Protecting People and the EnvironmentUNITED STATES NUCLEAR REGULATORY COMMISSION
Official Transcript of Proceedings
NUCLEAR REGULATORY COMMISSION
Title: Advisory Committee on Reactor Safeguards
486th Meeting
Docket Number: (not applicable)
Location: Rockville, Maryland
Date: Thursday, October 4, 2001
Work Order No.: NRC-039 Pages 1-265
NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers
1323 Rhode Island Avenue, N.W.
Washington, D.C. 20005
(202) 234-4433. UNITED STATES OF AMERICA
NUCLEAR REGULATORY COMMISSION
+ + + + +
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
486TH ACRS MEETING
+ + + + +
THURSDAY
OCTOBER 4, 2001
+ + + + +
ROCKVILLE, MARYLAND
+ + + + +
The Advisory Meeting met at the Nuclear
Regulatory Commission, Two White Flint North, Room
2B3, 11545 Rockville Pike, at 8:30 a.m., Dr. George E.
Apostolakis, Chairman, presiding.
PRESENT:
DR. GEORGE E. APOSTOLAKIS, Chairman
DR. MARIO V. BONACA, Vice Chairman
DR. DANA A. POWERS, Member
DR. WILLIAM J. SHACK, Member
DR. THOMAS S. KRESS, Member at Large
DR. JOHN D. SIEBER, Member
DR. F. PETER FORD, Member
DR. GRAHAM B. WALLIS, Member. ACRS STAFF:
DR. JOHN T. LARKINS, Executive Director
DR. PAUL A. BOEHNERT, Executive Secretary
MEDHAT EL-ZEFTAWY, ACRS Staff
DR. ROBERT ELLIOT, ACRS Staff
CAROL A. HARRIS, ACRS/ACNW
DR. JAMES E. LYONS, ADTS
SAM DURAISWAMY, ACRS
DR. SHER BAHADUR, ACRS
PRASAD KADAMBI
. I-N-D-E-X
AGENDA ITEM PAGE
Opening Remarks by ACRS Chairman . . . . . . . . . 4
Duane Arnold Core Power Uprate . . . . . . . . . .10
Readiness Assess for Future Plant Designs. . . . 100
and Staff Proposal Regarding Exelon's
Regulatory Licensing Approach for the
Pebble Beach Bed Modular Reactor
Action Plan on Steam Generator Tube. . . . . . . 202
Integrity
Proposed Resolution of Generic Safety. . . . . . 247
Issue 173A, Spent Fuel Storage Pool for
Operating Facilities
. P-R-O-C-E-E-D-I-N-G-S
(8:30 a.m.)
CHAIRMAN APOSTOLAKIS: The meeting will
now come to order. This is the first day of the
486th meeting of the Advisory Committee on Reactor
Safeguards.
During today's meeting the Committee will
consider the following:
The Duane Arnold Core power Uprate; and
the Readiness Assessment for Future Plant Designs and
the Staff Proposal Regarding Exelon's Regulatory
Licensing Approach for the Pebble Beach Bed Modular
Reactor.
The Action Plan to address ACRS Comments
and Recommendations Associated with the Differing
Professional Opinion on Steam Generator Tube
Integrity.
And the Proposed Resolution of Generic
Safety Issue 173A, the Spent Fuel Storage Pool for the
Operating Facilities; and the Proposed ACRS Reports.
A portion of this meeting may be closed to
discuss General Electric Nuclear Energy proprietary
information applicable to the Duane Arnold core power
uprate.
This meeting is being conducted in
accordance with the provisions of the Federal Advisory
Committee Act. Dr. John T. Larkins is the designated
Federal official for the initial portion of the
meeting.
We have received no written comments or
requests for time to make oral statements from members
of the public regarding today's sessions. A
transcript of portions of the meeting is being kept,
and it is requested that the speakers use one of the
microphones, and identify themselves, and speak with
sufficient clarity and volume so that they can be
readily heard.
We will begin with some items of current
interest. You all have this document with items of
interest, and it contains a speech by Chairman Reserve
on the growing area of radiation protection of
patients. He gave his speech at the IAEA general
conference of senior regulators.
And also there are some safety significant
findings that would be of interest to us when we
discuss the reactor oversight process. There was a
yellow finding at the plant and six white findings.
I found them very interesting to read for later this
afternoon.
The NRC is putting together the 2002
regulatory information conference, and requesting
suggestions for agenda items. If you have any, you
have by October 5th to send them to the staff.
And Dr. Larkins has a short announcement to make.
John.
DR. LARKINS: Yes. Good morning. Members
of the public and all non-NRC employees attending this
meeting to be escorted when leaving this floor. It
would be appreciated if people would leave at the end
of the session, such that an ACRS office individual
can escort you to the first floor.
An escort will be available at the end of
the session, or each session, to take individuals down
to the first floor on the elevator. Obviously, this
does not preclude individuals from using the
facilities, the restroom facilities and others on this
floor, as a guard will be stationed outside of the
meeting room.
We appreciate your cooperation in what we
hope will be a temporary situation. Secondly, I would
note for ACRS members and staff, that the Deputy
Executive Director for Management, Ms. Pat Nouri, will
give a presentation at one o'clock today in this
conference room on security issues at I and II White
Flint Complex, and other matters.
You are invited to either have your lunch
before or during this discussion, and she will answer
questions related to matters surrounding the operation
of the two buildings. Thank you.
CHAIRMAN APOSTOLAKIS: Two more items. By
a voice vote on September 26th, the Senate confirmed
former Commissioner Diaz to serve a second term as an
NRC Commissioner, through June 30th, 2006. Dr. Diaz
will be sworn in today for his second term.
And the last item is that because of
illness, our member, Graham Leitch, will not be
attending this meeting. He is at home recuperating,
and of course we all wish him a speedy recovery.
Now, back to the agenda. The first item
is the Duane Arnold Core Power Uprate. Dr. Powers is
the Cognizant Member. Dana.
DR. BONACA: I would like to make a
statement. I have a conflict of interest.
CHAIRMAN APOSTOLAKIS: So you will keep
quiet for a change?
DR. BONACA: I will try to keep quiet.
DR. FORD: On the Duane Arnold issue, I
have a conflict of interest being a former GE
employee.
CHAIRMAN APOSTOLAKIS: It is going to be
a quiet meeting. Dr. Powers.
DR. POWERS: Gee, I don't have any
conflict of interest. Sorry. Well, we are going to
discuss the application for a power uprate from Duane
Arnold. It has significance for us, and it is the
first that we have heard of many that we expect to
come in the future.
Many, if not all, of the BWRs, were
designed for much higher powers than they have been
operating for the last several years. They were kept
at a somewhat lower powers because of concern over the
ATWS and the reactor stability issue back a long time
ago, and part of that concern came from the ACRS
itself.
So it is somehow fitting that we should
hear about the relief from that concern, which has
come about in a generic way from General Electric.
When we go into this power uprate, you are going to
see that the applicant and the staff have addressed a
host of issues, quite a lengthy list of issues that
had to be addressed in looking at these power rates.
But in truth, they are relatively few that
are of particular concern. Those tend to be the ATWS,
the operator response times, material degradation, and
some infrequencies in the containment response
capabilities, and I think that is where the committee
should be focusing its attentions.
The application that is being made here is
not a risk informed application. This is a classic
deterministic application. But I think you are going
to find that risk is a language that has crept into
the classic deterministic analyses, and has played a
role.
And in some cases I might say the
applicant has done some very imaginative and creative
things in the use of risk, even though he has a
deterministic application here.
And I think it would be great fun to have
them back and go through in some detail the risk
portions of what they have done with their plant,
because I think we would find it interesting on what
they have been able to do, and what they have been
able to learn from risk.
But that is not our focus here. The order
of presentations today is going to begin with a
presentation by the applicant himself. We have given
him a very easy chore. We have only asked him to
compress 4-1/2 hours of a very detailed presentation
into 26 minutes or so.
And in looking at his view graphs, he has
done a manful job of doing this, but do recognize that
he is giving a synoptic account of all that they have
done; and that will be followed by the staff, who are
equally time constrained for their -- what amounted to
a little over 4 hours of presentation on their part
before the subcommittee.
So with that, I am going to introduce Ron
McGee to begin the presentations, and Ron, you will
introduce Tony and other people as they present.
MR. MCGEE: That's correct. Thank you,
Dr. Powers. Good morning. As he stated, my name is
Ron McGee, of the Nuclear Management Company at the
Duane Arnold Energy Center.
We have been asked to provide an overview
of the material that we presented to the subcommittee,
the Thermal Hydraulic Subcommittee, and as a
subcommittee members kind of test, we provided a large
amount of information.
And hopefully we can address any of your
questions. As you can see from the introductory
slide, our original rated thermal power was 1593
megawatts thermal, and we uprated the plant in the
early 1980s, and are currently licensed for operation
at 1658 megawatt thermal.
The application that we have before the
staff at this time is for operation at 1912 megawatts
thermal. We prepared a deterministic application in
accordance with the previously approved General
Electric topical reports.
Although the application was not a risk-
informed submittal, we have performed an impact review
utilizing our PRA, and I will point out how that
information was utilized in some of our upcoming
slides.
Concerning the modifications for power
uprate, we installed all of the necessary safety
related hardware changes during the 2001 spring
refueling outage. For the balance of the plant, we
have decided to implement the power uprate in two
phases.
They are the modifications in Phase I for
operation up to 1790 megawatt thermal, which were also
accomplished during our spring refueling outage. The
remaining modifications will be installed during a
future refueling outage.
Operator training. Once the engineering
evaluations were completed and the modifications were
designed, we began training the operators on the
impacts of power uprate.
In the classroom, we emphasized the design
basis changes, and explained the plant equipment
modifications. In the simulator, we showed them
static and dynamic examples of the most significant
changes, and then put the crews through routine
simulator accident and transient scenarios.
The PRA analysis and the engineering
evaluations each pointed out the importance of the
ATWS event as we mentioned earlier. Specifically, the
operators' ability to correctly inject standby liquid
in a timely fashion is critical to mitigate this
event.
So we emphasize this point in the
classroom, as well as in the simulator training
sessions. The revised PRA assumes a 20 percent
failure rate for injecting standby liquid.
In reviewing past performance records, we
found 58 evaluated ATWS, with a 100 percent success
rate for the operating crews completing this task for
injecting standby liquid.
DR. POWERS: This is a striking thing. I
mean, you go through the human reliability analysis,
and you come up with .1 and .2, and some sort of
failure rate. You get a hundred percent success rate
in your training exercises.
And I looked at some of the studies that
have been done in the past, and while I find for in
NUREG CR 3737 that they went through an analyses, and
they came to the conclusion that there was no
relationship between scores during training exercises,
and air rates by plants. Do you have any thoughts on
that?
MR. MCGEE: We merely used the information
provided to us by the training session as a benchmark
so that we were certain in our minds at least that the
failure rate that we had chosen was at least
conservative.
So utilizing the industry acknowledged
standard for error rates for operators, we were
confident that we were bounded by that particular
document.
CHAIRMAN APOSTOLAKIS: Where did that 20
percent come from?
MR. MCGEE: Brad, could you address that,
please, the 20 percent that we assumed in the PRA
analysis for the operator failure rate for injecting
standby liquid.
MR. HOPKINS: This is Brad Hopkins from
the NMC. The error rates that we used come from a
variety of industry standard methods for determining
human error probabilities. So we have a formula that
considers the complexity of the actions, and how time
they have to achieve the actions.
DR. POWERS: Can you give us an idea of
how the time changed, and from your originally
licensed power to the higher level, or the
intermediate to the higher, whichever one is easiest
for you to do.
Well, for the key operator action, or for
the ATWS event, is the time available for injecting
standby liquid control, and there we have values for
early injection, and one for late injection.
For early injection, the time decreased
from six minutes to about four minutes; and for late
injection, I believe from 15 minutes down to about 12
minutes.
CHAIRMAN APOSTOLAKIS: So what was the
failure rate when the time was 6 minutes?
MR. HOPKINS: You will have to excuse me.
I have a slide with that number on it. Just a minute.
CHAIRMAN APOSTOLAKIS: So are you planning
to talk about these things later?
MR. HOPKINS: Yes.
MR. MCGEE: We do have a slide
presentation on PRAs.
CHAIRMAN APOSTOLAKIS: So you know then
what question I am going to ask.
MR. HOPKINS: Yes. Here is the answer.
From .11 to .18.
CHAIRMAN APOSTOLAKIS: And we will have
some questions on that later.
MR. HOPKINS: Okay. Very good.
DR. POWERS: In my thinking on this
subject is the human reliability analysis has done its
job here, and that you record a relatively high
potential for failure in this operation, and
consequently you train on that.
And we can't just help feel that training
must help in keeping that number lower than perhaps we
calculated here. We may not know exactly what it is,
but at least we are providing experience and training
with this kind of an event.
MR. HOPKINS: That's correct.
DR. POWERS: That is my interpretation of
what has happened here.
MR. MCGEE: Unless there are other
questions at this point, I would like to turn the
presentation over to Tony Browning.
MR. BROWNING: Good morning. Once again,
my name is Tony Browning, and I am with the NMC at the
Duane Arnold Plant. I have the privilege this morning
to present to the ACRS the results of our thermal
hydraulic evaluations for EPU, our reviews of plant
materials in the EPU environment, and finally our
investigation into risk insights from EPU operation,
using our probablistic risk assessment methodologies.
Today I will briefly summarize our
evaluation in these key thermal hydraulic analysis
areas. For the ATWS EPU evaluations, we analyzed the
four bounding events identified by the generic studies
in the ELTRs:
The main steam isolation valve closure
transient, and the pressure regular failure open
transient, and the loss of off-site power transient,
and the inadvertently opened relief valve transient.
These evaluations were performed using NRC
accepted methods and assumptions. This is a
deterministic evaluation, with conservative
assumptions and acceptance criteria; as opposed to our
more realistic or best estimate evaluations performed
with our PRA models, which I will discuss later in a
presentation.
The purpose of this evaluation is to
demonstrate compliance with prescriptive hardware
requirements of the ATLS rule, 10 CFR 50.62, by
showing conformance to the underlying analysis basis
for BWRs as documented in the GE topical report NEDE-
24222.
As we can see the results were all within
their respective acceptance criteria, with margin.
Thus, the DAEC will continue to comply with the ATWS
without changes to the existing plant hardware.
Next I would like to discuss our
evaluation of thermal hydraulics stability. First, I
would like to start off with some background
information. The DAEC has implemented the stability
solution called Option 1-D.
The key point of the Option 1-D solution
is that it has been demonstrated that these plants,
through their inherent design characteristics, are
only susceptible to core wide or fundamental mode
oscillations, and not the regional or higher harmonic
oscillations.
This greatly simplifies the solution
approach. This solution utilizes a combination of
prevention with detection and suppression measures to
conform to general design criteria 12.
DR. POWERS: What this means is that you
are getting parallel channel flow problems?
MR. BROWNING: No. That's the regional
mode, where you have one side of the core oscillating
out of phase with the other side. Our plant is not
susceptible to that mode of oscillation.
It is only the fundamental mode where the
code is oscillating in unison.
DR. POWERS: It is sometimes called the
direct loop oscillation?
MR. BROWNING: Correct.
DR. POWERS: And it is an NED24222 that I
will find the mathematics on this?
MR. BROWNING: No, the General Electric
topical report that I referenced earlier was for the
ATWS evaluation. I'm sorry, but off the top of my
head, I don't remember the topic number for that
solution.
DR. POWERS: I have looked, and I cannot
find the underlying analyses that support your
contention.
MR. BROWNING: It is in one of the early
topicals, but I don't know it off the top of my head.
DR. POWERS: If you happen to find that
after we are done here, I would sure appreciate
looking at it.
MR. BROWNING: Very good.
DR. POWERS: I am willing to believe right
now, but --
MR. BROWNING: Yes. Prevention is
accomplished by establishing this exclusion zone right
here, this red line, on the power to flow map.
Operation is restricted in this region inside of here.
Thus, we prevent the oscillations by
affording the area of operation most susceptible to
instability, and we introduce a 20 percent margin by
using a conservative criteria of 0.8 for the
calculated K-ratio used to establish this boundary.
We introduce additional margin by
establishing this buffer zone, represented by the
orange line, by adding another 0.05 to K-ratio margin
to the exclusion zone, where operation is allowed only
when the SOLOMON software is available on the plant's
core monitoring computer.
The SOLOMON software is the same model as
ODSY, the frequency domain code used to calculate the
decay ratios used to establish the exclusion zone and
buffer zone boundaries.
So we are only allowed to operate in this
region between the two lines when SOLOMON is available
to provide the operators a prediction of their margin
to an unstable condition.
Otherwise, operation is prohibited in the
boundary zone region as well.
DR. KRESS: This is a picture of how you
start up and shut down?
MR. BROWNING: Correct.
DR. KRESS: Are there accident conditions
that will force you into that zone?
MR. BROWNING: Yes, there are. Any number
of transients, either single or dual pump --
DR. KRESS: Will take you down the mellow
line.
MR. BROWNING: -- will put us down into
this region, right. And the operators are instructed
any time they enter the exclusion zone to take
prescriptive measures to leave that zone immediately.
And if they notice any instability
condition on their in-core monitoring, they are to
SCRAM the plant immediately. Detection and
suppression comes from the flow bias neutron flex
reactor trip signal.
We validate this capability by
demonstrating analytically that any oscillation will
be suppressed by this flow bias SCRAM prior to the
fuel experiencing a transient that would exceed the
safety limit minimum critical power ratio.
While the major impact of extended power
uprate on thermal hydraulic stability is through the
introduction of MELLLA, which is raising the rated
load line from the black line to the blue line, which
expands the size of the exclusion and boundary zone
regions on the power to flow map, the operational
impact is acceptable as seen by this actual plant
start-up trace.
As we see, it is possible to maneuver the
plant around these zones, and thereby introducing more
margin in the stability.
DR. POWERS: When you plot power versus
flow, if I were to look at power versus time, would I
see a continuous curve here, or would I see a lot of
steps and overshoot and undershoots of that curve?
MR. BROWNING: I will let Steve answer
that, our plant operator.
MR. KOTTENSTETTE: I am Steve
Kottenstette, and I am an operations shift manager at
the plant. Normally over time, you will see us come
up in power and stabilize to do some annual requested
testing, and then go up.
We don't have overshoots or anything. We
just pretty much gradually go on up in power.
DR. POWERS: So where you come up and
enter your buffers or touch your buffer zone margin at
about 25 million pounds per hour, I would not see you
jumping in and out of that buffer zone?
MR. KOTTENSTETTE: No, you wouldn't.
Remember that the start up here is to our current
1658, and the power flow map that you have there is
actually what we will have.
DR. POWERS: I understand that. I am just
asking generically what the curve is. It seems to me
that the curve could have been drawn with a much wider
pencil if there was a lot of overshooting and
undershooting, and things like that. You are really
telling me that I am really looking at the outer
bounds on it?
MR. KOTTENSTETTE: Right.
MR. BROWNING: Thus, by establishing
conservative boundaries for the exclusion and boundary
zones and demonstrating the detect and suppress
capability of the flow by flux SCRAM to ensure that
safety margins are maintained under extended power
uprate operation.
And as seen from our start up example,
adequate operating margins exist under extended power
uprate as well. Now I would like to move on and
discuss the impact of a potential --
DR. WALLIS: While you have got this
figure up there, you are a hundred percent power, and
you are asking for 1912 megawatts, and it looks like
if it is a very small region of flow rate that you can
be in to be at the hundred percent power.
MR. BROWNING: Right.
DR. WALLIS: And you are right in the
corner of that graph, and the question is whether you
can really keep it as close as that without stepping
over some boundary. There is very little room for
error up in that corner.
MR. BROWNING: A very astute observation,
Dr. Wallis. Our reactor engineering crew is going to
be challenged to find drive patterns that will allow
maneuvering in here. Most likely what will happen is
that we will have to be very slow and deliberate in
this region to ensure that we don't encroach on the
boundaries.
And so again you are very astute. It is
going to be somewhat of a challenge to operate up in
this very tight corner of the power plant.
DR. WALLIS: And as your fuel burns up,
you will have different work patterns and so on in
order to maneuver in there.
MR. BROWNING: Correct.
DR. WALLIS: So you might well find
yourselves operating at 95 percent power for quite a
while until you learn how to get up there.
MR. BROWNING: Yes. We will need to have
some operating experience in this region in order to
better refine our capability there.
MR. MCGEE: And also -- this is Ron McGee
-- during our interim plant, or during our phase one
session, we expect to accumulate quite a bit of
operating experience because we will have a larger
flow window. We will be operating between our current
and the expected maximum allowable.
MR. BROWNING: Right.
DR. WALLIS: And if you look at what you
do now, that wanders around in an almost erratic way
as you search for broad patterns near the top in the
present plant.
MR. BROWNING: Yes, and also compensating
for Zenon as things build in. Now I would like to
move on and discuss the impact of a potential ATWS
instability during DPU operation.
As part of the closure of thermal
hydraulic stability issues, generic studies were
performed to determine whether the combination of a
core wide instability event with failure to suppress
the oscillations by a plant SCRAM because of an ATWS
event would lead to significant fuel failure.
And if so, to determine mitigating
strategies that would minimize these impacts, because
the first study did confirm that such an event,
assuming no mitigation at all, would lead to
unacceptable fuel cladding failure.
And a second study was conducted to find
mitigating strategies. The conclusion of the second
study confirmed that existing ATWS strategies
implemented by the BWR owners group emergency
procedure guidelines were effective in precluding
these fuel cladding failures.
And these being lowering the water levels
below the feed water sparger, which reduces the core
inlets subcooling, and which lowers the magnitude of
the power spikes during the oscillations.
And, second, and in the lower term,
injection of boron through the standby liquid control
system to completely dampen the oscillations.
DR. POWERS: When these studies were done
did they consider the power profiles similar to the
type that you will have once you start operating at
the higher power?
MR. BROWNING: Yes, and that is a good
seagueway. The generic studies were found to be
bounding upon Duane Arnold because they had previously
considered operation in the MELLLA region.
And they were also looking at peak bundle
powers significantly higher than what Duane Arnold
will be operating after an extended power uprate. So
it is the peak bundle power that drives the response,
and that was bounding upon us.
DR. POWERS: I guess what I am questioning
is whether the peak bundle power is really what is
limiting here, or does it make a difference how that
power varies along the length of the core?
And the reason for asking the question is
that it is fairly simple. You lower the water levels
so that the collapsed level is below the top of the
core. So you are relying on a certain amount of steam
cooling for the upper region of the rods.
Now, your upper regions of the rods have
a higher decay power than they would if you had a
classic cone type of power distribution. You have a
little different one now.
And so what I am asking you is does that
make any difference in this recovery process or is it
being looked at?
MR. BROWNING: Unfortunately, I am not the
person to ask that. We don't have our General
Electric experts with us today.
DR. POWERS: Gee, with all the conflicts
of interest at the table, maybe they could answer.
MR. BROWNING: One of the things that is
important in the extended power uprate for Duane
Arnold is that we are going to the GE14 fuel design,
which has partial length fuel rods.
So that tailors the power shape in the
upper region to keep it from being overly top-peaked.
So by that combination, I think if we did do the
investigation we would find that that would be the
factor that would keep us bounded by the study. And
you have stolen most of my thunder of this
presentation.
DR. POWERS: Keep covering your thunder.
Don't stop.
MR. BROWNING: And one of the important
points of this is what we have touched on; is that the
peak bundle power under extended power uprate is not
increasing from where it is today.
What we are doing is flattening the radio
profile and raising the average core power so that the
peak bundle response is not changing from where we are
today.
DR. POWERS: I have to admit that when you
first came into this application for a power uprates,
and you find that nothing seems to change, you are
wondering if it is done with smoking mirrors here.
MR. BROWNING: That is a astute
observation, because you find through this exercise
the parameters that do drive the response, and it
turns out in many cases that just basic power level is
not one of them.
But this is an area of containment where
we will see it. We will see it. So, for the DAEC
extended power uprate, we reanalyzed the containment
response using previously approved NRC calculational
models and assumptions for the FSAR events.
To illustrate the impact of EPU, let's
look at both the short term and long term cases for
the design basis loss of coolant accident. First, for
the short term response, the increased subcooling due
to EPU increases the blow down flow rate, which
directly drives the dry wall pressure response.
As we see here the dry wall pressure
increases slightly due to EPU, but we are not
increasing the reactor pressure, and the impact of EPU
is not dramatic in the short term. So, we have
learned it is the reactor pressure that drives this
response, and not the sub-cooling.
However, because the decay heat power
increases proportional to the increase in core thermal
power, the long term impact is a bit more noticeable
as you see here.
DR. WALLIS: But that temperature depends
upon the temperature that it starts out at doesn't it?
MR. BROWNING: That is correct.
DR. WALLIS: So how close is the initial
temperature controlled?
MR. BROWNING: That is controlled by
technical specifications in our license. We are not
allowed in steady state operations to go above 95
degrees unless --
DR. WALLIS: So this is calculated
assuming you are at 95 when you start?
MR. BROWNING: That is correct. That is
a conservative input measure.
DR. ROSEN: At the time that you reach the
215.3 degrees in the suppression pool what is the
pressure in the containment?
MR. BROWNING: At the corresponding time?
DR. ROSEN: Yes.
MR. BROWNING: Do you have that number,
Al?
MR. RODERICK: Not off the top of my head,
no.
MR. BROWNING: Do we have one of our back
up slides?
(Brief Pause.)
MR. BROWNING: We are looking it up for
you. We will move on then. This impact is also true
in the calculation of net positive suction head in the
emergency core cooling system, plus taking suction
from the suppression pool.
The hotter pool, due to increased decay
heat from EPU, leads to an increase in the amount of
over pressure required to ensure adequate MPSH. It
should be noted that Duane Arnold has always been
licensed to allow over pressure for meeting adequate
MPSH.
As we see here the DAEC's dependence is
not in the short term, but only in the long term. We
also see that we have adequate margin between the 5.3
psi that is required, and the 13.3 over pressure that
is available during the peak suppression pool
temperature for MPSH.
DR. POWERS: I think that gives us the
pressure doesn't it?
MR. BROWNING: It is similar, but this is
a slightly different analysis, with slightly different
assumptions. So it is not quite the answer you were
looking for.
DR. KRESS: Usually in the long term the
pressure corresponds to the saturation pressure from
the temperature in the pool. That is probably pretty
close.
MR. BROWNING: It is very close.
DR. POWERS: It seems to me that if I was
going to have to have over pressure giving me enough
net positive section head, I would want it in the
short term and not want it in the long term. And it
seems exactly the opposite here.
And things are going to degrade and I am
going to lose pressure and it is in the longer term
and not in the short term.
DR. KRESS: Well, I don't know when they
call to require that section head. It is probably
needed in both short term and long term.
DR. ROSEN: I think what they are saying,
Tom, is that it is not needed in the first 10 minutes.
It is beyond 10 minutes where it is needed.
DR. WALLIS: But the two are interrelated.
I mean, the amount of containment pressure and the
amount of MPSH you need is sort of interrelated,
because the temperatures are interrelated anyway.
MR. BROWNING: A lot of it has to do with
the density of the water in the pool, and so that is
what happens, and that is what you see in the longer
term.
It takes a while for the pool ot heat up,
and then at that point with the new strainer designs
that we have, and the assumption of the debris
loading, it is not until much later when we get into
a position where we require the over pressure in order
to meet MPSH.
All right. This is the time response in
hours for the event, and as we see here, the red line
is the required MPSH for the core spray pump, which is
bounded over the RHR pump, and we can see here the
time frame in which we need the over pressure as we
cross over atmospheric. So it is after about the
first hour into the event, and lasts until about 23
hours.
DR. WALLIS: That's what I mean about the
temperatures being interrelated. These curves all
have the same shape, and if you are going to change
one by something, then probably the others will change
as well. If you have some sort of containment, then
they will all change.
MR. BROWNING: Right, and you can see here
that this is the actual response, and it is
significantly more throughout the duration.
DR. POWERS: So we have about a one day
window in which we need over pressure.
MR. BROWNING: Correct.
DR. POWERS: And so we can't have any
degradation of the drywall pressure boundary during
this period, right?
MR. BROWNING: That is correct. One of
the assumptions that goes into this calculation -- and
Mr. Roderick can correct me if I am speaking out of
turn here, that one of the assumptions for this
analysis is we assume twice the tech spec allowable
leakage rate for the containment to do these
calculations. So it is a conservative calculation.
DR. POWERS: Well, why did you pick twice?
Why not 10 times?
MR. RODERICK: This is Al Roderick with
Duane Arnold. The assumption of a 5 percent leakage,
which is a little over twice the tech spec limit, is
consistent with the way that the containment analysis
was done when we established over pressure for the
original license. And that assumption was used at the
original plant licensing.
DR. POWERS: Well, you didn't really tell
me what the underlying scheme is.
DR. WALLIS: It is really a factor of two
because this is thermal hydraulics. If it were a PRA,
you would use a factor of 10.
