thing is to get this short term
safety assessment done, the process that we've
outlined involves identifying areas that are likely to
be the most susceptible and that's primarily going to
be based in this very short term on evaluating the
size of the welds, the temperature and the weld
materials. We felt that likely spots would be on the
Westinghouse and combustion plants to hot leg pipe
welds. At BNW is may indeed be the CRDM nozzle welds
at the top of the head, but they also have some other
pipe welds that they'll be looking at I believe.
Certainly not all the plants have the same
welds. The difference between the vendor designs, the
piping is completely different on the three plants, or
plant designs, and even within the Westinghouse fleet,
these welds have a wide variety of how they were
constructed. Some shop welds, some field welds, some
stainless steel, some inconel 182 butter with 82 weld
material, so there's a wide variety of those and we
have to go out and assess all of those.
One of the goals is to demonstrate that
most of the cracks will be axial or in the case of the
head penetrations they will be in the axial radio
direction as was seen at the Oconee.
MEMBER WALLIS: Why will they be axial?
MR. MATTHEWS: It's primarily because of
the stress field that the --
MEMBER WALLIS: The stress field stresses
it more highly in that direction?
MR. MATTHEWS: Yes. Well the stress is in
the circumferential making the crack --
MEMBER WALLIS: The flow direction has
nothing to do with it?
MR. MATTHEWS: No.
MEMBER WALLIS: Well flows have effects
around bends and things. Flows have some effects on
these things don't they?
MR. MATTHEWS: A little bit of flow
momentum I would imagine but I don't -- then that
would be taken into account. That's going to be
second order compared to the other stresses that are
driving these things.
MEMBER LEITCH: Are you going to -- can
you go back and identify welds where there was major
repair activity at the time of original construction.
Is that one of the things you're going to be looking
at here? It seems to me if that was not the prime
cause of this failure, certainly I think we would all
agree that it accelerated the failure in this
particular A hot leg. So can you go back and identify
those welds?
MR. MATTHEWS: The amount of data that's
available to each plant varies depending on, you know,
some of these plants are 20, 30 years old and their
construction records are sometimes hard to come by.
But what's available is available and will be looked
at by the individual utilities to see if there's
anything.
MEMBER WALLIS: So chemistry comes into
this propagation of the crack, chemistry is a factor.
MR. MATTHEWS: Water chemistry?
MEMBER WALLIS: Yes. And so there's a
whole lot of flow mechanics and diffusion processes
and things going on in these cracks. It's not just
stresses, it's everything else, too. I just wonder
how well that is understood. The biggest axial crack
with the water whipping by with some sort of flow
percolating around through the crack as well.
MR. MATTHEWS: These cracks are so very,
very, very tight. The water in those cracks is
probably --
MEMBER WALLIS: Well something has to go
up there, there's going to be corrosion effects.
MR. MATTHEWS: Yes, but it's a very
stagnant environment.
MEMBER WALLIS: So it's diffusion.
DR. FORD: If I -- maybe I could just help
you out maybe. In boiling water reactors where you
have an oxidizing environment, yes, the direction of
flow could be important. But, in fact, the water does
not enter into the crack very deeply and it becomes
more an academic exercise.
MEMBER WALLIS: What's in the crack?
DR. FORD: Well all this water but you're
talking about a replenishment of the water, and that
does not occur to any great extent in these tight
cracks.
The question of the PWRs, you're not going
to get too much flow effect, rate effects in this
reducing environment, if you do have a reducing
environment, and Dana's observation is an interesting
one as far as I'm concerned.
MEMBER SIEBER: Maybe I could ask another
question. It seems to me that 82/182 I think and
alloy 600 are all, as far as stress corrosion
cracking, are all dependent on temperature. And the
need is what, 608, 609 degrees Fahrenheit where higher
temperatures than that to correct growth rate
accelerates. It would seem to me, and I worked in a
plant at one time, where because of the finding of
some cracks, they reduced the temperature at the plant
by about 10 degrees which virtually stopped the growth
of the crack. Has anybody considered that as an
alternative to all these other things?
