ML082550409
ML082550409 | |
Person / Time | |
---|---|
Site: | Vermont Yankee ![]() |
Issue date: | 07/21/2008 |
From: | - No Known Affiliation |
To: | NRC/SECY/RAS |
SECY RAS | |
References | |
06-849-03-LR, 50-271-LR, Entergy-Applicant-E2-36-VY, RAS M-299 | |
Download: ML082550409 (59) | |
Text
0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 90 CHAIRMAN SHACK: We are just a little bit ahead of schedule, five minutes. But we'll go ahead and take our break until 10:45.(Whereupon at 10:22 a.m. the proceeding in the above-entitled matter went off the record to return on the record at 10:44 a.m.)CHAIRMAN SHACK: I think we can come back into session.Our next topic is the final review of the license renewal application for the Vermont Yankee Nuclear Power Station.And Dr. Bonaca is lucky enough to lead us through this again.FINAL REVIEW OF LICENSE RENEWAL APPLICATION FOR VERMONT YANKEE NUCLEAR POWER STATION VICE CHAIRMAN BONACA: It was kind of hard to keep FitzPatrick and Vermont Yankee apart.We met a month ago to review the application for license renewal for Vermont Yankee.And I believe we covered pretty much every item of the agenda having to do with license renewal.There was one remaining item that was left because of the time; we did not have
-'Ahd NEAL R. GROSS.COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C.- 20005-3701 www.nealrgross.com DOCKETED USNRC August 12, 2008 (11:00am)OFFICE OF SECRETARY RULEMAKINGS AND ADJUDICATIONS STAFF , _J J-*-ý kk-Q4 I oz, 9--
U.S. NUCLEAR REGULATORY OMSSION ,,eMaftrof
..IJU., " \44r., LL Docket No.j icial Exhibit No. 9 Z1&--l-OFFERED by, plicantlLicensee Intervenor
~NRb-i-I Other IDENTIFIED on-iL C Witness/Panel OEC ., Action Taken: (!DjlD REJECTED WITHDRAWN Reporter/Clerk.
m -,f -C-91 1 it has to do with the environmentally assisted fatigue 2 calculations.
3 I would just summarize very briefly what 4 has happened since. Entergy has chosen to address 5 environmentally assisted fatigue by demonstrating that 6 CUF and the most sensitive locations would remain 7 below one throughout the period of extended operation 8 considering both mechanical and environmental effects.9 The analysis performed by the licensee are 10 supported also by assumptions that will be monitored 11 and verified during the period of extended operation.
12 The analysis performed by the licensee had 13 confirmed ,that in all locations CUF is going to be 0 14 below one throughout the period of extended operation.
15 This staff however has challenged the use of the 16 simplified methodology used by the licensee for those 17 locations which exhibit geometric discontinuities or 18 no symmetric loads such as the feedwater nozzle for 19 example or the circulation out that nozzle and the 20 coarse spray line nozzle.21 At the request of the staff the licensee 22 has performed an analysis for the limiting location 23 which is the feedwater nozzle, using the methodology 24 at our command which is using ASME code Section 3.0 25 The analysis has confirmed that CUF will be below one NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 92 okay through the period of extended operation.
However I believe assuming the same environmental multiplier, the result with more analysis show a higher value of CUF though below one. And so the staff has requested the licensee to perform also the corresponding analysis for the two additional cases where there are geometric discontinuities or no symmetric loads and essentially the locations are the circulation outlet nozzle and the coarse spray line nozzle.Today I believe the licensee wants to present their methodology and make the case for the analysis they performed originally..
I believe the issue so far as the SER is closed in the sense that they have committed to perform the two additional analyses as requested by the staff.But we will hear both from the licensee and the staff about this contention and it's an important issue because it may affect other licensees that have performed calculations before using the same methodology used by Vermont Yankee.We would like to introduce and turn over to PT Kuo.MR. KUO: Thank you, Bonaca.Yes, this is indeed the last issue for the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.WASHINGTON, D.C. 20005-3701 www.nealrgross.com (202) 234-4433 93 0 1 2 3 4 5 6.7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0 Vermont Yankee license renewal application review.It has taken a long time, longer than what we would like to, but I think at this point we believe that the applicant has done what we have asked for, and we are satisfied with what they have done.We have supplemented our SER with our writeup. It's just I believe a week or so ago. And sent it to the committee members.I believe that right now with the additional calculations the applicant has done we consider this issue is resolved, and the applicant will first give you the story of how it is resolved, and the staff will also give you the reason, the basis of why we think this is acceptable.
Thank you.With that, applicant, please, take over.MR. DREYFUSS:
Good morning.Thank you, Dr. Bonaca, Mr. Chairman, members of the committee.
My name is John Dreyfuss.
I'm the director of nuclear safety assurance for Vermont Yankee.Before we get going with the presentation I do want to make sure that we introduce our Vermont Yankee and Entergy team here.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 0
94 1 First, I'd like to recognize Ted Sullivan, 2 our site vice president.
3 MR. SULLIVAN:
Good morning. I'd like to 4 thank the committee for allowing us to be here today 5 to continue the discussion on our license renewal 6 application.
And I'd like the team to identify 7 themselves, and then we'll turn it back over to John.8 John's our lead spokesperson.
9 MR. MANNAI: David Mannai, licensing 10 manager, Vermont Yankee.11 MR. RADEMACHER:
Norm Rademacher, 12 engineering director.13 MR. FITZPATRICK:
Jim FitzPatrick, design o 14 engineer.15 MR. STEVENS: Gary Stevens, structural 16 integrity associates, consultant to Entergy.17 MR. GOODWIN: Scott Goodwin, design 18 engineer.19 MR. METELL: Mike Metell, license renewal, 20 project manager.21 MR. YOUNG: Garry Young, manager of license 22 renewal for the Entergy fleet.23 MR. COX: Alan Cox, technical manager, 24 license renewal.25 MR. LOCK: Dave Lock, I'm part of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C: 20005-3701 www.nealrgross.com 95 1 Entergy license renewal team.2 MR. DREYFUSS:
All right, very good. Good 3 morning.4 Next slide, Beth, please.5 For the agenda for today we will go 6 through the environmentally assisted fatigue. And we 7 do recognize the last time we were here we went'8 through the rest of the SER and application and talked 9 about a lot of different issues.10 Our focus here on our presentation is as 11 requested on the fatigue issue.12 So we'll go through an overview of that, 13 some of the timeline, how we got to this point. We'll 0 14 talk about some of the bases, and go through both the 15 evaluation that we performed where there were 16 challenges from the staff, and confirmatory analysis.17 And just from a nomenclature standpoint, 18 I did want to mention, .a number of different terms 19 have been tossed out. What we will refer to during 20 the course of our presentation, we had original 21 analyses, for the license renewal we performed re-22 analysis.
I think we referred to that in the SER; you 23 may have seen the simplified analysis.
So we've 24 called it a re-analysis.
25 And then the confirmatory analysis that we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 96 I did I think is also referred to variously as the 2 updated analysis.
So for us re-analysis and 3 confirmatory and we'll step through that as we go 4 through the presentation.
5 I think the key thing to talk about is 6 that for the license renewal the confirmatory analysis 7 that we performed for the feedwater nozzle is the 8 calculation of record for license renewal.9 Additionally we'll talk about the license 10 condition.
We do have a license condition where we 11 will perform calculations, confirmatory calculations, 12 for the remaining two nozzles that were the subject of 13 the challenges, and we will perform those calculations O 14 prior to two years prior to entering into the extended 15 period of operations.
16 Next slide.17 From an overview standpoint we did, as far 18 as the full scope of environmentally assisted fatigue, 19 we did the locations that are identified in the 20 governing NUREG 6260, and that was the focus and the 21 basis for the calculations that we did do.22 Our original piping was designed to the 23 B31167 code so therefore we did not have the 24 calculations.