DR. POWERS: I understand. And if it were
metallurgy, we would use a factor of a hundred, right?
DR. BONACA: I have a question. Did you
have to make any changes to your EPGs?
MR. BROWNING: As a result of?
DR. BONACA: As a result of -- well, these
issues, but also power uprate in general.
MR. BROWNING: No, we did not. There is
subtle changes in some of the parameter graphs that
are driven by the decay heat levels, but the actual
flow charts with the operator actions and precautions
were not changed as a result of the EPGs.
DR. BONACA: When you went through a total
review of the EPGs?
MR. BROWNING: That is correct.
DR. POWERS: This requirement for over
pressure for net positive section head, however, is
not qualitatively different than what was required in
your original license, and there is simply a
quantitative difference?
MR. BROWNING: That is correct. The shape
of this curve is fundamentally the same as it was in
the original license.
DR. POWERS: Was the period of time that
you needed for the net positive section head, has that
changed?
MR. BROWNING: I can't answer that today,
Dr. Powers. We would have to go back and look, unless
Al has information on that. Were you able to do some
background there?
MR. RODERICK: This is Al Roderick. Just
from the standpoint of the increase in the power
level, the decay heat is going to be running you out
further.
So given that the pool temperature really
drives when you need the over pressure, I would say
the time from the time that the overpressure is needed
at 1593, compared to 1912, yes, we are going to need
a longer period of time.
DR. POWERS: The curves are all the same
shape. It is adding one thing. It is not like we
have a very long tail here.
MR. RODERICK: While I am up here, the
answer -- you asked what the containment pressure was
when we are at the peak pool temperature of 215.3 for
this event, in looking at the graphical results, we
are about at 20 psig for containment pressure at that
point in time.
MR. BROWNING: If there are no more
questions, I would like to move on, and we would like
to talk about ECCS analysis that was performed. The
DAEC has utilized the SAFER/GESTR methodology for ECCS
analysis prior to EPU and this is an entire change for
us.
Under the SAFER/GESTR LOCA methodology
approved under the provisions of SECY 83-472, for the
use of nominal or more realistic models, dual
acceptance criteria, are applied.
First, the licensing basis peak cladding
temperature is calculated using the required Appendix
K inputs to demonstrate conformance to the 50-46
acceptance criteria of 2200 degrees fahrenheit.
The second acceptance criteria is on the
so-called upper bound PCT, which is calculated from
the nominal inputs statistically adjusted for the
uncertainties in the models due to both generic and
plant specific inputs.
The resulting upper bound PCT is first
compared to the licensing basis PCT to demonstrate
that the licensing basis calculation is higher, and
does a bounding result.
And then we ensure that the upper bound
PCT remains within the 1600 degree fahrenheit limit
placed on the methodology, which ensures that the
results stay within the bounds of the test data on
which it is based.
DR. POWERS: Let me make this very clear.
Is there two different analyses, with different sets
of assumptions, going into them?
MR. BROWNING: Correct. And it is
graphically depicted here. This figure is a bit busy,
but it succinctly presents a number of key points
about the analysis.
First, we see that for small and large
breaks, they were analyzed to confirm that the large
break, or the dba case here, remains limiting under
extended power uprate conditions.
Next we see that the upper bound PCT is
indeed less than the licensing basis PCT, and below
its 1600 degree limit. In addition, we see that the
licensing basis PCT has significant margin to its 2200
degree limit.
Now, comparing the current and EPU
results, we notice that here in the small break LOCA,
we see a slight difference. But as we move up at the
licensing basis calculation, the EPU does not have as
big an impact on the licensing basis PCT.
And this is because the peak bundle power
is not changing from the pre-EPU conditions, which is
what drives the PCT calculation. Thus, we conclude
that under EPU, we have both substantial safety
margins as required by the regulation, as well as a
significant operating margin here. This concludes my
presentation. If there are no further questions in
this area --
DR. ROSEN: I do have a question. I would
like to come back to the question and answer we had a
moment ago about the peak pressure, and if you would
go back to your slide nine.
(Brief Pause.)
DR. ROSEN: And when we talked about the
suppression pool temperature in the long term, and the
EPU conditions of 13.3 degrees, it was offered from
the floor that that pressure is 23 psig in the
containment at that time.
But if you go to the next chart in the
long term, what we see there is that the pressure is
13.2 psig over pressure at the peak at the suppression
pool temperature.
And I don't understand the distinction
between those two numbers, and perhaps you could clear
that up.
MR. BROWNING: The 13.2 psig is the over
pressure required, and the absolute power is 28 pounds
at that point.
DR. ROSEN: But I am comparing the 20 psig
that was offered from the floor at that similar
condition, and there was a statement made that they
are not exactly comparable.
Hence, the difference between 13.3 and 20,
but I don't understand the reasons why they are not
exactly comparable. I don't expect those numbers to
be the same.
MR. RODERICK: This is Al Roderick. Those
are two separate analyses that use different
assumptions. The DBA LOCA is using assumptions that
will maximize pool temperature, and takes no credit
for heat syncs, et cetera.
When we do a containment analysis looking
at over-pressure, the assumptions that we make in that
model, while we have to balance the needs, we are
doing two things.
One, we are trying to maximize pool
temperature, and at the same time we are trying to
minimize containment pressure. So it is its own
containment analysis. So that is why you see a lower
pressure in the over-pressure for MPSH at 13.3 at its
peak temperature, versus the 20 pounds in the DBA
LOCA.
DR. ROSEN: It's a question of doing the
conservatisms differently because of the two different
acts or two different conditions. In one case, you
are trying to show conservatively that you have enough
MPSH, and hence you could up with a lower number.
MR. RODERICK: Correct.
DR. ROSEN: And in the other case, you are
doing the DBA calculations to look at the containment
response, and the margin to containment design
conditions.
MR. RODERICK: That's correct.
DR. ROSEN: So I understand now. So what
it is then is both of those analyses, if done
conservatively, so that you have an appropriate margin
of conservatism for the parameter of interest for that
analysis.
MR. RODERICK: That's correct.
DR. ROSEN: Thank you.
MR. BROWNING: Let's now move on to a
discussion on BPO and pictorial materials. Before I
begin with my formal presentation, we would like to
take this opportunity to address the subcommittee's
open item from last week's meeting.
For the record, we have included in the
handout package a copy of the written response that
explains in the increase in stress on the main reactor
flange from the increase in temperature due to EPU.
And I would be happy to address any
questions that the Committee might have on this
particular issue at the end of our prepared
presentation.
With that, I would like to begin our
discussion on the impact of EPU on materials, and
specifically our programs for addressing flow assisted
corrosion, otherwise known as erosion/corrosion and
in-service inspection of the reactor vessel internal
and associated piping systems.
DAEC has modeled its carbon steel piping
systems, which are susceptible to flow assisted
corrosion, using the EPRI CHEKWORKS package. The
changes in flow and temperature profiles due to EPU
have been modeled in CHEKWORKS and its resulting
predictions for wear are being incorporated into the
second and most important part of this program, which
is the actual inspection of piping for wall thinning.
These CHEKWORKS predictions tell us to
expect a slight increase in wear after EPU, on the
order of a half, to one-and-a-half mils per year. But
based on our base line from previous inspections, this
increase in wear should not cause a wall thinning
problem over the remaining life of the plant.
It seems that we have been blessed with
fat pipes, with wall thicknesses on the high side,
with the specified manufacturing tolerances. So we
have installed margin in this area.
DR. ROSEN: What are the most sensitive
components to flow assisted corrosion?
MR. BROWNING: It is generally the
chemistry and the temperature.
DR. ROSEN: No, but what components do you
find the biggest changes as a result of EPU?
MR. BROWNING: The feed water piping. We
have a table for that. This is the results on the
representative feed water piping that shows the
increases. So from this we could see where the
changes occurred in the parameters.
DR. POWERS: I think you will probably
have to translate for the committee what DLA-02-E14
is. The Committee reads these things very thoroughly,
but they probably just have a hard time recalling
those particular sections.
MR. BROWNING: These are particular
sections in feed water piping, and this is how we
designate them in our plant drawings so that we know
what we are talking about.
And this is an elbow, and as you can see,
it is in elbows that you would expect to see most of
the wear.
DR. ROSEN: This is main feed water piping
at the elbows, and --
MR. BROWNING: Correct. You can see the
current wear ways to predictions, and you can see the
predictions, and then from there you can see how many
bills of margin we have to the acceptance criteria.
So, they say we are blessed with fat pipe.
DR. WALLIS: How do you measure a minimum
thickness? You measure thickness at various places,
but to get a minimum, you have to measure everybody?
MR. MCGEE: They have several spots when
they pick a section of pipe, and they have a grid work
that they work out for the entire pipe basically for
--
DR. WALLIS: So it covers many, many, many
measurements?
MR. MCGEE: Yes, hundreds of measurements
in a small area.
DR. ROSEN: Now, you have a column called,
"Current Predicted Wear Rate." And looking at the 10
inch elbow, it is 5.9 mils per year. Is that at the
EPU, or is that prior to EPU?
MR. BROWNING: Correct. Here is the EPU
wear rate, a 120 percent prediction. So you can see
like we said about 1-1/2 mils in some of the larger
areas as being --
DR. WALLIS: What was the actual wear
rate from the measured values?
MR. BROWNING: This is the last --
DR. WALLIS: Yes, but what was the wear
rate actually? Was it much less than predicted?
MR. BROWNING: Yes, it should be
significantly less, and not having a prior inspection
before the current one, we don't have that data
available unfortunately.
MR. MCGEE: But the predictions versus the
actuals, they do compare those, and they found a very
good predictability of the wear rate. So it is very
close to the actuals that they find in the model.
DR. ROSEN: If you would look with me at
the 16 inch elbow, what you see is that there are lots
of margin as you point out, but that the rate for wear
for the EPU has gone up substantially from 3.6 to 4.9
mils per year, and that is about a 30 percent increase
in the wear rate.
MR. BROWNING: Correct.
DR. ROSEN: So it is having a rather
substantial effect on the feed water piping at these
locations.
MR. MCGEE: And not necessarily from the
flow. The temperature effect as we have discussed
previously at the thermal-hydraulics subcommittee,
depending on where in the pipe the temperature that
you get the best or the highest wear rate at, the
temperature may have moved up in the pipe or occurred
earlier in the pipe now.
So this particular pipe might have seen a
cooler temperature previously, but now because of a
power uprate and the feed water heating effect that
you get an increase in the temperature, and thus an
escalated wear rate associated with that temperature.
DR. WALLIS: Maybe I could point out to my
colleagues that the NRC is using a proprietary
description of this, and that does give the predicted,
versus the observed, in the corrections, et cetera.
So many of these questions are addressed
here, and I don't think it is going to be talked
about, but it is in this document.
DR. SHACK: Of course, they mis-labeled
the figures, whether it is an over prediction or under
prediction. But we can figure that out. Now, how
long is the period between inspections?
MR. MCGEE: Refueling outages
MR. BROWNING: As we try to stretch them
from an 18 to a 24 month cycle, and we are going to be
transitioning to 24 month cycles.
DR. SHACK: So you use up about 10 percent
of your expected margins or something like that?
MR. BROWNING: Yes. And we will now move
on. For the reactor vessel internals, and other
stainless steel components, those inspection programs
conform to the recommendations that a boiling water
reactor vessel internals project or VIP program.
So this program directs the scope and
frequency of the inspections to be performed each
refuel outage, and DAEC is a leader in the industry in
implementing the recommended inspection program. DAEC
is also a leader in the industry reactor water
chemistry, and being the lead plant for both hydrogen
water chemistry with crack or erosion/corrosion
verification, as well as the application of Noble
metals.
DAEC has performed two successful
applications of Noble metals to date. Because of our
good inspection results in the past, and by
maintaining our excellent water chemistry program as
we move into EP operation, and while we do not expect
to see any impact on IGSCC susceptible components due
to the uprate.
And while the core power is increasing
substantially, the increase in fluence on key
components, susceptible embrittlement or irradiation
as to stress, corrosion, or cracking, such as the
vessel walls, core shroud, or top guide, is not
increasing as dramatically.
This again is the influence of the partial
length full rods in the GE14 design. And with less
fission taking place in the upper port of the core,
where the fluence spectrum is hardest, we minimize the
impact of the increase in core average power and flux
profiles on that area of the vessel, and internals
most susceptible to radiation damage.
Thus, we believe that with our aggressive
monitoring programs for both fact and stress corrosion
cracking, we will ensure adequate safety and
operational margins are maintained as we implement the
uprate.
DR. FORD: My original questions at the
subcommittee meeting on this particular item really
related to the fact that the VIP inspection schedules
and deposition curves for degradation, those latter
were based on data produced with very low flow rates
in the laboratory.
And so the question really was how would you
expect those degradation kinetics to increase or
decrease with flow rate?
Now, those are not taken into account in
the VIP documents, and flow rate is a critical change
when you go to a power uprate in some components. So,
that was the question. How would the risk increase
because of specifically flow rate increases?
MR. BROWNING: In most cases -- for
example, in the core region, the flow rate is not
going to change at all.
DR. FORD: Correct.
MR. BROWNING: And so in those areas there
is no change at all. In the down come region, we are
getting a slight increase in the drive flow by about
.3 percent to overcome the delta-P as a result of the
increase in the power in the core.
So we are seeing a slight increase there,
but that has the most impact on the jet pump
assemblies, and not so much on the areas that we are
concerned about for cracking.
Now, those are just long term wear issues
and vibration issues which we did investigate and
didn't see any susceptibility there. And then the
third area of the vessel of course is in the upper
region, where the increase in the steam flow rate is
going up somewhat proportional to the increase in the
power level, an approximate 16 percent increase in the
steam flow rate.
So we do see an impact in the upper
regions, but we have investigated those as well for
the concern of the flow induced vibration issues, and
also the actual operating experience of those
components to the chemistry changes that they are in
so that you don't get quite the same protection from
the hydrogen water chemistry in that region as you do
in the lower part of the vessel.
So we have seen industry experience with
IGSCC on those components, and our inspection program
has factored that in from the GE seal and from the VIP
recommendations to inspect those components carefully
for those issues.
So that is the way that we have tried to
address it on our end, is through the inspection side
of the world to look more carefully, and be more
cognizant as we move up in the higher flow regimes.
DR. SHACK: And the VIP report on hydrogen
water chemistry does address flow effects. Now, you
may not agree with all of what they say, but they have
addressed the issue.
DR. FORD: I was being a devil's advocate
to a certain extent, because many of these aspects of
flow rate are well known in the technical community.
In fact, generally they decrease the cracking
susceptibilities.
And I just want to make sure that it goes
out in the public domain that these things have been
talked about in this committee.
DR. POWERS: I am going to move this along
now.
MR. BROWNING: We are ready to move into
the risk perspective.
DR. POWERS: That would be great.
MR. BROWNING: Very good. I would like to
now move into the insights that we have learned from
using our risk assessment tools. And as we said
earlier, although our application was not risk
informed in accordance with Reg. Guide 1.174., we did
apply our probablistic risk assessment tools to gain
valuable insights into the possible effects of this
extended power uprate on the operation of the Duane
Arnold Energy Center.
Although the results of both the Level 1
analysis of core damage frequency, and the Level 2
analysis of large early release frequency due to the
uprate, were not significant as defined by the EPRI
guidelines for evaluating plant changes. We did gain
valuable insights from this exercise.
DR. POWERS: Did you calculations include
that extended period of time on any effect that might
have the -- well, the extended period of time that you
need the net pressure suction head margin?
MR. BROWNING: I would turn to Brad, our
PRA expert.
MR. HOPKINS: This is Brad Hopkins from
NMC. We do factor in net positive suction head into
the PRA.
DR. POWERS: In the interest of time that
you need that, that enters into the --
MR. HOPKINS: In the Level 2, yes, we are
looking at a 24 hour period after the start of the
event. So, yes, we do factor in the likelihood of
pump failure due to inadequate net positive suction
head.
CHAIRMAN APOSTOLAKIS: What kinds of
uncertainties do you have there? I mean, you say 1.55
and 10 to the minus 5.
DR. POWERS: That is very accurate,
George. That is precise.
CHAIRMAN APOSTOLAKIS: This is 10 to the
minus 5.
MR. HOPKINS: This is Brad Hopkins again.
We have not performed a detailed uncertainty analysis
for the PRA. We addressed uncertainty in our initial
IPE submittal, with a sensitivity analysis, and for
the present power uprate study, we started the study
by selecting sensitive parameters, and sensitive
operator actions, and sensitive components.
CHAIRMAN APOSTOLAKIS: Well, Regulatory
Guide 11.74 requires the use of mean values. And you
don't know that these are mean values do you?
MR. BROWNING: Right. But as we said,
what we are looking for is insights. We are not
looking for a specific precise calculation of what the
actual core damage frequency is. We are just looking
for changes and what drives those changes to look for
the insights.
CHAIRMAN APOSTOLAKIS: And what are you
going to do with the insights?
MR. BROWNING: Well, first, and most
importantly, no new risk vulnerabilities were
identified in initiating event frequencies, component
reliability, or key success criteria. This is because
sufficient plant operating margins will be maintained
during the implementation of the uprate either through
inherent design margin, or modification to equipment,
such that the plant's overall reliability will be the
same as before.
The one area where we did see an impact to
the uprate was in the event timing, and in particular
where operator actions are important to success,
almost all of the change in core damage frequency and
large early release frequency is due to changes in
human error probability.
And especially in those events where there
is heavy reliance on operator actions, such as ATWS.
As you heard earlier in Ron's presentation, we use
this insight to tailor our operator training in both
the classroom and dynamic simulator scenarios to
ensure that the operators would continue to respond
successfully in these events after the uprate --
CHAIRMAN APOSTOLAKIS: I am just curious.
Suppose that PRA never had been abandoned, and you
were following the strict traditional deterministic
licensing approach, would this issue of operator
actions come up here?
Is there a part there that says to
calculate the time for operator action?
MR. BROWNING: It may not have been so
much driven by time, but I think as we would have gone
through the exercise of looking at the deterministic
evaluations, we would have also in parallel been
looking at the impacts in the emergency procedures
that the operators have to follow.
And from that side of the world, we would
have driven probably the similar conclusion that there
are certain key actions in the emergency procedures
that are important for the operators to take in a
timely manner in order to be successful.
So I think we probably would have come to
a very similar conclusion in this case, because ATWS
drives both of those, and what we are seeing here is
a corroboration of that knowledge.
DR. POWERS: I think it is a most obvious
conclusion when you say Dana Powers is going up, and
you say, well, since the plant is kind of fixed,
operator response time is going to be shorter just
because of higher power.
MR. BROWNING: Especially by those things
that are driven by decay heat.
CHAIRMAN APOSTOLAKIS: Okay.
DR. POWERS: And I will just interject
here. George, you didn't get to attend the
subcommittee meeting, but you might want to look at
some of the notes on that, because one of the things
that they did was use the PRA to look at all of their
operator actions and get a raw on them on risk
achievement work.
And then that gave them a prioritization
to go through and look at them, and I thought it was
a fun thing to look at.
MR. BROWNING: Right. And back to what
Brad was talking about with respect to sensitivity
cases as well. We also went back and looked at those
operator actions that were below the raw value of
1.06, and did some sensitivity cases there as well to
make sure that we weren't missing anything by setting
a screening criteria in appropriate levels.
So we look at all actions underneath that
screening criteria, and made some adjustments there in
sensitivity.
CHAIRMAN APOSTOLAKIS: Do we have these
slides?
MR. MCGEE: I can get you a copy, George.
DR. KRESS: In the subcommittee, we asked
what delta-CDF meant and the raw value corresponding
to it, and at one time I believe it was 1 times 10 to
the minus 6.
MR. BROWNING: Right.
DR. KRESS: I thought I remembered that.
DR. ROSEN: Here you have only discussed
a couple of the operator actions that were actually
examined and reported to the subcommittee. I think
there were a half-a-dozen key operator actions that
were examined.
Some of them, you know, have fairly longer
times available than the particular one that we
focused on here, which was the initiation of standby
liquid control within ATWs, which goes from 6 to 4
minutes.
But there are others described there,
took, George, that are not -- well, that one is the
most severe and the largest change.
MR. BROWNING: There were four key actions
that really came to the top of the list, and three of
them were driven by ATWS; and then the last one was
transients where the reactor stays at high pressure,
and the need for the operator to respond to pressurize
the vessel in a timely manner to get low pressure
ejection on it. And those were the key operator
actions that we learned from the PRA.
DR. POWERS: I am going to move you along
on this. We can talk PRA for a long time.
MR. BROWNING: And that is pretty much the
conclusion that we came to, and the insights that we
gained from the PRAs, and with that, that concludes my
portion of the presentation, and I am going to turn it
back over to Ron now.
CHAIRMAN APOSTOLAKIS: I am very curious
about how you quantified this thing. You have a
number of the available time that goes down from 6
minutes to 4 minutes, and the failure probability
increased from .11 to .18.
Now given the state of the art in these
things, this is really noise.
DR. POWERS: Well, I think that the issue
is bigger than that, George. I don't think that the
change from .1 to .18 is what is striking. What is
really striking is that we have a relatively high
probability of human error here, but we have a
database on simulator training where their rate after
50 or 58 tries, I believe, is zero.
CHAIRMAN APOSTOLAKIS: Yes, and that's why
I am curious.
DR. POWERS: And that is what is striking.
CHAIRMAN APOSTOLAKIS: And how old is the
methodology, and how all that stuff comes together.
It must be somebody's judgment at the end.
MR. BROWNING: I think that is a great
deal of it, and it comes down to expert panel type
judgments, where you have a number of people sitting
down and looking at video type of operator
performance, and doing the calculations, and factoring
all those pieces together to arrive at a conclusion.
I don't think it is strictly driven solely
by any one aspect. You are trying to model something
that is very complicated, and so you are trying to use
as much input as you can from diverse sources and
opinions.
CHAIRMAN APOSTOLAKIS: So if you were to
draw on uncertainty as an analysis to this, the .11
could be as high as .6?
DR. WALLIS: No, no, no, George. It is a
change, George, that they are talking about.
CHAIRMAN APOSTOLAKIS: No, it's not. They
are talking about increased from .2.
DR. WALLIS: Yes, but it is only
influenced by one small thing, and the change --
CHAIRMAN APOSTOLAKIS: It was .11.
DR. WALLIS: The change is what they are
talking about. The change is much more precise than
the uncertainties, and these absolute values.
CHAIRMAN APOSTOLAKIS: The change is more
precise?
DR. WALLIS: Yes.
DR. POWERS: I think it is one of the
esoterics of human error reliability that will arise
in our November meeting. I need to move us along
here.
DR. WALLIS: Could you bring up your
number seven very quickly for one minute, the number
seven slide. There is something different about your
slide and mine, and I just think for the record that
the arrow in my slide points to the blue line, and
whatever is going to go into the record should be the
right slide, which may be a copy of this one. Thank
you.
MR. MCGEE: Unless there are any other
questions, and I heard no open items for us, then in
conclusion the DAEC believes that our submittal has
shown that EPU will result in only a minimal increase
in risk, and that a substantial margin of safety will
be preserved.
The conclusion was confirmed by the NRC
staff's confirmatory analysis, and a review of
calculational results, on-site audits of design record
files, and our responses to many requests for
additional information.
As reflected in the draft SER, after
scrutinizing our limited deviations from previously
approved methodology and in general, a healthy
questioning attitude during its review of our
submittal, the NRC has confirmed our view that the
proposed power uprate complies with NRC regulations.
And that there is reasonable assurance
that public health and safety will be protected. We
believe the rigor with which we prepared our submittal
and the thorough NRC staff review have demonstrated
that operation at the uprated power level is
acceptable. We thank you for your time and attention
today.
DR. POWERS: Are there any other questions
for the applicant? If not, Ron, I thank you for your
presentation, and I will call on John Zwolinski to
begin the staff presentation.
MR. ZWOLINSKI: Dr. Powers, thank you for
recognizing me and our staff is prepared to go
forward. Good morning to all of the members. For the
record, my name is John Zwolinski, and I am the
Director of the Division of Licensing Project
Management in the Office of Nuclear Reactor
Regulation.
During last week's ACRS Thermal-Hydraulic
Subcommittee meeting, we made a presentation on our
review of the Duane Arnold extended power uprate.
I am here for two reasons. One is to
represent senior management and the support of our
staff, and secondly, to emphasize that the staff has
conducted a thorough review in all areas potentially
affected by this power uprate submittal.
The staff conducted its review consistent
with existing practices, including the lessons learned
from Maine Yankee. If you recall when I briefed the
subcommittee a few months ago, I went into great
detail on lessons learned from Maine Yankee.
I did not intend to repeat those. I did
want to take just a couple of minutes to reflect on
the process. I also reflect that Dr. Powers did
allude to the staff diving deeply into certain review
areas, and highlighted a couple of those that we
intend to speak to specifically today.
Those are ATWS, fatigue containment, and
things of that nature, but as a general reminder our
staff did undertake this activity using template
reviews. We relied on our standard review plan.
There was an extensive effort that our staff undertook
to ensure that we crossed all the T's and dotted all
the i's.
We got into the field as you have heard
with General Electric and with the licensee. Our
senior staff and management team that is here today
have been deeply involved with this activity.
We have relied heavily on GE topical
reports that have been reviewed and approved, and
presented to this committee in days go by. And
management and the Division of Engineering, through
Jack Throwsnider in the Division of Systems Safety,
and through Gary Hallahan, as well as in projects
myself, have all been deeply involved to ensure that
this project moves on smartly.
DR. WALLIS: This SRP is not specifically
for uprates is it?
MR. ZWOLINSKI: That's correct. It is for
individual sections or topics within various chapters
of the standard review plan. Although we reviewed the
information in many areas and on a licensing basis of
the Duane Arnold plant and beyond by use of risk
information, we will focus our presentation today on
the areas that we believe to be the most interesting
to the power uprate.
And we will also address the areas that
the ACRS expressed interest in. I would like to now
turn to our project manager, Brenda Mozafari. We are
essentially prepared to give about five minutes of
presentation each, and we will try to move quickly,
but obviously we want to take questions as
appropriate.
MS. MOZAFARI: Good morning. My name is
Brenda Mozafari. I am the project manager assigned to
the Duane Arnold at NRR, and I am going to tell you
briefly what our agenda is.
We have received the open remarks from
John Zwolinski, and I am going to provide an overview
of how the staff proceeded in their reviews.
The NRC staff evaluation is going to be
presented in large part by Ralph Caruso, who is going
to address the reactor fuel performance and in GE
audits that were performed to assist in the review.
Kamal Mandly is going to present some
information on the cumulative usage factors, and we
are going to have Rich Lobel present an evaluation or
the staff's evaluation of the containment response
evaluation by the licensee, and then John Zwolinski
will give some concluding remarks.
Okay. I think it is important to note
that the staff started out with a submittal by the
licensee that was provided to us within the guidelines
of ELTR-1 and ELTR-2 as the framework for the review.
The Monticello -- we previously approved
the Monticello safety evaluation as more or less a
template to kind of guide where the emphasis would be
put in the reviews and the depth of the reviews, and
plant specific design differences were addressed
within our draft of safety evaluation which you have
received.
Several additional submittals of
information were provided following teleconferences to
support NRC staff reviews. Now, in this I would like
to speak a little to this streamlined RAI process that
encouraged questions by our staff.
Although we have many additional
submittals of information along the way that we felt
were necessary to put on the docket, there were a lot
of telephone calls where we got clarifications to make
sure that we truly understood how the licensee
remained within the bounds of the generic topicals and
the analyses, and to make sure that we understood the
manner in which they performed calculations.
We were building our lessons learned from
previous uprates, and the licensee did make a
submittal that was informed by previous RAIs that have
been done on Monticello and Hatch,a nd tried to
address those up front, and provided us with detailed
lists of methodologies.
DR. WALLIS: I guess the streamline
process also allowed enough time to analyze the
answers to the questions, as well as allowing you to
pose questions?
MS. MOZAFARI: Right. And it allowed us
to have a good interchange of information to make sure
that there was true understanding. Statistically, we
ended up with a 120 page application that the licensee
submitted, which resulted in about 3,000 -- it will be
in excess of 3,000 staff review hours.
And we incorporated the staff review hours
and the inputs into approximately a hundred page draft
safety evaluation, which we did provide you.
And it is a work still in progress and we
recognize that it had a few quirks, but the outcomes
or the conclusions would not change.
CHAIRMAN APOSTOLAKIS: I was not present
at the subcommittee meeting, but I am trying very hard
to understand what these statistics are telling me.
More than 3,000 review hours. Okay. So what? I
mean, what does this prove?
MS. MOZAFARI: The staff will go into
that. What I am trying to say is that it wasn't a
cursory review, where we just looked briefly at the
submittal and said, oh, they need the generic
guidance.
We did go into -- and I will outline what
was done to substantiate what the licensee provided.
The review questions came from eight different
branches within the NRC, and there were 26 written
responses, and we had three general meetings with the
entire review staff.