MEMBER SHACK: People do it in steam
generators. That's generally a pretty drastic step.
MEMBER SIEBER: Yes, well you lose some
megawatts that way but a pipe break is a pretty
drastic thing, too. And given the choice, I would
rather lose a few megawatts.
MR. MATTHEWS: I don't think anybody's
considered at this point trying to reduce their hot
leg temperatures because of this. The drop may have
to be significant to get it down below that need that
you're talking about I would think.
MEMBER SIEBER: Well, it might have to be
below 600. On the other hand, for a lot of plants
that's seven or eight or 10 degrees.
MR. MATTHEWS: And for a lot of plants
it's more than that.
MEMBER SIEBER: Well, and so I continue to
question. You know, once you're above 610 as a hot
leg temperature, that means the reactor vessel head is
at the same temperature, well the inconel welds up
there that are also subject to the same kind of
cracking.
MEMBER SHACK: But you know he has a much
different problem than the steam generator people.
You know, they have typically much larger margins to
failure. A short crack in a steam generator gets you
a lot closer to failure than a short crack in a large
diameter pipe. It certainly could be done but it
certainly seems pretty far down on his list.
MR. MATTHEWS: And the temperature effects
are certainly going to be taken into account in our
assessment of susceptibility and crack growth. And
temperature is one of the factors in the crack growth,
too.
MEMBER POWERS: When you think about
activation energies for processes like crack growth
rates, you typically think about things with
uncertainties in the activation industry on the order
of five -- is that right?
MEMBER SHACK: Yes -- it is that much.
MEMBER POWERS: And so these temperatures
that like factors two or three on the crack growth
rate, so the difference is between the biggest between
two cycles and one cycle is kind of the input -- the
activation.
MEMBER SHACK: If all you were depending
on was the activation energy. But it's certainly
true that if you dropped the temperature 20C, you'd
get a lot. But you may not want to.
MR. MATTHEWS: Somebody asked earlier if
the crack growth curve that was used was essentially
the Peters-Scott model. Well that model was I guess
initially based on steam generators but it was
modified for the Alloy 600 head penetrations and that
crack growth model was used, the modified Peters-Scott
model was used for the susceptibility modeling that
we've done in the industry on the head penetrations
Alloy 600.
When we tested, in the test data that
we've seen on the Alloy 82/182 shows those crack
growths were, depending on the orientation with the
dendrites, five to ten times faster than the Alloy 600
crack growth rate.
And then the curve that the NRC used
bounded all of that so it was even more than that,
faster than the basic modified Peters-Scott model.
The short term assessment will demonstrate
a large tolerance for axial flaws and the
circumferential flaws. The stress analyses that we've
done indicate a preference for the axial cracking
because of the stresses in the welds and how they're
lined up. The flaw, as you saw on the plot they put
up, was limited to the axial length of the pipe weld,
which is just a couple of inches long. Basically,
that's based on the V.C. Summer experience, it stopped
when it hit the ferritic steel, it stopped when it hit
the stainless steel, and it was only the inconel weld
metal that actually experienced cracking.
The flaw in the CRDM nozzle at Oconee 1
also stopped when it hit the Ferritic steel. It did
propagate on into the Alloy 600 base metal of the
penetration itself.
And also the load limit fracture mechanics
analysis will show that there's a large margin to pipe
rupture.
MEMBER WALLIS: These are intents or
indications?
MR. MATTHEWS: We believe that this is
what they're going to show.
MEMBER WALLIS: But is it what you want to
show.
MR. MATTHEWS: No, we believe it will show
it. I mean a lot of this analysis is done --
MEMBER WALLIS: So this is based on
analysis having been done?
MR. MATTHEWS: The analysis -- a very
similar type analysis has already been done for V.C.
Summer and we're going to extend it to the rest of the
situation.
Similarly, for the circumferential flaws,
a large margin, but since they can go 360 they're not
limited by their axial. We'll demonstrate in this
case that leakage will be detected very easily from a
partial flaw while there is still a large margin on
the limit load.
