ML20137W898
| ML20137W898 | |
| Person / Time | |
|---|---|
| Issue date: | 04/16/1997 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-2098, NUDOCS 9704210065 | |
| Download: ML20137W898 (289) | |
Text
-
L Official Transenpt of Proceedmgs gmg;;
han nr s NUCLEAR REGULATORY COMMISSION BBillEO ER Ih20R 4R W . -
, munu ElB1 m.& p+ ;
i
Title:
Advisory Committet on Reactor Safeguards Materials and Metallurgy Subcommittee ?@ 9 7
W$n$yt
? sg . ? .
, ? .': .:$ A%;
TRO4 (ACRS) P ^3 v.
RETURN ORIGINAL k I
{ Docket Number: (not applicable) hs""E Sb'
_ 415-7130 l THANT.S!
' hgajM I Location: Rockville, Maryland i M3;
$29h N
y&lQ 3 Date: Wednesday, April 16,1997 1
?;$f*e b I fh y
A RS?W ovyaerx m
r S *W
_ fo : e ;. *e ca' e vo~ Tir.:ee -
lasi uned 1-
- Work Order No.: NRC-1080 Pages90-291 $ljl 9704210065 970416 -
h-20d PDR _ . _ . [
p._.,:.. _.:-,-~ -
NEAL R. GROSS AND CO., INC. O K9m(
Court Reporters and Transcribers ) t- i &
1323 Rhode Island Avenue, N.W. dd If : / ~.I!-
Washington, D.C. 20005 _
(202) 234-4433
(
90 1 UNITED STATES OF AMERICA
-w 2 NUCLEAR REGULATORY COMMISSION
' ~ '
) f 3 +++++
4 MEETING 5 ADV1SORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS) 6 +++ ++
7 WEDNESDAY 8 APRIL 16, 1997 9 +++++
10 ROCKVILLE, MARYLAND l l
11 12 The Subcommittee met at the Nuclear Regulatory 13 Commission, Two White Flint North, Room T2B3, 11545 s
14 Rockville Pike, at 8:30 a.m., William J. Shack, Chairman, 15 presiding.
16 17 COMMITTEE MEMBERS:
18 WILLIAM J. SHACK CHAIRMAN l
19 THOMAS S. KRESS MEMBER l
20 ROBERT L. SEALE MEMBER 21 22 23 24
,rx
) 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND ME., N.W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
91 l l
1 ACRS STAFF PRESENT:
1 l
,e 2 NOEL DUDLEY
)
'RJ \
3 CAROL HARRIS )
l 4 ,
1 5 ALSO PRESENT:
6 JACK STROSNIDER 7 TED SULLIVAN 8 PHILLIP RUSH 9 GENE CARPENTER 10 LAMBROS LOIS 11 CAROLYN FAIRBANKS 12 JACK HANSON 13 ANDREA LEE
( \
14 DEBORAH JACKSON 15 JOE MUSCARA 16 JIM MEDOFF 17 KURT COZENS 18 BARRY ELLIOTT 19 MIKE VASSILAROS 20 STAN ANDERSON 21 ROSS SNUGGERUD 22 MIKE MAYFIELD l
! 23 ED HACKETT 24
/m i \
N._/ 25 i NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS )
1323 RHODE ISLAND AVE., N W. l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
92 1 A-G-E-N-D-A f- 2 Acenda Item Pace l 3 Opening Remarks by ACRS Chairman 93 4 GL: SG Tube Inspection, Phillip Rush 95 5 GL: Reactor Vessel Head Penetration l 6 Gene Carpenter 120 7 Status of Palisades Reactor Pressure 1
8 Vessel (RPV) Activities, Lambrois Lois 136 9 Status of Draft Rege.latory Guide on 10 Neutron Fluence, Carolyn Fairbanks 155 l 11 Licensee Plans for Palisades RPV 12 Jack Hanson 161 4
13 NUREG-1511 Supplement 1, Andrea Lee 229 i
/-
\
\s/ 14 Reactor Vessel Integrity Database l l
15 Andrea Lee 247 l
16 DOE Reactor Vessel Annealing Project l 17 Update, Deborah Jackson 264 I 18 Discussion 289 19 20 21 l 22 23 24
.] 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 2344433
93 1 P-R-O-C-E-E-D-I-N-G-S l
g- 2 (8:39 a.m.)
3 CHAIRMAN SHACK: The meeting will now come to l 4 order.
, 5 This is the second day of the meeting of the 6 ACRS Subcommittee on Materials and Metallurgy.
7 I am Bill Shack, Chairman of the Subcommittee.
8 I have a conflict of interest with some topics of today's 9 meeting. During the discussion of those~ topics, Bob Seale 1
10 will serve as Chairman; basically, anything that says 11 " Steam Generators".
12 The ACRS members in attendance are Tom Kress l
13 and Bob Seale.
' (f~h I 1 i x/ 14 The purpose of this meeting is to continue l l 1 15 discussions with representatives of the NRC staff and l 16 Consumers Power Company concerning Generic Letters 17 associated with steam generator tube inspection 18 techniques, effective use of ultrasonic testing techniques 19 in inservice inspection programs, degradation of steam 20 generator internals, and degradation of reactor vessel 21 head penetrations.
1 i
22 We will also discuss the status of issues l
l 23 related to reactor pressure vessel integrity. The 24 Subcommittee will gather information, analyze relevant (G,) 25 issues and facts, and formulate proposed positions and l 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 94 l
l 1 actions as appropriate for deliberation by the full i
r~s 2 Committee. .
~
j 3 Noel Dudley is the Cognizant ACRS Staff 4 Engineer for this meeting.
5 The rules for participation in today's meeting l l
l 6 have been announced as part of the notice of this meeting l i
7 previously published in the Federal Register on April 4, l l 8 1997.
9 A transcript of the meeting is being kept and I i
10 will be made available as stated the Federal Register 11 Notice. It is requested that the speakers first identify l
12 themselves and speak with sufficient clarity and volume so l 13 that they can be readily heard. l i(D
\
' -)
14 We have received no written comments or l
15 requests for time to make oral statements from members of 16 the public.
17 We will proceed wath the meeting and I call 18 upon Phillip Rush of NRR to make the first presentation.
l l
19 MR. RUSH: Thank you. Good morning. My 20 name's Phil Rush and I'm in the Materials and Chemical i
21 Engineering Branch in NRR and I'm here to talk this t
22 morning about e proposed Generic Letter on steam generator i
l 23 tube inspection techniques.
24 I'll first start out with some general r~T
( ,) 25 background on my talk. Eddy current inspection techniques l NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 i
95 1 are the primary means of assessing steam generator tube l
,f-s 2 condition and that's true throughout the world and in the
(_) 3 U.S. as well.
i 4 Three elements of assessment include:
5 identifying the degradation -- that's detection; 6 characterizing it; and a third step you could take could 7 take would be sizing -- and that's primarily what I'll be ,
i 8 talking about today.
9 The objective is to identify the type of tubes 1
l 10 and then plug or repair them based on the inspection 11 results. What we're primary concerned with here is J 12 degradation caused by, either mechanical wear or it could l 13 be stress corrosion cracking.
,7
\ l
14 Stress corrosion cracking is really a primary 15 focus of what I'm going to be speaking on here today 16 because that is the most difficult mode of degradation to ;
17 size, if you will.
18 MEMBER KRESS: Does the eddy current identify 19 the stress corrosion cracking or do you have to do 20 something else to it?
21 MR. RUSH: Well, that goes in with the 22 characterization of the mode of degradation. There are 23 some means -- eddy current can do that to a certain 24 extent, although you know, bells and whistles don't go off n
( ,) 25 to tell you you've got a crack or whatever.
I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
1 96 1 You know, you could have a network of flaws 1
fS 2 that could make it appear like it's a volumetric )
Nj 3 indication, or it could actually be a volumetric l l
4 indication intergranular attack or something.
5 And as I mentioned, sizing is really what I'm 6 going to be focusing on here today because it is the most 7 difficult aspect, in my opinion, of eddy current 8 inspections. !
9 Just to give you a general overview of what I )
10 mean with eddy current sizing, what you see up here is a I l
11 lisisu figure. This might be a typical pattern that you 12 get during an eddy current inspection.
p- 13 MEMBER SHACK: That's the pattern you hope you i )
\"# 14 get.
15 MR. RUSH: Yes, this is -- I'm speaking more l
16 on a theoretical basis, yes. ,
1 17 ACTING CHAIRMAN SEALE: This is as good as it 18 gets.
19 MR. RUSH: This is as good as it gets. And if 20 you read your textbooks on eddy current, you know, they 21 say you might be able to relate a flaw depth to a phase 22 angle. And by phase angle we mean the difference between 23 this horizontal line and the sharp transition you get 24 there as the bobbin coil, in this case, would move across
(
i/
m 25 the center of the flaw.
NEA' R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
97 1 However, in reality, eddy current is sensitive
, g-) 2 to many other things: crack morphology, support V 3 structures such as all the other metal within the steam --
4 you know, around the steam generator tubes. All these 5 things are going to affect the response you get on your 6 lisisu pattern and very much complicate the ability to 7 relate this angle here with some flaw depth.
8 I guess the problem or issue that I'm going to 9 be talking about goes back to something we identified a 10 while back, where the utility was trying to put an 11 estimate of flaw depth on some steam generator eddy 12 current indications. And based on our knowledge, there 13 wasn't any method to do that at that time. We just
\'d 14 believe that the eddy current sizing uncertainty was much 15 tom large in order to do that.
16 The tech specs assumed an uncertainty on the 17 order of 10 percent. You know, that's just general rule-18 of-thumb I guess, that's used there, but to actually 19 demonstrate that, we haven't really seen that at this 20 time.
21 If you do have large errors in your eddy 22 current uncertainty, effectively to some difficulties with I
i 23 diupositioning degradation of the steam generator tubes --
24 I guess a recent example we have, that we've recently f%
(_, 25 countered was ANO Unit One was a -- they tried to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l
l 1323 RHODE ISLAND AVE., N W (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1
1
98 l
l 1 disposition indications based on eddy current depth sizing l
- 2 and later found out, after they were up and running, that 7s 3 those indicate -- the tubes that they used during the l
l 4 outage at which they did this -- the degradations weren't l 5 matching what they had in the eddy current sizing, so it l
6 led to some difficulty there.
7 MEMBER SHACK: Well, this confused me a little 8 bit when I read it. I mean, people disposition all the 9 time for less than 40 percent or greater than 40 percent.
10 MR. RUSH: They should -- see, that's the --
11 you really shouldn't be dispositioning less than or 12 greater than, even to that extent. If you go and there 13 was a bobbin coil and you're making an assumption based on fx
( )
N/ 14 the eddy current data, if that indication is less than 40 15 percent or greater than 40 percent, that in essence, is a 16 sizing technique.
17 Because really schat you have to do is, if you 18 detect anything that appears to be some mode of 19 degradation -- unless it's mechanical wear, obviously, 20 because they do a pretty good handle on that as far as 21 sizing -- you really have to go in with alternative eddy 22 current inspection techniques, or maybe even UT or l
l 23 something else.
24 MEMBER SHACK: Oh, I see. It's basing it
()
( ,/ 25 strictly on the bobbin coil call that you've got the
( NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
l l
99 l l r
l 1 problem.
1 I
i 2 MR. RUSH: Yes. That's primarily what I'll 7-~3; I
' \~J l 3 get into in a little bit more, l l
i 4 ACTING CHAIRMAN SEALE: You're going to tell l
5 us about this other elaboration of the technique to l l
l 6 resolve this issue, or at least try to --
7 MR. RUSH: I'll give you a little more l
8 background as to how we - you know, where we went with 1
9 this and what we've done about it.
10 ACTING CHAIRMAN SEALE: Okay.
i 11 MR. STROSNIDER: Phil, excuse me. This is 12 Jack Strosnider. I'd just like to add one comment on this l
l 13 discussion that you just had. I guess I characterize this
,,~ '
i[ )
- \- 14 issue of one of, the tech specs allow fou to size and l
15 repair anything greater than 40 percent.
16 But the assumption tnere is that you have a I I
l 17 qualified method, and what ve're really talking about here 18 is, have the licensees qualified a method for sizing? You 19 know, have they gone out and actually tested enough l 20 samples or done something that they demonstrate that they 21 can really do that sizing with some confidence?
22 MR. RUSH: Yes, I think that my last item here l
23 is really the focus of what our concern is. Is that the 24 bobbin coil isn't qualified, at least in our impression, rh
( ,) 25 for sizing stress corrosion cracking and intergranular l NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
100 1 attack. There's just too many complicated variables 73 2 there. That's not to say it can't be done but I don't
]
3 think there's been enough effort put into it.
4 These are just some of the requirements for 5 steam generator tube integrity and NDE inspections. The 6 first bullet here has to do with the structural leakage 7 integrity requirements for the general design criteria.
8 The second outlines some of the criterion for NDE --
9 specifically that they should be qualified, as Jack 10 mentioned just a moment ago.
11 And then a third tier we have to deal with is
> l 12 the tech specs which require flooding and repair of 13 degradation greater than 40 percent through wall.
/,_N
\- 14 I'll just present some further background on 15 this issue. As I stated earlier, we initially identified 16 this, you know, one utility using the practice of sizing 17 degradation in their steam generators. This was around 18 the 1993 timeframe so it was quite a while back.
19 And that led us to the question of whether or 20 not this was a unique plant in that, were they the only I l
l 21 ones out there doing it or was this an industry-wide ;
l 22 practice? l l
23 With that, we contacted NEI and EPRI, working ;
1 24 in conjunction with NEI, they performed a survey of the !
(,) 25 utilities out there which they presented to us. And we +
1 1
I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE , N W. l' (202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
101 1 identified some utilities identified in there that r~% 2 appeared to be doing some sort of sizing. It wasn't t 3 LJ 3 entirely clear because the information was held 4 confidential; they didn't identify specific utilities.
5 So we initiated our own effort to try to i
6 identify some potential -- some utilities that might be j l
7 employing that practice, and we came up with some examples l 1
8 -- which I'll go into in the next slide here.
9 I guess the first I'll just talk on was at 10 Sequoyah. They were using a practice where they allowed 11 tubes to remain in service if degradation was less than 30 l
12 percent through wall at bobbin coil. And as I mentioned 7s 13 earlier, any type of sizing -- whether you set the limit
( ) i
'# 14 at 40 percent, 30 percent -- that's still a sizing 15 technique.
16 You know, if you think you have degradation 17 there, there isn't the confidence in our opinion, that you 18 can adequately size it, even plus or minus 30 percent.
19 Since we first identified this, Sequoyah has changed their 20 practice and they no longer use this sizing technique.
21 Another example that we're aware of and we're 22 pursuing, is Three Mile Island where they used a 23 combination of voltage signal to noise ratio and 24 degradation of length to these position indications in
(_) 25 their steam generator. Their tech specs are based on a 40 NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
l (202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
102 1 percent through wall depth limit, so it's really difficult l
l ,s 2 to understand how their practice is consistent with what's
! \
, ( /
'~'
l 3 in the technical specifications.
i 4 There's other examples that I could go into 5 here, but these are just two that we came up with in our 6 investigation. Given that we're able to identify several 7 plants that were employing practices that we thought were i
8 somewhat questionable, we felt that a Generic Letter was i
9 appropriate to address the situation.
10 The focus of the proposed Generic Letter --
11 well, I've outlined that here -- just to notify the 12 importance of using qualified inspection techniques.
13 That's really nothing new -- we've done that in previous
,/T
\- 14 Generic Letters -- but I guess practices are still 15 continuing to use sizing techniques.
16 We requested that they provide information on 17 their sizing techniques and the basis for that, and you 18 know, the focus is really to verify their compliance with 19 Appendices A and C to 10 CFR Part 50 and the technical 20 specifications.
21 The Generic Letter was published for comment 22 on the last day o+ last year. As a result the time of l
23 period -- it was out for 30 days and has expired since l
24 that time. We only received.one comment during the 30-day
[^T 25 period, and that was from NEI.
( ,)
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
103 1 Their comment was a goed one, in my opinion, f~, 2 It concerned -- does overlap with what we're focusing on
( )
3 and the issues that were coming about in the rulemaking 4 process. We took a look at it but, in the end, we don't 5 think that it's valid because of two reasons.
6 One is primarily the timeliness of the issues.
7 We don't know how long it's going to take to get anything 8 together on the rulemaking. And then really, as you don't 9 really know what's going to come out of the rulemaking 10 effort at this time, whereas this Generic Letter is 11 focused. We know what its intent is.
12 We did receive one other comment that involved 13 some, I guess word clarifications. You know, there were O
kl 14 some discrepancies with the technical specifications with 15 regard to some of the things we had requested in the 16 Generic Letter, so we've modified it subject to fix those 17 areas. And my projections are that we should be able to 18 get this out sometime in May of this year. I i
19 And I guess that concludes my slides. Are 20 there any questions at this time?
21 MEMBER KRESS: When you get all this l
22 information back, which basically will tell you how the l l
23 industry is doing their sizing of the flaws, what will you l
24 do with that?
m
() 25 MR. RUSH: Well, I guess what I would do --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
104 1 you know, one, we want to find out who's sizing, and then 73 2 once we look at who is sizing, we'd like to see, you know, (s_)' 3 what are they doing and what's their basis for doing that?
4 And then I guess it would be a matter of reviewing it just ,
4 5 getting a handle on their qualifications, seeing if it 6 seems reasonable in that aspect.
7 MEMBER KRESS: And if it doesn't?
l 8 MR. RUSH: If it doesn't then we'd probably l 9 pursue it further. I'm not quite sure the vehicle by 10 which we use, though.
l l
11 ACTING CHAIRMAN SEALE: Jack, I'm interested l 12 in this because it strikes me that this is an example of a 13 potential problem, let's say, in the approach that you
! l
\~s' 14 folks are taking in connection with the other activity on 15 steam generator -- what was rulemaking at one time, anyway 16 -- where you're talking about the industry putting 17 together a program which meets a set of performance specs, 18 or requirements, and that you would not " pre-approve" the 19 individual program as long as you had assurance that you 20 met the program requirements that are in your NUREG, or 21 whatever. Is that a fair statement?
22 MR. STROSNIDER: Right. Yes, this is Jack 23 Strosnider. Make a few comments here. I think first to 24 put this in perspective, what we ought to recognize and
(
(3,) 25 what Phil was presenting here, is that first we have NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
r !
105 1 Appendix B which says that people should have qualified l ,c~\ 2 methods.
l l l % j; 3 ACTING CHAIRMAN SEALE: Sure.
4 MR. STROSNIDER: But, the industry has never l
1 5 defined anyplace, what a qualified method is, so it's 6 basically up to the licensee. I 7 ACTING CHAIRMAN SEALE: Yes.
8 MR. STROSNIDER: What most people have been 9 using are the EPRI guidelines that have been developed, 10 and they provide a performance demonstration actually, l l
11 with some statistical basis for what sort of operating 12 characteristics or reliability you should have in the 13 testing. So that's what most people are using.
7 ~3
)
14 And I would point out -- I think Phi' 15 characterized this as, some of things may not be qualified 16 in our opinion -- in fact, some of the methods we've seen 17 have not passed the EPRI qualification yet people are 18 applying them.
19 So we've never officially endorsed the EPRI 20 guidelines; that is, the NRC has not said yes, this is the 21 way to do it. On the other hand, we have tacitly, 22 basically if people are following that we said well, okay, l
- 23 that certainly provides some level of quality in these 24 inspections.
O
's ,) 25 So when you put all that together what you NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N W.
I (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
106 1 conclude, I think the bottom is -- as I said earlier --
r3 2 is that yes, we have Appendix B, use qualified methods, U 3 but we haven't told people what that is. The Regulatori 4 Guide that we've developed defines what we think people 5 need to do to qualify their methods. The trick here is 6 putting that in performance-based space. And I think 7 we've done a pretty good job of that.
8 They have basically built upon some of -- this 9 type of approach you heard on Appendix VIII yesterday with 10 regard to taking a certain number of blind tests and 11 looking at the statistics of it. We built somewhat on the
_ ' s a few areas there where we 12 EPRI guidelines.
- 13 thought improvementu ..ere needed. For example, in using
! )
\' realistic defects versus artificial EDM notches, for 14 15 example. So there's some areas there were we're still 16 having discussions with the industry.
17 But if you look what's in the Draft Regulatory 18 Guide, what it really says is, you need to do enough 19 testing, all right, of mock-ups or actual tube pulls from 20 the plant -- you know, however you're going to get these 21 data -- to characterize the statistics. And in this case 22 we are, in fact, indicating that people should l 23 characterize what their probability of detection is for 24 detection qualification, and some statistic related to f
(_},/ 25 sizing.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433 1
I 107 1 Then the way it's integrated into the overall 7- 2 assessment of the steam generator tubes -- we haven't
, > 1 1 \J 3 indicated in their, in the Reg. Guide, that you have to 4 meet a specific level probability of reliability in 5 sizing. But we said you have to characterize it and then 6 whatever it is, you'd have to factor it into your 7 integrity analysis.
8 So I think we've made that -- we've made an 9 attempt to make that as performance-based as we can. Now, 10 in implementation of this, our concept is that if a 11 utility commits basically, to follow that method for 12 qualifying their sizing method or for qualifying their 13 sizing or their eddy current methods, then we don't need
,. A<
t 5/ 14 that submitted for review and approval, j I
15 We understand the process they're going to go 16 through. We could follow up through our inspection i
17 programs to verify that in fact, they are using -- that 18 they have qualified the methods in accordance with, as I J 19 like to refer to it, the box that we've drawn in the 20 Regulatory Guide.
21 So that's the long-term solution that we're 22 driving toward for this issue. As Phil pointed out, l 23 there's timeliness in here though, with regard to getting 24 that out, getting it incorporated in licensee's programs, (G,) 25 and we don't want too many outage seasons going by where NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l 108 1 people might be using methods that really don't have the l
l >
I
- gs 2 sort of qualification that we think they need to have. l
< ( \ l L-) \
Now in some cases, we may determine -- when we 3
4 find out what people are doing -- that there is a !
l 5 technically acceptable basis. I mean, licensees have in i
1 6 the past said, oh yes, we pulled tubes, we did various )
1 7 things. But we won't know that till we see it. When we 8 get the response, we will review those and at a minimum, I
9 if we see something that looks questionable, we'll go back 10 and have more discussions with these utilities.
11 So this is really sort of a short-term effort 12 to address this issue. And yes, you're correct; it's tied 13 in exactly to what we're talking about in the --
(~-
\ -) 14 ACTING CHAIRMAN SEALE: Yes. Well really, the 15 success of that effort, it seems to me is hostage to being 16 able to work together between the agency and the industry 4'
17 on this kind of interaction. Because vou can't do 18 that, then there's really no basis to believe that a --
19 what I'll call performance-based rule now, in the sense 20 that you're talking about -- can ever deliver. I mean, 21 you have to be able to develop some confidence between 22 yourselves in order to be able to let this thing work.
l 23 MR. STROSNIDER: I agree. I think that 24 basically, when we get into performance-based regulation,
(
O) 25 we do need to lay down some groundrules for how it's going NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS
' 1323 RHODC ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
I 109 l
l 1 to work.
-s 2 ACTING CHAIRMAN SEALE: Yes. i L.)
3 MR. STROSNIDER: We need to have performance l
l 4 criteria that are tolerable, we need to have performance 1 l
I l 5 criteria that are measurable -- and that's what this I l
6 really gets to. If we define factors of safety that 7 people have to demonstrate on the steam generator tubes -- ,
l 8 in fact, that might be the easier part of it; saying 9 here's the performance criteria, here's the goal. How do l
10 you measure it? l l 11 This is the principal tool that they will use 12 for measuring. So we'd have to have confidence that this 13 is being done well. We have had numerous meetings and
, O kJ 14 discussions with the industry. If you look at the Draft 15 Regulatory Guide we've indicated where we have some l
16 exceptions.
17 For example, with the EPRI Guidelines, we've 18 called that out. EPRI is working on a revision to the 19 Guidelines, We've had some presentations where we know 20 they're addressing some of our concerns. If we've left 21 the option open that if in fact, the Guidelines are put in 22 a -- the revised Guidelines address the concerns that I
l 23 we've expressed, that it may be possible to reference the 24 industry guidelines as opposed to our Regulatory Guide.
( ) 25 I don't know how it's going to work out; I l NEAL R. GROSS ,
COURT REPORTERS AND TRANSCRIBERS l l 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234 4433
110 1 know the two are coming closer together. I don't know if 3 2 we'll ever get them exactly the same. But we have to have
( )
N.s 3 some understanding of how people are going to qualify --
4 ACTING CHAIRMAN SEALE: There's a lot at stake 5 for everybody here.
6 MEMBER KRESS: Jack, with respect to 7 developing the statistics say, with pulled tubes, to 8 qualify the technique, was it acceptable for the utilities 9 to pool their data if they're all using the same 10 techniques? You know, they would qualify the technique, 11 not the process they used.
12 MR. STROSNIDER: Yes, we have not ruled out 13 the possibility of integrated programs where yes, if k- 14 people could demonstrate that the morphology is similar, 15 it's the same sort of degradation -- in fact, if you look 16 at the voltage-based Generic Letter, that database is made 17 up of pulled tubes from a large number of plants, as well 18 as laboratory data. l 19 The other thing I have to point out here, l l
20 though, is that this approach is not a panacea, if you !
l l
21 will, because developing these qualifications can be very i 22 expensive. Pulled tube data is expensive.
i 23 We're seeing some situations now, for example, 24 with circumferential cracking at the top of the tube sheet l
(<m 25 where they're finding it difficult to grow in the l ( ,/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N W.
WASHINGTON, D C. 20005-3701 (202) 234-4433
( (202) 234-4433 l
l
111 l
l 1 laboratory, the same type of defects that they see in the l l rs 2 field within a reasonable timeframe.
- t )
I%J l 1
3 So developing the data and the statistics to )
1 4 support this can be difficult. That's the industry's 5 responsibility. We can provide a framework, but they 6 still have a lot of work to do to make it work. l l
7 The one other comment I'd make is, in case you 8 didn't recognize this, where people don'C have qualified 9 methods for sizing degradation -- which means that they i 10 can't really demonstrate they're satisfying the 40 percent l
11 -- the norm is to plug on detection, because there's a lot 12 of tubes being plugged because of that situation.
13 It's a difficult situation. You know, it may
[,, )
' \ 14 be very conservative but on the other hand, it's not clear l
15 what option you have when you can't really measure this.
16 ACTING CHAIRMAN SEALE: That's an awfully 17 slippery slope to get on.
18 MR. RUSH: One other comment on your concern 19 there, about pulling the data between different utilities.
20 That was attempted in the one example I alluded to with 21 ANO Unit One, where they used data obtained from another 22 plant.
I i
23 However, they were applying the sizing 24 technique to degradation in the upper tube sheet crevice,
- O
( ,) 25 and the data was obtained from free span degradation and
[
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
112 l
l 1 they were using a combination of voltage and things like
( ,. s 2 that. Which apparently, you can't take free span
! (
)
3 degradation and apply it to the crevice when you're using 4 --
l 5 MEMBER KRESS: That was the sort of thing I 6 was worried about, and how we can deal with that.
7 MR. RUSH: Yes, so there are -- and just a 8 glance at their sizing qualification appeared reasonable 1
9 because they had a lot of data there, but when you l l
10 starting looking further into the details of it, it broke 11 down, and that was apparently playe9 out in the pulled 12 tube verification that they did most recently. I 13 MR. STROSNIDER: Just to follow up on that a l /^i
! 2 14 little further -- and I think it's very interesting. The 15 thing that happened, EPRI came into us -- I guess six 16 months or so ago, or maybe nine months ago; I forget 17 exactly when -- but they basically had concluded that they 13 needed to dequalify some of the sizing techniques that 19 they had. And basically when they looked at it, it's 20 because of the way they had grouped data -- just what 21 we're talking about here.
22 And I think we really have to give them credit 23 for being that rigorous in their assessment in recognizing 24 and acknowledging that well, you know, this really wasn't p
( ,) 25 doing what we said it was going to do. It caused some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
! (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 l
113 1 interesting regulatory issues that we had to deal with, I
,s 2 but it still, I think it was the right thing to do. ,
l / \ '
l t / '
l '
3 MR. RUSH: Yes, t'nat was really -- just a I
4 correction, Jack -- that was on detection of degradation, )
5 that wasn't actually sizing.
6 MR. STROSNIDER: Oh, I'm sorry.
7 MR. RUSH: They dequalified it for detection, 8 you know, one coil, the rotating coil, and they also 9 dequalfied the bobbin coil for inspections inside crevice 1
1 10 regions.
11 MEMBER KRESS: Oh, the rotating coil was 12 actually dequalified? l l
13 MR. RUSH: One of them was. It was the .080
,m
'w- 14 inch pancake coil.
i 15 MEMBER KRESS: Okay.
16 MR. RUSH: And that was only for detecting 17 degradation at expansion transition regions.
19 MR. RUSH: Yes, the outeide diameter. Yes, l 20 it's very specific there.
21 MR. STROSNIDER: Yes, this is an interesting 22 comment, because when we talk about degradation-specific i
23 management, this is an example of what we're talking 24 about. You use the right size coil for the degradation
/' h
() 25 that's on the outside of the tube, in the right part of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS i 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 i
114 j
It's a complete process. It can work, but you 1 the tube.
,s 2 really have to pay attention to detail.
. \
/
\
3 ACTING CHAIRMAN SEALE: I noticed that one of the cases that you mentioned was the TMI. And normally we 4
5 think of Westinghouse and CE plants when we talk about 6 steam generators as being one family and B&W as being 7 another. I'd be interested -- is there anything you'd 8 like to maybe help us out to characterize the differences 9 between them?
10 MR. RUSH: Well, there are quite a bit of 11 differences, and --
12 ACTING CHAIRMAN SEALE: But I mean in terms of 13 the inspection techniques and so on.
'l 14 MR. RUSH: Oh, in terms of the general 15 inspection techniques, not really. But historically, B&W 16 once through steam generators have been pretty much 17 defect-free, if you will. They haven't encountered a lot 4
18 of degradation. The defective tubes, or the problem at 19 TMI was a result of a one-time event on the primary side.
20 It was just basically a mistake --
21 ACTING CHAIRMAN SEALE: Well, it was just heat 22 capacity, that's all.
23 MR. RUSH: But the age-related degradation 24 tho you get in Westinghouse and CE units hasn't really Q 25 affected the once through steam generators up until
(,)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433 s'
115 i 1 recently, and now we're really seeing a lot more signs of e- 2 --
t i
\.'"/
3 ACTING CHAIRMAN SEALE: Now they're growing 4 up, too.
5 MR. RUSH: Yes, they're growing up. They have 6 different materials in there. They have sensitized Alloy 7 600 material, whereas the other ones, you know, you have 8 low annealed and that's really what the concern is.
9 MR. SULLIVAN: We have an information notice 10 that is in draft form -- that should be coming out within 11 a month or so -- that discusses in fair amount of detail, 12 experiences with degradation that they're just now 13 starting to get into as Phil mentioned. And like I said,
- a)
's 14 that should be coming out within a month or two.
15 ACTING CHAIRMAN SEALE: Send Noel a copy of 16 that?
17 MR. SULLIVAN: Sure.
18 MR. STROSNIDER: This is Jack Strosnider. I 19 think one thing you ought to be hearing here, though, is 20 that B&W is not immune and they've reached that time where 21 there's some degradation that's beginning to occur, so 22 don't be surprised if you start hearing more issues with 23 B&W --
24 ACTING CHAIRMAN SEALE: Well, that's one of
,e V 25 the reasons I wanted to highlight it here for the --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS j
1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l
116 l
1 MEMBER KRESS: Does NRC have a continuing
<s 2 research program in this area of detection?
l [ \
~
3 MR. STROSNIDER: Yes, that's part of the 4 research program, and I think -- Dr. Muscara's here, if 5 you --
6 MEMBER KRESS: Okay, we'll hold off on that 7 one, then. Unless he wants to --
8 MR. MUSCARA: Joe Muscara with the NRC 9 Research staff. Yes, we do have relatively a new research 10 program with steam generators. One portion of it is it do 11 with inspection. We're looking into different aspects of 12 inservice inspection.
13 One is to quantify the reliability of current t ! )
N/ 14 techniques so we do have the information with POD and l 15 sizing accuracy. The other part is to evaluate new signal 16 analysis techniques that would help us impreve on the 17 sizing and detection.
18 And we're doing some of that work on our own, l
i 19 but we're doing some in cooperation with U.S. industry and 20 over the international community. And so we'll be getting 21 advanced techniques in from other sources that we will l
22 evaluate.
23 While I'm up here, I'd really like to mata 24 another comment; it has to do with qualification. I'd (O
(_,/ 25 like to stress an important difference I think Jack was
- NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS f 323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1
117 1 drawing between the qualification processes in place now
,- -) 2 with the NDE Center, and what we might be discussing in
( l 3 the Regulatory Guide or --
4 MEMBER KRESS: Yes.
5 MR. MUSCARA: And the rest, we would like to 6 have POD data and sizing accuracy data, is we want to 7 quantify that and have that known for the particular 8 inspector. I think that's important. We tend to get too 9 comfortable with the idea that, well this inspection 10 technique has been qualified; it's qualified, it's on the 11 book let's use it.
12 Well, what is it qualified for? When one 13 looks in the detail of how these things are qualified,
- i
' ' ' 14 right now a detection technique is qualified if detection 15 performance is 80 percent POD for flaws, or a 60 percent 16 through wall and greatcr. That means 30 percent of the 17 flaws, even through wall, passed this technique, while of 18 those that are 60 percent and less are possibly missed 19 completely. So it's qualified to do what?
20 What's important is that we need to know what 21 the POD is, and based on the criteria, based on the 22 performance criteria, doesn't meet that criteria. That's 23 what important. They take the 60 percent POD for some 24 flaws, it might be SCCs that have a high growth rate q
( j 25 that's really not a qualifier that it is test that's qual NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
118 1 before, it gives you a high probability of tube rupture.
. 2 And the same position with the sizing.
(3)
I mean, the sizing criteria right now is
~
3 4 qualified if your RMSE value is 25 percent or better.
5 Well, what we have done in evaluating RMSE on data we 6 obtained from the Surry generator, we find that a 20 7 percent RMSE is acceptable if you are trying to detect 8 flaws on an eddy current scale of 40 percent, and then you 9 would hope that those would be plugged before the flow 10 reaches 85 percent.
11 So a 25 percent RMSE is quite acceptable if 12 your plugging criteria is 40 percent of the eddy current 13 scale. But what you need to keep in mind is that these g
w) 14 days we're letting flaws progress beyond 40 percent. And 15 it is accepted because these flaws are quite short in most 16 cases, and they have high pressure holding capability.
17 But again, we need to be careful. What does 18 40 percent mean? What does 25 percent mean? It means 19 that you can detect a flaw with the eddy currency of 40 20 percent before it reaches 80 percent -- 85 percent through f 21 wall. So I think it's very important that we're asking 22 for the techniques to be quantified and identical measures 23 should be used in the integrity analysis and not a piori {
l 24 to assume that it's qualified just because it's gone l
/^x 0
( ,) 25 through the Center and it passes a POD with only 60 NEAL R. GROSS 4 COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1 1
)
119 1 percent of --
,--~s 2 MEMBER KRESS: Yes, I think that's a real t \
(j 3 important point.
4 ACTING CHAIRMAN SEALE: Yes.
5 MR. STROSNIDER: One other comment on it 5 because -- it's an area we've been, obviously thinking 7 about quite a bit. There's an interesting aspect of this 8 approach which is, there is an incentive for more data 9 that a utility or the industry can pull together, and the i
10 tighter they can make their confidence on this sizing, the 11 more benefit they get out of it -- the more tubes they 12 probably can service.
13 So we do see an incentive in there. From the p) i x' 14 industry's point of view I think it would come down to 15 some economic decision of, you know, do we think if we 16 pull enough tubes or generate enough data are we going to 17 be able to save enough tubes for how long in the steam 18 generators to make it work out? You know, that's their i
1 19 decision, but that's the framework and that's the 20 incentive we're trying to provide.
