ML20140E903
| ML20140E903 | |
| Person / Time | |
|---|---|
| Issue date: | 01/24/1992 |
| From: | Taylor J NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO) |
| To: | Fraley R Advisory Committee on Reactor Safeguards |
| Shared Package | |
| ML20007G200 | List: |
| References | |
| FRN-57FR47802, RULE-PR-100, RULE-PR-50, RULE-PR-52 ACRS-GENERAL, AD93-1-012, AD93-1-12, NUDOCS 9705010207 | |
| Download: ML20140E903 (1) | |
Text
{{#Wiki_filter:-- - - - .. M)3 -I UNITED STATES rp g NUCLEAR REGULATORY COMMISSION f g W ASHINGTON, C. C. 305$5 8 a 3 l January 24, 1992 1
\
i ' Raymond F. Fraley, Executive Director MEMORANDUM FOR: Advisory Comittee on Reactor Safeguards l James M. Taylor FROM: Executive Director for Operations
SUBJECT:
PROPOSED REVISIONS OF APPEN0!X A TO 10 CFR PART 100, SEISMIC AND GEOLOGIC SITING CRITERIA FOR NUCLEAR POWER PLANTS ' This is in response to your letter of DecemberQ4,1991, requesting that the pre-decisional document containing proposed revisions of Appendix A to 10 Part 100 be made publicly available as soon as possible. While we do not normally make pre-decisional material publicly available until \ after the Comission has had an opportunity to review it, weAccordingly, agree in this we case public release will make the ACRS review more effective. will place in the public document room as soon as possible a draft copy of This draft copy will be the same as that to be provided requested material.to the Comittee in preparation for the February 1992 Sub Comittee meetings. However, it should be recognized that this draft copy it, subject to change as a result of reviews by the Comittee and completion o the staff review and approval process. J s A xecutive Director for Operations cc: The Chairman Comissioner Rogers Comissioner Curtiss . Comissioner Remick Comissioner de Planque '
- SECY OGC t
a I L 9705010207 970422 oi-22-92 oi:ieru roz R
r ,, OFFICIALTRANSCRIPT OF PROCEEDINGS
~
GmGiNAL l ess -I
,. .. ...- ry+qb vP tl.
f $
~
v Y [f 8
&g f
ef p
$ U' t.
d * - 1; geng. Nuclear Regulatori ppmission
, _ Ad vi s o ry*C o mm i t t e o Reactor Safeguards' M.SJW Titlei . subcommittee Meetins on Extreme ,m,.
External Phenomena l
. s. ,
r -
, ., .Y ir
' Docket No. ,
L, '
\ (, . , _.,,- ,,
) ,
I.. . . . . Bethesda, Maryland fd I iOCARON: i ! DATE: Wednesday, February 5, 1992 PAGES: 'l - 283 . 4 !l
- l l 1 1 i l
ll E333YCe30p'-30tE!E i
' ' 'f 8" e t '*
,g **1
~0f"le Jie OT LHO LOE e.e-9
~ 1 ANN RILEY & ASSOCIATES, LTD. 1612 K St. N.W. Suite 300
) Washington, D.C 20006 (202) 293-3950
- % w e h w a;tz. 920205 -
l PDR ACRS ' i T-1889 PDR i
. . - - - . . ~ . _ . - . - _ . .
I I I ' l l PUBLIC NOTICE BY THE ! UNITED STATE NUCLEAR REGULATORY COMMISSION'S ADVISORY COMMITTEE ON REACTOR SAFEGUARDS i
)
l l l l DATE: Wednesday, February 5, 1992 i i The contents of this transcript of the proceedings of the United States Nuclear Regulatory Commission's Advisory
~
Committ'ee on Reactor Safe' guards, (date) Wedtie sday . February 5 . ~1,9 9 2 as Reported herein, are a record of the discussions recorded at the meeting held on the above date. This transcript has not been reviewed, corrected or edited, and it may contain inaccuracies. l l l l
1 _( 1 UNITED STATES OF AMERICA 2 3 NUCLEAR REGULATORY COMMISSION 4 5 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 6 Subcommittee Meeting on Extreme External Phenomena 7 8 9 10 Nuclear Regulatory Commission 11 Conference Room P-110 12 7920 Norfolk Avenue ( 13 Bethesda, Maryland l 14 15 Wednesday, February 5, 1992 16 8:30 O' clock, a.m. 17 18 The Committee met, pursuant to notice, at 8:30 19 a.m., C. Siess, presiding, where present were: 20 21 l 22 23 1 24 25 j
. .. ._ . .. . . . . ~ . - - _ - . - - .=- .- -.- .- .. . . - .
l l l 2 1 PARTICIPANTS: 2 3 C. Siess, ACRS Subcommittee Chairman 4 J. Carroll, ACRS Member 5 H. Lewis, ACRS Member 6 C. Michelson, ACRS Member 7 C. Wylie, ACRS Member 8 W. Lindblad, Invited Expert 9 D. Houston, Cognizant ACRS Staff Member 10 L. Shao, NRC/RES 11 A. Murphy, NRC/RES-12 K. Kenneally, NRC/RES ( 13 N. Chokshi, NRC/RES 14 D. McMullen, ACRS/RES 15 D. Bernreuter, LLNL 16 J. Sutton, YAEC 17 T. O'Hara, YAEC 18 .G. Bagchi, NRC/NRR 19 P. Sobel, NRC/NRR 20 R. Rathman, NRC/NRR
-21 22 l 23 l 24 l
25 l WM
.._ m _ . m.. . . . . - .. . _ . . .. __ _ . _ _ . . _ . _ _ . . _ . _ _ _ . . . . _ __ . _ _ . , _ , . _ . _ , . . _ . . _ .
L 1 I i l. 3 L 1 P'R O C E E D I N G S l l l 2 [8:30 a.m.] J 3 MR. SIESS: The meeting will come to order. This j l
'4 'is a meeting of the ACRS Subcomrittee on Extreme External 5 Phenomena. My name is Chester Siess and I'm chairman of the 1
.I
.6 Subcommittee. We have several other ACRS members in 7 attendance today.
8 On my left, Mr. Carl Michelson and then Mr. Wylie,
-9 Dr. Lewis, Mr. Carroll, and we also have, at the end of the 10 table, Mr. William Lindblad who is an invited expert 11 consultant to the Subcommittee. The purpose of the meeting i 12 is to discuss the staff's proposed position papers on 13 geological.and sejamological criteria.
(-
- 14 This involves proposed revisions to 10 CFR Parts 13 50 and Part 100, scur revised appendices and some revised 16 and new regulatory guides. Everything that we are 17- discussing today.has been placed in the public document room 18 as-of -- when? January 27.
< 19 The cognizant ACRS staff member for the meeting is 20 Mr. Dean Houston, sitting on my right. The rules for 21 participation in the meeting have been announced as part of 22 the notice published in the Federal Register on Wednesday, 23 January 22nd. 24 As usual, a transcript of the meeting is being , 25' kept and will be made available as stated in the Federal i I I
* ~
w - .y- w - w w. wy w _. --y- .__<, - --t'e y 7 *-*Tw+ f y7 m.
. - - . _ , . .. . - . . -- . - - . ~ . . . - . _ . . - . .. .. - . . - - -.
l 1 Register Notice. In order for the reporter to hear you, ! 2 it's requested that each speaker use the microphone-when he 3 speaks -- he or she si..aks -- and please give you name and 4 your af filiation when you start to talk. l 5 We've received no written comments or requests to 6 make oral statements from members of the public, although we T
.7 will have some planned discussion from non-NRC people 8 sometime this afternoon. I'm going to suggest, Andy, that
-9 you bring at least some of your people up to the table. I-10 think we've going to -- Dave Ward might show up later, so 11 you might save one seat down here, but there's at least four 12 microphones and four chairs. It makes it easier to see you l (, 13 and easier to use the mikes. i 14 Now, we have a large number of documents, all of !
15 which will be going up to the Commissioners together with a j 16 SECY sometime in the -- when are you planning to get them up l 17 there? Andy Murphy, NRC staff, Research; we
~
18 MR. MURPHY: 19 anticipate, if we have a favorable review from the Committee, the full Committee and this subcommittee, we l 20 21 anticipate that they'll be going to CRGR within the month 22 and everything goes favorably with them, within a month of l 23 that time, on to the Commission, so we're talking about two 1 i 24 months, if everything goes well from about this time to get l - 25 them to the Commission. ; I l l
.__q l 1 l
! l l t l l I 1 MR. SIESS: And then they'll go out for public l l l 2 comment? l 3 MR. MURPHY: Then probably another two to four l l 4 weeks after that, they'll be published for public comment. l 5' MR. SIESS: Now, there's a rulemaking. Has there 6 heen a notice of proposed rulemaking out; or is that 7 necessary? 8 MR. MURPHY: There has not been an advanced notice 1 9 of rulemaking and we don't believe that's necessary. l 10 MR. SIESS: Yes, okay, so this will -- when it i 11 goes out for public comment,.this will be the first chance 12 you will have, officially, to get public comment on o ( 13 everything? , i 14 MR. MURPHY: That's correct. ,
~
15 MR. SIESS: And it's your intention that the 16 proposed Reg Guides, the whole thing, go out as a package? i 17 MR. MURPHY: That's correct. . I 18 MR. SIESS: Okay, there's a revision to Part 100. i 19 MR. MURPHY: The whole package would include the 20 revision to Part 100, the Radiation Source Term, and the ! 31 geological siting criteria. At this time, it's planned to 22 go out as a single package for public comment. 23 MR. SIESS: Okay, and the geologic siting criteria 24 are all in Appendix A and B? 25 MR. MURPHY: A and B, yes. The engineering is in l i
.. - .- _ . . _ - ._ . .- . ~. . , .
I ( l( 1 Part 50. 2 MR. SIESS: Okay, and there's a revision to Part 3 50 which also, part of it, relates to the Source Term and 4 containment and part of it relates to seismic.
)
5 MR. MURPHY: Yes, that's correct. 6 MR. SIESE: Okay. Then there's an Appendix S to l 7 Part 50. 8 MR. MURPHY: Yes, which takes care of -- 9 MR. SIESS: Which takes the material that was in 10 Appendix A to Part 100?
)
11 MR. MURPHY: That's right. ) 12 MR. SIESS: Okay, I'm just trying to get the whole l( 13 picture. Now, Appendix A,.some of that is the Source Term 14 typ'e stuff and some is the seismic. The Committee 15 considered the Source Term stuff last month. In this I l 1 16 meeting, we're dealing only with the seismological part. t 17 Appendix A to Part 100 is slightly modified, ! 18 right? 19 MR. MURPHY: Appendix A itself is unmodified. It 20 will stay on the books as it exists. 21 MR. SIESS: I mean, there were no changes at all? 22 MR. MURPHY: No changes at all to Appendix A. 23 MR. SIESS: I'm mistaken then. Appendix B is new. That's correct. j 24 MR. MURPHY: f 25 MR. SIESS: And then we have four Reg Guides. l r
l l t 7 l 1 MR. MURPHY: That's right. 2 MR. SIESS: Last month, we looked at three of the ) l 3 Reg Guides, Draft Guides 1016, 1017, and 1018, and as I 4 recall, we had no serious technical issues on any of them? 5 MR. MURPHY: That's correct. There were some 6 requests for some editorial changes to all three documents. l 7 MR. SIESS: So, what I'd propose today is that you 8 update us on those changes. 9 MR. MURP3Y: Fine. l 10 MR. SIESS: At the beginning, and see if we could 11 put those out of the way. 12 MR. MURPHY: Okay, that would be our intent. 13 Roger Kenneally is prepared to make a presentation on what 14 we've changed and then, as desired, go through the 15 documents, line-by-line or page-by-page. 16 MR. SIESS: Then I would like to have the 17 Subcommittee revie~w briefly, the changes other'than the - 18 the changes to Part 50 and Part 100 which were relatively 19 brief things, and see if we have any problems with those. 20 MR. MURPHY: All right. 21 MR. SIESS: Then we'll get down to Reg Guide DG-22 1015. Rea ly, it's Appendix B and it's Appendix B of Part 23 100 and it's implementation. 24 MR. MURPHY: Fine. 25 MR. SIESS: This is 1015, right?
,._ . ., _. _ _ _ ~ . . _ _ _-._.-___. _ _ _ . _ . - _ . . _ _ _ . _ - _ _ _.. _ . _ . . _ _
]> 1 MR. MURPHY: Right. ( 3 MR. SIESS: That is roughly the order we take it 3' in. l 4 MR. MURPHY: We are prepared to address that, and 5 in that order. Does anybody on the Subcommittee have
~
6 MR. SIESS: 7 any questions about that? 8 [No response.]
'9 MR. SIESS: Once we get into the meat of 1015 10 we'll have a lot. I don't know what there is about seismic 11 that generates paper. My stack on Appendix A is just about 12 as large as my stack on the Diablo Canyon Long-Term seismic f 13 Program. And I don't know which took the longest.
( 14 [ Slide.] 15 MR. KENNEALLY: This morning what I'd like to 16 quickly do is update the Subcommittee on the changes that 17 the staff has made to the documents that they have reviewed l during their 'last meeting, and that is the meeting that took i
'18' 19 place on December 10th, 1991.
20 [ Slide.] al MR. KENNEALLY: The Subcommittee provided detailed 22 comments to the staff on the following documents. Two 23 regulations, one of them being Appendix B to 10 CFR Part 24- ~00, t and that is the citing criteria. It is designated as 25 Appendix B because the existing licensing basis will remain
. ._ j
d l I l l ( l 1 in effect, which is Appendix A. l 2 The second regulation is in Appendix S to Part 50. t ) 3 This was created by moving the' engineering aspects that were- [.
"4 in' Appendix A-to Part 100 into a new Appendices to Part.50.
5 In addition, the Subcommittee reviewed three 6 regulatory guides. Draft guide, which is what the DG stands 7 for, 1016, which is the second proposed revision to i 8 Regulatory Guide 1.12 on Seismic Instrumentation. In (. 9 addition, were two new Regulatory Guides, one basically on 3 ! 10 Plant Shutdown, which is DG 1017 and the last one, DG.1018 j 11 was the restart of a plant that had been shutdown by a 12 seismic event.. ] 13 . [ Slide.] 14 MR. KENNEALLY: In addition to'the comments that i 15 were received by this subcommittee, the staff had also
~16 benefitted from comments from the Office of General Counsel, 17 OGQ, and also the Office of Nuclear Reactor Regulation, NRR.
18 And what I'll briefly desc~ ribe to you is what has been '
! a 19 incorporated,. based on the combined input of these three 20 reviewing groups. And these were incorporated in the 21 package that had been sent-to you.
22 [ Slide.) 23 MR. KENNEALLY: With regards to the regulations, 24', we did not make any technical changes at all. We did l i i 25 numerous changes of editorial clarification or the like. i I l
i i 1 I l I l 10 1 One of the comments made was referencing a zero period , 2 acceleration made by this Subcommittee and others. We've 3 removed that from the text and now have replaced it with the 4 term of peak ground acceleration. We have not defined it. 5 It's a well-defined term. And we haven't elaborated on that 6 any further. , 1 7 A second comment was, within the text, there was a I 8 substitute iteration on seismic category one structures. It 9 appeared under the heading of the SSE, it appeared under a It also appeared in the definitions. So, 10 heading of soil. 11 we have done a nosting of that term. It has been identified 12 as Structure Systems and Components required to withstand 1
)
13 the effects of a safe shutdown earthqua' ground. motion and 4 ( 14 surface defamation. And a new definition has been added to 1 1 the test. And the repetitiveness has been repeated -- j 15 l 16 removed from two places in the regulation. 17 Also, with regard to the operating basis -
~18 earthquake, it has been clarified by indicating that it also 19 is a ground motion, not just an earthquake. And that has ;
20 been added to the text. 21 The documents that were distributed to the 22 Subcommittee for their review, and also the same that was I I l 23 put into the PDR have not have the benefit of the technical 24 editor yet. 25 [ Slide.)
l l 11 I I" 1 MR. KENNEALLY: Technical changes to the-1017, 3 Regulatory Guides were done principally in draft Guide 3 The staff -- the initial document reviewed by the 4 Subcommittee had taken an exception to a definition of felt 5 earthquake that was introduced in an EPRI report, NP-6695. The staff has removed that exception. It is now going with 6 the definition that was within the EPRI report. Basically, 7 8 it is saying that the operators will take action if there is 9 an activation of the seismic instrumentation or a consensus 10 of the control room operators. 11 Initially the staff had some concern about the
'12 operability of the instrumentation. It didn't want to see a 13 lack of attention by the operators because no
~( 14 instrumentation wa's activated. But, with careful reading of 15 the definition, it appears that it says for operable 16 instrumentation. And so we feel very comfortable with the 17 "or" in there and we haven't compromised .any of our
^
18 positions. 19 MR. LEWIS: As long as you're trying to be very precise about this, what's a consensus? ao 21 MR. KENNEALLY: They had indicated -- probably a 22 majority. The report didn't go on and identify it a lot. l l 23 MR. LEWIS: What they meant doesn't matter -- what i 24 you're writing -- what do you mean? MR. KENNEALLY: Yes. We would accept a majority 25 l a i
8 l l l 1 of the control room operators. 2 MR. LEWIS: You will accept? You're not being 3 very precise. 4 MR. KENNEALLY: That is true. 5 MR. LEWIS: Because you might not be around when 6 it's interpreted. l 7 MR. KENNEALLY: I understand. I probably would 8 not be the individual who would have to follow the 1 9 interpretation either. 1 10 MR. LEWIS: Why not make it precise? 11 MR. KENNEALLY: That's a good point. 12 The other technical change that the staff had (, 13 added was that after an earthquake, the licensee should use 14 a standard calibration that is used during the maintenance 15 of the instrumentation or an actual time history to redo the 16 spectra and calculate the cumulative absolute velocity. 17 Those are the two threshold criteria that I used to 18 determine if an OBE had been exceeded, and assure that the 19 instrumentation and the processing is providing the same 20 number that was used initially. We could then evaluate if 21 any drift or any changes had taken place. 22 This was common practice. Talking with the people 23 that have implemented this type of instrumentation, they are l 24 doing. And it is a little bit more of a clarification. l 25 Within the Regulatory Guides, the three of them, l
l h 13 l l 1 we have removed a lot of the commentary that appeared in the 3 regulatory position. It was pointed out correctly that it 3 was a very good explanation of what we were trying to i 4 accomplish, but it was not a position that in the past had l 5 gone into the discussion section, and we agreed and have
- 6 tried to catch most of that.
i 7 We have also revised the discussion section to l 8 provide commentary on the regulatory position. Again, the 9 three guides have not had the benefit of a technical 10 editing, as yet. 11 That is a quick summary of what we have done since l 12 the last subcommittee meeting. If there are any detailed ( 13 comments on those, we would be glad to , address them, your 14 review of them. It would probably be a little beneficial 15 for me if, with the Chairman's position, if I sat at the 16 table so I could have the documents'available. 17 MR. SIESS: Okay. Sit down. And I have a few
~
'18 minor things, questions an'd comments, and I'll start it off 19 and then let anybody else that has any chime in, and I'll 30 start with 1610, which is the first one.