That is what drove why we had to do o 25 these calculations.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 97 1 From a timeline standpoint in September we 2 completed the re-analysis as well as all the rest of 3 the work that we did on environmentally assisted 4 fatigue. There was an audit by the staff of those 5 calculations in October. And really during the course 6 of that timeframe, from October through January of 7 2008, a lot of questions back and forth, and a number 8 of different RAIs and audits that were performed 9 questioning the approach that we had taken.10 And the key challenge was how we treated 11 stresses at the blend radius for these three 12 particular nozzles, coarse spray, reactor recirc and 13 feedwater.
14 So what we'll do during the course of the 15 presentation is, we'll talk about what we did on that 16 reanalysis, and provide you with the basis for that.17 We will also talk about what we did on the 18 confirmatory analysis as well.19 We did complete -we had requested a 20 public meeting. And that public meeting was held on 21 January 8', where we defined what approach we took 22 with the reanalysis method. At that meeting we also 23 said that we were working on a confirmatory analysis.24 for the feedwater nozzle.O 25 We did complete that analysis on the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 98 1 nozzle and submitted that on February 1 4 th -I'm 2 sorry, January 30", Ray. And NRC, Dr. Chang, did an 3 audit of that calculation on Valentine's Day of 2008.4 We also submitted an amendment that 5 provided some chemistry data. That was one of the key 6 questions on how we treated the chemistry effects, and 7 how it may have influenced environmentally assisted 8 fatigue.9 So as far as basis for the evaluation, we 10 are consistent in our approach, consistent with the 11 Gall report. We did evaluate the specified locations 12 as I mentioned in the NUREG 6260, and the Fen 13 methodology that we used was appropriate and was 0 14 driven by the two cited NUREGs there for the different 15 materials, carbon steel and stainless.
16 Additionally we did use our as-built 17 drawings to do our analyses.
We used the design.18 MEMBER ABDEL-KHALIK:
How different are the 19 as-built drawings from the design drawings?20 MR. FITZPATRICK:
There is additional 21 thickness for -this is Jim FitzPatrick
-' the shell 22 has additional thickness in it from the design for 23 rolling, like a quarter inch, and the nozzles have a 24 little additional thickness from the original design 0 25 provided on the fabrication drawings.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com
% F 99 1 MEMBER ABDEL-KHALIK:
And when were those 2 as-built dimensions acquired?3 MR. FITZPATRICK:
They are on the GE 4 drawings of the design before the plant started up.5 MEMBER ABDEL-KHALIK:
Thank you.6 MR. DREYFUSS:
We did use design transients 7 versus the actual transients, so did not take credit 8 for any -we used the conservatisms associated with 9 design transients.
10 We'll talk a little bit more about cycle 11 projections, 'but we did project cycles for 60 years.12 We'll talk about some conservatisms that we have 13 inherent in those projections as well.O 14 We also assumed-15 CHAIRMAN SHACK: So when you say design 16 versus accident transient severity, it means you are 17 using the stresses from the design transient, not the 18 numbers of the design transient?
19 (Simultaneous speakers)20 MR. DREYFUSS:
And again we did assume full 21 uprate conditions for the 60-year period. We did do 22 the uprate in 2006.23 MEMBER ARMIJO: From day zero uprated 24 conditions, and put those into all of these analyses?25 MR. DREYFUSS:
That's correct. Assume from NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 100 1 1972 up to this point and through the 60-year period.2 Now we are going to talk about the 3 specifics of the evaluation itself, and Jim do you 4 want to talk on this a bit.5 MR. FITZPATRICK:
We used existing design 6 analysis for the RPB shell, the lower head, the 7 recirculant nozzles, and by the FEM to those existing 8 analysis, and for the fatigue analysis MB 3200 rules, 9 for three nozzles that entire original design fatigue 10 usage, we analyzed for new models, new analysis, for 11 the feedwater recirc outlet nozzles and the coarse 12 spray nozzles.13 MEMBER ARMIJO: Was the feedwater inlet.14 temperature changed as a result of the uprate?15 MR. FITZPATRICK:
372 to 392.16 MEMBER ARMIJO: Now is that change in the 17 conservative direction as far as this analysis is 18 concerned?
19 MR. FITZPATRICK:
It increases the stress 20 range from your normal operating down to your 21 injection.
Delta T goes from, instead of 372 to 100, 22 it goes from 392 or 394 to 100. It's a small increase 23 in range.24 MEMBER ARMIJO: Okay, thank you.0 25 MR. FITZPATRICK:
And then for the piping NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 101 1 we performed new ASME class I fatigue analysis for the 2 recirc RHR.3 On the reanalysis of the three nozzles, we 4 used 60-year cycles projected based on design 5 transient severity and the cycle. So basically 6 reviewed our design spec, and updated BWR for thermal 7 cycle definitions.
8 We had new answers, find out what models 9 are developed fbr these three nozzles using the as-10 built drawings and the material specs for each one of 11 these nozzles.12 Heat transfer coefficients were based on 13 the design report and design specifications.
14 A thermal stress response in the 15 reanalysis was developed from a step change in the 16 temperature.
And Green's function was developed from 17 that.18 Using the Green's function we developed 19 thermal transients, stresses, for each set of the 20. design transients for each nozzle.21 And we calculated component stress 22 differences.
This is where the difference between -23 we'll explain a little further on, but this is where -24 25 CHAIRMAN SHACK: Let me just come back to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS.
1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 102 1 your Green's function.
So you got your Green's*2 function essentially from a finite element analysis -3 MR. FITZPATRICK:
Yes.4 CHAIRMAN SHACK: -with a step transient.
5 MR. FITZPATRICK:
Yes, sir. And you pull 6 component stresses from there versus -it calculates 7 stress intensity.
And that has led to some confusion 8 before.9 Taking those, the thermal stresses, the 10 pressure stress intensities were directly from the 11 answers found with the models, and they were factored 12 to account for the actual pressure during the 13 transients, the unit load case and then factored up 0 14 for that.15 Adjusting intensities to detached piping 16 loads were conservatively calculated and added to the 17 other stress intensities for each transient and each 18 temperature.
19 The maximum stress differences from the 20 temperature transients were combined directly with the 21 stress intensities from the pressure stresses, and the 22 detached piping loads.23 And the ASME MB 3200 fatigue calculations 24 performed on the collective thermal transient stress 0 25 systems.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com
- o I 103 1 And that gets rid of the ASME CUI. Then 2 we used a bounding fatigue life correction factor for 3 all the transients, one bounding number applied to 4 that CUF for the entire 60-year operating period.5 And then the environmental CUF is that 6 bounding factor times the CUF.7 We had a list of-8 CHAIRMAN SHACK: One other -every time I 9 read the analysis it says, axi-symmetric ANSYS model.10 This is a nozzle on a cylindrical shell. Why is it 11 axi-symmetric?
12 MR. STEVENS: It's a simplification to -13 obviously when you model a nozzle axi-s~ymmetric you 0 14 treat, the vessel then becomes a sphere. So we also 15 had to apply a correction factor to account for the 16 ovalization of two intersecting cylinders.
17 And that's just a traditional way of 18 industry way of modeling these nozzles.19 MR. FITZPATRICK:
Did that answer your 20 question?21 Some of the conservatisms in the analyses, 22 the major ones -23 MEMBER ARMIJO: Just before you go on, the 24 bounding fatigue life correction factor, you say you 0 25 ,calculated from water chemistry conditions expected to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com
- o 104 1 occur over the 60-year operating period. But you have 2 had major changes with the water chemistry with 3 hydrogen implemented many years after. So which is 4 the water chemistry you used? Did you use the 5 appropriate water chemistry for the normal water 6 chemistry period, and a different water chemistry 7 correction?