21 MEMBER KRESS: The confidence bound is an 22 exponential function of a number of data points. You make 23 it to a point where you don't get much improvement.
24 MR. STROSNIDER: Yes, right. I mean, at some Ci
( ,/ 25 point there's not a whole lot of value in going after more NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
120 1 data, but if that's what the statistics show, then that's
,s 2 fine.
\
' ~ '
3 ACTING CHAIRMAN SEALE: Very good.
4 MR. RUSH: Thank you.
5 ACTING CHAIRMAN SEALE: Anything else? Thank 6 you very much. I must say, I enjoy these topics where we 7 don't feel like we're drinking from a fire hose. It's a ,
1 8 lot more fun to get into some of these issues a little bit 9 more.
I 10 Next is Mr. Carpenter. 1 l
11 MR. CARPENTER: Good morning. My name is Gene I'm here to talk to you today about General ;
12 Carpenter.
l 13 Letter 97-01 which was issued April 1st of this year.
( ) l
\' 14 Basically, it talks about the degradation of control rod l
15 drive mechanisms of the vessel head closure penetrations.
16 I was lead project manager on this issue and I work in the 1 17 Materials Branch under Jack Strosnider.
18 Today what I'd like to talk to you about, is 19 give you a little bit of background as to why we issued l t
20 the General Letter, some of the U.S. inspection experience l
21 with the PWR control rod drives. The present status as of 22 last month, actually -- now that we've got the Generic l
23 Letter out -- it's a little bit more that alighted there:
24 some of the public comments that we've received before we I'h issued the GL, what the GL's basically talking about, and
(,,/ 25
! NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
l (202) 234-4433 WASHINGTON, O C. 20005-3701 (202) 234-4433 l
i 121 1 then some of the staff conclusions that we have -- why we
,cy 2 issued it.
3 A little bit on the background. Primary water 4 stress corrosion of Alloy 600 components was first 5 identified as an emerging technical issue in 1989, and 6 that had to do with some of the cracking that was 7 discovered at Calvert Cliffs.
8 We went ahead and issued an Information Notice 9 90-10 later the next year, to inform the industry of what 10 we knew. And at the time we had told the Commission that 11 we didn't believe that this was an immediate safety issue:
12 mainly because the cracks were mainly axial in 13 orientation, they were exhibiting low growth rate, there
/ \
h
14 seemed to be a high flow tolerance of the material, and it 15 was unlikely to propagate very far.
16 But we did recognize that it was an issue and 17 it was one that bore closer watch. We also discussed 18 Alloy 600 cracking with Calvert Cliffs. As part of what 19 we decided to do at the time, we put together an action 20 plan and we met with the various PWR Owners Group 21 including Westinghouse, CE and B&W, and we also talked 22 with NUMARC -- now NEI.
l 23 Basically what we went through there was, we 24 discussed programs for investigating primary water stress
()
( ,) 25 corrosion cracking of Alloy 600, and they talked about NEAL R. GROSS
, COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
1 122 1 development of vessel head penetration mock-ups by EPRI, 1
- 2 some qualification of inspections, revision of generic 1 I o LJ 3 acceptance criteria, and performing actual inspections at ;
1 1
4 several plants.
l 5 We later, in that 1991 timeframe, cracking was 6 discovered at the Boujet 3 plant in France. Subsequently, 7 France, Sweden, Switzerland, Japan, and Belgium inspected, 8 basically all of their vessel heads, and they found 9 approximately 46 penetrations which equates to about 2 10 percent of all head penetrations that were inspected.
11 NRC staff reviewed the PWR Owners Group safety 12 analysis, or assessments that they sent in later in '92, 13 and concluded the vessel head penetration cracking again,
/,\
! )
\/ 14 was not an immediate safety concern, provided the Generic 15 Letter 88-05, inspections of the top head were performed, 16 and non-destructive examinations were also performed to i
17 ensure no unexpected cracking.
18 And we also recommended implementation of 19 enhanced leakage detection system such as they had in 20 European plants, and also in Japanese plants. The NRC 21 staff further asked NUMARC -- now NEI -- to coordinate all 22 the future vessel head penetration actions and activities.
23 And we continued of course, to meet with the PWR Owners k
24 Groups and with NEI.
,/ y
(_) 25 Some of the inspection experience we've had NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
123 1 here at the U.S. plants is that the methods were qualified
,s 2 by the EPRI NDE Center using eddy current, sizing by UT.
l( ~'
)
3 And the first inspections were performed spring, 1996, at i 4 Point Beach. They didn't find any indications in the 49 l
5 CRDMs that they inspected.
l 6 Oconee in the fall of '94 inspected and they 7 found 20 indications and one penetration, but these l
8 indications were fairly shallow.
9 ACTING CHAIRMAN SEALE: Earlier, in connection 10 with eddy current testing, we were reminded that the 11 qualification of the operator was an important element in 1
12 the overall validity of the test. Are you convinced -- or 13 satisfied I guess, is a better way to say it -- that the
(1 4 l
\ l
\/ 14 results here aren't just the fact that the guys at Oconee i 1
15 have got sharper eyes than the guys at the other plants? i l
16 That is, that they found 20 cracks and the 17 other people indicated none, but if the people who did the 18 Oconee inspection went to the other plant, might then find 19 some?
20 MR. STROSNIDER: This is Jack Strosnider, and 21 let me see what kind of response I can give you on that.
22 Unfortunately, Bob Herman's not here; he's the one who 1
23 actually went down and looked at the qualification of this 24 work. But my understanding is that looking at the
) 25 qualification of the NDE's methods at EPRI, that we i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l (202) 234 4433 WASHINGTON, D C, 20005-3701 (202) 234-4433 l
l
124 1 believe this was actually a fairly reliable inspection 7- 2 method. I don't think it had some of the ambiguities that l ( N~/ )
3 we see with the steam generator testing.
4 ACTING CHAIRMAN SEALE: So this is maybe a 5 bonified difference between a Westinghouse ~and a B&W?
6 MR. STROSNIDER: Well, or a difference between 7 this particular configuration --
8 ACTING CHAIRMAN SEALE: Yes, not only --
9 MR. STROSNIDER: Yes, but there's --
10 ACTING CHAIRMAN SEALE: -- of Oconee, but B&W?
11 MR. STROSNIDER: Yes, true. The other thing 12 that I would point out is that we have had people go out 13 and observe these inspections when they're in progress,
/ \
14 and some of the same people who witnessed some of this 15 down at EPRI. So they have some confidence that they were !
l 16 in fact, doing it the way it was qualified.
17 I'm not sure whether there really are a whole 18 lot of different teams doing this at this point, and I i 1
19 just don't know. It could be that there's not a whole lot I 20 of different contractors involved.
21 Jim, you need to use the microphone. Jim 22 Medoff has been following some of this. Maybe he has l
l 23 something to -- l 24 MR. MEDOFF: I'm Jim Medoff; I'm a Chemical j
/3
( ,) 25 and Materials Engineer for the staff -- Jack's staff. And NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 NHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
125 1 Oconee actually reinspected in 1996. I can't vouch for f-~s 2 the other plants but I know that when they did do their
- \
LJ 3 inspection for Oconee, Oconee prepared for their 4, inspections well in advance of performing them.
5 They came up and talked to us, had a 6 preliminary talk with us, then they took -- we went down 7 to the EPRI NDE Center and they showed us how they 8 qualified their techniques. And then they brought the 1
9 EPRI contractors up during the inspections and I went down 10 in the plant and observed the inspections there. So they 11 really did a very comprehensive job in qualifying the eddy 12 current techniques for doing the 1996 inspections, and 13 they performed eddy current and also penetrant testing.
p_
k
'- 14 Because they had a penetration where they 15 weren't sure -- they had indications they had a lot of 16 noise and actually they changed the penetrant testing 17 co npound and found out that a second penetration had 18 in:lications of cracking. But as you see on the viewgraph, 19 they were very shallow. UT couldn't pick up the --
20 cc uldn' t size the crack, or detect. But just to say that 21 they did do a lot of preparation.
l 22 ACTING CHAIRMAN SEALE: And they found cracks.
I l 23 MR. MEDOFF: Right.
24 ACTING CHAIRMAN SEALE: Shallow, but they s ) 25 found them.
i NEAL R. GROSS
(
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
126 1 MR. STROSNIDER: Yes, I guess that's one thing 7g 2 I wanted to point out here. When we talk about
)
u_/
3 indications -- yes, very shallow and these didn't require i
4 any repair, obviously, at this plant. And it's not clear 5 to us that this is actually the same mechanism that's been 6 seen in some of the other units, because these 20 7 indications -- this was in one penetration and it's not i
8 clear that this is the same sort of phenomena that's been 9 seen in some other units. They will be taking additional 10 looks at this, too is my understanding. ,
11 MR. MEDOFF: They're coming back in two cycles i
12 since the inspection -- I
,, 13 MR. STROSNIDER: Yes, so we'll get some more
( ) i 14 data. (
l 15 MEMBER SHACK: This is probably a Bob Herman l 16 question but -- I mean, these penetrations are in while 17 the vessels get their stress relief, so presumably this 18 stuff comes up with some sort of decorated grain 19 boundaries. Have people taken samples? I mean, do we 20 actually know what the metallurgical structure of these 21 things look like?
22 MR. STROSNIDER: There's been -- and you're 23 right, Bob Herman can give you a more detailed response.
24 I know that the Europeans have done some investigations of f
(_,/ 25 the material. I'm not aware of any material having been -
i
! NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASH'NGTON, D.C. 20005-3701 (202) 234 4433 l
l 127 l
1 _
73 2 MEMBER SHACK: That we've done here?
MR. STROSNIDER:
3 -- removed and examined in 1
I 4 the U.S.
5 MR. CARPENTER: Continuing on with U.S.
6 inspection experience. Palisades did a limited 7 examination of eight in-core instrument penetrations and 8 they found no cracking. D.C. Cook in the fall of '94, I l 9 found three clustered indications in one penetration, and 10 one indication had to be repaired -- or, the utility opted 11 to prepare it using a special method.
12 North Anna also examined some high-stress ;
13 areas on the outer ring -- again, a limited examination --
, ( 's
\' 2 14 and they also found no indications.
15 What we were doing at the time, last month 16 basically, was that we had inspected -- the NRC staff was 17 informed that Westinghouse, in some of the various 18 meetings that we had had with the PWR Owners Groups, had 19 developed a susceptibility model for Westinghouse vessel 20 head penetrations, and they were using that to determine 21 whether or not they should be inspecting.
22 The NRC staff was also informed that B&W had l
! 23 done the same -- Framatome Technologies for B&W plants --
24 and also that CE. So basically, all three of the PWR r8
(_) 25 Owners Groups had developed --
NEAL R. GROSS COURT REPORTERS AND TRANSCRlBERS 1323 RHODE ISt.AND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
128 1 MEMBER KRESS: Is this a sort of qualitative s 2 model that says if you had certain chemistry or you had 1( \
3 certain --
l 4 MR. CARPENTER: Well, they've used a variety l
5 of method and they've talked to us about it in meetings l
l 6 that we have had -- public meetings. But to-date, they 7 have not shared with us --
8 MEMBER KRESS: I see. You don't know what the 9 model is either?
10 ACTING CHAIRMAN SEALE: You don't know what a 11 susceptibility is?
12 MR. CARPENTER: Not as much as we'd like to.
13 Which leads us into the Generic Letter. As part of what
(
)
' J
'/'- 14 we had been requee'-ing, again, our first bullet here, none 15 of the PWR Owners Groups have submitted the models and 16 their assessments to the staff for review. And this is --
17 it's not because we haven't asked; we have, in multiple l
18 meetings. NEI --
19 MEMBER KRESS: They use that for prioritizing l 20 which plants they ought to look at first? Is that what --
21 MR. CARPENTER: That is what they've told us, l 22 yes. NEI submitted a white paper last year to the staff 23 that reviews the significance of SCC and PWR vessel head 24 penetrations and described how the industry is managing Q 25 this issue.
! NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l l
129 1 And one of the conclusions of that was that it i
I 73 2 describes an economic decision tool for PWR licensees to f )
w/
l 3 determine inspection need based on the evaluation --
4 probability of vessel head penetration cracking during the 5 plant lifetime.
6 Now, we went ahead and we put out a draft 7 Generic Letter last August for the public to comment on, 8 and we did receive ten sets of comments from industry, 9 NEI, and others. Basically, one comment requested an 10 extension of the comment period. We obviously granted 11 that.
12 One comment questioned that the proposed 13 Generic Letter was a backfit. It's not; it's a 50.54f
( \
\'s# 14 letter simply requesting information. And seven of the 15 comments endorsed NEI's comments, which were fairly 16 extensive and actually added some value to the letters, 17 that we certainly incorporated quite nicely.
18 However, several of them were editorial in 19 nature and those as I said, did strengthen the proposed 20 Generic Letter. Several comments pointed out the need for 21 the Generic Letter, and NEI disagreed with the staff's 22 request for information related to Corita. Zorita of l
l 23 course, was the Spanish reactor that had some cracking in 24 the upper head following some chemistry problems.
p
(,) 25 Following their initial set --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 l
! 130 1 MEMBER KRESS: What's the joke --
I l ,~ 2 MEMBER SHACK: It was a resin --
3 MR. STROSNIDER: Jack Strosnider. Probably --
l l
4 may be the understatement. They has some resin throughout 5 their system, and the significance was that the cracking 6 that developed in Zorita was circumferential in nature and 7 it was fairly significant.
8 And our whole assessment of the cracking 9 that's been observed in other units and the analyses that 10 have been done said that you should develop short, axial 11 cracks. Circumferential cracks have a much different l
12 significance with regard to potential failure and ejection 13 of one of these penetrations. So that's why we were l l ,
t /
14 interested in knowing if anybody had had any similar 15 problems of this kind, with regard to chemistry.
16 MEMBER KRESS: NEI thought it just wasn't 17 relevant to our plants? j 18 MR. STROSNIDER: I'm sorry, I didn't hear the 1
19 question.
20 MEMBER KRESS. Did NEI think it just wasn't 21 relevant to our plants? Is that why they --
l 22 MR. STROSNIDER: We have a representative l
23 here.
24 MR. COZENS: This is Kurt Cozens from NEI. In
,r~~
(_,1 25 response to the Zorita-type of data and the request from NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
131 1 the NRC to ask each licensee to look for that information,
,y
, 2 Zorita was a unique operator experience where they had let t i V
3 go several hundred liters of resin into their vessel, 4 which was permitted to stay in there for some extended 5 period of time.
6 The particular penetrations which exhibited 7 circumferential cracking were penetrations that were not 8 used -- they were capped, they were stagnated -- after the 9 licensee, I guess Zorita decided to shut own the plant to 10 take care of the problem. It sat in standby mode for some 11 extended period of time -- a number of months.
't 12 It was only in those particular penetrations
,_ 13 that Zorita found circumferential cracking. We didn't I i 14 believe -- NEI did not believe that the request was 15 consistent with the Generic Letter. First of all, the 16 draft Generic Letter was issued solely for primary water 17 stress corrosion cracking. This particular type of 18 cracking observed at Zorita was not a form of primary 19 water stress corrosion cracking; it was a totally 20 different mechanism.
21 Adding onto that, that internationally 22 approximately 90 plants have inspected there, had 23 penetrations. And none of those plants which are operated 24 in a manner very similar to the United States plants, had
("h
() 25 exhibited any circumferential cracking or degradation NEAL R. GROSS COURT REPCRTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C 20005-3701 (202) 234 4433
l l l
i 132 l t
- 1 modes similar to that experienced in Zorita.
l g3 2 Plus we don't believe that anybody's had resin
(~) 3 intrusions on the order of magnitude that Zorita )
l 1
4 experienced; which was again, several hundred liters of 5 resin into the primary systems which sat for months on i
6 end. j 1
1 7 So that's why we disagreed with the need to l
l 8 incorporate the Zorita exposure into the Generic Letter. l l
\ l 9 MR. CARPENTER: And basically, the reason that I l
1 10 we incorporated it was because we don't know if anyone has I
11 had that. We felt it would be a good idea to know that.
1 1
12 Following the conclusion of the comment
, r~N 13 period, shortly before we went to CRGR for the final )
\\ ') 14 acceptance of the Generic Letter, NEI submitted some 15 additional comments which basically still disagreed with a !
16 need for the Generic Letter.
1 17 One comment pointed out some of the 18 limitations that are inherent in the susceptibility models 19 that we have yet to see. And a lack of historic data was 20 also pointed out.
21 We felt overall, that the comments did support i
22 the need for further inspections.
23 The Generic Letter itself considers cracking 24 of the vessel head penetrations to be a safety cc; orn fer b
x 25 the long-term based on potentially exceeding the ASME Code NEAL R. GROSS i
COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W.
1
, (202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433 I
i
i 133 1 standards for margins of cracks that are sufficiently deep
,, 2 and continue to propagate. l N/ 3 And also by eliminating a layer of defense-in-l l
l 4 depth for plant safety. And what the letter is requesting 5 is information from PWR licensees to provide adequate 6 assurance that margins and defense-in-depth are being 7 maintained for the long-term.
8 The proposed Generic Letter also is l
9 requesting, in items 1 and 2, to verify compliance with 10 10 CFR 50.55a and Appendix A, General Design Criteria 14, and 11 determine if the imposition of an augmented inspection I l
12 program, pursuant to 10 CFR 50.55a (g) (6) (ii) is necessary. i 13 And further, the information requested, some
/~'s l
- 14 information on resin intrusions. My apologies. I should 15 have updated this slide. It is now -- items 1 and 2 are j l
16 now combined into item 1, and item 3 is now item 2. So if 17 you read the Generic Letter and you have some confusion, 18 that is my mistake.
19 And I seem to have left the final slide off, 20 in my office. Basically, the staff conclusions are that l l
21 an integrated, long-term program, including periodic
! 22 inspections and monitoring, is necessary, and that the 23 staff needs to better understand the industry's program 24 and its technical bases.
x ) 25 The NRC staff has concluded the vessel had NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 2344433 l
1 1
134 l 1 l
1 penetration cracking. It's not an immediate safety l
l 7
-. 2 concern provided the Generic Letter 88-05 inspections are l \)
3 performed and non-destructive examinations are performed l
4 to ensure no undue, unexpected cracking. And we recommend l
5 implementation of enhanced generic detection capabilities ,
l l l 6 by the industry. !
l I l 7 That concludes my presentation. Any questions !
l 8 I might answer? !
l 9 ACTING CHAIRMAN SEALE: Any questions? l I
10 MEMBER SHACK: What is the local leak I i 11 detection that the Europeans are using? l 12 MR. CARPENTER: It is a system called N-51, I )
l 13 believe. And as I understand it, it is basically a I 1
( m\
'- ' 14 monitoring system that will allow detection of up to a 15 half-gallon, or as little as a half-gallon a minute l l
16 leakage from the head. And my apologies but I really 17 don't have a lot of detailed data as to how they do that. ,
l 18 MR. STROSNIDER: This is Jack Strosnider. I 19 just think maybe a brief comment might be in order. It's 20 obvious from the presentation that this is a subject where 21 the industry and i;he NRC agreed to disagree with regard to 22 the need for this letter.
23 I want to point out that the real thrust of 24 this letter -- it's our understanding that the industry, g)s
(_ 25 based on our meetings, that they have a program, that they NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
I (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
135 1 have the susceptibility modals, and they are planning cs 2 inspections to verify those models, and it's driven by, kt') 3 among other things, at least some economic considerations.
4 The real thrust of this letter is that we want S an opportunity to look at those models and look at what 6 inspections are planning to verify the models. We have 7 reviewed the analyses in detail. That is one thing that 8 was submitted to us -- the safety assessment -- and we 9 agreed analytically, with the predictions that we 10 shouldn't see circumferential cracking of any significant 11 extent, and that's why we don't see it as a safety concern 12 at this point.
13 But we would like to be able to look at the gg
-' 14 model and verify that in fact, it appears to be reasonable 15 in terms of susceptibility and that the inspection that 16 they're planning is going to give us some assurance that 17 we'll be able to continue to have confidence in the 18 analyses we've looked at.
19 The important point I guess, saying it another 20 way is that, this letter doesn't say that every licensee 21 needs to go out and inspect every penetration. We 22 recognize the significance of this issue, we don't believe ti 23ditreallywarrantsthat, and it's g"ite reasonable to have 24 a generic industry integrated appr tch to this.
(,) 25 We just need, we feel, to have a better NEAL R. GROSS COURT REPORTERS AND TRANSORIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
136 1 understanding of it here at the NRC so that we can vouch f~x 2 to the reliability of that program. So that's really the
(' ' ~)
~
3 information that we're looking for in this letter.
4 ACTING CHAIRMAN SEALE: Any other questions?
5 Any comment to the industry in addition, at this time?
6 Well, let's see; the fire hose is big enough. I guess we 7 ought to take a break. Is there any reason we have to 8 wait any particular time to start again?
9 MR. STROSNIDER: Jack Strosnider. All our 10 next presenters are here today.
11 ACTING CHAIRMAN SEALE: They're all here 12 today. Okay, fine. Well, let's say we'll tentatively
,, 13 plan to start at ten minutes to ten, but if something I )
' 14 happens and the Palisades people won't be here but could l l
1 15 be here shortly, we'll check on that.
16 (Whereupon, the foregoing matter went off 17 the record at 9:39 a.m. and went back on 18 the record at 9:59 a.m.)
19 CHAIRMAN SHACK: Well, I guess we're ready to 20 resume and I guess our next presentation will be on the 21 status of the Palisades reactor pressure vessel 22 activities. Lambros Lois will make the presentation for l
l 23 NRR.
24 MR. LOIS: Good morning. My name is Lambros
/'~'T
(_,ri 25 Lois. I'm a member of the SRXB Reactor Systems Branch and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
137 1 I'm going to take you through the process and thought i 7s 2 process and how we arrived at the decision on Palisades
( )
v for the value of the fluence in the last -- almost a year 3
4 now.
5 The topics I would like to discuss -- I have 6 some background to tell you how we arrived at this 7 position. We reviewed the analysis and staff concerns, 8 the staff position which we have already have published, 9 and the results of that decision.
10 Back in 1992/1993 we reviewed a submittal from 11 Palisades as to the value of the fluence and we came to 12 the conclusion, and to the agreement I thought, that the 13 value of that fluence wouldn't allow them to operate in
,r) f /
\/ 14 1999. In 1996, about a year ago, Consumers Power 15 submitted a revised fluence -- which took us by surprise 16 in some respects -- but proposed a 25 percent reduction to 17 the value we agreed in 1992/93.
18 Anyway, we broke it down to about 8 percent of l
19 that reduction, proposed reduction, was due to physical 20 changes at the plant. There were three kinds of changes 21 which we'll see later on. There were another reduction of 22 ebout 12 percent which was due to averaging of 23 experimental data, and essentially adjustment of the 24 calculated value of the fluence.
/~T
( ,j 25 And then finally, another 5 percent which was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
138 l 1 due to, again, manipulation of the data via a code called ;
1
. 2 FERRET, which essentially goes through an adjustment of
/ \
1 l t
'~
3 the spectral for each ramification of the measurement 4 data.
5 Let's go back and go over the review and 6 analysis. The review focused on plant configuration, the ;
7 data changes, calculation methods, measuring the data and 8 corrections that were proposed, and finally, proposed 9 spectral adjustments.
10 One configuration included a SIMULATE-3 pin 11 power distribution. Before I proceed with this, let me 12 tell you that the staff has taken the position in the past 13 that we would accept calculated values of the fluence. 1
( I
's / 14 This calculation should have been based on a qualified 15 code and the NRC in the last 10 or 15 years has produced 16 benchmarks to qualify such codes.
17 These were done at Oak Ridge and are known 18 with PCA and PCF experiments. They were experiments which 19 were produced in simulated geometries in low power 20 reactors but adequate to measure the fluence, with the 21 exact, if you wish, exact to the measurement capability, 22 geometrical details. So we felt that those had quantified l
l 23 uncertainties, therefore a code could be benchmarked to 24 such experiments.
Lj) 25 Okay, let me now go back to the proposed NEAL R. GROSS COURT REPORTERS AND TRANSCRIE,ERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
i l 139 l
l I 1 changes in the plant configuration from Palisades. One of l
s 2 them was in the early -- the calculations of the early !
( ) I
\
L~/
3 cycles, presumably the fluences and the fluxes at the i 4 outer assemblies, which determine, by-and-large, the value j 5 of the fluence to the pressure vessel -- were determined 6 with diffusion theory codes and therefore we're some --
7 quite optimistic.
8 SIMULATE-3 is a code which does a much better 9 job and it's based on transport methodologies, has better 10 cross sections, presumably, and this is a code which we 11 have been using and has been accepted by the staff by a 12 long time. Those changes were acceptable to the staff. l 13 The second one was that if a downcomer
(~N
' '/ )
14 increased temperature. We went through that, we looked at !
15 the arguments presented, why the temperature in the 16 downcomer should be acceptable, and we found that it was a l 17 justifiable argument. Namely, the temperature would have 18 been lower, therefore, the density would have been higher, 19 and therefore the attenuation from the edge of the core to 20 the pressure vessel would also be higher. Therefore, this 21 will contribute to some extent, into a lower value at the 22 peak fluence of the pressure vessel.
l 23 Finally, the licensee has performed a large 24 number of measurements of the diameter of the vessel, and A
(_) 25 also of the vessel thickness. The original value was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4 433
140 1 taken from design drawings out was not the as-designed
,s 2 value.
( )
%_./
3 A number of measurements were large -- 152, in 4 that neighborhood, and so we felt that also those values 5 were acceptable and were justifiable, and therefore the 5 diameter of the vessel should be increased by about an 7 eighth-of-an-inch, and also the thickness of the vessel.
8 Now, the inside -- the vessel capsules of 9 course, are not going to affected by thickness of the 10 vessel but the cavity dosimetry which are extensive in 11 this case, might have had some sort of --
12 MEMBER KRESS: Were these measurements made at ;
i 13 the mid-plane?
14 MR. LOIS: These measurements -- no, my l
15 understanding is they were not made at mid-plane, but the i
i 16 licensee is here so they can probably answer that question l 17 more precisely.
18 So by-and-large, we felt that those proposed 19 changes to the plant parameters were acceptable.
20 Therefore, the value of the proposed fluence modification, 21 namely reduction by about 8 percent, was also acceptable.
22 This is 8 percent with respect to the calculated value.
l 23 I might also say that our calculated value 24 which we independently did at BNL, agreed quite well with
( ,) 25 the calculated value that Westinghouse had. We were able NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l
l 141 l
l 1 to do this because 1992/93 we also had disagreement as to fy 2 the acceptable value and at that time we set up an input
- t i /
3 deck, so this time was somewhat easier for us to go back l
4 and make the adjustments -- that they were rather minor i
l 5 with respect to the '92/'93 version -- and re-calculate 6 the proposed calculated value at the peak.
7 Now, the calculational methods that were used 8 -- we also used at BNL -- involved this P 3 -S, transport 9 approximation, their 2-dimensional, finite element code 10 which is generally acceptable. We also used cross 11 sections based on P 3-S 6 cross sections, which is also 12 acceptable and is also widely recommended by the staff, 13 and we're using ourselves, and of course, pin-wise source
(','s
\ /
14 distributions at the outer assemblies which as I said, 15 determine the value of the fluence by a large extent.
16 And finally, detailed dosimeter modeling, 17 which of course were used for the comparison of the 1
18 calculated dimensional values because the essence of our 19 difference with the licensee were right there as to 20 whether or not the calculated value was more reliable 21 compared to the measured value. So as I said previously, 22 we confirmed that the 8 percent reduction was an l
l 23 acceptable value.
24 However, the proposed reductions -- the large
/ ~s (x ,)' 25 reductions were the 12 to 5 percent which were based on l NEAL R. GROSS
! COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
1 142 1 the measured values of the fluence. Now, this topic is
,s
, 2 somewhat, quite considerably far away from the materials
, 1 3 and methodology -- environment is change of venue, so to 4 speak -- so let me tell you that the changes that we have 5 are quite difficult, the measurements are quite difficult, 6 and when we speak about the measurement, the content of 7 the measurement is very small part of the final value.
8 The remaining input for that measurement is 9 really a calculation, and one of the most important 10 aspects of that is that the measured values need a 11 spectrum to be interpreted. A measured spectrum doesn't 12 exist in the power industry, so to speak, because it's 13 very difficult to do.
F
\- 14 However, there are techniques which could do 15 that, and in a variety of experiments that exist in the 16 literature here and elsewhere, and abroad, the neutron --
17 I'm sorry, the proton recall method has been used, and 18 those spectra are far more reliable than spectra that can 19 be obtained from the more commonly used method of 20 unfolding. Namely, having the integral measurement of 21 foil, which then one attempts to interpret in terms of the 22 assumed spectrum at the position of irradiation.
23 And something else of course, is that the --
24 is that as you move farther and farther away from the edge CA
( ,1 25 of the core, the spectrum in itself gets modified. It's a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND ANE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
143 1 different value, different spectrum at the dosimeter; it's
, ,m 2 a different one at the inside surface, the pressure
( )
\ /
3 vessel, and even far more different on the ex-vessel 4 measurements -- namely the reactor cavity.
5 So one has to bear in mind that when we speak 6 about the measurement, all these complications play a 7 role. So we looked at the reductions and as I said, the 8 measurement to calculate the values they suggested the 12 9 percent reduction. The proposed mode or method if you 10 wish, is that the best estimate fluence -- which is 11 required, by the way, this particular case, by the 12 regulation 10 CFR 50.61 - says that the value should be 13 the best estimate.
,m i
'- Y 14 The proposed value is that the calculated 15 value should be multiplied by an M/C plus the spectral 16 adjustment, which is also a reduction of about 12 and 5 17 percent, which amounts to 17 percent. The database 18 consisted of in vessel dosimeters, cavity dosimeter 19 measurements, and the choice was of six different 20 dosimeters: it's an iron, nickel, copper, titanium, U-21 238, and neptunium-237.
22 As far as their energy response was concerned, 23 you have high energy -- which is copper and titanium --
24 and everything else low energy. In this particular case,
(
(_,) 25 high level energy are defined in the neighborhood of about l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 l
l 144 l
l l 1 4 to 5 MeV. There of course, there's another distinction
! ,s 2 one has to make with these dosimeters -- namely their i i( ) i
~#
3 half-life, or decay half-life, and how far they can see l
! 4 into the past irradiation.
5 They range from about 70 days to about 30 6 years. Their responses and their accuracy is not uniform, l
7 either. That's another complication but I guess we will l
, i 8 get some details. When we looked at the data that was l l
9 presented by the licensee, we were able to break them down 10 into high energy in vessel, low energy in vessel, and high 11 energy in the cavity, and low energy of the cavity.
I 12 Now, these groups as I have them down, as you !
13 see them, are not of the same, let's say, quality or 7_.
( )
's / 14 weight, and cannot be. There are several reasons for 15 that. The in cavity dosimeters, of course they have the 16 problems with the spectrum as I mentioned earlier, because 17 now you attenuate through an 8-plus inches of steel, it's 18 quite a bit.
19 The other problem is that the in cavity and 20 the ex vessel dosimeters are not the same number, and 21 they're not the same quality. Everybody knows in the i
22 industry that neptunium-237 has problems in the l
l l 23 measurement and in its qualification.
24 Likewise, there is a large number in this
,m,
(,) 25 particular case of iron nickel measurements which are in NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
k
- 145 1 the cavity Now, those measurements, although it's a
-m 2 large number of them, cannot statistically, at least in i
N" ,/
3 our minds, cannot statistically be assumed to be of the 4 same weight as those in the cavity, because essentially 4
5 it's the same measurement.
6 You ake the same measurement over and over 7 and over again, but when you count them, you can say, well 8 now I have 100 of those, because they're the same 9 position, the same thing, over and over again.
10 Anyway, with those -- those arguments it's 11 difficult to quantify, but regardless, if we just take the 12 nominal number of dosimeters that were presented, one 13 would get the high energy in vessel to indicate that
/\
i l V 14 there's no problem; that the calculated value and the 15 measured value should be roughly the same.
16 However, when we move to the low energy in 17 cavity and we put some great er f aith, those in vessel 18 rather than the ex core. Although, I must say, that we 19 have endorsed measurements in the cavity as the only 20 alternative for some reactors where in the future, should i
21 there be a life extension of the vessel, there will be no 1
1 22 other way of doing it, short of putting new capsules in 23 which are rather expensive and probably not a good choice 24 in that respect.
,O.
() 25 So that says that 14 percent difference in the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l
146 1 negative direction and then it would, the cavity, even the
,e y 2 high energy suggests that it's probably 9 percent i )
3 deficiency and at the outside would be probably around 15 4 percent deficiency.
5 Those deficiencies -- those differences, I 6 should say -- they seem significant to us, and we did not 7 feel that this was a consistent story. Let me give you 8 some additional values on that.
9 We looked somewhat farther. In other words, 10 what I'm trying to suggest here is, that the values for 11 the plant-specific measurements did not seem to be 12 concistent by themselves. Now, let's look at the
, 13 methodology that was used to produce that.
52 14 We looked at one plant, Plant A let's see, and :
15 Plant B, and Palisades. And we found out that taking the 16 iron nickel and -- for example, taking the iron nickel in 17 this particular case is that, because they are by-and-18 large, the largest contributors to the proposed 19 modification of the fluence.
20 We find that the same method yields an l 21 addition of 13 percent, thereabouts -- 11/12 percent in 22 this case. Another plant it's at 3 percent, and the i 23 Palisades it's a deficiency of about 14/15 percent.
24 So therefore, we come up with an inconsistency
(_,) 25 in the plant-specific data between high and low energies, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
{
l 147 1
1 inside and outside the vessel, and then inconsistencies i
i 2 across the methodology in itself. So that really we
, (7x l
, i
'~'
! 3 found, was very difficult.
4 Well, to put these data in a more graphical 5 sense -- let me show you his one. There's Plant A, Plant l
6 B, the uncertainties, and the Palisades. These somehow, 7 don't quite jibe. By the way, Plant A and Plant B just l 8 happened to be on my desk because I reviewed some other 9 issues with them. It was not that I looked to find other 10 examples.
11 There may be some other examples which using 12 the same methodology, may be farther apart; I don't know 13 that. But I just happened to have those values.
- i
'V '
14 I raised this issue with the licensee and 15 Westinghouse, and what I was expecting to find, to hear, 16 was some sort of a physical or physics-based reason which 17 would explain, well, here we know that in this case the 18 spectrum for such-and-such reason is this much or the 19 other, and taat's why I have this value here. This plant 20 responds differently because these two values are 21 somewhere in the other side.
22 Well, I know that these two plants, Plant A 23 and Plant B, do have thermal shields which Palisades 24 doesn't have. That may be a cause, but if that is the
(_) 25 case then the in cavity -- I mean, the ex vessel NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND /dE., N W.
(202) 234 4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
148 1 measurements then will be suspect, because then one can l 7, 2 suspect when they figure at cross sections to the pressure f i 1 3 vessel -- I mean, to the iron in the pressure vessel and 4 therefore had problems with the spectra in the cavity. l l
1 5 Which in this particular case formed the bulk of the i l
6 measurements in -- yes sir? l 7 MEMBER KRESS: Isn't the operative 8 uncertainty, the band about the particular point, though?
9 Isn't that the most important?
10 MR. LOIS: 'les it is -- well, it is --
11 MEMBER KRESS: It just -- this slide tells me 1
12 that -- we're looking at M/C -- it tells me that the C 13 spreads out much higher than the M, and it's the M error i
'w) 14 band it seems to me, to be the important one.
15 MR. LOIS: Not necessarily. Because as I 16 pointed out earlier at the beginning, the value of the M 17 has a large amount of C input -- with respect to those )
18 calculations. ;
1 19 MEMBER KRESS: You're saying'you can't really 1
20 separate it.
21 MR. LOIS: Can't really separate it.
22 MEMBER KRESS: But then I don't understand l
23 what that error band is on --
24 MR. LOIS: You have that many results give --
- y
() 25 find an average standard deviation, if that's all there is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l
1 149 l 1 to it.
l
- 2 MEMBER KRESS: Oh , that's what that --
i
, y!