31 This is a question on Page 6. Incidentally, I l , 22 think all of them have been greatly improved by editing. 4 23 They look more like Reg. Guides now. 24 In 1016 on Page 6, when it says the 4 . 25 instrumentation should have the ability to record the five j l
I l l I 14 l 1 seconds prior to seismic figure actuation, is that what's 2 done now? i 3 ,R. M KENNEALLY: Yes. 4 MR. SIESS: 'You have a built-in memory that 5 retains it five seconds running? MR. KENNEALLY: That's correct. That is.in the 6 7 existing instrumentation capability. l MR. SIESS: That's the only question I had on 8 l 1 Does anybody else have any questions about that? 9 that.
-10 [No response.]
11 MR. SIESS: b et of this is not new, am I right? j l That's correct. Moving of f 12 MR. KENNEALLY: ! 13 material, as I.had indicated. ' i(} - In 1017, again, I think it's 14 MR. SIESS:' Okay. 15 considerably improved. It sure reads a lot easier. 16 On Page 3 -- I'm willing to hand you my copy with 17 some editing. I don't know how good your editors are. But i8 "two criterion doesn't sound quite 'right to me. MR. KENNEALLY: Yes. 19 20 MR. SIESS: And the sentence that precedes that I 21 think is missing, I think your editor just missed it. I 22 have read it to say that shutdown of a nuclear power plant 23 would be required if the vibratory ground motion experienced
'34 exceeds that of the OBE ground motion.
25 I think'I could -- I read it three times to get l I
__._.,_..__.-____._____._-_.._____.__...._.__..__...__.,7 , I l l
- i l
15 1 that sense out of it. But except for that, I have no l 2 problem with that one. 3 I have one question, though. On Page 8, top of 4 the-page, 5.2, damage: to be shutdown if the walkdown 5 inspections discovery damage. Where is " damage" defined? 6 MR. KENNEALLY: Within the report, the EPRI NP-7 6695, they have identified a baselining of structure, 8 systems, components that are walked routinely during shift 9 operators. Their status is recorded, photographed, or 10 whatever. If they happen to be cracked in concrete already, 11 or anything that looked a little abnormal, it is noted, so l 12 that on a post-earthquake evaluation, that would be removed l ( 13 as n' potential earthquake initiator. 14 MR._SIESS: 'The question is, what is " damage"? 15 MR. KENNEALLY: Damage is also defined within the 16 context of the report, in their definitions and -- 17 MR. SIESS: This particular place references 18 'Section 432'of EPRI'NP-6695. Does that section deal with l
)
19 the walkdown inspection, or is that where damage is defined? 20 MR. KENNEALLY: That deals with the walkdown 21 inspections and refers to the 22- MR. SIESS: Could you put a reference in here? i 23 MR. KENNEALLY: Yes. 24 MR. SIESS: Because as I recall, the EPRI
- i. 25- documents got a very detailed description of damage by e
l
1 l s I l 16 1 categories. 2 MR. KENNEALLY: Yes. 3 MR. SIESS: That would indicate some agreement on j 4- what' damage is. I certainly think it's desirable to know i 5 what it looked like before. Otherwise, every crack in the i 6 concrete is going to be blamed on the earthquake. < i 7 Does anybody else have anything?' 8 MR. MICHELSON: It's just new damage, then, that 9 you're concerned with, not old damage. Right, Chet? 10 MR. SIESS: W, ell, damage as,a result of the ; 11 earthquake. I assume that if they do regular walkdowns, 12 that damage that would be significant to health and safety (N 13 of the public would have been fixed before the earthquake. . , l i 14 MR. CARROLL: Out of curiosity, what's the+ 15 industry practice in terms of keeping this baseline for 16 inspection up to date? MR. KENNEALLY: It has not been employed yet. The 17 18 documents were written in 1989. It is proposed as the 19 method to do. Some plants are starting to do it on a 30 volunteer basis as far as the instrumentation aspect goes. 21 MR. CARROLL: How about doing such things as 22 documenting cracks, and things like that? I 33 MR. KENNEALLY: The people that worked with NUMARC j 24 and EPRI in the development of the report benefitted from a l 25 lot of utility people, and there was a consensus on the l t I 1 l
. . . - . - . . - . . - - _ ~ . - _ _ . . - . . . . _ . . - . - . . . . -
l 17 [ 1 group.that developed a report that this approach would go a forward. And it doesn't appear to be a problem on l 3 implementation of it. To date, it is not actually in 4 existence. 5 MR. CARROLL: All right. 6 MR. SIESS: I think it's going to be difficult to ! 7 keep. people interested in looking for cracks before the 8 earthquake, when they probably never have had an earthquake 9 there, and after a few years, they haven't had one. j
'10 MR. CARROLL: I'll also bet that after the 11 earthquake, they will wish that they had done a better job.
12 MR. SIESS: You know that's true. And after the ( 13 earthquake, if somebody gets into. trouble, everybody else
)
~
14 will do a better job for a while. But it's really this l : ! 15 centingency-type thing. It's a very low probability event 16 that wa're talking about. It's going to be tough. l 17 It is not something I particularly worry about , l . 18 from our. point of view because if there was damage done by any definition whether it was there before or after, it l 19 20 doesn't make much difference to me.
' al The plant's going to get shut down until somebody l
32 decides how important it is. 33 The balance is going to be toward shut the plant 24 down first and think about it later, rather than argue about , 25 whether it should be shut down.
1 18 ( 1~ MR. MICHELSON: I have a general question about 2 these walk-down inspections. 3 Concrete structures oftentimes require seals to i 4 keep groundwater from getting into the interior of the structure or to keep floods out and so forth. How can you 5 6 tell the condition of one of these seals after and OBE, to 7 know whether it still would perform its desired function? l 8 Sometimes these things are almost impossible to 9 even view, let alone speculate as to whether they are still 10 good. I 11 It's not easy. l[ just wonder how do you know that 12 -- and some of these are very essential for certain types of l l 13 events such as floods. l( You might know after the nexh rain. 14 MR. SIESS: 15 MR. MICHELSON: Yes, you might know for such j 16 reasons but walk-down wouldn't help you much. MR. KENNEALLY: The walk-down wouldn't help you
- 17 I
18 much. As Prdfessor Siess pointed out, it would be another 19 indication that flooding or seepage had gotten into an area 4 20 that hadn't occurred before that you would have an i 21 indication.a 22 MR. SIESS: No, but if you thought about this a i 23 little bit, thought about seals, you would look for faulting 24 across a joint. If you looked at a wall and you had a construction joint there and it moved a quarter of an inch 2 25 4 4 i
19 I 1 one_ side,.the'next question would be is.there a seal in 3' there? Did it tear? I think if seals were listed,.you 3 know, as a type of thing people ought to be looking for, 4' there are ways of looking.
'S MR. MICHELSON: Do you measure the construction 6 gaps during a walk-down, for instance? That would be what 7 you would have to do, unless your eyeball is well-8- calibrated. You'd have to go measure the spacing if a joint-9 had shifted.
10 MR. SIESS:- Of course a joint's designed to -- the 11 seal should be designed to take some movement -- MR. MICHELSON:- It should, that's right. 12 A MR. SIESS: -- in that direction. Whether it'll ( 13 . N take faulting movement -- if it's an elastomer it may not. 14 15 MR. MICHELSON: But you can measure that too. That's obvious, you see. That's the 16 MR. SIESS: 17 one that will stand out but I think joints are things that, major joints where.there is not a structure or support are 18 19 the kind of thing I'd put on a walk-down. 30 It's the things -- if the earthquake's going to do 21 that you don't notice till the next time you need it. MR. MICHELSON: Is there actually a tabulation of 22 23 these structural joints so that the people that do walk-l- 24 downs ar'e aware that there is a watertight seal required at i- .35- that joint and therefore look for it? l 1
.u
l 20 ( 1 Even tabulating these so you look to see if 2 there's anything. unusual around that joint would.be helpful, i i l 3 I'm not sure those things are logged in. You l 4 know, I'm not sure an operator knows where all those are. l He doesn't pay much attention to structures. He likes pumps 5 6 and valves and pipes and so forth but he doesn't pay much 7 attention to seals that he can't even see in some cases but 8 the fact -- at least as a minimum there should be a log of 9 such seals so that after an OBE one at least tries to-10 determine if anything looks unusual. 11 MR. SIESS: Paint it on the wall. 12 MR. MICHELSON: But my experience has been people 13- arel.'t often aware of these until water starts coming in and l(- t
.then they check around and find out, yes, there's something 14 15 there that is not working anymore.
MR. SIESS: Okay, 1018. It sure got a lot 16 17 shorter. I guess my theory is anything that's shorter's 18 improved! No, I had no question on 1018. This is the plant 19 20 restart because it refers to the EPRI document which has a 21 great deal of detail on this. 22 MR. KENNEALLY: Yes. 23 MR. SIESS: What wasn't clear to me the first time I 24 I read it and it is not clear here either is what is not 25 endorsed in the EPRI document, but once I looked into the
i F i ! 21 . h 1- ' document I found that everything related to plant restart 2 was endorsed j 3 MR. KENNEALLY; That's: correct. 4 MR. SIESS: And the stuff that wasn't endorsed i 5 just wasn't' pertinent to this. 1 I 6 I guess it would have been better if'EPRI had put L 7 all that in one chapter or something. 8 Other comments on lois? Yes, Bill? ! l 9 MR. LINDBLAD: Could I go back to seismic 10 instrumentation, just for a moment? 11 MR. SIESS: Sure. Nothing's ever settled. ; 13 MR. LINDBLAD: Thank you. The operability of seismic instrumentation, it seems to me that.the key issue l( i.-- 13 . 14 is ground motion as distinguished from plant response. 15 It speaks of the maximum number of instruments being always available, both operating and shutdown. How do
~ 16 17 you picture this going into the tech spec? :
18 MR. KENNEALLY: That particular part of the , 19 instrumentation I believe would probably not even appear in 20 the tech spec. There will be -- MR. LINDBLAD: It will not be considered in LCO, a 31 22 limiting condition -- 23 MR. KENNEALLY: No, because some of that will come l 24 down just for maintenance in what may not be operable during 25 that particular time but the message the Staff is trying to 1 L . - . _ - -__
l l l 22 1 1 say is it probably would not be wise to have all your . 2 foundation level instruments and your free field instruments 3 out at the same time. 4 MR. CARROLL: That's what Item 3 on page 5 is 5 telling us. 6 MR. KENNEALLY: That is correct. l I 7 MR. SIESS: At one point you had something a lot i l 2 8 stronger. It sounded like you were saying that 1 9 instrumentation had to be operable always and the t l 10 implication was that there was going to be an LCO if it ! 11 wasn't operable but you were. going to shut down. f 12 Now those words got changed somewhere but there is t l( 13 a reference -- if you don't have the instrumentation working , 1 14 there was an alternative on deciding when to shut down. ] 15 MR. KENNEALLY: The Staff has developed an ) 16 appendix in 1017. 17 MR .- SIESS: That's in 1017? I remember it. I 18' just want to check. l 19 MR. KENNEALLY: Page 10. 30 MR. SIESS: Page 10. So you are encouraging i people to have it operating and this provision in here whicn al 22 --- what happens if you don't have it operating -- is a l l l 33 little further incentive to having it operating? l .i 24 MR. KENNEALLY: That's correct. 1 i 25 MR. SIESS: But it won't be an LCO and it won't be l 1 l
23 1 in the tech specs? i 2 MR. KENNEALLY: That's correct. 3 MR. LINDBLAD: Thank you. ; 4 MR. SIESS: Unfortunately, we have had a couple of l 5 nice' earthquakes when they weren't operating, not real nice 6 ones. 7 MR. MICHELSON: Let me ask you -- because 8 something doesn't appear in a tech spec doesn't mean that 9 there can't still be-a requirement specified elsewhere on 10 it. You've been doing that in the case of fire protection. 1 11 You're pulling the stuff out.of the tech specs and putting ' 12 them back in an'FSAR and so forth. But that still is a ( 13 requirement to do -certain things at certain times even 14 though' it's not in the tech spec. Is that comparable thing 15 being done in this case? 16 MR. KENNEALLY: I think the comparable thing there 17 would be the appendix that we just talked to. The position 18 the staff would have if the instrumentation were inoperable 19 is a tendency to be more anoservaite than what would happen 20 if the instrument still were operable. 21 MR. MICHELSON: This, of course, is a regulatory 22 guide, and everybody says regulatory guides are not 23 requirements. FSAR commitments are, and so in the case of 24 fire protection, they just moved them back, I think, to the 25 FSAR, but to get it out of the tech specs to keep from i
l l 24 [ 1 cluttering them up, but still making sure that certain 2 things are taken care of just as an FSAR requirement. i I 3 MR. CHOKSHI: There is the SRP Section 374, which 4 is specifically about seismic instrumentation, including l 5 maintenance surveillance, and that will still be required. 1 6 MR. MICHELSON: Of course, SRP is not a 7 requirement, either. 8 MR. CHOKSHI: Right. No. But the FSAR addresses 9 each section of the SRP. 10 MR. MICHELSON: Yes. A commitment to that in the 11 FSAR would do it, right, and that's what I was wondering l 12 here. I think you don't gain anything by falling back to ( 13 the regulatory guide, , but you would by falling back to an 14 FSAR amendment which made it a commitment, and that's the 15 way I would handle it. I wouldn't trust some other guy to 16 do it. 17 MR. KENNEALLY: I believe that an applicant would be citing the appropriate regulatory guide in their SAR 18 i 19 submittal. 20 MR. MICHELSON: Yes. Then it would become 21 binding. 22 MR. SIESS: Let's see. Suppose you don't have an 23 FSAR? Do we have an FSAR under Part 52? 24 MR. CHOKSHI: Yes. l 25 MR. SIESS: Okay. Okay. i
I l I 25 6 1 MR. CARROLL: That raises the question, how do 2- these requirements get backfitted. 3 MR. KENNEALLY: They do not. 4 MR. SIESS: These do not. These three reg guide 5 relate to Part 100, Appendix B. 6 MR. MURPHY: No. These three reg guides goes with 7 Part 50, Appendix S. MR. SIESS: Okay. Part 50, Appendix S. All
-8 9 right. This is strictly forward fit.
10 MR. MURPHY: That's correct. 11 MR. SIESS: See, the attempt here is at the 12 existing plants, nothing is going to change. 13 MR. BAGCHI: Can I address that, Mr. Chairman? ( . 14 MR. SIESS: Yes. 15 MR. BAGCHI: My name is Goutam Bagchi. I'm with 16 NRR. We are currently reviewing those two reports on the 17 request from NUMARC, and we are processing a -- proposing a 18 generic letter so that it would be available to the existing 19 plants, and we're encouraging people to use the new 20 instrumentation, but it's just an encouragement. l 21 MR. CARROLL: But it would strictly be voluntary. 1 22 MR. BAGCHI: Strictly voluntary. 23 MR. CARROLL: Incentive for a licensee to use it. l I 24 MR. BAGCHI: The only incentive would be that if 25 you didn't have the instrumentation, you are under stricter I i l l
i 26 1 requirements for shutdowi following OBE. 2 MR. SIESS: Where is the instrumentation described L 3 now? Do we have a -- 4 MR. BAGCHI: For existing plants? 5 MR. SIESS: Existing plants. 6 MR. BAGCHI: It's in the FSAR. j 7 MR. SIESS: It's always in the FSAR? There was i 8 never a reg ' guide or anything? 9 MR. KENNEALLY: Oh. Regulatory Guide 1.12. l 10 MR. BAGCHI: 1.12.
)
i 11 MR. SIESS: Okay. That's what I was asking. 12 MR. KENNEALLY: Revision 1, and there was a safety (s. 13 Guide 12. 4 14 MR. SIESS: Okay. 15 MR. MICHELSON: Have there been tech spec 16 requirements in existing plants? 17 MR. BAGCHI: Tech spec requirement is rather weak 18 for existing plants. It's only for maintenance. 19 MR. MICHELSON: Are you going to leave it there if 20 you adopt this alternate approach, or are you going to move al it back to -- 22 MR. BAGCHI: I don't think we're -- l l 23- MR. MICHELSON: Part of what you do, you've been l 24 moving som' of the other materials out of the tech spec. 25 MR. SIESS: It seems to me that's a great l l l
I 27 l( l 1 candidate to get out of the tech specs because -- 2 MR. BAGCHI: Tech specs are being revised. People 3 are adopting standard tech specs. Most of the time, they 4 are taking things out of it. I suspect that it'll be out of 5 it. ) 6 MR. MICHELSON: But when they take it out, it has 7 to go somewhere else if it's important. 8 MR. LINDBLAD: Into an administrative procedure. 9 MR. MICHELSON: Okay. l 10 MR. SIESS: All right. Now, I think that disposes 11 of these three reg guides which we could then -- my la inclination would be to tell the staff that those, as far as 13 we're concerned, are to go to the Commission in a CRGR, l( 14 which is their business. 'We're willing to see them go out 15 for public comment. 16 Now, two of these guides are based heavily on an 17 EPRI document.
^
18 MR. KENNEALLY: That's correct. 19 MR. SIESS: Has EPRI seen them? 20 MR. KENNEALLY: Just recently, when we put the 21 package into the public document room. 22 MR. SIESS: Now, of course, the point in that 23 would be to be sure that you haven't misinterpreted the 24 document and referenced it improperly, but then you worked i 25 on those EPRI reports, too. l l l
l I I i 1 28 l 1 MR. KENNEALLY: Yes, I did. l 2 MR. SIESS: So-there's no hidden meaning in there 3 that you wouldn't know about? 4 MR. KENNEALLY: Not to my knowledge, and I -- they 5 will have the advantage, of course, during the public 6 comment period to cite any discrepancy I might have seen or I 7 taken. 8 MR. SIESS: Okay. Well, let's go next -- 9 MR. CARROLL: Excuse me. 10 MR. SIESS: Yes. 11 MR. CARROLL: Just.for my edification, who 12 contributed to this effort? 13 MR. KENNEALLY: NPR in Washington was the major ~(.. . 14 contractor to it. Drs. Stevenson and Kennedy heavily l 15 contributed to the authorship. EPRI had set up a panel to 16 review it which consisted of NRO staff and numerous people 17- form utilities because it is really a systems approach to
'18 looking'at the problem, and they provided an extensive peer 19 review that lasted over nine months.
