ML18029A300
| ML18029A300 | |
| Person / Time | |
|---|---|
| Issue date: | 12/05/2017 |
| From: | Zena Abdullahi Advisory Committee on Reactor Safeguards |
| To: | |
| Abdullahi Z | |
| References | |
| NRC-3413 | |
| Download: ML18029A300 (104) | |
Text
Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION
Title:
Advisory Committee on Reactor Safeguards Thermal-Hydraulic Phenomena Subcommittee Open Session Docket Number:
(n/a)
Location:
Rockville, Maryland Date:
Tuesday, December 5, 2017 Work Order No.:
NRC-3413 Pages 1-85 NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers 1323 Rhode Island Avenue, N.W.
Washington, D.C. 20005 (202) 234-4433 OPEN TRANSCRIPT With BACKGROUND OPEN SLIDES
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 1
1 2
3 DISCLAIMER 4
5 6
UNITED STATES NUCLEAR REGULATORY COMMISSIONS 7
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 8
9 10 The contents of this transcript of the 11 proceeding of the United States Nuclear Regulatory 12 Commission Advisory Committee on Reactor Safeguards, 13 as reported herein, is a record of the discussions 14 recorded at the meeting.
15 16 This transcript has not been reviewed, 17 corrected, and edited, and it may contain 18 inaccuracies.
19 20 21 22 23
1 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION
+ + + + +
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS)
+ + + + +
THERMAL-HYDRAULIC PHENOMENA SUBCOMMITTEE
+ + + + +
OPEN SESSION
+ + + + +
TUESDAY DECEMBER 5, 2017
+ + + + +
ROCKVILLE, MARYLAND
+ + + + +
The Subcommittee met at the Nuclear Regulatory Commission, Two White Flint North, Room T2B1, 11545 Rockville Pike, at 1:00 p.m., Jose March-Leuba, Chairman, presiding.
COMMITTEE MEMBERS:
JOSE MARCH-LEUBA, Chairman RONALD G. BALLINGER, Member DENNIS C. BLEY, Member MICHAEL L. CORRADINI, Member
2 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 WALTER L. KIRCHNER, Member DANA A. POWERS, Member JOY REMPE, Member DESIGNATED FEDERAL OFFICIAL:
ZENA ABDULLAHI ALSO PRESENT:
JOSH BORROMEO, NRR CHRIS HOXIE, RES ANDREA VEIL, Executive Director, ACRS AARON WYSOCKI, ORNL PETER YARSKY, RES TAREK ZAKI, RES
- Present via telephone
3 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 CONTENTS Staff Preliminary Analysis of KATHY ATWS-I Temperature Data Peter Yarsky.................................5 Discussion ACRS Subcommittee Members....................64 Adjourn Jose March-Leuba.............................85
4 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 P-R-O-C-E-E-D-I-N-G-S 1
1:03 p.m.
2 CHAIRMAN MARCH-LEUBA: The meeting will 3
now come to order. This is a meeting of 4
Thermal-Hydraulic Subcommittee of the Advisory 5
Committee on Reactor Safeguards. I am Jose 6
March-Leuba. ACRS Members in attendance today are Ron 7
Ballinger, Dana Powers, Michael Corradini, Dennis Bley, 8
Walter Kirchner, and Joy Rempe. Zena Abdullahi is a 9
designated federal official for this meeting.
10 The research staff will brief us on the 11 preliminary assessment of the experiments conducted 12 of the KATHY thermohydraulic test facility. The main 13 body of this experimental data is to aid in the 14 development of the methodology to predict failure to 15 the weight and then assuring temperature excursion 16 during hours of instability power oscillations.
17 This subcommittee has heard proprietary 18 presentations by vendors on this same topic. And today 19 we're looking forward to hearing about the KATHY 20 experiments and the research preliminary assessments 21 of the results. And I understand in process on the 22 preliminary.
23 MR. YARSKY: Yes.
24 CHAIRMAN MARCH-LEUBA: Part of this 25
5 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 presentation will be closed to the public to allow 1
discussion of proprietary matters. We have one open 2
bridge line arranged for interested members of the 3
public to listen in. A separate and closed bridge 4
number is available for NRC staff.
5 We have received no written comments or 6
requests for time to make oral statement from members 7
of the public regarding to this meeting, but as 8
customary, the line will be open at the end at the open 9
section of the meeting to allow for spur of the moment 10 comments.
11 As the meeting is being transcribed, I 12 request that participants use the microphones located 13 throughout this room when addressing the subcommittee.
14 Participants should first identify themselves and 15 speak with sufficient clarity and volume so that they 16 can be readily heard.
17 Let me remind you now to please ensure that 18 all your devices have been placed in silent mode to 19 minimize disturbances within the meeting. We will now 20 proceed with the meeting and call upon Peter Yarsky 21 to assess the initial open portion of the meeting.
22 Go for it.
23 MR. YARSKY: Thank you. I'm Dr. Peter 24 Yarsky from the research staff. And before getting 25
6 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 into a discussion of the KATHY experiments, I wanted 1
to take a few moments to provide some background 2
information on MELLA+ ATWS instability and the 3
predicted failure to rewet phenomenology.
4 In this presentation I'll provide a brief 5
description of the MELLA+ operating domain, which I'm 6
sure many members are very familiar with. So I'll try 7
and do this very briefly. And if at any point this 8
gets very boring, you can ask me to speed up and skip 9
a number of slides. But I wanted to make sure that 10 everyone had some background before we started talking 11 about the experiment itself.
12 This includes a quick discussion of the 13 safety significance of the MELLA+ domain and what we 14 mean when we say ATWS with instability. We'll talk 15 about trace calculation results that have informed 16 recent license amendment requests reviews, as well as 17 what the NRC identified as a predicted mechanism for 18 potential fuel heat up during ATWS-I scenarios from 19 MELLA+ BWRs. And lastly, I'll touch on some 20 considerations that are important for doing plant 21 specific evaluations.
22 The MELLA+ domain represents an expansion 23 of the operating domain relative to extended power 24 operate for BWRs, and notably allows operation at high 25
7 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 power levels, 120 percent of the original license 1
thermal power level. But at low core flow rates about 2
80 percent of the rated core flow rate.
3 This operation at high power to flow ratios 4
introduces new aspects to the progression of 5
anticipated transient without scram events for BWRs 6
operating in this domain. In particular, ATWS events 7
are mitigated by an automatic trip of the dual 8
recirculation, a dual trip of the re-circulation pumps, 9
which causes a power and flow decrease of the pumps 10 run down.
11 The power in the reactor then increases 12 as a result of a reduction of feed water temperature.
13 When the core flow rate is low, this dual 14 re-circulation pump trip mitigating feature is less 15 effective and this means that there's a higher power 16 level in the core following the trip of these pumps.
17 What this means in the power flow operating 18 domain is that a plant operating at originally license 19 thermal power, which would be the point at 100 percent, 20 100 percent on this power flow map. If that plant were 21 to experience a ATWS and a dual re-circulation pump 22 trip, it would follow the black curve in terms which 23 trajectory on the power flow map and would wind up at 24 a given power flow, can I show with the pointer here?
25
8 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 No, I'm not able to do that.
1 CHAIRMAN MARCH-LEUBA: You have to use the 2
mouse to do that.
3 MR. YARSKY: Yes. I'm trying to use the 4
mouse, but it's not showing up on the screen. But 5
there's two points indicated at the end of the black 6
curve versus the red curve. If a plant were operating 7
in the MELLA+ domain, its evolution during this event 8
would result in a much higher power to flow ratio.
9 If we were to look at those points relative 10 to stability, what one would see is that the progression 11 of the event of an ATWS with a dual re-circulation pump 12 trip initiating from the MELLA+ corner would wind up 13 at the red point marked MELLA+. And at this point the 14 reactor is predicted to be much more unstable meaning 15 that it would cross the stability boundary much earlier 16 in the event.
17 And this can lead to what we refer to ATWS 18 wimp instability. So there's an ATWS event, and during 19 the progression of that event the reactor becomes 20 unstable and large amplitude power and flow 21 oscillations are observed.
22 MEMBER CORRADINI: Just, can you go back.
23 Just to be sure, go back a slide. The OLTP is where 24 we set the --
25
9 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MR. YARSKY: No, this power and flow map 1
that I'm showing is a little bit of a cartoon. It 2
doesn't represent any particular plant. The stability 3
boundary is set by actual -- essentially the geometry 4
of the core and the pressure losses in the core is a 5
big contributing factor to where the stability boundary 6
is. It's generally analyzed on a cycle specific basis 7
in the stability boundary moves.
8 MEMBER CORRADINI: Okay.
9 MR. YARSKY: So it's not meant to say that 10 the blue curve representing the OLTP line is in some 11 way related to the stability boundary. This is just 12 a --
13 MEMBER CORRADINI: That's just, okay.
14 CHAIRMAN MARCH-LEUBA: For reference, 15 OLTP stands for Original License Thermal Power.
16 MR. YARSKY: So that they intersect in this 17 particular figure is more a coincidence than that it 18 would be driven by some consideration of the ATWS event.
19 CHAIRMAN MARCH-LEUBA: Since I drew the 20 figure in my previous life, the intention I had there 21 is that originally the reactor would have been stable.
22 And as we start to add modifications to it, it became 23 more and more unstable.
24 MR. YARSKY: Right.
25
10 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 CHAIRMAN MARCH-LEUBA: If you lose your 1
pumps.
2 MR. YARSKY: Right. And if using this in 3
combination with this figure, what this also indicates 4
is that while a plant operating at EPU power levels 5
may experience some instability during the progression 6
of the event, at MELLA+ that instability could be 7
expected to occur earlier in the event. So there's 8
not just that the reactors were to become more unstable, 9
but even in going from EPU to MELLA+, you could expect 10 the reactor would become unstable earlier in the event.
11 So in terms of thinking about an ATWS with 12 instability, I wanted to give you an overview of a 13 typical event and then to provide some trace results 14 of analysis of that event. And what this event is that 15 we've analyzed previously is a turbine trip with a full 16 bypass capability. So a 100 percent turbine bypass 17 capability assume.
18 What will occur is, during a turbine trip, 19 the turbine stop valves will close and create a pressure 20 pulse that will result in an increase in the reactor 21 power level. As that turbine stop valve closes, this 22 will also initiate the dual re-circulation pump trip.
23 And the closure of the turbine stop valves 24 will remove source of extraction steam that's used to 25
11 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 supply steam to the feed water heater cascade. And 1
that loss of extraction steam will feed water 2
temperature to begin to decrease.
3 This particular event is expected to yield 4
unstable conditions and result in large amplitude power 5
instability. The operators will undertake two key 6
actions as part of the emergency operating procedures 7
to mitigate the event, one of which is to initiate the 8
standby liquid control system injection which injects 9
soluble boron into the reactor pressure vessel to shut 10 down the reactor core.
11 The other mitigating event is to manually 12 control the reactor water level to a low level. This 13 had the effect of removing the sub-cooling from the 14 core inlet flow. So reducing the water level so that 15 the injected feed water is coming in through the 16 spargers injects to a stem atmosphere so that that water 17 can heat up before it reaches the core inlet.
18 MEMBER KIRCHNER: How would the operators 19 physically reduce the water level?
20 MR. YARSKY: At the very end I was going 21 to talk about some differences between plant designs.
22 But in some BWRs the feed water pumps themselves are 23 motor driven. In which case the operators have direct 24 control over the feed water pumps in terms of the head 25
12 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 that's supplied by the feed water pumps. So that they 1
can reduce that flow rate in order lower the reactor 2
water level.
3 And they can use the feed water pumps as 4
a high pressure injection makeup system to maintain 5
the water level at a desired water level, based on that 6
plant's specific emergency operating procedure.
7 MEMBER KIRCHNER: But their standard 8
guidelines would always adjust the water level to above 9
the core height. Or what's the --
10 MR. YARSKY: Well, in the emergency 11 operating procedures, and this would be in a general 12 sense, each plant has its own plant-specific emergency 13 operating procedures. But you have different target 14 water levels. And depending on the event progression, 15 if things get worse, you would change what water level 16 you're controlling the level to.