Or the hydrogen water chemistry period?8 MR. FITZPATRICK-:
Did both, and Gary can 9 give you a detail on that.10 MR. STEVENS: We actually broke the 11 operating history up into three parts. The prior to 12 hydrogen water chemistry, or normal water chemistry, 13 where the factors, at least for 'the carbon and low 14 alloy would be much higher and the oxygen content was 15 higher.16 Then we had the operation that was post 17 hydrogen water chemistry implementation, with the 18 historical duty if you will or availability of the 19 system.20 And then in the future and what that's 21 projected to be. And that was based on water 22 chemistry guidelines that the plants are following.
23 CHAIRMAN SHACK: And you used bounding 24 strain rates for all these transients?
Or you O 25 actually tried to estimate strain rates?NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 105 1 MR. STEVENS: We used bounding strain rates 2 for everything.
3 MR. DREYFUSS:
And we will talk a little 4 bit more about chemistry during the course of the 5 presentation.
6 MR. FITZPATRICK:
Some of the major 7 conservatisms in the nozzle reanalysis.
The number of 8 transient cycles using analysis was greater than the 9 expected number of cycles for 60 years based on our 10 plant experience.
For example, heat up and cool down, 11 there were 300 cycles -heat up cool down for the 12 feedwater nozzle includes heat up and then a turbine 13 roll. It's basically the Imajor transient.
We used 14 300 cycles of that. To date we've had 95 over 36 15 years of operation, and the original design was 200;16 we don't even expect to hit that number, based on the 17 past history of 20 years of operation.
18 But the plant had more transients in the 19 beginning than they do in later life.20 The transient severity is a conservatism
.21 versus using actual transients.
We used the bounding 22 values, the pressure and temperature of the EPU for 23 the entire life, and the bounding Fen multiplier.
We 24 used values, the input stat, the temperature strain 0 25 rate, the sulfur content were chosen to maximize that.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 106 1 And that multiplier was basically applied 2 to all transient stresses, and that was the reanalysis 3 method that we used.4 MR. DREYFUSS:
We talked about the 5 chemistry itself. Bottom line is we chose our 6 chemistry factors conservatively, and chemistry 7 effects have been conservatively factored into the 8 analysis that we did.9 We did use the Fen factors from the cited 10 NUREGs. Additionally we selected the various 11 parameters that you see here in such a way as to 12 maximize the effects and maximize the contribution 13 that they had in 'terms of their effect on the O 14 environmental factors.15 So strain rates, temperatures', dissolved 16 oxygen, were all factored in that way.17 CHAIRMAN SHACK: Of course there is no 18 conservatism in that sulfur number since your sulfur 19 probably is well over .015. In the materials you 20 actually have in the plant.21 MR. DREYFUSS:
Correct.22 MR. STEVENS: Plus that particular 23 parameter tends to have less effect on the relations ,24 than some of these, oxygen and temperature and strain O 25 rate for example.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 107 1 MEMBER ABDEL-KHALIK:
Well, typically, how 2 long would these oxygen excursions last?3 MR. FITZPATRICK:
A couple of days when 4 there's the heating up, and you do a cycle flush, and 5 then you start heating the reactor up, conduits come 6 online. It takes awhile to get to the steady state on 7 the chemistry.
8 MR. DREYFUSS:
The startup might be over an 9 18-hour period, but getting it back to a stable 10 condition will sometime take a day or two.11 MEMBER ABDEL-KHALIK:
So the different 12 between the value that you used and the analysis, 13 which is the mean plus one standard deviation, the 0 14 difference between that value and the nominal value 15 for dissolved oxygen, what would that be in.16 percentage?
17 MR. FITZPATRICK:
It's a little different.
18 That number could be significantly higher, but there's 19 no transient occurring at that time. So looking at 60 20 years we tried to do a bounding number, a 21 representative number for all the transients expected 22 to occur over 60 years.23 MR. CHANG: If I may interject something.
24 The staff did a focused review of what they did, 0 25 especially in the oxygen content and excursion.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20605-3701 www.nealrgross.com
- o 108 1 Now this is a BWR, not a PWR. The PWR, 2 the maximum transients for the most critical 3 components is during the heat up and cool down. The 4 PWI especially the feedwater nozzle -now excursion of 5 the oxygen content occurred during the heat up, but at 6 that time there are no significant transients.
- So 7 even excursion rate is high, applied to -if you apply 8 to zero it's still-zero.
I don't mean zero; I mean 9 small number.10 MEMBER ARMIJO: So these excursions, these 11 oxygen excursions, really had a very small 12 contribution to the number that you used for the Fen?13 , MR. DREYFUSS:
Right, it did not 14 significantly impact it.15 So the summary here is that the cumulative 16 usage factors at Vermont Yankee under all analyses 17 that we did perform do remain below one for the full 18 60 years of extended operation with margin.19 I'll talk a little bit about the audit 20 that NRC performed of the calculations.
And the key 21 challenges really were when we had done the analysis, 22 we did the feedwater coarse spray and reactive recirc 23 nozzle corners. The challenges were at the nozzle 24 corners, the blend radius as it's referred to as well.0 25 And the methodology by which we treated NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 109 1 the stresses was really the key factor as Jim had 2 talked about as well. So we used component stresses, 3 stress difference versus the maximum stress 4 intensities.
And what it comes down to is the 5 treatment of sheer stress and are you neglecting sheer 6 stress using this methodology.
7 That was the challenge.
So we did submit 8 this amendment 33, based on or in response to an RAI.9 And we documented the evaluation that we had performed 10 and the methodology by which we had treated the 11 stresses versus the component stress difference.
12 And we did essentially a sensitivity calc 13 .that resulted in a change, a maximum difference 14 between the reanalysis that we had performed and the 15 sensitivity that resulted in a very small maximum 16 change, a .003 change which I think would have been 17 complete at that point. But we really only addressed 18 one element of the challenge.
19 And Gary, if you would explain a little 20 bit about that.21 MR. STEVENS: Yes, I think what we really 22 addressed in that response was the effect of sheer 23 stress.24 Another part of the challenge was on this, 0 25 it's been coined in several different ways, uni-axial NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 110 1 stress, one-D virtual stress. And I think what I'd 2 prefer to do is, we have a slide coming up where we 3 show the analyses we did side by side, and I can get 4 into a little more detail on that one.5 But for the purposes of this slide, I 6 think we generally agree that we might have satisfied 7 the sheer stress issue, but we didn't satisfy the uni-8 axial or one-D virtual stress issue.. And we'll talk 9 about that in a few more slides.10 MR. DREYFUSS:
And Jim, if you could step 11 us through the approach that we did here on the 12 confirmatory calculation.
'13 MR. FITZPATRICK:
We did a confirmatory 0 14 calculation on one nozzle, a feedwater nozzle. It's 15 the controlling nozzle, because it is the most severe 16 in design transients; had the highest fatigue uses of 17 the three nozzles in question.18 And we tried to put this, in simple terms, 19 basically it's cold return water and is the hot 20 vessel. That's why it is the more severe -the most 21 limiting nozzle.22 A number of design transients at two to 23 three times the number of transients for the other 24 nozzles. All the injections occur at that nozzle, 0 25 versus the other ones feeling just the environment in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com ill 1 the vessel.2 And industry experience has shown that the 3 fatigue usage is typically higher at the fatigue -at 4 the feedwater nozzle than any other nozzles.5 We used the same ANSYS finite element 6 model, the same transients, the same cycles, and the 7 same water chemistry that is the previous nozzle 8 reanalysis.