I would like to show you just briefly the
scope of the review. We did review -- and you can
read them -- in all these areas involving these
different branches, and from those branches, we
gleaned the fact that we had three characteristic
types of ways that we would evaluate things.
Site audits were performed by the reactor
systems group, and we reviewed specific calculations.
For example, electrical and stress calculations, and
Kamal Mandly is going to demonstrate the kind of
stress calculations we have looked at, and we did
confirmatory analyses, particularly in the containment
systems area.
And with that, I will turn it over to
Ralph Caruso, who is a section chief on the BWR
systems and nuclear performance section to present the
field specific issues.
MR. CARUSO: Good morning. This is an
introductory, and this is who I am. I don't have a
lot of slides this morning, and I am going to be doing
mostly talking from this one particular slide.
The staff review of the power uprate for
Duane Arnold was quite involved. This was a
significant review because the power level increase
that was requested was above any power level increase
that had been granted before.
Before we started the review, we had lots
of meetings with the Duane Arnold people, and I
believe they even came in and talked to the ACRS
before the application.
And one consistent question that came up
during this time is how can you do this. How can you
raise the power level in this reactor by 20 percent
and not change the pressure. Aren't all sorts of
things going to change, and how do you do this.
There was just a lot of incredulity
expressed by a lot of knowledgeable engineers about
how you could do this. Well, we talked to GE about
this, and it turns out that the way they actually did
it when it comes down to looking at the course is
pretty simple.
They flattened the flux shape in the core.
They got the peak bundle constant and they raised the
average bundle power, and that is a pretty simple
thing to do.
I mean, it requires some sophisticated
engineering by GE to do core design, but when you look
at it, fundamentally it is pretty simple, and they
were able to accomplish this within the existing
analysis methods that have been used over the years to
evaluate the performance of BWRs.
Now, that is what I am talking about when
I say in this first bullet that approved methodologies
were used for the safety analyses. BWRs are nice
machines. I wasn't originally trained on them, but I
have grown to appreciate them over the years.
They are basically channel reactors, and
they scale very well, and the analysis methods are
pretty simple because they just boil water. They don't
have these strange steam generators and funny loops
and stuff.
So if you can figure out a way to define
the flux shape and stay within the limits for the
individual channels, you can raise the power levels
with a reasonable amount of effort, and the staff
review of this reactor looked at these methodologies
to make sure that they were being done in accordance
with those methodologies.
DR. KRESS: And you get an increased steam
flow.
MR. CARUSO: That's correct.
DR. KRESS: And how do you maintain your
pressure constant with that?
MR. CARUSO: How do you maintain --
DR. KRESS: The head pressure being
constant.
MR. CARUSO: Well, the reactor pressure
remains the same. The pressure in the reactor vessel
remains the same, and you open the valve at the other
end a little bit more, and you redesign that valve,
and you redesign the turbine so that you can get more
steam flow.
DR. KRESS: So you redesign the turbine
and the control valves?
MR. CARUSO: Yes, but you maintain the
steam pressure in the reactor constant, and that makes
a lot of other things a lot simpler.
DR. KRESS: So it is not just simply a
flux flattening.
MR. CARUSO: I understand that, but I
think -- I started this because that was one of the
big questions that came up in the beginning was how
can you do this. People were just incredulous that
you could do this. And I have got to give GE credit
for this. They were very clever.
DR. WALLIS: So how do you know the flux
is flat when you are doing all this rod stuff and the
operation. Do they measure the flux profile in some
way?
MR. CARUSO: There are surveillance
requirements for operators to periodically measure
flux distributions, and to ensure that the reactor is
operating within the analyzed limits. Those are
important not just for safety purposes, but for
economic purposes.
DR. WALLIS: So they actually measure the
flux distribution?
MR. CARUSO: Yes.
DR. WALLIS: Thank you.
DR. KRESS: You would have a real problem
getting another 20 percent increase. You have already
flattened the flux about as much as you can do it.
MR. CARUSO: I don't know. I don't want
to speculate on that. I mean, I don't want to
foreclose the skills of GE's engineers.
DR. POWERS: Well, I don't think we should
take the view as this being a very simple change
because of some generic feature of the BWR. The fact
of the matter is that actually the boiling water
reactor is fairly complicated.
MR. CARUSO: I would not disagree with
you.
DR. POWERS: You have the void and
reactivity are coupled, and in a fashion that
inherently makes this system non-linear. So
stability, which we can analyze in a linear fashion,
becomes complicated when you are working with a non-
linear system.
MR. CARUSO: I understand that, Dr.
Powers, and I think I will be addressing some of this
in my fourth bullet.
DR. POWERS: Will you be able to tell me
where I can go to look and see if in fact 1-D is
applicable?
MR. CARUSO: Right here. And I have got
someone on the staff here.
DR. POWERS: So I will be able to look at
differential equations and see IKON values?
MR. CARUSO: I am not so sure about that,
but maybe I can point you in a direction.
DR. POWERS: Somebody I would hope would
point me, because I would really like to see the
differential equation analysis with this core, with a
flat and flexible --
CHAIRMAN APOSTOLAKIS: The stability
analysis --
MR. CARUSO: Well, let me think about
that.
DR. POWERS: Quite frankly, the stability
analysis has been done with enormously complicated
equations and are actually fairly simple equations.
MR. CARUSO: So I am starting out with the
point that the licensee used approved methodologies;
the SAFER/GESTR methodology for LOCA analyses, and
TRAC-G for stability, and ODYN I guess is one of the
codes.
There are a number of methodologies that
are used for various different Chapter 15 analyses.
This brings me to my second point, which is the
licensing limits for these analyses were retained.
The 2200 limit that is in the regulation
for LOCAs is still there, and the license showed you
that they continue to meet the 2200 limit. They
continue to meet the 1600 limit, the additional 1600
limit that is in the SAFER/GESTR methodology.
They continue to meet the 99.9 percent of
all rods do not undergo boiling transition, which is
the standard for AOOs. They continue to meet the ATWS
SLC injection requirements that are in the regulation.
They continue to meet the subsidiary
requirements int he ATWS analyses that they don't
exceed the 2200 degrees and the field geometry limit,
and the containment pressure limits. All of those
limits are retained.
DR. WALLIS: And these are all based on
calculations done by the licensee?
MR. CARUSO: Yes, or its contractor,
General Electric. And given the increase in power,
and the amount of the increase in power, the staff
decided that as part of its review to go out and
actually do an audit of these calculations.
We sent a team of four or five people to
GE-Wilmington for a week, and they went through the
detailed design calculations for this plant for a
number of these transients and accidents in order to
satisfy themselves that the methodologies were being
appropriately applied, and that the results were
acceptable.
During the course of that review, that
audit, we found a few problems with some different
aspects of the GE methodology that we described and
discussed with the subcommittee, and I think that we
discussed several months ago.
And that was useful. We found that
useful, and GE has resolved those issues, and we don't
think they stand in the way of this power uprate.
So based on the fact that they are using
the approved methodologies, and that they meet the
licensing limits, and which hasn't changed -- and I
brought that point up because there were some concerns
that we might have changed something like the fuel
burn up limit.
There is no change to the fuel burn up
limit. There were concerns that because it was a
power uprate that the fuel would be burned to a higher
level, and therefore, it would be in some sort of a
weakened state.
And the answer is, no, it's not. The fuel
is being burned within the limits of the methodology.
So those concerns turned out to be not well-founded,
and based on the use of these methodologies, and the
audits that we did, and the reviews that we did in-
house, we believe that this is a fully-justified power
uprate.
Now, the last bullet concerns ATWS issues.
There were several ATWS issues that came up during the
subcommittee meeting. One concerned -- I thought it
was Dr. Powers's radio peaking factor, and whether the
flattening of the flux shape resulted in a core that
was different than what was actually analyzed as part
of the ATWS stability bounding calculations.
And in talking to GE, we discovered that
indeed it is bounded by the analyses that were done
back then, and that flattening the flux shape does not
invalidate those analyses. I think today that you had
a question about an axial power shape.
DR. POWERS: Axial power shape comes up in
the ATWS recovery.
MR. CARUSO: Okay.
DR. POWERS: The first one that you are
raising I guess -- well, that was one of the
questions.
MR. CARUSO: Well, I can show you a power
-- well, I don't want to get too complex, but there is
a concern about operating at the MELLLA point, and
comparing the new operation at this MELLLA point here
to the original license value, and whether the core
state would be somehow different, and therefore
whether the bounding analyses didn't apply.
And GE talked about this at the
subcommittee meeting; that in any case, if you have an
ATWS and you don't do any mitigation, in both cases
the reactor ends up at the same stay point, which is
inside the instability region.
So the initial conditions are essentially
the same for instability. And therefore this
operation at this higher value is bounded by the
original bounding analysis.
And in actual point of fact, when you look
at ATWS and you look at the mitigation that occurs,
the plants don't go to that particular state point,
they actually drop down further on a natural
circulation line.
So we would expect that they would not end
up inside that instability region. The analyses
aren't --
DR. WALLIS: There is extra steam in the
lower level that reduces the power.
MR. CARUSO: That's correct. That's
correct. When you reduce the water level, you will be
reducing the reactor power. Now, you had a concern
this morning --
DR. WALLIS: That is below the stability
line?
MR. CARUSO: I don't show the stability
region here, but just looking at it and eyeballing it,
I believe it is below the instability region. Now,
you had a concern this morning about I believe decay
heat in the upper part of the bundle if you had
started with a core that had a top skewed power shape.
DR. POWERS: The question really is does
the -- well, when we looked at the ATWS recovery, part
of that involves bringing the water level so that the
collapsed level is below the top of the fuel.
MR. CARUSO: Right.
DR. POWERS: The staff resisted that
substantially when we discussed the recovery sequence,
or when it was first being considered. We have
actually agreed to that because there was adequate
steam cooling of the top of the core.
That was with an unflattened power profile
in the core. The question is that now that we have
flattened it, we have raised the amount of heat that
is available in that upper part of the core.
The staff is still comfortable with
dropping the collapsed water level below the top of
the core, and why?
MR. CARUSO: Okay. I have one of my staff
members here, Tony Ulyses or one of my former staff
members. He now works for the Office of Research
here, and he can talk about this a little bit.
MR. ULYSES: Yes, my name is Tony Ulyses,
and essentially, Dr. Powers, the question boils down
to the fact that when the water level was originally
developed -- and again this is referred to as the
minimum steam cooling reactor water level.
It was derived with the assumption of a
top feed axial power distribution, which at the time
was obviously not realistic, but it was done in order
to bound any future operations.
And that power distribution will continue
to bound the current operational strategies that are
being used right now and at the power uprate
conditions.
DR. POWERS: And where do I find in the
safety examination that the staff has done the
discussion that the staff looked at the flattened
power flow profile, and found that indeed it was
bounded by the original analyses and that the ATWS
recovery sequence was still applicable?
MR. ULYSES: In the SER that you are
looking at right now, the draft SER?
DR. POWERS: Yes.
MR. ULYSES: I would say probably nowhere,
having not actually written it myself.
MR. CARUSO: Off the top of my head, I
don't know that we explicitly addressed that
particular issue.
MS. ABDULLAHI: If I may interject. I
actually did the review for --
MR. CARUSO: Introduce yourself.
MS. ABDULLAHI: I am Zena Abdullahi,
Reactor Systems Branch. Dr. Powers, the submittal
itself did not actually consider or address
instability, because instability is considered by the
industry at large as a closed issue.
However, due to the concerns of the ACRS,
I did go through and looked at these topical reports,
and see if it is bounded. And I think that Mr. Caruso
has sections of it in which it talks about the power
shape in the topical reports itself.
What the draft SER basically says is that
we have not received any analysis, for specific
analyses, for instability, for ATWS instability, for
Duane Arnold. That's number one.
And, number two, we expect as an Option 1-
D plan if in fact you have the transients that create
the potential for instability, and that they would end
up being or have the potential for core-wide
instability.
Having core-wide instability, they would
take the mitigating factor, the mitigating steps, for
SLC introducing water level, and the analysis that was
done, could it be bounded for Duane Arnold.
And operators would also get -- in other cases, the
alarms would let the operators know that they are
experiencing -- the plant is experiencing instability.
DR. POWERS: I bet you when they go into
instability that there is a whole lot of information
going to the operators telling them that things are
not quite right. The question is whether the recovery
process still going to work.
MS. ABDULLAHI: There are a couple of
things that I could point out basically from the top
of my head right now without getting to you with the
details.
DR. POWERS: Well, what was in the part of
the examination.
MR. CARUSO: I am looking at the draft SER
here on page 60, and it is the section that talks
about aspects of ATWS instability and the EPU reports.
It talks about operator actions to mitigate ATWS
instability.
And the staff made a conclusion that the
mitigating actions that the plant will take with
regard to ATWS instability are exceptional. We make
that statement there. Is it your concern that we did
not specifically say --
DR. POWERS: Well, when I read statements
like that, I am not sure what you did. So, I am just
asking. It just seems to me that when somebody comes
into me and says I am going to do something to a BWR,
almost the first question that comes to mind is ATWS,
and the second question is ATWS recovery.
It seems like maybe it peculiar to me, but
those are the things that I promptly think about, and
I am asking what was done with the recovery sequence.
I think the previous speaker said that she looked at
it, and that is just is not something to worry about,
and I am willing to accept that.
MR. ZWOLINSKI: And it would strike me,
Dr. Powers, that would be the type of issue that we
would consider as follow to put in the final safety
evaluation.
DR. POWERS: There has got to be something
more than we looked at it and everything is okay.
MR. ZWOLINSKI: I understand your point.
DR. SHACK: There is another peculiar
sentence in the SER on page 61 that says that the
staff realizes that the EPU safety analysis did not
include a review of the applicability of the generic
instability analysis specified in EPU operation
involving a high density core MELLLA and GE14 fuel
design.
And my question is why didn't you? I
would have thought that was the number one item to be
addressed.
MR. ULYSES: Actually, Ralph, I can
probably answer that question. Essentially, if you
look at the instability mitigation strategies which
have been approved, they are intended to be reactor
design specific, and at the time of the submittal we
didn't have the actual specific reactor design to go
to the EPU conditions.
But that will be dealt with during the
reload calculational phase. What we did do was we
looked and we considered the applicability of Option
1-D, and we actually looked at the calculations which
were done by GE on our staff audit, and which was done
on what they called a representative core.
In other words, it is not the actual
reactor that they are going to be running, but it was
a representative core to demonstrate that they will
continue to be able to use Option 1-D.
Essentially, they show that there is a
-- that the core-wide instability mode will continue
to dominate. And all these calculations for reactor
instability based on the GE methods are done with
frequency domain methods.
In other words, they are calculating
basically to K-ratios. So they are not looking at
these instabilities in a time delay domain. They are
not using the TRAC code, for example, and those are
not being used in this case.
And those calculations were examined and
they were looked at, and there was a finding basically
that, yes, that basically from what we see right now,
we can conclude that they will continue to be able to
use Option 1-D.
However, it will have to be confirmed on
a cycle-specific basis, and that is the methodology
which is approved now, and that is the methodology
which will continue to be used in the future.
DR. WALLIS: I have a question, and maybe
it is for management. When we read these SERs, we
read the issues and then we reach the bottom line,
which as the staff stated is that this was acceptable
or something like that.
And then we hear in these presentations
that actually there is a technical basis for this.
Isn't there a paper trail somewhere where the staff
actually records the technical basis for accepting
these particular calculations by a licensee, whatever
the issue is?
MR. CARUSO: It depends on the issue. I
mean, the staff reviews -- well, we accept things
based on doing our calculations, which leaves a paper
trail, and by looking at the actual calculations that
the licensee does step-by-step.
DR. WALLIS: Well, if we are curious can
we find that paper trail? Can we follow it?
MR. CARUSO: Sometimes, but realize that
a lot of the time that staff decisions are judgment
calls.
DR. WALLIS: Well, that is what bothers
me. I mean, where is the part that was reassuring
when there was a technical analysis before.
MR. CARUSO: But in the end what it
absolutely comes down to is I have knowledgeable
engineers on my staff, and I count on them to have
good judgment. That is what it comes down to.
We don't have black and white criteria for
a lot of these things, because they are judgment. I
mean, I sit here and I listen to all 12 of you, and
you don't have black and white criteria. And you
disagree very frequently about what those criteria
should be. The staff is in the same position.
MR. ULYSES: Let me jump in here and say
that one thing that we can do when the staff reviews,
Dr. Wallis, is that we can look at the application of
an approved methodology, and when we do the review,
we confirm that the methodology will remain
applicable.
But that is something that we can do that
is black and white, and that is one thing that is
done, and that is something that was done in this case
as well.
MR. CARUSO: Okay. One last item that you
have on that was concern of the operator response
times, and Dick Eckenrode from the staff is able to
address that.
MR. ECKENRODE: First of all, are there
still concerns about operator response time?
DR. POWERS: You betcha.
MR. ECKENRODE: I am Richard Eckenrode
from the operator licensing and human performance
staff. As part of the human performance review, we
examined all of the risk important operator actions
identified by Duane Arnold as being affected by the
power uprate.
There were five actions which in time
available the operator complete the action was
reduced. Initiation of standby liquid control in the
ATWS events is the most limiting time dependent
action.
The time available to avoid emergency
depressurization will be reduced from 6 to 4 minutes.
The staff requested Duane Arnold to prove or provide
evidence that the operators could perform this action
successfully.
This is a critical task in the operator
requalification training and testing program. An
examination of the last four years of regular test
results showed that the ATWS scenario was exercised 58
times, with 100 percent success.
Since the ATWS EOP says to initiate SLC
based on suppression pool temperature approaching the
boron initiation temperature, it is not based on time.
CHAIRMAN APOSTOLAKIS: Well, let me
understand this. It was exercised 58 times you said?
MR. ECKENRODE: Yes, over the last four
years.
CHAIRMAN APOSTOLAKIS: And when you say
successfully, you mean it was completed within how
much time?
MR. ECKENRODE: That task -- that was what
I was just about to say now. The task was not timed.
The reason it was not was because the EOP indicates
that you want to do it before the suppression pool
temperature approaches the byte temperature. And so it
is not a timed item. You are watching the suppression
pool temperature increase.
CHAIRMAN APOSTOLAKIS: But that must have
something to do with the 6 and 4 minutes?
MR. ECKENRODE: Correct.
CHAIRMAN APOSTOLAKIS: So is it correct to
assume that it was done successfully within 6 minutes
roughly?
MR. ECKENRODE: Obviously, yes.
CHAIRMAN APOSTOLAKIS: And what does that
tell us about our ability to do it within 4 minutes?
MR. ECKENRODE: I have that, too, and am
coming to it. In fact, that is the next statement
really, is that it was estimated in those runs that
the action takes about 10 to 15 seconds.
DR. BONACA: It is not the action itself
it seems to me. It is deciding to do it.
MR. ECKENRODE: It is the decision, right;
the decision to do it, yes.
DR. BONACA: It took 10 seconds?
MR. ECKENRODE: No, the action takes 10 to
15 seconds. The decision is simply as the two
temperatures approach each other. Again, it is not a
timed decision.
DR. BONACA: So what you are saying is
that they really don't look at their watch.
MR. ECKENRODE: That's correct.
DR. BONACA: It's just that this
temperature is approaching, and then we trust that as
this temperature approaches this particular limit,
even within a minute, they will react appropriately?
MR. ECKENRODE: Correct. Correct.
DR. BONACA: Which is a significant
assumption here is it not?
MR. ECKENRODE: Yes.
DR. BONACA: I mean, you are saying that
this is really the driver. They see the temperature
going up, and whether it goes up in 10 seconds or 3
hours, they are going to do the right thing. And that
is exaggerating a little bit, but not much.
MR. ECKENRODE: At the high power levels,
in which most of these were run, the time that you are
talking about here is what is the 4 minutes.
CHAIRMAN APOSTOLAKIS: But the four
minutes is from the initiation of the ATWS event, and
not from the moment in which the temperature reaches
the point?
MR. ECKENRODE: That's correct.
DR. BONACA: And the question I have is
how long does it take from the beginning of the ATWS
event to the point where the temperature reaches that
point, and the operator takes the action?
MR. ECKENRODE: At the high power level,
it used to be six minutes.
CHAIRMAN APOSTOLAKIS: So, six minutes,
and now it is four.
DR. BONACA: Okay. I understand now. All
right.
CHAIRMAN APOSTOLAKIS: So they have to
decide to do it, and then do it within that period of
time.
DR. BONACA: And then do it within 15
seconds.
DR. ROSEN: They have four minutes to do
it, but once they have decided the question of
actually initiating it, that takes 10 seconds.
DR. POWERS: And as was said that is
hardly the issue. Let me ask you another question.
The licensee came in with their analyses and found
five critical operator actions. When you looked at it
independently did you confirm that there were just
five to look at, that merited being looked at?
MR. ECKENRODE: No, I did not do that.
The risk people did that, I believe.
MR. HARRISON: This is Donnie Harrison,
and I am in the PRA branch. What we did was after the
licensee submitted their information, they identified
five operator actions that had a raw value above 1.06,
which as we talked about before was equivalent to a
CDF impact if you assume that operator action failure
is 10 to the minus 6 core damage frequency increase.
What we did was that we went back to the
licensee and asked them if there were operator actions
that were below that criteria that may if you were to
look at combinations become more important, and the
licensee came back and identified one additional
operator action that was a little bit below the
criteria.
And its impact was identified that if you
assumed it failed as an impact on core damage
frequency of 5 times 2 to the minus 7, they also then
went back into all of their operator actions that were
screened out, and they doubled their operator action
failure rates.
DR. POWERS: What I am asking really is
you found their case so persuasive that you felt there
was no need to independently look at those?
MR. HARRISON: What I would say is that
first of all the PRA part of this is confirmatory to
gain insights. It is not the basis for the decision
making process. That first. Secondly --
DR. POWERS: Well, operator actions is
very much a part of this.
MR. HARRISON: Right, and the values that
they provided of a 20 percent chance of failure the
staff felt was a conservatively high number,
especially given the antidotal events from the
training simulators that they meet their criteria
every time they have done it in the last four years.
DR. POWERS: But from your own reports
there is no correlation between scores on training
exercises and operator errors.
MR. HARRISON: Right, and we are not
basing our answer on that. We are just saying that is
just information.
CHAIRMAN APOSTOLAKIS: Now, it would help
me to understand the situation here to know what is
the error forcing context in the initiation standby
liquid control?
MR. HARRISON: I'm sorry, but I didn't
catch the first part of your question.
CHAIRMAN APOSTOLAKIS: What is the error
forcing context? Have you heard those words before?
MR. HARRISON: I am not a human factors
person, but there is performance shaping factors and
on this one there is --
CHAIRMAN APOSTOLAKIS: So you are using
what, the level of stresses that are going out?
MR. HARRISON: This is not my analysis, and
I would actually turn to Brad Hopkins from the
licensee to address actually how they modeled that.
Again, I believe he referred to these as a group of
different human action models to come up with their
probabilities.
CHAIRMAN APOSTOLAKIS: But you approved
them?
DR. POWERS: Mr. Chairman, I am doing
grievous damage to your schedule.
CHAIRMAN APOSTOLAKIS: Yes, you are.
DR. POWERS: And I know that your generous
nature doesn't extend to me in that direction. I
think I am going to have to move this along here, and
I will turn to the speakers and ask if you can help me
any, because he gets violent with me.
MR. CARUSO: The next speaker is going to
be Kamal Manoly from the civil engineering and
mechanics section of the Division of Engineering.
DR. POWERS: I am I am the one that was
particularly interested in communicative usage
factors, and could I say that I think that this is not
an essential thing for the committee, and we can move
on.
CHAIRMAN APOSTOLAKIS: And that is no
reflection of you, Mr. Manoly.
MR. MANOLY: Thank you.
MR. LOBEL: My name is Richard Lobel, and
I am a reviewer in the plant systems branch in NRR.
And I have been asked to address the audit calculation
we did to look at the GE calculations for the Duane
Arnold Power Uprate.
We decided to do the audit for four
reasons. First, there was a large increase in power,
and we wanted to be sure that we understood the
behavior with that increase in power.
Second, there was a lack of a staff review
on one of the codes that is used by the licensee, and
we wanted to gain some confidence in the GE code.
There was also a desire to better
understand the input assumptions that are used by the
licensee; and finally there was the issue of the
credit that was taken for the containment accident
pressure, and doing MPSH calculations.
Four cases were considered by the
licensee, and we audited -- that is, we did our own
calculations for two of those cases as indicated by
the little hand-pointers.
The name of the GE code I put in
parentheses. The peak pressure calculation is a short
term calculation that is done with the M3CPT code, and
the peak wet well water temperature calculation was
done with a SHEX code.
And that is a long term calculation, and
the difference between short term and long terms is in
the type of assumptions that are used, and the way
that the two are modeled.
Another calculation that is done with SHEX
is the containment conditions for the MPSH pressure,
the MPSH margin calculations; and finally there is a
peak dry well temperature calculation that is done for
EQ purposes, and that is done with SHEX and a spread
sheet.
Okay. The calculations were performed for
the NRC staff by Information Systems Laboratories,
Incorporated, ISL. They did a very good job. ISL
used the NRC contained two code at our request.
ISL was also requested to use some
guidance that was developed by our Office of Research
for the use of contained 2.0 for design basis accident
calculations.
Contained 2.0 is a best estimate code, and
this guidance document was some work that was done for
research at our request to ask what assumptions should
be made to do a conservative calculation.
The guidance is similar to the licensee's
assumptions in many ways. Finally, and this is
important, we used the licensee's mass and energy as
an input. Mass and energy is important, and this is
where the effect of the uprate shows itself in the
containment calculations.
But we were interested in assessing the
containment codes, and in this case we weren't looking
at the behavior of the reactor and the blow down of
the reactor.
So we used the licensee's input for the
mass and energy.
DR. WALLIS: How many nodes are there in
this model for the containment?
MR. LOBEL: I believe -- Ben, do you want
to answer that?
MR. GITNICK: My name is Ben Gitnick, and
I work for Information Systems Laboratories, and I was
the engineer who performed the audit calculation for
the plant systems branch.
The audit calculation model as developed
by CND in this guidance that Rich just mentioned has
a four-node --
DR. WALLIS: Four nodes?
MR. GITNICK: Four nodes, and the reactor
program system is one node, and the dry well is one
node, and the suppression pool is another. In this
case though it would not really make much difference
because we are not looking -- and particularly in the
long term, we are looking more at the mass and energy,
and not so much --
DR. WALLIS: So it is a very simple
calculation, and probably a simple problem.
MR. GITNICK: Well, there is some great
complex mechanics going on in vent clearing, and I was
going to say the fourth node is the vent, and it has
a special model, which is sort of a response to the
vent clearing.
Containment, of course, is very flexible,
and you can set up as many nodes as you like, but we
are not looking at issues of stratification or mixing
as much as vent clearing and energy deposition in the
suppression pool.
DR. WALLIS: And another question is what
is the basis for the bottom line?
MR. LOBEL: I was told that you were all
given a copy of the contractor's report, and --
DR. WALLIS: You are not going to show us
the curves?
MR. LOBEL: I can show you the curves if
you want to see the curves.
DR. WALLIS: Well, they seem to know what
the basis for a good agreement is.
MR. LOBEL: This is a curve of the long
term pressure, and I have another one of the long term
--
DR. WALLIS: So is there some criterion
for how close they have to be?
MR. LOBEL: No, this gets back to what
Ralph Caruso was talking about. There is no criteria
for how close they have to be. Really what we were
looking for was that the curves basically would have
the same trend, and be fairly close int he absolute
values.
DR. WALLIS: And not step outside some
regulatory limit?
MR. LOBEL: Right.
DR. WALLIS: Well, that would be good to
show some limit. A limit is a criterion. A
regulatory limit is presumably a criterion.
MR. LOBEL: Well, the limit in this case
-- and I don't have the conversion off the top of my
head, but it would be 281 degrees fahrenheit.
DR. WALLIS: Which isn't shown here. Are
there some other kinds of units for temperature here?
MR. LOBEL: Right.
DR. POWERS: I have to say that this
report that you provided on this was very helpful to
understand the conclusions that you reached, and it
was the kind of thing that we were looking for in some
of the other areas.
MR. LOBEL: Thank you. I guess I am done.
I was going to say a few more things, but it isn't
necessary. Are there any other questions?
(No audible response.)
MS. MOZAFARI: I would just like to have
a few concluding remarks by John Zwolinski on behalf
of the NRR staff.
MR. ZWOLINSKI: I would like to thank the
committee for the opportunity to present our review of
a first of a kind extended power uprate to you.
The staff has considered this a first of
a kind application because of the magnitude of the
uprate. It is the first application to take a plant
to 20 percent over its original rated thermal power.
I would like to again emphasize that the
NRR staff has taken an extensive review of the Duane
Arnold Power Uprate request. All areas affected by
the power uprate have been reviewed and evaluated.
The staff has critically examined the
methodologies and their application for this power
uprate request. The staff has concluded that all
analytical codes and methodologies used for licensing
analysis are acceptable for this application.