17 MEMBER KIRCHNER: And had you analyzed 18 those scenarios --
19 (Simultaneous speaking.)
20 MR. YARSKY: We've analyzed a variety of 21 water level scenarios as part of a generic work that 22 we've done for MELLA+. So we were looking at a 23 representative plant that had some features of BWR/4, 24 or BWR/5 and we're looking some of these plant 25
13 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 differences. And so we looked at the sensitivity to 1
the results, depending on where the operators 2
controlled the reactor water level.
3 Initially what the operators will try and 4
do is to just lower the water level is it dropped at 5
least 50 below the feed water start. So, that's their 6
initial objective.
7 MEMBER KIRCHNER: And where is that in 8
relation to the core?
9 MR. YARSKY: So that would be maybe, like, 10 top of active fuel. I want to say plus --
11 (Simultaneous speaking.)
12 CHAIRMAN MARCH-LEUBA: Say that again 13 about the active fuel. Very high, right?
14 MEMBER CORRADINI: But because of the two 15 phase level, I'm still killing reactivity.
16 MR. YARSKY: Right, because you're 17 injecting into this two phase mixture above the liquid 18 level. What's happening is this that cold injection 19 is condensing steam in that space so that the liquid 20 water that's reaching the core has that sub-cooling 21 removed.
22 MEMBER CORRADINI: Okay.
23 MR. YARSKY: So the reactor operators will 24 attempt to lower the reactor water level to achieve 25
14 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 at least a two foot gap between the feed water injection 1
spargers and the reactor water level.
2 MEMBER CORRADINI: Three foot gap, this way 3
or that way?
4 MR. YARSKY: Two feet so that level is 5
below the spargers.
6 CHAIRMAN MARCH-LEUBA:
The typical 7
presentation of the EOP, the procedures says control 8
within this band. Top band is two feet below the 9
sparger, bottom band is top water fuel. And the 10 operator has freedom to put that wherever he wants.
11 And, I mean freedom to try to keep it stable like this.
12 MR. YARSKY: Yes. There may be other 13 things going on. Most operating are written to be 14 symptom based. So you're responding to symptoms in 15 your plant as opposed to a particular kind of event.
16 MEMBER CORRADINI: Okay. Thank you.
17 MR. YARSKY: Okay. What I would like to 18 do is present the results with trace analysis for a 19 particular representative case. This calculation was 20 performed for a generic BWR/5 plant model. We assume 21 a 100 percent turbine bypass capacity. The initial 22 core flow rate is 85 percent of rated. The power level 23 is 120 percent of the original license thermal power.
24 We analyzed this at a middle of cycle 25
15 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 exposure point, in particular the peek hot excess point.
1 And what we simulate are two manual operator actions.
2 The first is an attempt to control the reactor water 3
level. In this case, the operators strive to control 4
a reactor water level to the top of active fuel.
5 And we assume that the operators will 6
initiate the standby liquid control system at 120 7
seconds into the event.
8 MEMBER REMPE: So just to make sure with 9
all this generic stuff. My understanding was that what 10 you said earlier is what you've been analyzing a mixture 11 of a reactor that's part BWR/4, part BWR/5. Now this 12 is solely a BWR/5. So this is a new analysis for a 13 just a BWR/5 plant?
14 MR. YARSKY: Well, to clarify, what we had 15 done was we had built a generic model that had features 16 of a BWR/4 and a BWR/5 that we could turn on and off.
17 So the plant model had maybe two reactor core 18 oscillation cooling systems, one that would look like 19 a core oscillation cooling system for a BWR/4 and a 20 second one that would look like a five.
21 MEMBER REMPE: Okay.
22 MR. YARSKY: And then what we can do is 23 use flags to turn on and off features to make the model 24 look like a four or to make the model look like a five.
25
16 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 So in this case, it's that same generic 1
model that you're familiar with, but the options are 2
set to deactivate all of the BWR/4 features and to 3
activate all of the BWR/5 features.
4 MEMBER REMPE: Although it's not the topic 5
for today, but my understanding is you are now getting 6
ready to do a plant specific --
7 MR. YARSKY: Yes.
8 MEMBER REMPE: -- evaluation for a plant 9
that's coming in.
10 MR. YARSKY: Right. Exactly. Yes. And 11 at the end of this presentation I wanted to tease that 12 little bit.
13 MEMBER REMPE: Good.
14 MR. YARSKY: Not by presenting any 15 results, but at least by pointing out aspects of this 16 generic analysis where we looked at and identified some 17 features that are important to capture in a plant 18 specific analysis. So this analysis cannot really be 19 extended to every plant. There are aspects of each 20 individual plant that will need to be considered. And 21 we'll sort of talk about what those important 22 considerations are.
23 MEMBER REMPE: Good, thanks.
24 MEMBER CORRADINI: So remind me a couple 25
17 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 of key differences between a four and a five.
1 MR. YARSKY: So one of the key differences 2
is that in a BWR/3 or a BWR/4, the standby liquid control 3
system injects into the lower plenum of the vessel.
4 In a BWR/5 or a BWR/6, the standby liquid control system 5
injects into the high pressure, core spray, sparger 6
line which means that that injection is above the core.
7 But in the upper plenum above the core but below the 8
separator dome.
9 MEMBER CORRADINI: So, but from the 10 standpoint of modeling, can TRACE even know the 11 difference other than location elevation?
12 MR. YARSKY: Yes.
13 MEMBER CORRADINI: Because it's a lump 14 parameter calculation?
15 MR. YARSKY: If we want to talk about the 16 standby living control system modeling that we do, we 17 take into account that when the injection is into upper 18 plenum, it's being injected into a two phase upturbulent 19 environment. And what's being injected into the lower 20 plenum that is potentially being injected as, like, 21 a cold fluid into a warm fluid at lower flow rates.
22 And we have a methodology to account for 23 that difference, but I would, if we wanted to get into 24 the details of that, we should do so in the closing.
25
18 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER CORRADINI: That's fine. That's 1
fine. But it's an elevation difference and two phase 2
versus single phase.
3 MR. YARSKY: Right. In terms of the 4
standby liquid control system. There's also 5
differences in the reactor core isolation cooling 6
system. In the BWR/5, the reactor core isolation 7
cooling system injects also the upper plenum. But in 8
a BWR/4 it injects into the feed water line.
9 MEMBER CORRADINI: Okay.
10 MR. YARSKY: So there are differences such 11 as these that we reflect in the modeling. And so in 12 this case that I'm presenting, we're just doing a BWR/5.
13 And in this representative case, we just want to talk 14 about how we reach the prediction of fuel heat up and 15 what is the mechanism for this fuel heat up.
16 So this just sort of provides an 17 environment to discuss what's going on at a local level.
18 CHAIRMAN MARCH-LEUBA: Going back to four 19 versus five, if you inject boron from the top, you have 20 two options. You can mix, in which case it would mix 21 and go around and come back to the core. Or not mix 22 and stratify. But if it doesn't mix, it drops into 23 the core and it goes into the bypass region where it 24 effect is shutting down the core. Whereas when you 25
19 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 inject in the bottom of the vessel --
1 PARTICIPANT: It has the core?
2 CHAIRMAN MARCH-LEUBA: No, it stratifies.
3 (Simultaneous speaking.)
4 MEMBER CORRADINI: Sort of fall to the 5
bottom of the vessel and pool in the bottom head of 6
those --
7 CHAIRMAN MARCH-LEUBA: But the operator=s 8
solution is very cold and had like 20, 30 percent high 9
residue.
10 MR. YARSKY: Yes.
11 MEMBER CORRADINI: Okay. That helps.
12 But on the other hand though the way TRACE is formulated, 13 we have --
14 (Simultaneous speaking.)
15 MR. YARSKY: We have a special methodology 16 to account for the physics of a stratification and --
17 MEMBER CORRADINI: Other calculations?
18 MR.
YARSKY:
Right.
Well that 19 methodology is based on experiments that were performed 20 by General Electric. And so the nature of that 21 methodology if we wanted to get into it, we should 22 probably do so in a closed session.
23 MEMBER CORRADINI: That's fine. I just 24 wanted to remind myself. Thank you very much.
25
20 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MR. YARSKY: Okay. So I'm presenting here 1
a sequence of events. And eventually what I'm going 2
to show is a video of the event, a visualization of 3
what's going on. And we can step through key moments 4
in that. So it's not important to take in this whole 5
table right now. But I wanted to sort of give you a 6
preview to what I'm going to show you to put it in 7
context.
8 In this event, we're going to have a ten 9
second mal-transient. I'm going to show you a video 10 of the event happening in real time. So that this early 11 part, a lot of things are going to happen. So I want 12 to just sort of give you this background.
13 The turbine stop valves are going to close 14 at ten seconds. A turbine bypass valves are going to 15 open at 11 seconds. And feed water flow will start 16 to decrease around 12 seconds. And then what occurs 17 in the event after this initial, very rapid period where 18 the turbine stop valves close and the turbine bypass 19 valves open.
20 We then enter a somewhat slow phase where 21 the loss of extraction steam to the feed water heater 22 pass gates results in the steady, but somewhat slow 23 decrease in feed water temperature which will 24 eventually increase reactor power to the point where 25
21 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 the reactor becomes unstable.
1 This will happen around the 100 second mark 2
at which point we'll be able to watch evolution of the 3
instability, the nature of the instability. And then 4
following that period, the instability itself is 5
mitigated and the power oscillations are reduced by 6
a combination of the effectiveness of operator actions, 7
in terms of boron injection and reducing the reactor 8
water level.
9 CHAIRMAN MARCH-LEUBA: Are your feed water 10 pumps steam driven or --
11 MR. YARSKY: In this particular model, the 12 feed water pumps are motor driven.
13 CHAIRMAN MARCH-LEUBA: So is the control 14 of the feel water flow?
15 MR. YARSKY: Yes.
16 CHAIRMAN MARCH-LEUBA: But there is no 17 operator action in one second?
18 MR. YARSKY: Right. So this twelve second 19 period here, what's happening is the feed water 20 controller is still operating, and it's reacting based 21 on its three element control to the absence of steam 22 flow to the turbine.
23 CHAIRMAN MARCH-LEUBA: So you would expect 24 the operator to take action, but not that fast?
25
22 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MR. YARSKY: Yes. So when you talk about 1
the feed water flow starting to decrease, and Jose, 2
that's a very excellent point. In this particular 3
instance it is the automatic controller, the automatic 4
three element controller that's normally functioning 5
during plant operation.
6 It's still assumed to function before 7
operators have the opportunity to intervene. And in 8
this particular case, because it's a motor driven feed 9
water system, those controllers are still active.
10 Yes. And in this particular case it's reacting to the 11 decrease in steam flow.
12 So here's a plot of the transient reactor 13 power for this representative case when you can see 14 it's very early in the event. The closure of the 15 turbine stop valves causes this power pulse. We'll 16 skip ahead.
17 CHAIRMAN MARCH-LEUBA: The mouse not 18 working?
19 MR. YARSKY: The mouse is not working.
20 So I can see it on my screen, but I can't -- oh wait.
21 Oh. Oh there we go. Okay.
22 CHAIRMAN MARCH-LEUBA: It's extended.
23 MR. YARSKY: Yes, okay. So this point 24 here --
25
23 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER CORRADINI: It also disappears when 1
you're not using it.
2 MR. YARSKY: Yes. Okay. So now I can use 3
the mouse to point. So this power peak here, this 4
occurs because of the back pressure wave from the 5
closure of term and stop valves. There's very shortly, 6
very shortly thereafter the dual re-circulation pump 7
trip, which reduces the flow, causing the power to come 8
down. This happens very quickly.
9 Then there's a slow progression where you 10 can see the power level increasing. What's occurring 11 here is that the feed water injection is getting colder, 12 and that colder water is reaching the core. It's trying 13 to push the boiling boundary higher into the core and 14 this increases the reactor power.