MEMBER WALLIS: Are flaws necessarily
axial or circumferential?
MR. MATTHEWS: Well I guess they could be
diagonal, depending on what's driving in and what the
stress --
MEMBER WALLIS: Do they tend to go in
straight lines?
MR. MATTHEWS: Well, they're jagged
straight lines most of the ones I saw. There was a
circumferential flaw underneath at V.C. Summer that
intersected the axial flaw but it was up underneath
the ferritic part of the nozzle. And it grew for a
small distance and even that tended to turn in the
axial direction because of the stresses we believe.
DR. FORD: Sorry, did you say the crack
went into the ferritic steel?
MR. MATTHEWS: No, underneath it. The
ferritic steel is clad on the -- with inconel for part
way and it grew to the ferritic and stopped.
And then finally the short term safety
assessment will present arguments similar to V.C.
Summer's presentation on the January 18 meeting about
the pipe -- that are covered by defense and death.
And piping failure has been analyzed in the SARs and
there's systems in place to mitigate it. Also, visual
inspections for boric acid have been an effective way
of identifying leaks well before there's been any
structural margins affected anywhere.
MEMBER WALLIS: What is it you see when
you see boric acid?
MR. MATTHEWS: Pardon?
MEMBER WALLIS: What do you see when you
see boric acid?
MR. MATTHEWS: Actually you see the
powder.
MEMBER WALLIS: You see solid boric?
MR. MATTHEWS: Yes, solid boric. And
finding those boron deposits on the walk downs has
been an effective way, at least to date, of finding
flaws before there's any structural damage, structural
margin is significantly affected.
For the longer term, the assessment
committee will complete our scope definition,
identifying all the areas of concern. One of the
things we want to do is evaluate the generic
applicability of the hot leg cracking, one of the
elements, so that we'll be looking at finite element
analysis including operational and residual stresses.
We will assess the safety significance of
the issue for all of the components, and then we will
prioritize the locations based on safety significance
into capabilities and the actual experiences in the
field.
The assessment committee is also going to
be charged with determining if any new inspection
requirements are necessary and, if they are, such as
ISI frequency, perhaps the ten year frequency we have
on these may need to be modified. We'll be looking at
that. And they'll assess the research needs. Where
are the holes? They're defined where we need more
information and then define research efforts to get
that information.
One of the areas we'll be looking at
certainly is crack growth data available worldwide.
We have some data, we think there's other data
available in the world and we'll be gathering that
data and factoring all of that into our analyses.
VICE CHAIR BONACA: So when you're talking
about determining inspection requirements, you're
talking about refining the understanding of what
inspection you need to do to detect?
MR. MATTHEWS: I guess the way the
assessment committee would do it would say what do you
need to find and how frequently do you need to look
for it. And the inspection committee, which would be
the next one, would be defining what we've got to do
to find that kind of indication.
VICE CHAIR: Oh so you have -- okay. Yes.
MR. MATTHEWS: The next committee is the
inspection committee. The first thing they wanted to
do was get some guidance out for those plants with
spring outages. I believe that letter was signed
yesterday by Jack Bailey from TVA, the VP there who is
the chairman of the MRP. The goal was to develop a
consistent inspection approach. After looking at it,
the committee and the people that EPRI NDE center both
felt that for these nozzles the ID UT was still
considered the best available technique. It's
considered adequate certainly for the upcoming spring
outages for a couple of reasons.
V.C. Summer's inconel weld was a field
weld that was installed and had multiple repairs on it
done in the field. The ID contour on that weld was
not necessarily the best.
MEMBER WALLIS: Tell me more about UT. I'm
sorry, is this a thing where some diagnostician looks
at some picture? Or is it something where a computer
analyzes a picture, or a computer analyzes certain
facets of an image or what?
MR. MATTHEWS: At least for UT today, the
way it's being done is it's a trained technician
watches the instrument. These are automated
instruments that are --
MEMBER WALLIS: So it's as prone to error
as diagnostic X-rays in hospitals, where someone looks
at a picture and tries to see a crack.