3 MR. LOIS: Yes.
4 MEMBER KRESS: Okay, now I understand.
5 MR ., LOIS: That's the same -- yes.
6 MEMBER KRESS: And that's not a real estimate 7 of the error? lt's just --
3 MR. LOIS: Not really, yes.
9 MEMBER KRESS: Okay.
10 MR. LOIS: That point we're clear on.
11 MEMBER KRESS: Thank you. That's helps a lot.
12 MR. LOIS: Now, but all these and the 13 queations we heard from the licensee -- some of the
-- 14 questions were quite difficult, so we turned our attention 15 to see what other people were doing, particularly other 16 U.S. vendors and in other countries. And particularly we 1
17 turned to U.S. vendors. l 18 And as far as we know, nobody else applies 19 this methodology. And this methodology in particular I'm 20 referring to, taking the, what you would call the measured 21 data and averaging, and then say the calculation is not 22 good, and therefore we adjust everything to the measured 23 values.
i l
24 Then we looked at the German and the French (p _,
25 methodologies. Now, the French methodology happens to be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 ,
l 150 l
l 1 very close to what another vendor in this country, through
',s 2 I guess, corporate interaction -- they have a similar
)
/
3 methodology. However, they do not use calculated values.
4 They have produced a calculated methodology, 5 and from there on, they measured -- the plant-specific 6 measured values are no longer used. They're just going to 7 use to see whether or not the value is in the right, the 8 proper area of -- I mean, in the ballpark, but not to make 9 any adjustment beyond that.
10 Then we looked at the German methodology.
11 German in this particular case means all of the Siemens 12 values, KWU. And we found that likewise, do not use any 13 adjustments to the measured values.
O 14 However, I should hasten to point out that 15 both of them do not have those large differences that we 16 saw in the domestic, or more specifically in the 17 Westinghouse and of course Palisades, measured data. The 18 values are in much, much better agreement with that 19 calculation. So therefore, don't have motivation to 20 produce such -- to resort to such adjustments.
21 But specifically, I tried to see what they do 22 in the spectra point of it which is, could also be 23 significant. And nobody, at least as far as I know, l
24 adjusts the spectra to the calculate values, or use the
(~T s ,) 25 calculated values -- I'm sorry, use the measured values to NEAL R. GROSS COURT RFPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
151 1 adjust the spectra that they have.
< s. 2 In fact, there was a paper which was published v' 3 recently, and after that -- recently as last 4 September / October -- by a German called Polke, and 5 described BWRs and PWRs. One of the arguments that we 6 heard was that, well, why should the methodology show 7 differences between Plant A and Plant B and Palisades? We 8 say well, these are different plants.
9 Okay, well then, Mr. Polke comes along and 10 says, we have more than 20 BWRs and PWRs, and their 11 dosimetry is independent of reactor type, reactor location 12 -- I mean, dosimeter location, and type of dosimetry. And p_ 13 that really, makes it very difficult for us, for the
- \
\' /
14 staff, to come along and say well, now I don't care about ,
I 15 the differences in the plant-specific measurement results, 16 or in the methodology inconsistencies, and I'll go along 1
l 17 with that. That's extremely difficult to do.
18 I looked into the specific reasons for that 19 consistency, at least in their evaluation, and they said, 20 they attributed it to a number of factors. Number one was 21 that they have measurements which are performed for the 22 last 30 years or so, by the same people and counted in the
! 23 same laboratory. So in that sense they have consistency.
24 The also pointed out that their analyses is
,e s_) 25 performed with the same methods for the improved cross NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
! (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
152 1 sections, but again, by the same people who provide their l l
~s 2 intuition. It's probably needless to point out that those I \
3 calculations and the measurements, are extremely difficult i 4 to perform and results don't always come out as 5 consistent.
6 In contrast, the Palisades dosimetry results 7 depend on dosimeter type and dosimeter location, as I 8 pointed out to you earlier. So if for example, the choice 1
9 of dosimeters was different, the results would have been l l
l 10 different, and that's really not acceptable. I can't get l 11 along with that. ad also, there are no rules how one 12 would choose a sat of dosimeters. i l
1 13 Polke and company, they used iron and niobium, '
< i N' 14 and that's it -- these two. They said, iron and niobium l
15 cover the half-lives and the expanse of the spectrum we're l 16 interested in, and therefore, we don't need anything else.
17 Now, had they used copper and titanium and 238 18 and neptunium-237 and so forth, might have similar 19 difficulties. Their wisdom was that, no, no, these two --
20 we're thoroughly acquainted with them, we know exactly how 21 they behave, we don't need anything else. That was the 22 end of it.
23 Now, the French used a larger number of 24 dosimeters -- I think it's about four or sometimes five.
(^';
(_) 25 The French, again, they have another problem, another NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433 i
153 1 problem -- but in reality, as far as we are concerned, l
,_ 2 they rave the same location, the same dosimeters in all I
i
' 3 types of reactors. Not types in French reactorc from very 4 much this only the distance in radial direction that may 5 vary, but otherwise materials are just about the same and 6 we expect the same type of behavior.
7 But again, they have a consistency between 8 measured and calculated values, which we don't have. In 9 our case, in the Palisades case, it was called a bias. Of !
10 course, the popular concept of bias is that it is a system 11 deviation -- systematic error, if you wish. But then we 12 looked at the plant-specific measurements and don't seem 13 to be systematic. Then we looked at the methodology xj 14 deviations; don't seem to be systematic either. So we 15 cannot classify those as systematic error or as a bias.
16 This was the 12 percent. The 5 percent ,
1 17 spectral adjustment were based on the code called FERRET.
18 The code does have good theoretical background, but it l 19 does have numbers in it as far as the spectral adjustments 20 were concerned, which I really don't know whether they l 21 were related to the reactors in general, or Palisade a in 22 particular. l l
23 The code was never submitted for review, we 24 did not review it, we did not understand it, so we didn't
( ,) 25 accept that. However, regardless of whether we reviewed NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
1 154 1 the code or not or whether we accepted the results or not, 1
.s 2 the basis of the spectral adjustments were also the plant-I ')
3 specific measurements. And those, as we saw earlier, we 4 considered to be inconsistent, did not give us a story in 5 the same direction, and therefore we could not accept it.
6 So therefore, the staff had to deal with 7 plant-specific data which are not consistent. The second 8 one was, no physical explanation was provided for the 9 deviations that were present. Finally, the averaging -- l 10 therefore we found the averaging did not represent the 11 best estimate, and therefore we should rather stick with 12 the calculated value which just about everybody else is
)
13 using, and gives us a consistent story, at least in our l t ,
<- 14 expectations.
15 So therefore, the staff position was that the 16 safety evaluation submittal is -- the reduction based on 17 plant physical changes has been approved because we found 18 that it was justified because these were values which we 19 could account for and when counted in our calculations, 20 gave us the same value as we can get in the -- with the 21 licensee.
22 However, the reductions that were based on l 23 plant-specific data, the averaging and the spectral least 24 squares fitting, we disallowed that.
,\
) 25 The consequences of that decision is that the l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l
155 1 '92/'93 fluence value would allow operation till 1999. j l
s 2 The 8 percent reduction will delay that deadline to 2003. l
('^') 3 Now, as far as the Palisade is concerned, the licensee !
4 expires in 2007; however, they can recapture a 4-year i 1
5 construction permit, so therefore, all they have to do is 6 just ask for it. The license will then expire in 2011.
l 7 So the consequence of our decision is that the period j l
8 between 2003 and 2011 is not covered.
9 This is the end of my presentation. Any 10 questions? )
11 MEMBER SEALE: Are you going to stay around?
12 MR. LOIS: Yes sir.
13 MEMBER SEALE: I guess we'd like to hear the 7m
! )
v 14 other shoe drop.
15 MR. LOIS: Of course. Thank you, Mr.
16 Chairman.
17 CHAIRMAN SHACK: I guess our next presentation 18 is on a Draft Reg Guide that will resolve all these 19 problems in calculating vessel fluences. Ms. Fairbanks.
20 MS. FAIRBANKS: My name's Carolyn Fairbanks.
21 I'm with the Office of Research; I'm a Materials Engineer.
i 22 Today I would like to discuss the status of Draft Guide l
23 10.53, Calculational and Dosimetry Methods for Determining 24 Pressure Vessel Neutron Fluence.
(_,) 25 l'm the Project Manager for tnis task. I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
156 1 assumed this duty last July from Al Tibota who retired, i l
<~s 2 and the ten minutes given on the agenda is not a typo.
l V) 3 What I really wanted to discuss today was to brief people 4 on the current status of the Draft Guide and primarily 5 discuss and make people aware that there will be a new 6 section added to the Draft Guide.
7 MS. FAIRBANKS: The handout has an outline in 8 it. First I want to go over a few background items, 9 background and objectives, regulations that are related to 10 the Draft Guide and the development of the Draft Guide.
11 Just go over them briefly. Then we will discuss the 1
12 status of the Draft Guide and a new section in the j 13 schedule for completion.
- \ 1
' 'j 14 Reactor pressure vessel fluence is required in l 15 determining vessel embrittlement and operating life time.
16 Fluence is an important factor in several significant l l
17 regulations.
18 The current methods submitted by licensees 19 vary widely at this time. Basically, they are done on a l 20 plant-specific basis. So the objectives in the Draft 21 Guide were first to provide an acceptable state-of-'the-art 22 method for determining fluence by which we can ensure 23 accurate fluence predictions and quantify uncertainty, as 24 well as reflect present-day experiences and surveillance 7 ~.
(_, 25 report submittals.
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W.
! (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
157 1 Secondly, we wanted to standardize the vessel s 2 fluence determination methodologies and accomplish that by
\
\ /
\
'~'
3 providing a consistent set of guidelines for estimating 4 neutron fluence exposure to the reactor vessel.
5 This would provide the additional benefit of 6 eliminating unnecessary work for licensees and staff for 7 the submittals that are now given on plant-specific basis l
8 for reviews.
9 An additional factor taken into consideration 10 with the Draft Guide is that workstations are now 11 commonplace and they are capable of running the 12 sophisticated codes in reasonable periods of uime.
13 As a point of reminder there are three
, ', 2
-- 14 significant regulations related to Draft Guide 1053 in the l 15 use of fluence. Those would be Appendix G, the fracture i
16 toughness requirements in which fluence is used to 17 determine upper shelf energy.
18 The PTS rule in fluence is used to determine 19 adjusted RPTPS &nd is used in comparison in screening 20 criteria in the PTS rule.
l 21 Then Appendix H in which fluence is used to l
l 22 evaluate surveillance data.
23 A brief discussion of the development of the 24 Draft Guide. It was done with combined expertise from the r~b
(_,) 25 Brookhaven National lab, Oak Ridge National lab, the l NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l l (202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433 t
t
158 1 National Institute for Standards and Technology, as well 73 2 as input and guidance from the staff at the NRC.
i i
\~/ Could I ask a question?
3 MEMBER KRESS:
4 MS. FAIRBANKS: Yes.
5 MEMBER KRESS: If you have surveillance data, 6 are you talking about coupons that you take out and 7 naturally measure the shifts in the --
8 MS. FAIRBANKS: That is a part of the Appendix 9 H requirements but, maybe Lambros can elaborate a little 10 bit because I think --
11 MEMBER KRESS: But my question is if that is 12 what we are talking about, does the fluence determination 13 that one makes in relating- the shifts in the surveillance
,3 1
' / 14 data and using this same fluence determination to see what 15 happens to the vessel, does it cancel out?
16 MR. LOIS: This is Lambros Lois. It is a very 17 difficult question that you are asking because while we 18 have measurements in the surveillance capsule, we also 19 have some questions whether or not that measured value is I 20 the correct value.
21 MR. ELLIOTT: Barry Elliott, NRR materials 22 engineer. That is a good question.
I i
23 Most of the plants in America don't have the 24 actual limiting material in their capsule.
g~s
(,) 25 MEMBER KRESS: I see.
, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS i
1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
159 1 MR. ELLIOTT: So, the plants that do have it, s 2 it cancels out. In the case of Palisades, they did not
( )
3 have, at the time we did the evaluation, the limiting 4 material in their vessel. So, they were using generic 5 data to determine the amount of embrittlement and using a 6 fluence calculation to determine the calculate the amount 7 of total embrittlement.
8 MEMBER KRESS: Thank you, that helps a lot.
9 MS, FAIRBANKS: To continue then, the Draft 10 Guide reflects results from the lightwater reactor, the 11 pressure vessel surveillance, dosimetry improvement 12 program. It also reflects experience that we've gained in 13 performing the independent calculations of reactor vessel
- i
/ 14 fluences, input and experience gained by the contractors 15 and the staff that have been involved in this work.
16 Also, references, updated cross-sections, 17 specifically the ENDF/B cross-section libraries.
18 The Draft Guide itself contains a detailed 19 description of fluence calculation and measurement 20 methods, procedures for qualifying the calculations and 21 measurements and a table of specific modeling dosimetry 22 and qualification and reporting requirements.
23 There is an accompanying benchmark problem 24 which is NUREG/CR-6115. For the most part this is a
-rm s
(_,/ 25 requirement. The only special circumstances in which this !
l NEAL R. GROSS l COURT RdPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
160
! 1 would not be used would be if there were no reliable l
2 measurements available so that they could not benchmark l 7-(-)) 3 against a similar plant or to the specific reactor type.
l 4 We would expect that to occur very infrequently.
5 or in the case where the number and quality of 6 measurements allow determination of accurate calculational i
l i 7 bias, the bias can be applied directly to the predictive 8 fluence.
9 The current status of Draft Guide - 53 goes 10 back a few years. Work began upon request from NRR. The 11 initial version was designated Draft Guide 1025 with the 12 same title.
j 13 Draft Guide 1025 did go through the approved p.
/ \
' 'j
\ \
14 release reviews and it was released for public comment.
15 After the public comment period, there was a formal review i
l 16 meeting with held with industry. The comments regarding i
17 the Draft Guide were evaluated and where appropriate, were 18 incorporated.
19 Every comment was addressed and resolved ind 20 the resolution was documented.
21 The latest version of that incorporation the l
22 appropriate changes is designated with the same title as i
t l
23 Draft Guide - 1053. Recently there was a meeting with 24 industry to review this and the feedback at that meeting
(_,) 25 was basically very favorable.
l NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS
! 1323 RHODE ISLAND AVE., N.W.
l (202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-44'"
l
161 1 One point, however, that was brought up in
<3 2 that meeting and in the latest submittal of comments was a l r t i LJ 3 suggestion for a new section addressing the application of 4 Monte Carlo transport to pressure surveillance fluence 5 benchmarking problems. There was a consensus of agreement 6 among the NRC staff and the contractors on this work that 7 that should be included.
8 There will be corresponding changes made to 9 the benchmark problem as well.
10 The schedule, in order to accommodate the 11 inclusion of this new section, is that the responses 12 received from the last meeting that was held with industry p._
13 are being evaluated and, as appropriate, changes will be
( )
'\ / 14 made to the Draft Guide.
15 The Monte Carlo section is expected to be 16 completed in January of 1998 and final release of the 17 Draft Guide is anticipated for June, 1998.
18 Any questions?
19 PAUSE 20 CHAIRMAN SHACK: Our next presentation will be 21 from Mr Hanson of Consumers Power on the licensing plans 22 for the Palisades.
l 23 MR. HANSON: Good morning. I am Jack Hanson, 24 Director - Strategic Issues for Consumers Energy at the
,f'~)
(s,) 25 Palisades nuclear plant.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBE 3 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
162 1 I appreciate the opportunity to come talk
,3 2 about Palisades' reactor vessel and integrity status with l I i
~
3 you this morning.
4 I just wanted to jump in. Lambros kind of 5 touched on this; I wanted to put the background in place.
6 So you can understand what our current 7 situation is that our current plant license expires in 8 March, 2007; with the recovery of our construction period, 9 we could operate to 2011.
10 The current interim SER concludes that our 11 screening criteria is reaching 2003. Our submittal 12 concludes that the screening criteria is reached in 2012.
13 So, that is basically the background and the r~N
- 14 overall picture of where we are.
15 I would like to get into what we view as our 16 options to achieve the end of our license. We are 17 presenting these in what we believe to be the most 18 appropriate and most preferred order in terms of how we 19 would deal with this.
20 one of course, is furtner approval of our 21 April 4, 1996 fluence submittal. We are going to get into 22 that in much more detail, not quite as much detail as t
l 23 Lambros got into, and that will be in section three of 24 this presentation.
,r3
( ,) 25 The other options we are looking at, the next NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
163 1 option in line is to go to an ultra low leakage core and l
- 2 reduce the rate at which we accumulate fluence via core 3 management.
4 We have done a great deal in this area l
5 already. As you can see, these are our earlier cycles.
6 Cycle I and Cycle II were at reduced power operation.
7 This is flux at the vessel locations. These are our early l
8 cycles. We have reduced our flux levels at the vessel 9 wall by nearly a factor of four for operation.
10 MEMBER KRESS: How did you do that? Is that a 11 combination of things?
12 MR. HANSON: Core design.
13 MEMBER KRESS: Core design. You are still
/"N
\_ ')
14 operating at the same power level?
15 MR. HANSON: We are still operating at the 16 same power level. What we have done is we made some l 17 changes to our reactor protection system. When we were ,
1 18 first licensed we had no limits on axial shape index and 19 we had no way of monitoring axial shape index from the 20 reactor protection system perspective.
21 We changed our reactor protection systen, and 22 we essentially pushed the power into the center of the 23 core and away from the periphery of the core.
24 We believe that we can get another 40 per cent (g,) 25 reduction in fluence and that would get us to 2007 which i
l NEAL R. GROSS j COURT REPORTERS AND TRANSCRIBERS j 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
164 1 is our current end-of-license.
- (p) 2 To do that we will have to go to shield 3 assemblies in the periphery of the core, stainless steel 4 shield assemblies. And further push the power toward the 5 center of the core, actually reduce the core size, in 6 reality.
7 MEMBER KRESS: Can you do any axial shaping of 8 the power?
9 MR. HANSON: Not really. In terms of our 10 peaking factors, we don't have a lot of latitude there.
11 The design or our reactor we have part length 12 control rods for axial shaping. However, all of our 13 analyses are done assuming those are fully withdrawn. We p.
'v) 14 think the expense of trying to do that far outweighs any i I
15 benefit we would receive from that just because of the j l
16 complexity of the analyses when you get into that.
1 17 The cost to implement this ultra-low leakage l
18 core is around $2 to $4 million per cycle in increased 19 enrichment costs because to get the same life core with 20 the smaller core volume you need to go to higher 21 enrichments.
l 22 So, there is a significant cost to that sort i
i 23 of change.
24 The next process, and I assume that you have
/O
() 25 been briefed on this and have a lot of interest in this is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
l l
165 l 1 reactor vessel annealing.
l l
gy 2 When we got the SER in 1995 saying that we
- s x~
3 couldn't operate beyond 1999, we had been in the process 4 of developing this ultra-low leakage core because we had 5 an SER of 2004 at that time.
6 But it came back to 1999 and 40 per cent of 7 three years doesn't get you very much additional time. So 8 going to the ultra-low leakage core was dropped. We felt l 9 at that point that the only real option was to go with i 10 reactor vessel annealing because that would actually make 11 the vessel better and with the amount of time left in our 12 license, it was the only of achieving it.
13 We had submitted in April, 1996, a preliminary i
('T l x- 14 thermal annealing report per the Draft Guide which is now 1
15 Reg Guide 1.162. Our intent, the reason that it was 1
16 preliminary, is that our intent was to revise that thermal 17 annealing report to incorporate what we learned at the l l l
18 Marble Hill annealing demonstration.
19 When the screening criteria was extended to 20 2003, we put the annealing project on hold and at that 1
21 time detail engineering was approximately 70 per cent
- 22 complete on the project.
23 We currently have a schedule to restart the 24 project in early 1999 should we decide to do that. That rm
(_) 25 is when we feel we need to start back into the project to NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS
' 1323 RHODE lSLAND AVE., N.W.
l (202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
1 166 1 be able to anneal prior to reaching our current screen p)
(
2 criteria date.
3 Code Case NE57 was also for reactor vessel 4 annealing and was approved by the ASME. It has not 5 received endorsement by the NRC although it has been 6 submitted for endorsement.
7 That is an issue for us and making future 8 determinations of where we are going on this issue. That i
9 is a very important component of the regulatory risk from 10 our perspective. Having an approved code case, we put a 11 lot of effort in getting that code case in place with 12 ASME.
13 However, because we are not currently pursuing
! \
\ >
14 annealing, there is no effort going on within the staff to 15 approve the code case.
16 I talked about the Marble Hill annealing 17 demonstration. We were very involved in that 18 demonstration. I was the manager of the annealing project 19 at Palisades. Either I or my staff was there for all the 20 critical evolutions and for most of the entire annealing 21 demonstration.
22 That demonstration employed the same methods
- 23 and limits in terms of mechanical limits, the stress 24 limits, on the vessel and deformation limits that are l'%
(.,,) 25 proposed for use at Palisades through our thermal interim NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE IS!.AND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 23005-3701 (202) 234-4433 l
l 167 l 1 report.
,3 2 We had very favorable results. The prediction
' L~/ I 3 of displacements on the vessel, the stresses as read by 4 the strain gauges were very, very close to what was i 1
5 predicted by the models.
6 The estimated cost to complete an anneal is 7 $25 million from today. We have already spent in the l l
l
- 8 vicinity of $78 million on ancialing and we have another $
9 25 million to go to complete an anneal.
10 We are also very involved in supplemental 1
11 materials testing. We have installed two new supplemental )
12 surveillance capsules in the vessel surveillance capsule.
l 13 We installed those in 1995 in cycle 11. l x 4
'i ) l k/ 14 Those capsules contain weld materials that are !
15 representative of each of our reactor vessel beltline
- 16 welds. Our original surveillance program did not have i
t 17 representative material. It was not the same weld wire 18 heat for our welds.
l 19 It was a similar weld, but it was made after l
20 the vessel was made and did not use the same weld wire 21 lot.
l 22 One of these capsules will be removed at the i
i l 23 end of the current cycle which is May, 1998. The second 24 will be removed in 1999. These capsules have the O)
(, 25 potential to provide plant-specific credible surveillance l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
168 1 data for per the Rule. But until we see the results, we 7s 2 can't say that they will because it is dependent upon the 4
i s '/ results falling within a variance of the prediction.
3 4 We are also participants in a Pressure Vessel 5 Research Council program to help the industry to develop 6 an alternative method to directly measure fracture 7 toughness.
8 Clearly fracture toughness is a parameter of 9 concern. We infer fracture toughness from sharpie and 10 drop weight testing results on these materials. But 11 fracture toughness is clearly the issue.
12 We expect to have unirradiated fracture 13 toughness testing results fo. our base plate and our three k' 14 beltline weld materials within a year of this date. I 15 MEMBER KRESS: What is this alternative 16 approach? Is that the magnetic?
17 MR. HANSON: It is the precracked sharpie.
18 MEMBER KRESS: The precracked sharpie?
19 MR. HANSON: That is correct. This is an 20 effort with CE owners group and Babcock and Wilcox owners 21 group and also in conjunction with the Pressure Vessel 22 Research Council.
i l 23 The other option that we did look at or 24 consider was the Reg Guide 1.154.
.~.
(_,/ 25 MEMBER KRES5: This is not a technique that i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS
! 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 2344 433 1
l
169 1 you can apply directly to the vessel is what I was --
, 2 MR. HANSON: Excuse me?
\ /
3 MEMBER KRESS: This is not a fracture 4 toughness measurement technique that you can apply 5 directly to your vessel?
6 CHAIRMAN SHACK: It is a direct measurement of 7 fracture toughness of the specimen, not the vessel.
8 MEMBER KRESS: It is a specimen measurement. I l
9 MR. HANSON: Correct.
10 CHAIRMAN SHACK: Or a more direct more measure l
11 of fracture toughness.
1 1
12 MR. HANSON: What it is, instead of having to 13 come up with a 1-T specimen, it ia a way of measuring
'v-) 14 fracture toughness without having a specimen of that size.
l 15 Since we have sharpies available, that is what we are :
l 16 trying to use to measure fracture toughness.
17 We consider Reg Guide 1.154 --
l l
18 CHAIRMAN SHACK: Your new capsules are all '
19 sharpies again? There is no even subsize fracture 20 mechanics specimens in them?
21 MR. HANSON: No, they are all sharpies.
22 CHAIRMAN SHACK: They are all sharpies, okay.
23 MR. HANSON: We had considered subsize 24 fracture toughness specimen,. Because of the Rule not r3
(,) 25 recognizing fracture touchness n easurements, we felt we MAL R. GROSS COI%T REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
170 1 would be better served by using sharpies. And of course, s 2 it is very expensive to put vessel capsules in place and
)
's/
3 there is a very limited amount of space in those capsules.
4 Our program has enough material to get 5 irradiated, anneal and reradiated data from the specimens 6 that are in those capsules. In order to get that many 7 specimens in there, we ended up actually having to go to 8 specimens that are actually a third of the size of a 9 normal sharpie with the intent that we would reconstitute 10 those sharpies after we remove them from the capsule.
11 So, we already are doing some work to get the 12 number of data points that we were trying to get out of 13 those surveillance capsules.
( )
\ / 14 Any other questions on that?
15 MR. VASSILAROS: Those sharpies are still 10 16 by 10 cross-sections, one third the length?
17 MR. HANSON: Yes, they are just shorter.
18 MR. VASSILAROS: Mike Vassilaros from 19 research.
20 MR. HANSON: And it is all spelled out in our 21 thermal interim report, the configuration in the capsule.
22 MR. HANb7N: Okay. We did consider Reg Guide 23 1.154 analysis. We didn't see that as a viable option.
24 We locked real hard at the Yankee Rowe experience and we
(-
(_,) 25 came to the conclusion that the uncertainty of that i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234J433 WASHINGTON, D C. 20005-3701 (202) 234-4433
171 1 process was such that we felt at the time that it was I e3 2 better to spend our money to improve the vessel rather LY 3 than fighting over what the basis was for the Rule. 1 4 Of course, early plant shut down, that is an 5 option and that of course, will be driven by economic 6 issues. With electric utility restructuring, 7 deregulation, whether or not we continue to operate beyond 8 2003 is very dependent on our ability to compete in the i
9 deregulated environment.
10 This is a strategic issue for us. It has the 11 potential to have a very significant financial impact on 12 the company. We understand what the right answer is here.
13 Mr. Kress, you look like you are ready to ask i
~
14 a question.
15 MEMBER KRESS: Yes, as you might imagine, your 16 comment on 2.5 is very interesting to us. Is there any 17 significance currently in there, or will you reconsider at 18 some later time, possibly making a new Reg Guide analysis?
19 MR. HANSON: If the Reg Guide was changed to l
I 20 provide more specific direction in how to deal with it, 1
l 21 yes, we certainly would. I 1
22 The time that we really made this l l
l 23 determination was when we only had three years until the I 24 end of our screening criteria. We felt that there was no
[3 (x,) 25 way that we were going to get from here to there in that 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 i
172 1 amount of time.
7s 2 Now that we have more time I would consider
/ \
\ '~' /
3 looking at that. But there would have to be some 4 considerable effort coming from the staff's perspective to 5 provide better guidance to us as to what we would be 6 doing.
7 Going into it blind which we really feel is 8 the situation now isn't worth our while. There are other 9 issues that I think are more meaningful, other efforts.
10 Certainly fluence reduction, the flux reduction at the 11 walls is a much better option considering what you spend 12 doing all the analyses for Reg Guide 1.154 analyses.
13 MEMBER KRESS: That is what I find very (3
k- 14 interesting about that comment.
15 MR. HANSON: Okay.
16 MEMBER KRESS: It costs more to do this 17 analysis than it does to do a very expensive flux l l
18 reduction, l
19 MR. HANSON: Well, the analysis might take $5 l 20 million and annealing might take $25 million. But if you 21 have a 10 per cent probability of being successful with 22 the $5 million, you would much rather go with the 99 per 23 cent probability of reactor vessel annealing, improving 24 your vessel and having a vessel that is usable when you
,r y
(_) 25 are done.
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 173 f 1 MR. HANSON: Our preferred option is further
- 2 approval of our annealing submittal. We think that is the
)
l 3 right answer. Fluence reduction costs of option 2.2 will 4 start accumulating in September of this year. We will 5 have to commit to a fuel design for our next cycle thac 6 goes into the middle of next year.
7 MEMBER KRESS: Are you developing counter 8 arguments to the NRC's problems witn allowing parts of 9 that submittal. Are you developing counter arguments to 10 that?
11 MR. HANSON: Yes. I think we do have counter 12 arguments. I think presented those. I am going to go 13 through some of those here today. I was limited to 20 kl 14 minutes for presentation so I didn't have the time to go 15 into the detail that Lambros did.
16 1 do have some expertise here that will help 17 me get into some more details if you would like to. And 18 since we are ahead of schedule, we should have some more l 19 time.
20 MEMBER KRESS: Well, it is up to the Chairman.
21 CHAIRMAN SHACK: We do have extra time, i
22 MEMBER KRESS. I certainly think that would be 23 of interest.
l 24 MR. HANSON- Okay. Thank you.
'( m'~ ,) 25 CHAIRMAN SHACK: But you have already been NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
174 1 through the discursion with the staff and they have
,e -3 2 rejected your additional arguments? Or is this still in
( ') 3 process?
4 MR. HANSON: The first time that I had heard l
5 that they had rejected them was today in one of the last l 6 overheads that Lambros put up there. Up to that time, if 7 you look at our last SER it says specifically that it is )
8 an interim SER and they didn't have time to fully review 9 the best estimate issue that I will go into in more detail 10 here.
11 So, we are still working and they are still l l
12 reviewing our submittal. We are still working with the 13 staff; we met with the staff earlier this year to discuss N- 14 this issue.
15 We are still submitting answers to requests 16 for additional information on their part. So, we are 17 still under the assumption that they are still considering 18 our submittal.
19 MR. LOIS: Mr. Chairman? This is Lambros 20 Lois.
21 The staff evaluation states that we are 22 working on the FERRET portion, namely the spectral 23 adjustments. But that evaluation which should be 24 concluded -- we didn't have the time to issue it. But n
( ,) 25 essentially, it is a final evaluation that we have.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 2344433
175 1 MR. HANSON: Okay.
's 2 Again, I would like to get into some details
\ /
3 of our fluence submittal.
4 First off, why were we doing that? We were !
1 1
5 going down the path of reactor vessel annealing. You ask 6 why? Lambros already made the comment that it came as a ]
7 surprise to them when we resubmitted our annealing l
8 submittal.
9 The reason we did that, well there are a 10 number of reasons that I would like to go over here.
11 When we only had three years to prepare to 12 anneal, that was a real undertaking to try to get through 13 the regulatory issues and the technical issues associated
\d 14 with it and be prepared to do that prior to reaching the 15 end of the screening criteria.
16 In our 1992 PTS submittal, the technical 17 evaluation of the NRC's review of our fluence analysis, 18 they stated that there was a 10 per cent conservatism in 19 our calculations.
20 Lambros mentioned that they had already done a 21 transport calculation that mirrored our calculation back 22 in 1992 and that technical evaluation actually said that 23 from their calculations they thought there was a 10 per 24 cent conservatism in our calculations.
(-
() 25 We understood that and we knew that there was l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
176 1 some potential that we were not going to be able to anneal l
,cx, 2 before 1999. We felt that we would go forward with (v) 3 preparing to anneal, but that we should also be pursing i
4 fluence to gain just a little bit more time if we ran into 5 regulatory or technical issues that needed more time to be 6 sorted out.
7 So, that is why we chose to continue to look 8 at fluence.
9 A couple of other things happened. one was 10 that the new cross-sections that were developed by the 11 national labs, ENDF/B VI, became available in a commercial 12 version in 1993. That became available to us after that 13 1992 PTS submittal.
I \
'd' 14 So, that was something new that was available 15 to us. In addition to that, we had more ex-vessel and in-16 vessel dosimetry data.
17 We have as much dosimetry data as anyone in 18 the industry. We did get a whole other set of ex-vessel 19 dosimetry and one more in-vessel capsule following that 20 1992 submittal.
21 In addition to that, we determined early on, 22 going back to this curve, you will see that several of 23 these show cycles three through seven, combinations of 1
l 24 cycles.
A C) 25 Early on, we bunched a number of our cycles NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
177 1 together and used bounding inputs. As we went back and
,3 2 looked at the work that people had done, we said that it
( 'I 3 looks like there are some significant conservatism that 4 have been put in here by the fact that we had used those 5 conservative inputs that covered a whole set of cycles 6 that didn't have the same inputs.
7 Two of the cycles were operated at different 8 power levels, different inlet temperatures; inlet 9 temperature was the biggest change that occurred.
10 Those issues were what drove us to re-do our 11 fluence calculation. When we redid car fluence 12 calculation, we understood the impcrtance of this issue.
13 We went to our prime contractor who does most of the kJ 14 fluence calculations for US reactors, Westinghouse and had 15 them redo each cycle using cycle-specific inputs and the i 16 new cross-sections and got their results. l l
17 In addition to that, we went to AEA in the 1
18 United Kingdom and had them perform the same calculations, 19 only using an entirely different methodology, the Monte 20 Carlo methodology. As Carolyn mentioned, that methodology 1
21 is currently not accepted by the staff, but will be 22 considered to be included in the Reg Guide in the fluence 23 Reg Guide.
24 So, we redid our fluence calculations with rN 25 both the direct ordinate transport methodology currently is )
- NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 9 02) 234-4433 l
l
178 1 accepted by Westinghouse and we did it with AEA in the
,r - 2 United Kingdom using the Monte Carlo method.
i !
N_ /
The result that they got were extraordinarily 3
4 consistent. Lambros mad the comment that no other country 5 biases their calculations based on spectral measured data.
6 The United Kingdom does.
7 They don't use FERRET, they use a different 8 least squares methodology called SENSEC. But it is the 9 same sort of methcdology. It is a least squares 10 methodology.
11 They did use that for our analysis and came up 12 with the same results as Westinghouse did using the FERRET l 13 methodology.
\- I 14 MR. LOIS: Mr. Chairman? This is Lambros i
15 Lois. I 1
1 l
16 With respect to the Monte Carlo methodologies, 17 we did not say that we would not accept it. We said that l
18 at that time, about two years ago, we didn't have the 19 capability of review them. We would have to go into 20 different contractual arrangements, probably Los Alamos, 21 to have them reviewed
, 22 Because we didn't do this one at any other l
23 time in the past.
24 As far as the British data are concerned,
,n.
(, 25 their geometries are entirely different and also their l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
179 1 uncertainty limits that they have are much higher. I am 2 not sure whether we would approve of their methods used in 73 E
/
3 England.
4 Thank you.
5 MR. HANSON: Okay. As I will discuss they did 6 a detailed review of our transport calculations, agreed 7 with our transport calculations that they agreed with our 8 transport calculations that they supported an eight per 9 cent reduction in our fluence.
10 The associated uncertainty with that, as 11 agreed with BNL was 15 per cent which agreed with our 12 calculations.
13 The best-estimate methodology which is what we
-,\ 1 EJ 14 have been talking about is a -- and I would just like to 15 talk about that in general.
16 This a picture of our reactor vessel looking ;
I 17 from the top down. With the direct ordinate transport 18 methodology you have a deferred source here. Then you 19 model the geometry of the vessel, including the core 20 shroud, your core barrel, for other plants, the core 21 shield and the water in between.
22 You are able to calculation fluences at every 1
- 23 azimuthal position.
24 Now, what we have is surveillance capsules at
((3,) 25 these positions on the ID of the reactor vessel wall. Now l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
180 1 that is a little different from most of the other plants.
f~3 2 The plants that Lambros gave the data for that
! )
3 had the 1.13 and the 1.03, all those plants have their 4 surveillance capsules on the OD of their thermal shield.
5 Really a totally different position within the vessel and 6 a different position in terms of build-up factors and 7 spectral adjustments.