15 Eventually the reactor becomes unstable.
16 And then we can see is around the 230 second point, 17 the oscillations have been dying away or have been 18 damped by the effectiveness of the operator actions.
19 CHAIRMAN MARCH-LEUBA: I cannot see it 20 there already, but can you tell about the model 21 oscillation. Is it in phase, out of phase.
22 MR. YARSKY: Yes. This is why it's very 23 important. I'm pre-empting my next slide. So what 24 we see is that the oscillation mode is initially in 25
24 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 the core-wide mode, and that during the event a 1
bi-modal oscillation occurs. And this is a coupled 2
oscillation mode between the core-wide and the regional 3
mode.
4 And so this is just sort of where we first see 5
that oscillation using our visualization methodology.
6 But I want to show you is a video of this so we can 7
watch the event progression. So if you'll give me a 8
moment to do that.
9 MEMBER CORRADINI: But, as you're doing 10 that the operators start taking control over the 11 controller at what time?
12 MR. YARSKY: About two minutes.
13 MEMBER CORRADINI: Okay. So 115 seconds.
14 MR. YARSKY: 110 seconds for the feed water 15 control and 120 seconds for --
16 MEMBER CORRADINI: Or that the controller 17 is doing what it thinks it should do?
18 MR. YARSKY: Right. Yes.
19 MEMBER CORRADINI: What does it think it 20 should do?
21 MR. YARSKY: It thinks it should be 22 maintaining the reactor water level at the nominal 23 level. However, because of the three element 24 controller, there is going to be a delta level that 25
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 arises because of the air due to no steam flow.
1 MEMBER CORRADINI: Okay. Thank you.
2 MR. YARSKY: So it'll be trying to maintain 3
normal water level with some offset.
4 MEMBER CORRADINI: Thank you.
5 MR. YARSKY: So in the first ten seconds, 6
there's just a null transient. And then you'll see 7
the turbine trip come in. Power goes up and responds 8
to the turbine, turbine stop valve closure and then 9
comes down and responds to the duel re-circulation pump 10 trip.
11 At this point in the event there's, this 12 is a relatively slow portion in the event. What I'm 13 going to do is I'm going to advance the time a little 14 bit. So if you'll pardon my fast forwarding.
15 PARTICIPANT: My God.
16 MR. YARSKY: What I wanted to do is not 17 immediately go to the point of instability, but just 18 a few seconds beforehand. And the area of this movie 19 that's interesting to watch is the plot of the axial 20 power shade. So that's shown here. And what you can 21 see is that the axial peeking becomes stronger more 22 bottom peaked as the feed water temperature comes down.
23 And then we begin to see the onset of the 24 instability at this 100 second mark. And then each 25
26 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 one of these boxes represents one of the fuel assemblies 1
in the core and its height, of course, represents the 2
power in that fuel assembly. And so here, this is the 3
axially integrated power. So this is total assembly 4
power.
5 CHAIRMAN MARCH-LEUBA: Walt, Walt, Walt.
6 Microphone.
7 (Simultaneous speaking.)
8 MR. YARSKY: Yes. And as you can see there 9
are certain points where the power is reduced. That's 10 where the control blades are inserted in this middle 11 cycle point. It's around this point in the transient 12 that we begin to see the evolution of the bi-modal 13 oscillation where you can see this aspect of the 14 regional oscillation occurring. And it's here also 15 that we see the frequency of the oscillation double.
16 So this is how we know it's this non-linear modal 17 coupling.
18 What's notable is once the oscillation 19 develops this feature, even though it appears as though 20 the total core power oscillation magnitude is lower, 21 the oscillation magnitude in the hot assemblies is 22 actually higher.
23 MEMBER KIRCHNER: Let me ask you --
24 PARTICIPANT: Mic.
25
27 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER KIRCHNER: How competent are you 1
at what you're showing when it starts oscillating?
2 What's your model doing that it oscillates like this 3
across the core radially?
4 MEMBER BLEY: What drives it?
5 MEMBER KIRCHNER:
What drives the 6
asymmetry that was --
7 MR. YARSKY: So this is a well understood 8
feature of modal of kinetics. So if you were to look 9
at the different modes of a neutron flux, certain higher 10 modes can have their reactivity excited. And there's 11 a non-linear coupling.
12 So there's second order coupling between 13 different modes of a neutron flux. And it's just once 14 you get these huge amplitude oscillations that are 15 occurring in the fundamental or core-wide mode, there's 16 essentially reactivity spillover into these higher 17 order modes of the neutron flux.
18 And so that causes the power to increase 19 and decrease on different sides of the core. The flow 20 then responds in kind.
21 MEMBER KIRCHNER: How physical do you 22 think that is though, in reality?
23 MR. YARSKY: This regional model --
24 MEMBER KIRCHNER: Your model is doing it, 25
28 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 but the regional mode, what --
1 (Simultaneous speaking.)
2 MR. YARSKY: Yes. The regional mode is 3
something that --
4 MEMBER KIRCHNER: -- that would start it 5
off on one side of the core versus the other.
6 MR. YARSKY: Oh. In this particular 7
model, we force the plane of symmetry. So we did this 8
with channel grouping. But there's nothing to say that 9
it couldn't -- so we're trying to say like a north/south 10 oscillation. There's nothing to say it couldn't be 11 an east/west.
12 MEMBER KIRCHNER: Yes. That's my point.
13 MR. YARSKY: And east/west oscillation.
14 There's also the possibility that the north/south mode 15 and the east/west mode can become coupled and you can 16 have --
17 MEMBER KIRCHNER: I was hoping you would 18 tell me that your initial conditions weren't completely 19 symmetric.
20 MR. YARSKY: The initial conditions are 21 completely symmetric.
22 MEMBER KIRCHNER: Oh, they are completely 23 symmetric.
24 MR.
YARSKY:
They are completely 25
29 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 symmetric.
1 MEMBER CORRADINI: So then, I guess I 2
didn't understand your answer to Walt. What did you 3
mean by force? I don't understand what you mean by 4
force.
5 MR. YARSKY: So in this particular core, 6
there are 764 fuel assemblies.
7 MEMBER CORRADINI: Right.
8 MR. YARSKY: And what we need to be able 9
to capture is that the neutron flux on one side of the 10 core can be different than the other side of the core.
11 So in our neutronics model we simulate 764 unique 12 channels. But in the TRACE thermal hydraulics model, 13 what we do is we can group together two channels, one 14 with its symmetric sister. So we have 382 TRACE channel 15 components.
16 MEMBER CORRADINI: Right.
17 MR. YARSKY: But because we make those two 18 channel components of the same thermal hydraulic 19 condition, this would only allow the flow oscillation 20 or the difference in bumble flows to appear and across 21 only one of the axes.
22 MEMBER CORRADINI: So they wouldn't rotate 23 the way you would --
24 MR. YARSKY: Right. If we were to model 25
30 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 the core with 764 TRACE channel components, then we 1
would not be able to know a priori which one of these 2
degenerate modes would excite first.
3 MEMBER KIRCHNER: So, that's my point.
4 Yes.
5 MR. YARSKY: So in this case what we've 6
done is we've picked one and we said it will be the 7
north/south mode, even though the east/west mode would 8
be a degenerate. But it would have the same reactivity 9
characteristics, but we don't know which one would --
10 MEMBER CORRADINI: But, I guess to get back 11 to Walt's original question about what starts it off.
12 It would seem there's got to be a numerical reason 13 that it starts off. Like some tolerance that if I 14 torque down on the tolerance I would delay, or if I 15 allow the tolerance to be larger, I would start it 16 earlier. I would expect a numerical tolerance is the 17 kick off point.
18 MR. YARSKY: Yes. But we studied --
19 (Simultaneous speaking.)
20 MR. YARSKY: What we studied more recently 21 is that there is some uncertainty in when the regional 22 mode kicks in.
23 MEMBER CORRADINI: Okay.
24 MR. YARSKY: And we can improve the 25
31 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 reliability and consistency of our calculations by 1
using a feature in our methodology referred to as white 2
noise where we kind of just keep a small level of 3
excitation present in the calculation in all of the 4
different core-wide and first harmonic mode.
5 And so what this would mean is once the 6
conditions are right for that mode to be unstable, 7
there's a little bit of noise there to --
8 MEMBER CORRADINI: To kick it off.
9 MR. YARSKY: -- to keep it excited. So 10 once it becomes unstable, an oscillation would grow.
11 MEMBER CORRADINI: Yes.
12 MR. YARSKY: In these calculations what 13 we found is, because we ran these calculations by just 14 showing you without that feature turned on, that there's 15 about a 15 to 20 second delay of when that mode appears 16 relative to when we would expect that mode. If you 17 had a little bit of noise present --
18 CHAIRMAN MARCH-LEUBA: I mean, is not 19 really a delay, is that they start so small, you can't 20 see it. It has to grow. I mean, if you blow it up, 21 you see oscillations are growing.
22 (Simultaneous speaking.)
23 MR. YARSKY: Well, if you were to plot them 24 side by side, what you would see is the large amplitude 25
32 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 oscillations would be apparent about 15 or 20 seconds.
1 CHAIRMAN MARCH-LEUBA: And the noise exist 2
in the real reactors. It's there, so.
3 MEMBER KIRCHNER: Right. Well, my point 4
was going to be that in the real reactor, you've got 5
different burn up patterns. It's not symmetric like 6
it is in your analytical space for your initial 7
conditions. And that would be an obvious source for 8
starting a curtivation (phonetic) that you show in 9
numerically here.
10 So my question is to what extent are your 11 numerical oscillations that are going around bounded 12 by physics so that you're confident when you do an 13 analysis like that? And then compared to an actual 14 reactor where you will have non-symmetry that you're 15 bounding the kind of oscillations that you're seeing 16 around the core? Do you see what my question is?
17 MR. YARSKY: I think I understand. Let 18 me attempt --
19 MEMBER KIRCHNER: Yes.
20 MR. YARSKY: -- to respond. One of the 21 things that we do to justify the TRACE application to 22 these kinds of problems is to qualify or assess TRACE 23 against stability measurements. And in these kinds 24 of measurements, you're looking at these different 25
33 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 modes of oscillation, but in context where they're 1
stable. And can TRACE predict that harmonic contour 2
or that shape of the perturbation and the correct to 3
K ratio.
4 So when you have TRACE comparisons against 5
those stability measurements that we have confidence 6
that we're able to predict that behavior in the linear 7
stable range. When you get into this unstable 8
behavior, I think the only real plant data that we have 9
in an integral sense that we can use for assessment 10 purposes is the Oskarshamn event.
11 MEMBER BALLINGER: Yes. That's what I was 12 going to mention.
13 MR. YARSKY: Right. And in that one case, 14 we were able to represent the plant performance, the 15 plant data very well. The TRACE assessment indicates 16 very good agreement between our TRACE predictions and 17 the data measured during the event.
18 MEMBER KIRCHNER: Okay.
19 MR. YARSKY: So I would refer you back to 20 the body of assessment that we have for TRACE.
21 CHAIRMAN MARCH-LEUBA: Yes. But bottom 22 line is all these modes have been observed in real 23 reactors.
24 MEMBER KIRCHNER: No. I don't doubt it.
25
34 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 I just, you look at a video like that and don't convince 1
yourself that's actually what's going to happen.
2 CHAIRMAN MARCH-LEUBA: Well, no, no.
3 This is one particular, this is --
4 MEMBER KIRCHNER: This is kind of an 5
idealized case.
6 CHAIRMAN MARCH-LEUBA: As role TRACE 7
calculation, this is one particular implementation, 8
one particular observation. If you run into more, it's 9
going to be different.
10 MR. YARSKY: And we've run many, many of 11 these calculations with subtly different inputs, with 12 different noise features, with different sensitivities 13 to level. We've run a number of these different 14 calculations.
15 I'm presenting one here because when I talk 16 about the predicted heat up mechanism, I want to put 17 in context what's happening in the plant, what's 18 happening in the core, you know, what's driving the 19 event overall. So when I start talking about what's 20 happening around the core hotspot, you sort of have 21 an environment to talk about it.