The results of the deterministic analyses
have demonstrated the proposed increase in power level
at Duane Arnold is acceptable, and meet regulatory
requirements.
Based on the review the staff has
concluded that the proposed power uprate can be
approved at this time, and Dr. Wallis, your comments
are ringing in my ears, as well as the Committee's,
and Dr. Power's.
I can assure the Committee that the safety
evaluation will contain the appropriate technical
basis to support each section and thus the approval of
the power uprate. And with that, this concludes our
presentation. I would be happy to take any additional
questions.
DR. POWERS: Are there any other questions
posed to the staff in this area? Seeing none, I will
thank you, Mr. Zwolinski and Ms. Mozafari, and your
speakers.
And I will thank Mr. McGee and his team
for their presentation to the staff, and presentation
of materials to the Committee, and I will turn the
session back to the Chairman.
CHAIRMAN APOSTOLAKIS: Thank you very
much, Dr. Powers. We will recess until 10:55.
(Whereupon, the meeting was recessed at
10:35 a.m., and resumed at 10:55 a.m.)
CHAIRMAN APOSTOLAKIS: We are ready to go
back into session. The next item is the Readiness
Assessment for Future Plant Designs and the Staff
Proposal Regarding Exelon's Regulatory Licensing
Approach for the Pebble Bed Modular Reactor. And Dr.
Kress.
DR. KRESS: Thank you, Dr. Apostolakis.
I think the committee will find this session to be
quite interesting. We ought to view it as a sort of
an early interaction on this issue and something more
like a briefing more than anything else.
I don't think we will be charged with a
letter, but if we have any preliminary responses to
what we hear, then we can make some oral statements,
or we could write a letter if the Committee thought it
was necessary at this time.
But I think it is a little preliminary to
do that. We are going to hear two things. From the
staff, we are going to hear how ready they are to
attack the licensing or certification of these events
plans, and the plans for getting ready.
And from Exelon and also the staff, we are
going to hear about how to -- well, a possible
proposed approach for certification of the Pebble Bed
Modular Reactor itself.
And since most of our regulations are
highly biased by LWRs, the question is how do you
wedge a pebble bed modular reactor into that
structure. So I think that will be a very interesting
subject matter and very enlightening.
So with that, I guess I will turn it over
to whoever this guy is. Who are you and why are you
qualified?
MR. LYONS: In case you forget, my name is
Jim Lyons, and I would like to thank you for the
opportunity to come talk to you this morning.
DR. KRESS: Before we go any further, I
would like to note for the committee that one item on
our handout, Item F, is still very preliminary, and is
for internal committee use only in our handouts. So
treat it in that fashion, please.
MR. LYONS: I would like to talk just a
little bit about -- and I am going to try and go
through this quickly -- our readiness assessment that
we have done to assess the staff's readiness to
license and inspect the new reactor.
And what this paper is trying to do, and
it is up and getting ready to go to the Commission,
and it will probably go to the Commission next week,
and then it will be made public in a week or two after
that, after they have had a chance to look at it.
But back in -- and I will go to my next
slide, but back in February the Commission asked us to
assess the technical licensing and inspection
capabilities that the staff has to perform these new
reactor licensing.
And we were supposed to identify any
enhancements for reviewing or inspecting early site
permits, licensing applications, and the construction
of new plants.
They also asked us to assess 10 CFR Part
50 and 52, both of them, the regulatory
infrastructure, and whether that was sufficient for us
to move forward with these new plants.
As part of that we considered the
certified designs that are already in the rules, and
the certified designs that may come in in the future,
and the pebble bit modular reactor, which we are going
to talk more abut today. And other generation 3 plus
and 4 light water reactors.
We also were asked to provide schedules
and resource estimates to perform these reviews. I am
not going to get into those too much today. Those
will be in the report.
I guess I would like to say at this point
that some of those estimates are pretty high. They
are really based on the information that we had at the
time.
We are starting to get more information,
but as we get more information, we are going to be
updating those resource estimates and we are proposing
to the Commission that we come back to them like on a
6 month interval to let them know that as business
decisions are made, and plans are firmed up, what
really we are going to be working on, and what we are
not going to be working on.
To perform these assessment, we formed an
inner office working group with members from the NRR,
Research, NMSS, and the Office of General Counsel, to
look at these things.
And the group has had several licensing
scenarios for future applications based on the
information that we had gotten. We worked to estimate
the duration of the reviews and the resources that
would be necessary.
They used the results of critical skills
and resource survey that we presented to the staff.
We also had the benefit of industry plans and proposed
schedules that we heard in presentations or they sent
to us.
We looked at what it took to do previous
licensing and pre-application reviews, and previous
design certifications. We tried to factor in the
effect of complex technical issues or policy issues
that will be coming up.
And we also looked at previous resource
and schedule evaluations that had been done in the
late '80s and early '90s when we were in the midst of
doing the design certifications before. We had tried
to do some projections, and so that kind of made the
basis for where we were.
One of the things that we wanted to do was
identify the capabilities of the staff to perform
these reviews, and we identified skill gaps, or what
we are calling skill gaps, and what we are calling
skill gaps are areas where individuals with the
expertise we needed are either limited in number,
working on some other important issue that the agency
has in-house, or maybe not even in the office where
the gap exists.
We did not do a whole lot of looking
across offices. We did look at these mostly within
the offices that -- well, each office looked
individually.
We tried to identify people that were at
or near retirement and that we might expect to see
leave the agency within the next 6 to 12 months. And
that expertise that just doesn't exist in the staff at
all.
There were still gaps that we identified,
and kind of a big picture or manner identified on this
next slide. We see gaps in nearly all areas of the
site environmental reviews, and that is something --
DR. KRESS: A lot of those in the past
have been done by subcontractors.
MR. LYONS: Right.
DR. KRESS: And can you still call upon
those people?
MR. LYONS: We still can, and in fact we
have these gaps within the offices if you would, and
within NRR, but in the license renewal arena, we have
been doing environment reviews.
We have contracts set up, and I think we
have somewhere on the order of 140 contractors that we
have identified that could help us in this area. So
we think in the short term that some of these gaps
could be handled by contracting out to the National
Labs or other suppliers of environmental reviewers.
Let me go through some of the other
things. Historic and archeological resources that are
part of this site environmental review, and that is
something that we don't have here.
Financial analysts, and especially in the
anti-trust area. We have financial analysts that are
looking at license transfers, and we have some strong
people there, and we are starting to look at anti-
trust reviews, and that would be another area that
would be hard.
CHAIRMAN APOSTOLAKIS: Did the agency ever
have experts on the historical and archeological
resources?
MR. LYONS: I think they actually had
people that did those reviews back in the '70s, and
when we were doing that. We had people that would go
and look at historical records and look at
archeological records, and we had those people in-
house.
Obviously, we have not had the need for
that, and in some of the cases -- and the reasons that
we don't have these people available anymore is
because there really wasn't any work for them. So
they left and took on other jobs, and we never
replaced them.
And there are some fairly senior people
around that have been involved in some of these areas
that can help us, but they are not at the reviewer
level that we are going to need to do some of these
reviews.
And in environmental reviews, a lot of the
environmental reviewers and sitemologists, and
geologists, hydrologists, that were in NRR have now
shifted over to NMSS, and are working on Yucca
Mountain and that type of work. So there might be
some people to draw on.
DR. KRESS: I guess the natural question
is given that you have identified these areas where
there are gaps in skills, you are going to make plans
on how you are going to deal with those gaps?
MR. LYONS: That is one of the things that
we are working with the Office of Human Resources on,
on how we are going to bring people in, and what areas
should we be targeting to bring new people in.
And then again like I said, in the short
term, we can fill some of these gaps with contractor
resources. Obviously, that group is mostly for the
early site permits, which we would expect to be the
first areas that we would see.
DR. ROSEN: Jim, one of the critical areas
on the pebble bed will be fuel performance, and I
expected to see that as a gap here.
MR. LYONS: I think that is really the
next bullet, which is high temperature gas reactor and
graphite technology that the fuel aspects are going to
be very key to this. And in fact how you spend your
time looking at the various aspects of the plant.
DR. ROSEN: But my point is that there is
no bullet on the slide that --
DR. KRESS: Fuel performance.
MR. LYONS: Well, that's really what I
meant by the high temperature gas reactor and the
graphite technology. We also need help in metallurgy
and chemical engineering, and high temperature gas
reactor accident analysis.
And then when you look at inspections,
construction inspectors, and all the people that are
doing the construction inspections have now moved on
to doing other types of inspections.
We are especially light in the
geotechnical areas, and the same with the
environmental reviews. Also, as part of our
discussions with the regions, you might have some
fairly senior people that had construction inspection
experience, but now are in the region or somewhere
else, and it may be a little bit difficult to get
people to go to a site to start doing these
construction inspections.
The other thing that we identified was
that with construction that there may be large
portions of a plant that are built away from the site.
So having a strong site presence, you will
need that in some instances, but you also will need to
be sending inspectors to the fabrication facilities to
do some of the inspections there.
DR. KRESS: That is something that you
haven't done much of have you?
MR. LYONS: Yes, which we have not done
much of in the past, and we did have an inspection
program that did look at that, but again that has kind
of gone by the wayside. So that is an area that we
are looking at.
Moving on, the overall conclusions of the
report, and I will run through those, is that first of
all the licensing processes that are in 10 CFR Part
52, are ready to be used. That doesn't mean that they
couldn't be better, but right now we can license a
pebble bed modular reactor, and we can license any
other combined license that comes in.
We can use the procedures and the
processes that we have now, and I guess that is our
first point that I want to make. And with resources
that we have got lined up, we can see that we can
complete the current new reactor license activities,
which include the AP1000 pre-application review, which
I know that the Committee is seeing some work on.
And the PBMR pre-application review that
we are working on, and the rule makings that were
going on in 10 CFR Parts 51 and 52.
DR. WALLIS: You said that 10 CFR 52 is
ready to be used. What does that say about something
like containment for PBMR?
MR. LYONS: Well, right now there are in
the regulations and the general design criteria, and
they have criteria for containment. And as part of
this licensing approach, the discussion that we will
have next, we will talk about how Exelon is going to
take the group regulations that are in Part 50, and
how they are going to identify which of those
regulations they will meet, and which ones they will
meet in-part, and which ones are not applicable.
And that's really the whole crux of the
next presentation. So let me defer that to that. The
other thing on this Part 52 rule making that we have
ongoing right now -- in fact, we just put up on our
website last week draft rule making language to get
some early interaction with stakeholders.
We are working on preparing a proposed
rule making by the end of March of next year, and
those changes are really meant to address lessons
learned if you would from the design certifications
that we have done previously to clarify the
regulations and to make them easier to use.
And we have also had some initial
discussions with the industry on how we are going to
implement the early site permit, and the combined
license reviews.
So they are more trying to make the rule
a little more efficient and easier to use. So that is
what we are working on there. We also identified
research and additional infrastructure changes that
will make the reviews for early site permits, and
license applications, more effective and efficient,
and to in the long run maybe reduce some unnecessary
regulatory burden in that area.
For one thing, we were looking at is there
a another way or another regulatory framework that we
could have put in place that would be more risk
informed, and that would be technology neutral if you
will.
So those are all things that we are
looking at in the long term to try and do. We also
need to reactivate our construction inspection
program, especially the 2511, and I don't want to get
into a bunch of numbers.
But the 2511 portion of the construction
inspection program is the pre-construction
inspections. So those would be the ones that are
actually applicable to an early site permit review,
and the inspections that we would have to do there.
That manual chapter is no longer active
because we weren't using it. So we have to look at
how we are going to do the construction inspection
program to verify the inspections test analysis, and
acceptance criteria, the ITAAC that are put in place
as part of the combined license.
So there is a lot of work there just to
kind of construction inspection program back up to
use. As you know, and as you may know, I guess, the
House and the Senate have appropriated $10 million to
the NRC for next year or I guess for this year. It is
this year now.
And that is to work for future licensing,
but those bills have not gone to conference, and they
have not come out, and so we really don't know whether
we are going to get all $10 million of that, or we
will get any of it, or more, or what.
But right now it looks like they are in
pretty much agreement that we would get $10 million.
Certainly the events of September 11 may affect that.
So all the work that we have done on this
readiness assessment was really pre-September 11th,
and so we don't really -- other than to recognize that
it may change, you know, it is really not factored
into our readiness assessment.
But even if we get the full $10 million,
there is more work identified than there would be
resources to do. So obviously we would have to
prioritize what we are going to do and which
activities need to be worked on.
A lot of that has to do with how the
industry falls out in making business decisions that
are supposed to be made in the next 3 to 6 months. So
we will start to see really what applications we are
going to get, and what applications we are not.
So that is why we are saying that our
priorities are still evolving, and so we have kind of
given a first cut at this, and the report looks on a
project basis.
It gives resource estimates and duration
estimates on a project basis, and it doesn't break it
down on a year by year basis, because again we are not
sure exactly when a lot of these things are going to
start.
DR. KRESS: Would that $10 million,
assuming that you get it, go mostly to acquiring new
staff?
MR. LYONS: That would be another thing.
Some of it would be to bring in new staff, but I would
say that it is about -- I would say about a fourth of
that money would be bringing in new staff and about
three-fourths would be using contractors assistance
for the first year.
The budget looks as it goes out that the
Commission is looking at more money in the out years
and we are trying to prioritize that work, too. Just
to kind of give you -- this is something that is in
the report, but to kind of wrap this thing up, is that
future Commission correspondence, and the Commission
papers that we are working on, we are working on two
that are in due in November, legal and financial
issues that are based on a series of white papers that
Exelon has provided us.
We don't plan on coming to the Committee
with that paper. It gets into things like anti-trust
reviews, and the number of licenses, and what should
the annual fees be, and I don't think that you all are
really that interested in that.
You might be interested in seeing the
paper, and reading it, but I don't think we need to
give a presentation on that.
DR. KRESS: I think we would be very
interested in the second one though.
MR. LYONS: In the second one though,
Exelon's licensing approach, which is what we are
really talking about today, that paper will give you
an idea of where Exelon is coming, and where we are
coming from, for that paper.
DR. KRESS: Will we have that one on our
November agenda?
MR. LYONS: I think so.
MR. ZEFTAWY: We can have it for the
November agenda, provided that we can get it in a
couple of weeks before the November meetings.
MR. LYONS: I think we will need to work
with you on the timing of when we will have that
paper.
CHAIRMAN APOSTOLAKIS: Yes, we need it two
weeks before the meeting.
MR. LYONS: Yes, I understand that. I
remember.
CHAIRMAN APOSTOLAKIS: But now you are on
the other side.
MR. LYONS: We will give it to you the day
of the meeting. No, I understand. The other things
that we are working on now is AP1000, Phase 2 review,
which I think is already on the schedule.
And the proposed revision to Part 52 is
now looking for late March or early April; and we have
some work on alternative regulatory frameworks that
NEI is talking about presenting a paper.
Again, resources may cause us to not get
too far into that, and that may or may not happen in
June. But definitely I think the last two would be
ones that we would definitely come to you when we
start talking about technical issues and policy issues
for PBMR.
So we will get into things like
containment and some of the other technical issues on
how the fuels will be verified and fabricated, and
that sort of thing.
So those are other things that will be
coming to you. But that is mainly all I wanted to
talk to you about. I would like to turn it back over
to Exelon.
DR. ROSEN: Not so fast.
MR. LYONS: Okay. Sure.
DR. ROSEN: Could you go back to the first
slide. I think I have a simple slide or a simple
answer. What do you think the step requirements memo
is talking about in the very last bullet on that
slide?
And did your remarks cover that? Is that
the AP1000 that you are talking about there on that
slide?
MR. LYONS: To consider the certified
designs and other generation --
DR. ROSEN: The last bullet there, that
says other generation --
MR. LYONS: Yes.
DR. ROSEN: Do you see anything else
coming down the pike in the near term other than PBMR?
MR. LYONS: Yes. We are already in
discussions with Genotomics on the GTMHR, the gas
turbine modular -- well, I am trying to learn these as
they changed their name. Actually, it is from the
MHGTR that they had before.
So, the GTMHR, and they are looking at
coming in, and also Westinghouse is looking to come in
with the IRRAS plant. So that might be another one,
and there may be other vendors out there that would
seek for us to review --
DR. ROSEN: Well, do you want to
reconsider any of your questions about capability?
Imagine if you got all of that.
MR. LYONS: Well, that is one of the other
messages that I am trying to get, is if industry -- if
all the industry proposals, and plans come to
fruition, we are going to be very tight on resources.
And we are going to as an agency are going
to have to work with industry to try and prioritize
which reviews will go forward and which won't.
And obviously the first priority would
always go to someone who came in with a combined
license application, because that pushes you forward
to actually license a plant.
I think early site permits would also be
high on our list to review, but that is one of the
things that we are really looking at, is that it is
going to be tough if everybody comes in.
That is one of the messages that we were
trying to send to the Commission. Any other
questions? I will turn it over to Kevin Borton and
Exelon.
MR. MUNTZ: Good morning. I would to
thank the ACRS for this opportunity to discuss our
proposed licensing approach for the PBMR in the United
States. I would like to introduce our team. I am Jim
Muntz, vice present for Exelon for PBMR North America
Activities.
And next to me is Rod Krich, who is our
vice president for PBMR licensing; and then Mr. Kevin
Borton, who is our manager of licensing for PBMR. We
have Mr. Greg Krueger, from Exelon, who is our PRA
expert.
And Mr. Fred Silady from Technology
Insights, a consultant to Exelon. And Mr. Steve
Frantz, a consultant to Exelon from MLB. At this
point, I would like to turn it over to Kevin to start
through the presentation.
MR. BORTON: Good morning. As Jim said,
I am Kevin Borton, and I am the licensing manager and
in today's presentation we would like to present to
you the proposed methods and processes that we
presented to the staff that we feel could use to
assess the PBMR design in order to obtain an NRC
license.
We will also demonstrate these methods
with some examples from an earlier advanced gas
coolant reactor. We will describe our process which
we will use to judge the applicability of the current
regulations, and we will compare the approach with
some current regulatory practices.
Just an outline of this morning's
presentation. We will give you a brief summary of our
licensing strategy and what built up to our strategy.
We will go into some detail about the
elements of our licensing approach, and we will do a
high level summary of some NRC policies and practices
that we would like to compare this to.
Also, we will talk about some outcomes or
desires that we have for the pre-application
activities with the NRC, and with this last bullet in
mind, you just heard that the staff has identified
some technical issues that they would like to resolve
with us during pre-application space.
Our intention is to work with the staff
during pre-application to bring some of these
technical issues to resolution. However, what we are
presenting this morning does form the basis on which
we could work with the staff to resolve those
technical issues.
CHAIRMAN APOSTOLAKIS: You will come back
to the Regulatory Guide 1.174??
MR. BORTON: Yes.
CHAIRMAN APOSTOLAKIS: Okay.
MR. BORTON: This is just the outline of
what we have for today. Now, briefly our strategy has
been to work within the Part 52 process in order to
obtain a combined license.
We are building upon a design and review
of a South African demonstration plant. We are not
looking or seeking for any new rule making. We are
going to use the existing regulations.
Part of our strategy was to ensure that
there is a definable and stable licensing approach,
and this is the basis of today's presentation. Part
of the strategy was to develop a licensing basis
acceptance criteria, and will be building upon the
MHTGR work done in the 1980s by the DOE and reviewed
by the NRC.
Now, it will take a little time here, but
this is the basis or the biggest part of our
presentation today, is the elements of our licensing
approach.
The first element is the top level
regulatory criteria, and that establishes what must be
achieved in order to conclude that the public is
adequately protected.
Equally important is the licensing basis
event, which define the situations when the top level
regulatory criteria must be met. We will establish
some design specific regulatory design criteria, and
safety related equipment classification, which will
establish how it will be assured that the top level
regulatory criteria are met.
And then we will identify the conditions
for and special treatment of the equipment to assure
us how well the top level criteria are satisfied.
Now, if you take a look at the first four elements,
that could define the licensing basis.
However, in addition to that, we are going
to use those first four elements to determine the
applicable regulatory requirements which will
establish the scope of our application.
DR. ROSEN: I am going to have to
interrupt and ask a question about the fourth point.
When you talk about special treatment requirements and
defining how well they need to be applied, it seems to
me that is a little bit of a conflict with the prior
slide, where you talked about use of current
regulations.
Will you be seeking exemption from the
special treatment requirements in Part 50 as part of
this?
MR. BORTON: We are going to be addressing
the special treatment in there with our analysis and
I think when we walk through those steps later on that
you will see specifically how we are going to do that.
DR. ROSEN: Are you going to use 50.12 to
try to get an exemption, as has been done in light
water reactors recently?
MR. KRICH: In those cases where we need
to get an exemption from an existing rule, yes, we
would use 50.12.
DR. ROSEN: And that's what I thought you
could do, and I am not characterizing that is a good
idea or a bad idea, but I am just saying that it
doesn't seem consistent, unless you would say that
using 50.12 is using the current regulations.
MR. KRICH: Using 50.12 is using the
current regulation.
DR. ROSEN: Fine.
CHAIRMAN APOSTOLAKIS: On Item 2, I don't
understand the word when; define when the TLRC must be
met. Are there other situations when they are not
met? The word when throws me off.
MR. BORTON: We are going to be looking at
different frequency of events and categorizing those
frequencies, and then how we compare those different
regions against the top level regulatory criteria. So
we are looking at a number of criterion.
MR. KRICH: I think as we go through it we
--
CHAIRMAN APOSTOLAKIS: Will it be
replacing the current design basis events?
MR. KRICH: The licensing basis events
incorporates all the events that covers the
anticipated operational occurrences.
CHAIRMAN APOSTOLAKIS: I guess the word
when is not clear to me.
MR. BORTON: This next slide depicts the
overall approach in a model. As you can see on the
right side, there is a green area, and that is our
elements 2 through 4, our top level regulatory
criteria, aligned with the NRC mission and safety
goals.
Again, those elements were the licensing
basis events, and development of regulatory design
criteria, safety related equipment selection, and
special treatment.
As you can see the block towards the
center in the compare, we break these down into a
functional level, and the resulting criteria then
could be used to compare it against the current
regulations on the blue side there, also at their
functional level or underlying purpose, which goes
back to the question of 50.12.
DR. KRESS: This approach, which I would
call a risk informed approach, is going to rely
heavily on having a good PRA.
MR. BORTON: That's correct. And we are
going to talk about some of the attributes of the PRA.
DR. KRESS: Talk about the attributes of
that, and then how it is going to be reviewed and
qualified as a good PRA.
MR. BORTON: We will get through the steps
on those.
DR. ROSEN: Well, it will be more than
just the PRA, because this is a risk informed approach
and not a risk basis approach.
MR. BORTON: It's both.
DR. ROSEN: So it is going to require the
same kind of expert panel structure and protocols that
have been developed to make applications work on the
light water side as well, I think. Is that your view
of it, Greg?
MR. KRUEGER: Yes.
CHAIRMAN APOSTOLAKIS: Another thing. It
says licensing approach, NRC mission and safety goals.
As far as I know the safety goals are not the
regulation --
DR. KRESS: They kind of show up in
regulations here and there, and like the regulatory
analysis. They show up in 1.174.
CHAIRMAN APOSTOLAKIS: But the Commission
never said that if you meet the goals that you are
okay.
DR. KRESS: No, but this is a different
date and time, and a different reactor, and you have
to start somewhere.
CHAIRMAN APOSTOLAKIS: So it is the blue
boxes then --
DR. KRESS: Yes.
CHAIRMAN APOSTOLAKIS: And right now that
is what it is, right?
MR. BORTON: That's correct. Just the
lower part of that model shows that we will be going
through and looking for the regulations that are
applicable, and partially applicable, and not
applicable, or identify areas where this design
requires attention to that.
We will also be in a better position to
address any policy issues, and being armed with this
information, and being able to put together the scope
of our application.
DR. ROSEN: Does the PBMR specific block
imply that there may be some new requirements that are
not in Part 50 now?
MR. BORTON: Yes, that's correct, design
requirements.
DR. KRESS: Does this have to do with the
graphite, for example?
MR. BORTON: Yes. The first element is
the top level regulatory criterion, and this fits into
the establishment of a reference value to make sure
that there is adequate protection afforded by the
design.
We will look at the criteria coming from
the current regulations and guidance based on these
specific principles. We want to ensure that there are
direct statements of acceptable health.
In other words, ensure that the criteria
is fundamental to protection. We will look at it and
make sure that it is quantifiable, and measured or
calculated, and ensure that we can make an
indisputable conclusion or come up with unambiguous
conclusions.
And finally that it is an independent
reactor type site, and generic in terms, and to ensure
that is purely top level.
DR. KRESS: In terms of this independent
site, it sounds like a touch assignment to me. That
means that you are going to have to bound any
potential sites that you have because these
attributes, these top level criteria, are dependent on
population, and meteorology, and that sort of thing.
So in order to make them what you have to
meet in, say, certification, and site independent, you
will have to have sort of a bounding site or a
bounding meteorology, or something of that nature?
MR. SILADY: My name is Fred Silady. We
are looking at their current regulations and screening
them on these three criteria. When it comes to the
assessment, yes, at a particular site, and then there
are various approaches on whether to do it
individually or bound.
DR. KRESS: I see.
DR. BONACA: If you are working at
developing these criteria that you would propose to
the NRC, I guess since it is going to be an
independent reactor type, are you working with other
designers? Are there some industry standard
committees being formed to look at those issues?
MR. BORTON: No, what we are doing is we
are trying to look at establishing those top level
criteria, and I think the next slide gets a little bit
more into details of why we selected what we selected.
And again it is used as a reference value
to judge -- not to measure, but to judge the
acceptability.
CHAIRMAN APOSTOLAKIS: I see. All right.
MR. BORTON: So looking at those
principles, we did select these limiting top level
regulatory criteria for the PBMR. The first is 10 CFR
50, Appendix I, which is the annualized off-site dose
guidelines.
And 10 CFR 50.34, which are the design
basis accident off-site doses; EPA-400, which are the
protective action guideline doses; and the NRC safety
goal for individual prompt fatality risk.
So again these are the direct measures,
quantitative, independent limiting criteria used to
assess the design against. The second element is to
define what are the licensing basis events. So it is
when these top level criteria must be met.
So we will look at off-normal and accident
events for the PBMR, and Exelon will develop the
licensing basis events through our PRA, in which they
are collectively analyzed for demonstrating
conformance with the safety goals.
Now, thinking back, we have a top level
regulatory criterion, and in order to assess these
events in relation to those criteria, frequency
regions are necessary to frame the criteria in the
context of risk, and this is the three regions that we
alluded to before which have been identified.
The first region is the anticipated
operational occurrences region. This is where events
are expected once or more in a lifetime of the plant.
A plant is defined as having up to 10
reactors. The plant lifetime is assumed to be 40
years, and therefore a lower frequency of 0.025 times
10 to the minus 2 was selected.
Again, Appendix I was selected as the
criteria for this region, and it will be -- the
frequency of events, and the events will be identified
as families of events. They could exceed the Appendix
I criteria if certain equipment or design features had
not been put into the plant.
The consequences will be realistically
analyzed for compliance with Appendix I. The second
region is the design basis event region. These are
events of lower frequency not expected to occur in the
lifetime of a plant.
DR. POWERS: Excuse me, but I am not sure
how you do the frequency analysis, and maybe you can
help me a little bit. You quote .025 per plant, and
that can have 10 reactors at that plant, or is it .025
per reactor?
MR. SILADY: The .025 is per plant year,
and we are just making the point that for the PBMR
that a plant could have up to 10 reactors, but all the
assessments will be done on a per plant year basis.
DR. POWERS: Okay. So that each reactor
would have a frequency of .0025?
MR. SILADY: For independent events.
DR. POWERS: And are all of the events
independent?
MR. SILADY: Not necessarily in a PRA.
Earthquakes.
DR. POWERS: So do you calculate the
common mode --
MR. SILADY: Yes.
DR. POWERS: -- explicitly, or do you use
some sort of a beta factor or something like that?
MR. SILADY: The PRA is being done now,
and I think that they will use the best methods
available, which generally are the beta factor
approach.
DR. POWERS: What do you choose as the
beta factor for a common mode between two individual
reactors?
MR. SILADY: We will probably have more
interactions on how we do the PRA and so on. At this
point, we want to define the criteria, and we know
that we have to do it per plant year, because it is
kind of unique with 10 reactors.
DR. ROSEN: When you chose an example, you
chose seismic as affecting more than one unit at a
site, and that I think is fairly obvious to us. But
are there internal events at a plant with up to 10
reactors that could affect more than one reactor?
And that comes to the question of how much
isolation, how independent the individual units are
intended to be, and maybe that is a design detail and
I am way ahead of where you are.
MR. SILADY: Well, we would want to
consider it, and we are setting it up where we can by
making it per plant year.
DR. ROSEN: I am not sure you answered my
question exactly.
MR. SILADY: The answer is that there are
internal events. Let's say a common mode on a control
room or whatever, you can think of things, and the PRA
needs to do that.