20 MR. CHOKSHI: At least two to three SROs were on al that panel, and that input was taken, that, you know, how to 32 determine when to shut down and when not to shut down, what 23 we should be looking at. So this was a very integrated l 24 panel. They also had regulators, a couple of people with 25 that perspective on the panel. I
i 29 1 MR. CARROLL: Okay. All right. 2 MR. SIESS: The first page of the table of 3 contents is missing. I hope this isn't on where I got ever 4 other page. I thought it had the list of the panel but it 5 doesn't. 6 MR. KENNEALLY: It should. 7 MR. SIESS: Andy, your presentation deals with 8 basically Appendix A, right? 9 MR. KENNEALLY: It's the proprietary guide in 10 support of -- 11 MR. SIESS: Before you do that, I would like to be 12 sure we've cleaned up the other material that's involved in ( 13 the changes. That's Part 50, Appendix S, which we looked at 14 last time. 15 I had a comment last time that it needs. technical 16 editing and it's obviously gotten some but I think it still
- "17 needs.some technical editing.
18 MR. MURPHY: As.I understand it, 'most of the 19 documents have gotten a one-through by the technical editing 30 staff in Research. They are not finished with their work on al the whole package so your observation is correct. There has 32 been some changes but it's not complete at the moment. l 23 MR. SIESS: Have you got it in front of you? 34 Let's just glance at it and you can make and notice the i l, 25 kinds of things and see that they get them.
., . - - - . - - . - _ - . . - . - . - . . . . . _ _ - ~ - - . . ~ - . - . ~ . . . . . -
1 ,, 30 1
- l 1 MR. MURPHY
- Okay.
I 2 MR. SIESS: They are not issues as far as I am ! 3 concerned but they could be if I don't understand them l 4 right. l 5 This'is a draft, 10 CFR, Part 50, Appendix S. . , 6 MR. CARROLL: I have both comparative text and
? reduced text. j 8- MR. SIESS: Use the reduced text. The comparative ,
9 has got all the stuff that was thrown out and put somewhere j 10 91se but it's the one labelled Reduced Text and it is dated . 11 January 21, 1992 at the bottom of each page except the first 12 one, the cover page. l ( 13 On page 3, on lines 24 through 27, safe shutdown. 14 earthquake' ground motion, what we are talking about is
- 15. Appendix to Part 50, which deals with plant design, right? !
16 MR. MURPHY: Right. 17 MR. SIESS: That paragraph says " Required safety 18 functions of structures, syste'ms, and componerits shall be 19 assured during and after the vibratory ground motion 20 associated with the safe shutdown earthquake ground motion al for suitable analyses, testing, or qualification methods." 22 Is that clear enough language for a rule? I 33 " Suitable?" l 34 MR. CHOKSHI: I think this is sort of to give some [ I i , 25 leeway. The original, if I remember correctly, was very l L . .- _.
i l' 31 I , 1 specific -- dynamic analyses, qualification by tests, and .i I :2 words was leaving out a lot of other options which one can 3 use. 4 MR. SIESS: The thing is the safety function of J 5 nothing is assured by a calculation. I mean it would-be , 6 nice if it was, you know. It would make life easier for a ; 7 lot of people if all we had to do was calculate-something 8 and that assured safety. 9 MR. CHOKSHI: Analysis and design or something i 10 like that. i 11 MR. SIESS: I mean I can assure safety by design T I 12 without an analysis if I'm smart. It's just that " Safety ! 13 functions shall be assured through" -- you know, this is an 4( 4 ' 14 important statement because the safety function is 15 everything. 16 MR. MURPHY: Right. 17 MR. SIESS: You might think about that one. l 18 'MR. MURPHY: Yes. 19' MR. SIESS: In the next paragraph it says it's 20 " permissible to design for strain limits in excess of yield 21 strain in some of the safety-related structures."
' 22 That's not much help to anybody. Which "some?"
33 MR. MURPHY: Okay.
- 24 MR. SIESS
- Meaning you say all or you say none I 25 know what you mean but if you want to say it may be I
. ,. . _ _ _ l
l 32 i If ; 1 acceptable under certain conditions , ich we will define i 2 later. t 3 MR. MURPHY: Okay; specificatis1. 4 MR. SIESS: What did we say befers about this? 5 MR. MICHELSON: The same. 6 MR. SIESS: It's in the old Appendix A. 7 Anyway, make a note of that. 8 MR. KENNEALLY: Those are the words. 9 M R '. MICHSLSON: Same words. 10 MR. KENNEALLY: Those are the same words. 11 MR. SIESS: Thar. La line 32 it says "under the 12 ' postulated concurrent conditions" -- now I read that to mean ( 13 loading so are there any other conditions other than loadings that you are thinking about there? 14 15 THE REPORTER: Dr. Chokshi, can you use the m.'ke? 16 Thank you. 17 MR. SIESS: You're saying inelastic behavior can
'18 be considered but be sure you consider other things besides 19 the earthquake.
20 MR. CHOKSHI: Loads. 21 MR. SIESS: And if it is only load, I mean since 32 you are talking strain, I think it's load. 23 MR. MICHELSON: Loads. i 24 MR. SIESS: Might be temperature but again
' 25 " conditions" is a loose word.
l l
j 33 1 MR. MICHELSON: It's more than the earthquake 2 loads alone. 3 MR. SIESS: That is what it is saying -- other 4 things besides the earthquake but it is not just -- I think 5 they are thinking loads, accident loads or whatever. 6 At the top of Page 4, there's a reference to i 7 applicable stress and defamation limits. Now, again, we're 8 in a rule. Is that legal? I assume it's code stresses. Is 9 this the kind of reference you make where the code stresses 10 would apply? 11 MR. CHOKSHI: That's the intent. 12 MR. SIESS: Okay. We might check with the lawyers (, .. 13 on that. 14 MR. KENNEALLY: That'was in the. existing 15 regulation, of course. O,C has not reviewed it in detail 16 and won't comment. 17 MR. SIESS: Line three says the operating basis 18 earthquake ground motion shall be defined by response 19 spectra. I don't think that's right, because the response 20 spectra doesn't define a ground motion. 21 MR. CHOKSHI: It's the response to the ground 22 motion. 23 MR. SIESS: But you're trying to define the ground l 24 motion. 25 MR. CHOKSHI: Right. i i
l 34 l( 1 MR. SIESS: Usually a ground motion is defined by 2 a motion versus time. 3 MR. KENNEALLY: Right. 4 MR. SIESS: And you're saying it's defined by a , l 5 response. I know what you mean, but it ain't right. 6 MR. MURPHY: " Defined" is the wrong word. 7 MR. CHOKSHI: It can be characterized by response. 8 It's a unique characterization. 9 MR. SIESS: You need to think about it a little ] 10 bit. j 11 MR. CHOKSHI: Yes. , I 12 MR. SIESS: This is for design purposes. I 13 MR. MURPHY: Certainly. . I 14 MR. SIESS: Now, when you get into the 15 definitions, you don't say spectra. You say it should be 16 defined by spectra, but then one-third of the safe shutdown 17 earthquake ground motion level, now, that's not a spectra; 18 that's one-third of a level. Okay? 19 MR. MURPHY: Right, to the discrepancy. 20 MR. SIESS: In Line 21, Item 4, required seismic 21 instrumentation, suitable instrumentation, et cetera, so 22 that thee seismic response of nuclear power plant features 23 important to safety can be evaluated promptly, i l 24 I'm sure you mean promptly following an 25 earthquake. But I would think it needs to say it. This is I i
( 35 1 a rule. 2 Under the next item, surface deformation:< in the 3 event of such deformation, certain structures will remain 4 functional. 5 The lawyers aren't going to stand for that, are 6 they? 7 MR. KENNEALLY: I believe that was part of the 8 definition,we've taken care of on Page 2-B. 9 MR. SIESS: I think so. 10 MR. KENNEALLY: Perhaps we need a parenthetical , 11 reference. ;
)
12 MR. SIESS: Well, that's the kind of stuff I .( 13 caught in there. I think.it's more. legal -- 14 MR. LINDBLAD: Mr. Chairman'? 15 MR. SIESS: Yes, Bill. 16 MR. LINDBLAD: On surface deformation, is the 17 concept there that it must be permanent? Permanent seems to 18 he absent from the definition. But I think certainly 19 vibratory ground motion gives surface displacement and ; 20 deformation. But that is not what you mean. You don't mean 21 the temporal, but you mean the permanent. 22 MR. MURPHY: Permanent. 23 MR. LINDBLAD: Is there some reason why permanent j 24 is not included there? 25 MR.. MURPHY: Principally a matter of we thought it l
._ _ _.. . . _ _ _ . . _ . _ . m_. . . _ _ . . . .. . _ . __ ,
' ^ l l ., b L r > 36 1 was understood. But that would be the only reason.
-3 MR. SIESS: That's covered in 10lb?-
l l L 3- MR. MURPHY: Yes. ! 4 MR. CARROLL: There's a definition'of it on Page 1 5 2.
- 1
- 6 '
MR. SIESS: Of surface deformation? ) 1 7 MR. MURPHY: It doesn't say_ permanent, either. j 8 But th'ose things are viewed as permanent. They're not the 9 transitory dynamic aspects. l 10 MR. LINDBLAD: Well, I understand that. But will j
- 11. the reader of the regulation? 1 la MR. MURPHY: That's a valid point. !
( 13 MR. SIESS: Static wouldn't do it, because it 14 wouldn't be ' static while it was happening, it'would be I 15- static -- I'm thinking of static as opposed to dynamic. ! l I 16 MR. MURPHY: Permanent is better. 17 MR. SIESS: Permanent is not the best word,
- 1
* ' 18 either, because that' implies' forever.
- Semi-permanent.
1 19 (Laughter.] I 20 MR. SIESS: You get the point. 31 MR. MURPHY: Yes. 33 MR. SIESS: I am going back to the OBE again, on 23' Page 4 at the top. In Item F it says one-third of the SSE. , 24 The requirements associated with the OBE can be satisfied l 35 without the applicant performing the explicit response or l l
-*. a w y - - - - -w'* *-W=
. .- ~. . - - - _ - _ - - - . - . .
37 1 design analysis. 2 What requirements? 3 MR. MURPHY: You lost me for a second. Where were ; i 4 you referring to? 5 MR. SIESS: Top of Page 4. It's Line 7. It says 6 "the requirements." j 7 MR. CHOKSHI: I think the requirement was implied 8 by putting applicable stress and deformetion limits, that we
- 9. don't now have to show by doing calculations or design 10 checks.
11 The definition of OBE refers to the components 12 shall remain functional and within applicable stress and '( 13 deformation limits. . 14 ' MR. SIESS: Is that the only pos'sible 15 interpretNtion? 16 MR. CHOKSHI: I'm not sure, but I think that was 17 the intent, is that'you don't have to show that.
'18 MR.'SIESS: It s'eems to me you've gone to the 19 trouble to number an awful lot of things in this Reg. Guide.
20 But do the numbers reefer back to the requirements above? 21 And then when you get down to Line 11, we've got l 22 that suitable analysis and design, in this case. Before, we l l 23 had a suitable analysis, testing, and qualification methods.
'24 Now, we have suitable analysis and design shall be 25 performed. And I'm not sure what's a suitable analysis.
1 l
l l I ) 38 l' I think if you look at the language in there, 2- you'll see the things I'm talking about. 3 MR. MURPHY: Right. 4 MR. SIESS: Now, the' thing everybody 1Eoks at !
.5 there of course is the provision that says you can take a 6 third of the SSE, or a value greater than a third. But' 7 again, you have to demonstrate those requirements again, r
See them down there in Line 12? ! 8 i j 9 MR. CHOKSHI: Right. ! i I
- 10 MR. SIESS: That is the earthquake engineering
-11 criteria in Part 50.
12 Now, Part 100, what we've got there, and the most ,[ 13 recent copy I have of Part 100 is the October 11 , '91. Is
*4 that right? I did not get an update to that.
/'
10 MR. CHOKSHI: January 15. 16 MR. SIESS: .Oh , wait a minute. I'm sorry. I'm 17 sorry. Just a minute. i l 18 [ Pause.-] 19 MR. SIESS: Part 100, Appendix B, I've got. ) l 20 MR. MURPHY: Right. 21 MR. SIESS: But Part 100 itself, Subpart A,
~ 22 Subpart B. That's still as of October 11, '91? !
i 23 MR. MURPHY: I believe that's unchanged. l 24 MR. SIESSt That's what I assumed. l 1 25 Now, I want the subcommittee to understand, the E
~ , n , -
.,,ww--. ,- , , -
i t
- 1 revised Part 100, this is complete. Does everybody have it?
l 2 It came just'like this. That came in this package; we had 1 3 this last month. 4 Let me just' review briefly what it is. There's i 5 Subpart A, Subpart B, which is what -- Subpart A is the old l ! 6 Part 100. Essentially no change. 7 Subpart B is the new Part 100, effective this 8 date, whatever day it is. 9 Part 100 is divided into two parts. Subpart A to 10 take care of 100 plants that were licensed, and Subpart B to l 11 take care of the new ones. l l l 12 Subpart B has attached to it the new Appendix B. p( 13 Subpart A has attached to it the old Appendix A. l 14 Subpart A and Appendix A are in present l 15 conditions. Everything new is in Subpart B. And Subpart B, 16 in itself, is not really that much different than Subpart A, i7 except it references Appendix B instead of Appendix A. Is - ! 18 that right? 19 MR. MURPHY: Correct. 20 MR. SIESG: It deals not only with earthquakes, 21 but with other population density, et cetera, et cetera, and 22 related hazards, population, and su) forth. 23 Now, that then brings us to a really big chunk of ! 24 changing, and that's Appendix E, and Draft Guide 1015. 35 MR. MURPHY: That's correct.
l l l( 40 i i 1 MR. SIESS: And Appendix B, in itself, has not got 2 enough meat in it. We went through that last time,.and I 3 didn't find anything. There was a lot taken out of Appendix l l 4 A. l 5 MR. MURPHY: That's right. 6 MR. SIESS: What was'left was Appendix B plus this 1 7 Draft Guide. ! 8 MR. MURPHY: That's right. 9 MR. SIESS: Okay. Yes, Bill. 10 MR. LINDBLAD: In terms of scope of Appendix S and 11 what we've been talking about, is volcanic action a subset l l 12 of earthquakes, as seismically-induced floods are a subset ( 13 of earthquakes, or is volcanic action outside the scope of 14 earthquake engineering? 15 MR. MURPHY: What we've done at this time is the i 16 same thing that we've done in the past, and that was to 17 indicate that volcanic hazards be considered on a case-by- ! l'8 ' cace basis. 19 MR. LINDBLAD: And is that in the regulation, or l 20 is it just in the Draft Guide that comment is nade? l 21 MR. MURPHY: No, I believe the wording is in the , l 22 end of Appendix B, just the same as it was in Appendix B. 23 MR. LINDBLAD: Thank you. 24 MR. SIESS: Is it in Part 100 proper? 4 25 MR. MURPHY: For the moment, I have forgotten. 4 a
l 41 l 1 But it is specifically in Appendix B on Page 5. 2 MR. SIESS: Subpart B says population density, j 3 man-related hazards, and physical characteristics of a site, 4 including seismology, meteorology, geology, and hydrology. 5 Volcanos would come under geology, I guess. 6 Appendix B, now, is not just seismology. Appendix 7 B is seismological and geological. That's the heading on 8 it. Seismic and geologic siting criteria. 9 Okay. That brings us down to Appendix B. We 10 haven't talked about Appendix B yet at all, have we? l 11 I'm down to Appendix B, and Draft Guide 1015, l 12 which gets us down to here. (. 13 I will declare a 10-minute break. 14 [Brief recess.] 15 MR. SIESS: Andy, before you get started, I want l l 16 to try to set the stage, I guess. 17 Of course, the object of a subcommittee meeting is 18 to try to develop an understandi'ng andsome basis'for making 19 some recommendations to the full committee, because it's 20 unlikely that the full committee is going to take the time l 21 to go through everything the subcommittee has gone through, 22 although I can think of instances in the past where that's l 23 happened. l l 24 And the best way to get something before the full 25 committee in a reasonable length of time is to have the l
l 1 42 ' 1 issues clarified by the subcommittee, and the recommendation 2 from the subcommittee, and then presentations from the 3 staff, which can give the bases for those recommendations. 4 And I've found that, in some of the more difficult l 5 areas, like seismic design and geology and seismology, where 6 the knowledge of the full committee members is somewhat 7 restricted -- we don't really have any geologists or 8 seismologists on the committee -- I've always tried to have l 9 the thing clearly enough, understand it well enough myself l 10 that I can go in and do a pretty good job of explaining it 11 to the full committee, telling them-what they should listen 12 for in the staff's presentation, telling them what the ] 13 subcommittee has learned, and why it thinks we should do i ( 14 certain ' things, and then have the staff come in and make 15 enough of a presentation for the full committee, to help 16 them understand at least the basis for the subcommittee's 17 recommendation. 18 Right at this point,. I don't understand D' raft 19 Guide 1015, or portions of it, well enough to go to the full 20 committee. And so I'm hoping what we can accomplish today 21 is to dig into this far enough to get some understanding of 22 what's going on, what's proposed. l 23 Now, really there are two things that I'm having a 24 problem with. And one of them is not in the Draft Guide, 25 but it's in Appendix A. I < ~
_ _ . . _ . _ _ _ _ . . . _... _ .,__m _ _ _ _ _ . _ . _ _ . . . . . . _ . . _ . _ . _ _ . _ . _ . , l l V i ( 43 1 The-Staff is proposing, for the determination of , 2 the seismic design basis, shall we say, and that's really i i 3 what we're mostly concerned with here, a dual approach. ; 4 Something they've called a deterministic, which really'isn't 5 deterministic, because apparently nothing is deterministic 6 anymore, and a.probabilistic. A deterministic analysis, the 7 DA,.and the probabilistic analysis, the PA. You.do them l 8 both, and somehow you compare the results, and as a basis l 9- for that you come up.with a basis for, I guess for-setting a 10 safe shutdown earthquake, which is a number that is then 11 used in designing the plant. L 12 And, of course, until the plant is designed and 13 built, it doesn't represent any hazard to anybody. And the j ('
~
14 object is not an analysis, the object . is a design, to quote l 15 a famous structural engineer I knew. 16 But the first thing that isn't clear to me, and 17 I'm not convinced about, is why we're doing it two ways. 18 For years we got by selecting a seismic design basis by a 19 process very similar to what we call, what you not call the 20 deterministic approach. 21 When we originally wrote Appendix A, and I was 22 . involved in that, the main thrust of what we were doing in 23 Appendix A was trying to define the extent of the geological 24 investigations, because at that time, that was the biggest f' l 25 concern of everybody. People were going into completely new i l
_ _ , _ ._ ..-- _ ..-_-_- . . _ _ . _ . _ _ - . . _ . . . _ _ _ _ _ . _ . _ . _ . _ . _ _ _ = . _ ( 44 1 areas for siting, and they would do a geological 2 investigation. They would come to the staff and the staff 3 would say no, we want more. And they went through that 4 cycle a half a dozen times or more and finally got out far 5 enough that everybody'was satisfied. I l 6 And so one of the major objectives of Appendix A 7 ! was to define the scope and extent, and literally the 8 geographic extent, of the geological investigation. i 9 It succeeded very well in that. After Appendix A 10 was written, we never had any nore problems. People knew i 11 what to come in with. la What I am hoping we can do the rest of the day is ( 13 to understand why the staff thinks we need both a 14 probabilistic and a deterministic approach ~, how you relate
~
15 the two, and how I think it relates to the Part 52 or the ! l 16 certified plant designs with a pre-determined SSE. l l 17 I've asked staff to do two things. One is to take 18 an actual example, a site, and show how the SSE would be 19 determined for that site going through the various steps 20 required by the deterministic analysis, the steps being 21 required by the probabilistic analysis and ending up with an 22 SSE. I think you're prepared to do that for a site. 23 MR. MURPHY: That's right.