22 CHAIRMAN MARCH-LEUBA: And with that in 23 mind for the record, all these oscillations have 24 occurred in real reactors from the first three seconds 25
35 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 or so. Then a scram happens. This is a calculation 1
of what will happen if the scram were to fail.
2 MEMBER CORRADINI: Right.
3 CHAIRMAN MARCH-LEUBA: That has never been 4
reported.
5 MEMBER BALLINGER: The prime directive is 6
don't let it happen.
7 CHAIRMAN MARCH-LEUBA: Yes. That's the 8
moral of the story. Scram.
9 MR. YARSKY: Yes. So what I'm providing 10 here is a plot of the peak cladding temperature during 11 this event as a function of time. And there are two 12 key phases that I would like to discuss. The first 13 is the cyclic dryout/rewet phase. This is relatively 14 short in this event.
15 But what occurs eventually as the power 16 and flow oscillations increase in magnitude and in 17 particular the flow oscillation. As the flow 18 oscillation increases in magnitude, the hotspot and 19 the hot assembly will experience periods of dryout.
20 And this will occur when the local flow rate is low.
21 In that period of dryout, the fuel will 22 heat up a little bit. But the flow is oscillatory, 23 so it'll eventually return and then you'll have an 24 increase flow. And what that will do is it will rewet 25
36 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 that dried out spot, remove that heat and the 1
temperature will come down. And so as the event 2
progresses and oscillation magnitude is increasing, 3
we'll eventually hit a point where there is a cyclic 4
dryout/rewet.
5 What we then predict is that if this keeps 6
going that they'll be a temperature excursion. What 7
we show in this calculation is the temperature excursion 8
first to a little bit below 2200 Fahrenheit. However, 9
once the oscillation mode becomes this bi-modal coupled 10 mode where the regional mode kicks in and local power 11 oscillation magnitude increases, we see that PCT then 12 crossover above 2200 Fahrenheit.
13 MEMBER CORRADINI: So ignoring the blue 14 line, you always see this double bump because of going 15 from single mode to bi-mode?
16 MR. YARSKY: No. You do not always see 17 this bump.
18 MEMBER CORRADINI: So when you're doing 19 this, you don't see the bump, sometimes?
20 MR. YARSKY: It really does depend. You 21 may have this happening before, you may have the side 22 to side happen before the core-wide happens, so that 23 could be a possibility in which case you wouldn't see 24 25
37 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER CORRADINI: Okay.
1 MR. YARSKY: -- this bump. In this 2
particular case, I got to address the bump. The bump 3
is happening because that regional mode kicks in and 4
causes local oscillation magnitude to increase.
5 MEMBER CORRADINI: But we've entered a bad 6
region at more like at 120 seconds where cyclic 7
dryout/rewet stops.
8 MR. YARSKY: Right. At that point there's 9
a temperature excursion and it would be, what we would 10 say is that once that temperature excursion comes in, 11 it would be very difficult to reliably predict what 12 the PCT consequence would be.
13 MEMBER CORRADINI: So we've gone past the 14 point of no return.
15 MR. YARSKY: Right.
16 MEMBER CORRADINI: Okay.
17 MR. WYSOCKI: And that's it. This is 18 Aaron Wysocki, Oak Ridge. To go with what Pete said, 19 we've done simulations where it begins in phase and 20 remains in phase, where it begins out of phase and 21 remains out of phase and where they crossover like that.
22 So it really depends on the --
23 MR. YARSKY: Right.
24 MR. WYSOCKI: -- reactor and the event.
25
38 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MR. YARSKY: Right, but looking at what's 1
going on this local level of the hotspot, you go into 2
a period of cyclic dryout/rewet and then something trips 3
you into this temperature excursion. And where that 4
occurs is once the cladding surface temperature 5
reaches the minimum stable film boiling temperature, 6
TRACE predicts that that hotspot will lock into a film 7
boiling heat transfer regime. So it will have, like, 8
a very low heat transfer coefficient which then causes 9
the temperature excursion.
10 MEMBER KIRCHNER: Just to clarify, as Mike 11 was asking this, kind of right there when you're 12 crossing the blue whether it's this blue line or not.
13 MR. YARSKY: This part here.
14 MEMBER KIRCHNER: Yes. But just before 15 that, the double bump, that's the regional part --
16 MR. YARSKY: Yes. This is the onset of 17 the regional mode and with that --
18 (Simultaneous speaking.)
19 MEMBER KIRCHNER: I would ask you at some 20 point to consider just putting an actual asymmetry into 21 the initial conditions in the core and seeing what 22 happens.
23 MR. YARSKY: Right, if we were to --
24 MEMBER KIRCHNER: So that one could then 25
39 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 physically separate out how much of this is numerical, 1
and how much of it -- I'm not doubting that it's real 2
3 MR. YARSKY: Yes.
4 MEMBER KIRCHNER: -- and you're not 5
getting your good comparison in a macro sense, but just 6
put a perturbation somewhere in the core and see how 7
your local perturbations progress. I would expect a 8
very similar TRACE and such. But if it weren't similar, 9
then I would be concerned, right?
10 MR. YARSKY: Right. One thing that we 11 could do is to put a some sort of control rod 12 mal-alignment into the model. What this would have 13 the effect of is changing that harmonic shape. And 14 so it wouldn't necessarily be that the oscillations 15 would occur along the same line of symmetry.
16 So it's because the core is symmetric that 17 that harmonic mode symmetry is occurring in that 18 north/south, east/west plane. But if we were to say 19 put the control odds a little off kilter, it's likely 20 that that symmetry plane would rotate. So it's --
21 MEMBER KIRCHNER:
This experiment, 22 numerical experiment that I'm suggesting would then 23 give you a little more confidence about how much of 24 what when you're in the area is numerically induced 25
40 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 versus physically induced. Do you see where I'm 1
getting at? Codes will go into that, like you said, 2
you can put white noise in or something --
3 MR. YARSKY: Right.
4 MEMBER KIRCHNER: -- to stimulate it. But 5
otherwise, sometimes the reason you get these kind of 6
oscillations is that's the way the numerical solver 7
is going around the core.
8 MR. YARSKY: Right. Right.
9 CHAIRMAN MARCH-LEUBA: This oscillator 10 just to summarize it, they predicted by many different 11 codes and they even predicted by hand calculations.
12 MEMBER KIRCHNER: No, you're missing my 13 point. My point isn't that. I'm not questioning the 14 validity and it's been observed experimentally. What 15 I'm saying is you want to know what's numerically 16 induced versus what's physically induced.
17 CHAIRMAN MARCH-LEUBA: The fact that you 18 had the same behavior with hand calculations, it's 19 telltale that you are almost there.
20 MEMBER KIRCHNER: You're still missing my 21 point. My point is that numerically as the oscillation 22 progresses around the core east/south, north/west 23 whatever its doing, how much of that is your solver 24 algorithm --
25
41 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 CHAIRMAN MARCH-LEUBA: Your typically --
1 MEMBER KIRCHNER: -- and how much is --
2 CHAIRMAN MARCH-LEUBA: Typically you 3
start by changing Delta T, that=s the easy way to do 4
it. If it's numeric and you change, it pops up 5
immediately. And you --
6 MEMBER KIRCHNER: That will change the 7
overall behavior. You're missing my point still. You 8
want to separate out what your solver, your numerical 9
algorithm is doing in instigating the oscillation 10 versus what physical --
11 CHAIRMAN MARCH-LEUBA: When you do --
12 MEMBER KIRCHNER: -- trick point --
13 CHAIRMAN MARCH-LEUBA: When you do it with 14 hand calculations, there's no --
15 MEMBER KIRCHNER: I know you can make that 16 hand calculation oscillate. That's not the point that 17 I'm making. You've got a 3D representation, and what 18 you want to do is put a physical variance somewhere, 19 like burn up depletion or control rod as you suggested.
20 And then see how it goes around.
21 MR. YARSKY: Yes. Okay. I think that the 22 work that we've done with the noise sensitivity study 23 will likely address a lot of your concern here, and 24 it's something that we were concerned with as well.
25
42 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 I think to get at what you talking about, this effect 1
that there could sort of be like a ghost in the numerics 2
of code calculation itself.
3 I feel like we've done a lot of work to 4
address that, but not by inserting a spatial asymmetry.
5 It's something that we could do, but I can send you 6
or I can send the Committee references to work that 7
we've done to look at things like time step size, to 8
look at things like noise. Yes. I can forward those 9
along and maybe that could address your concern.
10 CHAIRMAN MARCH-LEUBA: Before you change 11 this slide, this is a phase calculation with a Tmin 12 correlation, correct?
13 MR. YARSKY: Yes.
14 CHAIRMAN MARCH-LEUBA: And the reason you 15 are, were going to want to the closed session and talk 16 about the KATHY experiment is because there is some 17 uncertainty on that correlation. Can you address that?
18 MR. YARSKY: In two slides what I wanted 19 to do is I have a slide that sort of steps through this 20 fuel heat up mechanism.
21 CHAIRMAN MARCH-LEUBA: Okay.
22 MR. YARSKY: And at that point I think that 23 would be an excellent point to talk about this topic 24 of Tmin.
25
43 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 What we found in the base case calculation 1
that I presented is that large amplitude regional power 2
oscillations developed with this model coupling and 3
that this results in high temperatures in certain 4
hotspots in the core with the PCT exceeding 2200 5
Fahrenheit. So we would diagnose this as a condition 6
of fuel damage.
7 We found that operator actions to reduce 8
level and operator actions to inject stem by the control 9
system were effective in reducing oscillation magnitude 10 and eventually shutting down the reactor. However, 11 these actions were not effective in a timely enough 12 manner to avoid this temperature excursion and this 13 local PCT exceeding 2200.
14 Based on our calculation --
15 MEMBER KIRCHNER: This is the value of what 16 you're doing. So you just said that for this stylized 17 reactor case that, what was it 120 seconds the one you 18 did?
19 MR. YARSKY: Yes.
20 MEMBER KIRCHNER: That's because that's 21 what's in the EOP, right?
22 MR. YARSKY: Well that is a plant specific 23 consideration of when --
24 MEMBER KIRCHNER: You just, well forget 25
44 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 the plants. You just took your model and you made that 1
point of injection 90 seconds.
2 MR. YARSKY: Right.
3 MEMBER KIRCHNER: Would it kill off the 4
oscillations and keep you below the 2200 threshold?
5 MR. YARSKY: I don't know if --
6 MEMBER KIRCHNER: You don't have to answer 7
me --
8 MR. YARSKY: If there is --
9 MEMBER KIRCHNER: It's the value of having 10 a model like this --
11 MR. YARSKY: Right.
12 MEMBER KIRCHNER: -- that you can look at 13 14 MR. YARSKY: You would be able to look 15 16 MEMBER KIRCHNER: -- that kind of thing.
17 MR. YARSKY: -- a changing the operator 18 action timing and if that operator action timing is 19 very early, what you may see is that that mitigating 20 action could be sufficient to preclude instability from 21 occurring at all.
22 (Simultaneous speaking.)
23 MR. YARSKY: -- the plant specific license 24 amendment request of like exceptionally fast operator 25
45 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 action.
1 CHAIRMAN MARCH-LEUBA: We have seen that 2
for MELLA classifications where one plant couldn't 3
survive 120 and they did it in 90.
4 MR. YARSKY: Right.
5 CHAIRMAN MARCH-LEUBA: And there was a lot 6
of operator training and study --
7 (Simultaneous speaking.)
8 MR. YARSKY: -- if the operators could 9
react --
10 CHAIRMAN MARCH-LEUBA: -- to ensure that 11 they could do 90.
12 MR. YARSKY: -- in that time frame.
13 CHAIRMAN MARCH-LEUBA: And neither were 14 able to. But they were able to do it in a 7 seconds.
15 MR. YARSKY: Yes.
16 CHAIRMAN MARCH-LEUBA: It was not, it was 17 borderline. I mean, the evidence, if I remember 18 correctly, it was like 60. But some of the tryouts 19 to 80, 85.