9 And the confirmatory analysis, you combine 10 six stress components for NB-32, 16.2. The sheer ii stresses are included for each stress.12 And as the fatigue analysis was done for 13 NB-32 2.4 for all the stress pairs, and this is the 0 14 same methodology used in the reanalysis.
15 CHAIRMAN SHACK: What is the difference 16 between the confirmatory calculation and the 17 reanalysis?
18 MR. DREYFUSS:
We are going to show that on 19 a slide. I make that very clear.20 CHAIRMAN SHACK: Not the difference in the 21 results. What's the difference in assumptions?
22 MR. STEVENS: Should we go to that slide 23 now? So this slide has the two analyses in parallel, 24 the reanalysis, and th6 confirmatory calculation.
O 25 And what's in bold we'll talk about is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 112 1 going to answer your question on what the differences 2 are.3 And I don't mean to simplify this 4 calculation, and this analysis; it's done in six 5 steps. We've simplified into six boxes, which in no 6 way indicates that there are six simple steps to this.7 It's an ASME code analysis, and 'there is a lot of 8 rigorous detail built into this.9 So let's start at the left, and we'll kind 10 of go through these both in parallel.
Because you'll 11 see a lot of the boxes are identical.
12 On the left we have 60-year cycles in 13 design transients.
That was the same and identical 14 for both analyses.
We assumed the same transients and 15 the same quantity; we didn't differ on those.16 We built an ANSYS finite element model.17 It was the same for both analyses.
There was no 18 different in model at all.19 The model how we used it was, and the 20 stresses we obtained, is where it was different, and 21 that's the next one. So for in both analyses we'll 22 take the simple part first, pressure stresses and 23 piping stresses -pressure stresses were determined 24 from that finite element model, pressure stress.0 25 intensity, and piping stresses were done by hand.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com , v 113 1 That was identical for both.2 Now let's go to the first box, and here's 3 where we have the first difference.
In the first 4 analysis rather than run all the transients, and we 5 have approximately 20 transients in the feedwater 6 nozzle -there's many and they are complicated
-7 rather than run all of those individually through the 8 finite element model, we used a Green's function to 9 generate the stress history for those transients.
10 That's -the Green's function is a well 11 known technique in most all college mathematical 12 textbooks.
I don't think there is any controversy in 13 how the Green's function generates stresses.
But 14 we'll talk about this uni-axial or one-D stress in a 15 minute, and that's really where the contention lies 16 there.17 But in the first case, the reanalysis, we 18 used the Green's function to generate stress histories 19 for all those transients.
That takes a significantly 20 less effort than running all those transients through 21 the finite element model.22 CHAIRMAN SHACK: But this is purely an 23 elastic problem, right?24 MR. STEVENS: That's correct, so Green's 0 25 functions would be appropriate for that. Everything NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 114 I is linear.2 Now in the second case, the confirmatory 3 calculation, we ran everything, all the transients 4 individually through the ANSYS finite element model.5 So up to now the only difference is, we used a Green's 6 function in the first case to generate stress 7 histories; in the second case the ANSYS finite element 8 model.9 To your point the two should be identical, 10 because everything is linear.11 So how did we combine -moving on to the 12 fourth box -how did we combine and determine maximum 13 stress intensities?
Here is where we get into some 14 esoteric differences between the two.15 I'll take the easy one first, which is the 16 lower one, the confirmatory calculation.
We basically 17 take for all those- transients, we get six stress 18 components out of the finite element program, X, Y, Z 19 and three shears. And we combine those for NB 32 16.2 20 of the code, which for every peak and valley you take 21 differences, in those six stress components, and you 22 rotate those into principal stress differences, and 23 it's stress intensity.
And you use that history, 24 resultant history, to calculate fatigue usage.2 25 What did we do with the Green's function?NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 115 1 We'll move, up to the reanalysis..
The Green'-s 2 function, what we did there is, the Green's function 3 itself, the stress history we got out of the finite 4 element program, we could have had six Green's 5 functions to use to generate six stress histories, six 6 component stress histories for all the transients.
7 What we took out of the finite element 8 program was basically the maximum stress difference, 9 which is essentially equal to the stress intensity 10 from the finite element program.11 So what we got from the Green's function 12 was a stress intensity history, and we used that to 13 integrate and come up with a stress intensity history 14 for all of the transients.
So I think you can see 15 that the simplification here that was made, and there 16 are several, we are obviously by using the maximum 17 stress component difference we are ignoring sheer 18 stresses.19 And in some of the responses to the RAIs, 20 and John mentioned on the one slide we showed the 21 sheer stresses were negligible.
22 But the other issue that we didn't address 23 in those RAIs is taking a single stress intensity.24 history and using that through a Green's function to 0 25 generate a stress intensity history for all these NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W: (202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 116 1 transients.
2 Is that identical or proximate or close to 3 taking all the six stress component histories and 4 doing differences and rotations into a stress 5 intensity difference?
I think there is where the 6 difference and the contention really lay was that 7 approximation.
8 Both of these analyses, the intent is to 9 do an ASME code fatigue calculation.
There was never 10 any intent not to do so.11 The difference in that step I think is 12 really key to our differences.
And obviously doing a 13 confirmatory calculation was intended to resolve that 14 issue, proof that how close these were.15 So after that step then we have a stress 16 intensity, history that was computed differently in 17 each of the techniques.
But given that stress 18 intensity history, the fatigue usage analysis was 19 performed identically between the two.20 There is a type on the slide here. It's 21 not NB 32 24, it's 32 22.4.22 MR. RADEMACHER:
So that is 32 22.24?23 MR. STEVENS: Correct., So that step is 24 identical between the two. And then the last step is O 25 -we get a fatigue usage out of that fifth box that we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 117 1 then apply environmental factors to..2 In the reanalysis, the first one we did, 3 the maximum Fen was applied to the total usage, to 4 come up with the environmentally assisted fatigue 5 number.6 In the confirmatory calculation a maximum 7 Fen was computed for each load there, where the only 8 thing that was taken into account was the temperature.
9 We took the maximum temperature of each load, put the 10 strain rate and the sulfur and all the other primaries 11 were the same. And good or bad the intention of that 12 difference there was to demonstrate yet another 13 conservatism built into the analysis.14 So the only thing different in the last 15 step, which is the environmental fatigue evaluation, 16 was one Fen applied to total usage in the reanalysis; 17 multiple bounding Fens applied to each load pair in 18 the confirmatory calculation.
19 MEMBER ARMIJO: So the more conservative 20 treatment was in the reanalysis?
21 MR. STEVENS: For that step.22 MEMBER ABDEL-KHALIK:
How much do the 23 material properties change over the temperature range 24 let's say for the feedwater?
0 25 MR. STEVENS: I can't give you a specific NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 118 1 answer, but generally speaking there could be 10 to 15 2 percent variation I the material properties over the 3 range of temperatures we are looking at.4 MEMBER ABDEL-KHALIK:
And how is that 5 accounted for in the analysis?6- MR. STEVENS: In the reananalysis we picked 7 bounding temperature properties.
Because of the 8 Green's function use, everything
-you do one run and 9 everything is constant.
So we tend to take the 10 bounding material properties and heat transfer 11 coefficients.
12 In the confirmatory calculation the 13 material properties are varied with temperature input 0 14 to the finite element program as well as heat transfer 15 coefficients.
16 And you are really touching on one key 17 element here, if you take these -we have identified 18 really just three bold spots where these analyses are 19 different.
We identified on an engineering level 20 20 differences in these two analyses, things like you 21 just mentioned, material properties; they were treated 22 differently.
Heat transfer coefficients were treated 23 differently.