DR. ROSEN: And you want a control room
for all these plants as I understand in your proposed
design?
MR. SILADY: Without getting into the
details of the design and going into that review, the
answer is yes.
DR. ROSEN: Clearly if you have one
control room for 10 units, you have chances of having
interactions.
DR. BONACA: I'm sorry, but I had a
question on the previous page. On page 8, this is a
category for anticipated operational occurrences, and
then the next one you are moving to design basis. Are
you planning to divide this operation on occurrences
also in families of groups?
MR. BORTON: That's correct.
DR. BONACA: So you will have additional
divisions inside?
MR. BORTON: That's correct, and we will
plot that for you in other slides.
DR. BONACA: So it will be equivalent to
the old ANSI standards with the help of a PRA?
MR. BORTON: Yes.
CHAIRMAN APOSTOLAKIS: When you say vents,
you mean initiate a new vent or the whole sequence?
MR. SILADY: It is a full sequence of
events.
CHAIRMAN APOSTOLAKIS: A whole sequence of
events.
MR. BORTON: Yes. So for the DBE, the
slide that is up there now, we looked at a lower bound
frequency of 10 to the minus 4 per plant year. With
events at 10 to the minus 4, you have a less than one
percent chance of it occurring over the lifetime of
the plant.
CHAIRMAN APOSTOLAKIS: So 10 to the minus
4 and the lifetime is 40 years?
MR. BORTON: Yes, 40 years. The criteria
was 5034 and was selected as the top level regulatory
criteria for this region, and it is the family of
events that could exceed that criteria if certain
equipment was not afforded by the design.
There will be mean values and uncertainty
ranges of consequences and are both evaluated to
provide high confidence compliance with this region.
DR. ROSEN: Why do you pick 40 years?
Forty years was an anachronism. Why not 12? Why not
47?
CHAIRMAN APOSTOLAKIS: Why not 60?
DR. ROSEN: Yes, why not 60?
DR. ROSEN: What is this magic 40?
CHAIRMAN APOSTOLAKIS: Sixty is good.
MR. KRICH: We agree that 60 is good, but
right now the way the rules are written and the
requirements of the law are, 40 years is what a
license can be given for.
CHAIRMAN APOSTOLAKIS: I see. Okay.
MR. KRICH: So that was the basis for
selecting 40.
DR. ROSEN: But you are going to use
50.12?
CHAIRMAN APOSTOLAKIS: They are changing
so many things.
MR. KRICH: We weren't anticipating
looking at that and extending the 40 years at this
point in time. We were just looking at the 40 years
at this point in time. But certainly if the NRC is
amenable to asking for a 60 year license, then that is
something that we should look at.
DR. ROSEN: I am not the NRC. I am just
one member of ACRS and I don't know what 40 means. I
never did.
MR. FRANTZ: This is Steve Frantz. The
Atomic Energy Act also specifies a 40 year period, and
that is set by law.
MR. MUNTZ: I think we are just looking to
what we have been working with, and we didn't go much
further than that.
DR. KRESS: You have to change so many
things that you might as well not fight every fight
right now. You can wait 40 years and fight that.
MR. BORTON: The last region here is the
emergency planning basis event region, and these are
events that are not expected to occur in the lifetime
of a fleet of plants. A lower frequency was selected
as 5 times 10 to the minus 7 per plant year.
That is consistent with meeting the prompt
fatality safety goal, and here consequences will be
realistically evaluated against the other criteria,
which is protective action guides and their dose
limits.
Having selected a top level regulatory
criterion and defining the LBE regions, and now we
could plot them, and the first thing you note about
this plot is the Y access is the frequency. The event
sequence mean frequency for a plant year.
The X axis is the consequences, and the
solid line going through the center there, the blue,
is the top level regular criteria. And below and to
the left, which is the acceptable region, and above
and to the right which is unacceptable --
DR. KRESS: I would like to note as an
aside to the Committee that frequently stated comment
that frequency consequences occurs could incorporate
the whole range of regulatory requirements is now
given -- this is a demonstration of that comment that
I have made several times. I just wanted the
committee to be aware of that.
CHAIRMAN APOSTOLAKIS: When you say on the
left event sequence mean frequency, what do you mean?
MR. SILADY: It is the initiating event
and any subsequent failures. It is the entire
initiating event frequency and all the probabilities,
the entire frequency of the end-state expressed on a
mean basis.
CHAIRMAN APOSTOLAKIS: But this curve is
supposed to be a complimentary cumulative curve. In
other words, you shouldn't go with individual
sequences here. You should go with a sum.
In other words, if I look at the curve, at
the dose, and I pick 10 to the minus 1, and I go up,
then all the contributions of 10 to the minus 1 or
less REM should have frequencies smaller than whatever
the number, 2.5 down to the minus 2.
This is an old interpretation of the
Farmer curve. It was misinterpreted at the beginning
that it applied to individual sequences. Now it
applies to the cumulative. Otherwise -- you know the
old trick. You can -- what is a sequence is not well
defined.
MR. SILADY: I agree with you with regards
to the comments on Farmers curve and so on, but what
we were trying to do here was look at each of the top
of the regulatory criteria.
Some of those are expressed in terms of an
individual event, and some of them -- like the safety
goal -- are cued -- and we trying to put everything on
one plot very simply here, and we have had some
difficulties as you can see by the footnotes as to
whether it is at the EAB or EPZ, or LPZ.
And some of these are expressed in whole
body, and some of them are expressed in total
effective dose equivalent. We understand that when
you get down close to the third region there that they
have to be cued for the safety goal.
CHAIRMAN APOSTOLAKIS: The third region?
MR. SILADY: The emergency planning, and
I agree with you and we will cum those. But with
regards to the design basis events, the derivation of
what the design basis events should be, and with
regards to the derivation of what the AOO should be,
we want to just look at it on a per accident family
basis.
CHAIRMAN APOSTOLAKIS: You can do that
separately, but this curve cannot be applied to
individual sequences simply because what is a sequence
is an ill-defined concept.
You can give me a sequence as you know
very well, or I can give you one and you can break it
up into 20 sequences, each one with 1/20th of the
original frequency. Now, the staff will never accept
something like that.
MR. SILADY: Right, and neither would our
peer review.
CHAIRMAN APOSTOLAKIS: And I appreciate
the difficulty you have, but it seems to me that this
frequency -- I mean, whatever else you have to do,
this curve has to be interpreted in a cumulative way.
MR. SILADY: It is cumulative for the same
consequence. It is an accident family. They have to
be summed for accidents, and I agree with that fully.
CHAIRMAN APOSTOLAKIS: Yes, that's what I
am saying. Now, Farmer himself, when he did this in
1967, was not clear. But his argument later was that,
look, whether you sum or not doesn't matter. There
will be 2 or 3 of those that really dominate. And I
think from that point of view that he was right.
But I think since we are starting here
fresh and anew, it seems to me that it would be wise
to do it correctly, and I appreciate the difficulties
that you will have with other sites, but this
particular curve -- and on another point, since again
we are starting fresh.
There have been variations of this that
some people have found convenient, and some people
have not. The variation is in -- I mean, the way that
you have it now, the original curve, you have an
unacceptable region and an acceptable region.
One could imagine that you have a light
blue curve that is below this one, and then you have
three regions; the unacceptable, the acceptable, and
the let's talk about it.
That gives you much more flexibility in my
view when you negotiate with a regulator. And that
curve I don't think you can have from the current
regulations, unless you can look very carefully.
But that probably will help you. Now, on
the other hand, you may not want to introduce too many
new things, but that is just an idea. Now, in
fairness, the Dutch did this for all their hazardous
facilities, and I understand that they are not doing
it anymore. So there must be a reason for that.
So I am giving you both sides of the coin,
but --
MR. SILADY: It is a good suggestion, and
I note that in the U.K. safety assessment that they
have that.
DR. KRESS: And the Swiss have a curve
similar to this which is the cumulative SC curve for
the --
CHAIRMAN APOSTOLAKIS: And they agreed to
that and so it has to be. This is a very good step
forward. I really like this.
DR. WALLIS: I have a comment on this
curve. If this were sort of risk mutual, you might
like it if it was proportionally 1 over X. I mean, if
we have this risk adverse approach, then the curve
would dip down faster on the right-hand side.
This is actually the opposite of risk-
adverse. You are allowing more, a very large risk,
and the very large events. You are high. So it is
not risk-adverse. It is quite the opposite. One
would expect it to be more risk adverse.
MR. SILADY: Can I make a comment on this?
That is a good observation, and one that we have
noticed as well. But all we are doing is taking the
regulations as we found them and plotting them.
CHAIRMAN APOSTOLAKIS: One other point
that will come up in the future I'm sure, is that as
you move to the right, the uncertainties in those
frequencies of course will increase as you are very
well aware.
And I wonder whether we need some guidance
as to how much of that distribution, the vertical
distribution and frequency, can be allowed to be above
the blue curve.
And it seems to me -- well, I don't expect
an answer today, but it seems to me -- unless you have
it. You do have an answer?
MR. BORTON: We are going to walk through
those and show some examples here.
MR. KRICH: We have established examples
for ourselves that we will discuss.
CHAIRMAN APOSTOLAKIS: That's good.
MR. BORTON: I think we covered everything
that was on this curve. The only thing again is that
LPZ and EPZ are assumed to be at the exclusionary
boundary, which gives us a little bit higher degree of
margin with the top level regulatory criteria.
So once we have this, we could plot or
populate the events derived from the PBMR PRA, or test
some deterministically generated events against this
type of plot.
CHAIRMAN APOSTOLAKIS: Now, let me
understand something else and maybe that is a question
for the staff. Are all the regulations embedded in
this? In other words, if I do this, and I show all my
sequences cumulative and so on are below the blue
line, are there any other regulations that I have to
meet?
MR. KING: This is Tom King from the
staff. Yes, the GDCs are not embedded in this. This
is a good approach to lay out the regulations that
have quantitative dose criterion.
But there is a whole other set of
regulations that are basically embedded in the general
design criteria that have to be dealt with as part of
this licensing approach. They will talk about it and
we will talk about it.
MR. KRICH: 10 CFR 20 and all the other
regulations, and we will talk about that later, but
this is really just looking at the off-site dose
affects.
MR. BORTON: But our overall approach does
have to look at the whole Part 50 regulations. So as
you see the fifth element and that we will go to later
on, using these as a top level regulatory criteria.
However, we will still have to go through each one of
the regulations and address each one of them since we
are not asking for --
CHAIRMAN APOSTOLAKIS: But you will
declare some of them as inapplicable?
MR. BORTON: That's correct.
MR. SILADY: That's correct.
MR. BORTON: And we will tell you some of
the criteria for that.
CHAIRMAN APOSTOLAKIS: Okay.
MR. BORTON: This was the PRA that we
talked about and the scoping requirements. There
needs to be a comprehensive treatment of initiating
events, sequences, and end states.
The PRA will include operational
experience from both light water and gas reactors from
here and overseas. We will address all modes of
operation, including shutdown in internal and external
events.
And the design characteristics that
support the use of an integrated event tree structure
from initiating events to end states for accident
family consequences and frequencies, including their
uncertainties.
CHAIRMAN APOSTOLAKIS: I don't understand
this last bullet. What do you mean by support?
MR. SILADY: Well, the integrated event
tree that will be developed will be like a level 1, 2,
and 3 PRA integrated together. It won't be separated
or split due to core damage --
CHAIRMAN APOSTOLAKIS: And this is similar
to what Sandia did in 1150.
MR. SILADY: Which is classes.
DR. ROSEN: And to take advantage of the
inherent features of the PBMR, I looked at some of
your documentation, and you talk about doing a level
three PRA and taking it all the way out.
And that raises the question of needing to
have a specific site or some sort of -- or not having
a site and identifying a specific site, and doing a
very bounding -- or taking a very bounding approach
which might penalize the design for most real sites.
MR. BORTON: That's correct. For the top
level criteria that we are using here today to assess
the design not against a site, but just looking at the
design, we will stop at the level two PRA, and that
will be the criteria which we use to assess against
the top level regulatory criteria.
And the MHTGR, the 1980s and '90s afforded
us some examples now to plot against this curve. As
you can see there the LBEs do populate all three
regions defining the events for those regions.
There are some events that do not result
in off-site releases. They are on the far left.
However, we can eliminate those since they are
corresponding functions of the plant that prevent them
from migrating or exceeding the Appendix I or 10 CFR
50.34 limits.
Again, you can see the uncertainty bands
here and they plan role in the classification of the
vents. DBE-6, the top arrow, and DBE-7, the mean,
fall outside the DBE range. However, their
uncertainty bands either fall in or are in close
proximity to the DBE, and that's why they were
described or classified as DBE events.
CHAIRMAN APOSTOLAKIS: Now, that is
something that bothers me. Pick any one of those --
DBE 11 or somewhere in there -- so that we can all see
it. The way it is presented is contrary or in
conflict with the intent of the criteria, because what
they are doing is they are keeping the sequence fixed,
and then they are saying given these sequence of
events, I am uncertain about their frequency, and I am
uncertain about the consequences resulting.
The intent of the curve though, the blue
curve you showed earlier, is not that. The intent is
that those are independent barriers, from the end of
the figure from the dose, and I go up, and all my
uncertainties are on the frequency.
So you have to take this and this, and
when you sum them up, then you have to do that, which
is done routinely in 113 PRAs by the way, with the
various contributors. Ultimately, your independent
variable is the dose.
MR. SILADY: I can see that there is still
a little bit of communication -- that I didn't
communicate properly. I thought we had it resolved,
but there is another aspect to it here.
The dose here in the DBE region, for
instance, the one by DBE 11, back in the MHTGR days it
was called 10 CFR 100 and now it is 10 CFR 50.34. It
is 25 REM whole body it used to be, and now it is a
total effective dose equivalent. It is for an event,
and it is not a cum for all your design basis.
CHAIRMAN APOSTOLAKIS: The sequence is
constant and then you have uncertainty regarding its
frequency and its consequences. So the blue curve is
different.
MR. SILADY: No, this is the blue curve.
It is just the blue curve from the '80s.
DR. WALLIS: It is the same, except --
CHAIRMAN APOSTOLAKIS: You see, in the
blue code, and we will go back. The way that I would
read this is that I will ask myself on the horizontal
access, what is the frequency of exceeding this number
of RADs, right?
MR. SILADY: We are not using this as a
complimentary cumulative distribution function,
because 10 CFR 100, and then 50.34 weren't set up that
way. It was before PRA.
So for the design basis accidents, the
traditional practice is that you take whatever the DBE
is and you compare it, and you come out with hopefully
a small fraction, different fractions at the
construction permits stage, and then as built, of that
dose.
And then you go to a different accident,
another DBE, and you compare it.
CHAIRMAN APOSTOLAKIS: So what is the
purpose of this curve then?
MR. SILADY: The purpose of the curve is
to help us figure out with PRA insights what the
corresponding DBE should be for this new kind of
reactor. It is not a complimentary cumulative
distribution function up there.
CHAIRMAN APOSTOLAKIS: The concept is that
if you choose those correctly, and decide to withstand
them for the things, and then if you did go into the
cumulative complimentary distribution function, the
anticipation is that you would meet that, and in my
mind that proposition has never been proved.
MR. SILADY: Well, we could plot the acute
and latent fatality safety goals throughout the entire
region, but they aren't nearly as limiting as 10 CDR
50.35 and Appendix I are.
The point being made over here is that
because of the regulations, and not a one over X
situation, or having any risk aversion in it. We are
taking what the current regulations and traditional
practices have been, and trying to find out if we use
our PRA insights what the right DBE should be, because
we have an opportunity here to set them correctly.
DR. KRESS: Essentially, this is a more
definitive way to establish DBEs. I think they used
judgement back in the early days to come up with the
DBEs with this thinking in mind, and without ever
really having to quantify.
MR. SILADY: And that's why we use that
selection criteria for making it quantifiable and
using the PRA quantifiable techniques, we will know.
DR. KRESS: And they may have missed some
of them back then. They may have had some that were
way out of bounds in terms of frequency and some of
them probably should not have even been considered.
But this to me is a more reasonable way of getting
them on the page.
DR. KRESS: If you are going to conform to
the design basis accident or concept, this is a
reasonable way in my mind to choose those things, and
hopefully if you choose them correctly, then you will
come up with meeting this cumulative distribution
curve.
CHAIRMAN APOSTOLAKIS: When you say -- and
let's take an example from a light water reactor.
That would be a small LOCA would it not?
MR. SILADY: Yes.
CHAIRMAN APOSTOLAKIS: And it would go to
the vent for a small LOCA and through all these
sequences, and add them up, right? And what I am
saying is why don't you add them up also across
initiating events and do it right?
MR. SILADY: Now, that is the part that I
agree with you that we are going to do. Let's take a
small example, like a PBMR, or a small primary coolant
leak, where you have forced cooling, and you release
circulating activity. You have to sum up all the ways
that you can get to that consequence phenomena.
CHAIRMAN APOSTOLAKIS: Right.
MR. SILADY: And another one where you
have a larger leak, and let's say a little more, plate
lifted off, and then you don't have forced cooling and
you have release from the core because the fuel is not
perfect, and initially particles are released, and it
comes out over 50 hours instead of immediately.
That is a different kind of phenomena
consequence sequent. You have to sum up all the
initiators for it, and it will have a different DBE
number. That is what is being done here.
DR. KRESS: I think what the problem is
that is having labeled those areas acceptable and
unacceptable. I don't think that is the right
designation for those. Those should be labeled
something else.
CHAIRMAN APOSTOLAKIS: We are talking
about two different curves now. I am confused on how
this one will lead to the other one, because unless
you do it cumulatively --
DR. KRESS: It is cumulative when you
determine the overall risk status.
CHAIRMAN APOSTOLAKIS: And the
acceptability is cumulative.
DR. KRESS: And that's why I say those are
probably misnomers, that unacceptable and acceptable.
MR. SILADY: There are two curves, and
there are acceptable and unacceptable on each curve.
The first curve that we are presenting here are the
regulations that are in the law, written in the law.
So, 10 CFR 50.34, we have got to meet it
on a per accident basis, and later there is another
acceptable and unacceptable, which is your safety
goal, and you cum them with complimentary cumulative
distribution function and assure that you meet that,
too, in all regions.
CHAIRMAN APOSTOLAKIS: And you are not
showing that today?
MR. SILADY: No, we are not showing that
today. This one is to derive the licensing basis.
DR. ROSEN: Let me check my understanding
here. In past licensing activities much of the
discussion revolved around such topics as let me
identify for you a very low probability event, and
having done that the argument becomes, well, that is
beyond a design basis.
To me what you have proposed here will
completely finesse that discussion.
DR. KRESS: That's exactly right.
DR. ROSEN: Because no matter what a
person puts on the table that is a design basis event.
But it may be that your 10 to the minus 8 or 9, or 10,
or 11 is below the X axis.
So you can tolerate any postulation in
terms of something happening within this framework.
It just ends up being of such a low probability that
it doesn't have any impact on the design.
MR. SILADY: I agree, but I just want to
clarify that he design basis region, what you design
for, is in the middle. And that leads to these other
elements that Kevin is going to get into, in terms of
what is safety related, and show that only with the
safety related equipment and so on.
And given that you design for that region,
the design has to be able to meet the safety goals and
the protective action guidelines at some distance in
the region below that.
And even beyond that, ACRS or the staff
may suggest other events that we both can mutually
agree are below even that, and we will have to look at
those, and our best estimate basis shows that the
residual risk is low. So there is some finesse here,
but we are still going to have a design basis region.
DR. ROSEN: Well, clearly, but my point
was that for things below the 10 to the minus 4, we
still have framework for discussion of them.
MR. SILADY: Exactly.
DR. ROSEN: And to come to a scrutable
decision that is joint between the applicant and the
staff, and the ACRS, that we have identified a
sequence that is plausible, albeit very low frequency
or low probability, which we know where to put on this
chart. And we know how to deal with it in the
regulatory aspect.
DR. BONACA: But I wanted to say that
except for the user PRA, you are refining with PRA
what has been done and designed in the past.
MR. SILADY: It is a hybrid, and it is
using the best tools that we have and that we know
today. And there are going to be uncertainties on
them. This is a new reactor, but the uncertainties
have to be treated as discussed.
DR. BONACA: I understand and you have a
much better way to go about identifying those designs,
those sequences, and having a basis for saying this
should be in it and this should not be in it, and
therefore defining what equipment is not going to be
qualified to meet those criteria, and so on and so
forth. So you have a structured approach with the
benefit of a sound PRA.
MR. KRICH: Exactly. Let me stop the
presentation now.
DR. BONACA: We will see, however, how
later on --
DR. KRESS: But the only part is the
regulations only deal with the very right-hand side of
that, in terms of frequency concepts; and the other
part of the blue curve as it has been defined, and I
think it needs defining as part of this exercise.
CHAIRMAN APOSTOLAKIS: Is there a document
that you can give us, or something that is detailed?
MR. BORTON: Details of what this approach
is?
CHAIRMAN APOSTOLAKIS: Yes. Do we have
that?
MR. BORTON: I have a letter that we sent
to the staff on August 31st.
MR. ZEFTAWY: I have the licensing
approach from Exelon, which is in the book, but it
does not describe the details of the special design
basis number 11, and how did he arrive at it.
MR. SILADY: You are correct that it
doesn't go into this example that is shown on this
chart, but it has references back to the publicly
available pre-application submittals, the MHTGR PSID,
that tell you what the events are and how they were
assessed, and it includes the PRA. But this is just
an example so that we could talk about what we intend
to do.
MR. BORTON: Again, the red arrows here
depict that there are required safety functions that
are necessary to keep the events in the acceptable
region to the right here.
And that is due to the design of the
plant. The MHTGR gives us another example here and we
expect the results to be similar to those for the
PBMR.
Down at the bottom here, the function of
the MHTGR that are required to meet the DBE limits.
So again there is radionuclide retention, control heat
generation, control of heat removal, and control of
chemical attack.
It is understanding these functions that
become the first step in determining the third element
of our approach, which is design criteria and
equipment classification.
DR. WALLIS: And chemical attack includes
air and water?
MR. BORTON: Yes, air and water. The
third element is really in two parts here. We talk
about something called regulatory design criteria, and
then we are going to talk about safety classification
of equipment.
And again this is how the top level
regulatory criteria are met. The first part of
element three is the regulatory design criteria. They
are qualitative function statements, developed with
risk insights, of course, because of the events which
were PRA driven for each required safety function.
And now these will supplement the current light water
reactor general design criteria.
DR. ROSEN: And again just to emphasize
that, that is PRA supported, but expert opinion, and
expert panel structured development. It is not just
the PRAs.
MR. BORTON: That's correct.
DR. ROSEN: Because you don't model
everything.
MR. BORTON: The PRA becomes an important
step as I get into looking at the regulations for
their applicability, and how they apply to the design.
And also bringing out -- this RDC though
are really intended to look at things that are not
currently in the regulations that are necessary, as
far as design criteria.
DR. KRESS: If you intend to use
importance measures for this determination of SSCs,
like it has been done in some of the risk-informed
applications, you will have to redefine those in terms
of does or fission product releases or something?
MR. BORTON: Yes. The second part of the
third element is the selection of the safety related
SSCs. These are the equipment relied on to perform
the required safety functions to mitigate or prevent
the DBEs, design basis events.
There is two steps in this selection
process, and one is real obvious, the consequence
mitigation. That assures that the dose criteria are
met.
The other one is for high consequence
preventions, which may or may not apply with doses
greater than the DBE criteria, and where we worry
about the frequency of the event migrating out of the
EP region into the design basis event region.
An example of how the MHTGR selected their
safety related equipment, as you can see here, this is
the function to remove core heat. They looked at four
systems available to remove the core heat.
Some of the systems were available and
some were not. However, the last two, the reactor
cavity cooling system, and the reactor cavity in the
surroundings -- the earth, the building -- were
capable of renewing the core heat.
The RCCS, however, was selected as safety
related based on the licensee's ability to demonstrate
its function over the lifetime of the plant.
Now, this process was performed for all
the required safety functions to mitigate design basis
events, and the results are shown on the next table.
Again, this is the MHTGR example.
And this shows the relationship of the
safety function with the safety related equipment.
DR. WALLIS: I'm surprised that it says
radiate heat from vessel. I would think that
compaction is such a big number and such a small or
low temperature on this thing isn't glowing red, that
radiation would be a small contribution to the heat
loss.
MR. SILADY: It is primarily radiation,
because the core heats up in the middle, and conducts
out to the side wall.
DR. WALLIS: This is called the vessel to
the outside world?
MR. SILADY: Yes.
DR. KRESS: As best as I recall, they
coated the outside of the vessel with --
MR. SILADY: There were discussions on the
MSTGR of increasing the humidicivity, but no decision
-- and it is not likely that that will be needed for
the PBMR for smaller power and power density.
DR. WALLIS: Well, humidicivity is an
awkward variable. All you need is a slight change in
the surface temperature and the humidicivity
is probably different.
MR. SILADY: True. It is going to have an
uncertainty band that you are going to have to look at
a .2 to a .6, or whatever, and show that it is still
acceptable.
CHAIRMAN APOSTOLAKIS: This looks like a
master diagram that you can use to define initiating
events, right?
MR. SILADY: It is a subset of it for that
which is required for the DBEs.
MR. BORTON: So having identified these
SSEs, now we can look at the special treatment to
ensure their performance, which is the fourth element.
The PBMR selection for the safety related equipment
will follow a pretty typical practice.
Again, we are going to look at the DBE
consequences and show that only using the safety
related equipment could mitigate those events. We
will classify the equipment during its design,
fabrication, operation, and maintenance, applying
special treatment to ensure its performance.
In the case of the PBMR, the special
treatment requirements for classified SSEs will be
developed based on the required functions for each
DBE. They have a clear road map now.
In this manner a clear basis will be
established for safety related equipment, selection,
and corresponding quality requirements over the life
of the plant.
DR. BONACA: I guess I don't understand
that very well. You are not bring PRA into this?
MR. BORTON: No, the PRA is in this. The
PRA was used to look at or to define the events. The
special treatment was looking at the functions. So
you have very clear linkages now between the PRA, the
selection of the equipment.
And now we could look at the special
treatment, saying under what conditions. What is the
performance parameters for those pieces of equipment
now for those DPEs.
DR. ROSEN: What are the critical
attributes.
CHAIRMAN APOSTOLAKIS: Well, that would
depend on the redundancy.
MR. SILADY: Oh, yes.
CHAIRMAN APOSTOLAKIS: So if you go
strictly by function, you may lose that benefit.
MR. SILADY: Yes.
CHAIRMAN APOSTOLAKIS: Just saying safety
related equipment for SSEs are the ones that support
essential safety functions is not good enough, because
some of these you have only one, and some of these you
have 10.
So they are not of the same value, and so
I am surprised that you are not saying that there will
be some sort of categorization using both safety
function considerations and PRA worth of some kind.
MR. SILADY: The process uses the PRA in
the front end to help fix the events, which are the
events.
CHAIRMAN APOSTOLAKIS: I understand that.
MR. SILADY: Then if there is any event in
that design basis region that if it were not for
something in the design that it would be unacceptable,
it becomes a design basis event.
And then you rerun all the design basis
events with only the safety related equipment that you
want to rely on, and when you do that, you find out
what the temperatures, pressures, loads, are that that
equipment has to be designed to.
And in that way you define the conditions
that it has to operate under. And then you say what
the performance requirements are, in terms of quality
and so on, in order to make the assurance that it is
going to be able to remove that amount of heat, and
stay within that temperature and so on.
So the PRA was at the front end, but as
soon as you get those events defined, it becomes more
of a deterministic traditional approach.
CHAIRMAN APOSTOLAKIS: But that's what we
learned from that application. You know, that a
diesel generator is an important component, but it has
a few thousand subcomponents.
The question is whether all of these
subcomponents also safety related, regardless of what
they do? The utility was complaining bitterly that
they shouldn't be.
MR. SILADY: It goes function by function,
and design basis by design basis event. And, for
instance, the reactor vessel has three functions that
it has to perform. It has to control chemical attack,
and it has to maintain core geometry, and it has to
radiate the heat away to remove core heat.
It is being made safety related for two of
those events, for two of those functions. For
example, maintain core geometry so you can get the
control rods in, or to remove the core heat.
So once then you look over the spectrum of
events, you get different conditions. Sometimes it is
pressurized, and sometimes it is depressurized.
And sometimes the initiating event was a
leak, and you go through that whole process and you
find what the requirements are on the reactor vessel.
And in a similar way for each of the functions you go
through it.
CHAIRMAN APOSTOLAKIS: And they cannot go
beyond that, because to go beyond that and do what was
done for STP, you need to have the procedures in
place, and you need to have operating experience, and
you need to have other staff to make judgments that
says that this component is not safety important as
the other component.
I mean, you are missing at the design
level from the mental elements for the --
DR. ROSEN: Well, clearly you are missing
the operational experience, but you are not missing
the ability to look at a system and say there is a lot
of redundancy here from a safety function, and taking
credit for that redundancy.