- 34 MR. SIESS
- What I didn't ask was that it be for a 25 site that does not have a plant at it.
l 1 I ,
_ _ . _ _ . ._ _ _ . _ , _ . _ . . _ _ _ = . - - . _ . _ . __ _. . . . _. _ _ . _ . . . __ l l l l !( 45 i MR. MURPHY: 1 We didn't do it that way. We did it 2 for a site that already has a plant at it. 3 MR. SIESS: Then the~next question is going to be, i 4 why couldn't you do it that way, and could anybody else do l 5 it that way? f I 6 MR. MURPHY: The difficulty right now in doing it j 7 for a plant that does not have a site at it is the extent 8 that the deterministic investigation would require. 9 MR. SIESS: How about the probabilistic? l l 10 MR. MURPHY: Not the probabilistic. We think we 11 can handle that well. 12 MR. SIESS: When you get through doing it for the l( 13 other, then you can explain how the probabilistic would be l - 14 done for a new site. 15 MR. MURPHY: Yes, sir. What we'll explain for the 16 existing site will be exactly the same as for a new site. l l 17 One exception -- that the EPRI or the Livermore hazard l - l 18 studies have'not been' execute'd at a new site. Thats the 19 only exception. 20 MR. SIESS: Yes, but isn't that just as big a 21 hurdle as the fact that you haven't done all the geology for 22 the site? 23 MR. MURPHY: No, sir, it is not, not in my i , 24 opinion. i 25 MR. SIESS: Okay. r
. . . _ .m . ._m. , _ _ _ _ ___ . _ _ . - - _ . . __ . _ - . . . . _ _ _ _ _
l I l l I
. 46 l
- 1- MR. MURPHY
- It's a reasonably simple task to do l
l 2 it for a new site. 3 MR. SIESS: That is what I want you to address. l 4 Okay. l l 5 Then the other thing I had asked is to show how l 6 the new methodologies might differ from the old at some of 7 the existing sites, and you have some data on that. Now, is l 8 that what you're prepared to talk about? l I 9 MR. MURPHY: Yes, sir, that's what we're prepared l 10 to talk about. l l 11 MR. SIESS: Okay. l l 12 MR. MURPHY: And as your questions arise, we can l l( 13 go into more detail than what I have in the viewgraphs. 14 MR.'SIESS: Okay.' Any other points anybody wants l 15 to make before Andrew starts? 16 [No response.] l 17 MR. SIESS: Okay. Onward. 18 MR. CARROLL: I guess -- Chet, I've recently been 19 reading about some experimental work that USGS has been 20 doing on the Loma Prieta earthquake that helps to answer 21 some of the anomalies that have been observed there. I 22 guess it starts with a theory that there are reflected waves l 33 off the Moho layer that amplify it certain distances and so
- 24 forth. I guess I'm just generally curious as to how that f
25 enters into nuclear plant siting.
A' J.
-1 MR. SIESS: Well, that brings up a point that I 2 didn't make that I think will come out, and that is.that the 3 proposed Appendix A, the part we're going to hear about is 4 -about 99 percent eastern US.
5 MR. MURPHY: That is correct.
- 6 MR. SIESS
- The staff says if somebody comes up with another plant out in the West Coast, we'll worry about l 7
-8 it when it happens.
9 MR. MURPHY: Not quite. We've got it covered. 10 MR. CARROLL: No, the words are in there, but does l l 11 -- is that phenomena unique to the West Coast? l 12 MR. MURPHY: The ground motion phenomena you're ( 13 talking about, no, sir, it is not. 14 MR. SIESS: No. I didn't'want to make that -- the ! 15 staff gives two examples of how you might do it for the West 16 Coast. One is Diablo Canyon and the other was what? 17 MR. CHOKSHI: WNP. . 18 MR. SIESS: WNP. Y e s .' Okay. 19 MR. MURPHY: Okay. 20 [ Slide.)
- 21 MR. MURPHY: Today's discussion will be directed 22 at this draft reg guide on seismic sources. Two important L 23 ' points on the viewgraph: obviously, my first name and a 24 telephone number so you can find me afterwards if you need i 25 to.
I
l l. I i 48 .[ l. l 1 (Slide.] 2 MR. MURPHY: A simple outline of the presentation. 3 We will talk about what we are-requiring by way of the dual j 4 ' analysis, and at the end of the presentation, we'll talk l 5 about why we're requesting the dual analysis, and then I l 6 will step you through the important steps that I am certain ( 7' have caused some consternation, particularly with reference 8 to Appendix C to this reg guide. I 9 As Chet has pointed out, we will have some initial 10 results and some comparisons. The initial results, the 11 plant that we will be talking about is the Vogtle site. l 12 That was not quite picked oc random, but just simply fortuitous that it was Vogtle. We feel that we could do the l( 13 14 same calculations for any of the other existing sites. 15 [ Slide.] 16 MR. MURPHY: This is a schematic.of the dual l 17 approach. If you want, we can put quotations around both 18' .the probabilistic and the deterministic analysis and say we 19 use them as defined here. 20 On the deterministic path, we proceed basically 21 through this path the same way we have in the past. There 22 is a little bit of change. Those changes are in definitions 23 and terminology. 24 The intent is that under the new Appendix B, you
- 35. would come up with explicitly the same results that you
l l f I 49 1 would if you went out and did these investigations under the i 2 current Appendix A. l l 3 There is a term here'in the second block under 4 deterministic analysis, " expected maximum earthquake". That 5 is a deterministic term. Our expert panel that we've been 6 working with has strongly recommended that we drop that 7 term, at least those words to define that term. We'll stick l 8 with the definition; we'll just give it a new name. 9 Their comment was that that phraseology is just 10 too probabilistic sounding. What we propose is replacing l l 11 that " expected maximum earthquake" with the terminology i 12 " deterministic source earthquake". It's strictly a matter 13 of terminology. The definitions that appear in the ( 1' . l 14 documents will stay the same, but we're just using different ' l 15 words to express it. l 16 MR. LEWIS: But it will have a precise definition? l l 17 18- MR. MURPHY: We figured it was precise when we l l 19 wrote it. We'll accept your comment on it when you have a 20 chance to look at it. 21 MR. LEWIS: How can you possibly commit yourself l j 22 to accepting my comment when you haven't heard the comment? 23 MR. MURPHY: I said accept it; I didn't say we'd 24 do something about it. 35 (Laughter.) i l l l
l l 50 t l 1 MR. MURPHY; We also come up with -- 2 MR. LEWIS: No, seriously -- we're joking, but j 3 there is a serious point here. 4 MR. MURPHY: Yes, sir, I understand. l 5 MR. LEWIS: Okay. 6 MR. MURPHY: We do come up with a new terminology 7 as well, " controlling earthquakes." There will be a figure 8 indicating what those are in a moment. i 9 The truly interesting side of this figure is the 10 probabilistic analysis. The firtt step in that analysis is 1 11 conducting, as they are today, either an EPRI or a 12 Livermore, and the choice is entirely un to the applicant. if 13 But once they make that. choice, they have to stay with criteria associated with ohe or the other. There is no 14 l 15 mixing back and forth. Either you do it EPRI or you do it i 16 Livermore, and you carry that all the way through. , 17 We will show you a comparison that is made with
'18 the existing operating power plants, and there will be a l
l 10 hurdle involved, and meeting the same probability exceedance 20 set-at a 50 percent level for those existing plants. That 21 process is described in Appendix B of the reg guide, and 22 I'll show you a figure on how we looked at that in a few 23 moments. i ' 24 Chet? i 25 MR. SIESS: But, Andy, when you say to do an EPRI < l
- _ i
?
E l
'51 1 or a Livermore hazard assessment, could you either now or a later explain just what you mean by conducting it?
i 3 MR. MURPHY: Yes, sir. j 4 MR. SIESS: I mean -- l 5 MR. MURPHY: Explicitly, what we mean is --; 6: MR.^SIESS: -- the panel of experts that was used I I 7' to conduct those assessments, right? t 8 MR. MURPHY: Would you repeat your question? ! 9 MR. SIESS: Conducting.an assessment, hazard ; l ! 10' assessment, required assembling a panel of experts, did it L 11 not? 12 MR. MURPHY: When the data bases were established L( 13 for the Livermore and th.e.EPRI, they involved assembling a 14 set of expert panels, both methodologies. We are taiking 15 about, in conducting an EPRI or a Livermore calculation, 16 just putting in some site specific coordinates and some site
- 17. specific determinations associated with the soil conditions 18 so that you are explicitly using the data' sets and the 19 ground motion descriptions that are already on the books.
20 You are not talking about going out at this time and 21 collecting any new data. < 22 MR. SIESS: And now the data-set consisted of i 23 what? Maps showing tectonic -- I 24 MR. MURPHY: Map showing tectonic provinces, 25 recurrence relationships with those. tectonic provinces,
.=e w,-- , , --.-,e .-._ e- __, c. y-, -_. .
- . . . - . .-. . . . _ - . . . . - _ . - . - - _ . - . . . ~ - - . . - . . . - - ~ . - . . . .
l 1 S2 (-
~
l 1 ground motion models, the whole data package. 2 MR. SIESS: They covered the entire eastern US? 3 MR. MURPHY: They did. j 4 MR. SIESS: It was only applied to particular 1 l l [ 5' sites, but it could have been applied to any site. 6 MR. MURPHY: It could be applied to any site. 7 'There are a few site specific pieces of information -- i.e., I 8 coordinates, soil and rock conditions and so forth -- at the 9 site, but the rest of the material is the existing material 10 today. 1 11 MR. SIESS: And there are people out there that 12 can do that? ( 13 MR. MURPHY: Yes, sir. Obviously, at this' time, !
'14 the folks at Livermore can do theirs, EPRI can do theirs,
- 15 and I believe there are two or three consultants that can l l
l 16 carry out the EPRI calculations as well. There is ! 17 documentation in the National Code Center for the Livermore. 1 18 We are talking'about providing an updated set'of 19 documentation and making it a little bit more user frier.dly 30 so that anybody that can get the code and put it onto a 21 machine can run those calculations. So it's, very 32 explicitly, it's making use of the codes and the data sets l 23 that exist today. i 24 MR. SIESS: Bill?
- 25 MR. LINDBLAD
- That would suggest, then, that l
l
. _. _ _ . . m. _ _ _ . _ . . _ . _ . _ . _ _ . _ . _ _ _ . . _ _ . _ . . . . _ _ . . . _ _ _ . _ . _ _ . . _
l l l l l l l 53 ( 1 there is a probabilistic zonation map available for each of 2 those projects, each of those studies, that is not disputed? 3 MR. MURPHY: No. There is -- okay. There is not 4' what I would call a probabilistic zonation map that gives l ( 5 you, as Ted Algermissen does, a plot of accelerations versus l
\
6 . frequency of returns, so forth. There are not those kinds 7' of maps for these studies. If somebody wanted to take the 1 8' time and the ognny, they could be generated. They would not l l 9 be undisputed. We are using these in a relative sense. If 10 those calculations were done for EPRI and then done for f i 11 Livermore and comparea, they would be different. No i la question about it. They would be different. That to some extent is immaterial for what we're t 13 l(
- 14 doing with them. As you'll see a little bit later, we're 15 using them in a relative sense. We're using all of EPRI or 16 all of Livermore to make these judgments.
17 MR. LINDBLAD: But do I understand that the expert 18 ' panel assembled for either one of those projects'has already 19 done the work -- 20 MR. MURPHY: Yes. 21 MR. LINDBLAD: -- in determining the distribution 22 --- , 23 MR. MURPHY: Yes. 1 24 MR. LINDBLAD: -- of probabilities for even new j l 35 sites, sites not -- i l-l
54 lI l 1 MR. MURPHY: The experts assembled their data, a their information, in a generic sense, not site specific, 3 and that information does exist today. 4 MR. SIESS: They have provided information on ! 5 sources. 6 MR. MURPHY: Yes, sir. 7 MR. S1ESS: Geographically defined sources, limits f 8 and what's the largest earthquake that could occur within 9 that area and its return period. Is that right? l 10 MR. MURPHY: That's correct. 11 MR. SIESS: All right. I 12 MR. MURPHY: Those sources were assembled 13 independcat of where nuclear power plants were located. { 14 MR. SIESS: To what extent could that information 15 be used in the deterministic approach? 16 MR. MURPHY: That information could be used and ! 17 the same people that helped generate that are the same l l 18 people that are going to be or have'been in the past f 19 involved in the deterministic approach. 20 MR. SHAO: Excuse me. All of this work has been 21 done to get the resolution chapter. EPRI and Sol Bernstein 22 and also NIC developed this. There were done a few years l 23 ago. We're just using the data. The existing data. l 24 We are not reinventing the wheel. 25 MR. SIESS: Okay. When you do the DA, I want the 4 l
i 55 l f( l 1 other little map with the five areas. I want to see how i 2 those might relate to the same things we got from the l l l l 3 probabilistic approach. 4 So just go ahead. 5 MR. MURPHY: Okay. We make a comparison of the i 6 probabilistic numbers with the existing plants, and then we do some manipulations which are described in Appendix C, and 7 8 based on those we are able to determine or compose a 9 controlling earthquake or earthquakes for the probabilistic 1G analysis. And here's this magic box called " compare". And 11 12 here we specifically make a comparison of the controlling ( 13 earthquakes derived from the deterministic process, and the 14 probabilistic process. 15 Looking at those results, we are also in the 16 position to compare the ground motions that are obtained 17 either from the controlling -- the probabilistic controlling 18 earthquakes and the deterministic control' ling earthquakes. 19 And based upon those comparisons, we are able to 20 establish, or if we need to compute a safe shutdown 21 earthquake ground motion. 22 This would be the ground motion that normally 23 would be available for design of the facility. 24 MR. SIESS: I am just trying to trace out the 25 regulatory positions. Part 100, Appendix -- Subpart B just i
m _. . _ _ . . _ . _ _ . . . . _ _ _ _ _ . . _ . _ . . _ _ .. _ __ _ _ . . . . - _ . . _ . _ . _ _ _ _ _ . _ _ . . . . . _ . . . L 56 l 1 simply refers to Appendix B. Is that right? 2 MR. MURPHY: I believe that's correct. 3 MR. MURPHY: There is nothing in that part that 4 says deterministic and/or probabilistic. 5 MR. MURPHY: Yes. I believe that's correct. 6 That's my recollection. 7 MR. SIESS: Now, where in Appendix B does it say 8- that there are two paths? { 9 MR. MURPHY: It does not say in Appendix-B that 10 there are two paths. It says in Appendix B that you must do l 11 a probabilistic analysis and.you must do a deterministic 12 analysis. 13 And then -- that's what it says there. , And then !( 14 in the reg guide we have this figure. 15 MR. SIESS: I'm not down to the reg guide. I'm
' 16 looking in the place that has Appendix B.
17 MR. CHOKSHI: That's on page 2, line 18,.Dr. t
' 18 Siess. It says "both deterministic and probabilistic 19 . evaluations shall be conducted".