20 MR. YARSKY: Right.
21 CHAIRMAN MARCH-LEUBA: So they couldn't 22 go much lower than 90. But it was very effective.
23 I mean --
24 MR. YARSKY: Right.
25
46 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 CHAIRMAN MARCH-LEUBA: -- surviving this 1
or not surviving that was.
2 MR. YARSKY: So from our TRACE result we 3
look at that sort of local hotspot to better understand 4
what TRACE was predicting in terms of the mechanism 5
of what leads to these, to this fuel heat up.
6 And what is occurring is oscillation 7
magnitude is increasing and we talked bout on that other 8
figure this period of cycling dryout/rewetting. And 9
as the oscillation magnitude continues to grow, the 10 amount of time that the hotspot remains in a dryout 11 condition increases as the flow oscillation magnitude 12 increases.
13 And so in each part of that dryout phase, 14 the peak temperature gets a little higher. And then 15 during the rewet portion, that heat can be removed, 16 but if some of it remains, then the average temperature 17 of the hotspot begins to increase.
18 And we refer to this mechanism as 19 ratcheting. And that average temperature will 20 continue to increase until it hits a point where the 21 cladding surface fails to rewet. And then there's no 22 more rewetting. And then that leads to the temperature 23 excursion.
24 In TRACE, and this is a, just like the trade 25
47 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 systems analysis code feature, that film boiling lock 1
is diagnosed by the temperature exceeding the minimum 2
stable film boiling temperature.
3 MEMBER CORRADINI: So when you cross that, 4
you switch key transfer modes in the calculation?
5 MR. YARSKY: Right. So once the 6
temperature crosses that line, you go from transition 7
boiling to film boiling and you lock into film boiling.
8 And so you'll stay locked in film boiling until you're 9
able to quench the surface. And then once you're locked 10 into film boiling, the temperature undergoes, you see 11 this temperature excursion.
12 So that's the mechanism by which the ATWS 13 and stability can lead to potential fuel damage is that 14 oscillations increase in amplitude, lead to cyclic 15 dryout/rewet. As the oscillation amplitude increases, 16 cyclic dryout/rewet leads to ratcheting. Ratcheting 17 eventually leads to film boiling. And once you're in 18 film boiling you undergo temperature excursion. So 19 that's the mechanism that we predicted with the TRACE 20 model.
21 MEMBER BALLINGER: So when you say fuel 22 damage, you're talking about increasing, going about 23 the 2200 degrees?
24 MR. YARSKY: Yes.
25
48 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER BALLINGER: What about ballooning 1
and things like that?
2 MR. YARSKY: So we are able to model --
3 MEMBER BALLINGER: By the time you get to 4
2200 degrees, I got a feeling that if it's a pressurized 5
rod --
6 MR. YARSKY: There will be a burst 7
population, and also we will have perforated a large 8
number of pins by this point. So every time you're 9
in dryout you're likely to, and perforate the pin in 10 some way.
11 MEMBER BALLINGER: You're not going to 12 shut off channels. Like ballooning.
13 MR. YARSKY: Well there will be a flow are 14 reduction if there is a lot of ballooning, and if that 15 ballooning and bursting is occurring in a co-planer 16 way. Right now what we're able to predict in TRACE 17 is there's a function of balloon in burst. We're able 18 to look at the change in oxidation.
19 For right now, that balloon and bursting 20 incidents doesn't feed back into the calculation in 21 terms of the change in flow area. So we partially 22 account for that, but we don't fully account for that.
23 MEMBER CORRADINI: But back to my original 24 comment. As soon as it starts creeping up, you're past 25
49 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 the point of no return and all bets are off as to how 1
we --
2 MR. YARSKY: Right. And keep in mind this 3
is a beyond design-basis event so that. So perforating 4
the rods is allowed. The real criterion, the 5
regulatory criterion here really is that the core must 6
remain in a cool-able geometry. And so, that's why 7
we're looking at fuel damage.
8 One thing we haven't looked at yet is the 9
population of fuel damage. So what fraction of the 10 core has undergone fuel damage and what does that mean 11 in terms of core cool-ability. That's something that 12 we haven't done in our research yet.
13 MEMBER REMPE: There were some --
14 CHAIRMAN MARCH-LEUBA: In the NRR side, 15 I remember that you were going to say that. In the 16 NRR side we did look at a small fraction of possible 17 places where you would have damage. So if only 0.1 18 percent of the core goes beyond 2200, it might be still 19 a coolable geometry.
20 MR. YARSKY: Right. That's something 21 that we haven't done in our confirmatory analysis --
22 (Simultaneous speaking.)
23 CHAIRMAN MARCH-LEUBA: -- so when does 24 that used that argument for MELLA+.
25
50 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER REMPE: Right and I thought it was 1
discussed, I can remember --
2 CHAIRMAN MARCH-LEUBA: Right.
3 MEMBER REMPE: -- what plant is was 4
discussed in because it was more of a vicious 5
discussion. That they decided that it was enough of 6
the core remained coolable that -- it was back when 7
Chris Jackson was here. But I thought, I mean that 8
was something we discussed and I thought there was some 9
basis it --
10 CHAIRMAN MARCH-LEUBA: But basically if 11 only one pin goes over 2200 and oxidizes on this occur, 12 it does not prevent core coolable geometry.
13 (Simultaneous speaking.)
14 MEMBER REMPE: Right. But more than one 15 was predicted.
16 CHAIRMAN MARCH-LEUBA: If it's ten, okay.
17 If it's 10,000, no.
18 MEMBER REMPE: Yes. Again, my memory, but 19 I thought research didn't support the decision by --
20 MR. YARSKY: Well it's in the research 21 confirmatory analysis. We don't yet have the 22 capability to predict damage population. So one thing 23 I think that would be necessary in order to do that 24 would be to model the assemblies with sufficient 25
51 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 resolution that you have the actual rod to rod 1
distribution of powers represented.
2 The reason for that being is because this 3
predicted fuel mechanism is, this predicted fuel 4
mechanism depends on this incidence of dryout and rewet 5
to get things started. So if your rod power 6
distribution is such that you have maybe a peak rod 7
that could be at a peaking factor of 1.2, but then you 8
have ten rods that are a peaking factor of 1.1, you 9
would probably would want to capture that those 10 additional ten rods at the 1.1 peaking factor in your 11 analysis.
12 Whereas, in the TRACE calculation now we 13 have essentially, like the bulk of the assembly is 14 represented with a couple or a few rod groups and then 15 you model the hot rod. I think we need to better capture 16 that rod to rod power distribution before we have 17 confidence in saying some given fraction of the core 18 has experienced fuel damage.
19 And so I think that for our confirmatory 20 analysis research hasn't developed that capability yet, 21 but I think that we will. And then we'll be able to 22 answer question of what is the predicted damage 23 population. The next step after knowing that 24 population would then be to make some sort of 25
52 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 determination with regards to core coolability. But 1
that'll be step two after we do step one.
2 MEMBER CORRADINI: But if I could just take 3
a step back from all these steps. There is some 4
temperature that's not the takeoff temperature, not 5
2200f. But currently staff is saying once I pass this, 6
we'll call intermediary temperature, I have no known 7
way back from it. Therefore, that's a loss of coolable 8
damage.
9 MR. YARSKY: Well what you could say, but 10 try and be very specific about --
11 MEMBER CORRADINI: I'm not trying to give 12 a number, I'm trying say there's something --
13 MR. YARSKY: Their licensee would be able 14 to demonstrate that they have met the core coolability 15 requirement. A way to make that demonstration would 16 be to show that they completely avoid this fuel issue.
17 MEMBER CORRADINI: That's one way.
18 MR. YARSKY: That's one. So if they're 19 able to demonstrate that they never get into this 20 mechanism during the event, that is a manner in which 21 they can demonstrate that they've met the regulatory 22 criterion.
23 MEMBER CORRADINI: But they may also argue 24 that, take Ron's suggestion that if I stable all 25
53 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 ballooning then --
1 MR. YARSKY: For instance, if they were 2
able to stay below CHF, if they were able to demonstrate 3
that the SLMCPR limit was met for the entire event, 4
then clearly they've avoided, they are still going to 5
meet the coolable geometry. So there's a way that you 6
can perform the analysis with a more stringent criterion 7
in order to demonstrate that you've met the regulatory 8
criterion.
9 CHAIRMAN MARCH-LEUBA: But just two 10 things. First, you can reach the -- the point of no 11 return where you're going to failure to rewet and 12 continue to overheat and never reach 2200 because your 13 power is not high enough. And you figure if we had 14 not moved into a --
15 MR. YARSKY: Right. So if you return to 16 this figure, we could postulate an event where you would 17 have a failure to rewet, but still not exceed 2200.
18 MEMBER CORRADINI: Are you still then --
19 MR. YARSKY: That would be a possibility, 20 but it would be very difficult I think given our 21 uncertainties in the models to say if I were to predict 22 a PCT of 1450 kelvin --
23 (Simultaneous speaking.)
24 MR. YARSKY: -- well, you know, how certain 25
54 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 am I or could I really say that that temperature is 1
less than 2200. Especially given considerations of 2
you know, hey what if because of a numerical effect 3
the oscillation started ten seconds earlier or --
4 CHAIRMAN MARCH-LEUBA: The other thing I 5
wanted to point out on this figure is that you have 6
an oscillation in the whole band, the 10 by 10 bundle.
7 But this is the regulatory flux of the hottest pin 8
on the bundle.
9 MR. YARSKY: No, it's actually more subtle 10 than that. I mean, for a cartoon type understanding 11 what's going on, you could think of this figure as the 12 hot spot, but it's not. This represents that each point 13 in time the highest temperature on any cladding in the 14 core. So it's searching. So it's really happening 15 around this bump is that this point is a different point 16 now. There's a different hot spot. So this is not 17 the temperature history of a particular spot in the 18 core.
19 From the standpoint of just understanding 20 the phenomenon, you could look at this figure and think 21 of it in that way and you would have the understanding 22 necessary I think to move on. But it's a very important 23 subtle point of this figure that it is not, it's not 24 a history of an individual spot.
25
55 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 CHAIRMAN MARCH-LEUBA: But it's not all 1
the rows in the bundle. It's like a small section of 2
the bundle.
3 MEMBER BALLINGER: Well, he could probably 4
interrogate --
5 (Simultaneous speaking.)
6 MR. YARSKY: Yes. What I'm saying is I 7
could do that on the rods, but I think that that exercise 8
will not be valuable until we have a significant number 9
of rod groups represented in every channel component.
10 Taken into account the distribution of rod powers --
11 CHAIRMAN MARCH-LEUBA:
TRACE has 12 typically 35/7 rod groups in the bundle, right?
13 MR. YARSKY: Right. I think we would want 14 to up that number significantly before we start talking 15 about damage population. So damage population I think 16 is the next step, but we need to model more to do it.
17 MEMBER BALLINGER: But if I might to help 18 because I asked about Walt, so I=ll point at Walt.
19 I'm still not clear about what Jose understands and 20 I'm not clear about the 3/5/7 rod itself in abundant.
21 The cartoon movie you showed us, and I'll go back.
22 Your seven hundred and something. You divided into 23 two so there three hundred and something. So let's 24 say you did the seven hundred and something, that's 25
56 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 a bundle.
1 MR. YARSKY: That's a bundle.
2 MEMBER BALLINGER: You see that bundle.
3 I am looking at seven representative heat structures 4
that I think to be rods and no more.
5 MR. YARSKY: Right. Correct.
6 MEMBER BALLINGER: That's what I want to 7
make clear about.
8 CHAIRMAN MARCH-LEUBA: That's what I meant 9
by programs.