So there is a hint here being offered to you by ACRS.
CHAIRMAN APOSTOLAKIS: It can't be just
the function.
MR. BORTON: What we meant by this third
bullet is that we will have the ability to do that
tracing back down to that level of detail.
CHAIRMAN APOSTOLAKIS: We will have other
--
DR. BONACA: I think that we have to argue
against the design line against GDCs. That is one of
the issues. What I am saying is that I think at the
design stage that it is hard to do --
CHAIRMAN APOSTOLAKIS: Well, that's where
people commit to things that they regret 15 years
later.
MR. BORTON: And that is the point of that
third bullet, is that we have the ability now to
analyze to that level.
CHAIRMAN APOSTOLAKIS: When in doubt, be
conservative.
DR. ROSEN: We are talking about one end
of the spectrum about things that are clearly safety
related and have important functions, and we are
urging you to think about redundancy and taking credit
for it.
On the other end of the spectrum, your
process seems to be very clear, and clearly able to
sort out the things that have no safety functions, and
not spend a lot of money and time on those.
MR. BORTON: Things won't be unambiguous.
DR. ROSEN: Yes.
MR. BORTON: Now, we have covered the
first four --
CHAIRMAN APOSTOLAKIS: Is this clear to
everyone?
DR. ROSEN: What is the question, George?
DR. BONACA: You said that the process is
very clear.
CHAIRMAN APOSTOLAKIS: Well, it is not
very clear. It is just clear. But I am not saying I
am objecting. I just want to read more about it.
MR. BORTON: Now, having covered the first
elements, and this brings us back to our model here,
and just quickly going over it now. We have looked at
the licensing basis criteria, and licensing basis
events, and functions and equipment, and the special
treatment.
The last element is to compare these risk
informed design criteria and functions with the full
scope of regulations in order to define the scope,
which will be able to obtain a license. And again we
believe that this could be done at a functional level.
And, number five, again it is elements 1
through 4, could be used to determine the applicable
regulatory requirements. We will have to establish
the logical rules for their selection, and Exelon
provided an initial screening and results to the NRC
in our letter on August 31st.
We also recognized by going through that
that we are going to have to do a more detailed
screening utilizing the four elements, and this is
what we talked about earlier, about how that supports
that in getting a more finer screen of the
regulations.
We also feel that this will give us the
ability to assess these regulations with consistency
and repeatability, and not subject to arbitrary
judgment. So what we used and essentially what this
slide is trying to say is that we use this systematic
logic diagram.
It's purpose was to determine what
regulations apply, partially apply, or don't apply.
However, it also has steps in it used to assess what
regulations and guidance could be used as guidance.
In other words, we didn't just throw this once we
determined that it is not applicable, we don't throw
it out. We look at it for guidance.
And again the first four elements of our
approach help us determine what guidance is there.
Certainly we also look at what guidance is necessary
that is not currently in the regulations, and we have
a bin for that as well.
The results of our preliminary screening
is that the majority of the regulations do apply,
either as required or required as guidance. And we
plan to once we have some more information from the
design, we will be able to do a finer screening and
share that with the regulators.
DR. POWERS: I guess I am a little
confused by the drawing. You look at regulation and
some of them are directly applicable to PBMR, and you
go down to what is to me the left side, you could use
not directly applicable and you throw away. There is
no route out of there.
MR. KRICH: Are you talking about
partially applies?
DR. POWERS: Well, what I am saying is
that I think it is this guidance business. For some
reason, they applied it to the -- say a PWR.
MR. KRICH: Well, maybe by example I can
maybe help answer the question. If we look at 10 CFR
50.46, which is the fuel requirements, or the
performance requirements for the fuel, that regulation
is written specifically for LWR fuel.
DR. POWERS: Right.
MR. KRICH: So we said, well, that clearly
doesn't apply. However, we are going to need to
develop the same type of performance requirements for
PBMR fuel.
So we said that then needs to go -- that
guidance still needs to go in there. We still need to
have something that applies to PBMR fuel, along the
lines of a 10 CFR 50.46.
DR. POWERS: So there is a third leg on
this someplace?
MR. KRICH: No, all the legs are here. I
guess I am not answering your question.
MR. BORTON: We have seven dunes, really.
The ones that are shaded are the bins.
DR. SHACK: Is there a bin for -- you have
guidance, but is there a bin or a new regulation is
needed, and that's what I don't see.
MR. BORTON: Well, right now we are not
going to ask for new rule making. It will be part of
the design application.
MR. KRICH: So our intent would be that if
in fact there is something that needs -- some
requirement that needs to be applied to the PBMR, we
would include it in our application. The NRC then
would include it in their safety evaluation report
back to us. So it would be imposed via that
mechanism, as opposed to there is a written rule.
DR. ROSEN: You didn't talk about the
deviation part of that block at all and I am
surprised. When you go down to the exemption request,
typically what kind of -- there are criteria in 50.12,
and which of those criteria do you think will be
exercised as part of this?
MR. KRICH: Well, it is hard to say. It
is going to be on a case-by-case basis, but I would
imagine that typically we can meet the requirement via
some other mechanism. That is one of the criteria in
50.12.
MR. FRANTZ: This is Steve Frantz again.
One area where we think we may need an exemption is
from the requirements on operator staffing in and
50.54.
Right now those requirements are general
and are not designed or specific state as applying.
Only LWRs. But in fact when you go back and look at
the basis for that regulation, they were developed for
LWRs.
And they probably are too stringent for
our pebble bed reactor or other kind of passive
reactor. So we are looking at possibly getting an
exemption from 50.54 requirements on operator
staffing.
And you would look at the special
circumstances in 50.12 and show that given a basis for
that rule that it really does not apply to --
MR. KRICH: That you could meet the
underlying requirement without meeting the exact
requirements in paragraph M of 50.54.
DR. ROSEN: Special circumstances apply.
MR. KRICH: Exactly.
MR. BORTON: Okay. Our presentation has
gone through the bulk now, and we are looking at the
next two sessions to have a quick comparison with the
NRC policy and practices, and specifically advanced
reactor policy, and risk informed guidance.
And then finally to cover some of the
objectives for our pre-application. The advanced
reactor policy, we again in our August 31st letter
provided to the staff a detailed comparison of the
policy, and we concluded that the PBMR meets this
policy.
Some of the high level things that popped
out of the policy are the early interactions. Of
course that is what we are doing with the staff right
now in seeking their agreement on this process.
The same level of degree of protection,
and we utilize the current regulations, and we develop
our top level regulatory criteria from the current
regulations.
The proposed specific review criteria or
novel regulatory approaches. I think we meet that
with our design criteria, regulatory design criteria,
and with our risk-informed approach.
And finally providing enhanced margins of
safety and/or utilize innovative means to accomplish
their safety functions. The design of the gas reactor
is noted in the policy statement as being innovative,
using policy -- I'm sorry, passive systems.
And of course our discussion about meeting
PAGs at site boundings will result in enhanced safety
margins.
DR. ROSEN: I would point out that the
safety criteria say at least the same degree of
protection, and it underlined that on your chart, but
you didn't say that in your words.
MR. BORTON: I'm sorry. The next slide is
a comparison with the risk-informed changes and
Guidance Document 1.174. The first thing to note is
that it is applicable to light water reactors and
license amendments.
However, the principles we felt provided
useful guidance. We also sent a letter to the NRC in
detail providing how we meet this Reg Guide as far as
its principles. Some of the things that we
highlighted was defense in depth philosophy that will
be retained.
We look at providing prevention,
termination of events, and mitigation of consequences,
as well as providing physical multiple barriers
through our design. And we do have a balance between
prevention and mitigation.
Some of the other areas that we didn't
touch on earlier is monitoring. We looked to monitor
fuel performance with on-line refueling. The
important systems like the ARCCS system, we will be
looking to monitor that in its performance, and of
course reactor neutronics.
CHAIRMAN APOSTOLAKIS: You said that you
had a balance between prevention and mitigation. Can
you elaborate on that?
MR. BORTON: Yes. I think we have another
slide here. Page 56, towards the end. If you recall
before, we were looking at that chart at the
consequence -- what we talked about before is that we
look at not only the consequences in ensuring that the
safety functions could ensure that the events do not
migrate to the right as you look at this plot.
We also look at the frequency as well. We
talked about unique situations, and if you have a high
consequence in a very low probability area, or low
frequency, we still have to ensure that we can
maintain that frequency through the design in order
for it not to become more frequent and result in
exceeding the DBE region.
MR. SILADY: Basically, when you go to
look at the balance between prevention and mitigation,
you have to look at each situation individually. In
one case a particular SSE can serve a preventative
role, and in another event, it can perform a mitigated
role.
And so you really have to go in to each
particular accident family and say what are the SSEs
that are preventing this event from occurring, and
given that it has occurred, what are the SSEs that are
mitigating the consequences.
And as you go down from top to bottom in
the risk chart, you will see this dual nature; that
one particular SSE will be a preventive measure in one
event, and be a mitigative in another.
And so it is just a question of taking a
very careful look at the high risk events, and seeing
which ones -- what the composite nature is over the
spectrum.
MR. BORTON: The last slide that we have
here is our outcomes for the pre-application
activities, and working with the NRC. We were looking
for agreement on the top level regulatory criteria as
the limiting values.
Agreement on the risk-informed LBE
selection process. Agreement on the process for
equipment classification and the development of RDCs.
Comments and feedback on our approach to
special treatment. Agreement on the process of
determining the PBMR applicable regulations, and the
reasonableness of a preliminary set of regulations.
And finally, comments and feedback that
our approach is consistent with the NRC current policy
and practices, and specifically in these last two
areas.
Now, once this licensing approach is
mutually agreed upon, it will form the basis in which
we can work with the staff to resolve the other
technical issues during the pre-application phase, and
that concludes our presentation for this morning.
MR. KRESS: Thank you very much.
CHAIRMAN APOSTOLAKIS: When were you
looking for or by when would you like these
agreements?
MR. BORTON: We were looking to seek
agreement from the NRC in the SECY that is coming up
in November, around that time frame, so that at the
end of the year we can make our decisions on whether
there is a stable platform to move forward.
MR. MUNTZ: We would expect the pre-
application phase to extend until next September, and
we would still be expecting to have that type of
interface.
DR. KRESS: I think we now turn to see
what the staff's perspective on this approach is.
MR. KING: We are going to have a joint
presentation from NRR research, who are working on
this jointly. Eric Benner from NRR and Prasad
Kadamibi from Research.
DR. KRESS: I propose in the interest of
time that you skip the introductory slides that have
already been covered pretty much, and then go to the
slides --
MR. KING: Go right to the slides that
talk about staff perspective.
DR. KRESS: Yes.
MR. KING: And what you are getting is a
work in progress. You are getting a status report.
We have not finished yet.
DR. KRESS: That's understood.
CHAIRMAN APOSTOLAKIS: And are we going to
have a Commission paper?
DR. KRESS: You are going to have that
Commission paper in when, November?
MR. KING: The Commission paper is due at
the end of November, and we would like you to look at
that and give us feedback on that after your November
meeting is what we would like. We are not ready to
ask for it now.
MR. ZEFTAWY: You said the end of
November?
MR. KING: The paper is due to the
Commission at the end of November.
MR. ZEFTAWY: And we will get it the last
week in October?
MR. KING: We will get it to you as soon
as we can. It is written, but it is being edited and
comments incorporated, and so forth. So we have a
package prepared. We will get it to you as soon as we
can.
But in general this whole PBMR, we have
many more interactions that we need to have with you,
and this is not going to be the only topic that we are
going to talk about.
CHAIRMAN APOSTOLAKIS: Do you have
supporting documents that explain this approach?
MR. KING: We have received from Exelon on
August 31st a fairly thick package that explains this
approach. You should have it. I gave copies to
Medhat.
CHAIRMAN APOSTOLAKIS: Do you have it?
MR. ZEFTAWY: No. Are you talking about
the one for next week maybe?
MR. KING: That is the same one. That is
the same document.
MR. ZEFTAWY: Okay. That is the one in
the book.
CHAIRMAN APOSTOLAKIS: Okay.
MR. KADAMIBI: Basically, you have heard
that we are talking of course on treads on the ground
that have already been used in the past. There was
extensive work done on the MHGTR, and the staff put
out a lot review documentation, and we are treating
this as a run of the mill application of the
regulations basically.
CHAIRMAN APOSTOLAKIS: Go to the one that
says, "Staff Perspectives, General. Appears to be a
reasonable and structured method for screening
regulations."
MR. KADAMIBI: Yes, okay. That is number
seven. Well, at the level that we are talking about,
I guess what we are trying to point out over here is
that the four boxes that are covered by the
possibilities in the screening process seem to cover
it all.
That its regulations apply, and that they
are partially applicable, don't apply, or they may be
PBMR specific requirements which we might include in
the license condition, or tech specs, or things like
that.
But in terms of -- you know, what the
method does not offer, and what we find is that it
doesn't really offer a way to bring to the surface
safety issues that may not have been dealt with in LWR
space, which is when all the regulations were done.
But there isn't an automatic process to
bring up potentially significant issues, but that is
the sort of thing that we will have to cover as part
of or as we apply the top level regulatory criteria,
and go through the licensing basis events.
We would need assurance that in fact that
is a sufficiently comprehensive and complete set to
support the regulatory decisions. But the other point
that we wanted to make over here is that there isn't
right now on the table anything that is a substitute
for cool damage frequency and/or any kind of a large
release.
But one could foresee that there would be
things like just a calculated peak temperature within
a few -- a pebble bed. You know, that would serve as
a surrogate in the same way that CDF has served.
CHAIRMAN APOSTOLAKIS: Why? Why would you
need that?
MR. KADAMIBI: Well, the actual core
damage that I guess is in a light water reactor does
not apply in a pebble bed.
CHAIRMAN APOSTOLAKIS: I understand that,
but why are you looking for something to replace it
with?
MR. KING: It gets back to your question,
George, of the balance between prevention and
mitigation.
CHAIRMAN APOSTOLAKIS: Okay. That's what
I wanted to hear.
MR. KING: And what we saw on the curve,
the blue curve, was -- well, after you go through the
accident, here is what you get off-site basically.
CHAIRMAN APOSTOLAKIS: That's right.
MR. KING: It was nothing that dealt with
the prevention piece, and so that is the issue.
MR. KING: If the Chairman of the
Subcommittee agrees, can you tell us where you
disagree with what we heard, because a lot of this
stuff -- well, of course you have to be perceived as
being independent, but are there any points where you
disagree with what Exelon presented, or you don't
really disagree, but you really want to think about
it?
MR. KADAMIBI: I guess the level of
agreement at the high level, in terms of where the
four boxes in fact cover the range of eventuality, we
agree there. But actually when you come down to what
regulations apply or don't apply, I think we may have
significant disagreements.
CHAIRMAN APOSTOLAKIS: Sure.
MR. KADAMIBI: And we haven't really gone
through that. We haven't really gone through on our
own either that or a similar process.
CHAIRMAN APOSTOLAKIS: That's fair enough.
MR. BRENNER: I think what we see is that
the licensing approach provides a very good construct
by which the applicant and the staff are going to be
able to bring safety issues to the table, and discuss
those issues.
While not detailed in the licensing
approach, we have had a number of meetings with
Exelon, and we have pointed out to them that one of
the things that the staff is going to need to do is be
able to bring new tables, events, and put them on the
table to see where they fall given the events they
have selected.
And they fully agree, and they say that is
an inherent part of the licensing approach.
CHAIRMAN APOSTOLAKIS: And you are
comfortable with the blue curve?
MR. KADAMIBI: Yes, as a starting point
for the discussion.
CHAIRMAN APOSTOLAKIS: Oh, that is such an
answer. Oh, you have been here before. Very good.
DR. WALLIS: What he said before I thought
was significant. I mean, what we seem to be going
through here is Exelon tells you how you should
license their reactor, and I think you ought to be
telling them.
And you were saying that you have not yet
gone through a process like theirs to decide how you
would license the reactor.
MR. KADAMIBI: That's right.
DR. WALLIS: And doesn't that have to be
done rather than just accepting what they asked for?
MR. KADAMIBI: Well, the stage at which we
are does not really reflect an acceptance of
everything. It is really an acceptance of the
validity of the approach, where you go through a
screening of the regulations, and you develop some
kind of an objective basis for judging acceptability
and unacceptability, and trying to deal with those
issues that are going to be very difficult.
And there is a proposal on the table, and
we have got to begin somewhere, and we are going
through it. We only got their application -- I mean,
their submittal -- on August 31st, and so we haven't
really had much time to --
CHAIRMAN APOSTOLAKIS: I don't know how
much the fact that you have to write something by
November and when you received it at the end of
August, continues to the continued assurance of
regulatory independence.
I mean, I agree with Professor Wallis that
the approach may be technically sound, but boy, this
really doesn't look very good. They are telling us
how to license their reactor.
MR. KADAMIBI: Well, we are very sensitive
to that perception, and at the same time the
Commission has asked us to engage in early
interaction.
CHAIRMAN APOSTOLAKIS: And that's very
good.
MR. KADAMIBI: And so we are engaging in
early interaction, and we will maintain independence
and bring up the kinds of questions that -- well, I
think the basis for our regulatory review will to some
extent naturally bring up -- such as Reg Guide 1.174.
If we look at really how it applies in
terms of defense in depth, Reg Guide 1.174 lays out
seven attributes that we would look at, and we could
use those as guidelines.
One of the things that we noticed is that
they only used six of those seven. Now, because this
is an ongoing interaction, we don't necessarily know
why only 6 of the 7 were chosen.
But probably there is no reason to exclude any of
those seven.
DR. BONACA: Another thing is that I don't
think that they are proposing regulation. I think
they are proposing a way in which they can license
this plant under existing regulations it seems to me.
DR. KRESS: That's my opinion, and I don't
think you should let the concept of lack of
independence color your view too much. I mean, if you
agree with the proposed approach as having a lot of
merit, I would urge you not to let the concept of a
perceived lack of independence color that.
MR. KADAMIBI: Well, we have to assure
ourselves that we have a sound basis for that.
DR. KRESS: I think you are going to have
to look at it and see if it is a sound basis, and is
going to protect the health and safety of the public.
MR. KING: Ultimately that has to be the
staff's licensing criteria.
DR. KRESS: It will be the staff's
licensing criteria, no matter where it comes from.
MR. KING: Whether the bright idea comes
from the staff or somebody else is secondary.
DR. KRESS: That's right. You certainly
don't want to dismiss a bright idea just because it
came from outside.
DR. POWERS: When you think about this
licensing process, do you think about it as here is a
site with a control room and a reactor, or do you
think about it as a site with a control room and 10
reactors?
MR. KADAMIBI: We haven't really gotten to
that point yet. I myself have not given that any
thought. I don't know about Eric.
MR. BRENNER: And we early on, there was
some discussion between us and the potential
applicant, and the applicant stressed that they wanted
to focus on the approach, versus the design.
So to that extent, we have tried to not
look at particular aspects of the design, but look to
make sure that the approach can handle questions like
that of, well, okay, because this is different from
how we have maybe licensed plants in the past, will
there be a way for the staff to interject those issues
as the applicant is working through the licensing
approach.
And that sort of thing has been the focus
of our judgment of acceptability, versus
unacceptability, for the approach; as opposed to
acceptability for --
DR. WALLIS: Well, you knew that you had
to face this licensing of unusual reactors, and it
would seem to me that the staff would know it before
it saw anything from industry to look at the
regulations and say how are we going to do it.
And to have some ideas generated here, and
not coming all from outside, about how to license
these new reactors. I have not seen any independent
assessment.
All your presentation here is based on
ideas that came in from outside. Weren't there some
ideas here before you got ideas from outside?
MR. BRENNER: Yes, and two of the things
that we have looked at already is the previous
licensing of gas cooled reactors that the NRC has
done, and we looked a lot at --
DR. WALLIS: So you had your own blue
curves and things like that, or some sort of curve,
before you saw these that came in from outside?
MR. KADAMIBI: The way that I would put
it, Dr. Wallis, is that even if we used the curve as
some kind of value in what we are doing, what we use
it for could be quite different from what they used it
for.
DR. WALLIS: Did you have any ideas before
ideas came in from outside, and what did they look
like?
CHAIRMAN APOSTOLAKIS: Did the Option 3
report help you at all?
MR. KADAMIBI: Yes.
CHAIRMAN APOSTOLAKIS: And would that
serve a purpose of the thinking that Dr. Wallis wants?
Was there anything there that would apply to these
reactors?
MR. KADAMIBI: Well, Option 3, I believe,
was to support rule making.
CHAIRMAN APOSTOLAKIS: Well, 1-174 was to
support the request for changes, but now it is used
for other things.
MR. KADAMIBI: Well, we have gone through,
I think, a relatively and systematically way to find
the principles in Reg Guide 1.174, and I think we see
a very clear application of some of the basic
principles.
In terms of the Option 3, I think the
concepts of prevention and mitigation, and how one can
use quantitative analyses will be very useful. But
exactly how they might be useful, I don't think we
have come up with that yet.
MR. KING: The Option 3 work is very
useful. Option 3 categorized events by frequency
categories. That thinking applies to the thinking we
heard from Exelon.
You know, the same kinds of questions that
we had to wrestle with there are the same kinds of
things that have to be addressed in this proposal; the
balance of prevention mitigation, and that stuff
applies.
CHAIRMAN APOSTOLAKIS: Yes, but there is
no discussion of it here, and that is what Dr. Wallis
is saying. I mean, all you are doing is you are
responding to what --
DR. BONACA: Well, I want to say that for
older plants that you could take right now the
combination of ANSI standards, Chapter 15, and SECY
goals, and draw the same curve that they have, and it
would be with a band around it because they wouldn't
have some foundation.
But all I am trying to say is that this is
not a revolutionary approach. It is an approach that
has been used before. You have not seen a blue curve,
but you saw pieces of it in different portions of the
regulation that you had to meet in order to license
plants.
So I am saying that I don't think it was
a strikingly or radically different approach.
DR. KRESS: But it puts it all together
though.
DR. BONACA: It puts it all together.
DR. KRESS: And in fairness to the staff,
this particular presentation was supposed to be what
was their perspective and response to the Exelon
proposal. We weren't asking them to say what else is
out there that they might do. I am sure that they
will think along those lines.
MR. KING: Or if we took a clean sheet of
paper what would we come up with.
DR. ROSEN: I think that Prasad made that
very clear in his remarks, that they are thinking
beyond what the Exelon people put on the table.
And then the statement of what else does
the staff -- what things will fall outside this
protocol if that needs to be brought to the table. If
you just use the protocol that they suggested and put
blinders on, clearly you may miss some things.
And I think that Prasad was quite clear that that is
not what the staff was doing.
CHAIRMAN APOSTOLAKIS: Well, Exelon, in
their last slide, stated that they wanted agreement
from the staff on six processes. You know, an
agreement on the process for equipment classification
and the development of RDC, et cetera, and that is due
up to the Commission by the end of November. That
SECY will do that, will address these?
MR. KING: That SECY will go as far as we
can go in November. We may have some IOUs in that
SECY for follow-up activities.
CHAIRMAN APOSTOLAKIS: You guys need at
least a month for reviews. So, we are talking about
lightening speed here in approving all these things or
agreeing.
DR. ROSEN: But where does the ACRS get
involved?
CHAIRMAN APOSTOLAKIS: We don't. I guess
they are going to give it to us at the last moment.
MR. KING: Well, we will get the draft
SECY hopefully in a few weeks. We would like to get
your views on it. As I said the SECY will go as far
as it can go.
The reason that we picked November was
because when we first sat down and laid out our pre-
application plans and discussed it with Exelon, they
were looking for feedback as soon as they could get
it.
We thought that the earliest that we could
get them something would be November. It may not be
everything that they want, but we are trying to be
reasonable, in the sense that they have decisions to
make.
And they are looking for feedback, and we
are trying to get them whatever we can get them by the
end of the year. It gives the Commission a month to
look at it.
DR. BONACA: I have a question that I
would like to ask you here. Regarding the blue curve,
we heard from Exelon that it is not a frequency
consequence curve.
But in your presentation, you are calling
it a proposed frequency consequence curve. I mean, is
there confusion there on what it is and how they are
using it?
MR. KADAMIBI: Our view on this is that
the actual numbers on there may or may not mean a
whole lot other than representing something that is of
a fundamental regulatory value, which is that the
higher the consequences, the less likely it should be.
And this generally represents that concept. Also --
CHAIRMAN APOSTOLAKIS: There is some
contributions or on individual sequences, and that is
a conceptual problem that has nothing to do with the
numbers.
DR. BONACA: Absolutely.
CHAIRMAN APOSTOLAKIS: The frequency
consequence curve means cumulative. Now, if that is
the wrong interpretation and is interpreted in a
different way --
DR. WALLIS: They are not talking about
concepts. Everybody understands concepts. We are
talking about hard criteria that you are going to
apply to license or not license a reactor.
MR. KADAMIBI: Well, that curve I don't
think will be the hard criterion.
DR. WALLIS: Well, what are they, or what
ideas do you have about what they might be?
DR. KRESS: What is already in the
regulations.
DR. WALLIS: What is it that you need to
invent or change, or whatever?
DR. KRESS: The design or the selection of
the design basis and events that will have to meet the
regulations. That is what this is about.
MR. KING: They have not selected a design
basis and --
DR. WALLIS: Did you know the blue curve
before it was shown to you by Exelon?
MR. KADAMIBI: It was part of the MHGTR.
It wasn't new and the concept has been around.
DR. WALLIS: So the idea that they are
showing you something is wrong. You knew this before
they showed it to you?
MR. KADAMIBI: I got it from NUREG 13-38,
which is where the staff reviewed what MHGTR had
proposed, and we are offered is something that the
staff has spent a considerable amount of time
reviewing in the past. So we want to take as much
advantage of that as possible.
DR. WALLIS: Well, we knew all of this
beforehand because you had been through it before, and
therefore, you are in a good position to evaluate it.
So, I wouldn't do away with the impression that it is
something that came all from outside in some way.
MR. KING: If you were expecting we were
going to come in here with a design basis, the
accidents, the GDCs, and all the other criteria that
we have now decided to apply, we are not ready to do
that.
DR. WALLIS: No, I just wanted to get away
from the impression that Exelon is telling you how to
do the regulation.
DR. KRESS: Exelon is telling them what
the regulations already consist of.
DR. WALLIS: Well, why do they have to
tell you?
DR. KRESS: Well, they knew that. They
are just putting it down on paper so that it is a
point of focus.
MR. KING: What they are telling us is
that how they propose to go through the regulations
and decide what applies and doesn't apply to PBMR.
DR. KRESS: Yes.
MR. KING: Their ideas as to how they
would like to do it and how they would like us to
agree with the way they would like to do it. Our job
is to take a look at that and say does that sound
reasonable or not, or do we have another way that we
think it ought to be done, and that is what we are
doing.
DR. ROSEN: I think that Exelon is free to
propose anything that they want, but one thing that
Exelon can't do is license a reactor. That can only
be done by the Commission, and that is what we are
doing. That is what we are looking at.
CHAIRMAN APOSTOLAKIS: I guess Graham's
point is different though. He is saying that instead
of starting that way, and where you have a proposal
from the applicant, and you say, gee, here I agree or
there is another way, he is saying why haven't you
thought about other ways before you got that. I think
that is the thrust of his question.
DR. KRESS: Yes, but once again, this is
the way in principle that they have been licensing
reactors for years, and it is just putting it down in
a systematic focus and basis.
DR. POWERS: The difficulty that I am
having here is that this is exactly the way they have
been licensing reactors for years, and should that be
the way that we do things?
I mean, shouldn't we say what is the
safety that we are trying to achieve, and then define
what regulations you need to achieve that? And then
see which ones you have and which ones you don't have?
DR. KRESS: Yes, and to complete that
picture, I would have liked to have seen the
cumulative curve --
DR. POWERS: I am not wild about that, but
--
DR. KRESS: But that would have defined
what we were trying to achieve, and this is a way to
achieve that, although there is a disconnect between
them. It is not clear how this leads to achieving
that other one.
But it is the same thing, and it is not
clear how the design basis accident now leads to
achieving the safety goals. And you can meet those
rules and regulations by many different paths.
CHAIRMAN APOSTOLAKIS: I think you are
going to have a hell of a problem with defensing that;
picking it out of 1.174 is --
DR. KRESS: I think the issue is going to
come down to what in the heck do you mean by defense
in depth.
DR. POWERS: Well, you have the challenge
of with your frequency consequence curves, there are
some high frequency things that you are going to allow
to occur, upside events that you are going to allow to
occur.
And it seems to me that I may be
substantially more conservative than you and say that
I don't want that kind of thing to occur, period.
And I think there is more than just
defense-in-depth that is going to be a problem.
DR. ROSEN: I see another big problem, and
that to me is the risk matrix. I have to define a new
set of risk matrix, and --
DR. POWERS: Well, I think that CDF is
shot here.