20 MR. SIESS: That is the scope of what? 21 MR. MURPHY: The scope of the criteria that -- the 22 scope of the requirements. 23 (Pause.] 24 MR. SIESS: Skip that part with the introduction 25 and the scope of the investigation. I am looking now under l l
l ( i > 1 l 57 1 1 required investigations. 2 Wouldn't that be the place.to look? 3 MR. CHOKSHI: Yes. That's also lines 28, 29 and 4 30. 5 MR. SIESS: Line 28. Yes. It says, "to support 6 both probabilistic and deterministic determinations required 7 by these criteria". 8 Now, that suggests that somewhere else 1t's 9 required by these criteria, and I'm looking for the place 10 where it says thou shalt do both. l 11 MR. CHOKSHI: It's only in the scope. 12 MR. MURPHY: In the scope. Maybe what you're l l( 13 indicating is it may be appropriate that we put in an ,.w 14 caplicit statement in the requirements tha't a probabilfstic 15 and deterministic analysis is required. ; 16 MR. SIESS: It makes it a little more obvious and 17 then the need to justify that requirement become.c a little 18 clearer. 19 MR. MURPHY: Okay. 20 MR. SIESS: I don't know how much the Commission 21 looks at reg guides, but presumably the Commission does look 22 at regulations. 23 Okay. Let's go ahead then. t 24 You think you've made it required. Somewhere. That was definitely our intent. We 25 MR. MURPHY: l l l
_ . . _ . - ~ . l 58 1 will look at that and see if it's there, and obviously we 2 will get comments from the lawyers as to whether we'.re 3 requiring it or not. 4 MR. SIESS: And you're going to tell us why it's l 5 required? l 6 MR. MURPHY: That's the end of our presentation l 7 today. 8 MR..SIESS: Okay. All right. Well, most of your i 9 presentation is how you do it. You'd be better off to start 10 off with why. 11 MR. MURPHY: Okay.- 1 12 MR. SIESS: Or you think it's easier to explain l( 13 why after we see what? s. l ! 14 MR. MURPHY: I would hope so, but I felt initially l I that the areas where we were having difficulty were in the i 15-
- 16 Appendix B and C of the draft reg guide, so I intended to 17 address.those items first, and then after --
l ' i 18 MR. SIESS:' I'm still having difficulty getting to 19 the need for two ways to do it. L I 20- MR. MURPHY: If you want, we can take this out of 21 order. It's up to you. 22 MR. SIESS: Yes. I keep having a feeling that the 23 decision to do it both ways was made semewhere up front , l j 24 without deciding why. 25 And then after you got through, you figured out {i I I i l
l I 59 I, 1 some pretty good reasons for doing it that way. 2 MR. LEWIS: I'm a little concerned about that too, L 3 Chet. And as I read some of the justification language, l l 4 there's a place in the reg guide that describes what the 5 defects of a deterministic analysis -- limitations, I'm 6 sorry: -- and what the limitations of a probabilistic l 7 analysis are. 8 And it lists among the major limitations of a 9 probabilistic analysis that it gives some weight to the 10 tails of the distributions. But I regard that as one of the I 11 strengths of a probabilistic analysis. I 12 So you are requiring people to do a probabilistic 13 analysis in order to extract from it a single controlling ( l 14 earthquake and throw away the tails, which are the great 15 strengths of a probabilistic analysis in such a way that you 16 can get a number which can then be compared with a number l 17 that comes with a deterministic analysis, which is.in turn 18 produced by exper opinions avera'ged in some bizarre way, 19 which I have never entirely understood. I sort of agree with Chet. One should lay out the 20 al philosophy of what one is doing, and then do the arithmetic. 32 MR. SIESS: You know, I may be wrong on my 23 chronology, but we've been talking about revising Appendix A 24 for what -- 15 years? 25 MR. MURPHY: Twelve that I am specifically aware i ,
60 1 of. 2 MR. SIESS: Twelve that you are sure of. 3 MR. LEWIS: It's a moment in history time. 4 MR. SIESS: And I'm sure that when.we starting 5 looking at our problems with Appendix A, none of them had to 6 do with probabilities. ! 7 We were hung up -- we were using tectonic 8 provinces and all sorts of geologic terms, a lot of which we i 9 had invented, and some that were obsolete. I 10 And of all '.ne things we wanted to fix , none of 11 them had to do witt. probabilities of exceedance. 12 MR. MURPHl: That's correct. ( 13 MR. SIESS: And the changes that have been made in and in Part 100 dealing with the deterministic 14 Appendix B, 15 part, have fixed all of those things. 16 MR. MURPHY: No, sir. They haven't fixed all of i 17 those things. What.we are looking for -- and I guess we , J 18 will get started in the rationale for it; going to the l 19 probabilistic approach -- is that we are looking to the 20 probabilistic material to provide us with a solution, a way 21 to look at many of the problems that were associated with i 22 the deterministic process. I 23 The deterministic process was not explicitly able i 24 to address the uncertainties, the variations that were out 25 there in the opinions that were being expressed by the l l s
_..g 1 l l l 61 ( l experts. The people that were being used as consultants and 2 as employees of the utilities to get the applications 3 submitted and accepted. 4 MR. SIESS: As of what time -- what point in time I 5 are you talking about? l 6 MR. MURPHY: From my own experience, we are f f 7 talking in '70s. , l ! 8 MR. SIESS: How did we manage to license those i 9 plants and pick SSEs that -- that seemed to have led to l 10 plants that have very nice margins of safety? l 11 Every plant we've ever had to do a seismic margin 12 analysis has shown very adequate margins, which is very 13 comforting. ( 14 And yet all those plants were -- we picked the SSE ! 15 by a method you say that was wrong? MR. MURPHY: No, sir. We're definitely not saying 16 17 it was wron,g. l l 18 What I'm saying is that it was -- if you want, , l 19 fraught with difficulties and uncertainties, and it was a process that took significant iterations. That it wasn't 20 21 as you indicated, when Appendix A was initially written, i l 22 there were iterations involved in how far out we went with 23 our geological and seismological investigation. 24 As I understand the difficulties, because I wasn't 25 present for all of them -- there were difficulties back and i i i
i i 62 l
\
1 forth on selecting the ground motion, debating as to whether 2 or not this earthquake was part of this province or.not, 3 whether or not this province boundary extended within 4 12 kilometers of the site, or whether we could restrict it 5 at 100 kilometers away. l 6 MR. SIESS: And those are still there. I 7 MR. MURPHY: To a large extent, some of those 8 debates are still there. We are looking to the 0 i 9 probabilistic analysis to put those debates into -- I'll 10 call it a proper perspective. i 11 MR. SIESS: Have you got an example that you could l 12 show us where we can see how the probabilistic analysis will f (- 13 resolve issues in the licensing process and/or lead to a I f 14 safer plant?' Which are not the same thing. i 1 15 MR. MURPHY: That's correct, l 16 MR. SIESS: By any means. f l 17 MR. MURPHY: I have,the example for the Vogtle l 18 site'that we did, in part for this presentation. And we 19 have the information that we did for three other sites. t I 20 MR. SIESS: Was Vogtle a problem in picking the 21 SSE? 22 MR. MURPHY: There -- 23 MR. LEWIS: Phyllis Sobel is reviewing it. 24 MR. SIESS: I can't hear you. Use the mike. 25 MR. MURPHY: Phyllis Sobel, who has worked on this l 1
63 1 revision with us, was a reviewer, and maybe she can offer a 2 comment as to whether or not there were problems with that. 3 With that site explicitly. 4 MS. SOBEL: This is Phyllis Sobel. There were no 5 problems using the deterministic procedure at that site. We ; l 6 did also employ probabilistic procedures to compare the l 7 probability of exceeding a safe shutdown earthquake at 8 Vogtle with that of other sites. 9 MR. SIESS: Why did you do that? 10 MS. SOBEL: It was a procedure that we were 11 developing at that time. We were using probabilistic i 12 calculations at operating power plant sites to compare ( 13 probabilities of exceeding the SSE at new sites under 14 review. 15 MR. SIESS: Was it helpful there? I mean, could 16 you not license the plant without doing that? 17 MS. SOBEL: According to the regulations, we did 18 not have to use that procedure to license the plant, and'it 19 didn't change the safe shutdown earthquake ground motion. 20 MR. SIESS: What was it, incidentally? 21 MS. SOBEL: What was? 22 MR. SIESS: The SSE. 23 MS. SOBEL: The SSE was 0.2g anch to a reg guide 24 1.60 response factor. 25 MR. LEWIS: And what was the probability of l
k 64
$I ? 1 exceeding that?
2 MS. SOBEL: I'm sorry. I've forgotten that 3 number. 4 MR. SIESS: We remember the SSE, but we don't 5 remember the probability of exceedance. 6 MR. LEWIS: Well, if we know the SSE, we know the 7 probability of exceedance. We just don't seem to remember e 8 it. 9 MR. SIESS: The SSE affects how you design the 10 plant. The probability of exceedance doesn't. MR. LEWIS: I understand that. But I can still 11
- 12 ask what the probability of exceeding the SSE is.
1 MR. SIESS: I know. But the point -- my point was
;( 13 14 everybody remembers the SSE for a plant.
But we must understand that at the 15 MR. MURPHY: 16 time Vogtle was done, the probability of exceedance was l j 17 simply a tool that was used to assist the staff. It wasn't required. It wasn't specified in the l 18 19 regulation. And it's not recorded in the SER for that 20 plant. 21 MR. fiICSS: Is it part now? 4 22 MS. SOBEL: It's part -- This is Phyllis Sobel. 23 I'm sorry. We did record our analysis of the use of j 1 24 probability in the safety evaluation report. 25 MR. SIESS: But now you said it wasn't required. l
l o 1 ! 65 lI l 1 It wasn't a part of the regulations. You proposed to make j l 1 2 it that. 1 3 MR. MURPHY: That's correct. 4 MR. SIESS: The question is why? What I want you 5 to do is to justify, convince me and anybody else that needs ; l l 6 convincing, why we need the probabilistic approach. 7 One, to make plants safer or to make them safe I 8 enough, or two, to make them easier to license. i l 9 MR. MURPHY: The intent is principally to make 10 them easier to license. The intent in going into revising 11 Appendix A was not to change.the safety or the safety 12 margins associated with the next generation of plants. l l( 13 , We're doing that based upon the Commission policy l 14 statement that the current plants are safe enough. l 15 So our intent was to maintain that level of I 16 safety, and then the intent -- the specific intent of the 17 regulation was to address and alleviate the problems in 18 licensing the facility. l 19 MR. SIESS: So what you need to show me, then, is j 20 where you get hung up if in the licensing process -- if all i 21 you have is the deterministic approach and you don't have 22 the probabilistic. 23 Before you get to the point where you needed to 24 ask 100 more questions or whether you needed to require 25 another six months of exploration.
66 1 In other words, where it helps you in making a 2 decision that you've had enough information from the l 1 3 licensee to set an SSE. 4 So that's what I'd like to see. An example of l 5 that. Hal, did you have something? 6 MR. LEWIS: I have a slightly differant concern, 7 Chet. But I share that one. 8 , And that is, I'm sort of stunned by the comment l 9 that one can calculate the probability of exceeding the SSE
=
10 internally but it wasn't recorded. That's a little bit l 11 strange. 12 MR. SIESS: She said it was. l? MR. LEWIS: I'm sorry. It was recorded. 1
! l
- 1. 4 MR. MURPHY: Phyllis says it was recorded.
15 MR. LEWIS: Okay. ; 16 MR. SIESS: She corrected herself. It is on 17 record. But the thing -- I know a lot of 18 MR. LEWIS: 19 information that goes into the probabilistic approach that 20 does not go into the deterministic approach. 21 My objective is to get better assurance about the 32 safety of the plant. Not to make it easier to license or 23 harder to license or whatever. Or even to follow tradition. 24 I'm a little fuzzy about what information goes 25 into the deterministic analysis that is not available for i I' l
. . . , . ~ . . 4 . -= - -. .-,.n - - - # - a s - .- .-~=a+. - . a- a -.
I i l 1 l l l l { the probabilistic' analysis, and I would appreciate being i 1 i I 2 updated on that point. 3 MR.' MURPHY: You are interested in knowing what 4 goes into the -- 5 MR. LEWIS: The deterministic that is not part of 6 the probabilistic. Or cannot be made part of the 7 probabilistic. 8 In other words, what does the deterministic give - 9 - use in informatien that is not used in the probabilistic? 10 I'm just asking out of ignorance. 11 MF. MURPHY: Okay.. 12 MR. SIESS: I think he -- I think he's going to tell you it's the,same information. It's just used in a ( 13 V s 14 dif erent way, but -- MR. LEWIS: Let's see if he says that, in which 15 16 case you know what my next question would be. 17 MR. SIESS: It's probably about at the point now 18 where we ought to let him ge\. on with trying to describe 19 these two things. 30 Keeping in mind that, as far as I am concerned, 21 you are on the defensive. 22 MR. MURPHY: I feel it! 23 [ Laughter.) 24 MR. SIESS: You're going to have to convince me i 25 that there's a good reason for doing the probabilistic 1 l
-..~ - _ _- .- - _ _ __ ~ ..-._ -_ . - - . - . _ - . - . .__- .- . . - ..- - .
! i l I 68 'f 1 thing, because -- i 2 MR. LEWIS: So you are going to rule that.he ! 3 shouldn't answer the question. t MR. SIESS: Oh, no. No. I think he should, but I 4' ' i 5 'think he -- l 6 MR. LEWIS: I thought that's what you were doing. ! 7 MR. SIESS: I think he can bring it out better by l l 8 going througn the process and showing you and sort of keeps 9 that' question in mind. 10 MR. LEWIS: So you are going to rule that he l 11' shouldn't answer it at this time? 12 MR. SIESS: Well, no. He can answer it at this ( 13 time if he wants to. . id MR. LEWIS: Do you want to? t l 15 [ Laughter.] 16 MR. MURPHY: If I understand the question, you l 17 want to know what is used in the deterministic that doesn't l 18 go into the probabilistic? 19 MR. LEWIS: Yes. That's right. 20 MR. MURPHY: What is not -- 21 MR. SIESS: And he has predicted what your answer 22 will be. So we'll see whether it really is. I 33 MR. MURPHY: The databases that are associated 24 with the two of them are basically identical. The 25 difficulty is the deterministic process does not have l i , - _ _
l 69 l ( 1 provision for handling the uncertainty associated with the 2 databases that are used. 3 So that if there is an outlier opinion about the l 4 size of the earthquakes that can occur in a tectonic 5 province, that has -- in the deterministic process, it is 6 going to be a fully enveloping process -- that opinion must 7 be taken into consideration. That earthquake may be 8 . extremely large,,and to the rest of the community, maybe too 9 large. It may be totally unrealistic. 10 The deterministic process, in my mind, does not 11 have an adequate way of addressing those outlier issues. 12 MR. LEWIS: Yes. But that's the opposite of the l 13 question I asked. l(
. ~
14 MR. MURPHY: I understand that. l 15 MR. LEWIS: Okay. 16 MR. MURPHY: And what I am saying is that what is l 17 in the probabilistic that cannot go into the deterministic , ! 18 is thai, information associated with the uncertainty'in those 1 19 databases. Those information that are used for the 20 calculation. 21 MR. LEWIS: But that's -- again, the opposite of 22 the question I asked. 23 The question I asked is what is in the deterministic that is not in the probabilistic. I know 24
, 25 about many things in the probabilistic that are not in the
t l l 70 1 deterministic. 2 MR. MURPHY: There is nothing in the deterministic 1 3 that is not available in the probabilistic. l 4 MR. LEWIS: Okay. Fine. That's the answer to the
! i 5 question.
6 In which case, the next question is obvious. But l 7 that we'll answer as we go along. That's the way I want to ! 8 structure it. 9 MR. SHAO: I think you need to answer Chet's 10 question. Why the present approach is not adequate. f 11 I think that really is the answer to Chet's 12 question. MR. MURPHY: That is part of the answer to chet's ( 13 l 14 question, yes. 15 MR. SIESS: I think I can hear the words, Larry, 16 Andy. But there are 110 plants out there licensed, safe 1 i 17 shutdown earthquakes.were determined by essentially the , j 18 det'erministic process, which we have improved on in the new I 19 Appendix. 20 An awful lot of those plants have had their 21 designs reviewed, their seismic margins calculated, and it 22 turns out that they are pretty good plants. They have 1 I t ;23 pretty. good design based on that. 24 So I want to know where we went wrong that we are 25 going to fix. What's going to be different? i I
l i l (. 71 I L 1 MR. MURPHY: What I am trying to say is that what 2 we did-in the past is not wrong in the technical sense. We 3' did a good job. You did a good job. The other members of l 4 the NRR staff did a good job in selecting and approving the ! 5 SSEs for the plants that exist. We are not trying to change that aspect of the 6 i 7 regulations. We want to come back to the same point. What 8 we are trying to do is, in our minds, is to make the process f 9 more stable, easier to be actually executed, so that there 10 would be to some extent less controversy and less l o i 11 discussion, less argument about the results of the l'2 utilities' investigations for the application, and i,n the (;> 13 way the staff looks.at them and reviews them.
- .~ , , .
14 Tha' intent is to improve the process, the t 15 bureaucratic part'of the process, rather than specifically . change the science or the engineering that;was done. That's 16 17 what I'm trying to convince you of, that, yes, there is ? l 18' nothing wrong in a generic sense with the plants that were l l 19 licensed and technically, the way they were licensed. What 20 we want to do is improve the bureaucracy of how they were 31 licensed. 22 MR. SIESS: So I guess what you're saying is that 23 when you get into an argument about a particular seismic' ! 24 source, which we did with Charleston, for example, -- 35 MR. MURPHY: Very definitely.
. . - ._ . - = .
l~ l L i. 72 i l i 1 MR. SIESS: -- you can get yourself out of that l 2 problem by going to a probabilistic approach and say, "Yes. 3 Well, we're not quite sure what that is over there, but if 1 ! 4 we pick this earthquake, the probability of exceedance of , I ! 5 this earthquake is no worse than that for half the other l l 6 plants out there." l 7 MR. MURPHY: That is part of it. 8 MR. SIESS: It's no better than -- it's worse than , l l l 9 half of them. 10 MR. MURPHY: That's correct. I mean, that's what 11 happens if you pick that level. But we are doing some l 12 sensitivity studies on that particular point to see how critical that picking the 50 percent is. It doesn't seem to i( 13 14' be ultra-critical. 15 But your point is correct that what we are 16 intending to use, the probabilistic analysis is to put that specia.1 earthquake that somebody has found or postul,ted' a 17 , 18 into the proper perspective of all.the other earthquakes 19 that are out there. That may be done at the time of 20 licensing of the facility, or it may be like the Charleston 21 earthquake and its reconsideration, it may be done after the 22 plant has actually been licensed. I l 23 We need to know whether or not that extra 24 earthquake, that additional earthquake, has any significant l 25 influence or safety significance to the facility as proposed l l l
. __ _ ._ . _ . - _ _ _ . . ~ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ __._ _
1 l 73 1 or as built. So that's what we're using, in my nind, a
- 2 principal focus of the probabilistic analysis.
3 MR. SHAO: Can you give the basis for that? Do 4 you remember the Perry earthquake? 5 MR. MURPHY: Yes, I remember the Perry earthquake. 6 MR. LEWIS: Could I ask one question just for my 7 information? I know so little. How many events have there 8 been ever in which a plant has been challenged,by an SSE and 9 survived? How many times has that happened? 10 MR. MURPHY: You mean in the United States? l 11 MR. LEWIS: Yes. That's my country. 12 MR. ROTHMAN: May I say a couple of things? My i( 13 name is Robert Rothman and I'm from the NRR Structural and - l
~ . . .
14 'Geosciences Branch. 15 There have been cases where the response spectrum 16 of a plant, the SSE response spectrum has been exceeded in 17 the United.. States -- The response spectrum or the ground l i 18' .MR. LEWIS: f 19 motion? l l 20 MR. ROTHMAN: Well, it's the response spectrum of l 21 the ground motion compared to response spectrum that was 22 used for the design of the plant, the SSE response spectrum. 23 MR. LEWIS: No. I thought that the SSE was 24 defined in terms of ground motion. 25 MR. ROTHMAN: No. It's defined in terms of a l l
74 I 1 response spectrum that reflects the ground motion. 2 MR. LEWIS: It's calculated in terms of response 3 spectrum. l i 4 MR. ROTHMAN: You take -- you can use -- you use 5 -- the SSE is defined by a response spectrum that envelopes l l 6 the -- which represents the respected ground motion. 7 MR. LEWIS: That's right. 8 MR. ROTHMAN: But it's not a specific time history 9 that's used for the -- 10 MR. LEWIS: I see. So every time people tell me l l 11 the SSE is .2g, they are lying. ! l 1 12 MR. ROTHMAN: No. It's .2g anchoring some l l (, 13 response spectrum.
- 14 MR. LEWI'S: Okay. Fine. Fine.
15 MR. ROTHMAN: For a standard spectral shape, that i I 16 SSE PGA is the -- ' 17 MR. LEWIS: Okay. Fi.n e . In that case, how many i
'18 ' events have there been?.