8 It is a very different whether you are on the 9 front side or the back side of the material where the i 10 neutrons are going through.
11 The majority of the data, most of the plants 12 in this country are Westinghouse plants, that is the l l
1 13 configuration for Westinghouse surveillance capsules. ;
{,)
\/ 14 We are a CE plant. Our surveillance capsules l l
15 our in-vessel capsules are at the wall; there is very i
l 16 little spectral difference between our surveillance 17 capsules and the ID of the vessel wall because of the 18 positioning of our capsules.
19 We also had ex-vessel dosimetry, the symmetry 20 in the cavity that Lambros referred to.
21 The best-estimate methodology is such that you 22 do have to use calculations and the measurements of the 23 different dosimeters that are in there. Those six 24 different dosimeters measure different specific spectrums.
.,o
( ,) 25 The issue is that you are able to come up with NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLANJ AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005 3701 (202) 234 4433
1 181 j 1 your best estimate of your fluence using the least squares r3 2 methodology that looks at both calculated and measured and
\
~~
3 comes up with the best estimate.
4 That is what we are proposing in our 5 submittal. That is what the Draft Reg Guide calls for and l
6 it is also the basis for the inputs for Reg Guide 1.99 to 7 predict what your reference transition temperature shifts 8 are.
9 Those are best-estimate fluences.
10 So, the methodology here is you develop your 11 direct ordinate transport model to give you your flux l l
12 disposition in the azimuthal positions, then you bias that 13 to a best-estimate fluence where you have measurements I',_,i/
\~' 14 because you only have measurements at specific azimuthal 15 positions. j 1
16 That is the issue and that is what we are 17 proposing to be done with our data.
18 The results of our best-estimate calculations 19 indicate that transport calculations should be reduced by 20 approximately 17 per cent. Which is about 1 sigma on the 21 calculation direct ordinate transport. It is a little bit l
22 over. One signa is 15 percent; this is just slightly over 23 17 per cent of the direct transport calculations for 24 fluence.
l3 The use of best-estimate is consistent with
(_) 25 l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 2344433
182 1 the Rule. Most if not all the data points used in
,t ,s 2 developing the Reg Guide 1.99 curves, used a best-estimate i
~
3 fluence, not a direct ordinate transport fluence. That is 4 a difference.
5 I am not sure what the French and Germans do, 6 but if the inputs into your curve are calculated, then it 7 is appropriate to use calculated.
8 If the inputs to your curve are best-estimate, 9 then it is appropriate to use best-estimates when you go 10 back into that curve.
11 MEMBER KRESS: Have you identified the 12 physical reason for the 17 per cent bias?
13 MR. HANSON: We have --
{]
' 14 MEMBER KRESS: Or have you just observed that 15 it is there?
16 MR. HANSON: We have observed that it is 17 there. We do have some significant differences. Some of 18 them I have talked to. This is back in one of your 19 attachments. I anticipated that might be one of your 20 questions.
21 These are the differences, I have covered some 22 of them. You know, when you are saying that we are l 23 different, we are, but we are not that different from 24 other plants and I will show you a curve that shows all (3
( ,) 25 the data point as opposed to the two that happen to be on NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
! (202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
l l 183 1 Lambros' desk.
l ~s 2 Unlike other reactor vessela, Palisades has no t s l \ .)
~
One of 3 thermal shield. That is a significant difference.
4 the big problems that was found in the ENDF/B IV cross-5 sections was that our cross-section was found to be l 6 considerably different.
7 How much closer are we now? I think we are 8 obviously closer but there is still some potential 9 differences there. Our in-vessel capsules are the inside 10 diameter of our vessel instead of the outside diameter, 11 like Westinghouse plants.
12 Palisades has a demonstrated two per cent bias l 13 in feedwater volume.
I
- (
/~h i
\/ 14 When I talked about that inferred source term, 15 when you take that source term you make the assumption i
l 16 that the plant is a full thermal power.
17 We determine our thermal power by use of 1 18 feedwater flow measurements. We brought in some 1
19 ultrasonic flow monitoring to look at our feedwater flow l 20 and we found that we had a 2.3 per cent bias in our 21 feedwater flow. That our feedwater flow was actually 2.3 l
22 per cent than what our feedwater flow venturies were l
23 telling us it was.
24 So, that accounts for two per cent of that (n) 25 fluence difference because we are actually operative at a
! NEAL R. GROSS
. COURT REPORTERS AND TRANSCRIBERS
! 1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
184 l
1 2.3 per cent lower power level than the direct ordinate l
fs 2 transport calculations assume.
( \
~
3 We also have a considerable difference in our 4 core design from a Westinghouse plant. We have cruciform l
l l
5 control blades versus CEAs or RCCAs. That is a 6 significant difference.
l 7 We have narrow and wide water gaps in our core 8 and we have a very little leakage core design, as I 9 already showed you.
10 And our T-inlet is lower than most PWRs. We 11 operate a T-inlet in the 533 degree range versus the 555 12 degree range that most of the other plants in the country 13 do.
p
\- 14 So , there is some significant differences in 15 Palisades that can explain why we are dif fere'it .
16 Now, can we tell you exactly what per centage 17 of that 17 per cent is due to each of these? No, we 18 can't.
19 MEMBER KRESS: Have you attempted a 20 sensitivity analysis of the code calculation?
21 MR. HANSON: Stan, can you help me out on that
, 22 one?
1
! 23 MR. ANDERSON: As part of the --
24 MR. HANSON: This is Stan Anderson. You can k ,) 25 just speak into the one of the microphones there.
! NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
I (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
185 1 MR. ANDERSON: As part of the qualification on
, f3 2 the calculation of methodology, one has to evaluate what l N. )
the potential uncertainty in that calculation is. That is 3
1 4 where that 15 per cent came from.
5 A part of that is a sensitivity study 6 involving such things as what is the effect of potential l
7 variations in water temperature, the potential effect of 8 uncertainty in the source, what is the potential effect of 9 an as-built versus a design dimension in a thermal shield l 10 where you are actua31y building to tolerances.
11 MEMBER KRESS: But it doesn't seem to address 1
12 these items here.
13 MR. ANDERSON: Some of them would be. For
(--
i i
- 14 instance, the temperature effect would be covered by some 15 of the sensitivity studies. The actual differences in the 16 core design itself, probably not.
17 MEMBER SEALE: Well, have you looked at the j l
case where you had the capsules on the core side of the l 18 19 water gap as opposed to the vessel internal diameter side 1
20 of the water gap? l I
21 MR. ANDERSON: Most of our data is that case.
22 MEMBER SEALE: I mean in your calculation.
23 Have you looked how much the calculation would be skewed 24 by that?
,r q,) 25 MR. ANDERSON: By the location?
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
186 l 1 MEMBER SEALE: Yes.
! y3 2 MR. ANDERSON: We have done sensitivity
! t
! \_/
3 studies on the capsule geometry and placement, yes.
l 4 MEMBER KRESS: See, what we are looking for is 5 something that we could put our finger on and say the
- 6 reason that you have this bias is this and we can evaluate 7 it and say yes, this is a good reason.
8 These things, of course, would make a 9 difference, but we don't know how much or which it is and 10 I don't think the staff would know either.
11 MR. HANSON: That is something that we will 12 consider.
13 MR. ANDERSON: Can I just make a philosophical
(,,Nf
\# 14 comment? When we do the calculation, and I think we are 15 agreed, that calculation, because of the uncertainties in 16 the input of that calculation, has an uncertainty at the 17 one sigma level of about plus/minus 15 per cent.
18 What we find looking at a total body of data, 19 is that some of that uncertainty is due biases and there 20 are two categories there.
1 21 There are methodology-related biases that 22 would apply across the board. That would cover something 23 like your transport cross-section. ,
1 24 There are also plant-specific such as what is
/ )
( ,/ 25 the real thickness of your thermal shield; are you at the NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
187 1 upper end of the tolerance, the nominal or the bottom?
,s 2 In addition, there is random variation because t i
~
3 there are uncertainties in the way the plant was operated.
4 Remember now, we are doing one calculation to cover an 5 entire cycle, so there is a lot of averaging that goes on 6 in the calculational input.
7 There are uncertaint'ies in your ability to 8 make dosimetry measurement anc there are uncertainties in 9 the cross-sections of the dosimeters.
10 All of these contribute to a statistical 11 variation of what is going on. We believe the 12 measurements in a plant-specific environment react to the 13 things that are going on in that plant. They see the
! i
\> 14 changes over the cycle, the differences in geometry and 15 they are tell us something.
16 Now, can we go back and specifically say that 17 that change is due to this that or another thing? That is 18 very difficult and perhaps in possible. But we do believe l 19 that they are telling us something about the relationship 20 of that plant and its nominal calculation.
21 MR. LOIS: Mr. Chairman, this is Lambros Lois. ;
O 22 Dr. Kress asked a crucial question, really, we j i
23 understand what is the physical basis for those i 24 differences. The article we made previously on the
(~' !
(_,T/ 25 blackboard, there are differences in the Palisades plant NEAL R. GROSS COURT REPORTERS AND TRANSCR;BERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, O C. 20005-3701 (202) 234-4433
188 1 and the Westinghouse background or basis plants. We don't 2 see that this is a justified observation, given that f-ss f I
~
3 others have consistent measurements to calculate the 4 results between BWRs and the variety of sizes and 5 structures of, and designs as you'd say, of PWRs, such as 6 the Zimas KWU data.
7 The question is not really whether the 8 calculational procedure has followed the mechanism of the 9 proposed Regulatory Guide, the question is who understands 10 what is the physics behind it?
11 This is a precsure vessel and these are l l
12 inconsistencies. As they are in plant specific data, l l
13 inconsistencies as they have been manifested in the ;
l l'i-Jh 14 methodology, difficult for the staff to find otherwise.
15 So far we've been asking the same question 16 that Dr. Kress asks and we have not seen yet a 17 satisfactory response.
18 This is not really whether or not this is a 19 best estimate. Yes, it is. The mechanistic procedure it 20 followed is not adequate. We don't feel that this is the 21 case. Thank you.
22 MR. HANSON: Okay, let me -- I did want to go 23 over this. This is a curve of our Reg. Guide 1.99 curve 24 as shows the expected reference transition temperature for
(>
s ) 25 a certain amount of influence for our -- th'is is for our NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 2344433
189 1 base plate.
p- 2 When you look at this, this curve was based on Q' 3 best estimate fluence inputs, not direct ordinate 4 transport calculations. When this data was used and input 5 to the belt, that curve for the Reg. Guide 1.99, the best 6 estimate fluence was used to input.
7 When we used our previously accepted fluence 8 numbers and we plotted out our data points, those data 9 points were the diamonds.
10 When we readjusted our fluence using 11 methodology we've just done with the ENDF/B-VI, and 12 advising to the best estimate values, this is where our 13 data points moved. This data pointed probably would be N- 14 better represented to be about 4.4 and move to there.
15 There were some other issues with why this one was out 16 where it was.
17 This moved there and the early data point did 18 actually move a little bit further away from the curve, 19 but the uncertainties here, when you have very low -- that 20 was damaged because of the inaccuracies of the sharpie 21 testing. We had a lot of variance at your low fluence 22 levels that can't always be explained.
23 When you look at this, you'll see that our 24 data points did move far closer to what was predicted by ;
<^ l
( ,w) 25 the Reg. Guide, by using our current fluence calculations l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 j
190 l
l 1 of the submittal and the basis for the rule, and the basis f3 2 for the Reg. Guide is best estimate fluence, not
\
l 1 l Y-)
l 3 calculated fluence. If you built your regulations on 4 calculated fluence, then it would clearly be appropriate l l
i 5 to only use calculated fluences in inputting to the term l
6 what you're shifting reference transition temperatures.
7 With using a best estimate fluence, and these 8 best estimate fluence data points were used, were created
)
9 using a FERRET / SAND 2 code, a least squares methodology ;
I 10 similar to the current methodology we're using.
l 11 So yes, mechanistically, it would appear that 12 this is the right way to do the -- the right way to 13 determine what your, what the condition your vessel is.
/, \
'! J
- / 14 The issue still remains is that we are somewhat of an 15 outlier, but not a significant outlier.
i 16 When you look at the overall data, this is 17 data not for all plants, obviously, because not all plants 18 have redone their fluence calculations using the ENDF/B-19 VI. Very few have, actually, and this is a great majority 20 of the ones that have. These are -- this is a data base 21 of 21 plants that Westinghouse performs fluence 22 calculations for that have redone their fluence analyses l
l 23 using the ENDF/B-VI and the FERRET /SS methodology.
24 Of these plants, this is the distribution of
,q s ,
) 25 their -- what we call versus a measured over calculated.
i l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
191 l
1 It's really a best estimate over calculated, because there
,s 2 is an element of the calculation that goes into that best i )
3 estimate.
4 So this is a best estimate over calculated !
5 ratio and you'll see that the actual average is not 1.0, 1 1
6 but .95 and you'll see that there is not -- each one of l
7 these data points represents a plant, one of the 21 8 plants.
9 Now Palisades is this plant, so we are on the 10 low side, but this is the one sigma level, so we're just 11 outside of one sigma and you'll see that there are four 12 other plants that are very close to us, not the 1.13 that e
13 Lambros was talking about and the 1.03 that Lambros was l r ~~x !
I i k/ 14 talking about.
15 MEMBER SEALE: Are any of those plants 16 Combustion plants?
17 MR. HANSON: No. These are all Westinghouse 18 plants, except for -- that's my understanding.
19 MR. ANDERSON: That's correct.
20 MR. HANSON: Thank you.
21 MEMBER SEALE: So none of these plants would 22 have their --
23 MR. HANSON: All these plants have thermal 24 shields and all of them have their surveillance capsules
,,m I ,) 25 on the OD of the thermal shield and nearly all of them NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
1 192' 1 operate at 555 degrees.
,y
, 2 MR. ANDERSON: Not all.
3 MR. HANSON: Not all, but the vast majority of 4 them do. So this is the overall distribution of those 21 5 plants that have been redone by the ENDF/B-VI by 6 Westinghouse using the ENDF/B-VI and like I say as you can l
7 see we do have the lowest BE/C, but it's only Reg. .83.
8 The average in .92, .95.
9 So we are .12 different, but the one sigma is 10 very similar, very close to that .8.
11 MR. LOIS: Mr. Chairman, this is Lambros Lois, 12 again. I was not aware of those data, however, I may 13 assure you that none of these plants that they are close
.\ ') 14 to Palisades came in for a regulatory decision and they 15 come in, probably we will not get a favorable response.
16 The plants that came in with difference 17 between calculated and measured values, those are the 1.13 18 and the 1.03 were approved on the basis of the calculated 19 value, the accepted zalue was conservative, rather than 20 that the measured value was taken into account in this 21 case.
22 Thank you.
23 MR. HANSON: However, there were several 24 plants that did end up with -- when they redid their gw
(,,) 25 ENDF/B-VI calculations, their fluence calculations went NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
1 l
! 193 l
l 1 up, but if you did take into consideration the B/C l
7- 2 correction, it brought them back down to what they had L)) 3 been originally accepted on the docket with NCR and for 1
4 those plants their reannounce was accepted without an SER 5 being issued, so in effect, it was accepted that those 6 fluence calculations were biased down by the B/C factor.
7 MEMBER KRESS: The database goes into the 8 embrittlement change in the RDF and the upper shield 9 energy. Does it use calculated fluences?
10 MR. HANSON: No, it uses best estimate 1
11 fluences. 2 l
1 12 MEMBER KRESS: It's best estimate. And l 13 they're based on the old ENDF/B or the new one?
14 MR. HANSON: They are based on the old ENDF/B 15 cross sections, however, we did do a sensitivity analysis 16 on the changes. We went back and we did a very detailed 17 analysis of why did our fluence change and what we saw was 18 that there was a -- that the ENDF/B-VI cross sections 19 caused a 1 or 2 percent increased influence in one area 20 and about a 1 percent decrease in another area and when I 21 say 1 to 2, that's because of ineffective different 22 azimuthal positions differently, but, so the net change 23 was really very close to zero from just the ENDF/B-VI 24 cross section change.
f (q) 25 Some of the things that the ENDF/B-VI cross NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHlWGTON, D.C. 20005-3701 (202) 234-4433 l
194 1 section change did for us is -- what it did for us is we
,-~ 2 had a large bias between -- we had the .83B/C for our in
.Q/
3 vessel capsu;.es, but our ex vessel capsules were like .75 4 or I'm sorry, Ross, can you give me the right numbers.
5 MR. SNUGGERUD: I'm sorry, I don't remember 6 the right numbers, but it was closer to like 1.9. My name 7 is Ross Snuggerud. I believe that you look back at our ,
8 1992 and 1993 submittal, the in vessel capsules had biases 9 around the .90 and the ex vessels were closer to 1.
10 MR. HANSON: Now, some of the issues that we 11 had there --
J l
12 MR. ANDERSON: I'd like to differentiate 13 something on this change from ENDF IV to ENDF VI. It was l
(\
t /
l
\- 14 the transport calculations, the cross sections that are )
l i
15 used in the calculations that were changed. The dosimetry 1
l 16 cross sections changed very, very little. If you go back 17 to previous comparisons, typically you would find 18 calculations were low measurements by about 15 percent.
19 They were biased upward. It was recagnized the 20 calculations were low.
21 There's now a change that increases the 22 calculations approximately 20 percent. The measurements 23 from ENDF IV to ENDF VI virtually have remained the same.
24 Hence, we're now seeing a calculation that's much closer
(~\
(_,1 25 to measurement but as the data base showed, you're seeing NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
195 1 like a 5 percent difference now as opposed to a 15 percent es 2 difference with ENDF IV.
[
s h
s 3 So it was a calculation that evolved, not 4 interpretation of the measurement.
5 MR. HANSON: And one of the things that also 6 did for us was getting to earlier is that when we got 7 done, we did do a number of vessel wall measurements, a 8 large number of vessel wall measurements in preparation 9 for the reactor vessel annealing. We felt we needed that 10 for our stress calculations.
11 When we factored that in to the direct 12 ordinate transport calculations, and we used the ENDF/B-VI 13 cross sections, we then came up with a consistent BE/C for
- i
\/ 14 the ex vessel and the in vessel. Before that, there was a 15 big -- there was a very large skew. So the difference in 16 the vessel wall and the difference in the ENDF/B-VI helped 17 give us a more consistent picture of what's between those 18 two capsules.
19 One of the other issues that Lambros brought 20 up and he had the box with the four squares and greater 21 than 4 MeV and less than 4 Mev, he said that well the 22 lesson for 4 MeV fluence had .86 BE/C for the in vessel 23 capsules and the greater than 4 MeV had 1.0, well, there's 24 only about 5 percent of the fluence that's above 4 MeV.
( ,/ 25 And what the fear of program, this best NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
196 1 estimate -- I'm sorry, the least squares adjustment
,-s 2 procedure does is bias your answer so that the dosimeters i 1 O 3 that measure the greatest amount of your neutrons are the 4 ones that you put the most confidence in as determining 5 what your actual fluence is. You don't put the same 6 weight on dosimeter result that only measures 5 percent of 7 your neutrons versus one that measures 60 percent of your 8 neutrons.
9 That's part of what this least square s 10 adjustment procedure does for him. So showing you that l
l 11 greater than 4 MeV has a BE/C of 1 versus the other ones 12 that have .86, that's 5 percent of the data that you're 13 dealing with, versus the other 95 percent, so it doesn't
( )
\/ 14 have a lot of meaning to look at it from that perspective.
15 MEMBER KRESS: Should I assume equal weight to 16 the damage that is done?
17 A neutron does damage whether it'c 4 MeV or 18 less 4 Mev or is that a nonlinear function?
19 MR. HANSON: I can't really tell you.
20 MEMBER KRESS: It seems to me like the 21 weighting factor ought to be on damage and not the number 22 of neutrons.
23 MR. ANDERSON: Ultimately, if you look at what 24 the least squares procedure does, it takes your calculated (n,! 25 value along with the uncertainties associated with it.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
197 1 The cross sections that relate to calculation to the 7s 2 dosimeter measurement and their uncertainties and your raw
('~' )
3 measurement of its uncertainties and if you have 6 foil 4 measurements essentially you have 6 equations relating 5 various portions of the energy spectrum.
6 It does a best fit to that set of data and 7 comes up with a new energy distribution that best matches 8 not only the measurement, but the measurement and the 9 calculation and from that if one were to use site 10 displacements per atom as a measure of damage, it uses 11 that best fit spectrum to calculate then the overall 12 damage that occurs.
13 MEMBER KRESS: Okay, that sounds like it would t' c ,\
'w #
14 be an appropriate way to do it, yes.
15 MR. HANSON: Okay, then my last slide was that 16 the NRC evaluation of our best estimate methodology in our 17 4/4/96 submittal we were told that it had not been 18 completed due to time and resource constraints. I guess l 19 what we're hearing today is that that has been completed 20 since that time.
21 I think we've already covered some of this, 22 based on an evaluation of the data from the 21 reactors, 23 the BE/C bias is .95 with a standard deviation of 9. The 24 BE/C for Palisades is .83 which differs from the data base
(~1
(,) 25 as citing more than one signal. The plant specific bias NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
( 198 l 1 should be used when the adequate number of data points are l ,c 3 2 available and that's exactly what the Reg. Guide says for 1
('~') 3 fluence and like I said we have physical measurements of 4 fluence in our vessel within 1 or 2 of anybody in the 5 country. 1 l
6 In summary, I'd like to say that we really are 7 interested in the correct answer. I think we've shown l l
8 that by the number of things that we've done. We've gone i 1
9 through complete fluence analyses, reanalyzing our fluence l 10 by two different vendors using two different 11 methodologies.
12 We've put additional surveillance capsules in. l 13 We've put additional in vessel dosimetry capsules,
/ '4
'- 14 capsules that didn't have any sharpie specimens, but only 15 had dosimeters be put into our vessel. We put two sets of 16 ex vessel dosimetry in. We're very involved with the 17 industry's efforts on this issue. We really are looking 18 for the right answer, not just the answer that gets us to 19 2011.
20 Now we'd &ike to get to 2011 and we think 21 that's the right answer. Both 10 CFR 50.61 and a draft 22 Reg. Guide call for the use of best estimate round l 23 conservative or bounding values for pressure vessel 24 fluence.
n
!. ,) 25 The rule has margin terms built int it and l NEAL R. GROSS
( COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
I i
i l 199 1 those margin terms are there to account for the 73 2 uncertainties in the analyses. If you use -- if you then ,
N]
l 3 pile on the conservative inputs into your analyses and 4 come up with something more conservative than a best 5 estimate fluence, you're really accounting for the 6 uncertainties of the analyses twice. l l
1 7 And the default conservatism rather than the l
l 8 best estimate places an undue burden on Palisades in 9 several areas of fluence reduction and annealing or plant 10 shutdown and it creates additional risk. Fluence 11 reduction -- the basic end point of a fluence reduction is 1
12 you end up operating your core at a higher power level.
p_ 13 You're still within your design limits. You're still j
\'-) 14 within your peaking factors and within your license, but l
j l
15 you're operating in a higher power density in that core, l 16 because your core is smaller.
17 Now, we believe that's an acceptable risk, but 18 it's still going in the direction of -- it's not going in 19 the' direction you want to be going. But it is an 20 acceptable risk.
21 Reactor vessel annealing, there's some 22 technical risks associated with it, but I think we're well
! 23 on our way to dealing with all those issues and I have 24 confidence that a reactor vessel can be annealed in this
/~N kj 25 country safely and with the right output.
l NEAL R. GROSS
! COURT REPORTERS AND TRANSCRIBERE 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
200 1 However, the dose that it takes to do a
,. 2 reactor vessel anneal, we predict to be in the 200 R
! )
~ ~ '
/
3 range. That's greater than the median annual dose for an 4 operating reactor in this country.
5 If you don't need to anneal, if your reactor 6 vessel has adequate fracture toughness to be operated to 7 the end of the license, if you don't need to anneal, 8 annealing is now ALARA, in no shape or form. If you don't 9 need to do it, you shouldn't do it because there are other 10 issues here that you have to deal with when you go to do 11 an anneal that aren't very appealing.
12 Plant shutdown, certainly that reduces the 13 nuclear risks to the public, but that power will be i )
' - 14 replaced by an plant that will operate far less 15 environmentally benignly than Palisades does. Our impact 16 on the environment is far smaller than the alternatives in 17 our region that I'm aware of.
18 So I believe that there's additional risks to 19 default to a conservative answer in this issue and those 2C are some of the risks that are there, based on the options 21 that we see available to us to continuing operating our 22 license.
! 23 MEMBER KRESS: Let me ask you a question about 24 options. It seems to me like they're very limited. You
, 25 either reduce the fluence, you recalculate using the NUREG NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
201 1 Guide, you shutdown or you thermally anneal.
,- 3 2 MR. HANSON: Correct.
I \
~'
3 MEMBER KRESS: That's a limiting number of 4 options. Isn't there some other option that says, 5 recognizes that the PTS rule has conservatisms and it is 6 just a screen, these screen values for when you reach the 7 criteria or screens and raises a flag. Shouldn't there be 8 some other option that says okay, we know that we've 9 hoisted this screening flag and that tells us we ought to 10 be more vigilant in how we operate our plant. There could 11 be administrative things we do differently. Should there 12 be another option or is there possibly another option that 13 could be considered?
/m
\
5 /' 14 MR. HANSON: I think what you're talking about 15 is the Reg. Guide 1.154 analysis and that's clearly what 16 the intent of the Reg. Guide 1.154 analysis was. And 17 you're right, that would be a very attractive option if it 18 was clearly defined what you needed to de to be successful 19 going that route. That's really the intent of the Reg.
20 Guide 1.154 is exactly what you just stated.
21 MEMBER KRESS: Yes, I guess we haven't 22 reviewed that Reg. Guide yet, have we?
23 CHAIRMAN SHACK: That's the old PTS rule.
24 MEMBER KRESS: That's the old rule. That's a pq
( ,) 25 reanalysis.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4 433 WASHINGTON, D C. 20005-3701 (202) 234-4433
202 1 CHAIRMAN SHACK: That's just a reanalysis.
g_s 2 MEMBER KRESS: And I was thinking something
- ~.
\") 3 different than that.
4 CHAIRMAN SHACK: You wanted another option.
5 MEMBER KRESS: I wanted another option.
6 MR. HANSON: What do you have on your mind 7 that --
8 MEMBER KRESS: Well, I believe that the PTS 9 rule is a screen, that it says okay, I've raised a flag 10 now that my plant is --
l' CHAIRMAN SHACK: Deserves some attention.
12 MEMBER KRESS: Deserves some more attention, 13 but that attention doesn't mean that I have to really get rm (x >) 14 below the PTS screen over my lifetime.
15 It just means I have to watch what I'm going a l l
16 little more and be sure I don't thermally shock the 17 vessel, for example, or do something administrative 1y to 18 watch for -- do a little more inspection or something that 19 is not quite such a burdensome option as the ones we look 20 like we have.
21 I mean shutting down the plant is a pretty 22 severe option, but I don't know. Maybe this is something 1
23 I ought to address to the staff instead of to you, is --
(
24 MR. MAYFIELD: Thank you, yes, we'd like to
!q Q 25 answer that. We certainly agree with I think everybody l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
l WASHINGTON, D C. 20005-3701 l (202) 234-4433 . (202) 234-4433
203 1 that's locked at the Reg. Guide, that the way it was used 2 for the Yankee Rowe evaluation, demonstrated something i i 3 that didn't work very well.
4 However, the way the Guide was used for the 5 Yankee Rowe evaluation was not the way the original 6 authors of the Guide intended that it be used. I base 7 that on discussions with the author of the Reg. Guide.
8 Rather, they had intended the Guide be used 9 very much as you're describing as a tool to look at 10 options for changes to the plant, changes to the 11 operation, is flux reduction going to get you there, well, 12 if flux reduction is really going to get you there, you 13 wouldn't have gotten into the Reg. Guide analysis, safety I i N' 14 injection.
15 There's a laundry list of potential changes 16 you could make that say well you're going to operate above 17 the screening criterion, but the probability of failing 18 the vessel is acceptably small.
19 The difficulty is that -- and I think that as 20 we've looked at the Guide, when you get into the mechanics 21 of using it, it gets to ne very cumbersome and there are 22 things that are not laid out very well and we certainly 23 would agree with Mr. Hanson and we have had several public 24 meetings on this where the staff wholeheartedly agrees
/N
( ,)
25 that Guide needs to be worked on.
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l (202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 I
l
204 1 We've told the commission we're going to go y3 2 work on it and that activity continues to this day. It's
\ )
3 not a simple problem. It gets very sticky when you get 4 off into these things.
5 The other thing we've started looking at is, 6 in fact, the PTS rule itself and is there a basis for 7 changing that? And if so, what would it look like? And 8 that's a dialogue that the staff is having internally. We 9 haven't gotten far enough with it to have a public meeting 10 with it on the subject. We haven't gotten far enough with 11 it to have a public meeting with it.
12 On the subject, we haven't gotten far enough
,_ 13 with it to come before this committee and talk about it.
( ,
'# 14 It's something that we're starting to explore. But the 15 notion of Reg. Guide 1.154 was to use as a tool to look at 16 what changes you might make.
17 Unfortunately, in the Yankee evaluation, the 18 embrittlement estimates were so high when'they started the 19 analysis, it was almost a foregone conclusion, you weren't 20 ever going to pass the test. I'm not sure where you would 21 end up with Palisades. Obviously, Jack and the Consumers 22 people who looked at it more carefully than we have, but 23 the notion of the Guide was as a tool.
24 MEMBER SEALE: That's helpful, but I would
(-
( ,)
25 make one other comment. I would argue with you on one NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
205 1 statement you made. You made the comment that you're
,7 3 2 having to pay the conservative price twice, as you went N]
3 through this discussion, that is, you're having two kinds 4 of conservatisms. But that's not paying it twice.
5 MR. HANSON: It's accounting for it twice.
6 MEMBER SEALE: That's right, but the basis for 7 establishing one level of conservatism was not on the same 8 basis as it was for establishing the other one, that is, 9 the flux or the fluence calculation is one conservative 10 estimate.
11 Then there was another one that went into the 12 original analysis and so you're really paying two kinds of 13 conservative prices, I think, if you want to look at it i
14 that way. I mean you've got a conservatism in the fact l l
l l
15 that you're not being allowed to do your best estimate l 16 assessment.
17 That's not the same one that was originally in l 18 the assessment.
l 19 In the rule, you say that the rule originally 20 said you do best estimate. You're not being allowed to do 21 that.
22 MR. HANSON: Right.
l 1
23 MEMBER SEALE: But the rule was already 24 conservative.
(7,
, () 25 MR. HANSON: Correct.
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l WASHINGTON, D C. 20005-3701 (202) 234-4433 l (202) 234-4433
206 1 MEMBER SEALE: That's what I'm saying, see.
,- 2 There was that conservatism that was in the rule and now
( )
3 you have a different conservatism which is the not being 4 allowed to do the best estimate assessment. They aren't 5 necessarily equal.
6 MR. HANSON: No, but --
7 MEMBER SEALE: They do add.
8 MR. HANSON: The conservatism in the rule took 9 into account what they expect the uncertainty of the best 10 esti nate , not the calculated, but the best estimate 11 fluence analysis would be, and that was built into that 12 margin term.
13 MEMBER SEALE: I agree. All I'm saying is
,q
~) 14 that it's a trivial point.
15 MR. HANSON: And our best estimate fluence, 16 let me throw something else up here real quick. l l
17 MEMBER SEALE: I'm not disagreeing with you. !
l 18 I hope you understand that.
19 MR. HANSON: Okay. Let me throw something 20 else up here anyhow. l 21 MEMBER SEALE: That's all right, fine. You've 22 got the time.
I 23 MR. HANSON: Lambros made the comment that we 24 didn't have a clear bias. I couldn't quite follow what (O) 25 your logic was when you said that our, the B/Cs of our l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
207 1 surveillance capsules were not consistent and didn't show
,- 2 clear bias.
( )
~
3 MR. LOIS: Jack, at that point I was 4 commenting on the consistency or the appropriateness of 5 the methodologies. The methodology does not show a bias 6 in the sense that it is a systematic error. In other 7 words, there are plants that clearly are above and there 8 are plants that are clearly below.
9 So in that sense, there is no bias, even 10 methodology. That is what I was commenting on. And then 11 I'll come back with comments in the best estimate.
l 12 CHAIRMAN SHACK: That just sort of addresses a 13 question that I was going to ask. You made the comment j
+
/3
\ 'k 14 that you weren't going to -- you didn't like the notion of l l
15 a correction based on a single plant. ;
l 16 MR. LOIS: That's right.
17 CHAIRMAN SHACK: NN it's a largel data base 18 where you averags for M/C over a much larger data base.
19 MR. LOIS: That's right.
20 CHAIRMAN SHACK: But you're still not l
! 21 comfortable with the notion of using that number for an 22 adjustment because some of them are above and some of --
l 23 even though the average seems to show a bias.
24 MR. HANSON: But clearly, the bias is plant
[ h
(_,/ 25 specific.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASH lNGTON, D C. 20005-3701 (202) 234-4433 l
i
208 1 MEMBER SEALE: That's -- yes, that's weird, 1
l 2 actually. I mean that says that plants are specifically
, i
),
V 3 different from each other.
4 MR. LOIS: But it has been shown that they're 5 really not. Others have been able to obtain consistent 6 dosimetry data across a variety of all kinds of plants and 7 there are also other vendors who do not assign or use 8 plant specific values for the evaluation fluence.
9 MR. HANSON: But you agree that isn't what was 10 used to develop the rule, that the rule did use best 11 estimate ;
12 MR. LOIS: Let me -- I have a limited --
l l
13 MR. HANSON: With the FERRET / SAND 2 least l \ ! ,
\'
14 squares adjustment. l l
15 MR. LOIS: I have a little bit perspective in 16 the whole thing. I am addressing it from the perspective 17 of 10 CFR 50.61. I am not allowed to reach into the 18 conservatism of the rule if there is any built into it and 19 say well, I'm going to decide this way, the other way 20 because of the conservatism of the rule. I have to obey 21 by the letter of 10 CFR 50.61. Now this is not the issue 22 of best estimate.
23 This is an issue of consistency of the data.
24 I, as I stated previously, you have followed the mechanism Q' 25 of the draft regulatory guide as to estimating of using
! NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS i
1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C 20005-3701 (202) 234-4433
209 1 the best estimate. That's not the issue. The issue is is 2 this a best estimate or not and my view from the plant (n)
\J 3 specific measurements, as well as the performance of the 4 methodology, this is not a best estimate. In that 5 context, not that I want to go into the conservatisms 6 built into the rule. That's somebody else's business and 7 I cannot, I am not allowed to go into that and as a matter 8 of fact, we would not be approved by anybody from my 9 management.
10 Thank you.
11 MR. HANSON: But you agree that that's a best 12 estimate that was used in development of the rule and 13 we're using the same best estimate that the rule is based
( \
\d 14 on.
i 15 MR. LOIS: No, the rule assumes a consistency )
16 of data. If the data did not point in the same direction, ;
17 that's not best estimate.
l 18 MR. ANDERSON: Could I pose a question for 19 thought? l 20 MR. HANSON: Yes, Sam.
l l 21 MR. ANDERSON: Relative to consistency of l
l 22 data. If -- let's hypothesize that we believe the l
23 calculation produces a best estimate and we believe that 24 the uncertainty is associated with that as plus or minus in (s) t 25 15 percent and I've heard no one in this industry say it's
- NEAL R. GROSS
! COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 i
210 1 any better than that. And if we could make a very em 2 accurate measurement, would we not expect those
( ) l 3 measurements to fall within a plus or minus say 30 percent 4 of that calculation?
l 5 I think what we're observing is statistically )
6 consistent with the uncertainty that we assigned to these 7 calculations. And in fact, they show tighter bands than i 8 plus or minus 15 percent. If you look at the entire data l l
9 base, it was a plus or minus 9 percent. The Palisades is l
10 plus or minus 7 percent.
l 11 MR. HANSON: Yes, there was some assumption 12 here that this was the same graph I put up and we jus:
13 like it, earlier, it's not. l n
i \
\J 14 CHAIRMAN SHACK: It's a different one.