10 MR. YARSKY: Right.
11 MEMBER BALLINGER: But TRACE thinks of it 12 as a funny looking heat structure.
13 CHAIRMAN MARCH-LEUBA: Yeah, but --
14 MEMBER KIRCHNER: And for this particular 15 example, how much of the core is above that threshold.
16 CHAIRMAN MARCH-LEUBA: Yes. That's why 17 he said he doesn't know.
18 MR. YARSKY: I don't know. I don't --
19 (Simultaneous speaking.)
20 MR. YARSKY: That's something that we are 21 going to look at analyzing in the future to be able 22 to answer that particular question. What fraction of 23 the core experiences this condition? We don't have 24 that yet.
25
57 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER KIRCHNER: And then I would ask you, 1
don't answer me or anything like that. But I'll ask 2
you, are you aware of the draft RIA Reg Guide that is 3
out for, on it?
4 MR. YARSKY: I'm not immediately familiar 5
with it.
6 MEMBER KIRCHNER: Do I have the right 7
person, Clifford?
8 MR. YARSKY: Paul Clifford?
9 MEMBER KIRCHNER: Yes. You need to look 10 at that because then he has draft criteria for what's 11 to find this core coolability, et cetera.
12 MEMBER BLEY: Three activities --
13 MR. YARSKY: Yes.
14 MEMBER BALLINGER: But it's based on an 15 enthalpy insertion, right?
16 MEMBER KIRCHNER: No. It's also 17 temperature and core coolability and how that is defined 18 and whether you get to ballooning or not.
19 MEMBER BALLINGER: Yes. That's for sure 20 there.
21 MEMBER KIRCHNER: I just ask you to look 22 at it because --
23 MR. YARSKY: Yes. Certainly if there's, 24 if we can translate something like the flow area 25
58 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 reduction due to ballooning and burst. Or the 1
population of rods above a certain temperature or above 2
a certain enthalpy, if that can be translated to a 3
coolability determination, there's information about 4
how to do that already.
5 (Simultaneous speaking.)
6 MR. YARSKY: That's step two that we were 7
talking about earlier.
8 MEMBER KIRCHNER: But I think that's where 9
that particular Reg Guide is going. That's how they're 10 going to define core coolability. So a certain number 11 of that --
12 MR. YARSKY: Yes, I'll certainly look that 13 up.
14 MEMBER KIRCHNER: That are impacted by 15 reaching whatever threshold he is using in that Reg 16 Guide.
17 MR. YARSKY: Yes. That could certainly 18 be useful. I think once we expand on methodology, look 19 at many more rod groups then that's a natural 20 progression.
21 CHAIRMAN MARCH-LEUBA: Food for thought.
22 This expansion of a group may not even be necessary 23 to put into TRACE. It can be a post processing result.
24 Once you have the flow of the pressure of the void 25
59 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 of the bundle, you can just put this rods, second rod 1
on this. You don't have to complicate TRACE more than 2
it already is.
3 MR. YARSKY: Well, when we get to the 4
preliminary analysis I think we'll show you some of 5
what's happening in the bundle and these phenomena are 6
pretty complex and in many ways subtle.
7 So I would really like to leave it up to 8
the code to really analyze that. And I would even go 9
as so far as to strap in power reconstruction onto it 10 as well.
11 CHAIRMAN MARCH-LEUBA: That would be 12 great. But you're talking many, much CPU.
13 MR. YARSKY: It's true. It's true. But 14 I think if that's the question we want to answer. Now 15 if licensees continue to present results where they 16 have very quick operator action timing, and we'll get 17 to a slide to talk about other things that are specific 18 for a plant's specific evaluation.
19 If they continue to provide analysis that 20 demonstrate they avoid this heat mechanism altogether, 21 then the subtler points of that may not be --
22 PARTICIPANT: Be necessary.
23 MR. YARSKY: -- may not be necessary. But 24 if we are going to have to answer the questions like 25
60 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 if two percent or three percent of the core is damaged, 1
does the core remain in a coolable geometry. I think 2
we need to do more as terms of our methodology before 3
we can be prepared to perform that kind of confirmatory 4
analysis.
5 CHAIRMAN MARCH-LEUBA: I remember being 6
on that side of the table making the argument that you 7
are much better off taking positive steps of training 8
the operators to lower the fuel water faster than 9
sharpening your pencil and doing better calculations.
10 And I know you're in the business of doing 11 calculations. But from the point of your sector CO-of 12 the reactor, I'm much more happy training my operators 13 to lower the water level in 90 seconds then you saying 14 that you have now had operators that trust me, 120 is 15 okay.
16 MR. YARSKY: Yes. But I think those of 17 you who appreciate this from the perspective of Office 18 of Research, we have to look at and do the confirmatory 19 analysis that's necessary based on what the applicant 20 proposes, or what the licensee proposes.
21 (Simultaneous speaking.)
22 MR. YARSKY: And so that's really not up 23 to the Office of Research.
24 CHAIRMAN MARCH-LEUBA: I'm sorry. Maybe 25
61 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 we can finish today sometime however we move, because 1
the important part, the interesting part is in the 2
closed session.
3 MR. YARSKY: In the closed session. I'm 4
going to spend one minute on the slide. Less than that.
5 Thirty seconds. I just wanted to go through a list 6
of those parameters that can really change the analysis 7
on a plant specific application just to give you an 8
awareness that, you know, if you see a different 9
application, these results can be very different 10 because all of these factors can have a significant 11 impact on how the event progresses and then what the 12 consequences of the events are.
13 Of course the fuel in quartazine 14 (phonetic), that's going to affect the stability 15 characteristics of the core. It's going to affect the 16 relative stability characteristics of the regional 17 versus the core-wide mode. The turbine bypass capacity 18 is a factor.
19 In our analysis we looked at different 20 turbine bypass capacities and found there are certain 21 competing effects with respect to turbine bypass 22 capacity. That this can have an impact on the event.
23 One of the most significant is manual operator action 24 timing. And I think we've talked about this over the 25
62 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 course of the presentation.
1 There's also the impact of the Slick 2
injection location and boron enrichment. We talked 3
about a little bit. There are advantages to upper 4
plenum injection over lower plenum injection for 5
instance.
6 If the feed water pumps are motor versus 7
steam driven can have a significant impact. Steam 8
driven feed water pumps will trip when they lose, when 9
the turbine trips which will give a benefit in terms 10 of the early reactor water level response.
11 There's also the issue of the feed water 12 heater cascades. We talked about how feed water 13 temperature is what sort of brings the power up and 14 leads to the instability of that. How much thermal 15 inertia is in that feed water heater cascade can have 16 a significant impact.
17 Lastly, just to mention it. In the case 18 of Nine Mile Point 2, if you design and implement 19 automatic protective features, that of course will have 20 a significant impact.
21 So in summary, to put what we're going to 22 talk about in the closed session into context is that 23 for MELLA+ BWR plants, instability during ATWS events 24 is likely to occur and expected to occur earlier in 25
63 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 the event.
1 If large amplitude power oscillations 2
occur during an ATWS-I event prior to mitigation, the 3
fuel may experience cyclic dryout/rewet. During that 4
cyclic dryout/rewet phase, if temperature ratcheting 5
occurs we predict that there could be a temperature 6
excursion leading to fuel damage.
7 And that's all I have for the open session.
8 CHAIRMAN MARCH-LEUBA: Great. So at this 9
point we are going to close this session and we'll not 10 have any phone open any longer. So I want to ask for 11 questions in the room. Somebody wants to make any 12 comments or questions.
13 And is the phone line open? It's supposed 14 to be. Anybody on the phone line would like to make 15 a comment? Another question?
16 (Off microphone comments.)
17 CHAIRMAN MARCH-LEUBA: Can you verify it's 18 open? Because I don't think -- when it's open, you 19 can hear.
20 Okay. The line was open. There is nobody 21 listening. So at this point we are going to close the 22 open session and we are going to take a 15 minute recess 23 until 2:25.
24 (Whereupon, the above-entitled matter went 25
64 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 off the record at 2:26 p.m. and resumed at 4:13 p.m.)
1 CHAIRMAN MARCH-LEUBA:
And we're 2
definitely on the record now. This part is open 3
session. We are going to describe the conclusions.
4 We just opened the phone line. If anybody 5
is on the line, can you please identify yourself?
6 (No audible response.)
7 CHAIRMAN MARCH-LEUBA: Even though you 8
hear crackles that means nobody on the mind but -- is 9
anybody on the line? We still have it open. This is 10 open session.
11 Peter, continue with the conclusions.
12 MR. YARSKY: Okay, thank you.
13 In terms of conclusions, I just wanted to 14 provide a brief summary of what we had talked about 15 up until now. The first is that the KATHY test loop 16 can closely match expected prototypical conditions of 17 ATWS-I stability.
18 The NRC test bundle that we used in the 19 conduct of these tests include many features of modern 20 BWR fuel assemblies.
21 And we performed tests both with and 22 without simulated neutronic feedback and this was to 23 determine the conditions for instability, cyclic 24
65 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 dryout/rewet, and failure-to-rewet conditions.
1 These, of course, being prototypic for modern fuel 2
assemblies operated and experiencing conditions with 3
MELLLA+ ATWS-I.
4 Based on our preliminary analysis and 5
comparisons of the temperature data from the KATHY 6
tests, it appears to indicate that a failure-to-rewet 7
temperature is in reasonable agreement with a model 8
of homogeneous nucleation temperature plus contact 9
temperature.
10 And until the research staff can complete 11 our TRACE assessment and complete a more thorough 12 analysis of the experimental results, we have 13 recommended using a Tmin model based on homogeneous 14 nucleation plus contact temperature in TRACE 15 predictions of ATWS-I consequences for MELLLA+ BWRs.
16 Our area of future work is to conduct the 17 inverse heat conduction analysis of the KATHY heater 18 rods and this will be so that we can determine heat 19 transfer coefficients during the nucleate boiling, 20 transition boiling, film boiling, and quenching regimes 21 that occur during the event.
22 As applicable, we will provide new 23 correlations based on our improved understanding of 24
66 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 the mechanisms affecting failure to rewet.
1 We will perform the TRACE assessment 2
against the experimental data.
3 And lastly, we'll be able to conduct 4
sensitivity analyses using TRACE with modifications 5
so that we can look at the impact these models of Tmin 6
and film boiling heat transfer have and determine if 7
alternative models to those that are currently present 8
in TRACE provide a better fit to the data under these 9
conditions.
10 CHAIRMAN MARCH-LEUBA: Are you able to 11 find TRACE to use the homogeneous nucleation for Tmin?
12 MR. YARSKY: Yes.
13 CHAIRMAN MARCH-LEUBA: And including the 14 quenching of Tmin? Because right now TRACE only 15 quenches on axial conduction.
16 MR. YARSKY: It would --
17 MEMBER CORRADINI: It switches.
18 MR. YARSKY: Yes.
19 MEMBER CORRADINI: It switches at this 20 temperature threshold.
21 CHAIRMAN MARCH-LEUBA: I'm not sure it 22 does.
23 MR. YARSKY: It would go into transition 24
67 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 boiling.
1 CHAIRMAN MARCH-LEUBA: And can you refine 2
that? Because I remember a lot of discussions with 3
one of the vendors about how the axial conduction was 4
working. I know that's how they do it.
5 MEMBER CORRADINI: Oh, you have to have 6
only the axial conduction model on, you think?
7 CHAIRMAN MARCH-LEUBA:
The axial 8
conduction is the only one that gets you off -- gets 9
you down to --
10 MEMBER CORRADINI: But once you pass you 11 12 MR. YARSKY: Jose, we had a closed session, 13 where we could have freely discussed different models 14 of quenching. And I am not comfortable in an open 15 discussing the specifics of how other --
16 CHAIRMAN MARCH-LEUBA: But of TRACE.
17 MR. YARSKY: -- vendors --
18 CHAIRMAN MARCH-LEUBA: I am asking about 19 TRACE.
20 MR. YARSKY: Even without the fine mesh 21 axial conduction model, the heat transfer regime would 22 go from film boiling to transition boiling.
23 CHAIRMAN MARCH-LEUBA: But how do you cool 24
68 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 it?