MR. MATTHEWS: Well it's not a picture
though, it's a trace on an oscilloscope.
MEMBER WALLIS: Looks for some anomaly?
MR. MATTHEWS: It's looking for any kind
of anomaly and the data that's taken on these is a
digital form of data, with hot leg alphas anyway if
they're done from ID, it's an automated exam where the
data's gathered and stored and digitized and can be
then reviewed.
MEMBER WALLIS: And can they zoom in or
something? I mean if he thinks there's something
there can he get a magnification of the signal and
things like that?
MR. MATTHEWS: Well, you can go look at it
closer but I mean all the data's there available for
him to look in as much detail as is available.
MEMBER WALLIS: I'm just wondering if
greater attention to detail in the inspection would
buy you some better assessment.
MR. MATTHEWS: We think absolutely and
that's one of the things that we're working, that's
one of the recommendations that we're putting out is
to enhance the awareness of those inspectors to the
kinds of anomalies that led to missed indications at
V.C. Summer.
MEMBER POWERS: And Indian Point and a few
other places. I mean there's been pandemic missing of
indications here.
VICE CHAIR BONACA: I mean I'm somewhat
disturbed by the top bullet, the UT is still
considered the best available technique.
MR. MATTHEWS: That's today.
VICE CHAIR BONACA: I understand but you
know the whole experience we saw in the presentation
says that eddy current was an important complementary
technique to identify indications that are seen as
significant enough to say you should go only one
cycle.
Now if this is true, it has implications
for the other plants too, and it's hard to take then
at face value the statement that ID UT is still
considered the best available technique. I may just
have trouble in accepting both statements at the same
meeting.
MEMBER POWERS: Separated in time they're
okay.
(Laughter.)
MR. MATTHEWS: The inspection committee
and the people that -- everybody there with the
inspection had worked with the results from V.C.
Summer. They were also aware of the indication, or
the information out of Ringhals that those guys, the
eddy current didn't even see some of those flaws that
the UT did see.
VICE CHAIR BONACA: And that's why I used
the word complementary. That together seem to really
yield some information.
MR. MATTHEWS: The problem the industry
has with UT at this point is it's never been used
except at Ringhals and at V.C. Summer. We don't
really know what's in the VNC loop, we haven't really
got a clue in my mind what's there. There's
indications there. They haven't been proven, we don't
know what it is. And to jump in there with an
unproven technique on a plant that's down for a
regular ISI and say, okay, you find a scratch in
there, now you're limited and you've got to come back
and do the cold leg again next cycle.
We felt it was a little premature for the
industry to jump to something like that.
VICE CHAIR BONACA: But for your plant,
still you have restrictions based on ET.
MR. MATTHEWS: Well, it's not my plant,
it's V.C. Summer. No, I'm from Southern Nuclear.
VICE CHAIR BONACA: Oh I thought -- I'm
sorry, all right.
MR. MATTHEWS: The way the mergers are
going I don't think it's my plant.
VICE CHAIR BONACA: I was referring to
that, jut I confused the two, all right.
MR. MATTHEWS: Another thing that the
inspection committee did do is there was a mock-up
available that the EPRI NDE center had of this type of
weld, with imbedded flaws. Now admitted they were
fatigue flaws but it's the best we've got right now.
Some of those flaws are very shallow and we
recommended that the plants, and there's a very
limited number of plants with iconel welds here on the
hot leg that are going in for inspections this spring,
I think there's actually only three plants with
iconel.
And we recommended that those plants have
their vendors perform a demonstration of their
techniques on the EPRI mock-up, and all those vendors
have done so and, as a result of doing those
demonstrations, there have actually been some
modifications to the procedures that were being used.
New transducers and new scanning gains on the
instruments to get a more sensitive examination.
Also, we said that we were going to
enhance the awareness of inspectors and not just be
willing to say, well, I got 90 percent coverage on
that weld, that meets the code. If you get a lift off
do what you can to remedy that situation. And be
aware of what a lift off looks like on the data and
see if there's not something you can do about it.
MEMBER WALLIS: How long does it take to
do this inspection? Have you got something which is
going around and traversing on some track?