15 MR. HANSON: So I do want to cover this. This 16 is the B/C for each of our individual dosimeters, both in 17 vessel and ex vessel. You'll see that while the variance 18 in B/C for the plants has a standard deviation of 9 19 percent, the standard deviation of our B/C for our 20 dosimeters in our plant was less than that, was 7 percent 21 and if you're looking to see that there's no systematic 22 errors here, there's no -- and it clearly shows a l 23 consistent bias.
24 MEMBER KRESS: Let me ask you, ask it another
(~~
(/ 25 way, I guess. For the looks at the difference between the i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISi.AND AVE., N.W.
j (202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234 4433 l
t
211 1 bias and the BE/C, and if one is asking to take full g~3 2 credit for that bias in terms of adjusting the fluence, l )
L~ J 3 the implied assumption is that the BE does actually 4 represent the true fluence.
5 MR. HANSON: No, the assumption that it 6 represents the fluence that you use to determine what your 7 transition temperature does. That's the point I'm trying 8 to make.
9 MEMBER KRESS: Aren't you really interested in 10 what the true fluence was?
11 MR. HANSON: No, I'm interested in what the 12 transition temperature is of the material in their after 13 vessel. I really don't care what the fluence is. I care
- )
k/ 14 about what the transition temperature of that metal is and i
15 --
16 MEMBER KRESS: So when we look at this, we 17 should ask whether or not that best estimate number you 1
18 have is the same best estimate that actually goes in to j i
l 19 the --
20 CHAIRMAN SHACK: The Reg. Guide.
21 MEMBER KRESS: The Reg. Guide. And now I 22 still have the same question though. That is -- we'll say l i
23 that is in the assumption in trying to take the full 24 credit for the bias and what I'm hearing from the staff is
.~
( ,) 25 no, it's not the same best estimate that goes into the l
l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
212 1 Reg. Guide.
,s
, 2 It's because there is a large mecsure of i )'
\/
3 calculation involved in that best estimate and that the 4 calculation and process that you use for your plant 5 specific doesn't capture all of the biases that are 6 captured in the best estimate calculation that went into 7 the Reg. Guide. That's what I'm hearing. Your plant 8 specific calculation doesn't really capture everything 9 that went into the rule itself.
10 MR. HANSON: Oh, I think it clearly does. The 11 discomfort among the staff, and I'm speaking for them, but 12 -- and they'll get a chance to speak for themselves I'm 13 sure is that we have considerably different B/C, but the t i A/ 14 fact is the way we calculate our BE our best estimate is 15 consistent with the way not 100 percent. It's not the 16 exact same code, but it was a fair to say slash 2-code. I i
17 haven't gone into the details between those, but they're 18 both least squares methodology using the averaging, maybe 19 Stan can get into a little more detail, 20 MR. ANDERSON: Yes, when Bob Simons at Hanford 21 redid the data base, he used the SAND / FERRET code which is 22 the same code that we actually got from Hanford that we i 23 use now. He requested calculated input spectra from the 24 varicus vendors as input to the specific surveillance ex
(_) 25 capsule evaluations. I know Westinghouse supplied those.
i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
213 1 I cannot say whether other vendors did or not, but the 73 2 dosimetry evaluation procedure that he used there, we
! 1 3 essentially adopted. So it's the same code that's the 4 same approach.
5 Lambros has referred to numbers in the code.
6 The code itself does a standard least squares algorithm.
7 It's mathematical. The input to that are three pieces of 8 data. A calculated spectrum that comes from the transport 9 code and that would have changed from ENDF IV to ENDF VI 10 because you're doing calculations differently.
11 It requires the raw reaction rate measurements 12 that you make that come from the laboratory. Those are --
13 were measured once. When he redid it, you could not go
\J 14 back and measure them again from 20 years ago because they 15 decayed away. Those procedures are very well known and we 16 can shown measurement consistency over the years.
17 The third piece of information are dosimetry 18 reaction cross sections. At the time that Simons did the 19 prior analysis, he used ENDF V dosimetry data from the 20 ENDF files. Today, we use the library that was prepared 21 under ASTM auspices by Pat Griffin at Sandia Labs. It's 22 primarily an ENDF VI library, although he reevaluated the l
l 23 Japanese data bases and an Oak Ridge data base and the 24 ENDP VI library and produced an evaluated set of cross p
( ,/ 25 sections. Basically, those are the three pieces of data, l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 i
214 l
i 1 if you will, that go into that evaluation. There's no fT 2 built in libraries or built in data or built in numbers in 1
O 3 the code per se.
i 4 MR. HANSON: Does that answer your question?
5 MEMBER KRESS: That helps. It really didn't 6 fully answer it though.
7 MR. MAYFIELD: Let me take a little different 8 crack at it. When the fluence values in Reg. Guide 1.99, 9 Revision 2 is based on 177 data points. The fluence 10 values, as I understand this, the fluence values were not 11 quite as clean as lifting them from the surveillance ,
l 12 reports over some subsequent assessments made and
_ 13 adjustments made to try and come up with a best estimate.
i' ') 14 Those best estimates are not -- were not l
I i
1 15 arrived not through the same kind of process that's being I i
16 used today. I think if we went back in and could revisit I 17 the whole thing, we probably would do something much more 18 akin to what's on the table today with some of the nuances 19 that have to get sorted out, but the process would be much 20 cleaner, I think, than what we did several years ago.
21 One of the issues, the staff is struggling 22 with, having nothing whatsoever to do with Palisades 1
l l 23 directly is can you use the revised cross section 24 libraries and current state of the art to estimate fluence
/~N k -) 25 for the vessel in a best estimate sense and then tie NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
215 1 embrittlement estimates and what is the RT NDT or the l
<w 2 transition, temperature and material? Make those 1
( )
%J 3 estimates using the process that's in Reg. Guide 1.99.
4 There is plainly a difference.
l 5 Is that difference important? And that's 6 something we're looking at as part of the research l
7 program. That's one of the questions as we've considered 8 a third revision to Reg. Guide 1.99, what do you do with 9 that ugly little piece of information, is it important?
10 We haven't sorted it out yet, quite frankly.
11 We've marked forward just like we were satisfied with what 1
12 the answer was, but we have this side discussion about is i 13 this the right thing to do?
]
l f3 L
l l
\/ 14 We haven't brought that to the public arena jn l
' l 15 a particularly forceful way and we certainly haven't l
16 settled it on the staff. !
17 Does that get more of what you were --
18 MEMBER KRESS: Yes, I think that addresses 19 more specifically what --
20 MR. MAYFIELD: I certainly wouldn't want to 21 tell you we have the answer today.
22 MEMBER KRESS: Yes.
23 MR. MAYFIELD: It is something we're talking 24 about.
(~'
(,,T) 25 MR. HANSON: Mike, you said, you implied they l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l
i 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 i
1 l
l l
216 1 didn't use the same least square?
.3 2 MR. MAYFIELD: There was some subsequent 1
( )
~ ~/
3 adjustments made and unfortunately Boda and --
4 MR. HANSON: But they were made by Simons. l l
5 MR. MAYFIELD: The fellow from -- I think he 6 was from Louisiana at that point. I think there was some 1 7 subsequent adjustments made. Williams, I believe. Neal ,
8 Randall and company got involved in this and they were I 9 making best estimate adjustments to the best estimate 1
10 Simons calculation as I understood this. We can go back 11 and dredge up the history, but it wasn't quite as clean a 12 process as I think we would like to have today.
13 MR. HANSON: We went back and we saw that they o i k/ 14 use the Simons data and he did use the least squares 15 methodology --
16 MR. MAYFIELD: But there was apparently some 17 additional adjustments made.
10 MR. HANSON: But we did note that there was 19 some additional --
20 MR. MAYFIELD: There was some tweaking around 21 with it and unfortunately the people that can tell you 22 what adjustments were made and why aren't --
l 23 MR. HANSON: They were very minor adjustments.
24 MEMBER SEALE: Which Williams was this?
,\
(u./ I 25 MR. MAYFIELD: Mark Williems. But that l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433 l
i 217 '
1 brings, how major or minor were they, to be determined.
,- s 2 It's a fairly small data base and we've been a bit
. x
' ) 3 bothered by the fact that that was going on and how do you 4 tie that to current methodology and current cross section 5 library.
6 MR. HANSON: Now in the development to the new 7 Reg. Guide 1.199, Kurt, are there any easements done for 8 you? The inputs to him are best estimate based on FERRET.
9 MR. MAYFIELD: The inputs to him are the 10 values that were submitted in surveillance capsuled 11 reports that have been updated based on inputs from the 12 industry through this ASTM review process. And one of the 13 things we are working and hoping to hear more from the o
f )
> 14 industry, unfortunately, Kurt Cozens is hiding behind that 1
15 column and I can't quite see him, one of the things that l 16 we're hoping to hear back from the industry is a little l
17 more information abut exactly the basis for each and every l l
l 18 one of those points and that's something that I think 19 we're a lot closer than we were, but I can't tell this 20 Committee that we have the answer for each and every one 21 of those data points.
22 I think before we went a lot further with 23 publishing a proposed revision to the Guide, we're going 24 to have to resolve some of those questions and make sure
(,)
25 that we can document and have documented the basis for t HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433 l
I 218 1 each and every one of those points.
l l ,3
^ 2 Quite frankly, I don't want to get to this l U 3 point five years from now where we've got a guide on the 4 street and whether it's Palisades or some other plant l
5 where we're in here worrying about the basis for data 6 points that are in the Reg. Guide. I want that documented l
7 so there is no question and I don't have to worry about 8 whether Neal Randall is dead and buried, to be able to 9 resolve these questions. I don't ever want to get in this 10 situation again. i 11 We think we can pin down what we have. WE're 12 confident that it is certainly a conservative process in j l
13 the Reg. Guide. When you start getting outside of that l
) l 14 and doing other things, we've got some questions that have 15 to be answered.
16 I don't know whether that helps or hurts, but 17 it's --
18 MR HANSON: What's the schedule on the 19 development of the Reg. Guide 1.99, Kurt?
20 MR. MAYFIELD: There's an ASTM meeting next 21 week --
22 MR. VASSILAROS: 21st in Boulder.
l 23 MR. MAYFIELD: 21st in Boulder, Colorado where 24 there will be some discussion of the technical work that's g
(_) 25 been done to develop those curves. Then the staff has to l NEAL R. GROSS
! COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234 4433 i
i 219 l 1 sit down and answer some of these other questions about do 2 we really want to move forward with this, do we have !
f' 3, t l N/ )
3 sufficient documentation. I would hesitate to try and l
4 give you a schedule today because those questions may 5 become the rate limiting factor in coming up with another
- 6. revision of the Reg. Guide.
7 MEMBER KRESS: The bulk of the data that goes 8 into the screening criteria, does it comes from 9 surveillance specimens that come out of actual reactors or 10 were these developed in test reactors?
11 MR. MAYFIELD: The data bare for the screening 12 criteria came exclusively from survc111ar.ce data. It was
,._ 13 an interim version of what is Reg. Guide 1.99, Revision 2
/
N'- )\ 14 today.
15 MEMBER KRESS: Surveillance. l 16 MR. MAYFIELD: Surveillance. We have never 17 used test reactor data explicitly in any of these fits.
18 It's used as supplemental information to gain 19 some confidence in trends we see, but it's never used 20 explicitly.
21 MR. STROSNIDER: I'd like to make one comment 22 in this area. This is Jack Strosnider. I think to some i
23 extent it supports what Mr. Hanson has been saying and 24 also I think it's important for the Committee to recognize
/~
(_T ) 25 that one of the important things in doing the pressurized NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
220 1 thermal shock assessment is that you do it from an g ~s 2 integrated perspective and a consistent perspective.
G 3 And the reason I want to make this point is we l
4 talk about a schedule for the revision of the Reg. Guide 5 and that's probably ahead of a lot of those other things 6 and we have to be very careful about taking a new 7 embrittlement trend curve and saying now we're going to go 8 apply that to measure where we are with regard to the 9 screening criteria because the screening criteria had 10 certain margins, etcetera, built into it based on the old 11 methodology. So when you change one part of the problem, 12 you need to look at the whole thing.
13 So on the one hand, yes, it needs to be done
--]
14 consistently and I think there's no question about that.
l 15 On the other hand that creates some other problems because l 16 it makes it difficult to work one piece of this problem at 17 a time and you have to be very careful even when you get 18 into plant specific evaluations, doing something 19 specifically here that's not consistent, so two inch 20 short, if you will, that it adds a great deal of 21 complexity. I think you can see that from the 22 presentation.
23 MEMBER KRESS: If -- let's look at the 24 screening criteria. If the data base that goes into it 25 comes out a specimen, it pulls out of the actual NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234 4 433
1 l 221 ;
1 surveillance, you measure the embrittlement compared to i
,s 2 its original shape, the change, the shifts. That's what r i
%J 3 you -- you go ahead and measure that with the sharpie 4 meter or something, then you calculate the fluence based 5 on the operating history of that plant.
6 If the calculation of procedure that you use 1
7 to calculate that fluence that went into that data base, 8 it's different than what he does then I think you have to l
9 say you don't accept what he's doing. He has to do his l 10 fluence calculations the same way it was done in these 11 data reports.
12 MR. MAYFIELD: And the question is how 13 different can it be?
/~') !
\ 2 14 MEMBER KRESS: How different can it +, okay. j i
15 MR. MAYFIELD: And that's the part that we're l
16 sorting out. Does it have to be identical? Well, I'd 17 like to answer that's no.
18 MEMBER KRESS: It can be different.
19 MR. MAYFIELD: How different can it be and 20 still provide you a legitimate basis for comparison?
21 MEMBER KRESS: And that goes back to my 22 question on the sensitivity calculation.
23 MEMBER SEALE: So you do that in the ENDF/B-IV 24 calculation which is probably more like what was used to
' (~h
() 25 calculate --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 23^-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 I
222 1 MEMBER KRESS: Yes, it has a higher fluence l
7- 2 though. l N,)]
l 3 MR. HANSON: But that's what I told you '
1 4 before, is that we tried to address that issue in our 5 fluence analysis, that if there are some things different, ;
l 6 we were asked to use the ENDF/B-VI versus the ENDF/B-IV.
l l 7 And Labros said in his presentation that the l
8 recommendation of the staff to the utilities is to use the l I
9 ENDF/B-VI. However, when we did our fluence submittal, we 10 did analyze how much difference -- or actually it was an j 11 RAI to our fluence submittal, that we did analyze how much l
l 12 impact did that have on our fluence calculation versus if
,\
13 we would have done it with the ENDF/B-IV. That's what I
.6 *
'\ Y 14 was saying earlier that we believe that the net difference l
15 was --
l 16 MEMBER KRESS: It's not that big there.
17 MR. HANSON: -- was in one area it appeared to l
l 18 increase it by 1 to 2 percent and in another area it l I
I l 19 reduced it by 1 percent for a net change of less than 1 l
l 20 percent and overall net change on the impact of the actual i 21 calculation.
l 22 So the difference between our methodology 23 today and the methodology back then as best as we can )
24 determine, there was no real significant difference. We i
(m) _, 25 did our best to do the same type of method, the same l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
223 1 methodology as Simons did, with the least squares fit.-
1 7y 2 We did recognize that there was a slight
- e 3 difference between the numbers that were actually used in 4 Reg. Guide 1.99 than what Simons put out when we looked at l 5 the data, but we also knew that the people that put those 6 curves together and Simons worked in the same office and 7 he was continuing to work on that and we thought that i 8 there might have been some update in his analysis.
9 MR. MAYFIELD: There may well have been, but 10 there was another piece of it as well.
l 11 MR. HANSON: Yes, I wasn't aware of the 1
12 Williams work. I 13 MR. MAYFIELD: And how that all factored in, i
(
14 I'm afraid some of the details were not documented as well 15 as we would like today.
16 MR. HANSON: But the reality is that I believe 17 that we have as close, we have used as close of a 18 methodology as you could, as you can, as anyone has in 19 industry to use, to come up with the same basis for our 20 fluence evaluation as was used in the Reg. Guide.
21 MR. ANDERSON: Let me also add back in late 22 1995 or so we redid all of the Westinghouse surveillance l
l 23 capsules with ENDF/B-IV technology so that they were all 24 done on a consistent basis.
I
(_j\ 25 As we've gone through these 21 plants or so NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433 i
224 1 that Jack has shown you, essentially all of that data for r~N 2 those 21 plants has been redone to ENDF/B-VI. The
(
)
/
3 calculations have changed. The dosimetry evaluation and 4 the best estimate value that we would report for that l l
1 5 capsule changes by less than 2 percent because the 6 measurements don't change. Evaluations of the 7 measurements don't change. It's the BE to C ratio that's 8 changed.
9 We have always reported in surveillance 10 reports a best estimate value. Those remain stable.
11 MR. HANSON: And I would like to point out 12 that the use of ENDF/B-VI, now we're going back to the
,_ 13 original curve I put up here, that the use of the ENDF/B-
[ )
\ 14 VI -- no, I'm sorry. Wrong one. Sorry.
15 MEMBER SEALE: It's a horse of a similar i 16 color, but not quite the same. l 17 MR. HANSON: The point that I was trying to l 18 make is that the average B/C for those 21 plants is now 19 .95. Prior to the use of the ENDF/B-VI, that average B/C l 20 was 1.15, in the area of 1.15. So it used to be 1.15 and 21 we were asked to use a BE/C at 1.15 in our 1992 PTS 22 submittal by the staff because that's what everybody else 23 had and when it was redone for these 21 plants, at least 24 to incorporate improvements in the ENDF/B-VI over ENDF/B-p
(_j 25 IV, that BE/C dropped.
NEAL R. GROS 5 COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
225 1 The average BE/C for those plants dropped from
(, w) 2 1.15 down to .95 which then became in our state almost the
'~'
3 same. What was ours before we used the ENDF/B-VI, Stan, 4 do you remember?
5 MR. ANDERSON: .93.
6 MR. HANSON: Okay, it was .93. Now it's .83, 7 but theirs dropped considerably to come closer to ours 8 after the use of the ENDF/B-VI. So there's -- I guess the 9 point I'm making is that the ENDF/B-VI helped make the 10 BE/C closer to 1 and it also made the rest of the plants 11 and industries BE/C look more like ours, but we were the 12 only ones that were less than 1 before, but it was that 13 change in the cross section of an iron that had a big 14 impact and because we had two inches less of iron between 15 our core and our vessel wall, there was a big difference 16 there.
17 It changed everyone else's a lot more than it 18 changed ours because of that change in the iron cross 19 sections and in fact everyone else had a thermal shield 20 and we didn't.
21 So that's why we tried to make the point that 1
22 there are differences, physical differences between our 23 vessel and our core design than there are -- and that does 24 explain, but not to a specific degree why our BE/C may be 1
/~h
( ,) 25 different.
l i
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. ;
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
226 1 But when you look at the BE/C calculations for s 2 our plant, the determined uncertainty of those fall within i \
\
/
~'
3 that assumed in the rule. The rule assumes a 20 percent 4 uncertainty and our uncertainty of our best e' stimate 5 fluences is less than 20 percent and I believe it's in the 6 15 percent, Stan? 15 percent.
7 And the margin terms in the rule account for 8 20 percent uncertainty. So that's the point I'm trying to 9 make.
10 Any other questions or comments?
11 MR. LOIS: This is Lambros Lois again. At the 12 risk of echoing my own arguments that you've heard over 13 and over again before, that's really not argument. The e s
\ 'l 14 argument is the quality of the data that we're looking at.
15 The 20 percent that's in the rule, I calculated it to BNL 16 back in the early 1980s. At that time the capability and 17 the results of the dosimetry measurements were far more 18 apart than they are today. We are really improving 19 dosimetry data, has been improving.
20 However, the 1 sigma deviation that was shown 21 on the data that's Plants A, B and Palisades that I showed 1
22 you before, indicate that the change that is claimed in 23 the adjustment is way outside of the 1 sigma deviation for 24 the measurement and in that sence we are having a problem.
(_,/ 25 But again, it is a matter of consistency of the plants,
)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W. i (202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 I l
227 1 specific data and then a consistency of the methodology in
~s 2 itself from plant to plant. Thank you.
/ i
~~
3 MR. MAYFIELD: Mr. Chairman, this is Mike 4 Mayfield from the staff. I'd like to try and clarify 5 something that Jack said on his third slide, having to do 6 with Code Case N-557 on thermal annealing. He said 7 something to the effect that there's no effort on the 8 staff to review and approve this code case.
9 A process for us reviewing, approving code 10 cases has to do with when ASME publishes the things in 11 various supplements. Code Case N-557 was published in a 1
12 supplement that we got some few months ago. It's normal 13 process it would have been reviewed and nominally approved l t/ 14 in what would have been revision 14 to Regulatory Guide l
15 1.147. ;
1 16 Because of the importance of that code case, 17 we've pulled it out of the particular, its normal 18 progression. It will be picked up in revision 13, 19 principally because Palisades, from what we can tell had 20 backed away. We decided to not hold up revision 12 to 21 that Reg. Guide and try and pick up 557 at that time.
22 The current process for Reg. Guides, these 23 code case endorsement Reg. Guides has changed a bit and we 24 now have to go out for public comment. Previously, they
(T
( _,
/ 25 were simply published for use. There was a change in the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 228 l
1 process and we're now forced to go out for public comment
, fs 2 with a draft revision.
l t )
1 %J 3 The current process would have that draft 4 revision going out at probably December of this year. We ;
5 received some urging from the Commission to look for ways 6 of expediting some of these review and approval procedures l 7 and that's something the staff is exploring so we may get 8 it out ahead of that schedule for public comment, but 9 should be the latest will be December of this year.
10 I know that's something that you probably l 11 weren't aware of. Hopefully, it will make the people that 1
12 worked trying to get the code case out feel a little .
1 1
13 better about it, but it's not a matter that is being l l
'/ 14 ignored. That was, it will be picked up and actually it's i l
15 an accelerated process, 16 MR. HANSON: Thank you.
17 CHAIRMAN SHACK: Well, I think we better take 18 a break here for lunch since we're down to 45 minutes for 19 lunch anyway. We'll be resuming at 1 o' clock.
20 (Whereupon, at 12:25 p.m., the meeting was 21 recessed, to reconvene at 1:00 p.m.)
i 22 l
23 24
()h
( 25 i NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
229 1 A-F-T-E-R-N-O-O-N S-E-S-S-I-O-N
(~x 2 (1:09 p.m.)
t .
%/
3 CHAIRMAN SHACK: We're ready to resume our 4 subcommittee meeting and our first presentation this 5 afternoon will be on NUREG-1511, which is this reactor ;
6 vessel -- reactor pressure vessel status report, and Ms.
7 Lee will be making the presentation. ;
1 8 MS. LEE: I'm Andrea Lee and I work in 9 Materials and Chemical Engineering Branch, and I'll be 10 presenting information on NUREG-1511, the reactor pressure !
i 11 vessel status report. )
i l
12 I'd like to start by giving a brief I
,- 13 introduction, and then I'll give some details on Generic
\
'-) 14 Letter 92-01, Revision 1, Supplement 1. And also, the 15 staff completed some plant specific pressurized thermal 16 shock evaluations, so I'll present the results and 17 conclusions.
18 And the structure of the reactor vessel 19 integrity data base has changed since the initial issuance 20 in August of 1995. And in this presentation, I'm going to 21 focus on the development of the RVID; and in the next 22 presentation, I'll project the new Access version on the 23 screen; and I'll conclude by summarizing some plans for 24 future activities.
(s,/ 25 NUREG-1511 was originally issued in December NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
230 1 of 1994, and the report summarized the staff's review and I
,3 2 responses to Generic Letter 92-01, Revision 1. And the i 1 s /
'~'
3 report stated that Beaver Valley Unit One and Palisades 4 were projected to exceed the pressurized thermal shock 5 screening criteria prior to expiration of license.
6 And that conclusion was based on data that the 7 staff had at the time, and the report stated that the 8 conclusion was subject to change based on the availability 9 of new data. Supplement 1 to the NUREG was issued in 10 October of 1996, and that basically summed up the staff's 11 basis for issuing Supplement 1 to the generic letter, and 12 it described the status of licensees' responses to the 13 supplement.
g i )
\_/ 14 It also discussed licensees' compliance with 15 the pressurized thermal shock rule, and it outlined the 16 staff's development of the reactor vessel integrity data 17 base. And I know that the subcommittee has seen the 18 August 1996 tape of the Commiasion brief on annealing, and 19 also Deborah Jackson, the next presenter after me, will 20 present the annealing demonstration project, so I'm not 21 going to be covering annealing in my presentation.
22 There was some issues that were identified by l
23 the staff after the issuance of Generic Letter 92-01, 24 Revision 1. And specifically, different chemistries were
(
(_)) 25 being reported for welds that were fabricated using the l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l
231 l
1 same heated weld wire.
-3 2 Also, licensees had not considered all of the
( ) \
\' ,/ 1 3 data that's relevant to their RPV integrity evaluations in l l
l 4 their initial assessments. Greater than expected l
l 5 variability was also identified for welds that were l 6 fabricated using copper coated electrodes, and the staff i
l 7 also discovered the RPV embrittlement is subject to change '
8 by small changes in chemical composition.
9 The Supplement 1 to the generic letter was 10 issued in order to gather information to assess the above-11 mentioned issues. Supplement 1 to the generic letter 12 requested that licensees identify, collect, and analyze
~
13 any and all data that's available for their RPV integrity
/ 'T
~. 14 assessments.
15 And it also requested that licensees determine 16 the need to use the Regulatory Guide 1.99 Revision 2 ratio 17 procedure. And basically, that procedure says that if 18 there's clear evidence that the chemistry of the 19 surveillance data differs from the chemistry of the vessel 20 data, which is often the case, then adjustments have to be 21 made in the embrittlement calculations to account for 22 those changes.
23 In terms of status of responses, all licensees 24 have responded at this point to the Supplement 1 and the r^
( )T_
25 staff sent out administrative close out letters to l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, O C. 20005-3701 (202) 234-4433 I
232 1 acknowledge the issuance -- the acceptance of the
,m 2 responses. The Combustion Engineering Owners Group and iI
'% )
3 the BWR Owners Group have yet to respond.
4 Now the owners groups have developed programs ,
5 which involve extensive searches of fabrication records 6 and logs, and the last of the responses from the owners i l
7 groups is expected by summer of this year. Assuming that 8 we get the responses by summer, we could finish our j 9 evaluations by the end of this year.
l 10 And to give you an example of some of the l I
11 programs, combustion engineering searched through 450 l 12 boxes of fabrication . cords and logs and came up with the l i
13 methodology to get a best estimate chemistry from
- 14 pedigreed data.
15 MEMBER KRESS: Owners groups are not required 16 to respond to generic letters, are they?
17 MS. LEE: That I don't know. l l
18 MR. STROSNIDER: This is Jack Strosnider.
19 The request to respond goes to the licensees.
20 However, one of the main points here is making sure that 21 all licensees are aware of all of the data that are 22 available in reference to their vessel. So it has to be
! 23 done in an integrated way. Basically what the industry 24 approach to doing that was to form up owners groups.
() 25 These are vessel owners groups, so these are -
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
m 1 -
233I 2
MEMBER KRESS:
3 These are different.
STROSNIDER: -
4 people who have vessels manufactured by combustion engineering 5
design, NSSS, okay.
not necessarily CE 6 And they pulled together and pooled their resources to go back and do thi s sort 7 of evaluation and put all their data into a common d 8 ata base.
MEMBER KRESS: Thank you.
9 MS. LEE:
10 The Babcock and Wilcox owners group responded that the reports that have bee 11 throughout n submitted the years on the docket make up th i 12 the best estimate chemistry e r basis for 13 and they also responded that the ratio procedure is not needed t 14 data. o be used with their
, 15 16 And the staff has reviewed these r eports. We broke it up into an effort of three pe 17 ople and reviewed an i
extensive number of reports.
18 There were questions that were forwarded to the owners group ,
19 audit conducted at and there will be an 20 the Framatone facility in Lynchburg May 19th through the 21st.
21 The licensee -
22 CHAIRMAN SHACK:
23 They didn't need the ratio J
procedure because their surveil e ~ --
233 1 -
2 MEMBER KRESS: These are different.
3 MR. STROSNIDER: -- people who have vessels 4 manufactured by combustion engineering, not necessarily CE 5 design, NSSS, okay. And they pulled together and pooled 6 their resources to go back and do this sort of evaluation 7 and put all their data into a common data base.
8 MEMBER KRESS: Thank you.
9 MS. LEE: The Babcock and Wilcox owners group 10 responded that the reports that have been submitted l
11 throughout the years on the docket make up their basis for 12 the best estimate chemistry, and they also responded that 13 the ratio procedure is not needed to be used with their 14 data.
15 And the staff has reviewed these reports. We I
16 broke it up into an effort of three people and reviewed an i 17 extensive number of reports. There were questions that ))
18 were forwarded to the owners group, and there will be an l 19 audit conducted at the Framatone facility in Lynchburg May i
l 20 19th through the 21st. :
1 1
21 The licensee --
l 22 CHAIRMAN SHACK: They didn't need the ratio {
i 23 procedure because their surveillance specimens were so 1t 4 24 close to the actual material?
+
25 MS. LEE: Well, the consensus was that --
i NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 5 1323 RHODE ISLAND AVE., N.W.
t (202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
234 1 exactly what you said. But there are some inconsistencies 1
rw 2 from what we saw in the reports, which is why we're going l l t ,i l x_/ i 3 to go out and do this audit. l l
4 There are about 28 units that submitted new i 1
5 data, both PWR's and BWR's. But the licensees concluded 6 that these data don't have any effect on the previously 7 submitted RPV integrity evaluations. And as I said, it l l
8 will be part of my conclusion that the staff contin'ues to l 9 review the new' data. ,
1 10 CHAIRMAN SHACK: That would confuse me a 11 little bit too because in the initial evaluation that was I
12 Beaver Valley and Palisades that had the problem, then i
1
,s 13 there's something about the exception six RPV's.
r 1 l
14 MS. LEE: Right. I 15 CHAIRMAN SRACK: When did that change from two 16 to six occur?
17 MS. LEE: Okay, what happens is when new data 18 is submitted, surveillance capsules are pulled. Those 19 data have to be rolled into the calculations, determined 20 for credibility, what have you. With the generic 21 assessment that you're talking about that the staff 22 conducted -- and I go into this a little bit more later, 23 but I can answer it now.
24 After the Palisades review, there was greater
( i
( 25 than expected variability looked at due to the evaluation
(_/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344 433 WASHINGTON, D C. 20005-3701 (202) 234-4433
235 1 of the steam generator data, so that prompted the staff to i
- 2 do a generic assessment. So we took generic chemistry
\~)
3 values for sets of data like combustion engineering, B&W, 4 and then we used increased margin terms.
5 That was a very conservative analysis. It 6 didn't address any plant specific information. It was 7 just really done to show that there's enough time to do 8 the plant specific evaluations. That's where the six 9 plants came from after that assessment.
l l
10 So -- and that's why it's always stated that, ;
l 11 you know, the conclusions are subject to change due to new 12 data, and that's a very true statement.
13 The Electric Power Research Institute
(~N!
\
\ 14 developed RPV data, and that's an access data base. It 15 has the same type of set up as the reactor vessel 16 integrity data base in terms of putting data in a 17 consolidated manner.
18 It has the reactor vessel integrity data base 19 as a subset, and it also has a screen that compares what's 20 in the RPV data to what's in the RVID so that data 21 inconsistencies are easily identifiable. There are 22 differences. RPV data is basically a flat file with no i 23 calculations, and the RVID is the relational data base 24 with calculations throughout.
,CN
(,,) 25 And I'll get into a little later about how l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l l
236 1 we're going to try to match those up.
l l .
s 2 10 CFR 50.61 is the pressurized thermal shock
! ,i l
V 3 rule, and the screening criteria is given in terms of a 1
4 reference temperature RTPTS. That quantity is obtained by 5 adding an initial reference temperature to the change in 6 the temperature due to radiation plus a margin.
7 And that margin is added to account for 8 differences in the initial reference temperature, copper, i
9 nickel, fluence, and also a calculation of the procedures.
10 The screening criteria for plates, forging, and axial 11 welds is 270 and for circumferential welds it's 300 .
12 As I mentioned earlier when I was answering 13 your question, there was a generic assessment that was
's I 14 done by the staff, and that was directly out of the 15 Palisades review. When the information and data from the 16 retired steam generators were evaluated, we discovered the 17 larger than expected variability.
18 So the staff assessed all PWR RPV's and used 19 the generic chemistry values and increased margin. And 20 the result was that no unit was projected to reach the PTS 21 screening criteria before seven years from 1995. And in 22 fact, most of the vessels were projected to reach the 23 screening criteria after the expiration of license.
24 And as I mentioned, this was an extremely
, r 25 conservative evaluation, and it was done just to show that
(_h HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
i
237 1 there is time to do plant specific evaluations. And 73 2 again, it didn't consider any plant specific data.
t !
~
3 In terms of the plant specific PTS evaluations 4 that were conducted and reviewed by the staff, Palisades 5 and Calvert Cliff's Units One and Two accounted for 6 variability -- coil to coil variability by using a 7 weighted average as opposed to a simple average to 8 determine the best estimate chemistry.
9 The staff concluded that the Palisades RPV in 10 this 1995 SER would satisfy the criteria of the PTS rule l
11 until late 1999. And as you know, subsequent to the SER, !
l 12 the thermal annealing report was submitted in I believe 13 October; and in April of '96, the reduced fluence
- n\ l Y2 14 evaluation. So that result has changed.
1 15 The Calvert Cliffs Unit one and Two vessels j 1
16 were projected to remain below the screening criteria 20 ;
17 years after expiration of license. And also, the Ginna 18 plant specific evaluation showed that the vessel would 19 remain below the screening criteria after expiration of 20 license.
21 And this was an important conclusion because 22 the staff's generic assessment projected Ginna to be the 23 first vetael that would reach -- that was the seven year 24 point -- would reach the screening criteria in that seven
( ,j 25 years from 1995.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
l 238 1 But again, when the plant specific data were s 2 evaluated, it brought the conclusion to the fact that it (N /
}
3 would remain below the screening criteria expiration of 4 license. And the key to remembering all of this, and I've 5 said it probably like three times now, that the 6 conclusions are based on data that's available.
7 And if there's new surveillance data, which is 8 a common occurrence, or new best estimate chemistry data, 9 these conclusions can change.
10 MEMBER KRESS: When you do generic 11 calculation, what do you use for the plants? Do you l
12 actually go --
13 MS. LEE: What we took is a 20% of the mean n
( )
'/
' 14 value which was consistent with what was done with the PTS 15 rule.
l 16 MEMBER KRESS: Did you actually get the number 17 from the plant specific --
18 MR. ELLIOTT: No, excuse me. What we did was 19 we determined from generic -- Barry Elliott. i I
20 What we did was the generic data we used was l l
21 the generic copper and nickel values for different types l 22 of plants. And the uncertainty in the fluence values was l l
l 23 the generic value, 20%, as we discussed. And then what we 24 do is we do a Monte Carlo simulation of the -- of a vessel f3
(_) 25 using these generic copper, nickel, and fluence l l
NIL 4L R. GIUDSS i COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W. l (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
239 1 uncertainty; and we figure out when a generic vessel would r~3 2 reach the screening criteria.
I i
\, _)
3 Then we go back and look and see when the 4 actual vessels would reach it -- reach that value. So 5 this generic value gives you a generic fluence value which 6 --
7 MEMBER KRESS: There's a number used for all 8 the plants?
9 MR. ELLIOTT: No, no; this is just a generic 10 value to determine what the effect of increased 11 variability of copper and nickel would be. It is not in 12 the PTS rule. It was just a scoping calculation to give 1
13 us a handle on a conservative estimate of how bad things I
/ \ l
! /
14 could be if we decided that all the vessels had very wide 15 variability in copper and nickel. j 16 MS. LEE: Right, but the 20% was what was used 17 in the rule. That was the point I was making.