Twenty differences between the 24 reanalysis and the confirmatory calculation really 0 25 that were levels of conservatisms built in to the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 119 1 analysis, approximations using a simplified approach 2 versus a very detailed approach.3 So we did not go through exhaustively a 4 parametric study to understand which of those 20 items 5 caused the differences between the two. We were 6 satisfied at the end that the final result we got was 7 the same, usage factor less than one with margin.8 MR. DREYFUSS:
Do you want to move on to 9 the results?10 MR. CHANG: Before moving on, could I put ii in a couple of comments?12 I think Gary have summarized what you call 13 the reanalysis and what you call the confirmatory 0 14 analysis very nicely.15 But I'd like to bring out a couple of key 16 points that can facilitate going right through the 17 heart of the issue.18 Actually applicant submitted two 19 reanalyses.
One was submitted by amendment 31 which 20 is dated 9/17. The second refined analysis was 21 submitted December 1 1 th; that was submitted by 22 amendment 33.23 So those two I call them just reanalysis.
24 And then there is a final confirmatory
-you call 0 25 final confirmatory analysis submittedby amendment 34 NEAL R. GROSS COURT'REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 120 1 on January 30, '08.2 Now in our final SER, submitted to the 3 ACIS and it was issues, we call that. analysis as 4 analysis of record for the feedwater nozzles. Why?5 That's the point I'd like to point out. Missing this 6 phase, this is the opportunity, you may keep in mind, 7 reanalysis, analysis of record, which is not the case.8 The -now let's call that analysis of 9 record. The analysis of record took all the unknowns 10 out of the place. You use six components, stress, 11 including sheer stress and nominal stress. Only thing 12 is you approximate the header effects by a spherical 13 header. That as Gary said is a very standard 0 14 industrial approach.
We buy that.15 The difference comes that the reanalysis 16 did not analyze every transient.
From the base 17 transient case, and finite element results, from that 18 base case you project it to the other transient 19 stresses by the Green's function.20 I don't dispute the Green's function 21 methodology at'all; I love it. The only way is, how 22 do you apply it? Now you apply it by six components, 23 or you apply it by one-D virtual stress.24 The reanalysis -still have the one-D 0 25 virtual stress there. But the analysis of record do NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 121 1 not have that.2 So let's for the time being call the 3 analysis of record close to the reality. The outcome, 4 you don't see it at the amendment
- 34. Because 5 amendment 34 seems to indicate the analysis of record 6 always give you a lower answer. That means the 7 reference analysis is conservative.
8 But that is deceiving, because if you use 9 the same Fen as you used in the refined analysis, the 10 CUF will be higher. As I report it, as the staff 11 report it in the final SER, that number, the CUF, will 12 be .893. It's not .353 anymore.13 So in other words~the analysis of record 14 gives you higher CUF for everything the same 15 condition.
16 In other words the refined analysis can be 17 conservative, can be not conservative; can be 18 conservative by a factor of two; and also can 19 underpredict by a factor of two.20 For that reason we don't call that the 21 refined analysis or analysis of record. But for 22 Vermont Yankee the feedwater nozzle, the final 23 analysis, additional analysis, or whatever you call 24 it, still give you at least 10 percent margin to the 0 25 code CUF limits.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 122 1 For that reason I feel comfortable.
Now 2 as long as you make this as the analysis of record.3 For the future if you want to adjust anything you base 4 it on that. You don't back to the refined analysis.5 On the same basis if this can produce 6 results like this, the same or similar results can 7 also be produced.
I'm not sure, because I didn't do 8 that analysis on the other two nozzles.9 For that reason we asked them to perform 10 similar analysis for the other two nozzles. When all 11 this is completed, we have three analyses of record.12 Those are fully justified.
13 VICE CHAIRMAN BONACA: What I would like to 0 14 point out, however, is that this calculation results 15 seems to be consistent with the one that was in the 16 SER. So we would like to understand it better.17 In the SER you asked the licensee to use 18 the same maximum Fen.19 MR. KUO: Right, what we consider that is 20 acceptable is what the applicant calls confirmatory 21 analysis.22 VICE CHAIRMAN BONACA: Yes, but here in the 23 confirmatory analysis I see the result being 0.35, and 24 you are quoting .893.025 MR. CHANG: Eight nine three, we have both NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com
- .
123 1 numbers reported in the SER, so it's on record that 2 the analysis of record, using the maximum Fen, you 3 will get .893. But you use 24 different values of Fen 4 which is appropriate, you will get .353.5 In other words, the .353 is not wrong;6 it's just compare the earlier analysis and the newer 7 analysis.
The earlier analysis may not be 8 conservative.
It depends on the final analysis which 9 we know is right and conservative.
10 CHAIRMAN SHACK: What you are arguing is 11 that his stress analysis could be nonconservative, and 12 he covers that up by using a conservative Fen, but 13 clearly his overall calculation is conservative but 14 he's piling it up in different ways, and I guess the 15 question is, is that always going to be the case?16 It's certainly true in these two situations.
17 MR. CHANG: Normally staff do not second.18 guess what the future outcome will be. But since this 19 feedwater nozzle, the CUF, is five to 10 times higher 20 as compared to the others, I would imagine the other 21 two nozzles when you complete your analysis give us a 22 good foundation to work for the future. This number 23 will also be. good.24 MEMBER ARMIJO: I'm a little confused.
The 25 mechanical analysis I think, the confirmatory NEAL R. GROSSI COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 124 1 calculations were done by the methods the staff was 2 comfortable with and were done with a lot of 3 conservatism as pointed out in some of these charts.4 In addition they applied a more realistic 5 Fen for different periods as opposed to the original 6 reanalysis approach.
But still.conservative.
7 So I don't know, and there's a big 8 difference in CUF, right, .35 versus .89, that's a 9 very big difference.
So' what does the staff consider 10 to be the official number for CUF for this nozzle?11 MR. CHANG: .353.12 MEMBER ARMIJO: Okay.13 VICE CHAIRMAN BONACA: Because in the SER 0 14 you state very clearly that any request of the 15 licensee to use a maximum Fen, and you got the value 16 of .89, okay, still using the confirmatory calculation 17 now it ends out to .89, and you are saying because it 18 is higher than what you calculated with the reanalysis 19 which was .64, then the analysis of record has to be 20 the one with the higher value.21 So here we are talking about apples and 22 oranges. I mean I'm trying to understand what is the 23 confirmatory calculation result, and what is the basis 24 for forcing them to use the highest Fen? I mean O 25 that's probably the best question.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 125 1 MR. CHANG: As many people call the fatigue 2 analysis, it's a black box. You can turn out 3 different results depending on the level of 4 sophistication that goes in there.5 The first step we are trying to establish 6 is, is the Green's function methodology or the 7 confirmatory analysis methodology, which is correct.8 We say the confirmatory analysis 9 methodology is correct. That's the purpose of 10 bringing the .893 up.11 VICE CHAIRMAN BONACA: But you told me that 12 0.35 in the confirmatory analysis calculation is 13 cOnservative;,that's what you said.14 MR. CHANG: They are realistic.
15 Realistically speaking, the refined analysis do not 16 have to use Fen equal to 11 to all the transient 17 pairs. If you make every assumption the same, 18 confirmatory analysis will get you lower results.19 MR. KUO: Just like you said, Dr. Bonaca, 20 comparing this two analyses here is comparing apples 21 and oranges, because the numbers involved are 22 different in terms of Fen.23. For the reanalysis that they used, okay, 24 they used a bounding Fen value for all transient 0 25 pairs. But for the confirmatory analysis as they NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 126 1 called it they used Fen, maximum Fen for each 2 transient pair.3 VICE CHAIRMAN BONACA: that is what I 4 understood.