MR. MATTHEWS: It's a robot arm typically
that hangs off the vessel. It is done simultaneously
with the belt line weld exams on the vessel by the ten
year ISI. And it will go in and has a sled that will
--
MEMBER WALLIS: Is this a day long
operation type thing?
MR. MATTHEWS: I imagine it's at least
that. I'm not sure how long it takes. Not per nozzle
I would imagine it wouldn't take all day.
MEMBER SIEBER: About 12 hours including
moving into position.
MR. MATTHEWS: Per nozzle.
MEMBER WALLIS: Presumably the signal --
everything comes out as stored information so it's
available at any time.
MR. MATTHEWS: For these particular
nozzles and these particular welds it is. There's
other inconel welds that are not done today with
automated techniques where that's not true. But for
these hot leg and the cold leg nozzles off the vessel,
those are typically automated exams from the ID.
Back on the, well I guess I've mentioned
the three plants that have inconel welds that are
doing 10 year vessel ISIs we understand geometry is a
very big issue here, the ID geometry in the contact of
the sleds. Looking at it, those three plants, the
inconel weld was a shop weld not a field weld and we
firmly believe that the ID geometry for the inconel
part of this weld will be in much better shape than
the situation at Summer. I'm not sure we have ID
contours but typically those shop welds are in much
better shape when we ID than the field welds.
MEMBER SHACK: Can you try to finish up in
five minutes?
MR. MATTHEWS: Oh, okay I'll hurry.
Other things that we're recommending is
that we enhance the sensitivity of the boric acid walk
down and enhance the awareness of the operations and
chemistry people looking for small changes and
unidentified leakage and possible, or notifying them
where these 82/182 welds are.
Longer term actions of the inspection
committee, they need to evaluate the need for
alternative and new techniques and we'll be doing
that. We'll be looking at the evolving capabilities
of the vendors over the years and what new techniques,
if applicable, could be applied.
We'll be looking internationally at what
techniques are available. There may be something
overseas that some of those vendors are using, and we
realize we have to address in the little bit longer
term the geometry concerns for the ID.
We'll evaluate the data we get out of the
spring outages and feed that back in to the fall
plants for further recommendations.
Also, we'll be defining what additional
mock-ups are needed. We'll work with the vendors on
delivery systems, such as the smaller transducer
packages or better articulating tools, and coordinate
the demonstration of capabilities with the PDI -- as
they're developing their mock-ups for qualification of
inspectors. That's a fall 2002 requirement that
bimetallic or dissimilar metal welds be examined by
qualified people.
And then through the NDE center we'll be
providing training and expert help where it's needed.
And, finally, evaluate the impact on risk
informed ISI.
But the industry experience is an integral
part of that risk informed ISI process, so experience
here will have to be factored in, and there's a
required feedback loop in the risk informed ISI
process that takes industry experience and phases it
back in and into the program for those plants that
have already implemented risk informed ISI. And
they'll have to be assessing that impact on their
programs.
Finally, the repair and mitigation
committee's running a little bit behind the others.
It's not quite as urgent for us to be addressing that.
They'll be meeting in March and they will assess the
need for improvements in the repair and mitigation
processes. That will depend to some extent on what
comes out of the assessment in the inspection
committee.
What the repair group will be looking at
will be prioritizing the locations based on repair
mitigation inspection perspective which could be quite
different than a safety perspective. They'll look at
the likelihood and consequences of a failure or a
leak, how difficult is it to implement a repair,
trying to assess where we might need to work with
vendors to come up with better ways to repair or
mitigate the situation.
They're going to create a matrix by
assessing the existing technology, look at what would
be involved in the qualification and demonstration of
a new technique, and where there's any kind of code or
regulatory compliance or involvement, we'll be getting
the NRC and the code people involved early in the
development of those processes.
What is our schedule? We've scheduled a
technical working meeting with the NRC on March 23.
In that meeting we will be going over the detailed
approach of the short term safety assessment and
hopefully by that time we will have many if not all of
the results available, possibly not in their final
form. And we'll solicit their feedback on that short
term safety assessment. The plan is to get that to
them by the end of the month.