18 MR. ELLIOTT: Right, as far as the fluence.
19 MS. LEE: Okay.
20 MR. ELLIOTT: But we didn't have a -- we don't 21 calculate for each vessel of fluence. We calculate a l 22 generic fluence for a particular type of vessel, and then l
23 we compare that to what the actual vessels have to get 24 when the reach the screening criteria.
O
(_) 25 MEMBER KRESS: Okay, thank you.
NEAL R. GROSS
- OURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(
(202) 234-4433 WASHINGTON D.C. 20005-3701 (202) 234-4433
l 240 1 MS. LEE: The reactor vessel integrity data 1
g 2 base or RVID was developed because there was just massive O 3 amounts of information after the Generic Letter 92-01 l
l 4 responses were received. So the staff wanted to have a 5 comprehensive way to put all this data in one place so 6 that future reviewers like myself, because I wasn't '
1 1
7 involved in this effort when it first started, would be I
1 8 able to go back and see what information was submitted in l
1 9 response to the generic letter. !
10 And there were data from various documents 11 included like pressurized thermal shock reports or 12 pressure temperature limit reports and also responses to 13 request for additional information. It uses the i f
'V 14 Regulatory Guide 1.99, Revision 2 calculational 15 methodology throughout.
16 The data base was originally issued in August 17 of 1995, and it was distributed on diskette via mail.
1 18 Administrative Letter 93-03 announced the availability of l
l 19 the data base to the industry and to the public. I took 20 over as technical monitor in October of 1995. And the 21 structure of the data base -- some of the structure and 22 the user's manual were revised in the data base in zipped 23 form -- the diskettes in zipped form were reissued on the 24 World Wide Web in June of 1996.
s V 25 And Revision 1 to the administrative letter NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l l l 241 )
- i l 1 announced the availability of the data base. And for l l
I l
7\ 2 individuals that don't have access to the Web, diskettes
( )
l
' x/
3 are still available; and I have mailed those out, and we I
, 4 will continue to do that in the future.
I I t 1 5 Once we get all the data, all the responses, I l l 6 we can add any new data to reviews we've already done.
7 And we plan to update the reactor vessel integrity data l i
l 8 base and then reissue it. But there may be, you know, l
9 some amount of time longer than we would hope before tha* i i
10 happens. l l
11 But me, personally, I feel that the data base )
i 12 should be updated before it is reissued again on the Web 1
13 or elsewhere. So the staff is going to work together to f -~3 l 4 i
\/ 14 get the information addressed, assessed, and to update the 15 data base.
l 16 In terms of future activities, as I mentioned, 17 the staff will continue to review the data that we've 18 received from Generic Letter 92-01, Supplement 1. And 19 also, there are some -- what I would call data 20 inconsistencies that have been existing since Generic 21 Letter 92-01.
22 So the entire package will have to be revised.
23 And when that happens, then we can reissue the data base.
24 I will review any data that comes from the owners gro'ap.
/m 25 And again, that's expected by summer of this year. We'll
( _)
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., %W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
242 1 update the RVID, as I said, after we complete all reviews.
7y 2 And it's not just a review; it's a review, t i
%J 3 update, and a QA of what goes in the data base, so we want 4 to make sure that it's right before it goes out again.
5 Once discrepancies between the RVID and the industry data 6 base are resolved, and that could be a major effort, then 7 we could establish an NRC approved data base.
8 And our long term goal, and we've discussed 9 this before with NEI, is to have this data base maintained 10 by the industry and audited by the NRC by comparing what 11 is in the data base to docketed information. That's a 12 long row before that happens. We'd have to make sure that 13 the inconsistencies are resolved.
'- # 14 MEMBER KRESS: Let me ask you a question. l l
15 My view of data bases in general is that this 16 is a collection of data that people can access and all of ,
1 l
17 them use. But it seems to me like all of these data !
l l
18 you're talking about are plant specific data. l 19 MS. LEE: The data that is cont'ained in the 20 data base?
21 MEMBER KRESS: And if a licensee's going to 22 access it, all he's going to use is the part of it that's 23 relevant to his plant.
24 MS. LEE: Well, there's -- and I'm going to
,y is,) 25 show this next. There is a very important part of this NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005 3701 (202) 2344 433
243 1 data base called the user defined report. It was called l
gs 2 query in the old version of the data base. And that
(
u'j
)
3 allows licensees to do cross cut searches of data.
4 So if you had a weld fabricated with the same l 5 heat of weld wire as another plant, you can go across the 6 data base -- l 7 MEMBER KRESS: Oh, that would be useful.
1 8 MS. LEE: -- and get that report and use it.
l l
9 So it's very valuable for people to search other plants.
l 10 In fact, I think that may be the most useful 11 tool in the data base.
12 MEMBER KRESS: Yes, that helps me. That helps 13 me a lot.
/^s
\ -} 14 MS. LEE: Okay. I'll take any more questions 15 and then I'll move on to showing the data base on the 16 screen.
17 MR. STROSNIDER: Andrea, this is Jack 18 Strosnider. Before you move on to demonstrating the data 19 base, if there are any questions, I just want to make a 20 few comments.
21 I wanted first to emphasize and hope that the 22 committee recognizes the magnitude of some of the effort 23 here. With regard to the initial review of Generic Letter 24 92-01 -- this was after Yankee Rowe -- we asked people to l
(y
( ,)
s 25 give us all the information we'd need basically to asses NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
244 1 their reactor vessels.
,m 2 It took roughly five people a year. We i )
's J ,
3 assessed every reactor pressure vessel in the U.S. And as '
4 Andrea mentioned, we got a large amount of data, and that 5 was a very intense effort. One of the things we 6 recognized was that with all these data, we should have 7 them in a data base.
8 And that allowed us for the first time --
9 although some of the reviewers had noted in the past some l 10 differences between reported values from licensees. When 11 we had it all in this data base, we were able to perform 12 this cross cut that Andrea just said is so important. I
, 13 And what we saw there is that for what we
/ ;
i 14 referred to as sister welds, that is plants with welds 15 made from the same weld wire, that we were getting 16 different reported chemistry values. And the difference 4
17 in the values that were being reported was more than you 18 would expect just from measurement error or variability in 19 the material.
20 There were differences that could make 21 significant impacts on PTS evaluations and that sort of 22 thing. And what this really gets down to, when we were 23 talking earlier about margins in the rule -- when the rule 24 was developed, we used their ability in chemistry, copper (3
C) 25 in particular, that had been taken from a section of a NEAL R. CROSS COURT REPORTERS /,ND TRANSCRIBERS 1323 RHODE IS! AND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
245 1 weld where the variability was measured.
~s 2 That weld -- actually, there were several
( )
3 welds that were looked at. But it turns out they were all 4 made with one coil. What happens though is when you look 5 at the variability from coil to coil, even though it's the 6 same heat, the copper coating process could have a much 7 larger variability from coil to coil.
8 That's what we were seeing. And that's where 9 the larger variability came from, and that's the issue we 10 were trying to address.
11 The other thing that became clear when people 12 were reporting data to us and that there were these large 13 differences is that at least there was our concern that t )
'/ 14 they weren't sitting down and sharing all their data. And 15 we confirmed that in fact that was the case.
16 There were some licensees that weren't aware 17 that another licensee might have data relevant to their !
18 plant, which is another important aspect of this data base i 1
19 and one of the reasons we've made it public. So we had a l 20 large effort doing that. The generic letter supplement, 21 on the other hand, has caused a large effort by the l l
22 industry now.
23 Things have been, frankly, somewhat quiet in 24 the vessel area. We've been doing some plant specific p
(,) 25 reviews in that. But generically, it's been rather quiet NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
246 l
l 1 for over the last year. The reason is that the industry -
- 2 - it acnears to be quiet. But actually, there's a lot of N.) 3 work going on behind the scene.
l l 4 The industry spent a lot of resources going 5 back and reviewing the original fabrication data and i
6 comparing from plant to plant, and we know this based on l
7 some of our interactions at various meetings. The 8 response to the supplement will hopefully be an integrated 9 data base in which all the licensees have considered all 10 the data relevant to their plant.
11 And one of the things we'll do is we'll update i
12 this data base and we'll go through and we'll look for any i 13 discrepancies. Now that's not to say everything has to be I
.\~/ 14 exactly the same. There might be some legitimate l
15 explanations for some of that, but we'll have to '
16 understand it.
l 17 And so we are looking now at another major 18 effort when we get these responses middle of this year to l
19 go back and potentially reassess every vessel in the l l
20 country again. I think it will be much easier because we 21 now have the data base in hand. We've been through it 22 once before.
23 But that's kind of the big picture. Like I l
24 say, it's a major effort which, hopefully when we've
(~'N
( ,) 25 completed this review, we will have a base line data base NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
L (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 I
247 l
l 1 and a base line assessment of all the plants in the i
l!
I I
f3
'x,_,/
s 2 country with regard to the current regulations.
3 And that's our goal, and it will probably take 4 us into next year, you know, before we're going to be able l 5 to complete all those reviews.
l l
6 Now Andrea will probably show you some l
l 7 carefully selected searches that --
l 8 MEMBER KRESS: That work.
9 MS. LEE: Just to give you some background, l
l 10 the data base was originally programmed using Fox Pro l
11 software which is Dos-based, and it's been transferred 1
12 over virtually -- the contract is virtually complete to 13 Access which is Windows-based. And I think it's a lot
/~
14 more user friendly than the older version.
15 I say virtually complete because the manual l
16 still has to be drafted. And there is training scheduled 17 for May 15th in the auditorium downstairs. There will be 18 an industry session and a short session for the staff to 19 address updates and LAN considerations because it will be 20 available on the LAN.
l 21 Also, what I'm going to be showing you, this 22 is in the full rights update mode. And what I mean by 1
1 23 that is currently what's being worked on is when I log 24 into this data base, there will be a user name requirement
(~'N
\ ,) 25 and a password. The public will have a user name that i NEAL R. GROSS
! COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2344 433
l 248 1 just gives them the rights to view, and any buttons i
'e3 2 related to updating or anything like that will disappear.
1
( \~)
3 You have to have the unique user name and l
l 4 password combination to be able to update, and there will 5 be even limited staff that will have that so that we're l
6 consistent with making changes.
7 This is a 496, and it's -- you've just go --
8 you know, it's going to be a little slow, so you just have !
9 to deal with it. l 10 MEMBER KRESS: Could you make that thing do 11 flips or something? 1 1
12 MS. LEE: No.
13 Now at the top level, or what I call the top i i
\'- 14 level of the program, there are two sections. There's a 15 plant information section and a report menu, and I'm going ,
1 16 to start with the plant information. To get to the l
1 17 listing of plants, you select here. I 18 MEMBER KRESS: Try Palisades.
i I
19 MS. LEE: Now the plants are listed here along 20 wit., important information like manufacturer, expiration 21 of license date, docket number; and they can be sorted l l i 22 according to any of the column headings. If you wanted to l
l 23 know which plant is going to expire first, you would 24 choose that as the sort criteria.
(_) 25 If you want to get into looking at the details NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
249 1 of the beltline materials, you go to the next level. What g 2 you see first is general plant information such as vessel N-] thickness, vessel radius.
3 And you also see references for 4 the data. And there's a drop down list here. If you want l
5 to see additional notes -- not every plant has additional 6 notes, but this one does.
7 And there's also a section for information on l
l 8 the limiting closure flange material. And this level has 9 two different sections as well. If you select this, you 10 get a listing of the beltline materials that you can go 11 into later and look at details.
12 If you select here, the listing changes into i
_ 13 surveillance data. We'll go back and start with the
- i. !
14 forging plates and welds. You can either access the list 15 like this which I didn't feel was as user friendly, so I 16 had the contractor add, once you're within the screen, the 17 buttons that you see at the top backward and forward so 18 you don't have to actually go back out to the list to get 19 to your next material.
20 And the way that the data base is set up now 21 is that the left side has all of the input data and the 22 right side has all of the calculated data. And the l
l 23 benefit to doing this -- let me go Lack. This isn't the 24 first one. This isn't my sample material, so I'm going to
/~'s
(_,) 25 go back to my sample material.
NEAL R. GROSS I
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433 l
f 250 l
1 The benefit of doing this is that if you make l
,y 2 a change -- and this was not the way that it was set up
! )
3 before. If I were to make a change in the initial 4 reference temperature -- if you notice now, RTPTS is S 102.9. I wanted to change it to three. You see the 6 automatic result of what your change does throughout.
7 And these colors aren't showing up that well,
~
8 but the contractor color-coded it so the three quantities 9 that I mentioned in my other presentation, the initial 10 reference temperature, the delta and the margin are 11 yellow; and it adds up to the RTPTS value which is in 12 blue.
13 Here you have a drop down list that allows you p
I 14 to select your method of determining your initial 15 reference temperature. And here your other inputs such as 16 fluence -- these two quantities go into calculating the 17 margin. The reason why they're blank is, if for any 18 reason the staff needeG to do an override of a value, they 19 would be -- the values would be entered in this section.
l 20 And then when you select the appropriate l l
21 method, that value is reflected over on this side. And l
! i l \
- 22 here you have information on unirradiated upper shelf ;
! l 23 energy, and there's a drop down list as well to pick the 24 method.
g
( ,) 25 And as I mentioned, over here in the
! NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
I 251 1 calculated data, one important aspect of this section is
,ss 2 the chem;stry factor. And there are four different
- ( )
3 methods that could be used. The table method is selected 4 here. If we determine that a plant had credible 5 surveillance data and we wanted the plant to use the ratio 6 procedure after we've done an evaluation, they would 1
7 select this method.
8 If they wanted to do surveillance data without 9 using the ratio procedure, you'd select this and the 10 number changes according to what you select. As I 11 mentioned, if you need to override the value, you'd select 12 the method here, put the number here, and it would be 13 reflected on this side. ,
/,_N l I }
k/ 14 Down at the bottom, you have the upper shelf 15 energy at expiration of license and the percent drop. And 16 this value comes from a programmed method of transferring .
1 1
17 the reg. guide graph that appears in the regulatory guide. !
18 So it's an algorithm that inserts these numbers into these 19 fields.
20 And you can -- with any screen that I'm going 21 to show you, you can print it. If you want it to print 22 out for an evaluation of this particular page, you could 23 print it. You could save it. You don't need to have the 24 save button. I just had them put it there because as soon
() 25 as you close out of this form, any changes that you were NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 252 1 going to make are automatically saved.
f~s 2 Here this may seem strange to have a new t
x y.-
3 component button; but in our evaluation, the staff has 4 split the plants up, and we each have about 20 units.
5 And there are cases where the licensee may have listed 6 values separately that we group together, welds with the 7 same heat.
8 So if we want to separate it out, you select 9 this. And if you go through this process and hit 10 continue, it will give you a screen just like this and you 11 can fill in the blanks. Here when we go over to the 12 surveillance side, now what you see is a listing of the 13 surveillance data.
/
-' 14 If you go into the detail of that screen, you 15 come up with general information for this particular 16 surveillance data section. You have your fluence, your 17 measured delta RTNDT values. And the important points I l l
18 think on this screen are this button, this button, and 19 this one and this one.
20 And they interact in the way that I'm about to i
21 explain. When the staff assesses surveillance data, we go 22 by the regulatory guide credibility criteria. And if the 23 data is deemed to be credible, then the licensee could cut i
i 24 that margin term that I mentioned in half.
- \ ,/ 25 So in this example, what we call the residuals NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISt.AND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (207) 234-4433
i i 253 1 in the reg. guide are the predicted delta values minus the l
,y 2 measured delta values. For plates, must be less than or b l V
3 equal to 17; and for welds, less than or equal to 28 to 1
4 satisfy that criterion.
I 5 In this case, ais 11.9, the absolute value of ,
l !
l I l 6 that, 11.9, would be less than 17. So in this case, it's l l
l 7 designated as yes because it's credible regarding -- with I ,
8 regard to that regulatory guide scatter criteria. But if l 9 you went to another material or another surveillance set ,
1 l
10 -- this isn't showing up; but in this case, it's 21.3.
11 It's greater than the 17 for the plate. So )
i 12 what this is, it's actually red. The previous one that
' ,, 13 said yes was in green. This is in red, and it says no. )
/\
) 14 So what the staff could do in this case is look at the l
l l
15 group standard deviation. And what that is, is the ;
1 16 residual values that predict minus measure, the set of 17 data for representing this particular heat.
18 And if the standard deviation of this set is 19 less than the criteria which is 17, the staff could assess 20 this data and say that it can be used in the embrittlement 21 calculation. But in this case, the margin term would not 22 be cut in half. It would be a full margin term utilized l
l 23 because it didn't meet that first level of credibility.
24 And these are decisions that we can make. And
/-n
(_) 25 what this is is a tool that will help us make those NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
254 1 decisions. And here, for whatever reason, if the staff gS 2 decides that this data base is not credible for use once O 3 we've gone through these types of exercises, you would 4 select here and say no, that the data is not going to be 5 used in the chemistry factor calculation, anu the:. it will 6 be reflected throughout the data base that this piece of 7 data is not being used.
8 Now let me warn you. This is going to take a 9 second. But before I leave this plant, what I call the 10 plant information section, I want to show you another 11 feature that's been added that will be very helpful to the 12 staff, and it's a new feature in this data base that
,- 13 wasn't in the old one.
i l
\ l
' If you go to a material that's using 14 1
15 surveillance data to calculate the chemistry factor -- and l
16 I just know that this one does, and that's why I picked it 17 -- you come up with an additional feature, a button, that 18 allows you -- here it's saying it's using surveillance 19 non-ratio as a method.
20 If you click on this button -- this is what's 21 going to take a second. What it's doing is it's going in, 22 pulling from all the numbers that it has, doing 23 calculations, and then hiding the screen that you see 24 here. What you're going to end up with is selections to
( ,/.
25 show you exactly what numbers were used to go into these NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344 433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
255 1
1 calculations for ratio, chemistry factors, and basically
- 2 every calculational procedure that's in the data base.
l wi 3 Okay, say you wanted to see exactly how the 4 chemistry factor was calculated using the surveillance 5 data, you would select chemistry factor. Here it gives l l
6 you the formula that was used, it gives you the pieces of l l
1 7 data that were used, and it does the sum of the squares 1
8 method. And to me, this is a valuable -- it can also be i 1
9 printed.
10 This is a valuable sheet for us to do our 11 assessments. And when we have errors, whether they're 12 slight or great between what we see from the licensee and 13 what's in the data base, we can see exactly where the
'\ J 14 problems occur. And this also is -- I guess it would be 15 called real time because if I were to went out -- go out 16 and make a change in copper and nickel or fluence, every 17 time I pull this up, those changes will be reflected here 1
18 and you'd see the resulting chemistry factor.
1 19 The reason why the ratio says one here is for l
20 programming considerations. This is surveillance non- l 21 ratio. So for non-ratio, it's just like saying that the l l
22 ratio is one. If I were to go out and say surveillance j 23 ratio, then the appropriate ratio would be reflected here.
24 Here, this is kind of an outline for someone
( ,j 25 that just came into the data base and doesn't really know NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344 433
256 1 what calculational procedures are occurring. This 7s 2 outlines everything that is being calculated in the data
(' ' ' ~ )
3 base. And it also, for whatever material you're looking 4 at, gives you the quantities that go into calculating 5 RTPTS.
6 And again, this can be printed as well. And a 7 button that will get a lot of use is the new component 8 bu-t"n for surveillance data. As new capsules are pulled 9 and data is assessed, we will use this to put the new data 10 in. And again, if you choose, you know, whatever type you 11 want and you go through and hit continue, you'll get a 12 screen like this that allows you to put all the data in.
13 Okay, now that's the plant information section e
k' 14 from the top level. There's also a report menu section, 15 and this section has a query that I was talking about. It 16 has a pressurized thermal shock summary report. You can 17 either select all plants and each of the plants will be 18 reviewed on the screen one after the other -- but for 19 demonstration purposes, I'll just pick one.
20 And wnat you see is a report with the 21 components that go into calculating RTPTS. If you wanted 22 to see what it's going to look like fu?1 screen, you just 23 click your mouse to see how the report will print out on 24 the screen. But it has all the components, all the g~s 25 methods; and what I showed you in the individual plant
()
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4435
257 1 screens is what is represented here.
7s 2 If any changes are made in that plant screen,
() 3 they'll be reflected in this report. You can either print 4 the report from here or return back to the screen that I 5 just showed you.
6 MR. STROSNIDER: Andrea, let me make a quick 7 comment. You happened to pick a boiler there, and they're 8 not subject to RTPTS evaluations.
9 MS. LEE: Oh, did I?
10 MR. STROSNIDER: However, that same 11 information is useful in pressure temperature limit 12 evaluations.
13 MS. LEE: And actually, --
[}
14 MR. STROSNIDER: I think you had Big Rock l
1 15 Point.
16 MS. LEE: I picked Big Rock?
1 17 It should say at the top of that report 18 pressure temperature limits report. Because, you know, we 19 could change the name of that button, but --
20 MR. STROSNIDER: I'm not sure if --
21 MS. LEE: It says PTS. Okay.
22 Well, that's a note.
l l 23 MR. STROSNIDER: We're still debugging the 24 system.
/ 6
() 25 MS. LEE: Okay. The upper shelf energy 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 l
l 258 1 summary report is the same set up. You pick the plant, go ,
l
,S 2 to the preview selected plant, and you get the same type i N ,) l 3 of report. But it has the parameters that went into 4 calculating the upper shelf energy. l 5 And one of the important features of this data 6 base is the user defined report or the query. For a 7 plant, you can keep the defaults for vessel design or you 8 can choose from a list. The same with all the quantities.
9 But I'm going to select PWR for the type of reactor. I'm 10 going to select welds. And for heat wire, I'm going to 11 select 10120.
12 If I put a check mark in this box, then the
,_ 13 results will be sorted in terms of alphabetical order of
('-) 14 the plants. If not, it will be sorted by whatever 15 criteria I pick first here. So the first column will be 1
l 16 copper, second nickel, and third I'll pick RTPTS. So once 17 you have all your parameters set up, you can run the l 18 query.
19 Now one feature that was added to this data 20 base that wasn't in the old one is a summary of whatever 21 selection criteria that you pick. That way you can see 22 what you put for an input and what you get for the output.
l l
l 23 And in this case, this shows you the units that have this 24 heat, and it gives you copper, nickel, and RTPTS for the
(_) 25 selected parameters, i NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS j
1323 RHODE ISLAND AVE., N.W.
! (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
259 1 MR. STROSNIDER: Andrea, could I interrupt you
,r' 2 for just a second?
O) 3 MS. LEE: Uh-huh.
4 MR. STROSNIDER: This is an interesting one 5 that you'll notice that the copper values -- yeah, these 1
6 are very low copper plants. I'm wondering if you could go 7 in and take a CE plant. j 8 MS. LEE: Okay.
i 9 MR. STROSNIDER: Maybe take a couple. Because '
10 I'd like to show -- this data base will still show some of 11 the discrepancies that we saw in the original 92-01 ,
l l
12 responses. Because that won't all be updated until we get
,_s 13 through this next review effort.
i s
\i 14 MS. LEE: Barry, off the top of your. head, do l
15 you know a good --
16 MR. ELLIOTT: Use Palisades, W5214.
17 MR. STROSNIDER: This is my human computer 18 next to me.
19 MR. ELLIOTT: This is not pre-rehearsed, so we 20 don't know.
21 MS. LEE: It's down at the bottom? There.
22 Okay. So there are some differences here.
- 23 MR. STROSNIDER
- Here you can see some 24 differences. These are not -- we see more significant q
(_,/ 25 differences than what's reported here. But it just gives l
NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., il W.
i (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 i
260 1 you an example of, you know, the kind of cross cuts we can 2 do.
O 3 When we first put this data base together, we 4 took ten welds. I think they were CE fabricated welds.
/
5 And we found more than half of those had what we 6 considered significant differences in reported copper 7 values. Ten different heats that had sister plants.
8 MS. LEE: And again, as I just showed you, you 9 could either go back to requery to the screen that you 10 left, or you can go back to the main screen or print. ,
11 And the last thing I want to show you is
(
12 what's called the look up tables. And one of the comments 13 when I first met with the contractor was that he wanted 14 everything to be available to the user. If they wanted to 15 see what the chemistry factor tables from the regulatory 16 guide looked like, he wanted it within the data base.
17 So they've included the Regulatory Guide 1.99, 18 Revision 2 chemistry factor tables for base metals and for 19 welds. And they also included this percent drop upper 20 shelf energy table. And what this represents is the slope 21 and the Y intercept that is contained within the graph in 22 the reg. guide.
23 So this is just kind of a numerical 24 representation of what that graph has in the regulatory -
25 guide.
NEAL R. GROSS -
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
261 1 And that's all of the features. I'll take any s 2 questions.
i
\
i
\
~'
3 MR. STROSNIDER: Andrea, I just wanted to make 4 one other point. I don't want to interrupt the commitCee, 5 but could you show how the data that are in this, how you 6 can go back to the reference document?
7 MS. LEE: Uh-huh.
8 MR. STROSNIDER: This was one of the very 9 important things that we wanted to include in this data 10 base. As Andrea indicated, three, four, five years from 11 now, we're not sure who might be doing these reviews. I i
12 We've included in this for all the raw data that goes into 13 it, you can go back and find out what report it came from.
('~h
, 1 N/ 14 All right, so we wanted an auditable data 15 base, and that's something that we think -- I think strived hard to accomplish th'at .
I 16 17 MS. LEE: There is two places where the 18 references come into play. What we have here is, for any 19 report that you're looking at, rather -- whether it's 20 pressurized thermal shock or upper shelf energy, the same 21 references are at the bottom of that plant.
22 And the reason why is when I was at the main 23 screen and I went over in this area and showed that box, 24 changes can be made in one place, and the changes are
/ \
( ,) 25 automatically reflected in both of the reports. But this NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, O C. 20005-3701 (202) 234-4433
262 1 has all of the reference information here.
f~s 2 And this, you know, can be traced back to l l \
\ !
^'
3 whatever documents or reports that the data came from.
4 And part of our update effort will be to update the 5 references, you know, delete as appropriate if we need to; 6 but we just have to be very careful about not just getting 7 caught up with just revising the numbers.
8 We need to revise the references as well.
9 MR. STROSNIDER: A final comment, I guess, is 10 one of the issues that came up when we first put this data 11 base together was some of the calculational algorithms ,
l 1
12 that are in here. And I think it's important to point out '
13 that we initially put this together with the idea that it 1 \
(~') .
ij 14 would be a licensing data base.
l 15 And by that, I mean that values of chemistry 16 or RTPTS or upper shelf that are docketed with the NRC and 17 reviewed and approved as a licensing value, those are in 18 here. All right, there are also, however, as Andrea 19 illustrated, some calculated values. All right, and those 20 go through some standard algorithms and, you know, where 21 we've put in equations for the tables and the reg. guide 22 and that sort of thing.
23 Those are good tools for us to be able to do 24 our reviews, but those are not licensing basis numbers g-() 25 unless, you know, we put a reference in there which l NEAL R. GROSS
! COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 263 l 1 basically says it's been docketed and reviewed in an SER. l l
i
! 7 -~s 2 But we have maintained some of these calculational --
1i )
i V 3 we've had quite a few tools in there that will help us do 4 our reviews.
5 And I guess a final comment, you can see it's 1
1 6 regulation of operating reactor pressure vessels, our 1 7 assessment. This is a huge bookkeeping effort. All 1
3 right, and it's not necessarily high tech in concept, but 9 it takes this sort of data base and a lot of attention to i
10 detail to keep it all straight. l 1
11 And you know, this is a tool we put together )
l 12 to do it. We've emphasized, like I say, making the thing 13 auditable; and also, the QA that goes into, making sure l l' 'e
14 the data are right. And Andrea broke some ground, I
- 15 think, with our computer people making sure we could get ,!
l l 16 this thing out on the Web and stuff so it will be 1
17 accessible to everybody.
18 MEMBER KRESS: Are there other data bases out 19 there that you guys -- not you guys, but the NRC has that 20 I could access and just get general design characteristics 21 of all the plants like how many loops it has and volume it l
! 22 has and --
23 MR. STROSNIDER: I'm not certain. I don't 24 know if anybody is aware of whether the NRC has it out. I
,7
() 25 can tell you that -- I can tell what my branch is doing.
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
l 264 1 The other data base that we're putting together is a steam
,r-y 2 generator data base which does have, at least for the l
\-]
3 PWR's, that information in it.
l 4 I'm not certain that there's anything out
- 3 there right now. I just -- I don't know.
6 CHAIRMAN SHACK: It looks like we're ahead of 7 schedule again.
8 MEMBER KRESS: Yes.
9 CHAIRMAN SHACK: Debbie's here though.
10 Our next presentation is on the DOE reactor 11 vessel annealing project update and I guess the NRC 12 interaction.
7, 13 MS. JACKSON: Good afternoon. My name is i
\' # 14 Debbie Jackson, and I'm a mechanical engineer in the 15 Electrical and Materials and Mechanical Engineering Branch l 16 in the Office of Research. And today I'm going to give 17 you an update on the DOE annealing demonstration project. j l
18 This is an outline of the topics I'll cover 19 today. I'll go over the objectives, the NRC's objectives, 20 a summary of events to date, the ADP program success, l l
21 status of both Marble Hill and Midland, and a few lists of 22 items that the NRC staff still has to do for Marble Hill.
23 These were the DOE objectives for the DOE 24 annealing demonstration project. Right now we refer to
, ,, 25 this as the DOE annealing prototype demonstration project.
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS
! 1323 RHODE ISLAND AVE., N W.
i (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 I
l 265 1 We had a meeting with DOE last week and they informed us
,3 2 that they have inserted a new word, so they now refer to
( )
3 this as a prototype demonstration project.
4 And annealing of vessels has been performed on 5 the Russian design, the VVER, and it's been completed 15 6 times. It was done twice on one particular vessel. It's 7 never been completed in the United States successfully on 8 a U.S. design. So that's why DOE was interested in going 9 through and testing the annealing process on a U.S.
10 design.
11 MEMBER SELE: I understand they've annealed 12 all of the VVER-450's. Is that --
13 MS. JACKSON: They've done 14 vessels. One )
0 1 V 14 was done twice. The number of --
15 MEMBER SEALE: Yes. But the 450 megawatt ones i 1
16 are the ones they've done them all I think now, haven't l
l 17 they?
l 18 MS. JACKSON: Right.
19 MR. HACKETT: This is Ed Hackett from the 20 staff.
21 Bob, I don't think I can answer you that 22 they've done them all. They have -- as Debbie said, 23 they've done 14. Unit three was done twice. And the most 24 recent one is Loviisa Unit One in Finland.
/%
25 MEMBER SEALE: Thank you.
l
{)
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433 l
I 266 1 MS. JACKSON: This is a list of the NRC l
,s 2 objectives that we had when we were participating in the
>( )
'~'
3 DOE prototype demonstration project. We were not a 4 participant in terms of sponsoring funding. We were l
5 strictly an observer. There were no NRC inspections 6 performed on the Marble Hill site.
7 We specifically just observed the installation 8 of equipment and the process that went through during the i I
9 annealing process. We did have confirmatory 10 instrumentation. We had specifically two thermocouples :
I 11 that were on the nozzle support. We had four strain i i
1 12 cauges at the RPV nozzles and four displacement gauges at i l
13 the base of the reactor coolant pumps in the steam l r i
\/ 14 generators.
t 15 This next --
16 MEMBER KRESS: Did you guys do any pre-17 predicting of the --
18 MS. JACKSON: Excuse me?
19 MEMBER KRESS: Did you all do any pre-20 predicting before the test of the displacements and 21 stresses and --
22 MS. JACKSON: Yes, there was models. Yes, l
i 23 there were models that were done before the test.
24 MEMBER KRESS: Did the NRC do them or did DOE?
(-
( ,! 25 MS. JACKSON: They were done by DOE.
l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE , N.W.
l (202) 234-4413 WASHINGTON D C. 20005-3701 (202) 234-4433
267 l 1 We also had a model, but we didn't have as t
l
,s s 2 much instrumentation as the ASME DOE group had.
((Q )
l 3 And this is a list of the summary of events to i
4 date. The DOE funded the two projects. It was Marble l
5 Hill and Midland. And the Marble Hill and Midland were 6 both funded by two separate consortiums. The Marble Hill
- 7 consortium consisted of DOE, Westinghouse, the ,
t 1 1
8 Westinghouse Owners Group, Cooper Heat, and Parsons Power.
I 9 The Midland consortium consisted of also DOE,
- 10 Sandia, and Mott, which is a Washington consortium, and ,
! l l
11 NPR associates outside of -- they're located in l
l 12 Alexandria, Virginia.
13 Now I would like to show you a short video of i t .,
l \ '} 14 the actual annealing at Marble Hill.
j l
1 l
15 MEMBER KRESS: Marble Hill is the gas fired 16 annealing?
17 MS. JACKSON: Yes, that used the gas fire with l
18 the propane.
19 Marble Hill is located in Washington, Indiana 20 right outside out Louisville, Kentucky. This is Unit One, 21 and the other one was Unit Two. These are thermal 22 barriers that were inserted in each of the nozzles to 23 prevent the air flow down through the nozzles. One was 24 specifically designed for use in Palisades.
A
(_) 25 It was a remotely installed barrier. The NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISt.AND AVE., N.W.
, (202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
\
l l 268 I I ;
l 1 other ones are just insulation sandwiched between two j l
l I r~.s 2 pieces of metal, as you can see. And these are just all l l 1 1 LJ l 3 cables. This is one of the spot welded thermocouples.
4 This is inside of the vessel.
l 5 They did NDE measurements on the inside of the l
6 vessel. I'll talk about that a little later.
7 CHAIRMAN SHACK: This is a what, a two looped, 8 a three looped --
9 MS. JACKSON: This is a four loop Westinghouse 10 plant. And this is outside -- we're outside of the vessel 11 area. Some of the reactor coolant loops. You can see 12 some of the instrumentation. This was the heater. This p_ 13 was a five zoned heater. And this was the duct work that
(
)
14 was connected to the propane tanks.
15 This is a lifting rig that was used to lift 4 1
16 the heater into the vessel. You can't see it, but there 17 was a convection barrier that's sitting up vertically. l l
18 During the anneal, it was down in a horizontal position.
19 MEMBER KRESS: When you say indirect fired, 20 does that mean the -- you heat something that then 21 radiates the heat to the wall?
1 22 MS. JACKSON: Thest are the -- the propane was 23 piped up to burners, and the 1.rners were at 100 , and then 24 that forced the air down into the heater exchanger that C\
(_,) 25 you saw that went actually into the vessel.
! NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS i 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C. 20005-3701 (202) 2344433
269 l
l 1 MEMBER KRESS: I see.
r')
2 MS. JACKSON: Now these were the burner
(' 3 controls. There were six of them. This is a burner right 4 here. There were six. Five because they had five zones 5 to the heater, five for inlet, and there was one extra.
6 MEMBER KRESS: What was the target temperature 7 for this?
8 MS.. JACKSON: It was 850 . I have a slide that 9 gives the range of the temperatures and the gradients.
10 This is inserting the heater into the vessel. This is a 11 convection barrier if you can see it, the white that's 12 around the edge of the vessel.
,_ 13 And these were guides that helped us lower it.
(\' i 14 These were the -- there were ten connections for the duct I i
l 15 work, five inlet and five outlet, each going to a j 16 different zone of the heater. Now this is the connection 17 -- this is the gas going into this -- this is the burners.
18 This is all of the duct work. These are the five outlet 19 pipes from the heater -- excuse me, five outlet ducts from 20 the inside of the heater.
21 The ones on the bottom, if you can see them --
22 there were actually two levels; five on the bottom, and l 23 five on the top. These were inlets going into the top of 24 the heater and five coming out. And they were all
/~'s
(_) 25 insulated. There was a problem with the duct work. It NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
270 1 took a bit longer to put in place than they had s 2 anticipated.
l )
iJ 3 This is some of the data we collected.
4 Is that Cooper Heat or was that the -- that 5 was Westinghouse, okay. Cooper Heat was the actual 6 contractor for the heater. They recorded their own data 7 also.