And you said you have to assume the same.5 Fen for both methodology if you want to compare 6 results.7 MR. KUO: If they were to use the same 8 bounding Fen for all transient pairs, using the 9 methodology in the confirmatory analysis, the number 10 would have been .893.11 VICE CHAIRMAN BONACA: Okay, that's why you 12 are talking about -13' MR. CHANG: Dr. Bonaca, Robert Schu, who 14 used to be on my staff and is fairly involved on this 15 topic, he may supplement some of the points.16 VICE CHAIRMAN BONACA: No, I understand 17 now. But go ahead.18 MR. SCHU: May I say something?
Because 19 basically when you are doing the fatigue analysis 20 you've got to calculate the stress. And right now the 21 stress implemented by the applicant is not correct.22 Compare -it's not adequate, because everybody believe 23 the ANSYS result is adequate.
So we asked the 24 applicant to compare their methodology with the ANSYS 0 25 analysis.
The result, there is no way they can match.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 127 1 So from that analysis record point of 2 view, their Green's function, any time they do a 3 Green's function analysis, they've got to redo the 4 traditional ANSYS analysis.5 And actually the traditional ANSYS 6 analysis will create the correct results and that's 7 NRC accept.8 MR. CHANG: The traditional ANSYS analysis 9 will create reasonable results. That result could be 10 higher; it could be lower. But that's reasonable.
11 That's correct. That's why we think our -that's will 12 be our future basis.13 We want something to be correct.14 MR. DREYFUSS:
Garyk if we could summary?15 MR. STEVENS: Okay, let's forget abou8t the 16 sixth box here, which is the environmental fatigue, 17 and let's look at the fifth box, which is, we've got 18 the stress history. We calculated fatigue. And let's 19 write some numbers down and put everything in 20 perspective.
21 CHAIRMAN SHACK: That is the CUF in error 22 if we just quite at the fifth box.23 MR. STEVENS: We will compare apples to 24 apples here, which is CUF from each analysis prior to o 25 an application of environmental factors.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com
- ° 128 1 Okay, the top box, the CUF for 60 years 2 from the reanalysis was .064.3 The bottom analysis, fifth box, the CUF 4 for 60 years was .089. The difference between .025.5 If we applied the same environmental 6 factor to both. fo those numbers, the difference in the 7 magnitude would be identical to comparing those two 8 numbers. So if I decided the environmental factor is 9 11, and I applied them to both, the ratio of the two 10 would be the same.11 So comparing apples to apples here, the 12 confirmatory calculation, .089 versus the reanalysis 13 of .064, as I mentioned before there were 20 some odd 14 differences built into these two calculations, any one 15 of which could havecontributed to that difference.
16 The use of a single stress intensity 17 history could be one. The material properties varying 18 with temperature could be one. The heat transfer 19 coefficients varying. Any of them. We did not do 20 exhaustive analysis to determine which one contributed 21 how much.22 So I think what the staff is saying is 23 that that increase is what has led them to the license 24 condition for the other two nozzles.0 25 MR. CHANG: You are correct.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 129 1 MR. DREYFUSS:
This is what took from 2 September or so up to this point, going through this 3 and trying to. address staff questions on it.4' It became clear to us that a simpler 5 approach is to go with the confirmatory approach.6 That is why we did that for the feedwater nozzle, and 7 we do have that license permission.
8 MR. CHANG: when all the three nozzles were 9 done, the three confirmatory analyses would become 10 three analyses of record; that's important.
11 MR. MANNAI: This is Dave Mannai, licensing 12 manager. I'd like to make one point, because I did 13 sense a little bit of concern on the part of the 14 staff, the ACRS committee.
The bottom line is, we 15 agreed with the NRC in their request to make the 16 confirmatory analysis the analysis of record.17 When we had performed the calculation.and 18 then subsequently the NRC staff had ordered that 19 calculation, they looked at our methodology, and they 20 did not disagree with the fact that for the 21 confirmatory analysis that the maximum Fen factors had 22 to be chosen for each transient, but that was a more 23 realistic use of that calculation that was wholly 24 appropriate as Dr. Chang said a month ago.0 25 And so if you stop in the middle of it NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS
.1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 130 1 you'd say oh there is this big difference.
But as the 2 analysts went through and our own folks reviewed that 3 and then subsequently the NRC staff reviewed it, there 4 were no concerns with the use of that calculation or 5 those assumptions that were used.6 MEMBER ARMIJO: Okay, so there is no.7 disagreement with the staff on the use of bounding 8 Fens for each transient pair as the right way to go;9 correct?10 MR. MANNAI: Right.11 MR. KUO: It is more realistic.
The reason 12 that we want to make this so-called confirmatory 13 analysis as the analysis of record is to prevent 0 14 future readers getting the wrong impression.
The 15 original reanalysis is still the right reanalysis that 16 we accept.17 MR. CHANG: If you only read this analysis 18 result once, you want to read the right one. You can 19 skip all the intermediate steps.20 MR. DREYFUSS:
Okay, next slide.21 These are the results, we've talked about 22 them. And the next slide.23 I'll speak a little bit about the license 24 condition.
As discussed, the confirmatory analysis 0 25 for the feedwater nozzle is complete. ,It is the calc NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 131 1 of record.2 The reanalyses performed for coarse spray 3 and reactor recirc outlet you can see the CUFs 4 adjusted for environmental factors here. The .17 and 5 .08, we fully anticipate that as we perform the 6 confirmatory calculations, that we will again be below 7 one with plenty of margin, and that in fact the 8 feedwater nozzle is the controlling nozzle for us.9 The license condition itself is, we will 10 perform the confirmatory analyses for coarse spray and 11 recirc outlet no later than two years prior to going 12 into the extended period of operation.
13 MEMBER ARMIJO: If you are already tooled 0 14 up for this analysis work, why don't you just do it?15 MR. DREYFUSS:
There is some additional 16 work to do, there's resources, there's modeling work 17 that needs to be done. We will be getting to work on 18 that. We just don't have those analyses complete yet.19 Our intention is that we will be working 20 on these during the course of this year, and getting 21 that work complete.22 VICE CHAIRMAN BONACA: Thank you for the 23 presentation.
It was clear, and we begin to 24 understand what's happening here. And now we go to 0 25 the staff presentation, right?NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 132 1 MR. ROWLEY: Good morning. My name .is 2 Jonathan-Rowley, and with me I have Dr. Kenneth Chang.3 And we will discuss the environmental fatigue issue as 4 it pertains to the Vermont Yankee safety evaluation 5 report.6 Next slide. I'd like to give you a quick 7 recap of this discussion from the February 7 th, HRS 8 meeting. We talked about the resolution of this 9 concern, and the included license renewal, the license 10 condition that we have applied to Vermont Yankee.11 Next slide.12 As you can recall Vermont Yankee revised 13 their application to use the fatigue model for their 0 14 management of fatigue for the extended period of 15 operation.
The corrective action element of that 16 program allows them to do a reanalysis of components.
17 They submitted those reanalyses to the NRC that 18 included incorporated environmental fatigue on 19 September 1 7 th, 2007.20 We performed an audit of those reanalyses 21 on October 9 th and 1 0 th. We asked six audit questions 22 during that audit. One was not answered to our 23 satisfaction, so we made that an RAI; we sent that on 24 November 2 7 th, 2007.0 25 The response to that RAI came back on NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 133 1 December 11, 2007.2 We had some discussions about this RAI.3 There were some differences in nomenclature and other 4 things that we couldn't quite work out, so we decided 5 to have a face-to-face meeting on January 8 th, which 6 was a public meeting on January 8 th, 2008, at that 7 time they agreed to submit a confirmatory analysis of 8 the feedwater nozzle. Next slide.9 That analysis was to include benchmarking 10 for the Vermont Yankee's feedwater nozzle using axi-11 symmetric on that element model, taking full care of 12 all stress components of the nozzle using ANSYS code 13 for all defined transients; demonstrated that Vermont 14 Yankee specific benchmarking calculations bound the 15 coarse spray and the recirculation outlet nozzles, 16 calculated fatigue usage factors were done by ASME 17 code Section 3, and they can compare the results to 18 the previous calculations to determine if they were 19 conservative or not. Next slide.20 On January 3 0 th Vermont Yankee submitted 21 those what we called -a terminology change -updated 22 analysis, which is one and the same with the 23 confirmatory analysis.