DR. ROSEN: Clearly, but I didn't hear
much thinking about that from either the licensee or
the staff.
DR. KRESS: Well, the matrix may be on the
bottom curve there; frequency of which you exceed a
certain dose there.
DR. ROSEN: But now the ACRS is defining
risk as --
DR. KRESS: Well, these guys know that is
--
CHAIRMAN APOSTOLAKIS: You know, I am
really curious. Maybe it sounds like a crazy idea,
but I would really like to understand or know how
many times has industry and regulators which have been
surprised by operating experience in the last 40
years, because that tells me a lot about defense-in-
depth.
DR. POWERS: Can you ask the question
again, George?
CHAIRMAN APOSTOLAKIS: How many times have
we as a community been surprised by the operating
experience.
DR. POWERS: A bunch.
CHAIRMAN APOSTOLAKIS: Well, I would like
to understand that. I mean, that would be a nice
little project, because that tells me how cautious I
have to be for the future, which means defense-in-
depth.
The words are coming are out with great
difficulty, but it is the structure of the approach to
defense-in-depth.
DR. POWERS: It is the correct approach.
DR. BONACA: Mr. Chairman, we are
hopelessly late.
DR. KRESS: Can you guys summarize?
MR. KADAMIBI: Well, if I were to
summarize, I would go to slide 12. This is really
what -- this really captures many of the concerns that
I have heard expressed over here and these constitute
the central challenges that I think we face.
And we are cognizant of it, and the burden
is heavily on the staff in order to deal with these
issues. At this point, I don't think we have really
seen enough about the substance of the repeat PBMR
design in order to be able to say very much about any
of these.
DR. WALLIS: Well, I think you have a
great opportunity here to relate the top level
regulatory criteria, and the real question of adequate
safety and all that to the way that you apply them to
this new thing.
You have a wonderful opportunity to make
things more rational than they were in the past. I
think just responding to someone else's idea just
isn't good enough.
CHAIRMAN APOSTOLAKIS: What do you mean by
enhanced level of safety margins?
MR. KADAMIBI: Well, that is an
expectation I believe of the Commission, in terms of
we don't require or we don't really require a higher
level of safety, but through applying concepts of
simplicity and passive systems, and things like that,
we would expect that there would be an enhanced level
of safety.
CHAIRMAN APOSTOLAKIS: Okay. Thank you,
Gentlemen, and thanks to Exelon for their
presentation. We will reconvene at 5 minutes past
2:00.
(Whereupon, at 1:05 p.m., a luncheon
recess was taken.)
. A-F-T-E-R-N-O-O-N S-E-S-S-I-O-N
(2:05 p.m.)
DR. BONACA: All right. Our Chairman had
to leave for a few minutes, and so I will be chairing
this part of the meeting. Right now we are going to
review the action plan to address ACRS comments and
recommendations associated with the differing
professional opinion of steam generator tube
integrity, and I will turn to Dr. Powers, who is the
Cognizant Engineer.
DR. POWERS: The Chairman of the
Subcommittee will present a summary and that Chairman
is Dr. Ford, formerly of General Electric.
DR. FORD: I will do exactly the same as
you did. I will handle the Chairmanship.
DR. POWERS: I thought you were going to
offer an opening summary.
DR. FORD: I will let you do that.
DR. POWERS: Well, being caught completely
flat-footed here, that the objective is for the staff
to come up and discuss a little bit on what they are
doing in their steam generator action plan.
I understand that this is an action plan
that has existed for some period of time, and has been
augmented by the staff to address some of the comments
and recommendations we made in the report that the
Committee endorsed on the differing professional
opinion concerning steam generator tube integrity.
I think that you will find that the staff
has gone more beyond than just looking at some of our
explicit recommendations, but rather has very
carefully scrutinized the report, because in many
places in the text we come along and say here are some
thoughts and comments on this, and they have taken
them to heart.
And they have come up with a plan that
seems to address most of our comments. But what I
don't know is who is the speaker is going to be. And
let me go on and say that we did have a subcommittee
meeting on this, and a substantial portion of the
plan, or some portion of the plan, has been relegated
to research to address.
And they have some very exciting results
that are going to be presented to the Committee, and
more for interest than they are for reviewing. So,
with that --
MR. SULLIVAN: Well, all I am going to do
is make some opening remarks. There is not too much
that I have prepared on this little sheet that you
didn't already go over, but since it is very short, I
will just go quickly through it again.
My name is Ted Sullivan, and I am the
section chief in NRR responsible for steam generators.
In November of 2000, NRR prepared a steam generator
action plan that addressed the activities that we were
going to undertake in response to the Indian Point-2
lessons learned task group report.
And the related OIG event report on Indian
Point-2, and other ongoing activities related to NEI
97-06. So that is what we did in late 2000. Then by
the time of May, after we had had time to study in a
fair amount of detail the recommendations in the ACRS
report related to the DPO, which is NUREG 17-40, we
expanded the action plan to address the activities
that we were going to undertake in response to those
recommendations.
The major activities as Dr. Powers
indicated are being undertaken -- I think you did
indicate this, but they are being undertaken by the
Office of Research.
We understand that as a couple of people
have said already that the purpose of this portion of
the meeting is to understand how NRC is responding to
the recommendations of the ACRS on the DPO.
And what we intend to do for the rest of
this hour is to present a summary of the work on the
near term research milestones on the actual plan, and
that is what Dr. Powers was alluding to.
DR. POWERS: One thing that I would
correct you about is that I don't attach anything
minor -- well, the word minor does not come to mind
when I think about those activities that NRR is
responsible for in the action plan in responding to
the recommendations.
MR. SULLIVAN: I would agree, but I think
that in terms of resource expenditures --
DR. POWERS: That may be true, but the
resource expenditures I agree with you, but I think
the things that are going on to look for significant
deviations from the expected linear response, I think
that is an extraordinarily important activity, and
minor is just not a word that I would attach to it.
MR. SULLIVAN: Okay. I would agree with
that. I also wanted to let you know that we are not
going to hit the entire steam generator action plan
today or even all of the DPO issues.
But there is a description of the
milestones related to all of the action plan, but in
particular the DPO, on a web page that has
specifically been created for the steam generator
action plan.
And the staff's progress on those
milestones is updated quarterly. So if you want to
track that between meetings that is available. And I
guess with that, I would just turn this over to Joe
Muscara, who is going to lead off these presentations.
DR. POWERS: Never let it be said that a
metallurgist can't handle a few engineering activities
here.
MR. MUSCARA: Good afternoon. I guess by
way of introduction, some of the things that I was
going to say have already been said, but maybe I will
just repeat them.
One of the points of interest was that
when the staff reviewed the RES report, we didn't just
look at the recommendations. We studied the report in
detail, and when ever we found errors where there was
an interest or a lack of information, we decided to
address those issues also.
The report actually helped us to focus the
research and to get support for it. I would also like
to point out that much of the work was ongoing, and is
ongoing. So they have helped us to address in the
near term some of the issues that were represented in
the report.
As has been mentioned, we do have an
action plan, and this is updated monthly. So the
current status of the work to address the DPO issues
is available in the action plan.
What we were planning on doing today was
to essentially address some of the recently completed
research, and also to talk about some of the near term
milestones.
So I will be discussing the work related
to materials and inspection, and Chris Boyd will give
us an overview of the thermal-hydraulics work.
And he will also address some results from CFD
calculations that he has completed recently.
Now, Milestone 3.1 in the action plan
deals with understanding any possible crack growth
during a main steam line break. The work that we are
planning on doing here, we will be looking at some
thermal-hydraulics evaluations to calculate the loads
that are experienced by the tubes in these conditions.
This will be conducted by the staff. We
will also be looking at work that has already been
performed and that is available in dockets. Based on
this review and in conducting thermal-hydraulics
evaluations, we will come up with some upper-bound
estimates of the loads.
And then with these kinds of input, we
will evaluate the growth of existing range and types
of cracks to see whether these cracks will propagate
under the steam line break loads.
We will also be estimating the loads that
are required to propagate a range of cracks, so that
we can get an idea of the margin for propagating these
flaws over and above the main steam line break loads.
At this point if we find that there is a
great deal of margin, I think we do not need to do any
further work to better define the loads, but if we do
not have large margins, we will then do additional
work in the thermal-hydraulics area to better define
the loads.
Much of this work will be completed by the
end of the next calendar year. Once the evaluations
have been completed, we will also conduct some tests,
including both the pressure stresses and axial and
bending stresses, to essentially validate our findings
from other core results.
DR. POWERS: Isn't it a case that the
loading on the tubes is simple; that we have complex
multi-dimensional loading that complicates these
things?
What the subcommittee was thinking of were
in terms of the support plates, oil canning, and even
painting. And we looked upon those as very
complicated loads that might be difficult to assess
the magnitude of strictly by analysis.
MR. MUSCARA: Well, we will have to
monitor that situation. I think the worse condition
is when the tube is locked between support plates.
And then we should be able to get some input from the
thermal-hydraulic calculations on the kind of force
that the support plates are experiencing at a
particular location.
And then we will evaluate what kind of
loads resulting from that, including body weight
cyclic loads. And then we will evaluate the growth of
cracks.
And then in the testing itself, we will
try to simulate the kinds of loads that we predicted
in the tubes.
DR. POWERS: I know it is probably too
soon to ask the question, but I will ask anyway. You
have two tube support plates, and if they go into
oscillation, they don't need to be in-phase do they?
MR. MUSCARA: I suppose not.
DR. POWERS: And that would create just a
horrific situation I think.
MR. MUSCARA: Well, those are the kinds of
things that we have to be looking at.
DR. SHACK: And they are sort of coupled
by about 5,000 tubes though.
MR. MUSCARA: Well, there are assumptions
that the overload is transmitted to one tube, but many
of these tubes are a lot.
DR. BONACA: So now some tubes may be a
lot and some may not.
MR. MUSCARA: Right.
DR. BONACA: So you will have to look at
what some range of sensitivity might be.
MR. MUSCARA: Some range, and we have to
make conservative assumptions about the numbers of
tubes that may be a lot.
DR. WALLIS: Has anybody got any estimates
of these loads so that you know what sort of thing you
are dealing with before you go on to something more
complicated?
MR. MUSCARA: There have been several, and
maybe a couple of submittals from the utilities, where
they want to take advantage of the support plate being
logged, and they are requesting the use of a higher
voltage and voltage lows, and in that work they have
conducted a number of analyses to try and predict
those lows.
And we will be reviewing those, and in
addition we will be doing our own hydraulic
calculations to see how close these are.
DR. POWERS: Well, even relatively simple
axial and bending loads have been done as part of
setting up the alternate repair criteria.
MR. MUSCARA: I think in that case that
they are assuming that the support plate was not
there. So they do not have to go through that
exercise.
In Milestone 3.2, we were interested in
evaluating the effects of jets impinging on adjacent
tubes, both under severe accident conditions, and
under steam line break conditions.
Last year, we presented some results to
the ACRS on work that had been conducted with respect
to the erosion of tubes due to the erosion impact from
severe accidents, and we had concluded at that time
that the degradation was very minimal and that this
was not a concern.
And the ACRS recommended that we should
possibly run some longer term tests. Our original
tests were 10 minute tests, and so we have since
conducted some tests of a duration of 30 minutes.
And we have found that we have exactly the
same data, and so even with the 30 minute testing, we
still have very low rates of erosion under those
conditions. And the rates range from somewhere
between 2 and 5 mils per hour. So we still that is
not an issue.
We conducted a number of tests to evaluate
jet impingement on the steam line break conditions.
Much of that work is completed, and we are doing some
validation work right now on real cracks.
But I would like to show you some of the
data from those tests since you have not seen it
previously.
DR. WALLIS: How do you characterize the
jet? It is a two-phased jet?
MR. MUSCARA: A two-phased jet. Last
year, we presented work on CFD evaluations to get the
properties of the jet velocities, and temperatures,
and we made use of that information.
DR. WALLIS: Well, did the tests make an
effort to model the upstream conditions for the jet
and everything, and to get the jet velocities and
quality, and everything right?
MR. MUSCARA: In the severe accident work,
we used the rig at the University of Cincinnati, where
we were essentially using a burner. We injected
particles into this burner, and we worked on a range
of velocity conditions; from very little, and up to
about a thousand feet per second or meters per second.
We also found in that work, and based on
the CFD conditions, that when the jet impinges on the
tube, the velocity is down pretty much to nil. In our
evaluations, we assumed the velocity of 200 meters per
second. So we were conservative in that respect.
DR. WALLIS: Isn't that jet a steam water
jet?
MR. MUSCARA: No, under severe reactor
conditions it would be an aerosol, and materials
evaporate from the core.
DR. WALLIS: Okay. So you have to model
that somehow.
MR. MUSCARA: Right. In a steam line
break, when we have a blow down facility, we reproduce
the conditions inside the tubing and outside of the
tubing.
Of course, on a steam line break, there is
no atmosphere on the outside. So we have conducted
those tests under 2400 psi pressure, at a range of
temperatures.
And we find that the amount of erosion is
dependent on the temperature, and it is dependent on
the amount of sub-cooling. The greatest amount of
erosion occurs about 280 degrees centigrade, which is
about the cold wet temperature.
We would not expect to see -- well, cracks
are less likely in the cold leg, and around 300
degrees C where we have the hot leg temperatures, the
erosion rates are quite diminished.
DR. WALLIS: Do you understand the reason
for this dramatic bump in the curve; the mountain of
erosion that occurs over a narrow range of
temperature?
MR. MUSCARA: Again, we need to look that
this is a two hour test, and at 2400 psi, and the peak
is about 27 or 28 percent.
DR. WALLIS: But why does it suddenly
change under a certain temperature?
MR. MUSCARA: Well, as the temperature
goes up, we are starting to get some flashing, and so
the jet dissipates.
DR. WALLIS: So that is the place where it
changes its two-faced conditions or something like
that?
MR. MUSCARA: We think so, because based
on evaluating the condition of the surface, we would
see a larger area that is being affected. Also, as we
penetrate into the tubing, then there is water present
at the bottom and that acts as a cushion.
So the tests that we conducted for longer
times would not get any greater penetration than --
DR. WALLIS: Is this a condition where the
jet forms bubbles, which then collapse by cavitation
on the target?
MR. MUSCARA: No, we don't believe that is
the case under these test conditions.
DR. SHACK: Well, I have a conflict of
interest here, but at the 300 and 320, the jet will
flash obviously given enough time under all these
conditions, because it is under pressure and
temperature.
But since it is a transient thing, we can
sort of see from the impact area that it hasn't
flashed at 280 or lower temperatures. There is no
flashing that goes on. The diameter of the impact
area is the same as the diameter of the exit hole.
And at 300 to 320, it is flashy. You
know, the impact area is spread out.
DR. WALLIS: And so that is more benign?
DR. SHACK: That is more benign. You
know, we are getting small droplet impacts. The guess
is that we are getting a short of bundle cavitation
damage there at the peak, and when it drops off again,
we are not getting the cavitation damage.
But it is very difficult to really know
anything except that the peak goes up and the peak
goes down.
DR. WALLIS: But there is a small window
there where you will form bubbles in the jet and it
can collapse on the target. You have to have just the
velocities and temperatures just right for that to
happen.
DR. SHACK: And it seems to be that degree
of sub-cooling that you just happen to get in the cold
leg?
MR. MUSCARA: We are currently conducting
additional tests at different pressures, and so we are
trying to understand this better.
DR. WALLIS: Is there literature on this
bubble formation cavitation in a jet which is close to
flashing? Is there literature on that?
DR. SHACK: We can't find any.
DR. WALLIS: I remember that I thought
about it in various circumstances, and I don't know of
any literature either.
DR. WALLIS: There is lots of literature
ont the droplet impact, but --
DR. WALLIS: No, that is not it. It is
the bubbles that have to form, and then they have to
collapse.
DR. SHACK: And it clearly is a lot more
dangerous.
MR. MUSCARA: We are planning on providing
some topical reports by the end of the calendar year
to describe the work on the jet impingement under
severe accident conditions, and also on main steam
line break conditions.
So there should be a lot more detail
available at the end of this calendar year. Milestone
3.6 of the action plan addresses the issue of the POD.
And the ACRS made some comments as to the possible
better use of other parameters rather than a constant
POD.
Fortunately, we have been doing work in
this area for a number of years, and we were trying to
quantify the ability of flood detection using current
techniques and commercial teams for realistic kinds of
flaws.
We had completed work in the past, but we
recently completed some of the analyses of this work,
and I thought it might be useful to show you some of
those results.
The POD that we are using now is a
constant number of .6 POD for any size or any voltage
flaw. Without going through a great deal of detail,
what I wanted to show you is that we now have data on
the POD as a functional flaw depth, and as a
functional voltage; and four different kinds of flaws
in different locations; ID and OD SCC at the support
plate, and at the tube sheet, and the free span.
And besides depth, flaw depth and boltage,
we were also evaluating the data against the parameter
MCP. This is a fracture mechanics parameter, and MCP
describes essentially the stress consideration in the
ligament at the cracked tip.
MCP is a functional of both the flaw
length and depth. So it is very strongly dependent on
flaw geometry, and of course this dictates the failure
of pressures of these tubes.
And so we have also plotted the
probability of detection for flaws against this MCP
parameter, and a MCP value of about 2.3 is that
location where the flaw can no longer meet 3 delta-P
So what we can see from these foils is
that for MCP that around 2.3 probability detection is
fairly high. So a flaw that would fail under 3 delta-
P would be quite detectable.
And just to give you a feeling for the
team performance, the last view graphs are based on
the results from 11 teams, and in these graphs we are
showing the performance on a team by team basis.
DR. WALLIS: So the probability of
detection is very dependent on the people?
MR. MUSCARA: That is what this graph
shows for some conditions. You will notice --
DR. WALLIS: Especially the one on the
bottom right.
MR. MUSCARA: Right. At the top, we have
the cracking and tube support plate cracking, and
these are fairly common cracks, and so you expect to
see these quite often.
What we see at the bottom is work on POD
on the free span, and of course we don't get a lot of
flaws in the free span, and so sometimes these confuse
the inspectors, and they are not willing to call it a
flaw, and they think it is something else.
In the bottom right we are showing
information on the POD for the tube support plate
location from stress corrosion cracking on the inside
diameter. Many of our flaws that are stress corrosion
cracking on the inside diameter are also accompanied
by denting.
Now, the denting produces a fairly large
complex signal, which makes the detection more
difficult. So there is a signal there, but the
inspector is to decide if this is a flaw or not. Is
it a flaw, or is it the signal from the dent.
And clearly some inspectors do quite well,
but then some other inspectors don't do quite as well.
So we can see the range of performance we can get from
team to team, which is quite useful, for example, for
Monte Carlo evaluations.
DR. WALLIS: From 10 percent to a hundred
percent. In one case it is a huge variation.
MR. MUSCARA: Well, it is a large
variation, but again we need to look at flaws of a
depth that might be of concern. So a 40 percent flaw
for early detection for the worst team is not very
good, but a tube can withstand -- a tube with an 80
percent through flow and 85 percent through flow can
still withstand a steam line break.
DR. WALLIS: How shall I interpret that
these points are in relationship to the solid curve
then?
MR. MUSCARA: I'm sorry, but the solid
-- these are all different symbols for different
teams.
DR. WALLIS: That's right.
MR. MUSCARA: So a curve is a particular
team, and a given series of points is another team.
DR. WALLIS: For the solid curve?
MR. MUSCARA: For example, the green curve
--
DR. WALLIS: Well, it misses all the
points in there?
MR. MUSCARA: No, the points are just a
different team.
DR. WALLIS: That's right.
MR. MUSCARA: This just fit a logistic
fit, and so the points are only describing not the
data points, but the team performance. So the dash
curve is a particular team, and the green curve is a
team, and the triangles is another team.
DR. WALLIS: The green curve is a team?
MR. MUSCARA: Yes, and that team performed
quite well.
DR. SHACK: That is the POD for the best
team, and the dash line is the POD for the worst team;
and rather than putting 11 curves where you run out of
ways to distinguish them, we use symbols for the
intermediate teams. And also it confuses you, Graham.
MR. MUSCARA: One thing that is worth
mentioning is that the worst team was not always the
worst team. That is for different flats and different
locations. There was no consistency, and the worst
team is not the worst.
DR. WALLIS: You should use this method
for the ACRS members on what should go on a letter.
DR. POWERS: But no member is ever wrong.
So there is no flaw.
DR. WALLIS: Sometimes the teams --
DR. POWERS: Low detection rate.
DR. ROSEN: Are you claiming that one of
our members is flawless?
DR. POWERS: All of our members are
flawless.
MR. MUSCARA: This particular view graph
is not really related to materials issues, but it is
the task on the item of spiking, and I guess very
briefly what I want to say is that the staff has
conducted a review, and it was completed this summer.
We plan on developing a staff position on
this by the end of the year, and then to provide this
to the public, and to have public comment, and then
finalize our position with respect to ACRS comments.
The last area that I would like to address
is clearly this is not an internal milestone. Noticed
that we finished up in '95 or '96, but work will be
going on in this area beginning the next calendar
year.
And this has to do with getting a better
understanding of stress corrosion cracking phenomena.
So we are interested in finding out better information
on cracking initiation, and crack growth, and crack
evolution.
A particular interest of ours is to really
understand crack evolution, because as cracks progress
from their infant stage, where we have many small
cracks with ligaments in between, they eventually get
to a point where the ligament is small, and it no
longer provides any strength. And then the cracks
will join up.
The reason that the voltage based criteria
works right now is because cracks really are in a
infant stage, and these cracks, although they are many
cracks, have ligaments, and they exhibit very high
burst pressures.
But we need to understand better when we
start losing the ligaments, and when we need to use a
different structural integrity criterion. So we want
to understand both the initiation, the evolution as a
crack changes from small cracks to larger cracks, and
of course the crack growth rates.
This work is really in the planning
stages, and not a lot of detail is available yet, but
those are the key features if you want to study them.
DR. POWERS: Is this all focused on the
600 alloy, or is it also looking at 690?
MR. MUSCARA: Thank you for reminding me.
We would like to of course look at 600 because we have
a lot of field experience with this material, and so
we understand its behavior somewhat in the lab, but we
must also understand its behavior in the field.
Along with these tests we will be
conducting 690. And the idea here is to be able to
understand the behavior of 690 in the field by
understanding its laboratory performance, as compared
to 600.
DR. POWERS: It seems to me that we really
need some technical guidance on that. There are a lot
of licensees coming in with 690 steam generator tubes,
and they are saying, gee, scratch our inspection
intervals because this material is more immune, and
has more sensitivity maybe is the right word to stress
corrosion cracking.
And it seems to me that we really need a
technical foundation for deciding what to do there.
MR. MUSCARA: Yes. One of the things that
we will be looking at for 690 is that we clearly know
that the material cracks in the laboratory. And it
cracks under conditions which may not be atypical of
steam generator conditions, because it cracks under
fairly neutral conditions in impurity environments,
such as things with copper, sulfate.
And so we know the material cracks, and we
want to be able to bound the conditions under which it
cracks. One area of concern for us is the inspection
interval.
There are two items of concern. One is
that we want to stretch -- maybe industry wants to
stretch the inspection interval, and secondly, we are
inspecting a small sample. There is maybe 20 percent.
If we stretch the inspection interval, and
some cracking is going on, and we don't catch it on
time, and if we use a small inspection sample, again
we need to have a great deal of degradation in the
generator before a small sample picks up the problem.
So we need to be careful about the length
of the inspection interval, and the sample size that
we are using. Clearly the material behaves better in
the laboratory than 600. So it probably has a longer
useful life.
But we don't know that we can make it
through 40 years. So it may be reasonable for the
first inspection cycle to have a longer length, but I
am not so sure that after a certain amount of time
that they should not be inspecting the same kind of
frequency as the materials that we are inspecting now.
DR. FORD: Joe, before you get off that
one, could you comment on the relationship between
this task, 3.10, which is more of a quantitative task,
with that of 3.8, which is looking to see whether
there is a linear bounding relationship.
MR. MUSCARA: There is a great deal of
confusion with this topic. We had a very good write-
up in the ACRS report, and clearly we agree that
stress corrosion cracking is not a linear phenomena.
We know this.
What has turned out to be linear is the
correlation between the voltage growth rate -- well,
since voltage does not trap crack size, it cannot
track cracking rates. So the phenomena is not
changing. It is only linear. The voltage seems to be
linear.
So we will try and get a better
understanding but I think the answer is that the
voltage is not tracking what we are looking at. It is
not really tracking crack growth rate.
And since there is a lot of scatter in the
voltage correlation with crack size or burst pressure,
you can draw almost any correlation through that data,
and right now we have a linear correlation of time
with voltage growth rate.
But that does not mean that is a linear corrosion
between time and crack growth rate.
DR. WALLIS: If the end result is to
predict as a function of time, when you have a near
tube rupture event, are you saying that any current
will never meet that criteria?
MR. MUSCARA: No, what I am saying is that
the voltage will not meet that criteria. But we
presented also last week some work, but in more detail
last year.
We have been developing some techniques in
the laboratory for accurate sizing of flaws. I think
the critical parameters really are the flaw geometry,
the flaw length and depth.
If you can't accurate measure these by any
current, then you can't accurate predict the burst or
the failure pressures, or the burst pressures, of the
integrated tubes.
The work that we have done so far has gone
quite well in predicting the failure of tubes that we
have tested in the laboratory.
By using this advanced sizing technique,
and then from that data predicting the burst, and we
run a test, and sure enough we are within 200 psi of
the burst pressure in many cases.
So the key here is not that any current
can't do this job. It can, but you have to use the
right parameters. And the right parameters in my view
are the flaw size, and the flaw shape, and not the
voltage, which does not relate to flaw tightness, or
length and depth, and so on.
DR. WALLIS: So there is not a big jump in
application technology? You are not talking about
changing the whole NDE industry on its head.
MR. MUSCARA: We in fact are using the
same probes that industry is using, 10-K probes, and
what we are doing is data analysis, which is fairly
different from what the industry is using.
We are doing this right now in the
laboratory, and so it is not a user friendly technique
at this point. But there is a lot of interest from
the industry. EPRI is interested in this technique,
and we recently had an e-mail from
Westinghouse. They want to come in and look at the
technique, and be able to use it.
So the technique needs to be improved from
a human factors point of view to make it easier to
use, and we would like to make it so that there is not
too much dependence on the uprater.
Right now we have a very smart guy doing
the evaluations and he uses his knowledge, along with
what he has programmed to come up with the right
answers.
If the industry uses it, they we have to
do a bit more work in making it more user friendly.
And it may take some time, but it is not a huge jump
at this point in the technology.
I think at this point that I am finished
with the remarks that I had planned on making, and
unless you have questions, we can turn it over to
Chris Boyd.
DR. POWERS: Are there any more questions
on the metallurgical aspects of the problem? Seeing
none, we will turn to some aspects of the thermal-
hydraulics issues, and some results that I think the
Committee will find interesting on some of the
progress that research has been making in the area of
computational fluid dynamics.
We have commented several times in our
research reports that we thought that this was an area
that the agency could use and would profit, and the
speaker will give us some idea of the progress that
they are making. Thanks, Joe.
MR. BOYD: My name is Chris Boyd and I am
going to talk to you today about the Division of
Systems Analysis and Regulatory Effectiveness Programs
in this area. And I am going to give just a quick
overview of the entire division activities, and then
focus in on some of the CFD work that has been done.
This overview was given in more detail
last week. So based on recommendations from the ACRS
subcommittee, looking at the DPO, there were two areas
that our division is focusing on, and these are in the
action plan, Items 3.4 and 3.1.
Item 3.4 has to do with just developing a
better understanding of the behavior of the tubes
during these severe accident conditions. I am going
to go into a little bit of that.
And then Item 3.1 is evaluating the
potential for damage due to rapid depressurization,
such as a main steam line break. So in Item 3.4,
looking at the tubes in the severe accident
conditions, the major components of this research in
our division are system level code analysis. It is
the SCDAP/RELAP work.
And that is under way now, and there is a
report that is just finishing up that is covering a
lot of sensitivity studies and plant design
differences, and things along those lines, using the
workhorse code for the thermal-hydraulics.
And then we are looking with computational
fluid dynamics at the inlet plenum mixing in
particular, and trying to enhance what we understand
of that mixing, looking or starting from the test data
that we have, and then trying to enhance that, and
that is what I am going to talk about today.
And then there is some additional
assessment of the 1/7th scaled data, and we are
looking at new experimental data, and possibilities
for that. As far as the rapid depressurization goes,
and its impact on the tubes, this work is scheduled
for completion at the end of the next calendar year.
They are looking at the pressure loads on
support plates and tubes and flow induced vibrations.
This work is just now in the early formulation stages,
and it was presented last week. But again this was
just some preliminary thoughts.
Now I am going to focus in on some of the
work. This is one of the task items, which was the --
DR. POWERS: Chris, going back to the 3.1
task, I understand that it is just in the formulation,
and right now that formulation is focusing on
analysis.
The complaint or the issue that comes up
in this is that it is a very difficult analysis to do.