8 MR. HANSON: This is Jack Hanson. They were 9 able to monitor each of those thermocouples that you saw 10 in that vessel. They had an instant read out of rates on 11 those thermocouples. They had all the averages for the 12 axial levels and radial levels.
13 MS. JACKSON: What you just saw was the duct
,e
\ t
\/ 14 work after it had been cut off from the top of the heat 15 exchanger.
16 CHAIRMAN SHACK: Where would the control 17 thermocouples be for a real annealing? Would it be on the 18 ID of the vessel?
19 MR. HANSON: This is Jack Hanson.
20 They will be on the ID of the vessel. They 21 . t rill be on retractable arms that will cone up and place 22 the thermocouple onto the wall. They will be welded 23 thermocouples. We use both in the demonstration so we can 24 show the correlations factor between the welded
,i 3
(,) 25 thermocouples and the retractable thermocouples.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
271 l
l 1 We call them switchblade thermocouples.
,3 2 MS. JACKSON: Which are right here. Everybody i \
3 will see them as they pull the heater out of the vessel.
i 4 They were on different sides, and this is one of them. It 5 comes down and touches the inside of the vessel. They're 6 all along this part of the heater right here. That's a 7 close up.
8 MR. HACKETT: And those actually have two 9 thermocouples; one that's actually on the wall, and one 10 that's on the back of the insulation block. Because they 11 are closer to the heat source, they are affected by the i
12 difference in distance from the heat source and so we put ;
13 an insulation block behind them and another thermocouple )
l
?"% 1 1
\ >l 14 on the back of that insulation block and that gives us a l
15 much better correlation factor. l 16 Much more predictable. The problem being, of 17 course, is you can't go down in your vessel and --
18 CHAIRMAN SHACK: It's a little tough to get 19 down for that.
20 MS. JACKSGN: Right. In an actual anneal, the 21 tank wouldn't be open like that the way it was here.
22 This slide here lists the success of the 23 Marble Hill annealing. One of -- on my other slide, I had 24 a list of the participants, and that was one of the
( ,/ 25 successes. We were all able to work together, and that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N.W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l 272 1 was also essential to the success of this.
,_s 2 One of the -- the vessel deformations, they
'~
3 did pre and post anneal dimensional measurements and pre 4 and post anneal NDE measurements. So that's why they were 5 able to get the information on the vessel flaws and the 6 deformations. And the concrete temperatures were within 7 acceptable ranges. It never exceeded 150 .
8 There were a couple of problems at the Marble 9 Hill plant. One of them I briefly mentioned was the duct 10 work. The duct work, when it arrived on site, it was in 11 ten foot lengths and it needed to be welded together with 12 collars. And what was pointed out was that this would 13 have exposed people to a lot of radiation during an actual 4
.k/ 14 anneal.
15 So one of the suggestions maybe for Palisades 16 would be to have the -- if they actually use this gas 17 firing method was to have the duct work come to the plant 18 in longer lengths so they wouldn't have to take so much 19 time welding the duct work together. They actually did 20 NDE on the welds for the collars, and they had to end up 21 getting an additional shift of pipe fitters.
22 And there was a problem with the Polar crane. l 23 And speaking with one of the gentlemen who had been on 24 site since Marble Hill was built, he said they always had n
(_) 25 problems with the Polar crane. So it wasn't a problem NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
273 1 with the Polar crane just sitting. So there were a few 7e) 2 problems identified that might be specific to Palisades.
N,]
3 MR. HANSON: This is Jack Hanson again.
4 Because of the schedule of this project, there 5 was very little preparatory time. The heat exchanger was 6 built in less than time effort, and we really didn't have 7 the time to prefabricate any of the duct work. So we 8 really -- it was an on site effort getting the duct work 9 together.
10 We would not do it that way at Palisades. We 11 would do as much prefabrication work outside of the 12 radiation area as possible.
p_ 13 MS. JACKSON: Thanks.
/ 'a
- ' -- ) This is the status of the Marble Hill as of
- 1. ' l 15 today. It was completed in July, and the industry report 16 is being worked on now. It will be completed by the end 17 of the year. It's been funded by EPRI. And the DOE 18 steering committee, they're going to meet sometime this 19 week and they're going to decide on a final format for the 20 report.
21 And one important point for the report is that 22 it will allow the first commercial anneal not to have to 23 rely on corporate memory. It will have all of the 24 information documented right there in the report.
p
(_, 25 Now as for Midland, with the DOE funding cuts, l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W, (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
274 l
1 the Midland project has been delayed. But the site is !
- x. 2 ready, and it's been secured. It's been cleaned. It's I%.;)
3 ready to begin the anneal if they were in fact able to get 4 the funding from Congress.
5 The heater test report has been completed, and 6 the heater has also been tested. So it's already sitting 7 in Russia waiting to be shipped to the United States. I i
l 8 have an overhead of the actual heater. This is a radiant i 9 heater -- electric radiant heater that would be used on 10 the Midland vessel.
l 11 And these are the activities that the NRC 12 staff is presently working with. We're working with Oak
- 13 Ridge on the data from the instrumentation, and we also l
/ 3
]
14 plan on having a NUREG completed by the end of the year to 15 document all of the information on the annealing.
16 MEMBER KRESS: Will these projects be useful 17 in estimating the costs of -- comparative costs of the two 18 techniques?
19 MS. JACKSON: Yes, the approximate -- well, in 20 Midland, they won't -- they don't have a total cost 21 because the project didn't go to completion. But it gives 22 an estimate on the cost. But it doesn't take into account 23 the ALARA issues because neither plant was actually 24 irradiated, so there's no ALARA issues taken into
,rh
(_) 25 consideration.
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W.
! (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
275 l
1 And there wasn't any work done on the recovery m 2 for the vessel material, so we have no idea what that s
' '~
3 would cost either.
1 l
4 MR. HANSON: I can tell you that the -- this !
5 is Jack Hanson -- that the Marble Hill annealing l
6 demonstration project was in the six million dollar range. I 7 There's a lot of analyses that you have to do for -- to )
8 operate a plant that you don't have to do for this )
9 demonstration. ,
l 10 There are things that, if we saw there was no ;
l 11 benefit to, we didn't spend the money on. j i
l 12 MEMBER KRESS: I see. i I
13 MR. HANSON: We were very pushed to get as
( 1
\ '- 14 much as we could for the least amount of dollars to i
15 complete the project.
16 MEMBER KRESS: So it wasn't really a cost 17 demonstration. It's more so --
18 MS. JACKSON: No, because I think Midland to 19 date has spent about $2.9 million dollars. So it's not a 20 representative price.
21 That concludes my presentation. Are there any 22 other questions?
23 CHAIRMAN SHACK: Just in connection with the 24 ASME code case, what does that actually give you, o
(_,) 25 allowable temperatures and stresses for it?
l NEAL R. GROSS l
COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
276 l
1 MR. HACKETT: This is Ed Hackett from the l
l 73 2 staff.
i 1
' ~
3 That's exactly right, Bill. What pretty much 4 was in the ASME code case as opposed to our reg. guide --
5 you know, Jack and I were both involved with putting 6 together criteria for stress -- allowable stresses, i
7 deformations, temperature limits, temperature ranges and 8 excursions.
9 Those are the types of things that are in 10 there in terms of criteria and limits. As opposed to in 11 our reg. guide, we basically say these things need to be 12 addressed. The ASME code says here are the ranges you 13 need to stay within to have a success.
m U- 14 CHAIRMAN SHACK: And so there was no need in I l
15 this thing to -- and they talked once upon a time about 1
16 cutting the coolant piping, but this is -- everything is 17 attached here? I mean, nothing was --
18 MR. HACKETT: This is Ed Hackett again.
19 It's an interesting question. At Marble Hill, l
20 everything was attached. One of the debates with the 21 Midland anneal was the fact that the loop piping was cut 1
22 at the hot leg. That was a B&W type design. So they were t
l 23 cut at the candy cane type position.
24 That was one of the things NRC objected to in
- "'y
( ,) 25 terms of getting -- limiting the useful information you NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
277 1 could get from that anneal compared to Marble Hill. We i
,- 2 felt strongly that you'd obviously get a lot more better l
' (/
3 information to have the loops connected.
l 4 I think there were also windows cut in the i
i 5 cold legs at Midland too. So there were modifications to l
6 the loop piping for Midland, but Marble Hill was intact,
! 7 you know, as the four loop system instead we described.
, 8 MR. HANSON: This is Jack Hanson.
- 9 To understand the difference between the two l
10 demonstrations, it wasn't just the difference of the 11 heating method. The reactors at Midland are B&W design,
! 12 and those vessels are skirt supported which makes a very l
13 large difference in the way they behave during the heating i )
( 2 14 process and the stresses that are involved in the coolant i l
i t
l 15 piping.
1 l
l 16 The preliminary analysis showed that they l 1 l
17 would -- that they felt that they might exceed the ASME 18 code case allowable stresses if the coolant piping was l 1
19 still attached. Because of the skirt mounted support and 20 the candy cane design of the hot leg, they were still 21 working on it. They were trying to sharpen their pencils 22 on their analyses to see if they could reach it.
l l
23 And they were also looking at some other 24 issues about actually heating the skirt itself to try and
')
( _/ 25 reduce the stresses. Palisades reactor vessel is one inch
! NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433 l
278 1 --
,m 2 CHAIR 14AN SHACK: It's a CE column now, right?
' (V) 3 MR. HANSON: Excuse me?
4 CHAIRMAN SHACK: Palisades is what, a CE --
5 sort of the long columns; is that what supports it?
6 MR. HANSON: It's supported underneath the 7 nozzles. One hot leg and two cold leg nozzles. The 8 Marble Hill vessel is also nozzle supported, and it has 9 four supports versus the three that Palisades has.
10 Palisades has two hot legs and four cold legs.
11 Our cold legs are about the same size -- within an inch of 12 the same size of the cold legs at Marble Hill. However, 13 our one hot leg is considerably larger than the hot legs
( a V' 14 at Marble Hill because they have two and we have one.
15 But our reactor vessel is within one inch of 16 the diameter of the Marble Hill vessel. Our cladding or 17 wall thickness are all very, very similar. The geometry 18 of the Marble Hill vessel is very, very similar to 19 Palisades and that was one positive aspect of this 20 demonstration.
21 Now as far as the heating method, the truly --
22 now whether you heat it with electricity or whether you i
1 23 heat it with gas has very little -- there's very little 24 difference other than there are some issues about the --
C
(,h/ 25 having the gas on site from a safety analysis perspective.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 i
l
279 1 It is fairly large amounts of gas. We can
-,e3 2 limit the amount of gas on site, and we're trying to deal L !
3 with that from a 50.59 safety analysis issue, and we're 4 still -- we were still in the process of determining 5 whether or not that was an unreviewed safety question that 6 would have to come up.
7 MEMBER KRESS: Are we trying to validate a 8 particular code like ODINNA to calculate the stresses and 9 to show that the calculated values fit pretty good with 10 the demonstration, then you'd use the code to calculate 11 the plant specific values?
12 MR. HANSON: We used -- we did use a code, a
,_ 13 specific code. It was really more a methodology. I can't
/ i
-' 14 remember -- I'm blanking out on the name of the code.
15 It's a very normal code, i 16 CHAIRMAN SHACK: ABACUS?
17 MR. HACKETT: This is Ed Hackett. I believe 18 we did us ABACUS, Bill, for Midland. I don't recall 19 exactly what Westinghouse used. It might have been WECAN.
20 It's very similar to WECAN.
21 MR. HACKETT: I don't recall the exact code 22 either.
23 MR. HANSON: ANSIS -- yes, it was ANSIS.
24 So we used an ANSIS model. And the
()
(m ,/ 25 predictions came out very good. We were -- while we were NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
280 1 in the heating process, we were able to follow what we
- 2 expected for the deformations and for the strains --
3 strain gauges. There were also strain gauges on that 4 vessel at the critical area.
5 And the strains that we saw and the 6 deformations we saw were amazingly close to what we had l
7 predicted for the temperature distributions that we saw.
8 So it wasn't -- we're not qualifying a -- the purpose of 9 this was not to quality a specific code, but to validate a 10 methodology.
11 And we feel we were very successful at i 12 validating a methodology.
1 13 MR. MAYFIELD: This is Mike Mayfield from the j
(~'\ l
\~ - 14 staff.
15 Let me add to that a little bit. I think the 16 NRC staff had a slightly different agenda in what we were 17 doing with our independent analysis. We took a look at 18 this through Oak Ridge as our contractor, and we asked 19 them to, in many ways, duplicate what the DOE folks were 20 doing.
21 And the notion was to get a look at what 22 differences you can get when you bring two sets of l 23 analysts to the table and turn them loose and see what 24 happens, how much difference is there, what kind of impact
~
/ '%
! ) 25 do modeling assumptions make, how -- you know, do we end
! HEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l
281 1 up with night and day kind of questions?
fs 2 From my perspective, what we found is that
)
3 yes, you'll see some differences, that the analysts had a 4 great time arguing and feuding about the nitty-gritty 5 details. At the end of the day, there were some issues 6 that we probably would want to talk about a little more; 7 but the bottom line was they got sensibly the same answers 8 and the results from the demonstration sensibly tracked 9 the prediction -- the pre-anneal predictions.
10 I think there were a couple of lingering 11 issues that we've been looking at, some streaming above 12 the thermal barrier. I think we had a somewhat higher 13 gradient than had been anticipated. And we've been 14 looking at that and whether that is something we would l
15 want to look at more closely in the future. l l
16 But I think the sense of it, at least that I I 17 came away with, was that we got remarkably good agreement l
18 given the differences we've seen in other areas. Given 19 this is a fairly complicated thing to do, we got 20 remarkably good agreement from two independent analyses 21 using two different computer codes.
22 CHAIRMAN SHACK: And it's the computer codes l 23 that are showing you have the problem at Midland then that 24 you might have to -- that's why they went and cut the
,/~3
( ,) 25 piping or the piping was already cut?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISI AND AVE., N.W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
282 1 MR. HANSON: That's correct. It is the code 7~s 2 that would predict those stresses. However, that piping i \
\s/ 3 was cut through the financial considerations to be able to 4 write off the plant when it was canceled.
5 MR. MAYFIELD: This is Mike Mayfield from the 6 staff.
7 The steam generators at that plant have been 8 sold and they're actually setting them under a nitrogen 9 blanket. The pipe was cut and there's a blind flange 10 installed in the piping so the generators are isolated.
11 There was a fair bit of discussion about whether the 12 piping could or should be reattached.
13 I think in an ideal world we were all in
.r N i )
\' 14 agreement sure, you'd like to reattach it. Based on l l
15 conversations I had with the DOE staff involved, they 16 weren't interested in buying the steam generators. There 17 was a limit to even their deep pockets. So the notion --
18 you know, as the dust settled, that simply was not a 19 practical issue.
20 Once we all finally understood that, we moved 21 forward. And I think we could gain useful information 22 from that second demonstration even with the piping not 23 reattached. There would be some lingering questions that 24 could be resolved if it could be reattached. But it's
) 25 just not a practical matter.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l I
283 1 CHAIRMAN SRACK: You just analyze it the way
,fm 2 it is.
- i 3 MR. MAYFIELD: Just analyze it the way it is, 4 exactly.
5 Now there was some dialogue about well, could 6 you put some insulation around the pipe and basically 7 thermally reattach it, and somewhere along the line some 8 one of us finally said why, there's a blind flange 9 installed; where is this air going to go? So that's all 10 pretty much dissipated, that dialogue.
11 If the annealing goes forward, I think that 12 the plan that the DOE folks had would -- we're going to be 13 in reasonable agreement. And I think that the analyses --
)
(
\ 14 the numbers escape me, but both sets of analysts agreed i
15 there were high stresses. They were starting to mire down 16 on exactly what the number is.
17 I don't know how much credence you'd give ,
1 1
18 either set of them when you're down picking the right 19 number, but there was agreement that there were high --
20 predicted to be high stresses in some of these locations.
21 And in that part, there was agreement. It was getting 22 down to I think what's the number, and I don't know that l 23 you're ever going to settle that.
24 CRAIRMAN SHACK: Have you done your analysis f)
!.s. ) 25 of Palisades to see whether -- you do have that huge hot NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l 284 1 leg compared to a Westinghouse plant. Have you done the f3 2 preliminary analysis to see what the stresses look like
( )
x_/
3 there?
4 MR. HANSON: This is Jack Hanson.
5 Yes, we had completed that analysis. That was 6 part of our thermal -- our preliminary report. And the 7 stresses -- we had a considerable margin from the code i
8 case allowable, and that was assuming some very large 9 gradients. And we were -- the big values of this l
10 demonstration was the operational issues that we resolved l 11 and we learned how to control some of the things.
12 It was very interesting. Early on in the p._-
13 process, at the lower temperatures, convection took over
' ,] 14 and we were chasing the heat as it rose up the chimney ,
1 15 between the heat exchanger and the vessel wall. And that 16 created some gradients early on larger than we expected to 17 see.
18 But then as we got up in higher temperatures 19 and radiation started overcoming that convective heat 20 transfer issue, and then the -- well, we learned a lot 21 about how the vessel would react during the heating 22 process.
23 And we will be able to control those gradients i
l 24 much lower. When we did our analyses, we didn't bound to O
l (_,) 25 say as long as we keep the temperature gradient less than l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4 433
285 1 470 between the middle of the hot leg and the top of the
,w 2 beltline, our stresses will be within the code allowable.
l )
3 Now we did the same thing for Marble Hill.
4 But we were able to maintain those temperature gradients 1
1 5 below, and we could have done better if we had had the j l
6 operational experience that we gained from Marble Hill l 7 before we started instead of at the end. I I
8 So there were some real positive things from 9 an operational perspective that would really help anyone 10 do an anneal in the future based on what we learned. And 11 it may also affect some of the design of the future -- i i
12 there may be some design differences in terms of 13 convective barriers. I
/~N I
\s '
14 And actually, we've also looked at l l
15 supplemental heating of cartain areas. Convection took 16 over, so the top of the vessel heated very quickly and the 17 vessel was real slow early on. We could save a 18 conciderable amount of time by putting heaters on the 19 bottom of our vessel.
20 The bottom of our vessel's not insulated. It 21 would be easier for us to go on and put heaters on the 22 outside of it to heat that up to reduce those gradients in 23 that area. That also is a critical area for us because we i
i l
24 have a flow skirt -- not a -- a flow skirt, not a support
()/
(_ 25 skirt, but a flow skirt in the inside of our reactor NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
286 1 vessel that's welded to the lugs at the bottom of the l 2 vessel.
/-mi
\'~') 3 And there's some high stresses in that area l
4 too by reducing the temperature gradients in that area 1 5 also. So there are a lot of things we learned !
l 6 operationally that would make this go much smoother and i
7 keep us much further away from design limits. l 8 MEMBER KRESS: How long did you leave the 9 vessel at temperature?
10 MR. HANSON: This is Jack Hanson again.
11 We left the temperature -- we left the vessel i 1
1 12 at temperature for the full seven days. l l
13 MEMBER KRESS: Seven days. l
(~h -
l l \
'wl 14 MR. HANSON: All the annealing data -- when 15 they looked at the data base for anneals to determine what 16 the materials recovery -- anneals property recoveries are, 17 they kind of picked one hour, one day, one week as kind of 18 how they did the -- set their data points.
19 And so you find out that the vast majority of 20 the data was at seven days. So we felt that although you 21 can extrapolate between those one day and one week points, l
22 we felt that it was best to -- it was our intent at the l
t 23 anneal of Palisades to go the full seven days.
24 Although there would be considerable (O,) 25 statements to go less, we felt that it was more consistent NEAL R. GROSS COURT REPORTERS AND TRANSChiBERS i 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
i 287 I
1 with what the real data was that was available. So we I g- 2 also feel that's bounding. Once you get up to seven days, i l
~
3 an additional day is a very, very small percentage of l
4 increase. And the real majority of your recovery occurs 5 in two to three days. !
l 6 And then after that, it really tapers off in l 7 terms of how much more transition temperature --
8 MEMBER SEALE: What fraction expectation would l l
9 you have in seven days, 50%? !
I 10 MR. HANSON: Eighty percent. In the range of 11 80-90%. We have done some annealing recovery 12 measurements on our plant on our materials, and -- well, 13 our base plate and our surveillance weld which is not a f
\p )
14 credible surveillar weld. And we didn't see quite as 15 much recovery as was predicted by the reg. guide 16 correlation.
17 So whereas the reg. guide predicted around 90, 18 we saw more on the 80 as a pretty round number. I haven't 19 looked at this closely in the last four or five months.
l 20 So we did see some reduction; however, we did propose in l 21 our thermal annealing report a statistical means of .
I 22 accounting for that difference and to bias down the 23 prediction from the curve to account for our actual 24 measurement -- the fact that our actual measurements did j l
(~')s
(, 25 show less recovery. l
)
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS j 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005 3701 (202) 234 4433 i
288 l
1 We had proposed a means by which we would then 1
i l 73 2 bias down the predicted -- the recovery prediction.
(_) 3 MEMBER KRESS: Do you basically go back to 4 some point on the original curve in terms of transition 5 temperature and the slope, the change with respect to 6 additional fluence -- you actually replace your spot on 7 the curve somewhere, is that what the recovery does?
8 MR. HANSON: That's what the regulations would 9 drive me to do. And that's a conservative estimate.
10 Because what it's assuming is that your re-embrittlement 11 rate will mirror your original embrittlement rate. And 12 there's a lot of data out there that suggests that the re-13 embrittlement rate will be somewhat less.
p l I
\/ 14 MEMBER KRESS: Okay.
15 MR. HANSON: But we also had in our program t.o 16 gain some actual re-embrittlement data through our 17 supplemental surveillance program that is covered here 18 today. That's far enough off that when you gain 80%, how 19 fast it comes back with -- even using what we call lateral 20 shift method.
21 You just take the curve and you shift it 22 laterally. It would be good for another 20 years for a 23 reactor vessel at our curve.
l l
l 24 MEMBER KRESS: That would cover relicensing, l
(m i) s 25 wouldn't it?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
289 l
l 1 MR. HANSON: les, it would. i 1
,e 3 2 MEMBER SEALE: Get well in a hurry.
t 4 Q) 3 MR. HANSON: Of course, there's a lot of other 4 issues.
5 CHAIRMAN SHACK: Any further comments or 6 questions?
1 7 MEMBER SEALE: What does the staff expect us 8 to do with all of this? I 9 CHAIRMAN SHACK: Grow wiser.
10 MR. STROSNIDER: This is Jack Strosnider.
l 11 I guess we don't really expect anything in l 12 particular, although we recognize that ACRS has requested
,_ 13 copies of generic letters in the future and that, you
/ 1
--' 14 know, you may want to be briefed on those in the future.
15 This was one of the first times we've done 1
16 this, so I am curious with some feedback if this was the l 17 level of presentation that you expected or any feedback 18 for future presentations of this type. But beyond that, 19 you know, we're not expecting any letters or anything to 20 support anything.
21 What we discussed, I think it was largely an 22 update of some of the technical issues and programs they 23 have ongoing, which I think was useful because I think it 24 reflects on some of the other issues that we're discussing s_-
) 25 that you will be asked to comment on.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344 433 WASHINGTON, D C. 20005-3701 (202) 234-4433
290 1 I think at this point, if there's any comments s 2 with regard to future presentations of this type, that
( )
v 3 would be appreciated.
4 MEMBER KRESS: I thought they were just at the 5 right level.
6 MEMBER SEALE: Yes, I think, you know, we've l
7 got some learning to do; but in general, I believe the 8 desire is to try to work more in parallel as we've done in 9 some other areas on these things rather than waiting for 10 things to get pretty well in place and then suddenly you 11 go to the ACRS and find out there's some wild ideas that 12 you hadn't thought about.
13 And so, you know, getting at it early is a r~w i
('/ )
14 good idea. And I think I learned a lot, but I've got to i 15 think about it some more to decide what it was. But l
16 anyway, -- l 17 MEMBER KRESS: Are you interested in ACRS's 18 opinion on the -- how much fluence change would be allowed 19 for Palisades? Is that something you guys are going to 20 work out yourselves?
21 MR. STROSNIDER: Well, you know, the review is 22 still in process obviously; and if it's in an area that 23 you have interest in, we could obviously -- you know, f 24 obviously keep you informed on the progress there.
(_/ 25 MEMBER SEALE: Well, this is an interesting NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20e05-3701 (202) 234-4433
291 1 c.ise where there are elements of risk in the decision.
, ~s 2 And I think we've seen some things here recently that have i )
3 to do with the role that risk-based decisions might make 4 in what we do.
5 CHAIRMAN SHACK: But at the moment, I don't 6 think we're prepared to offer a comment.
7 MEMBER KRESS: I wouldn't --
8 MR. STROSNIDER: I wasn't sure if you wanted 9 to offer that comment right now or not.
10 MEMBER KRESS: No.
11 MR. STROSNIDER: We're just going to take a 12 little study.
13 MEMBER SEALE: I don't have anything else.
I I L/ 14 CHAIRMAN SHACK: If there are no further 15 comments or questions, I think we can adjourn the meeting.
16 (Whereupon, the proceedings were adjourned at 17 2:41 p.m.)
18 19 20 21 22 23 24 O.
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
_ ..~_ _ __ _ _
O O
CERTIFICATE This is to certify that the attached proceedings before the United States Nuclear ,
Regulatory Commission in the matter of:
i Name of Proceeding: MATERIALS AND METALLURGY l SUBCOMMITTEE Docket Number: N/A Place of Proceeding: ROCKVILLE, MARYLAND l
were held as herein appears, and that this is the original ,
transcript thereof for the file of the United States Nuclear Regulatory Commission taken by me and, thereafter' reduced to ;
typewriting by me or under the direction of the court ]
1 reporting company, and that the transcript is a true and accurate record of the foregoing proceedings.
/t AM/L l dORBETT'NINER" Official Reporter Neal R. Gross and Co., Inc.
i i
i i
l I
i
't u .
NEAL R. GROSS COURT REPORTERS AND TRANSCRIDERS
[
1323 R1lODEISLAND AVENLT,NW (202)234-4433 WASHINGTON, D C. 20005 '(202)234-4433
- ._ .-- . -- -. ~ . . ._ . -. . . - - _-
t INTRODUCTORY STATEMENT BY THE CHAIRMAN OF THE
! MATERIALS & METALLURGY SUBCOMMITTEE 11545 ROCKVILLE PIKE, ROOM T-2B3 l [~'\
\~ / ROCKVILLE, MARYLAND APRIL 15-16, 1997 '
l The meeting will now come to order. This is the second day of the I
meeting of the ACRS Subcommittee on Materials & Metallurgy. I am William Shack, Chairman of the Subcommittee. I have a conflict of interest with some topics of today's meeting. During the discussion of those topics, Dr. Robert Seale will serve as Chairman.
, The ACRS Members in attendance are:
Thomas Kress and Robert Seale.
The purpose of this meeting is to continue discussions with representatives of the NRC staff and Consumers Power Company concerning generic letters associated with steam generator tube inspection techniques, effective use of ultrasonic testing techniques in inservice inspection programs, degradation of steam generator internals, and degradation of reactor vessel head penetrations. We will also discuss the status of issues related to reactor pressure vessel integrity. The Subcommittee will gather x_, information, analyze relevant issues and facts, and formulate proposed positions and actions as appropriate, for deliberation by the full Committee.
Noel Dudley is the Cognizant ACRS Staff Engineer for this meeting. l The rules for participation in today's meeting have been announced ;
as part of the notice of this meeting previously published in the '
Federal Register on Aoril 4. 1997_. ,
A transcript of the meeting is being kept and will be made available as stated in the Federal Register Notice. It is requested that the speakers first identify themselves and speak with sufficient clarity and volume so that they can be readily heard.
We have received no written comments or requests for time to make oral statements from members of the public.
(Chairman's Comments, if any) l(
l We will proceed with the meeting and I call upon Phillio Rush of NRR to begin.
O O O '"!
i l
NRC STAFF PRESENTATION TO THE ACRS 1
l
SUBJECT:
PROPOSED GENERIC LETTER ON STEAM l GENERATOR TUBE INSPECTION TECHNIQUES
. DATE: APRIL 16,1997 PRESENTER: PHILLIP J. RUSH MATERIALS ENGINEER IV SION OF ENrlN RNG i 415-2790 SUBCOMMITTEE: MATERIALS AND METALLURGY j l i
STEAM GENERATOR TUBE INSPECTION TECHNIQUES EDDY CURRENT (EC) NDE METHODS ARE THE PRIMARY MEANS OF ASSESSING CONDITION OF STEA'll GENERATOR TUBES
= DEFECTIVE TUBES WITH REDUCED MARGINS FOR STRUCTURAL AND LEAKAGE INTEGRITY ARE PLUGGED OR REPAIRED BASED ON EC INSPECTION RESULTS i
- IN ORDER TO DISPOSITION EC INDICATIONS PER THE TECHNICAL SPECIFICATIONS, ALL TUBES CONTAINING DEGRADATION WITH DEPTHS IN EXCESS OF THE REPAIR LIMITS MUST BE PLUGGED OR REPAIRED (2)
o o o ;
STEAM GENERATOR TUBE INSPECTION TECHNIQUES :
.____1 ;_4 0 0 k D if f . . _ _ _ _ _ _ _
_ Span:9 Rot: 278 FLAW DEPTH = f($;i ,
~
~
.V pplM ax l Vm axllG An.llI801
.N p p ==.0,5 2V_1.18 d e g._5 6 %IW ACTUAL DEPTH OF DEGRADATION MAY DIFFER FROM ESTIMATES MADE USING EC TECHNIQUES-CRACK MORPHOLOGY SUPPORT STRUCTURES (TUBESHEET, TUBE SUPPORT PLATES)
CONDUCTIVE DEPOSITS TUBE GEOMETRY CHANGES (3)
- O O O .
i ;
. STEAM GENERATOR TUBE INSPECTION TECHNIQUES
- t i .
NRR IDENTIFIED UTILITIES SIZING INDICATIONS BASED ON ESTIMATES OF FLAW DEPTH FROM BOBBIN COIL EC DATA i I
BASIS FOR DEPTH-BASED REPAIR LIMITS IN TECHNICAL SPECIFICATIONS ASSUME NDE UNCERTAINTY ON THE ORDER OF 10-PERCENT INACCURATE SIZING OF DEGRADATION USING EC TECHNIQUES COULD RESULT IN OPERATION WITH TUBES WITH REDUCED MARGINS FOR STRUCTURAL AND
~
LEAKAGE INTEGRITY
. IN GENERAL, THE BOBBIN COIL IS NOT QUALIFIED FOR DEPTH SIZING SCC /lGA .
(4)
O O O .
STEAM GENERATOR TUBE INSPECTION TECHNIQUES
- 10 CFR PART 50, APPENDIX A - GDC 14 AND GDC 31 REQUIRE THE REACTOR COOLANT PRESSURE BOUNDARY ~
TO HAVE A LOW PROBABILITY OF ABNORMAL LEAKAGE, l OF RAPIDLY PROPAGATING FAILURE, AND GROSS l
RUPTURE
- - 10 CFR PART 50, APPENDIX B, CRITERlON IX, " CONTROL OF SPECIAL PROCESSES"
" MEASURES SHALL BE ESTABLISHED TO ASSURE THAT... NONDESTRUCTIVE TESTING, [IS] CONTROLLED AND ACCOMPLISHED BY QUALIFIED PERSONNEL USING QUALIFIED PROCEDURES..."
^
TSs CURRENTLY REQUIRE PLUGGING / REPAIR OF SG TUBES WITH DEGRADATION OF DEPTHS >40%
~
O O O i
STEAM GENERATOR TUBE INSPECTION TECHNIQUES
- ISSUE INITIALLY IDENTIFIED DURING DISCUSSIONS ON PAST INSPECTION PRACTICES USED BY ONE UTILITY IN RESPONSE TO NRC CONCERNS NEl/EPRI PRESENTED
- RESULTS OF A SURVEY OF VENDORS AND UTILITIES A LIMITED NUMBER OF UTILITIES RESPONDED INFORMATION WAS INCOMPLETE
. NRR ATTEMPTED TO INVESTIGATE THE SCOPE OF THE ISSUE - SEVERAL LICENSEES WERE IDENTIFIED THAT CURRENTLY USED EC DEPTH SIZING TO DISPOSITION SCC /lGA SG TUBE DEGRADATION (6)
7 STEAM GENERATOR TUBE INSPECTION TECHNIQUES
- EXAMPLE: SEQUOYAH DEGRADED TUBES ALLOWED TO REMAIN IN SERVICE :
IF INDICATIONS SIZED AT < 30% TW
" DEEPER" FLAWS REINSPECTED AND PLUGGED IF CONFIRMED BY RPC PROBES
. EXAMPLE: TMI LICENSEE USES EC SIZING TECHNIQUE BASED ON INDICATION VOLTAGE, S/N RATIO, AND LENGTH RELATIONSHIP TO TS REPAIR LIMIT IS UNCLEAR i
INDUSTRY GUIDANCE ON SG TUBE EXAMINATIONS ADDRESSES DIFFICULTY WITH SIZING INSERVICE TUBE DEGRADATION (7)
o o o STEAM GENERATOR TUBE INSPECTION TECHNIQUES
. PURPOSE:
NOTIFY ADDRESSEES OF THE IMPORTANCE OF INSPECTING USING QUALIFIED TECHNIQUES l
REQUEST ADDRESSEES ~TO PROVIDE INFORMATION ON SIZING TECHNIQUES USED AND THE BASIS FOR THE ACCEPTABILITY OF THE TECHNIQUE VERIFY COMPLIANCE WITH APPENDICES A AND B OF 10 CFR PART 50 AND THE TS
. PROPOSED GL PUBLISHED IN FEDERAL REG / STER ON 12/31/96 FOR 30-DAY COMMENT PERIOD t
(8)
O O O 4
STEAM GENERATOR TUBE INSPECTION TECHNIQUES ONE COMMENT RECEIVED DURING COMMENT PERIOD NEl " PURPOSE [OF GL] CAN BE ACHIEVED MORE EFFECTIVELY THROUGH GUIDANCE BEING DEVELOPED AS PART OF NRC-INDUSTRY INTERACTIONS ON THE SG RULE..."
RULE / REG GUIDE WILL ADDRESS THIS ISSUE BUT MORE TIMELY ACTION IS APPROPRIATE ONE OTHER COMMENT SUBMITTED AFTER COMMENT PERIOD RESULTED IN A CLARIFICATION OF THE REQUESTED INFORMATION ISSUANCE OF FINAL GL IN MAY 1997 (9)
l .!
NRC STAFF PRESENTATION TO THE ACRS f
SUBJECT:
DEGRADATION OF CONTROL ROD DRIVE MECHANISM AND OTHER VESSEL CLOSURE HEAD PENETRATIONS l DATE: APRIL 16,1997 l
)
PRESENTER: C. E. (GENE) CARPENTER, JR.
LEAD PROJECT MANAGER MATERIALS & CHEMICAL ENGINEERING BRANCH DIVISION OF ENGINEERING !;
415-2169 SUBCOMMITTEE: MATERIALS AND I METALLURGY
o o o ;
. BACKGROUND
- PWSCC of Alloy 600 components in PWRs identified to the Commission as emerging technical issue in 1989
- NRC staff issued IN 90-10 to inform industry of issue Discussed alloy 600 cracking at Calvert Cliffs i
4 t
,I
i 0- o o q BACKGROUND, CONT.
e NRC staff met with WOG, CEOG and B&WOG Discussed programs for investigating PWSCC of Alloy 600 Assessed possibility of cracking of CRDMs and VHPs in respective plants
.* Cracking was found in CRDM penetration in France in 1991 France, Sweden, Switzerland, Japan and Belgium inspected
- 46 penetrations had flaws
O O O ~!
. BACKGROUND, CONT.
e Staff reviewed PWR Owners Groups' safety assessments Concluded VHP cracking not immediate safety concern provided GL 88-05 inspections performed and non- i destructive examinations performed to ensure no unexpected cracking l
NRC staff recommended implementation of enhanced !
leakage detection e !