19 MR. ROTHMAN: Well, there have been a couple of 20 events where -- 21 MR. LEWIS: Two? 22 MR. ROTHMAN: Well, I can think of Perry and VC 23 Summer before it was an operational plant. It probably at 24 certain frequencies exceeded. They were very short duration 1 25 events -- I'm talking on the order of a second or two -- i l l
L l i . ) i ! l [ 75 1 with high frequency ground motion. So it didn't have any l 1 2 effect on the structures or components. ! l 3, MR. LEWIS: Okay. So that's happened twice. I l l 4 MR. ROTHMAN: But'there have been some overseas j 5 where there have been cases, I believe, but I can't document ] 6 .them for you. ) l 7 MR. LEWIS: Well, let's not worry about that for
- 8. the moment.
39 MR. ROTHMAN: I'd like to answer a couple of the 3 10 questions that have been asked, try and help out a little 11 bit. I'm not a probabilist, so I'm not trying to defend the l i 12 probabilists, but when we do a -- ( 13 MR. LEWIS: You know, I.wish people wouldn't br,ag !' -14 'about w' hat'they don't know.
- 15 MR. ROTHMAN: No, it's just I'm saying --
16 MR. LEWIS: It's just a bad habit. t 17 MR. ROTHMAN: I would just like'to tell you how it l
- l . ,
can help us, and there have' been cases where a certain 18 l I 19 region of the country, where the licensee has proposed a 20 certain magnitude earthquake, and possibly the staff has 21 agreed with it, where other people have suggested that there 22 might be larger earthquakes. 33 Now, the staff, in doing its deterministic ! .34 analysis, implicitly actually considers what could be a I 35 larger earthquake, and if they don't think the likelihood is i
m =__ ._ _ . _ _ _ . . _ _ _ - - . - i l l l ,[ 76 i 1 very large, they exclude those earthquakes. 2 We can't, from a scientific point of view, say 3 that you could never have that earthquake larger than the l ' 4 one we select, but wo make some kind of a judgment on what 5 we think is the largest that can occur. It's an implied ; 6 probabilistic analysis, it's not a formal. 7 However, the probabilistic analysis can allow you i
)
i 8 to consider these large earthquakes, give them some formal j 9 probability and see if they do contribute to the~ hazard at ! 1 ! 10 the plant, so that you can take into consideration larger l 11 events than you would consider in a deterministic analysis, I
\
l 12 and that's the way it would help us in making these !( 13 decisions, and we have done that on a few cases, not 1 14 formally'but informally. 15 In fact, if you remember, in the Diablo Canyon 1 16 review that we just completed this last year, the utility ! 17 did do a probabilistic analysis to see which of'the faults i l l'8 and their earthquakes contributed to the hazard at the 19 plant. We found that the Hosgri was the dominant, and that 20 the other nearby faults contributed very little, and so we 21 were able to put that into proper context. 22 We had the deterministic understanding that they l j 23 didn't contribute much, but this way we had a formal way of ' 24 putting them into proper context to complete the review. 1
, 25 You asked the other question of how the i
l i ( 77 ) 1 information from the deterministic would be used in the 2 probabilistic. You've got to realize, right now, our ! 3 probabilistic -- ) l 4 MR. LEWIS: I didn't ask that. ! 5 MR. MURPHY: The opposite. l 6 MR. ROTHMAN: You asked the opposite. How the 7 information from the probabilistic would be used in the 8 deterministic or how the information would be -- 9 MR. LEWIS: No. I asked what information is used j 10 in the deterministic that is not available to the I i 11 probabilistic. j 12 MR. ROTHMAN: Okay. One of the -- and let me put 13 it in the context of what we're talking about right now. 14' Right now, we're talk'ing about using t'he Livermore and the 15 EPRI programs which are basically regional
- 16 characterizations. The detail of the geology is not of a l
l 17 level that we would ask for a specific study at a site, if 18 we had a site. That kind of work was not done'. They are 19 regional studies. The data bases are frozen. 20 If we were to go in and site a plant someplace 21 now, the detail of the knowledge, the information that would f 22 be required for the deterministic study would be a lot ! l l 23 greater for the specific region around the site. Within, 24 say, 25 kilometers of the site, we would require detailed 25 geologic study. I l i _ .
- ( 78 1
1 That would not be factored into the data base in 1 l 2 the probabilistic studies because those are already I l 3 -- the data bases exist and they are frozen, they wouldn't 4 be modified. So there would be more detailed knowledge on 5 the site itself in the deterministic study. , l 6 MR. LEWIS: That's in the practice. I understand 7 that. But in principle, those studies would be available to 8 the people who are doing the probability of exceedance 9 curves for the probabilistic if they wanted to do it better. 10 Is that right? 1 11 MR. ROTHMAN: Yes. 12 MR. LEWIS: Okay. 13 MR. ROTHMAN: They would be available to them, but (- .
'14 at this time, what we're'saying and what we're putting down I i
15 here on paper now is that those data bases are frozen, 16 what's in them now, and that if the applicant were intent on l t i 17 following this reg guide, that data base -- that information l l 18 would have to be factored in for all eastern sites, or else 19 they would not be following the guidance that's provided 20 here. 21 MR. LEWIS: Actually, when I asked how many times 22 -- just to clarify why I was saying this -- when I asked how 23 many times the SSL has been exceeded and the plant has 24 survived, it was to address the sense which I seem to find 25 that we seem to have done okay on the first 110 plants since l I , i i
._._.~._ _ ..___.__ _ .. _ _ _ ,
l \ l i i l 1 they are still standing, and, in fact, a large earthquake is 2 a. low probability event. You know, there's a little bit of l ) l l 3 Russian roulette about it. You can't really say that you've i ! 4 done well just because you've survived a short fraction of 5 the return time of a.large earthquake. That's why, l l 6 presumably, we're doing these things. i- ! 7 MR. SIESS: We have also got seismic margin 1 8 studies -- 1 9 MR. LEWIS: Yes, I understand that. 10 MR. SIESS: -- that would indicate how much larger
'll they could stand.
L 12 MR. LEWIS: Yes. I l { 13 MR. SIESS: And we've got a great many non-nuclear
. plants thNt were not designed for any seismic that-w'ithstood 14 15' pretty good earthquakes.
16 MR. KENNEALLY: No, no, I'know that, Chet, and'in l 17 fact I've always felt that we've tended in this business to 18 exaggerate the threat of earthquake. Simply because the l 19- uncertainties are larger, we're trying so desperately hard l 20 to be conservative. l 21 MR. SIESS: We have compounded our conservatisms. 22 MR. LEWIS: That's correct. You know, that's my I 23 general view. l 34 MR. SIESS: Yes. I i-25 MR. LEWIS: But since we're doing it, we ought to l l i
- ( 80 1 do it right.
2 MR. SIESS: But Andy has accepted -- has said that 3 this is not going to make plants any safer; it's just going 4 to make them easier to license. I guess there are two steps 5 to the licensing process. One is what goes on between the 6 staff and the applicant, but in the past, we've had another 7 step which involved other groups than either of those either 8 before a hearing board or in a court. 9 I'm not sure that what makes it easier for the 10 staff and an applicant to agree on is going to make it 11 equally easy for a licensing board or a court to settle an 12 argument. But I'm not a lawyer, so maybe I shouldn't bring i 13 that up. 14 Carl?
.15 MR. MICHELSON: Yes. Since the Perry earthquake 16 was brought up as an example, something has bothered me for 17 .some time -- maybe you could give me a brief explanation as 18 to why I shouldn't be bothered -- and that is the earthquake 19 as an earthquake was almost a "No, never mind" in terms of 20 surrounding countryside, if I understood the earthquake l 31 correctly. Yet, for that particular plant at that site, it 22 was a rather significant building response.
23 Then I wonder, well, is the design basis 24 earthquake properly selected for that site or why such a 25 small earthquake would create a large but brief response,
. - _ . =-. . - -- _.- - -
( 81 i 1 and should I be uncomfortable about the design basis, then? 2 MR. BAGCHI: Dr. Siess probably has good answers l
\
3 for that, but I would simply observe that there is a ! 4 resonance phenomenon that one should not ignore. That ; 5 earthquake had energy in the particular frequency range. , 6 i That's how it got amplified. But it didn't do any damage. l 7 MR. CHOKSHI: Acceleration response was higher. l l 8 The displacement -- l l 9 MR. MICHELSON: Okay. Let's assume for the moment 10 since the earthquake per se was sort of a "no, never mind" 11 for the area, how much larger earthquake might I get in that 12 area and how would the building respond to it? You know, Ix 13 have we properly selected response spectra for the -- MR. 14 BAGCHI: Based on the analysis, it's 10,000 year return 15 period. 16 MR. MICHELSON: Well, we've already had that one, 17 of course, so we won't get another one on the average for 18 10,000 years or something? 19 MR. CHOKSHI: I think the answer to your question, 20 if you compare the design spectra, the earthquake spectra, 21 actual earthquake, you will see the amount of energy going 22 into the design is much larger than the short spike. So the 23 design spectra is far more conservative in that sense. 24 MR. MICHELSON: We just happen to have hit an 25 unusual earthquake. I
, _ . ~.. _ _ _ . _ _ _ . _ _ _ __ _ _
,( 82 l' ! 1 MR. CHOKSHI: Yes. And it had a one frequency -- ! l 2 MR. MICHELSON: Now, since it was not fully 3 anticipated, it was really unusual. Why do I believe that l l 4 we're smart enough to know that we won't get a more severe l 5 response the next time from a relatively earthquake, but ! 6 bigger than the one that was experienced this time? j 7 MR. CHOKSHI: I think our design spectra will l 8 cover a lot of those kind of events. ! l' 9 MR. MICHELSON: Okay. In other words, I shouldn't l , 10 be uncomfortable even though we've got a very interesting l l 11 response from a building for what was apparently a very 12 modest earthquake. l( . 13 MR. SIESS: You know, I think if you hadn't made f4 an elaborate dynamic analysis of that building, you would' 15 have never thought that earthquake did a darn thing. I 16 MR. MICHELSON: That's probably right. 17 MR. SIESS: And it didn't. 18 MR. MICHELSON: Except to certain components in' 19 the building, and they did survive all right. 20 MR. SIESS: It reached a value equal to what they 21 analyzed for a part where it didn't make any difference. 22 MR. CHOKSHI: May I add one thing? In 1986, there
- 23 was a symposium on Appendix A, and I think, as Andy l
24 mentioned, the issue of user probabilistic conception in 25 Appendix A licensing was debated at that at great length. l I
- .. .~. .. .= _ _ . - - - - - . -.
l 1 l l 83 [ 1 Just to -- I think just what you are saying, Dr. l 2 Siess, and I think Andy also has a quote from that, but what j 3 they said at the meeting -- I am reading from the summary of 4 the '86 symposium, which was three days, and the probability I 5 was a large part of that discussion -- basically, they are 6 saying that one aspect of the existing regulation is its , 7 sensitivity to changes in the perception of the seismic 8 hazard. l 9 There was a question of the stability during the 10 licensing process, but even after the plant is built, the ; 11 plant is vulnerable under current regulations if something la happens. l . (, 13 Now, occurrence.of a medium to large earthquakes ) l 14 in the tectonic province of a site can lead to an updati'ng i 15 of the maximum earthquake. This process is unstable, which, ) l i 16 in the opinion of most participants, could be stabilized by 17 using probabil.istic techniques. i l 18 Now, that was -- what I want'to convey is that l l 19 this issue has been around for a while. It is not something l 20 that we have just come up with. It has been debated at a ! 21 number of locations. 22 MR. SIESS: Well, I guess what you're saying is 23 the Charleston issue was settled on a probabilistic basis. 24 MR. MURPHY: That's correct. j l i 25 MR. SIESS: If Charleston was free to move up and ; l i l
( 84 1 down the Eastern Coast of the United States, I mean not only 2 were there going to be nuclear plants we should worry about, 3 but whole cities, including the District of Columbia. I 4 wouldn't want to be in this building if the Charleston 1 5 earthquake hit. l 6 MR. MURPHY: Right. , 1 7 MR. SIESS: I mean, I wouldn't have wanted to have l 1 8 been in the capitol building before they fixed the north l 9 front of it, or whatever, the west front. I know what its 10 shape was. 11 But again, we sort.of got away from Charleston 12 partly by the paleoseismic studies which suggested that we i ( 13' were right in the first place that it was local, and partly 14 by the probabilities, right? 15 MR. MURPHY: That is correct. The Charleston -- 16 MR. SIESS: And we didn't really make anything 17 safer. i. 18 MR. MURPHY: Not specifically -- no, we did not 19 make anything safer. l 20 MR. SIESS: I mean -- 21 MR. MURPHY: Nothing was changed. i 22 MR. SIESS: And I guess we satisfied ourselves 23 that we didn't really do the world a lot of harm by citing
- 24 plants along the coast here.
35 MR. MURPHY: That's correct, we did not. i s
+
l \ L l l I l 1 i 85
- (
l 1 MR. SIESS: Now, if we didn't have any plants 2 along'the East' Coast, would we have reached the same 3 conclusion? 4 MR.' MURPHY: If'we didn't have any plants, the ! 5 only thing you can say if we didn't have any plants on the 6 East Coast, we probably wouldn't have carried out this 7 analysis and be in the position we are today. If they l '8 weren't there, USGS would not have brought up this 9 difficulty. 10 MR. SIESS: Well, you know -- j 11 MR. MURPHY: Let me back up a little bit. Back in i
' 12 '82, when we started the -- started and got these two
(" t 13 seismic hazard probabilistic -- probabilistic seismic hazard I 14 studies started, they were part of a two-pronged effort to 15 resolve the charleston issue. The first was called short-16 term, which took ten years, and that was specifically these l 17 two probabilistic studies to indeed put the Geological 18 Survey's opinion in'to' a proper ' perspective. ; 19 The second was the long-term deterministic actual l 30 studies of the geology and geophysics and seismology of the
~
21 area to see whether or not the Charleuton earthquake could 32 be located in size and so forth, and that was the material , 23 in the same ten-year time frame that produced the ; 24 information about the paleoseismicity area. 25 So indeed, it is something the Commission has done I I
~c ,e
I i y 86 1 ; 1 in the past by way of a two-pronged.-- a probabilistic with 2 quotations and a deterministic with quotations -- approach 3 to resolving a problem. 4 MR. SIESS: Okay. I'm hcaring interesting 5 arguments, and I just want to mention one thing, and that is ! 6 I appreciate hearing from people from NRR and I hope they l 7 will continue to participate because these ere the people 8 that are going to be licensing plants. These people at the 9 table are the ones that are writing the regulations, but 10 they are the ones that are going to be applying them. 11 MR. MURPHY: Let me raise my hand and say 12 explicitly, Phyllis Sobel -- and I won't say she agrees with y 13 every word we have written -- has been participating with , 14 us. She is with NRR. And Bob Rothman/ who is not so l 15 explicitly in the group that is writing it, has been 16 participating in many informal discussions of this process. , i 17 So the folks in NRR are definitely involved in writing this
-t 18 regulation and the reg guides.
19 MR. SIESS: I have heard a couple of comments now l I 20 about, you know, how this is applied and how it's useful to I 21 them, and I think that's important, and I'd like to continue 32 to hear it. 23 I guess I just want you to continue, because I'm i f ! 24 still trying to figure out just how much of an effort is
. 25 required to do both the deterministic and the probabilistic.
. l l l 1
f l l l 87 A. I 1 I can see some advantages in the probabilistic. I'm not 2 sure they are all that separate, coming down two columns. I , j 3 think there is -- 4- MR. MURPHY: There is feedback. 5 MR. SIESS: I think they may be just looking at 6 the same thing from two different directions. But let's go 7 ahead now.
}
f-l 8 MR. LINDBLAD: Chet, could I -- 9 MR. SIESS: Yes, sure. 10 MR. LINDBLAD: There have been some sites that 11 have been rejected on the basis of permanent ground , 12 displacement. Is it intended that probabilistic approaches l( 13 might facilitate licensing where there is uncertainty associate'd with permanent ground deformation? 14 15 MR. MURPHY: I cannot specifically answer -- 16 MR. SIESS: Which one? l, 17 MR. LINDBLAD: I guess I'm talking about Malibu 18 and Bodega. MR. SIESS: Tnose were seismic, I thought. They 19 20 are both before my time, but -- t 21 MR. CHOKSHI: Yes. The process here is for the 32 ground motion, not displacement, the vibratory ground 23 motion, what is being described. ! 24 MR. LINDBLAD: Is there any uncertainty associated 25 with permanent ground deformation and geologic studies? i 4
88 ( 1 MR. CHOKSHI: Yes. 2 MR. LINDBLAD: But that will be handled strictly l
~
3 on deterministic approach, are you saying? 4 MR. CHOKSHI: Or it could be on a case-by-case. l 5 basis because it's such a, you know, a very -- l 6 MR. LINDBLAD: Deterministically, but not I 7 probabilistically. 8 MR. SIESS: GETR,was seismic, wasn't it? 9 MR. MURPHY: Yes. l 10 MR. SIESS: Or was it? 11 MR. MURPHY: I'm not sure what you nean by was 12 seismic? 13 MR. ROTHMAN: GETR the issue was displacement at ( 14 the foundation. 15 MR. SIESS: Yes. And was that due to earthquake 16 or was it due to something else? 17 MR. ROTHMAN: It was due to an earthquake fault 18 displacement at the foundation -- through the foundation of That is what the issue was. It wasn't vibratory 19 the plant. 20 ground motion. I al MR. SIESS: No. We're talking about i 22 displacements. 23 MR. ROTHMAN: Yes. 24 MR. SIESS: Yes. 25 MR. SHAO: I don't think this is covered by that
89 1 definition. l l 2 MR. MURPHY: That is not an issue that would be 3 solved by the probabilistic analysis. 4 MR. LINDBLAD: So in your flow diagram, I should l 5 read this strictly from the vibratory ground motion -- l 6 MR. MURPHY: Yes, sir. t 7 MR. LINDBLAD: -- not from other -- Tsunami or 8 deformation. 9 MR. MURPHY: No, not Tsunami, not permanent ground l l 10 displacement or anything. 11 MR. SIESS: That's right because everything will 12 end up being in terms of spectra. 13 MR. ROTHMAN: I don't think the fact that the reg l( L 14 guide does not specify displacement, I don't see why, if a , i 15 utility wanted to come in with some kind of a probabilistic l 16 analysis, we wouldn't review it. We're not precluding that; 17 it's just not spelled out in the reg guide. i I 18 MR. LINDBLAD: And you don't have a previous study 19 to compare it with. 20 MR. ROTHMAN: No, we don't. 21 MR. SIESS: Everything we're talking about in this 22 chart, for example -- and most of what follows -- is dealing i 23 with some numbers that go into the front end of an analysis. 24 And when you ask a questfon about the ground 4 35 displacement of Tsunami's they haven't got any computer 4
_. . . - ~ - - - i 90 ( 1 programs to put that into. l 2 So those will be handled case by case. 3 MR. MURPHY: Right. l 4 MR. SIESS: You can laugh, but that's the way the l 5 process operates. 6 MR. MURPHY: That's right. 7 MR. SIESS: We argue about what the number is that L 8 goes into that dynamic new pipe, or whatever it's called. l 9 You know, everybody's got one now, that will calculate 10 everything that will happen, assuming that you put the right 11 number in it. 13 Fine. Go ahead. ; I l( 13 [ Slide.) . l 14 MR. MURPHY: Okay. And as I understand it, wh'at l l 15 you would like to do now is in effect skip down to viewgraph ! 16 number 25. : l l I 17 This is the material that I had prepared to l
'18 address the question of why we are going'into a dual i
l 19 approach, requiring both probabilistic and deterministic 20 analyses. I 21 MR. SIESS: Now, that's what we've been talking ( 23 about for the last hour. i 23 MR. MURPHY: That's right. 34 MR. SIESS: But these are the words that are going i ) 25 to go to the Commission. To justify this. t l
I l l 91 lt 1 MR. MURPHY: Words like this is what's going to l 2 end up -- I believe it will be in the statement of 3 considerations that will go to'the Commissioners, to CRGR j 4 and then to the Commissioners, as to why we are into this 5 dual approach. 6 MR. SIESS: Just for the sake of argument, or just 7 to set the stage, let's assume that the ACRS writes a letter 8 to the Commission saying that we have reviewed all this and 9 we think most of it's pretty good. 10 But we just don't see any real good reason for 11 having this dual approach. l 12 Now that goes to the Commission and they probably
- g 13 will write back and -- DDO, and say, respond to the ACRS, or 14 something.