1 MR. YARSKY: In the transition boiling 2
heat transfer regime, you would have an improved heat 3
transfer coefficient.
4 CHAIRMAN MARCH-LEUBA: In transition, 5
yes. But if you are above Tmin in TRACE --
6 MR. YARSKY: Yes.
7 CHAIRMAN MARCH-LEUBA: -- now you are 8
flooded with water. You will still have film boiling.
9 MR.
YARSKY:
Right.
So there's 10 conduction but there is also film boiling heat transfer.
11 So for instance, as temperature increases, you still 12 get improved heat transfer because the film boiling 13 heat transfer coefficient is not zero.
14 CHAIRMAN MARCH-LEUBA: Okay.
15 MEMBER CORRADINI: I think what he's just 16 asking is if I had the heat flux on constantly, there's 17 no way to cross back over in that regime.
18 CHAIRMAN MARCH-LEUBA: That's correct.
19 MEMBER CORRADINI: That's what he was 20 asking. If I pass the boundary and I have a re-flood, 21 there's no way to go backwards.
22 MR. YARSKY: Right. Right. That would 23 physically be the case if there was no --
24
69 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 CHAIRMAN MARCH-LEUBA: Well, when we do 1
the detailed modeling with TRACE of your data, I saw 2
some of your data that keeps the temperature of 3
thermocouples high when in reality, it went down faster.
4 So and that is -- I don't know what 5
mechanism it is but what is the axial conduction that 6
TRACE already has, it is an expensive --
7 MEMBER KIRCHNER:
It's not axial 8
conduction.
9 CHAIRMAN MARCH-LEUBA: It's not?
10 MEMBER KIRCHNER: You've gone into a flow 11 regime where there's ample amounts of water and you'll 12 get a collapse of film boiling en masse.
13 CHAIRMAN MARCH-LEUBA: But we are talking 14 about Tmin.
15 MEMBER CORRADINI: But I think in his case, 16 again, I'm just -- in his -- his -- yours -- KATHY case, 17 I think my memory is one of the models in TRACE allows 18 for the fact there's a mass flow rate tenet. Since 19 you've got this oscillatory mass flow, you come back 20 into a regime where you cool and you --
21 MR. YARSKY: Yes, well like for instance 22 23 MEMBER CORRADINI: That's what I thought.
24
70 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MR. YARSKY: -- you would go from a 1
dispersed flow film boiling heat transfer regime to 2
an inverted annular film boiling heat transfer regime, 3
which would affect a change in the heat transfer 4
coefficient once you had a re-flooding condition.
5 So even with just film boiling --
6 MEMBER KIRCHNER: You don't need to have 7
film conduction model for that. It's not even --
8 MEMBER CORRADINI: I mean it would drag 9
down the speed of calculation.
10 MEMBER KIRCHNER: Right.
11 MEMBER CORRADINI: You could have it 12 highly hyper-nodalized but I think you'd still --
13 MEMBER KIRCHNER: When you're in those 14 conditions, it's not like the wet front just comes right 15 down from the top or the bottom. You just collapse 16 the film boiling regime because of the presence of 17 water.
18 CHAIRMAN MARCH-LEUBA: Well that's what 19 the channel is. I'm asking what the numerics do. I've 20 seen him not do that.
21 MR. YARSKY: For that to occur in TRACE, 22 the inverted annular film boiling heat transfer 23 coefficient would have to be sufficiently high to cool 24
71 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 the temperature below Tmin before you have film 1
collapse.
2 CHAIRMAN MARCH-LEUBA: That's correct.
3 And --
4 MR. YARSKY: And there's, of course, axial 5
conduction can be modeled in different resolutions in 6
TRACE. You can have the fine mesh axial conduction 7
model or you can have a course mesh axial conduction 8
model.
9 CHAIRMAN MARCH-LEUBA: Just keep it in 10 mind. Let's not discuss it. Let's not fix the models 11 here but keep it in mind.
12 MR. YARSKY: Yes but one of the things that 13 we will be doing, as we do the TRACE assessment, is 14 we will be looking at you know what is the effect of 15 having the fine mesh axial conduction, fine mesh axial 16 conduction with very fine nodes, coarse mesh axial 17 conduction, and no axial conduction.
18 CHAIRMAN MARCH-LEUBA: The bottom line is 19 that the time after scram, TRACE has to model it, too.
20 When you do your evaluation, let's not 21 model just the time before the scram on KATHY but the 22 time after the scram, when it re-wets. TRACE has to 23 follow that. It would be nice if it did.
24
72 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MR. YARSKY: It would be very nice if TRACE 1
did. In our current preliminary analysis, of course, 2
our focus is on diagnosing and understanding the 3
incidents of failure to re-wet.
4 What happens afterwards is something that 5
we will have to look at as future work. If there are 6
phenomenon that occur on a very localized level, like 7
for instance, quenching, and quenching ends up being 8
very important, then the KATHY facility experimental 9
data, because of how far apart the thermocouples are, 10 might not be the best data to support some sort of 11 conclusion where we start rethinking quenching.
12 CHAIRMAN MARCH-LEUBA: Well you cannot see 13 the front --
14 MR. YARSKY: Right so a different 15 experimental basis, if we wanted to really study 16 quenching behavior might be appropriate.
17 CHAIRMAN MARCH-LEUBA: Yes, don't ignore 18 the time after scram on your evaluation. And again, 19 this is --
20 MR. YARSKY: We won't. But if there is 21 an indication that TRACE is conservative, that may be 22 a conclusion in itself.
23 CHAIRMAN MARCH-LEUBA: Yes.
24
73 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MR. YARSKY: We may not have sufficient 1
data to justify changing the quench model because the 2
thermocouples are arrayed in such a manner that it may 3
be difficult to infer information about a quenching 4
using this particular experiment.
5 CHAIRMAN MARCH-LEUBA: As long as we don't 6
ignore the data and we address it, even if it is by 7
waving your hands, I am happy.
8 MR. YARSKY: Okay.
9 MEMBER REMPE: On your future work, I don't 10 see anything about what you discussed earlier today 11 about input on what debris coolability means. And is 12 that because maybe it's not so important?
13 MR. YARSKY: Oh, one thing that the Office 14 of Research is looking into is developing a capability 15 in TRACE to calculate the population of the core that 16 has experienced damage.
17 MEMBER REMPE: Right.
18 MR. YARSKY: That's not within the scope 19 of this. It would be -- so this project is being driven 20 by using these requests from the Office of NRR.
21 Separate to that and part of a long-term research 22 strategy is to develop this capability in anticipation 23 of future incoming requests.
24
74 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER REMPE: I'm just wondering if there 1
will be that many. I said the wrong word. I said 2
debris coolability but for core coolability. I'm just 3
wondering if some of those requests would go down, in 4
light of the work. I mean I would think that that issue 5
might not come up as much, in light of some of the work 6
that's done. But you think it will still be important 7
for --
8 MR. YARSKY: Well --
9 MEMBER REMPE: Because if that was high, 10 it might not be so important anymore.
11 MR. YARSKY: In my thinking, at least 12 observing how things have gone from MELLLA to EPU to 13 MELLLA+ is that based on say an incomplete understanding 14 of these phenomena led us to look at intermediate 15 criteria, for instance, looking at temperature criteria 16 short of the regulatory criterion of core coolability 17 to say if you can meet say some limits on temperature 18 like PCT limits, then you're able to demonstrate 19 compliance with the regulatory criteria. But this has 20 resulted in certain applicants, for instance, changing 21 operator action timings and then making regulatory 22 commitments to adhere to more restrictive operator 23 action timing requirements.
24
75 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 So, I could foresee in the future analyses 1
a pencil sharpening being done to further expand an 2
operating domain or to further relax those conditions, 3
in which case the real criterion, the real regulatory 4
criterion is that core coolability question.
5 MEMBER REMPE: Right but there's only, 6
again, that many -- there is a limited number of plants 7
even coming in for EPUs that are BWRs.
8 MR. YARSKY: I think most of the BWRs went 9
for EPUs, except for the twos. I imagine we would see 10 a similar influx for the MELLLA+s.
11 MEMBER REMPE: Really? Because I guess 12 I was thinking that there wouldn't be that many more.
13 The ones who have come in for EPUs have done it, pretty 14 much.
15 MR. YARSKY: Yes.
16 MEMBER REMPE: And then there's a limited 17 number that have come in for MELLLA+ and those are kind 18 of going a certain way with methods they're selecting, 19 et cetera. So I just am wondering if they try and expand 20 that region, expand the flow region, I guess there might 21 be more. I just was curious.
22 MR. YARSKY: Yes, it's just my gut feeling.
23 It may be completely inaccurate. I don't know.
24
76 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER REMPE: Yes, the other thing is with 1
the future work, there is this plant-specific 2
evaluation that will be supporting a plant coming in.
3 I think you said it's on our agenda for April. It's 4
coming into you guys sooner, right?
5 MR. BORROMEO: Oh, we have it and we're 6
currently reviewing it now.
7 MEMBER REMPE: And I'm wondering, with 8
respect to interacting with us, would we best do the 9
results of your analysis as part of our evaluation in 10 April or should we plan to have a discussion earlier 11 to talk about some of the insights from your 12 plant-specific analysis?
13 MR. YARSKY: We were discussing having 14 research meet with the Thermal-Hydraulic Subcommittee 15 in advance of the LAR but I think that that's something 16 that we need to continue to discuss internally and also 17 with the ACRS staff to figure out which is the best 18 plan, in terms of schedule and everything else.
19 MEMBER REMPE: If you think it's a half 20 day, then definitely, I think it ought to be done 21 separately. If it's an hour discussion, we can work 22 it all in together. But I just was curious now because 23 I'm involved in the one that's coming to us in April 24
77 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 and I wondered if we should be --
1 MR. YARSKY: I think the staff needs to think 2
about it a little more in terms of -- especially in 3
light of the presentation we've given today and other 4
presentations that we've made, if we might not just 5
be repeating ourselves if we come in for half a day.
6 MEMBER REMPE: Okay. Well I definitely 7
would think I would want to have an hour during those 8
discussions of the plant.
9 MR. YARSKY: Oh, certainly. Yes, I think 10 it's something that just needs to be sorted out. The 11 direction we were leaning in was to have a separate 12 meeting with the Thermal-Hydraulic Subcommittee.
13 MEMBER REMPE: Okay.
14 MR. YARSKY: But I don't think any final 15 determination yet has been made with respect to the 16 ACRS schedule.
17 MEMBER REMPE: So Mike and Jose --
18 CHAIRMAN MARCH-LEUBA: That would be a 19 generic presentation on TRACE calculations?
20 MR. YARSKY: No, a presentation on the 21 Brunswick plant-specific confirmatory analysis.
22 CHAIRMAN MARCH-LEUBA: But only TRACE 23 calculations.
24
78 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MR. YARSKY: Only TRACE calculations only 1
Brunswick.
2 CHAIRMAN MARCH-LEUBA: That would be nice.
3 MEMBER REMPE: When is it?
4 MR. BORROMEO: April 20th.
5 MEMBER REMPE: That's our subcommittee 6
meeting. That's actually on the application, the LAR.
7 So you're planning to fold it in is what I'm hearing.
8 MR. YARSKY: We're leaning separate.
9 MS. ABDULLAHI: And this one is AREVA.
10 Okay, Brunswick is AREVA methods, right?
11 MEMBER REMPE: Right.
12 MR. BORROMEO: But AREVA and GE methods 13 together. So that's what makes it fun.
14 MS. ABDULLAHI: Which part of it are GE 15
-- oh, we can't talk about it.
16 Okay. Just for the ATWS-I part, it is 17 AREVA, I think. And so there will be something 18 different about it.
19 MR. YARSKY: Well, the plant-specific 20 confirmatory analysis performed by the Office of 21 Research would not address any of the -- like we would 22 be presenting our results, right?
23 CHAIRMAN MARCH-LEUBA: So if I have a vote 24
79 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 on that, if we could have his presentation in the middle 1
of the Brunswick MELLLA+ presentation with the 2
applicant, and with the vendors, and everything, we 3
won't have enough time. You see how this thing goes.