Most of our codes have never been designed to get
these kinds of vibrations and dynamic effects.
Are we at a place now where we can start
talking about what kinds of experimental data would be
needed to validate these codes. I understand that it
is iterative with the kinds of things that the
previous speaker was discussing on what the magnitude
of the loads is.
But can we talk about the types of
experiments, or is that down the road a ways?
MR. BOYD: I think it is down the road a
little ways. They are still evaluating what code they
might use. So ar far as benchmarking the code if you
don't have the code yet, and as far as early ideas of
doing hand calculations --
DR. POWERS: I had a little hope for hand
calculations in this field. It's just that it is a
tough calculation to do without having some
experimental validation to have any confidence in what
you got.
MR. BOYD: I would agree. Most codes fall
into that category.
DR. POWERS: Except for chemical codes,
and that goes without saying that they are all correct
or all wrong.
MR. BOYD: Well, I will give you a quick
overview, and I will cover this with the slides. But
what we are trying to do is enhance our understanding
of the inlet plenum mixing. I am showing here that
thermal hydraulics reacted to this particular severe
accident scenario.
This is where the core is uncovered, and
where it is single-face steam, and the loop seal is
plugged. So flow through the hot leg goes through
part of the tubes and into the outlet plenum, and it
comes back through the remaining tubes, and mixing in
the inner-plenum, and this counter-current flow sets
up in the hot leg.
This is modeled with SCDAP/RELAP, a lump
parameter code. One of my slides has a noting diagram
for that. This areas of interest that I show in the
SCDAP/RELAP code is essentially three nodes, and they
nodes have fixed mixing parameters that are set based
on generally the 1/7th scale experiments that were
done in a Westinghouse type facility.
So the idea Here is to look at those
mixing parameters and look at the effect of different
things on those mixing parameters. For one with CFD,
we want to go full-scale, full-pressure. We want to
look at the effect of a leaking tube.
We want to look at the effect of different
inlet geometries, and that is what we are trying to
accomplish. The first step will be to benchmark the
CFD code though to see if it can pick up the right
behavior, and whether it is a useful tool. And that
is what we are going to talk about.
But again the background here is that the
thermal-hydraulic predictions ultimately come from
SCDAP/RELAP code, ad=nd that is our workhorse code.
The tube temperature predictions that come
out of that are going to be influenced by these fixed
mixing parameters, and these fixed mixing parameters
come from a limited set of experimental data.
DR. WALLIS: You had a picture that you
just showed of a steam generator with a flow going in
one direction with some of the tubes, and then the
other direction and the other tubes?
MR. BOYD: Right.
DR. WALLIS: This sort of situation is
usually prone to historesis, but it depends on the
past history which one is going where, and you can't
just look at it and say 50 percent of the tubes are
going one way and 50 percent of the tubes are going
the other.
And maybe because of past history, you
have got something of a 70-30 distribution or
something. I think it is not so easy to know how to
set up the program.
MR. BOYD: We are not setting it up and
specifying which tubes are in an up and down flow.
DR. WALLIS: Well, then how you set it up
initially, and then it evolves into something.
MR. BOYD: It evolves into something. I
will say that from at least the 1/7th scale that they
interrupted the entire flow pattern by opening a valve
at the pressurizer line.
Their experience was that the overall flow
pattern, and the number of tubes set back into its
condition fairly quickly after shutting the valve
again.
If you can imagine that you have got half
of them going one way, and the other half going the
other way, and let's say you want to go to, say, 48
percent and 51 percent.
How do those ones that were going this
way, how do some of them decide then that I have got
to reverse? It is not so obvious how that happens.
I would assume that some are just teetering on the
edge and ready to go either way.
MR. MUSCARA: Well, I would assume that
some would be obviously teetering on the edge. The
assumption would be nearly stagnant given that there
are up flow and down flow.
DR. WALLIS: Well, in a continuum like
that it may be earlier to handle.
DR. SHACK: I remember at the subcommittee
meeting that you said that you diddled the conditions
at the core, and you always managed to sort of set up
a kind of a stable profile, and your counter-current
flow.
MR. MUSCARA: That's right.
DR. SHACK: Did that affect your fractions
at all? I mean, when you changed that, did you --
MR. BOYD: No, changing those conditions
didn't really impact the fractions. That was a very
-- it did not impact it significantly. What happened
was that there were other parameters to change the
impact in a much greater sense.
So I found that to be a very limited
impact and that gave us confidence in setting those
boundary conditions, and that was the point that I was
making.
MR. BOYD: All right. I talked about the
background, and the bottom line is that these
SCDAP/RELAP bins are relying on mixing parameters
which come from 1/7th scaled data which we are going
to try and bolster the confidence in that data.
Why use computational fluids. Well, it is
less expensive than experiments. We can go to full
scale and full pressure. We are going to have a
direct resolution of mixing. We are not setting fixed
mixing parameters.
And we will then be able to extend our
data within the MHTGR effects and tube leakage effects
in a much wider number of variations than we could in
an experiment.
DR. WALLIS: Well, let's go back to this
again. The mixing is dependent upon the flow in the
tubes, and presumably you have jets coming out of some
tubes and so on. So the level of turbulence in the
lower or left plenum is a function of how the flow is
coming out of the tubes isn't it?
MR. BOYD: If I say no to mixed mixing
parameters, we are not specifying the flow in the
tube. We are letting the equations --
DR. WALLIS: But the equations don't model
many tubes. They just lump them together. If you
limp them together, you get a very different --
MR. BOYD: Well, at 1/7th scale, we have
got a tube for every tube. There are only 260 tubes
here and I am modeling every one of them as an
individual channel.
DR. WALLIS: And in the real generator,
you can't do that.
MR. BOYD: I have my ideas on tube
modeling if you want to go into that.
DR. WALLIS: Well, at least you realize
that you have to do it.
MR. BOYD: Just to show a quick flow
physics comparison. This is really if people
understand this, but on the CF approach the hot leg,
what I am showing is that we are predicting the direct
counter-current flow; and in the SCDAP/RELAP
calculations, you have to set up before the run, and
run two pipes, and one carries one fixed temperature
flow in, and the other carries a fixed one-dimensional
flow back.
In the inlet plenum, in computational
fluids, we are modeling the rising buoyant plume, and
letting the turbulent mixing happen, and the lump
parameter code, we have got these three volumes, with
fixed flow co-efficients to set up the mixing as you
see on the three blocks on the right.
And then as far as the tubes go, we have
got the advantage where we are going to directly
predict the number of tubes in up flow where in the
SCDAP/RELAP runs that is a fixed parameter again, and
we had basically one set of tubes all the same
temperature going up, and one set of tubes, all the
same temperature, coming back.
So we will also be able to get the
temperature variation, tube to tube variations, which
could give us some insights.
DR. WALLIS: Again, in CFT, usually you
have this K esplon model or something for turbulence,
which was not developed for these conditions, and
density stratification we know dampens out turbulence.
MR. BOYD: Right.
DR. WALLIS: So you need a different model
if you are going to do it right.
MR. BOYD: What I did was that I looked at
several turbulence models. I used K-Esplom because it
is fast for a while, but I didn't want to present that
here, just because I knew that you guys would pick on
me if I produced K-Esplon results.
But I looked at a few other turbulence
models, and they didn't make as big an effect as I
thought.
DR. WALLIS: Did they take account of the
Richardson number type thing, the buoyancy and killing
the turbulence?
MR. BOYD: Yes, there are options for
that. The one that I ended up with was a Reynolds
Stress Model, and so a second order model, modeling
each of the components.
DR. WALLIS: Was buoyancy in it or it
doesn't does it?
MR. BOYD: Well, that is implicitly.
Buoyancy is implicitly in every one, because I have
got variable properties and gravities, and all the
bells and whistles were on.
But I was surprised in that it did not
make as big of a difference as I might have suspected.
It did actually look better in some areas, the second
order of turbulence though.
DR. WALLIS: It certainly looks far better
than lumped parameters.
MR. BOYD: A little better than lumped
parameters. So CFD is going to provide an improved
understanding at 1/7th scale, and when I say improved
understanding, we have got data, but we have got
limited data.
There is just a handful of thermal-couples
no velocity measurements, and mass flows are inferred
from external energy balances. So with CFD, you can
fill in some of the gaps, and you can take a look at
the errors that you might or the uncertainty that you
might get by just measuring with four thermal-couples.
And assuming that you have measured enough
in the hot leg to understand the full profile, and
things along those lines.
DR. WALLIS: It is not a predictive tool
then, and you just use it to understand data?
MR. BOYD: Well, in this case, I am saying
one of our action items was to assess the 1/7th scaled
data. So this is a tool that helps us assess that.
At this point though we are not really
interested in this 1/7th scaled data so much as we are
the full scale data. And at this point, I am showing
results against 1/7th. This was our first step.
DR. WALLIS: And you eventually want to
predict full scale?
MR. BOYD: That's right, and we are
working on that now and starting to build the models
for full scale. Then extending the full scale, the
first question --
DR. WALLIS: Full scale means predicting
full scale without experiments?
MR. BOYD: That's right. We will be
making predictions at full scale, that's right, down
to the third decimal place.
Does scale affect the mixing parameters,
and that is our first question. And at full scale,
that is where I am going to look at the effect of tube
leakage, and the effect of inlet geometry variations,
as opposed to doing them at the small scale, where we
are not really interested.
And then we can look at tube to tube
variations. The schedule and the approach validate
the technique to see if it is a valid approach, and
with 1/7th scaled data we have done that.
Extend the predictions to full scale, and
that is starting up now, and then complete these
additional studies, and that is next July. So we will
take a quick look at what we did at full scale.
We have got two measurements at full
scale. This is the course mesh and the other mesh is
actually still running several million cells. This is
about a million cells for a hot leg and steam
generator. They are about 5 centimeters a piece.
And cutting that down to a 2 centimeter
cell case, which is what is running now, makes the
mesh obviously a lot tighter, and a lot more
cumbersome. But that gives you an idea of what the 3-
D mesh of each of the tubes being modeled.
I look at the results in two ways,
qualitatively and quantitatively. Qualitatively, they
gave a handful of observations from the test, and
things such as the flow coming in to 60 percent of the
hot leg, and exiting the hot leg covering about 25
percent of the hot leg area, and sloping interface, of
course.
This was picked up in the test. The plume
or the rate of drop off of temperature in the plume
roughly matched what was predicted. The temperature
dropped through the tubes. Again, similarly matched
all the counter current flows.
So on a qualitative basis, I am just
saying that we picked up the global flow phenomena
that was observed and could be talked about in that
test series.
Quantitatively, when we go down to the
parameters we are interested in to put into
SCDAP/RELAP, this is one of the tests that we ran.
This is actually the worst set of results. We do have
a little bit better, but these aren't that bad.
Generally, I would say they are within 10
percent on most of the parameters. We have
overpredicted in this case a number of hot tubes, the
number of tubes in upflow by approximately 10 percent
of the tube sheet.
That is probably one of the larger errors,
and that is something that I am looking at with
sensitivity studies on the tube model. As far as
things like mixing fraction and recirculation ratio,
which are direct parameters that we care about int he
SCDAP/RELAP runs, we generally are within 10 percent.
And in this particular case, there is a
model in the recirculation ratios that is as high as
15 percent off. Given the data and the limited
measurements that were made, I would say that we are
within the uncertainty of the data on this.
So in general I am saying that the code
has done a decent job. We are getting the big
picture, and I think it is going to be a useful tool.
And when I told you that we were about 10
percent of the tube sheet off in the upflow, this is
a direct picture of that. The outside dotted line is
the fluent prediction of the boundary between up flow
tubes and down flow tubes, the inner two lines are the
boundary based on the data.
Now, that is a band because they didn't
have every tube monitored. So somewhere within that
band, is where the region of upflow and down flow
shifts.
They reported the outer band, but I put in
the inner band because that was just as likely in my
mind.
DR. WALLIS: Westinghouse gets fewer --
MR. BOYD: That's right. So the results
of the validation, generally I am saying that we are
within 10 percent of the Westinghouse 1/7th scale. We
have picked up the global flow parameters that we
expected.
And I think given the uncertainty in the
experimental data, we are doing very well here. And
work on full-scale predictions is under way. And I
think I am repeating myself here.
So I am saying that the CFD technique has
been demonstrated to be applicable for the prediction
of these mixing parameters, and for this kind of
counter-current flow situation.
This work provides a level of confidence
that CFD can be used to go beyond the experimental
data to conditions not explicitly covered, such as a
little variation in the inlet conditions or a
variation in the height of the hot leg relative to be
the tube sheet.
And further analysis as planned that full
scale, with tube leakage, and these geometry effects,
and other sensitivity studies.
DR. POWERS: It is just an exciting
progress and I look forward to what comes out of it.
It is a new level of understanding of what is going on
in these flow calculations, and it would be delightful
to have those variations that you are talking about,
because that has been the subject of endless amounts
of speculation and hand-waving, and it would be nice
to have some reliable calculational results in that
area.
DR. WALLIS: You said there was a run
running now? I just wondered if it will be over
before we leave on Saturday.
DR. KRESS: Actually, it won't.
DR. POWERS: That was the right answer;
whether it was going to be done or not. All right.
Any other questions on the thermal-hydraulic aspects
of the action plan?
It is my impression that the staff has
bent over backwards to respond to our recommendations,
and it is my impression that we have helped the staff
in acquainting the Commission with the fact that they
need to fund research in these areas if they want the
level of understanding that they would like to have.
And I see us producing an improvement in
the state of the art in many areas, and certainly in
the metallurgical areas, and then from this I see in
the computational fluid dynamics areas.
So I think that truthfully this has been
win-win situation in producing this review. Thank you
very much. I will now turn the meeting back over to
you, Mr. Chairman.
CHAIRMAN APOSTOLAKIS: With that, we will
take our break, and meet again at 3:20.
(Whereupon, the meeting was recessed at
3:05 p.m., and was resumed at 3:20 p.m.)
CHAIRMAN APOSTOLAKIS: The last
presentation of the day is on Proposed Resolution of
Generic Safety Issue 173-A, Spent Fuel Storage Pool
for Operating Facilities, and Dr. Kress, you are up
again.
DR. KRESS: I was looking through David's
slides, and he goes into the background pretty well,
and so there is no need for me to make the background
statements that I was going to make. I will just let
him make them. So I will turn it over to him, and let
him go ahead.
MR. DIEC: I appreciate that, Dr. Kress.
Good afternoon. My name is David Diec, and with me
today is Steve Jones from the Office of NRR. We also
have the technical staff who are sitting in the
background and who are available to answer any
questions that you may have.
For today's presentation, I will be
discussing the purpose of the presentation itself, and
go over a little bit of background of what GSI-173A is
about.
And the staff evaluation of plant specific
issues. Steve Jones will then discuss with you the
basis for the closure of this GSI and make a final
conclusion.
The purpose of the presentation today is
two-fold. We are addressing the recommendation that
you made in the June 20th letter, and certainly we are
seeking for your closure letter on this issue, as it
has been a long time since we last talked to you.
The information that is being discussed
today has been presented to you so many times in the
past. We came before you to discuss about our action,
task action plan earlier, as early as 1994, and to
present our findings in 1996, and to as recently as
last year, to discuss about our proposed resolution
for a plant specific issues.
In the June 20th letter, the Committee
raised a concern whether the screening criteria that
we used were appropriate for potential plant specific
evaluations for spent fuel action and its risk at
operating facilities.
The impetus of the concern was that the
criteria in the Reg Guide was derived from prompt
fatalities at facilities, and it is probably
appropriate for an operating reactor source because it
was driven by a steam oxidation condition.
However, for the spent fuel pool accident,
and you may be looking at the source term is different
and involved with a large amount of releases of
Ruthenium and fuel fines.
As a result of that, the Committee made
the recommendation that we defer closing out this GSI
until the technical study that we conduct on
decommissioning plants is complete, and consider
developing appropriate screening criteria for
regulatory analysis of the spent fuel pool accidents
at other facilities.
In way of background, this GSI was a
genesis of the report filed under Part 21 in 1992 by
two contractor engineers who performed work at the
Susquehanna plants.
These engineers contended that the
Susquehanna spent fuel pool failed to meet regulatory
requirements with respect to sustained loss of pool
cooling function after the loss of off-site power or
a loss of LOCA event.
And they sustained that boiling could
cause failure of equipment necessary to mitigate
accidents, and to safely shut down the plants. That
was the genesis of that.
In 1993, we formulated a task action plan
to resolve issues associated with the spent fuel pool
storage, and to ensure the reliability of the decay
heat removal capability, and maintenance of the
inventory in the pool.
The task action plan was completed in
1996, and we briefed you of our results, and at that
time after the completion of the briefing, the
Committee asked us whether or not we would like
anything; i.e., a response letter.
In hindsight, we should have said yes.
But it seemed unnecessary at that time, and so that
leads us to today's presentation. The task action
plan looked at three generic areas; coolant inventory,
and the ability to provide and maintain inventory in
the pool; and to detect temperature and fuel
reactivity; and the ability to maintain fuel in the
sub-critical and the bar flex integrity in the spent
fuel racks.
Just a note on the fuel reactivity. We
addressed this issue separately as part of the generic
letter in the 96-04 response. So it was not
considered in this broader scope.
In doing the implementation of the action
plan, we visited a number of plants, and we reviewed
plant specific design features that addressed these
two areas, the coolant inventory and coolant
temperature, and also the reactivity issue a little
bit.
And we concluded that these plants
conformed to the current regulations. However, we
identified a number of plant specific issues that need
further regulatory evaluation; and in June of last
year, we came before you to discuss specifically how
we would resolve those.
I am going to go over the probabilistic
assessment that we used at that time. We performed
the probabilistic assessment of these issues by first
conducting a screening analysis, using plant specific
design and operational information.
And two endstates were chosen to test the
design features under the evaluation for inventory
endstate corresponding to a loss of coolant to within
a one foot level above the top of the spent fuel pool
rack was used.
For issues involving boiling, an endstate
corresponding to a sustained boiling in a pool for
greater than 8 hours was used. The endstates
represent conservative points in the sequence, where
public health and safety was assured.
DR. BONACA: If I could ask a question and
if you could go back a moment to that, and you are
talking about consistent with regulatory guidelines,
and it seems as if there is one set of requirements to
which all these power plants are adhering to.
And in the license renewal, we have seen
a lot of different requirements, especially for older
plants, that seem to have more or less regulatory
requirements imposed on the equipment.
So the question that I am asking here is
this implied that it is just one set of requirements
and applying to all pools, such that you can perform
a generic PRA screening analysis?
MR. DIEC: We look at the number of plants
and we choose a specific group of plants that
represent specific plant issues, using this criteria.
So I am not sure it is a blanket approach in this
case.
DR. BONACA: Well, let me give you an
example. We have some plants where we were told that
probably none of the cooling equipment in the spent
fuel pool is in their design basis.
Therefore, they only commit to maintaining
the liner of the pool and the emergency injection and
both of them are self-degrade and they are in the
scope of the license renewal.
And we were left with some surprises on
that. We are reviewing now an application where the
cooling system is in the scope of license renewal, and
therefore the heat exchanger and pumps, et cetera, is
in the scope.
So we have a sense that there is a very
different regulatory requirements applying to
different plants, and that's why I was asking these
questions about this generic analysis.
MR. JONES: That in essence was the
genesis of this issue. Most of the design features
that we evaluated were called for in later design
guidance, but we are not implemented at a lot of the
earlier plants because their construction permits
predated that guidance.
Things such as -- well, like safety
grades, spent fuel pool cooling system, and seismic
makeup lines, and things of that nature. So the
purpose of the evaluation was to determine if we were
justified in imposing backfill requirements to upgrade
those systems.
DR. BONACA: So actually what I am raising
here is it was the center of this issue, and you are
going to talk about that.
MR. DIEC: After performing the analysis,
we concluded that if we can see the loss of the spent
fuel coolant events were less than 1E minus 6 per
year, and there was no regulatory action justified for
these plants, group of plants.
The frequency of suspended boiling was
found between 1E minus 6 per reactor year, to 1E minus
5 per reactor year. We conducted a further evaluation
of those plants and concluded that no action was
justified.
The evaluation that we looked at took into
consideration the licensee's voluntary actions to
modify their plant piping system designs to install
additional low level alarm in the pool, and a switch
that will give an indication in either the control
room or at the local station.
And also beefing up their operating
procedures to make sure that the operator is aware of
the onset of the loss of cooling events.
DR. BONACA: So the first bullet is
applicable to all power plants?
MR. DIEC: All power plants that fall into
the loss of cooling events.
DR. BONACA: And that includes the spent
fuel pool coolant event in their licensing basis. I
am trying to understand what that means.
MR. JONES: These were essentially
screening analyses of certain plants that had
vulnerabilities to loss of coolant events because of
the way their spent fuel pools were configured or the
reliability of their makeup systems.
And in the second case there were
vulnerabilities with the spent fuel pool cooling
system. Like it did not have an on-site source of
power available to the pumps and heat exchanges.
DR. BONACA: So the conclusion for the
first question was that in no case -- I mean, the
frequency of loss of spent fuel pool coolant was 1E
minus 6 for all plants?
DR. KRESS: I interpreted that to mean
that these individual plants would make a judgment,
and if their frequency was that, then no action was
required. If it wasn't, then you would have to go
further with it.
DR. BONACA: So this was the conclusion.
All right.
MR. JONES: We made an effort to select
the most vulnerable plants and evaluate them on a
plant specific basis the probability of these two
events or two endstates.
And it was a bounding assessment for all
the plants that had some of the design features that
we were evaluating.
DR. ROSEN: And a definition of the
endstate is within one foot above the top of the rack?
MR. DIEC: Correct.
DR. ROSEN: And this was based on an
analysis of operational experience?
MR. DIEC: Yes.
DR. ROSEN: You are saying that the
frequency has been less than 1E to the minus 6 based
on operating experience?
DR. KRESS: No, it was calculated --
MR. LEE: This is Sam Lee. I was the risk
analyst that worked on this, and Dr. Bonaca, you are
correct. It is per plant frequency estimation, and
this is all a calculated number based on operations,
per se.
DR. BONACA: Thank you.
MR. LEE: Well, it is based on operations
as is, but it is all based on risk analysis, or what
I would call a probablistic analysis, because the
endstate that we had used was one foot above the
stored fuel.
DR. WALLIS: You have to have millions of
years of experience, and it has got to be calculated.
MR. HUBBARD: This is George Hubbard with
the plant systems branch, and I just wanted to clarify
one thing. This was the risk analysis that was done
by the staff and by the contractor through site
visits.
They went out and they looked at the
issues, and we performed the screening analysis that
we are talking about here. It was not a licensee, per
se. It was that we got a contractor and went out and
did the work ourselves.
DR. ROSEN: What does the operating
experience tell you? We have had a hundred plants
running for 20 or 30 years, but we have never had one
of those obviously.
MR. JONES: Right. None of these
endstates has been experienced in the industry to
date. There have been long term loss of coolant
events, on the order of 24 hours, or more, but those
were at times when the heat load in the spent fuel
pool was such that -- well, loss of coolant events
have been relatively limited.
We did look at issues such as what
occurred at Connecticut Yankee, and also we considered
the freezing event at Dresden, which was resulted in
Bulletin 94-01.
And although it didn't directly involve a
spent fuel coolant system, it was identified at that
time that there was a vulnerability to a rupture in
the spent fuel coolant transfer line that could have
drained the spent fuel pool at Dresden Unit 1. But to
answer your question though --
DR. ROSEN: That no operating experience
would contradict this conclusion.
MR. JONES: Right.
MR. HUBBARD: This is George Hubbard again
from Plant Systems. Right in this same time period,
AEOD did a study that looked at operating experiences
for temperatures and of the height or the level of the
water level, and the water level dropped.
I was just glancing through here as you
were talking, and it looks like there were a few in
which the duration of the loss of coolant was like
three in greater than 24 hours, in which the loss of
coolant lasted for longer than 24 hours.
As far as the temperature increase, there
was one time where it got to 50 degrees, and this was
looking at our operating experience, and when we were
doing this, we were aware of this data as this report
was in early '97, and we completed the plant specific
backfits.
And a report to the Commission went up in
September of '97. So the operating experience was
available to us, and I have not found the exact
probabilities that they used from the AEOD experience.
But that was taken into account, the
operating experience, and your comment that the
operating experience doesn't contradict this is true.
DR. KRESS: Let me ask you about your two
screening goals. It basically implies that if you
meet these goals that you would meet the Commission's
safety goals, and there is an order of magnitude of
difference between these two, and the second one still
meets the safety goals because during that time period
you have operator action that could mitigate it so
that you could get another factor of 10 out of that.
MR. JONES: Right.
DR. KRESS: Is that a correct
interpretation?
MR. JONES: Yes.
MR. DIEC: We take into consideration a
voluntary action as well and operator actions in this
phase.
MR. JONES: These were essentially
selected as more easily modeled endstates.
DR. KRESS: Oh, yes. If you go any
further than that, then you have lots of physics. And
if you don't meet these, there is no reason to go
ahead.
MR. JONES: And also the difficulty in
assessing the probabilities of recovery when the time
frames get out to an order of a day or more.
DR. BONACA: So here the criteria is 8
hours? So as long as you can recover before 8 hours
of boiling that's a success?
MR. JONES: Right. If it is extended
beyond 8 hours, then it is a questionable recovery.
MR. LEE: This is Sam Lee again. If I may
add, the reason behind estimating the frequency of
sustained blowing, if you remember back in the early
'90s with the Susquehanna situation, where a boiling
pool could affect the ECCS of the other units.
And that was the basis for why we
conducted this analysis. After we did the analysis,
or when we went to the plant as George said during the
analysis, we found or discovered the physical layout
of the plant was such that it was not like
Susquehanna.
That even if you had sustained boiling,
that the steam environment would not impact the ECCS
of the other units. So it really became a non-issue
for us at that point.
CHAIRMAN APOSTOLAKIS: Do you remember
what the dominant contributors to these were? Were
they seismic events?
MR. LEE: Yes. A seismic event is one of
the major contributors, yes.
DR. ROSEN: When we talked about
decommissioning plants, we talked about loss of spent
fuel pool coolant events, and we were told at that
time that the two most likely sequences to get there
were seismic and sabotage.
MR. LEE: I will talk about one particular
plant that we have data for, and the loss of inventory
-- and this is through a break in the pipe -- was one
of the dominant contributions, and the other one is
the earthquake.
CHAIRMAN APOSTOLAKIS: Okay. Why don't we
go on.
MR. DIEC: Well, at this time, I am going
to turn it over to Steve Jones, who will discuss the
basis for closure.
MR. JONES: As you are aware the staff
completed a study of the events of the decommissioning
of plants, and that study established a pool
performance guideline regarding the frequency of
events leading to a potential fire in the spent fuel
pool or really fuel uncovery at 1 times 7 to the minus
5 for reactor year.
Using a source term that included
consideration of large releases of Ruthenian and fuel
fines, the staff evaluated consequences for a pool
with approximately 30 days of decay from the most
recent discharge.
That study demonstrated that the
quantitative health objectives were met for then
frequencies below the plant performance guideline,
both in terms of prompt fatalities and latent cancer
fatalities.
Since the screening criteria that were
used in the plant specific studies were more
conservative than the pool performance guideline of
the decommissioning study, we considered that that
would demonstrate that the quantitative health
objectives were met for all the plant specific issues
that were evaluated.
DR. KRESS: Let me ask you a simple
question about that. The previous study with the
NUREG 17-38 was dealing with decommissioning plants
that didn't have a reactor there.
MR. JONES: That's correct.
DR. KRESS: So only the risk associated
with the suppression pool had to be met in order to
meet the safety goals. Now you are talking about a
subset of sequences at an operating plant.
And it is not appropriate to say that the
subset has to meet the safety goals, because safety
goals have to be met by the pool summation of all the
sequences.
So when you say it is an order of
magnitude less than this screening criteria, I
consider that appropriate, an appropriate choice. It
is not overly conservative. It is an appropriate
choice for a subset of sequences.
So I would not want to see you go up to 10
to the minus 5, for example, and say we will use that,
because this is for an operating reactor, and you have
to keep it down so that it adds in to the other risks.
That was just an observation and a comment.
MR. JONES: I understand your point. We
concluded that additional screening criteria were not
necessary and that as Dr. Kress mentioned the absolute
frequency values were roughly on the order of a
magnitude lower.
And in addition the endstates were
substantially more conservative, and the pool
performance guidelines looked at a complete loss of
inventory in the fuel pool, and our endstates were
stopped at either sustained boiling or loss of a
significant inventory, but still the fuel being
covered. That concludes our briefing.
DR. KRESS: Are there any comments or
questions that the Committee has? Seeing or hearing
none, I think you made your case clear, and we will
have a letter on the subject. And you brought us back
on schedule, and thank you for the presentation. It
was short and sweet and to the point, and we
appreciate it.
CHAIRMAN APOSTOLAKIS: All right. Thank
you.
(Whereupon, the meeting was recessed at
3:50 p.m.)