We're trying to arrange a visit to the NDE
center by the staff. We'd be working with the staff
and, as Gene said, the staff has already been down and
audited the demonstration of the technology, or the
inspection tools at FTI.
The short term assessment inspection to be
completed in March. The inspection guidance, like I
said, I believe that was issued yesterday. Longer
term, the assessment inspection efforts for June time
frame involve evaluation of the spring 2001 inspection
results and assessment of all the 82/182 welds in the
plant, in the primary system, not just the ones that
we think might be the most likely.
And then even longer term there'll be
continued assessment of all the Alloy 600 applications
inspection and repair mitigation technology and
whatever research efforts we and the staff have to
come up with.
Finally, in conclusion, MRP has taken the
lead for the industry in developing an industry plan
here. We firmly believe this is not really a near
term safety issue because of the margin available in
these welds to failure of the piping. Visual
inspections for boric acid have been effective and
they are effective at finding leaks before there's any
structural integrity threatened. Pipe welds are
covered by defense and death approach has been
inherent in the nuclear industry all along, and we're
performing the short term assessment to demonstrate
that we can continue to operate.
MEMBER WALLIS: I want to ask you about
this effectiveness. Now at Summer, boric acid was
used to find the leak, right.
MR. MATTHEWS: Yes.
MEMBER WALLIS: Suppose it had not been
found for another period of time, how long can it go
on before something worse happens?
MEMBER POWERS: One, two, or more cycles.
MR. MATTHEWS: The leak at V.C. Summer was
through a very, very small pin hole where the crack
had finally made it to the OD and it was a very, very
small leak. It was 1.2 GPM. Before the crack gets
anywhere near a crack size that could threaten a
rupture of the pipe, you'll be leaking tens of gallons
of water per minute and you'll easily pick that kind
of thing up.
MEMBER WALLIS: Yes, but that's not really
addressing the first question. I mean are you saying
more visual inspections are effective, they're only
going to be effective if they're caught on time, early
enough. And if you inspect and you don't see boric
acid and then you wait for so long, it must not grow
in that period of time. How fast is it, I don't have
a feel for how fast it would grow if you hadn't
detected it.
MR. WAGONER: Cycles and cycles and
cycles.
MEMBER WALLIS: Many cycles before there's
a big leak?
MR. WAGONER: Yes, sir. I'm Vaughn
Wagoner from Carolina Power and Light. And the point
is that the, and I'm going to use some round numbers,
but a half an inch a year, okay, three-quarters of an
inch for an 18 month cycle. So you've got two or three
inches of weld metal in axial direction and it's going
to stop on both ends, theoretically you could run for
ever. So you've got numbers of two, three inches
that you'll be able probably to have a discernible
leak and you've got even in the circumferential
directions you've got tens of inches of flaw
capability before you ever get there.
So you've got cycles and cycles and cycles
of margin, even if you're in a circumferential
direction, which is the only one we're really worried
about it for a catastrophic failure. And those are
round numbers, but I think it's the order of
magnitude, I mean it's in the ballpark of what we're
talking about.
MEMBER WALLIS: Is that something the --
agrees with.
MR. WICKMAN: Keith Wickman, NRR.
Critical crack size both axially and circumferentially
are very large. Okay. So on the face of it, yes.
VICE CHAIR BONACA: So really boric acid
will be identifying those before leakage?
MR. MATTHEWS: No. The boric acid comes
from the leak.
VICE CHAIR BONACA: No, I understand that,
I'm saying that you would find it through an
inspection, walking down, you see boric acid and
that's on inspection before you find it through
unidentified leakage. What you're telling me is that
--
MR. MATTHEWS: That's a distinct
possibility. That's exactly what happened at V.C.
Summer.
VICE CHAIR BONACA: Because the growth is
so slow.
MR. MATTHEWS: That's exactly what
happened at V.C. Summer. And we do boric acid walk
downs every outage and they're pretty thorough and
we're enhancing the awareness of the people that are
doing those boric acid walk downs to be sure you trace
it back and don't assume it was a valve. Trace it
back and make sure you know where it's coming from.