4 MEMBER REMPE: Well okay, we've gone 5
through a lot of power uprates in the past, right?
6 CHAIRMAN MARCH-LEUBA: Yes, the other part 7
about power rate will go fast. It's his part.
8 MEMBER REMPE: Yes, well, again, if we're 9
going to do it before April 20th, it's not on the agenda, 10 to my knowledge. And so that's something we ought to 11 get going soon because April's not that far away, 12 frankly, for a schedule change.
13 MR. YARSKY: Well, certainly we have to 14 complete the confirmatory analysis. There's a lot of 15 work to be done there.
16 MEMBER REMPE: So yes, how soon --
17 MEMBER CORRADINI: I'm not sure, Joy --
18 I'm not sure why inserting it at this point is necessary.
19 I think we have the general direction. You want it 20 because of why?
21 MEMBER REMPE: Well, they are going to the 22 effort to do a plant-specific Brunswick evaluation.
23 Do we want to have it before the power uprate discussion 24
80 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 on April 20th or not? That's why I'm bringing it up.
1 MS. ABDULLAHI: Is that EPRI or MELLLA+?
2 MR. YARSKY: MELLLA+.
3 MEMBER REMPE: You're right, it's a 4
MELLLA+ but they are using the AREVA methods --
5 MR. YARSKY: I think it's the Power Uprate 6
Subcommittee.
7 MEMBER REMPE: Yes, it's the Power Uprate 8
Subcommittee but it's a MELLLA+ evaluation for 9
Brunswick, which already has its EPU.
10 And
- again, if they've got this 11 plant-specific analysis, it sure seems like that we'd 12 want to hear about it as part of our review of the 13 MELLLA+, right?
14 MEMBER CORRADINI: And Jose doesn't think 15 it will fit within the time period of --
16 CHAIRMAN MARCH-LEUBA: It's not a one-hour 17 presentation. See how it went here with the --
18 MEMBER CORRADINI: But wait. It doesn't 19 have to be. We're now planning the future. I'm not 20 sure if this is the appropriate way to do it. But why 21 does it have to be like he did it here? It doesn't 22 have to be.
23 MEMBER REMPE: Well, I don't know how much 24
81 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 material there will be and I just am bringing it up 1
because we need to have it on the agenda.
2 MEMBER CORRADINI: Well I mean reverse 3
this thing. Schedule it in, you have an hour, they 4
figure out what they want to present in the hour, and 5
that's the end of it.
6 MS. ABDULLAHI: But are they going to keep 7
the schedule?
8 MR. BORROMEO: We're committed to April 9
20th or 19th I think I just saw in my email.
10 MEMBER POWERS: It is fairly well-known, 11 Mike, it is fairly well-known that predictions are 12 difficult, especially about the future.
13 MEMBER CORRADINI: Okay.
14 MEMBER REMPE: You are still completing 15 the Brunswick analysis. It's not done yet?
16 MR. YARSKY: The Brunswick analysis is not 17 complete at this point.
18 MEMBER REMPE: Do you have an estimated 19 date for when you think it will be?
20 MR. YARSKY: We have a revised estimated 21 date that we have to finalize.
22 MEMBER CORRADINI: So are you telling --
23 this is going to be part of the staff's analysis anyway 24
82 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 to accept what's being done anyway, right? So does 1
that imply that everything's going to be delayed?
2 MR. YARSKY: No.
3 MEMBER CORRADINI: Oh. All right.
4 MEMBER REMPE: So you may not have it done 5
before the staff finishes the April --
6 MR. BORROMEO: Oh, no, they'll have it.
7 Research has committed to having it done before the 8
April meeting.
9 MR. YARSKY: Right.
10 MR. BORROMEO: And I think we said sometime 11 of early next year, right?
12 MR. YARSKY: Yes.
13 MEMBER REMPE: And it will be used in your 14 evaluation that you're going to be presenting to us.
15 MR. BORROMEO: Yes.
16 MEMBER REMPE: So we definitely want to 17 have something at the meeting on it, if we don't do 18 it separately.
19 MR. YARSKY: Absolutely.
20 MEMBER POWERS: Can I put my vote in for 21 sometime after May?
22 MEMBER REMPE: No.
23 MR. YARSKY: It would certainly -- I think 24
83 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 it would not fruitful for us to present the analysis 1
after the April meeting.
2 MEMBER REMPE: I would think so. But I just 3
was trying to get some facts here. And so again --
4 MR. YARSKY: We are not far enough along 5
in the analysis for me to provide a hard date of when 6
the analysis will be complete. After the Thanksgiving 7
holiday, we have had several meetings to align the team 8
that's working on it. Because this is a team that is 9
almost essentially the entire Reactor Systems Analysis 10 Group working on different aspects of it.
11 When we have a better idea of the date, 12 we will be working with the ACRS staff to schedule time 13 to talk to you. And what we had been discussing with 14 NRR and internally was having a separate meeting with 15 the Thermal-Hydraulics or Power Uprate Subcommittee 16 to present the plant-specific results in advance of 17 the LAR meeting.
18 MEMBER REMPE: March looks pretty open for 19 the calendar for ACRS and so I'd shoot hard for that 20 but you'd need to get stuff to us then ahead of time.
21 MR. YARSKY: Right. We need to complete 22 the analysis. There are other scheduling things so 23 we just need to consider the staff level.
24
84 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 CHAIRMAN MARCH-LEUBA: You have the 1
Christmas break to work on it.
2 MR. YARSKY: I'm going to be honest with 3
you, Jose. I am not doing the heavy lifting on the 4
Brunswick plant-specific analysis. So I am kind of 5
plugged into that activity but I am not the technical 6
lead on any of the --
7 CHAIRMAN MARCH-LEUBA: So you are not 8
opposed against them working over Christmas.
9 MR. YARSKY: So I would not be the person 10 that would be working over the holiday.
11 MEMBER REMPE: Anyway, I just was curious 12 because I would like to kind of understand the timing.
13 MR. YARSKY: Absolutely. And as we get 14 a better idea on schedule, we'll get on the calendar.
15 If the scope of what we're coming up with 16 that we would want to present starts to look like it 17 would fit in an hour, then maybe we could reconsider 18 the meeting with you guys separately beforehand. But 19 it seems the indication I'm getting from Jose is to 20 meet with you beforehand anyway.
21 CHAIRMAN MARCH-LEUBA: I can see this 22 going half a day easily. I cannot see how we can do 23 this in one hour.
24
85 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 MEMBER REMPE: The MELLLA+ discussions are 1
getting easier. The issue with this is the --
2 CHAIRMAN MARCH-LEUBA: I don't have any 3
problem with MELLLA+. I have got a problem with TRACE 4
calculations. They will be interesting. Everybody 5
likes them.
6 MEMBER REMPE: I appreciate the time to 7
discuss it because it will draw our attention to what 8
we need to figure out what we're doing.
9 MEMBER POWERS: It definitely sounds to 10 me like June subcommittee would be really good.
11 MEMBER REMPE: And we can just delay your 12 retirement because Shirley has asked me to find a way 13 to do that.
14 CHAIRMAN MARCH-LEUBA: Okay, I am going 15 to use the power of the office. I am going to call 16 it closed. Since we already called for comments 17 earlier, we don't have to do it now.
18 So this meeting is adjourned.
19 (Whereupon, the above-entitled matter went 20 off the record at 4:35 p.m.)
21
BACKGROUND ON MELLLA+, ATWS-I, AND FAILURE TO REWET PHENOMENOLOGY Dr. Peter Yarsky US NRC Office of Nuclear Regulatory Research 1
OUTLINE
- 1. Description of the maximum extended load line limit analysis plus (MELLLA+) domain.
- 2. Safety significance of the MELLLA+ domain.
- 3. Description of ATWS-I.
- 4. TRACE calculation results.
- 5. Predicted mechanism if fuel heat-up.
- 6. Plant-specific considerations.
2
MELLLA+ DOMAIN
- Maximum Extended Load Line Limit Analysis Plus (MELLLA+) is an expanded BWR operating domain allowing high thermal power (120%) at low flow (80%)
- MELLLA+ operation introduces new aspects to the progression of Anticipated Transient Without SCRAM (ATWS) events 3
SAFETY SIGNIFICANCE OF THE FCW
- During ATWS events, the reactor power is decreased by a trip of the recirculation pumps (2RPT).
- The power and flow decrease as the pumps run down.
- Power then increases due to a decrease in feedwater temperature.
- When the flow rate is low (80 %RCF), the 2RPT becomes less effective in the reduction of gross core power.
4
OPERATING DOMAIN AND RPT 5
OPERATING DOMAIN AND RPT 6
OVERVIEW OF ATWS-I
- ATWS event considered is a turbine trip event with turbine bypass capability (TTWBP).
- The TTWBP results in a pressure pulse, a trip of the recirculation pumps, and a loss of extraction steam to the feedwater heater cascade.
- The TTWBP ATWS is expected to yield unstable conditions and large amplitude power instability.
- Operators control reactor water level and inject boron through the standby liquid control system (SLCS) to mitigate the event.
7
ATWS-I RESULTS Representative Case:
Generic BWR/5 model TTWBP with 100% bypass capacity Initial core flow rate is 85% rated Initial power is 120% of originally licensed thermal power (OLTP)
Core exposure is peak-hot-excess (PHE)
Operators attempt to control reactor water level to top of active fuel (TAF) starting 110 seconds into event Operators initiate SLCS at 120 seconds into event 8
PHE ATWS-I CASE - SEQUENCE OF EVENTS 9
TRANSIENT REACTOR POWER 10 Oscillation growth begins around 100s Oscillations are damped by operator actions. They are indiscernible after
~230s
BI-MODAL OSCILLATIONS 11
VISUALIZATION PHE - Case 2 12
PEAK CLADDING TEMPERATURE RESULTS 13 PCT exceeds 2200°F after onset of non-linearity. High PCT observed before that point.
Cyclic dryout/rewet 1478K
BASE CASE CONCLUSIONS Point in cycle studies confirm that PHE is the most limiting state-point Large amplitude regional power oscillations develop (modal coupling with frequency doubling).
High amplitude power oscillations (local) results in calculation of high PCT
(~1700 K [2600 °F]).
Operator action to reduce level Effective in reducing FW flow, limiting increase in core inlet subcooling and eventually eliminating inlet subcooling.
Operator action to inject boron through SLCS Effective in suppressing power oscillations and reducing core power level.
14
PREDICTED FUEL HEAT-UP MECHANISM
- Oscillation magnitude increases and the fuel undergoes periodic dryout/rewet cycling.
- As oscillation magnitude continues to grow, the rewet period of the cycle becomes insufficient to remove all of the energy accumulated in the fuel during the dryout period. This is accompanied by a ratcheting of the fuel temperature upwards after each dryout/rewet cycle.
- Once temperature ratchets up to the minimum stable film boiling temperature, the cladding surface locks into film boiling heat transfer.
- Once locked in film boiling, and while reactor power is high, fuel temperature excursion occurs.
15
PLANT SPECIFIC FACTORS AFFECTING EVENT PROGRESSION AND CONSEQUENCES
- Fuel and core design.
- Turbine bypass capacity.
- Manual operator action timing.
- SLCS injection location and boron enrichment.
- Feedwater pumps (motor vs. steam driven).
- Feedwater heater cascade thermal inertia.
- Automatic protective features (i.e., NMP2).
16
SUMMARY
1.
Instability is likelier to occur and is expected to occur earlier for MELLLA+ plants during ATWS, meaning that operator actions designed to mitigate the instability may not be effective before large amplitude power/flow oscillations develop.
2.
If large amplitude power/flow oscillations develop during the ATWS-I event, prior to mitigation, the fuel may expect cyclic dryout/rewet.
3.
During the cyclic dryout/rewet phase, the temperature may increase due to ratcheting, resulting in the fuel failing to rewet.
Failure to rewet leads to excessive fuel heat-up and fuel damage.
17