15 So the issue will be raised. You have already ; 16 raised it, in effect. l 17 MR. MURPHY: Yes. You started out by saying obviously we l 18 MR. SIESS: l l 19 have to justify this. 1 30 I'm now trying to make it a little stronger. And 21 that was hypothetical. I don't know what the Commission's l 22 going to do. l l 23 MR. SHAO: There is one rationale that is not 24 here. With the deterministic approach you can always come 25 up with an earthquake larger than what you calculated. l l l l
-i J
j 92 I I 1 And just.like a large earthquake, you say, , l 2 somebody's-going to say, what is the basis for not l 3 designing? i j 4 The only way'you can get rid of'it is by the l 5 probabilistic approach. l 6 MR. SIESS: You are saying that the old maximum 7 credible earthquake is an unsatisfactory basis for design, , l 8 because people's leve) of credibility varies? I ! l 9 MR. SHAO: Yes. But some geologists can say that j i 10 there's such an earthquake. In the SSE, how can you say you l l 11 don't consider that. ! la MR. SIESS: Well, you mean that there's a greater ( 13 one that's credible to him that wasn't to you? . ! + l 14 MR. SHAO: Yes. 15 MR. SIESS: Now, if you ask him to put a ; 16 probability on it, you think he'll agree with you on what l
- l. 17 the probab.ility -- so far,.you're just talking about the 18 size of it.
19 MR. SHAO: Yes. L 30 MR. SIESS: Everybody else has decided it's a 6 21 and this guy comes along says, gee, it could be a 7. t 23 Now, how are you going to use probability to 33 decide that it can't be a 7? Who's going to set the 34 probability? i 35 MR. SHAO: Dased on the study that we had.
1 1 I[ 93 i I.know, but then somebody can disagree 1 MR. SIESS: l 2 with that, can't they? j 3 MR. MURPHY: Very definitely they can. And ) 4 obviously there have been disagreements-back and forth a 5 between EPRI and Livermore. i
- 6 MR. SIESS
- Do you think it's easier to agree on 7 probabilities that are magnitudes?
8 MR. SHAO: The answer to the question here is we i 9 have looked at these big earthquakes and it's not that we 10 have ignored this big earthquake, but we used probability to 11- say that this earthquake is very low. The probability is l 12 very low. ( 13 MR. SIESS: .Do,you respond to this man by saying, , . s , , 14 we didn't say it couldn't be a 7, we just naid the 15 probability of a 7 is too low. 16 MR. SHAO: Too low. Yes. If you only use the i 17 deterministic, you cannot get rid of this question. Just
. ~
18 like we had the Charleston earthquake question ten years 19 ago. 20 MR. SIESS: You couldn't get rid of somebody 21 else's question or your question? 22 MR. SHAO: Oh, no. In the -- people tried to have 23 a contention. 24 MR. LEWIS: I have still a little bit of unease 25 about'the idea of comparing the probabilistic and the
l i i[ 94 l 1 deterministic approaches, .because many of the words I hear l 2 in defense of the deterministic approach have kind of . l ! ! 3 probabilistic undertones. i 4 They're based on sort of what one could call 5 " visceral" probabilistic analysis, or something like that. 1 6 And the output of a real probabilistic analysis is 7 a probability distribution, which is principle, just cannot 8 be compared with an effort that results in one number. 9 So what is being done to compare them is to l 10 conceal the probabilistic origins of the deterministic 11 analysis, and emasculate the probabilistic consequences of a l 13 probabilistic analysis, so that you end up with comparable l 13 outputs, which can be compared with each other. (...
- 14 And that's -- it leaves me feeling a ~little bit l
15 hungry -- that's there's more there t.bnn you are using. 16 MR. CHOKSHI: I think you made a very good point, i 17 and intent here is a procedure. But the basic. intent is l 18 'that'-- when you do the probabilistic analysis, you have a 19 lot of information. Not in terms of these scenarios, but 20 what are the dominant earthquakes, what frequency, and that 21 all is there. 22 And this is just a procedure to come up with a l 23 design ground motion, because you are @aing to come up with 24 one motion for design. 25 But I think, to me, the most important benefit of l l l
i 95 t ( 1 doing probability.is to have all that information. Not this 2 -- this is a design tool. 3 MR. LEWIS: That mirrors my concern. That one is 4 taking this mass of information -- and maybe it's the word 5 " compare" that is bothering me, because you're really not 6 comparing them, you're emasculating one of them by pulling l out numbers that you can then put against the numbers in the 7 8 deterministic. 9 MR. CHOKSHI: And in that context, do you mention 10 about the tails? l 11 MR. LEWIS: Yes. 12 MR. CHOKSHI: What do we -- I think we probably i 13 didn't.say it properly. What -- suppose the probabilistic ( 14 comes out an unusually large event? 15 And then that gives you a way to go back and 16 determine. The deterministic thinking says, my best
. 17 judgment doesn't accept that.
I 18 MR. LEWIS: 'Uh-huh. 19 Then you can go back to what assumptions in your 20 probabilistic analysis drew those results, understand them, 21 and you have better information to then judge. 22 But to help you have all that information. j 23 MR. LEWIS: Sure.
- 24 MR. CHOKSHI
- And that was the intent of saying f 35 why we need both.
i
i' l l l( 96 l l 1 MR. LEWIF: Yes. I also recognize that in the end 2 a licensing decision is yes and no. ! 3 So, one can have probabilistic analyses in one's I 4 ears. One still has to extract an answer from them. 5 But the question is, at what stage that's done. l 6 And I'm not -- I'm still uneasy about the idea of comparing l l 7 these two. 8 MR. SIESS: Why don't we -- 9 MR. MURPHY: We resolved the Charleston earthquake 10 really by two -- by both probabilistic and determinittic 13 analysis. ! 12 MR. SIESS: Andy, I think it would be helpful to 13 go ahead and give us an example of the deterministic 14 approach. 15 MR. MURPHY: Okay. ) 16 MR. SIESS: Now, you've got a slide onthe 1 17 schematic definition of controlling earthquakes Is that 18 basically the' deterministic approach? [ l 19 MR. MURPHY: Right. l l 30 MR. SIESS: Can you explain it in terms of that 21 slide? Because I don't think you've got another slide. 22 MR. MURPHY: No, sir. I don't. But I think from 23 the slide that you are talking aboui', I can explain what is 24 done. 25 MR. SIESS: You can also, I think, explain where i l
._. _ - . , _ . _ _ _ _ .._ _ _ _ _ __ __ ._ . _ ~ _
l [ 97 1 the proolems are with this. 2 MR. MURPHY: That's correct. I think we can do ' 3 s ame of that here. l 4 MR. SIESS: Bring out the problems. l Those e f l 5 problems that are solved by probabilistic approhch. I 6 MR. LEWIS: Chet, that's fine, but, you know, i 7 since he didn't talk about his viewgraph 25, I just want to l 8 stipulate -- 9 MR. SIESS: He will. l 10 MR. LEWIS: He will? i l 11 3 SIESS: He will. 12 MR. MURPHY: I will. ( 13 [Slids.] 14 M R' . SIESS: I want'him to come back to that. 15 MR. LEWIS: I won't be here. 16 MR. SIESS: Write me a note, and I'll be sure that l 17 it gets covered. l 18 MR. MURPHY: Okay. As it says, this is what we l 19 are using as a schematic definition of the controlling 20 earthquake. 21 What we have shown here -- obviously the nuclear 22 power plant site to start with. We have something tnat we 23 have called seismic zone one, which is basically the 24 background zone that the nuclear power plant is sited in. 25 Then we have -- at this state we don't show any e
- ...-._ _ . _ _ _. . _. . . _ . - _ _ . ~ . - _ - _ . ~ . _ _ . - _ , _ . _ . - . _ _ _ . _ _ . _ _ . . _ . . _ . . _ _ . . _ .
l i [ 98 ! I specific boundaries to that zona. 2 Then we show three other seismic zones that have i 3 been defined on the basis of seismicity and geology. The j 4 classical work that'has been done in siting nuclear power 5 plants in the past. , I 6 MR. SIESS: And how are those different than the i 7 zones that are used in the seismic margin? 8 MR. MURPHY: They are different from those'in the 9 probabilistic. l 10 MR. SIESS: Yes. 11 MR. MURPHY: In that these zones have specific la boundaries associated with them. They have specific,
. 13 basically single maximum earthquakes associated with them.
14 Where available, they have single recurrence in' formation 15 associated with them. 16 They are singularities. They -- 17 MR. SIESS: Where do they come from?. 18 MR. MURPHY: Pardon? 19 MR. SIESS: Where do they come from? 20 MR. MURP1fY: Basically they came from -- 21 MR. SIESS: No. I've got a new site. I'm going 23 out -- the sites located 30 miles northwest of Indianapolis, 23 Indiana. 34 MR. MURPHY: Okay. You go out and -- 25 MR. SIESS: What do I do? l l l
=
99 l( 1 MR. MURPHY: You hire your crew of geologists, , l 2 seismologists, geophysicists, and you basically go through ! 3 the existing procedures to find in Appendix A, to define i 4 what the seismicity and what the geology is of that area.
]
5 And then -- l 6 MR. SIESS: They would ignore all the stuff that l 7 was done in the seismic margin studies? l l 8 MR. MURPHY: No. They would definitely not ignore 9 that. And if they did ignore that information, you can be 10 sure that the staff would specifically say, we have this 11 reference here. What use did you make of it? Or, why I 12 didn't you use it? ; 13 The same as we've done, NRR has done in the past. i l ('s l e 14 They would be required to explain why a piece of information 15 was either deleted or used, and based upon that information, l 16 the utility -- the applicant -- would put together something l
- 17. similar to this.
i 18 MR. SIESS: And this would be -- these would be. 19 smaller areas than those areas that are defined for the l 20 seismic margin studies? l ! 21 MR. MURPHY: They might be. Or they might be 22 larger. We're not -- 23 MR. SIESS: They're different? 24 MR. MURPHY: We're not -- 25 MR. SIESS: You know. I'm trying to see what's
' l i
l l l 100 ( 1 the difference between these " seismic sources" and those -- 2 I don ' t kns ',' what you call ' ems -- that I have'seen on maps 3 of the experts from the Livermore and the EPRI studies. i 4 What are those called? I 5 MR. MURPHY: To answer Hal Lewis' question, 6 basically this represents -- could represent one of the 7 opinions that went into the probabilistic analysis. , 1 : l 8 This.might be a,particular firm's best estimate of 9 what the seismic sources look like in this particular 10 region. 11 MR. SIESS: This is one man's opinion? I l 12 MR. MURPHY: This is one man's or one company's l l 13 opinion as to what's there. l l 14 This'would be in'that case, then we contrast it l 15 with the opinion expressed by the total of the EPRI and the 16 total Livermore hazard studies. ( 17 MR. SIESS: And there's no probabilities 18 associated with th'is? He-doesn't have to'say, I think -- 19 MR. MURPHY: No.
.20 MR. SIESS: This is an area, this is the maximum 21 earthquake I would expect from that area?
22 MR. MURPHY: That's right. ! 23 MR. SIESS: Okay. 24 MR. MURPHY: Did -- 25 MR. LINDBLAD: Andy, as you show this diagram, i t'
l l ! 1 l [ 101 l 11 though, you have four component responses that -- l l l l 2 MR. MURPHY: You're not supposed to look down i l 3 there quite yet.
- 4. MR. CHOKSHI: Explain. l l
5 MR. MURPHY: That's -- we haven't quite got to it 6 yet. Okay? 7 And in carrying out the deterministic analysis for 8 the seismic' source 1, which includes the site, today's I 9 procedure would be to ask for a site-specific set of ground l 10 motions associated with the earthquake here. l 11 And basically, what happened, the earthquake -- l 12 the largest earthquake here would be moved to the site 13 vicinity. (( -
. And i'f there was a set of records that could 14 15 reproduce what that maximum earthquake would be in this l 16 particular geologic and geotectonic area, the specifications l
17 of the site -- those would.lue first used to specify - to 18 develop the spectra for the seismic source on'1. j
-19 MR. MICHELSON: Is it correct to assume that that .
l 30 was the process, then, for Perry? That you moved it down to al Charleston and over to Brunswick, and so forth? 22 MR. MURPHY: No.
-23 MR. MICHELSON: Or did I misunderstand your l
- 24 presentation?
l 25 MR. ROTHMAN: You misunderstood. Okay. Perry --
,...e n
. - - . - - - . .~...-_ -- - - ..... - ._ - ...... - ... - .. - . _ _. - - _ .
l f c 102 1 the Perry nuclear power plant is in a seismotectonic 2- province that is characterized under the current Appendix A 3 called the central stable region of the United States. 4 What we have done in the central stable region is 5 taken the largest historical known earthquake that we have 6 not been able to associate with a particular seismic source, l 7 and that happened to be the Anna earthquake -- Anna, Ohio l I. '8 earthquake of 1937, I believe. 1 9 MR. MURPHY: Uh-huh. 10 MR. ROTHMAN: And we have assumed that that 11 earthquake could occur at the site. At the Perry site. 12 MR. MICHELSON: For this drawing here, then, these l l i ( 13 are rather closely located -- 14 MR. ROTHMAN: We don't have any scale on'it. 'You 15 could think of this as being maybe a 1,000 kilometers, each 16 side. 17 MR. SIESS: He's not talking about that right now. 18 - MR. ROTHMAN: But what he's saying -- - i 19 MR. MICHELSON: But when he envelopes it -- he's 20 talking about all of them. What he's saying -- l l 21 MR. SIESS: One at a time. 22 MR. ROTHMAN: Let's assume -- let's say the Perry L 23 site. We would assume -- and it was assumed -- that an i 24 earthquake the size of the Anna, Ohio earthquake. k 25 MR. SIESS: That's source 1? l t
i f ( 103 , t l 1 MR. ROTHMAN: That's the central stable region of i l 2 the United States. 3 MR. SIESS: We are trying to relate this to the l l 4 chart? l l ! 5 MR. ROTHMAN: That's right. 6 MR. SIESS: That's source 1. l 7 MR. MICHELSON: That would be source one. ! .8 -
. MR . ROTHMAN: Correct. We would assume that that' 9 earthquake could occur in the site vicinity.
10 MR. SIESS: Okay. 11 MR. ROTHMAN: And habitually, what we have done is 12- we have asked to develop site-specific response factors ( 13. . based on.that magnitude earthquake -- 14 MR. SIESS: You're getting ahead of the question. 15 Please. 16 MR. ROTHMAN: Okay, i 17 MR. SIESS: Stor l 18 MR. ROTHMAN: Yes. ! 19 MR. SIESS: What was source 2? 20 MR. ROTHMAN: I don't -- 21 MR. SIESS: Where there any other sources for 23 Perry that -- 23 MR. ROTHMAN: Well, I mean -- I don't think that i ! 24 there are any other sources close enough to Perry that would , 25 have contributed.to the ground motion higher than assuming
i 104 1 that the Anna, Ohio earthquake would occur. 2 But let's assume that we would consider, say, the 3 Clarendon-Linden fault of New York State as a source. 4 That's in northern New York State in the Attica, New York l 5 region. 6 We would have assigned a maximum magnitude
- 7. earthquake to that source, assumed that it could occur on ;
l 8 the closest approach of that source to Anna, Ohio, and then i 1 9 done a ground motion attenuation to estimate what the ground 10 motion would have been in Ohio. l 1 11 So that you can assume that that was maybe source I 12 2. ; l 13 MR. MICHELSON: These other sources are, though, (l 14 in the same tectonic province? 15 MR. SIESS: No. 16 MR. ROTHMAN: Well, they might be or they might be 17 other tectonic provinces. 18 MR. MICHELSON: Why, for instance,.wouldn't you l 19 assume that the Charleston earthquake might be moved? 20 MR. SIESS: You could. But well -- 21 MR. ROTHMAN: In past licensing we have assumed 22 that the Charleston earthquake could occur in -- would only , 23 occur in the region of Charleston, South Carolina, and 24 that's several thousand kilometers away from Anna, Ohio and 25 it wouldn't have any effect on it. I i i l 1 I
\
l( 105 1 The ground motion would be nil. 2 MF. MICHELSON: I misunderstood. I thought you 3 were moving the plant site to each of these four. 4 MR. ROTHMAN: No. We move the source to the plant 5 site, or to the closest approach -- 6 MR. MICHELSON: Well, if you move the source to 7 the plant site -- 8 . MR. SIESS: Don't move the source to the plant. I 9 Bob, sit down. You're confusing us. You don't move the ! l 10- source to the plant site. L 11 Source 1, which is.the whole cockeyed area -- but 12 you've got no way of locating -- pinning an earthquake down. ( 13 You put that.one next to the site. You never put it right 14 smack right under it. Fifteen kilometers or whatever. l l 15 All right. These other sources have boundaries. ' l l 16 You put.the earthquake at the closest to the site that you 17 could get it, and then attenuate it by the attenuation t 18 formula --
'19 MR. MICHELSON: By the distance.