That kind of thing. So if there is a leak and it's
been going on for any -- or leaked out any significant
amount of boric acid, we feel quite confident that
we'll find those on the walk downs.
MEMBER SHACK: Coming back to Graham's
question though a little bit, I mean circumferential
cracks are a little more difficult to deal with
because when you have stress corrosion cracking and
you have residual stress pattern, you can at least in
fusion systems where you get very large aspect ratio
cracks, and he's certainly right that if you get a
through wall crack of X inches it will take you a long
time to grow that way.
If the crack is sort of growing at a very
large aspect ratio when it comes through the wall,
then things can get more exciting, which is why the
NRC hasn't allowed leak-before-break in systems that
have been susceptible to stress corrosion cracking.
MEMBER WALLIS: Aspect ratio I mean
there's a long base to the crack, a little tip up here
so --
MEMBER SHACK: Well, relatively shallow
and a long length. The axial flaw are really much
easier because they do, they butt up against the
stainless steel on the one end and the ferritic vessel
on the other, and they're sort of stuck there.
MEMBER WALLIS: Are they stopped forever
there?
MEMBER SIEBER: Pretty much.
MEMBER SHACK: Forever as long as, you
know, on the scale that we're interested in things,
yes.
MR. MATTHEWS: And we'd certainly find
those before they corroded away the nozzle from boric
acid from the OD I think.
Interim inspection guides for near to term
plants has been issued. We will revise that later as
we get more information, as we have better handles on
technology. The longer term assessment in the Alloy
600 and 182 and 82 welds in the PWR primary system and
including inspection repair and mitigation, we'll be
looking at all these things in a little longer term
and we intend to keep the NRC staff fully informed of
everything we do and as we're going along.
Basically, any more questions. Did I do
it in five?
MEMBER SHACK: Close enough.
MR. MATTHEWS: Okay.
MR. BATEMAN: This is Bill Bateman from
the staff. If you don't mind I'd like to make a
couple of comments quickly because I know it's getting
close to lunch time.
I'm the chief of the branch that had to
make the decision as to what to do as the result of
can Summer restart or not. And at least from a code
perspective, the affected weld was totally replaced so
we got out of the code realm when they cut the whole
weld out and put in a whole new weld. So from that
perspective Summer was in compliance with the code.
Regarding the other two indications on the
cold leg, those welds, even assuming a 2:1 aspect
ratio did not achieve 10 percent depth in the pipe,
which would have required a flaw analysis by the code.
So by code requirements there is no requirement to do
a flaw analysis because the assumed depth of that eddy
current indication was not at a 10 percent depth.
However, we took the bounding crack growth
rate and applied it to the one-eighth inch assumed
depth to that crack to assure that that crack would
not exceed the 75 percent through wall. And the time
we came up with was about one cycle.
So those are the conservatisms that the
staff used in coming to the conclusion that it would
be all right for Summer to restart. I want to try and
firm that up. There seems to be some skepticism I
think in terms of our rationale.
So, again, in terms of the code, there was
never an issue with code. Everything was totally in
compliance with the code, we went beyond the code and
basically with the bounded crack growth rate analysis
to make our determination.
And, again, with respect to the crack
growth rate analysis, there's not a lot of data, but
the data we did have we reviewed with our own
expertise and we went to National Lab and got their
advice in terms of whether or not this was a valid
crack growth rate, and whether or not our assumptions
were valid. And all the feedback we got led us to use
the data that we did with confidence that we would not
have a problem prior to an inspection after one cycle
of operation.
MEMBER SHACK: Any additional comments?
No. Mr. Chairman.
VICE CHAIR BONACA: With that I think we
will take a break for lunch and we will come back at
one o'clock.
(Whereupon, the above-entitled matter went
off the record at 11:56 a.m.)
Page Last Reviewed/Updated Monday, August 15, 2016
Page Last Reviewed/Updated Monday, August 15, 2016