NRC staff asked NUMARC/NEI to coordinate future VHP actions
~
~
1 0 O O
-i l
l U.S. VHP INSPECTION EXPERIENCE e Inspections for detection used eddy current, sizing by UT Methods qualified at EPRI NDE Center
- e First U.S. inspection of VHPs ,
(Spring 1994) at Point Beach i
i .
No indications were detected in 49 CRDM penetrations e Oconee (Fall 1994) revealed 20 indications in one penetration Indications were very shallow
~
O~ O O ~!
m U.S. VHP INSPECTION EXPERIENCE, CONT'D e Palisades limited examination of eight m-core instrumentation penetrations found no cracking i
l e D. C. Cook (Fall 1994) revealed three clustered indications in one penetration One indication repaired using special method e North Anna Unit 1 (Spring 1996) examined some high-stress areas on outer ring CRDM penetrations No indications obse-rved j
o o o ;;
PRESENT STATUS !
6
- NRC staff informed that Westinghouse had developed susceptibility model for Westinghouse VHPs
- NRC staff was informed that Framatome Technologies, Inc.
also developed a susceptibility model for CRDM t penetration nozzles and other VHPs in B&W reactor vessel 1 designs.
- NRC staff was informed that Combustion Engineering had performed an initial susceptibility assessment for CE
~
PWRs.
j
O O O ~l PRESENT STATUS, CONT'D
. t
- None of the PWR Owners Groups have submitted models and assessments to the NRC staff for review.
e NEI submitted white paper that reviews significance of PWSCC in PWR VHPs and describes how industry is managing issue.
Describes economic decision tool for PWR licensees to determine inspection need based on evaluation of l probability of developing VHP cracking during plant lifetime. ;
O O t t PUBLIC COMMENTS
- NRC staff received 10 sets of comments from industry, NEI and others One comment requested extension of comment period -
granted One comment questioning if GL is backfit - OGC determination is not Seven comments endorsed NEI's comments
_9_
PUBLIC COMMENTS, CONT'D e NEI's comments mostly editorial in nature Several editorial comments strengthened proposed GL Several comments point out need for requested information NEI disagreed with staff's request for information related to Zorita t
~
O O O .
PUBLIC COMMENTS, CONT'D '
e NEI submitted additional set of comments Offered no new information and still disagrees with '
proposed GL ,
Comment pointed out limitations of susceptibility model Comment identified lack of historic data Comments support need for inspections
O O O ~,
PROPOSED GENERIC LETTER l
l l l
l e NRC staff considers cracking of VHPs to be a safety l l concern for the long term based on.
i Potentially exceeding the American Society of Mechanical Engineers Code for margins if cracks are sufficiently deep and continue to propagate during. subsequent operating l' cycles, and Eliminating a layer of defense in depth for plant safety e Proposed GL is requesting information from PWR licensees to provide adequate assurance that margins and defense-in-depth are being maintained for the long term.
O O O .
PROPOSED GENERIC LETTER, CONT'D
- The information requested in items 1 and 2 of the GL needed:
To verify compliance with 10 CFR 50.55a and 10 CFR Part 50, Appendix A, GDC 14, To determine if the imposition of an augmented inspection program, pursuant to 10 CFR 50.55a(g)(6)(ii), is necessary e The information requested in item 3 of the GL relates to the
- potential for domestic resin intrusions, such as occurred at Zorita.
j
STAFF CONCLUSIONS
- NRC staff believes that an integrated, long-term program, including periodic inspections and monitormg, is necessary :
i staff needs to better understand industry program and its technical basis .
- NRC staff concluded VHP cracking not immediate safety concern in 1993, provided GL 88-05 inspections performed and non-destructive examinations performed to ensure no unexpected cracking e
NRC staff recommended implementation of enhanced generic detection j
O O O -
NRR STAFF PRESENTATION TO THE ACRS
SUBJECT:
PALISADES: Pressure Vessel Fluence Reevaluation DATE: April 16,1997 PRESENTER: Lambros Lois PRESENTER'S TITLE: Sr. Nuclear Engineer Reactor Systems Branch Division of Systems Safety and Analysis Office of Nuclear Reactor Regulation PRESENTER'S TEL. NO.: (301) 415-3233 i
i
n V
Discussion Topics 1 Background 2 Review and Analyses 3 Staff Concerns .
4 Staff Position O 5 Results O
- .- = _ _ - -._ - _-- _______
i 1
!O l
! 1. BACKGROUNQ i
i i In 1992-93 the staff reviewed l and approved a Fluence value l which allowed operation to
! 1999.
! In 1996 CPC submitted a revised l fluen e which proposed a 25%
lO
- reduction.
4 i The reduction consisted of:
j 8% due to plant data,12% due to M/C bias, and 5% due to spectral adjustments I
4
!. O
i
!O
! 2 REVIEW AND ANALYSES ~
i ,
i i Review focused on:
l' l 2.1 Plant configuration data changes i
4 i 2.2 Calculational methods '
i 1 iO i 2.3 Measurement data and l l Corrections I J
i 2.4 Spectral Adjustments i !
i \
i l i l l :
i l
iO a
l i
I
i .
)
i
)O i
l 2.1 Plant Configuration i
included:
j SIMULATE-3 pin power distribution l reduced downcomer temperature g increased vessel radius, and 1
l increased vessel thickness 1
l O
i i
lO 2.2 Calculational Methods l Calculations performed at BNL 3 using:
l ;
1 DORT P3-S 3 transport
- BUGLE-93, (ENDF/B-VI cross se tions)
O Pin wise source distribution !
l Detailed dosimeter modeling confirmed the fluence value and the 8% fluence reduction attributed to plant changes O
O 2.3 Adjustments Based on Measurement Data BE Fluence =
((M/C) + Spec Adj} x (Calculated Value;>
M/C = 12% reduction Spec. Adj = 5% reduction O
M/C data bas, e:
1
- in-vessel dosimeters
- cavity dosimeters
- high energy (E>4 MeV:l
- low energy (E <4 MeV) i
i O
Dosimeter Threshold Energy (MeV)
E > 4.0 MeV E < 4.0 MeV 5
0 1.00 0.03 0.86i0.02
,E b
0.91i0.03 0.85 0.07 Q
M/C Data Versus Location and Dosimeter Threshold Energy ;
}
}
lO d
O A
cn PLANT A 1
5 PLANT B 1.20 -
w PALISADES y O
J V g (Fe, Ni) o-g 1.10 -
k z
o p- o (Fe, Ni) p v
s 1.00 3 (Cu, Ti) 1 O
> PLANT J
O.
O T
w 0.90 -
l E w (Fe, Ni) o CC
\
D CD
<C y 0.80 -
! 2 v
- - ------r
Comparison of < M/C > Measurement-to-Calculation Bias Based on Fe-54 and Ni-58 in-Vessel Capsules for Palisades, Plant A and Plant B i
Plant Capsule M/C < M/C > o-Plant A i
(Fe-54, Ni-58)
A1 1.16 A2 1,15 A3 1.15 A4 0.99 1.113 + /-0.07 Plant B (Fe-54, Ni-58)
B1 0.86 B2 1.09 B3 1.12 B4 1.05 1.03 + /-0.10 Palisades (Fe-54, Ni-58)
A-240 0.88 W-290 0.87 W-290-9 0.84 W-110 0.86 0.863 + /-0.015 g G #
- o 2.3 Other Methodoloales -
i
{ Review of other methodologies
] indicates that no other US Vendor nor .
the French or the Germans use plant j specific surveillance data for plant l specific applications 1
i l French and German surveillance data
! are mostly consistent with their
<O j calculations.
i i
l l
4 lO 4 _
O 2.4 German Dosimetry Survey l l
A recent paper by E. Polke covering more than 20 BWRs and PWRs, reported that the M/C (about 4-8%) is independent of reactor type, dosimeter location and type of dosimeter.
O in contrast the Palisades dosimetry results depend on dosimeter type and location. Th'us, different choice of dosimeters will result in different j average values i
I l
)O
l~
!O 2.5 Spectral Adjustments
- are based on the FERRET code which has not been reviewed nor approved j by the staff.
Regardless, it would still be based on plant specific data, which do not constitute a, reliable data base. '
O :
O
O 3 STAFF CONCERNS Plant specific data are not consistent 1
No physical explanation of the deviations has been provided O Averaging does not represent the l best estimate value 1
I i
1 O
I~
!O l 4 STAFF POSITION l
l In the staff safety evaluation of the j revised fluence submittal; i
i j the reduction based on plant !
l' physical changes has been j approved :
O
- the reductions based on plant specific data averaging and spectral least square fitting have ;
been disallowed i i
O
i i
'O I 5 RESULTS
! The 1992-3 fluence value allowed
{ operation to 1999. The 8% reduction allows plant operation to 2003. ;
l I l Plant license expires in 2007, if the
! construction permit time is claimed the license will expire in 2011.
O
. L O
~*
O O O .
NRC STAFF PRESENTATION TO THE ACRS
SUBJECT:
DRAFT REGULATORY GUIDE DG-1053, " CALCULATIONAL AND DOSIMETRY METHODS FOR DETERMINING PRESSURE VESSEL NEUTRON FLUENCE" DATE: APRIL 16,1997 PRESENTER: CAROLYN J. FAIRBANKS MATERIALS ENGINEER ELECTRICAL, MATERIALS & MECHANICAL ENGINEERING BRANCH DIVISION OF ENGINEERING TECHNOLOGY OFFICE OF NUCLEAR REGULATORY RESEARCH 415-6719 SUBCOMMITTEE: MATERIALS AND METALLURGY
l . :
OUTLINE 1
1 ;
l
- BACKGROUND AND OBJECTIVES l
i i
i 1
- REGULATIONS RELATED TO DG-1053 i l \
, i i
- DEVELOPMENT OF DG-1053 :
l i
- STATUS OF DG-1053 I
- NEW SECTION AND SCHEDULE i t
l l
?
ACRS i
BACKGROUND AND OBJECTIVES
- Reactor Pressure Vessel fluence is required for determination of the vessel embrittlement and operating lifetime.
- Current methods submitted by licensees vary widely.
- Provide an acceptable state-of-the-art method for fluence determination.
Insure accurate fluence predictions and quantify uncertainty.
Reflect present day experience and surveillance report submittals.
- Standardize vessel fluence determination methodologies.
Provide consistent set of guidelines for estimating neutron fluence exposures to reactor vessels.
Eliminate unnecessary work by licensees and staff for submittal-specific reviews.
i ACRS (3) ;
~'
O O O REGULATIONS RELATED TO DG-1053
- Appendix G,10 CFR Part 50, " Fracture Toughness Requirements."
Fluence used to determine upper-shelf energy.
- 10 CFR 50.61, " Fracture Toughness Requirements for Protection .
Against Pressurized Thermal Shock."
Fluence used to determine Adjusted RT ers -- comparison to screening criteria. l 1
- Appendix H,10 CFR Part 50, " Reactor Vessel Surveillance Program Requirements."
Fluence used to evaluate surveillance data.
l i
ACRS (4)
O O O .
DEVELOPMENT OF DG-1053
- Reflects results of the LWR Pressure Vessel Surveillance Dosimetry improvement Program.
- Reflects experience in performing independent calculations of reactor vessel fluences.
- References updated cross-sections (ENDF/B-VI cross-section libraries).
- Detailed description of fluence calculation and measurement methods.
- Procedures for Qualification of calculations and measurements.
- Table of specific modeling, dosimetry, qualification and reporting requirements.
- Requires calculation of NUREG/CR-6115, " Pressure Vessel Fluence Benchmark Problems for Methods Qualification." .
ACRS (5)
O O O .
STATUS OF DG-1053
- Requested by NRR.
- DG-1025 completed.
- NRC pre-release reviews.
ACRS Subcommittee ACRS Committee 4 -
- NRC release for public comment.
- Formal review meeting with industry. ,
i
- Industr~y comments evaluated and incorporated where appropriate.
Resolution of Comments completed.
- DG-1053 available through PDR.
- Meeting with industry to review DG-1053.
ACRS (6)
O O O .
NEW SECTION FOR DG-1053
- Application of Monte Carlo transport to pressure vessel fluence benchmark problems.
- Corresponding changes to NUREG/CR-6115.
SCHEDULE
- Evaluate industry responses and incorporate changes. (June 1997)
- Completion of Monte Carlo section. (January 1998)
- Final release of Regulatory Guide. (June 1998)
ACRS (7)
O l l
Palisades Reactor Vessel Integrity Status l
l l
Presentation To:
Materials and Metallurgy Subcommittee Advisory Committee on Reactor Safeguards April 16,1997 l
l l Jack Hanson, Director - Strategic Issues l Consumers Energy l
!O
l ,
1.0 Background and Current PTS Screening Criteria Date O i.i Current giant iicense expires warch 2007.
1.2 With recovery of construction period, the plant license expires March 2011.
1.3 Current SER concludes screening criteria reached in 2003.
1.4 Consumers' submittal conclude screening criteria reached in 2012.
l l
l l
l J
i '
iO ,
i Palisades RV Integiity Status 1
1
. 1 I
2.0 Options Available to Reach EOL l
l O 2.1 Further NRC Approval of the 4/4/96 Submittal I
A. Section 3.0 provides more details.
\
2.2 Fluence Reductions via Core Management l l
l A. Much has already been achieved. (Figure 1) l B. l Shield assemblies are being considered for the next cycle, Cycle ;
L 14.
i C. Estimated cost to implement is $2-4M per cycle.
i lO r
I l
l l
l l l
t
!O Palisades RV Integrity Status 2
1 2.3 Annealing i
A.
- Q 4
The preliminary Thermal Annealing Report was completed 4/29/96 cer Draft Regulatory Guide DG-1027 (now RG 1.162).
i
- B. When the screening criteria date was extended to 2003, the ;
annealing project was put on hold. At that time detailed ~
engineering was approximately 70% complete.
C. A restart of the project in early 1999 could occur to anneal in 1
j December 2002. '
! D. Code Case N557 for RV Annealing has been approved by ASME, but has not received endorsement of the NRC.
! E. The Marble Hill Annealing Demonstration employed the same
- method and limits proposed for use at Palisades with very favorable results.
.O L
F. Estimated remaining cost to complete an anneal is $25M.
1 l
i l
3 4
j i
O Palisades RV Integrity Status 3
2.4 Supplemental Materials Testing C A. Palisades installed two new supplemental surveillance capsules at the end of Cycle 11 in 1995.
j B. These capsules contain weld materials representative of each reactor vessel beltline weld.
l C. One of these capsules will be removed at the end of the current i
cycle (Cycle 13) in May of 1998. The second will be removed at l the end of Cycle 14 in December 1999. These capsules have the
{
potential to provide plant specific credible surveillance data.-
1 D. Participant in Pressure Vessel Research Council program to help ;
j the industry develop an alternative method to directly measure '
fracture toughness. l 2.5 Regulatory Guide 1.154 Analysis l A. This does not currently appear to be a viable option.
I 2.6 Early Plant Shutdown l A. Always an option.
2.7 Preferred Option A. Further NRC approval of the 4/4/96 submittal.
B. ' Fluence reduction costs of option 2.2 will begin to accumulate l 9/1/97.
! l
, i l Palisades RV Integrity Status 4
3.0 Consumers 4/4/96 Fluence Submittal O 3.i Bacxground A. NRC Technical Evaluation of Palisades 6-5-92 PTS submittal noted fluence calculations had 10% conservatism; reference NRC 9/2/94 transmittal of the Technical Evaluation Report. Viewed as a i fall back if annealing was delayed beyond 1999.
B. Wanted to evaluate the impact of ENDF/B-VI which became available aner Palisades 6-5-92 PTS submittal.
C. More ex-vessel and in-vessel dosimetry capsules were available following Palisade.s 6-5-92 PTS submittal.
D. Determined that the grouping of several cycles in the fluence calculations had resulted in conservatism.
O e. wonte Cario caicuiations were aiso performed, by AEA, to provide an independent evaluation of the plant's fluence. The AEA results supported those provided by Westinghouse.
O l
Palisades RV Integrity Status 5
3.2 NRC SER Provided Interim Approval f .
O A. NRC performed a detailed review of transport calculations.
4
- 1. Concluded that the new transport calculations supported an 8% reduction in fluence with an associated uncertainty of 4
approximately 15%. -
i j B. Best Estimate Methodology (Figure 2) i 1. Least Squares Adjustment Procedure is used to combine calculations and measurements to produce a Best Estimate i [BE] 6f pressure vessel fluence.
- 2. Results of our Best Estimate evaluation indicate that the transport calculation should be reduced by approximately 17% or 10. (Figure 3)
O 3.
The use orBest Estimate is consistent with the ruie. west. if not all, of the data points used in developing the RG 1.99 curves used a Best Estimate fluence based on a Least Squares -
Adjustment Procedure. (Figure 4)
- 4. NRC evaluation of the Best Estimate methodology in our l 4/4/96 submittal has not been completed due to time and resource constraints.
i t
i t
1 Palisades RV Integrity Status 6
\ .
3.3 Differences O A. oirference Between eaiisades [BE]/[C) Bias and other eiants 1.
Based on an evaluation of the data from all 21 reactors ,
reevaluated using ENDF/B-VI, the set of[BE]/[C] biases average 0.95 with a standard deviation of 9%. (Figure 5) .
- 2. The [BE]/[C] bias determined for Palisades is 0.83. This differs from the data base average by slightly more than lo.
- 3. The plant specific bias should be used when an adequate number of data points are available. '
3.4 Summary A. Palisades is interested in the correct answer.
(v3 ,
B. Both 10 CFR 50.61 and Draft Regulatory Guide DG-1053 call for the use of Best Estimate [BE] rather than conservative or bounding values for pressure vessel fluence.
C. A default conservative answer, rather than a Best Estimate, places an undue burden on Palisades in the areas of fluence reductions, annealing, or plant shutdown and creates additional risks.
l
}
+
!U i
Palisades RV Integrity Status 7 l
i l
l l
l Figures l l
l i
l l
i I l l
l l
r i
d M
i 1
Palisades RV Integrity Status
A ,
i
!O t
i Figure 1 - Palisades Flux Reduction By Cycle l Representative of the best estimate fluxseen in a octant of the core.
I j
5.s 5.0 _
Cycles 3 through 7 have been
! combined for clarity.
4.5 -
1 1
l o 4.0 .
i u; O
l [ 3.5 .
bC l
! .E 3.0 _
$ 4 i o i
- W 2.5 -
l A '
Cycle 3-7 l [
i j 2.0 ,
l Cycle 1 l Cycle 2
- 1.5 -
Cycle 8 1.0 _ Cycle 10 Cycle 9 Cycle 11 0.5 I I I I t i i i 0 5 10 15 20 25 30 35 40 45 Angle in Degrees O
O O O 4
Figure 2 Accelerated A-60 Wall W-80 Reactor Vesse; 90*
Wall W-100 Core Support Banc -
Wall W-110 Core Shroud s 30- iso-
_N Ex-Vessel Removal End of Cycle 8 (09/90)
End of Cycte 9 (02/92)
End of Cycle II (05/95)
/ Axial Welds s
340* - -
-- ry v 5s-Joe-290- *=*
(08/83)(02/92) (Olns)
Wall W-280 Wall W-260 l
e
9
! Figure 3 Individual Capsule Biases for Palisades
!O i
[BE/ Calculation] Ratio for 17 Measurement Points l Comparisons Based on Flux (E > 1.0 MeV) j 1.10 r
i i
1.00 -
E
! 0.90 _
- _O g =
m .
a 8 5'0.80 e s a s ...............................................
\
W m
l . 8 0.70 -
4 -
.T t n
(
- 0*60 -
- Average
- Avg. +7%
j .
" Avg. -7%
0.50 i
t ABCDEFGHIJKLMNOPQ
! Measurement Point iO
- Palisades RV Integrity Status
O O O Figure 4 L-T 260 l' u_ 240 RT g 220
',_... ,_ .-- RT-17 2 200 ., O
,- '. 0 -""~',_,
cm ,'
S
...... R T+ 17 u 180 -
F 160 -
,e'f O- New Fluence x /
140 -
+- Old Fluence 120 -
0 2 4 6 8 Fluence (E19 n/cm^2) 4 Palisades RV Integrity Status
Figure 5 Average Biases for Individual Plants O
V
[BE/ Calculation] Ratio for 21 Plant Data Base Comparisons Based on Flux (E > 1.0 MeV) 1.20 1.10 -
s
-..--a-----
S y----------------.
g 1.00 -
8 2
l 2 e O n i a -
e .
- m e O e 0. 0 i" a .
I e ..............................W......'.........
m w
a a
0.80 -
l l
0.70 f Average
-
- Avg. +9%
Avg. -9%
0.60 ' ' ' ' ' ' ' ' ' ' ' ' '
! ABCDEFGHIJKLMNOPQRSTU Plant identification s
iO Palisades RV Integrity Status l
l
r l
O l
l l
t Attachments i Additional Information l
l l
l l
i
'(3 lV l
l
- 1 l 1 l
l l
i I
4 lO i
Palisades RV Integrity Status
~
Differences O A. Unlike the other reactors in this comparison (Figure 5), Palisades has no thermal shield.
! B. Palisades' in-vessel capsules are on the inside diameter of the vessel i
instead of the outside diameter of the thermal shield like Westinghouse plants.
C. Palisades has a demonstrated 2% bias due to feedwater fouling.
D. Palisades' core design differs from standard PWR core designs.
- 1. Cruciform control blades versus CEAs/RCCAs l
c 2. Narrow and wide water gaps C)
- 3. Very low leakage core designs !
- 4. Tini,is lower than most PWRs b)
Palisades RV Integrity Status A-1
i Neutron Flux Measurements i
i O A. A totat or l7 muitigie foii sensor sets inciudins hoth in-vessei end ex-vessel locations have been analyzed to date. As much as anyone in the
- industry.
l 1. One core support barrel mounted in-vessel capsule.
- 2. Three reactor vessel wall mounted in-vessel capsules.
- 3. Thirteen ex-vessel capsules located at the midplane of the reactor core. 4 i
B. Figure 3 shows consistent Palisades Best Estimate / Calculated [BE]/[C]
comparisons for all locations and irradiation times.
I O C .
Figure 5 shows [BE]/[C] fluence results derived for other reactors by our l
fluence vendor. l 1
l l
l l
L iO l
Palisades RV Integrity Status A-2
- . _ - -. . - . .-. -.~. - - - . .- - - .-.
DG on Fluence Determinations l
l O A. overaii oninion: we are currentiv in comgiiance with the DG.
l B. Methods used for reaction rate determination, transport calculations, and least squares adjustment evaluations are all based on consensus ASTM Standards that were developed from the NRC sponsored Light Water l Reactor Pressure Vessel Surveillance Dosimetry Improvement Program l (LWR-PV-SDIP).
C. Bench marking to plant surveillance data is allowed per the latest DG version issued for comment.
- 1. Palisades has an adequate amount of data to provide statistically ,
- meaningful results.
l -
2.
. The aRTsm predictive equation in the PTS Rule is based on ;
surveillance data from commercial power reactors; reference RG l 1.99 Revision 2, section B.
l l
l l
i Palisades RV Integrity Status A-3
1 Internal Capsules ;
A240 (30 ) Cycle 1 & 2 W290 (20 ) Cycles I through 5 ;
i W290-9 (20 ) Cycle 9 l
W110 (20") Cycles I through 10 New to 4/4/96 Submittal Cavity Capsules Cycle 8 (16 ) i Cycle 8 (26 )
Cycle 8 (39')
Cycle 8/9 (6 ) v l 1
Cycle 9 (16')
Cycle 9 (26 )
Cycle 9 (39 ) -
l Cycle '10/11 (6 ) New to 4/4/96 Submittal Cycle 10/11 (16') New to 4/4/96 Submittal
- Cycle 10/11 (26') New to 4/4/96 Submittal
. Cycle 10/11 (36') New to 4/4/96 Submittal l Cycle l0/11 (39 )
New to 4/4/96 Submittal Cycle 10/11 (24*) New to 4/4/96 Submittal l
l i
4 lO 1
Palisades RV Integrity Status A-4 ,
O O O 4 NRC STAFF PRESENTATION TO THE ACRS
SUBJECT:
NUREG-1511, SUPPLEMENT 1, " REACTOR PRESSURE VESSEL STATUS REPORT" DATE: APRIL 16,1997 PRESENTER: ANDREA D. LEE MATERIALS ENGINEER MATERIALS & CHEMICAL ENGINEERING BRANCH DIVISION OF ENGINEERING 415-2735 SUBCOMMITTEE: MATERIALS AND METALLURGY
o o o
~
OUTLINE
= INTRODUCTION 4
- GENERIC LETTER 92-01, REVISION 1, SUPPLEMENT 1
=
PRESSURIZED THERMAL SHOCK (PTS) EVALUATIONS REACTOR VESSEL INTEGRITY DATABASE (RVID)
FUTURE ACTIVITIES i
1 (2)
O O O -
I INTRODUCTION i
. NUREG 1511 -- ORIGINAL " REACTOR PRESSURE VESSEL STATUS REPORT" ISSUED DECEMBER 1994 SUMMARIZED THE RESULTS OF THE STAFF'S REVIEW OF LICENSEE RESPONSES TO GL 92-01, REVISION 1 " REACTOR VESSEL STRUCTURAL INTEGRITY" BEAVER VALLEY AND PALISADES WERE PROJECTED TO EXCEED THE PTS SCREENING CRITERIA PRIOR TO EOL ,
i - NUREG 1511, SUPPLEMENT 1 ISSUED OCTOBER 1996 DESCRIBES:
- STAFF'S BASIS FOR ISSUING SUPPLEMENT 1 TO GL 92-01, REVISION 1 AND STATUS OF LICENSEE RESPONSES TO THE SUPPLEMENT LICENSEE COMPLIANCE WITH PTS RULE STAFF'S DEVELOPMENT OF RVID !
ACRS (3)
O O O -
GENERIC LETTER 92-01, REVISION 1,. SUPPLEMENT 1 ISSUES IDENTIFIED SUBSEQUENT TO GL 92-01, REVISION'1 ISSUANCE DIFFERENT CHEMISTRIES FOR WELDS FABRICATED FROM THE SAME HEAT OF WELD WIRE LICENSEES HAD NOT CONSIDERED ALL DATA RELEVANT TO THEIR RPVs WELDS FABRICATED FROM COPPER-COATED ELECTRODES HAD LARGER THAN EXPECTED VARIABILITY IN CHEMICAL COMPOSITION RPV EMBRITTLEMENT IS SENSITIVE TO SMALL CHANGES IN CHEMICAL COMPOSITION OF BELTLINE MATERIALS GL 92-01, REVISION 1, SUPPLEMENT 1, " REACTOR VESSEL STRUCTURAL INTEGRITY" ISSUED MAY 19,1995 TO OBTAIN INFORMATION NEEDED TO ADDRESS ISSUES OUTLINED ABOVE ACRS (k
O O O
-l GENERIC LETTER 92-01, REVISION 1, SUPPLEMENT 1 (CONT.)
REQUESTED THAT ADDRESSEES IDENTIFY, COLLECT AND !
ANALYZE IMPACT OF ANY NEW DATA REGULATORY GUIDE 1.99, REVISION 2 RATIO PROCEDURE
- STATUS OF LICENSEE RESPONSES TO GL 92-01, REVISION 1, SUPPLEMENT 1 ALL LICENSEES HAVE RESPONDED; CEOG/BWROG HAVE YET TO RESPOND OWNERS GROUPS HAVE DEVELOPED PROGRAMS; FINAL RESPONSES EXPECTED BY SUMMER OF THIS YEAR LICENSEES CONCLUDED THAT DATA HAVE NO EFFECT ON PREVIOUSLY SUBMITTED RPV INTEGRITY EVALUATIONS l ELECTRIC POWER RESEARCH INSTITUTE (EPRI) DEVELOPED j RPVDATA ACRS (5)
O O .
O -l PRESSURIZED THERMAL SHOCK (PTS) EVALUATIONS SCREENING CRITERIA GIVEN IN TERMS OF A REFERENCE TEMPERARURE, RTers -
RTers = RTwortui +MTers + MARGIN SCREENING CRITERIA ARE 270 F FOR PLATES, FORGINGS AND AXIAL WELDS AND 300 F FOR CIRCUMFERENTIAL WELDS.
SUMMARY
OF STAFF'S GENERIC ASSESSMENT DURING THE PALISADES PTS EVALUATION--STAFF EVALUATED ALL PWR RPVS USING GENERIC CHEMISTRY VALUES AND INCREASED MARGIN TERMS NO UNIT WAS PREDICTED TO REACH THE SCREENING CRITERIA IN LESS THAN 7 YEARS FROM 1995 EXTREMELY CONSERVATIVE--DID NOT CONSIDER PLANT SPECIFIC DATA
\
ACRS (O)
~
O O O PLANT SPECIFIC PTS EVALUATIONS PALISADES AND CALVERT CLIFFS ACCOUNTED FOR COIL-TO-COLL COPPER VARIABILITY BY USING A WElGHTED AVERAGE !
TO DETERMINE A BEST-ESTIMATE COPPER STAFF CONCLUDED THAT THE PALISADES RPV WOULD SATISFY THE REQUIREMENTS OF THE PTS RULE UNTil LATE 1999 STAFF CONCLUDED THAT THE CALVERT CLIFFS RPV WOULD REMAIN BELOW THE SCREENING CRITERIA 20 YEARS AFTER EOL FOR UNITS 1 AND 2 GINNA PLANT SPECIFIC PTS EVALUATION SHOWED THAT THE RPV WOULD REMAIN BELOW SCREENING CRITERIA AT EOL i KEY TO REMEMBER: CONCLUSIONS ARE BASED ON AVAILABLE CHEMISTRY AND SURVEILLANCE DATA, AND COULD BE SUBJECT TO CHANGE AS NEW DATA BECOME AVAILABLE.
ACRS (7)
O O O l
l REACTOR VESSEL INTEGRITY DATABASE (RVID) l
SUMMARY
OF FEATURES DEVELOPED FOLLOWING NRC STAFF REVIEW OF RESPONSES TO GL 92-01, REVISION 1 DATA FROM VARIOUS DOCUMENTS INCLUDED USES REGULATORY GUIDE 1.99, REVISION 2 CALCULATIONAL METHODOLOGY ORIGINALLY ISSUED IN AUGUST 1995--ADMINISTRATIVE LETTER 95-03; REVISED AND RE-ISSUED ON THE WORLD WIDE WEB IN JUNE 1996--REVISION 1 TO ADMINISTRATIVE LETTER FOXPRO TO ACCESS SOFTWARE TRANSFER VIRTUALLY COMPLETE STAFF WILL UPDATE DATA AND RE-ISSUE ACCESS VERSION ON THE WORLD WIDE WEB ACRS (8)
_ - - - -.7.
FUTURE ACTIVITIES CONTINUE REVIEW OF DATA SUBMITTED IN RESPONSE TO -
GL 92-01, REVISION 1, SUPPLEMENT .1 REVIEW NEW BEST-ESTIMATE CHEMISTRY DATA TO BE SUBMITTED BY INDUSTRY'S OWNERS GROUPS--REVIEW WILL INCLUDE REASSESSMENT OF EACH AFFECTED RPV UPDATE RVID SUBSEQUENT TO COMPLETION OF ALL REVIEWS :
ONCE DISCREPANCIES BETWEEN RVID AND THE INDUSTRY ,
DATABASE ARE RESOLVED, AN NRC APPROVED DATABASE CAN BE ESTABLISHED LONG TERM GOAL: APPROVED DATABASE MAINTAINED BY THE INDUSTRY WITH NRC OVERSIGHT--NRC WOULD VERIFY UPDATES BY COMPARISONS WITH DOCKETED INFORMATION
[
ACRS (b)
0 0 0' L$
NRC STAFF PRESENTATION TO THE ACRS l
SUBJECT:
DOE REACTOR VESSEL ANNEALING PROJECT UPDATE i
DATE: APRIL 16,1997 i
PRESENTER: DEBORAH A. JACKSON MECHANICAL ENGINEER ELECTRICAL, MATERIALS, & MECHANICAL ENGINEERING BRANCH DIVISION OF ENGINEERING OFFICE OF RESEARCH 415-5887 SUBCOMMITTEE: MATERIALS AND METALLURGY i
t-
O O O- ~.
OUTLINE
-e OBJECTIVES OF THE DOE ADP e OBJECTIVES OF THE NRC
SUMMARY
OF EVENTS TO DATE e ADP PROGRAM SUCCESS
e NRC STAFF ACTIVITIES RELATED TO MARBLE HILL 2
O O O- [ ,
i OBJECTIVES OF THE DOE ANNEALING DEMONSTRATION PROJECT !
e DETERMINE RPV DIMENSIONAL STABILITY e VERIFY LACK OF DAMAGE ON KEY COMPONENTS
- ESTABLISH INSTRUMENTATION AND DATA ACQUISITION NEEDS FOR MODEL VERIFICATION t
- PROVIDE AND BENCHMARK 3-D THERMAL / STRESS MODEL e DEVELOP TECHNICAL BASIS TO RESOLVE REGULATORY CONCERNS e PROVIDE REALISTIC COST DATA FOR HEATER AND MODELING e GAIN ALARA INSIGHTS '
3
O O O- ~i t
NRC OBJECTIVES e INDEPENDENT REVIEW AND EVALUATION OF DOE ADP 1
e INDEPENDENT VALIDATION OF ADP ANALYSES t e INDEPENDENT EVALUATION OF ADP MEASUREMENTS AND INSTALLATION OF CONFIRMATORY INSTRUMENTATION l
e INDEPENDENT VALIDATION OF THERMAL AND STRESS ANALYSIS MODELS
.i t
t i
4 r
_____.____m__ _ _ _ _ _ _ _ _ _ _ __ _ _ _ . _
O O O ll
SUMMARY
OF EVENTS TO DATE .
i e DOE FUNDED TWO PROJECTS FOR THEIR ADP e MARBLE HILL AND MIDLAND WERE FUNDED BY A CONSORTIUM e NRC SIGNED AN MOU WITH DOE REGARDING NRC i PARTICIPATION e 'NRC WAS AN OBSERVER IN THE DOE ADP i t
e INDIRECT GAS FIRED AND ELECTRIC RADIANT HEATING METHODS 5
i
O O O--.
ADP PROGRAM SUCCESS ,
i e DEMONSTRATED THE VIABILITY OF ANNEALING U.S. i DESIGNED NOZZLE SUPPORTED RPVs e NO VESSEL FLAWS INTRODUCED OR ENHANCED e NO PERMANENT VESSEL DEFORMATIONS e NO EXCESSIVE STRESSES TO REACTOR COOLANT LOOPS e CONCRETE TEMPERATURES WITHIN IN ACCEPTABLE RANGES 6
O O O- . ;
STATUS OF MARBLE HILL ADP e SUCCESSFULLY COMPLETED IN JULY 1996
- INDUSTRY REPORT FUNDED BY EPRI WILL BE COMPLETED BY END OF CY97 e DOE /ASME STEERING COMMITTEE WILL AGREE ON THE '
REPORT FORMAT e FINAL REPORT WILL REFERENCE AND CAPTURE INFORMATION SO THAT THE FIRST COMMERCIAL ANNEAL WILL NOT HAVE TO RELY ON CORPORATE MEMORY i
7 i
o o o ;;
STATUS OF MIDLAND ADP ,
o COMPLETION OF THE PROJECT HAS BEEN DELAYED
- PROJECT MAY BE TERMINATED DUE TO FUNDING !
RESTRICTIONS
- PRESENTLY THE PROJECT IS IN A POSITION TO ANNEAL IF CONGRESS PROVIDES DOE FUNDING e HEATER TEST ASSESSMENT REPORT IS COMPLETED e HEATER IS DISASSEMBLED AND IN STORAGE IN RUSSIA,
. READY TO BE SHIPPED 8
O O O l:
NRC STAFF ACTIVITIES RELATED TO MARBLE HILL e VALIDATE DOE RESULTS WITH NRC CONFIRMATORY INSTRUMENTATION e NRC NUREG WILL BE COMPLETED BY THE END OF CY97 9
t
____ ______ __ ____________ -_______-__-__ _- ___ _ __________ _