They proved to us that they 24 used the same parameters, same data, methodology, as O 25 agreed upon.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com
% I v 134 1 And the last slide, what was stated during 2 the January 8 th meeting; determined'that the CUFs were 3 the safe ends and then rated lower than the previous 4 analysis.5 Next slide please. Supplemental 6 information was submitted to us on February 5th to 7 demonstrate that the updated feedwater analysis bounds 8 the recirculation cutlet nozzle, and it described how 9 the water chemistry effects were accounted into this 10 analysis.11 Next slide. We performed an audit on 12 February 1 4 t, Valentine's Day, and we discussed the 13 things listed here. And I would like Ken Chang to,.14 talk about what we did at that audit, 15 MR. CHANG: I will not follow these slides.16 Instead I will go through the process of how we 17 performed the audit.18 The audit, the main purpose to address the 19 concerns expressed during the previous ACIS meeting.., 20 So really it's the chemistry, effect of chemistry on 21 this EF analysis.22 So we spent a good time of the day 23 reviewing the absorbed oxygen content, the strain 24 rate, the temperature, the surface content, those 25 parameters that they used in the.confirmatory analysis NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 135 1 or the analysis of record.2 Those parameters were properly used, like 3 the dissolved oxygen is average plus one standard 4 deviation.
And then we asked about whether any 5 excursion was there, the excursion happened during the 6 heat up. During the heat up process we found that the 7 feedwater nozzle don't have any significant 8 transients, although it doesn't bound the oxygen level 9 during the heat up, so that doesn't really matter.10 And we also looking at the strain rate, a 11 low strain rate to bound the value, to bound the Fen 12 value, was used all along.13 And the temperatures, we assumed using 550 0 14 degree Fahrenheit for the nozzle, which is also 15 bounding.16 For the surface content, for stainless 17 steel, surface content is not one of the parameters 18 evaluated by NUREG CR 5704. But for the carbon steel, 19 .015 percent was used to have the maximum impact on 20 the Fen.21 We also look at how they performed this 22 confirmatory analysis.
The. confirmatory analysis and 23 the reanalysis use the same model, the axis-symmetric 24 finite element model, for which the branch site is 0 25 exact. You find the axis of symmetry.
You do a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 136 1 revolution around it. But on the header pipe, on the 2 header side, you can only simulate with either the 3 flat plate or with a sphere.4 Normally people do simulate the header 5 effects by using a spherical header. The right way to 6 do is to use two times the actual' radius for the 7 sphere. That way you simulate to accurately predict 8 the pressure stress.9 For Vermont Yankee there was a model using 10 1.5 radius already done, so I don't dispute that, 11 since they adjust the pressure stress by another 12 factor of 1.33, four thirds. Now four thirds times 13 three halves, that's a factor of two,. That is a 14 typical number being used by the ASME stress analysis 15 simulating the 3-D effects.16 We also look into what Fen value we used.17 That has been already discussed in quite detail. I 18 really fully endorse them of using 24 training pairs 19 to come up the total CUF, and 24 Fens were calculated, 20 one for each training pair. That is the most 21 complete analysis I've seen so far. I hope we can 22 make this as analysis of the future, as a general 23 case.24 Now, the -another question was asked 25 during the early meeting was how was film coefficient NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 137 1 calculated?
The film coefficient was calculated 2 correctly even including the gap between the thermal 3 sleeve and the nozzle wall. They estimated how does 4 the gap open or close, and calculated some film 5 coefficient to simulate inside of thermal sleeve, 6 between the thermal sleeve and the nozzle wall, and 7 after the nozzle wall. So that analysis was quite 8 accurate, and even by today's standards it's still 9 very good.10 Other transients:
the two analyses use the 11 same transients; otherwise you cannot compare.12 Transients got to be the same. Cycle got to be the 13 same. Same training curves. Same number of cycles 14 was used in the refined analysis and in the 15 confirmatory analysis.16 External piping loads, here is a little 17 deviation from the traditional MD 3200 analysis as 18 compared to this. Although Vermont Yankee did not 19 apply the external piping loads in a 3-b way, but they 20 calculated a stress intensity based on the external 21 load.22 And that external load was added, that 23 stress intensity was added, to the stress intensity 24 calculated for the thermal transients.
After that 0 25 stress intensity was calculated add on top, that is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 138 1 known to be conservative.
2 K sub e, ASME code requires elastoplastic 3 cycling penalty factor. In old analysis normally 4 people have K sub e equal to one. We look into it, 5 and for the feedwater nozzle, K sub e the worst 6 combination K sub e equals to 1.115. So in other 7 words this 11.5 percent penalty on that underlying 8 stress before you go into the -- allow the cycle to 9 stress to the allowable cycle curve. That is also 10 appropriate.
11 Young's modulus, ASME curve, the fatigue 12 curve, is based on certain Young's modulus. When you 13 are performing analysis 'you have to adjust your 0 14 Young's modulus to the ASME code value. That was done 15 also properly.16 Six stress components, although it's not 17 a true 3-D analysis, but six components was used. For 18 the thermal transients, those components, in 19 particular the unit stress giving small or big is 20 included in their confirmatory analysis.
That is, to 21 us that's acceptable.
22 Seismic loading, seismic is one of the 23 transients.
Seismic, you cannot put on the 3-D 24 analysis and put in six components, because you don't 0 25 even know what it is. However, the seismic loads are NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 139 1 small compared to similar transient loads. And 2 seismic loads, when seismic load occurs, the strain 3 rate is high, Fen is low. So by not considering the 4 seismic load in the combination, produce conservative 5 results.6 Cycles: the two analyses use the same 7 cycles, the same transient cycles. That is 8 appropriate.
9 So based on these descriptions we felt 10 through deeper review and through the cooperation of 11 the applicant, by bringing two suitcases of material 12 into NEI, downtown office, we reviewed there; we are 13 very satisfied.
14 If you can make this as analysis of record 15 for the feedwater nozzle, we say, we.have no further 16 questions.
17 On the same basis there are two other 18 nozzles, could result in a similar way. So we say, if 19 you perform this kind of confirmatory analysis as 20 described above, then you heard it twice already. You 21 heard it from the applicant; you heard from me. If 22 you do that kind of analysis for the two additional 23 nozzles, our confidence level also goes up for those 24 two nozzles. So the whole issue will be resolved.25 Now I really want to thank the applicant NEAL R. GROSS K COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 140 1 for performing this analysis, because this, let me 2 remind you, yesterday we talk about whether on the 3 nozzle, they are one location or two locations or 4 three locations which you need to study.5 This nozzle, the plan radius is not at the 6 safe end. Yesterday you hear about safe end. You've 7 got to evaluate your pipe to the nozzle well, you've 8 got to evaluate the safe end. You've got to judge 9 whether you have similar sleeve or not. You've got to 10 evaluate the plan radius.11 It happens to be for this nozzle the plan 12 radius is the highest to CUF location.
Did you see 13 that yesterday?
I don't. That's why we insist on 14 performing similar. analyses for similar kind of 15 conditions and terrains.16 That concludes my presentation.
17 VICE CHAIRMAN BONACA: Could you go to page 18 nine?19 MR. CHANG: Page nine?20 VICE CHAIRMAN BONACA: Here you are talking 21 about previous analysis.
Is this the reanalysis?
22 MR. CHANG: Previous analysis means the 23 reanalysis.
The September 19 and December 11.24 VICE CHAIRMAN BONACA: Okay. We got an 0 25 explanation of what we meant by reanalysis and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 141 1 confirmatory analysis.
So the October analysis now is 2 the confirmatory analysis.3 MR. CHANG: One and the same.4 VICE CHAIRMAN BONACA: That's what I 5 thought.6 MR. CHANG: Updated analysis, the 7 confirmatory analysis, and the analysis of record, 8 those three are equal right now.9 MR. SHUN: I am sorry, Ken, why do you say 10 these three are equal? I thought they are different.
11 Reanalysis is reanalysis; normally reanalysis is -12 they are not equal.13 MR. CHANG:. What Jonathan call is update 14 analysis, and what applicant call as confirmatory 15 analysis, we call them analysis of record.16 MR. KUO: I would personally suggest, let's 17 not confuse the issue. We, at least from staff's 18 point of view, we stopped using the term, confirmatory 19 analysis.
We have the analysis of record.20 MR. CHANG: I agree.21 VICE CHAIRMAN BONACA: Are we disagreeing 22 with the previous statement, that previous analysis 23 means reanalysis?
24 MR. ROWLEY: No.0 25 MR. CHANG: For the feedwater nozzle, there NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 142 1 is only one analysis of record; that is submitted on 2 January 30, '08.3 VICE CHAIRMAN BONACA: Still it says, the 4 confirmatory analysis which now has become the 5 analysis of record.6 MS. FRANOVICH:
If I may, this is Ronnie 7 Franovich, the reason that this has been such a strong 8 view by the staff is that we are establishing a new 9 licensing basis for license renewal, and so being very 10 clear on what the licensing basis is for this issue is 11 really important for the future regulation of the 12 facility.13 I wanted to answer one question by the 14 gentleman, why wouldn't they do the analysis now for 15 the other two locations.
The end of the current -the 16 period of extended operations really begins in 2012, 17 and so two years before that would be 2010. So it 18 won'.t be but for another couple of years th8at we will 19 get that analysis in for the other two locations.
20 Just wanted to clarify that too.21 MR. ROWLEY: All right, next slide please.22 Our conclusion is that the feedwater 23 analysis is the analysis of record, as performed
'in 24 accordance with ASME code Section 3, the coarse spray O 25 and the reactor circulation nozzle analysis will be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 143 1 performed according to the fourth condition which is, 2 next slide, that the licensee perform and submit to 3 the NRC for review and approval an ASME code analysis 4 for the reactor circulation and outlet nozzle and the 5 coarse spray nozzle at least two years prior to the 6 extended period of operation.
This analysis shall be 7 the analysis of record for these two analyses.8 VICE CHAIRMAN BONACA: Now on the 9 conclusion on the second bullet, did you say that the 10 CUF was calculated in accordance with ASME code 11 Section 3. But the analysis was also in conformance 12 with the ASME code Section 3?13 MR. ROWLEY: The entire analysis -the 14 entire updated -well, confirmatory analysis, yes.15 VICE CHAIRMAN BONACA: Yeah, the claim was 16 made that the original analysis was also conforming to 17 ASME code Section 3.18 MR. CHANG: to be precise, that should be 19 performed according to the ASME code without using 20 Green's function.
methodology.
21 VICE CHAIRMAN BONACA: Yes, okay. They 22 stated the same thing. So that is not the 23 distinguishing attribute 24 CHAIRMAN SHACK: Well, just to defend the 25 poor Green's function -here for a second, poor Mr.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.neatrgross.com 144 1 Green, the Green's function is fine. It's how they 2 combine the stresses after the use the Green's 3 function that is the problem.4 MEMBER BLEY: Calling that the Green's 5 function method is not right.6 (Simultaneous speakers)7 CHAIRMAN SHACK: I did have a question, if 8 I could ask Gary Stevens, this came up.9 Does the location of the maximum fatigue 10 usage change when you do the individual transients, 11 decay Fen? You find that the actual location of 12 maximum usage has shifted? You didn't look at that?13 MR. STEVENS: We did. I'm trying to figure 0 14 out the best way to answer your question without 15 confusing the whole room.16 The answer would be no, but what location 17 we looked at we built into this going into the 18 analysis.
And there were several considerations.
19 First and foremost would be looking at 20 what the original design analysis tells us about where 21 the high usage location is. And that's an appropriate 22 technique
-23 CHAIRMAN SHACK: Well, no, when we say high 24 usage location, I mean are we talking nozzle or are we 0 25 talking finite element location, et cetera.NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 145 1 MR. STEVENS: I'm not sure I understand 2 that question.3 CHAIRMAN SHACK: You get a different usage 4 factor for every finite element in this whole axis-5 symmetric model, and I'm assuming the number you are 6 quoting here is the highest usage factor for any given 7 element that you are looking at.8 MR. STEVENS: That's right. We based our 9 selection process on really three things: maximum 10 stress, which is going to give us high usage factor;11 we also need to look at different materials.
Some of 12 these nozzles have stainless steel safe ends and low 13 alloy steel nozzle forgings which have different Fen 14 factors associated with them. And we also have to 15 look at chemistry, as in water chemistry.
16 An example there would be the feedwater 17 nozzle where the incoming feedwater stream, the oxygen 18 content is significantly different than it is in the 19 vessel. So the environmental factor for the safe end 20 would be drastically different than it is for the 21 nozzle forging.22 All that was built together, and that's 23 why for each of these nozzles we take two locations, 24 the limiting location in the safe end, and the 0 25 limiting location in the nozzle forging. And that is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 146 1 a composite of all those factors going together, that 2 collectively this gives us -between the two locations 3 we've covered the maximum possible usage factor for 4 the whole component.
5 If I -I would come up with a different 6 conclusion if the chemistry was constant for all 7 locations, the material was constant, I might pick one 8 location in a safe end, in a PWR for example, 9 especially where stratification loading is present, 10 and it drives you back to the safe end.11 In this situation here, with different 12 materials and different chemistry, we chose to 13 evaluate two locations to bound it.14 MR. CHANG: Dr. Chang. If I may supplement 15 what Gary says. You vary two locations, but when they 16 say safe end, actually they evaluate three locations 17 in the safe end, the pipe end, the pipe to nozzle 18 weld; and the transition.
Consider, next to that 19 transition there is a thermal sleeve which can change 20 temperature diffusion pattern.21 So one location covers three areas which 22 they did not advertise.
I just tried to clarify.23 MR. ROWLEY: So that ends our presentation 24 unless there are more questions.
0 25 VICE CHAIRMAN BONACA: Thank you for your NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 147, 1 presentation.
And are there any questions?
Or 2 further comments?3 I guess~not, so I'll give it back to you.4 CHAIRMAN SHACK: Gentlemen, I think we can 5 break for lunch until 1:15. And again I'd like to 6 thank the licensee and the staff for very interesting 7 presentations.
It did help clarify an issue that was 8 quite confusing.
9 (Whereupon at 12:04 p.m. the proceeding in 10 the above-entitled matter went off the record to 11 return on the record at 1:15 p.m.)12 CHAIRMAN SHACK: We can come back into 13 session.O 14 Our next topic today are some selected 15 chapters of the SER associated with the ESBWR design 16 certification applications.
And Dr. Corradini will 17 lead us through that.18 MEMBER CORRADINI:
Thank you, Mr.19 Chairman.
I'll just give a short reminder to the 20 Members about where we are in this. So the purpose of 21 this portion of the meeting is to review four chapters 22 of the design certification document and the 23 associated SERs that we have talked about in 24 subcommittees.
Those chapters of the SERs are 0 25 chapters 9, 10, 13, and 16, with open items related to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com