20 MR. SIESS: Out to the site. But you never -- 21 nothing moves. 23 MR. MICHELSON: Okay. I misunderstood the 23 earlier -- 24 MR. MURPHY: Okay. Now, if-that's understood, 25 what's done -- to some extent, we'll call it a mental f i i i
l I l [ i 106 1 exercise. j 2 This doesn't necessarily have to be done on paper. 3 That you take and you use the earthquake from source 1 to j 4 obtain a spectra'that would be associated with that -- the 5 largest earthquake in that source. 6 In effect, you put it down on a piece of paper. ! 7 Then you look at the other sources in the area. 8 In this particular example, we have taken source 1 and 9 source 4, which is this one here -- you've taken the largest 10 earthquake in source 4, you move it to the closest approach, ! 11 you attenuate its ground motion to the site, and in la effect -- then you can come up with a spectra associated 13 with that. ( ' 14 M R'. SIESS: 'Now, how'do you get the spectra? Is j 15 that the reg guide 1.60 spectra for the ground motion you 16 got when you attenuated to the site? I 17 MR. MURPHY: That is one way of doing it. Other l . l 18 ways of doing it are to take actual ground records from 19 earthquakes that have approximately the same geotechnical 20 characteristics as the site that you've got. 21 MR. SIESS: Like what? We were criticized for 22 using West Coast time histories on East Coast plants. How , 23 many East Coast -- 4 l 24 MR. MURPHY: Often West Coast earthquakes time t 25 histories were used for East Coast plants. i 1 t-l ,\ L - ._ _ s
. . . -- . . .- - . . ~ - . . - - . -._.- . -
1 l i 4 107 1 MR. SIESS: Yes. It wasn't qualified by -- you 2 l mean rock or soil, or somet ing? 3 MR. MURPHY: At this' stage, by rock or soil. 4 ! MR.'SIESS: So, each of those little broken lines 5 you've got down there is a spectrum that may be obtained l 6 from actual records or may be just arbitrary? From reg 7 guide 1.60? 8 MR. MURPHY: Actual records. I wouldn't say ' 9 arbitrary. 10 MR. SIESS: Well, 1.60 is arbitrary. l 11 MR. MURPHY: Okay. If 1.60 is arbitrary -- l 12 potentially arbitrary and you simply take a look at these. 13 Plot thrsse and examine which ones of these earthquakes, l 14 .which opectra in these earthquakes wou'ld have the great'est 15 effect on the facility. l 16 MR. SIESS: And you envelope it? 17 MR. MURPHY: No. What we're saying is that if you 18 decide -- well', for th'e new Yegulation, we're'saying that 19 you can envelope them, or you can actually use the 20 individual spectra that were developed for these controlling l 21 earthquakes. 22 MR. SIESS: For what? For designing the plant? 23 MR. MURPHY: For designing the plant. If the 24 engineers that are designing the plant can make specific use 15 of these individual shapes, that's their option. i (
t [ 108 1 MR. SIESS: Now, you're speaking as a geologist. 2 MR. MURPHY: It's a' difficult -- more difficult j 3 task than enveloping -- but it's their option. 4 MR. SIESS: Would a structural engineer say that 5 for me? 6 MR. CHOKSHI: No. I.think for the most likely 1 7 outcome, it will be something like reg guide 1.60. You will 8 assure that it does envelope all the ground motions, and use 9 that. 10 I don't think the engineers would prefer to use 11 two different ground motions. 12 MR. SIESS: I don't see how.they could. 13 MR. CHOKSHI: The other -- one more clarification 14 that you would -- 15 MR. SIESS: For design, you would come up with a 16 time history that would give you a spectrum -- 17 MR. CHOKSHI: Some smooth spectrum. You 18 wouldn't -- 19 MR. SIESS: An envelope -- what's labeled here as l 20 the SSE ground motion. I 21 MR. CHOKSHI: I think that could be exactly what 23 practically will occur. l 23 One more thing. This spectra. One way of getting 24 it is the assembly of time histories, but there are also
- 25' attenuation relationships one can use.
l
l I l ll 109 ! 1 One doesn't have to use 1.60 or some arbitrary l 2 shape. There are other ways of computing those spectra in 3 the ground motion. So -- I don't want to leave the l 4 impression that it's either time histories or 1.60. 5 MR. LINDBLAD: But are there not probabilistic
- 6 inputs into those responses? Aren't they generally mean 7 plus one sigma?
j, 8 MR. CHOKSHI: Well -- okay. But that's more of 9 a -- you mean one ground motion or something like that. 10 MR. LINDBLAD: The response. 11 MR. CHOKSHI: The response. Yes. It's one mean 12 plus one standard deviation. l (- 13 But it's not a fully probabilistic, but 14 'considering different models, and different -- you know. If ! 15 you take, for example -- if you take ten time histories, 16 which represents a source and a distance, you calculate the 17 spectra and take the mean plus one standard deviation 18 response, which is not probabilistic in the sense of doing l 19 some other things. j 20 Allowing for radiation, the record for l 21 earthquakes, and going to some different ground motion 22 models. It's a statistical approach. Just thinking of mean 23 plus one standard deviation. 24 MR. LINDBLAD: You are distinguishing between i 35 statistical and probabilistic? 4 + l
. . _ - . _ . . . _ _ . - - - . - - ~ - - - . . - . . --
l t l ( 110
- 1 MR. CHOKSHI
- Yes. That's not in terms of 2 modelling. Yes. It's not like having an --
3 MR. LINDBLAD: How about stochastic? Is it 4 intended that there be a margin between those and the SSE, j 5 Jim? You've shown a margin, but that's not really intended? 6 MR. MURPHY: Yes. That is not really intended. 7 MR. CHOKSHI: It's schematic. 8 MR. SIESS: When you were saying you don't have to 9 use 1.60, I think -- I'm trying to remember the case. 10 Clinton -- there's a little bit of New Madrid that gets into 11 Clinton at what? The low frequencies? Long period stuff? 12 MR. ROTHMAN: Yes. It would be done at one or 2 f( 13 hertz, I imagine. 14 MR. SIESS: Okay. So that's what you mean. You 15 might take that earthquake, but it would only affect that 16 part? Or would that be taken care of by the way it's shaped l 17 on here? 18 MR. CHOKSHI: Right'. 19 MR. MURPHY: That would be taken care of by -- 20 where that earthquake would influence the spectra. And a I 21 distant earthquake, like the New Madrid at Clinton, we would 22 anticipate that that would be in the long period end. 23 Low frequencies. 24 MR. SIESS: Okay, now. That's the end of the -- 25 you have described all of the deterministic selection of an
L 1 ( 111 1 SSE? 2 MR. MURPHY: That's in very simple terms,'I 3 hope -- that's what is done on'the deterministic side. l 4 MR. CARROLL: Now, we are talking here, applicants 5 under Part 52? 6 MR. SIESS: Not necessarily. 7 MR. CARROLL: Not necessarily? 8 MR. SIESS: There is nothing in the Appendix _that 1 9 limits it to Part 52. l l l 10 MR. MURPHY: We're talking future applicants. ' 11 MR. CARROLL: But suppose under Part 52 I come in, 12 pick a site, and apply for a combined operating-license, and 13 in the course of all that, we agree on what this curve ought t 14 to look like, right? 15 MR. MURPHY: Right. 16 MR. CARROLL: How can it ever be changed in the 17, future? Have I got that locked in forever more? Or, what 18 would happen if -- 19 MR. BAGCHI: Mr. Carroll, there may be some 20 misconception here. Part 52 deals with early site permit 21 only. And COL would revert back to Part 50. 22 So that doesn't have to be changed. i 23 MR. CARROLL: Well, what I am really asking is, 24 can -- what has to happen before the staff can invalidate l 35 the response factor that they have agreed to in the
( 112 1 licensing process? 2 MR. BAGCHI: The COL is a site-specific license. 3 MR. CARROLL: Yes. 4 MR. BAGCHI: So that does not have to be under 5 Part 52 requirements. 6 MR. SIESS: Are you talking about an early site 7 review? 8 MR. CARROLL: No. I'm talking about -- 9 MR. MURPHY: You are talking about basically the 10 same thing as would happen today. What has to happen. 11 MR. CARROLL: Yes. 12 MR. MURPHY: And basically, some piece of
- 13 significant new information has to come to the surface.
14 Which may mean Chet's Washington,'D.C. earthquake 15 in 1995. If we have a magnitude 7-1/2 earthquake under 16 Washington in 1995, we are going to have to examine 17 potential effects of that kind of an earthquake on the 18 plants out there. 19 MR. SIESS: As far as I'm concerned, you could 20 have it anytime after April 1st. i 21 [ Laughter.] 22 MR. BAGCHI: Are we understanding your question 23 correctly? 24 MR. CARROLL: I think so. 25 MR. MURPHY: So that we are looking for a -- if
l i 113 ( 1 what has been approved under Part 50 for a new facility, you 2 are looking for some significant piece of information to 3 basically -- to change our outlook on the information from i 4 beforehand. ( 5 MR. SIESS: Maybe -- you weren't asking about an 6 early site review? 7- MR. CARROLL: No. 8 MR. SIESS: You were asking about what? A i 9 standard plant reference? l 10 MR. CARROLL: Yes. 11 MR. SIESS: Certified design? 12 MR. CARROLL: Yes. Certified design. I. 13 MR. MICHELSON: Are you talking about a COL? l( I 14 MR.' CARROLL: I was talking about an applicant 15 getting a combined operating license. { 16 MR. SIESS: Yes. But not necessarily with an 17 early site review? It could have been a site review just ! 18 yesterday. 19 MR. CARROLL: Well, I don't care how he got it. 20 At that point -- 31 MR. SIESS: And now you're worried about the
-22 change between --
23 MR. CARROLL: No. He and the staff have worried 24 about an earthquake response -- 25 MR. SIESS: You are worried about the time between 1 t
- _ . . . ~ . . . . ~ - . . . . . . . - - - - - -- - .- -
l l
> 114 1 the COL is granted and the plant's operating.
, 1 l 2 MR. CARROLL: No. After the plant is operating. ; 3 MR. SIESS: After the plant is operating. That's 4 no different than any other plant. i 5 MR. CARROLL: That's what they told me. That's l 6 it. 7 MR. LINDBLAD: Now, I am trying to understand your 8 picture, and the title of the figure. 9 Do I understand that when we are talking about 10 deterministic versus probabilistic that this is the way we l l 11 would envelope for sources in a deterministic viewpoint? 12 But it could be that the SSE ground motion could l( 13 be less than these two peaks under a probabili.stic approach 14 if we'showed that sources 1 and 4 were rare events? 15 Is that correct? 16 MR. MURPHY: That is correct. But under the 17 probabilistic approach as it's set up today, it's -- 18 realistically, it's unlikely that this represents an exact l 19 duplicate of one of the maps that we used in the 20 probabilistic analysis. 21 So that you probably are not going to be able to 22 point specifically at source zone 4 and say what you have 23 just done about that maximum earthquake determined from the 24 deterministic analysis. 25 There's -- l l
I i 1 115 { 1 MR. LINDBLAD: Are you saying that the 3 probabilistic analysis can raise ground motion but it can't l i 3 lower ground motion? l l 4 MR. MURPHY: The probabilistic analysin has the 5 potential of all three being exactly the same, being higher, 6 or lower. 7 And when we get to it, I'll show you the cases 8 that we've done so far, and I'll say that'the results so far 9 indicate that the probabilistic and the deterministic 10 analysis are in the same vicinity or neighborhood. 11 MR. SIESS: But if in the deterministic analysis l 12 source 4, which just looking at this thing, is a fair , ( 13 distance away, so it must have had a pretty good earthquake j 14 assigned to it -- but if the geologist, seismologist says, 15 yes, that could -- that could generate a 7.2, but we think l 16 the likelihood is very small. ; 17 That couldn't be taken into account. 18 MR. MURPHY: Not as we're set up today. Not with 19 the deterministic process at that specific site. I 20 MR. SIESS: So what we are supposed to do for each 21 of those sources is determining the maximum possible 22 earthquake, maxinum credible earthquake? , i I 23 MR. LEWIS: Controlling earthquake. 24 MR. SIESS: No, no. Controlling earthquake is 25 something else. They have to identify for each of those i
. .--- . . . . .- . . . . . ~ - . - . . . -
I 1 116 [ , 1- areas -- what do you call it? Expected maximum earthquake. I MR. MURPHY: Expected maximum earthquake. And we f 2 1 3 really should give you the reference. You can read the 4 exact words that we have proposed today for that earthquake. ! j 5 MR. SIESS: The maximum expected earthquake is 6 what you determine for each of those zones. l MR. LEWIS: But I thought you said you were going 1 7 8 to change that name? 9 MR. MURPHY: Yes, sir. We are going to name that 10 name. i 11 MR. CHOKSHI: The definition will be the same. j 12 MR. MURPHY: At this stage, we are proposing to , R change the name from " expected maximum earthquake" to -- 13 ( , l 14 MR. SIESS: To unexpected? l No! No. Whoa. Never. To something l 15 MR. MURPHY: l 16 we are calling a " deterministic source earthquake". l .17. MR. SIESS: Okay, now. Where is the definition? l . MR. CHOKSHI: On page B-2. Beginning line 35. I 18 19 MR. LEWIS: But it is the expected maximum l l l 20 earthquake? MR. CHOKSHI: -It is. The name will change. 21 22 MR. LEWIS: They are just going to hide that one
- 23 layer further down.
24 MR. MURPHY: Well, at this time, the expected 25 maximum earthquake is only used -- it's a term only used or i
117 ,( 1 associated with the deterministic analysis. l 2 MR. LEWIS: I understand. I
- 3 MR. MURPHY
- Okay.
It better be. l 4 MR. LEWIS: 5 MR. SIESS: Where it is? Page what? 6 MR. CHOKSHI: B-2. 7 MR. SIESS: Of what? l 8 MR. CHOKSHI: Of Appendix B. ;
)
9 MR. SIESS: Appendix B. Page 2. 10 MR. CHOKSHI: The last paragraph. j 11 MR. MURPHY: He's looking in the reg guide. Oh, I'm looking at the reg guide. I 12 MR. SIESS: 13 should be looking at what? l (- .
)
14 MR. CHOKSHI: I was looking at the rule. 1 15 MR. LEWIS: Page 15 of -- shall I just read it to 16 you? 17 MR. SIESS: Yes. 18 MR. LEWIS: It says, "An expected maximum 19 earthquake" is an earthquake that can reasonably be expected 20 to occur in a given seismic source and a current tectonic EMI is characterized by its magnitude. EMIs are 21 system.
.32 not necessarily associated with a given return."
23 MR. SIESS: But the operatives are " reasonably be l 34 expected to occur"? 25 MR. LEWIS: That's what it says. , 1
._._______m_. _ . _ _ __. _ . - _ . _ _ . . . . _ _ _ _ _ _ _ . - . . _ . . _ . . . _ . _ _ _ _ _ _ . . _ .
i l 1 118 ( { 1 MR. SIESS: And all that has to happen is for the
, 2- applicant and his engineers and the staff and their-3' engineers to agree on what's reasonable?
l 4 MR. LEWIS: What matters is what the staff things t
'5 is reasonable. Whatever you end up finally calling it.
1 They usually agree.
- 6 MR. SIESS
- 7. MR. MURPHY: Okay. Any"further questions on this l 8 viewgraph?
4 9 MR. SIESS: Well, I think so. The probabilities 10 associated with it explicitly -- except I don't know how 11 anybody is going to determine the maximum that can
- 12 reasonably *oe expected without somehow thinking l 13 probabilistically. l
( 14 MR. LEWIS: Well, of course. They put in the 15 exact word, saying it has nothing to do with the return 16 period, but reasonably expected clearly has to do with the l 17 return period. 18 So,-expected to beiong? They really are 19' obfuscating something whereas the job in doing this should 20 be to clarify it. 21 MR. MURPHY: Okay. 22 MR. LEWIS: I think.
..23 MR. SIESS:- And really what you are saying now, 24 getting back to this question of where the probabilistic 25 approach helps you.
l
I l l l l f 119 l 1 If, after you and the licensee have -- as ] 2 reasonable people have decided what you think is the l l 3 expected maximum earthquake there, if some intervenor comes l 4 in, and he's unreasonable, so he expects a higher l l 5 earthquake. ! 6 And so now you're going to say I'll get my experts l 7 to decide on the probability of that earthquake rather than j 8 on the reasonableness of the people who selected it. 9 But what's to keep him from getting his experts to 10 disagree on the probability? j
.11 MR. MURPHY: Okay --
l l 12 MR. SIES3: Before all were disagreeing with was i Why can't you have a similar disagreement l( 13 on the magnitude. 14 on the probability? 15 MR. MURPHY: Okay. With the probabilistic 16 analysis -- 17 MR. SIESS: You weigh them, right? 18 MR. MURPHYi You weigh them to some extent.- But l 19 the probabilistic analysis may already have taken into 30 consideration -- this specific earthquake. 21 Our experts, Livermore's experts, or EPRI's 22 experts, may have indeed said that there is some probability 23 that this size earthquake that has been proposed by the 24 intervenor -- ! 25 MR. SIESS: But you see, this only takes care of l l
t ! l t 120 1 the case where the difference of opinion has to do with i 3 magnitude. j 3 When you already took it into account, you 4 assigned a probability to it. 5 MR. MURPHY: We assigned -- in that case, we j 6 had -- l 7 MR. SIESS: Suppose the other person assigns a 8 different pro.bability to it? You can't -- if all he',s j
\
9 willing to say is yes, that could be, you didn't think of j l' 10 the biggest earthquake -- I can put a 9 there, and the I I l 11 biggest thing you thought of.was an 8. 12 And you could say, oh, no. I thought of a 9, but ( 13 I assigned it a low probability. He says, oh, but I 14 assigned it a high probability. 15 Now how have you solved your problem? 16 MR. MURPHY: Depending upon how high a probability 17 has been assigned, it is possible to inject this earthquake 18 into the probabilistic analysis that has been carried out, 19 and determine the sensitivity of the results. The hazard-20 curve that comes out. The spectra -- to that earthquake. 21 Obviously, it can go either one way -- it either 22 is significant to that seismic hazard curve and the L 23 'resulting spectra, or it is not.
- 24 And then, if it is not insignificant, you do again
' 25 have to address that.
. - . . - . . - . -. -._ -.= . - . .- - -
I i l 121 I 1 And again, you are going on the weight of the 2 evidence. If in effect one expert's opinion versus'16 3 experts' opinions in a process'that has gotten a review, and i 4 to some extent, an acceptance by the scientific community as 5 a whole. l 6 MR. SIESS: Well, yes. That's all very good to --
'7 but I'm not sure it's good in court.
8 MR. LEWIS: I don't understand that at all. ! 9 Because if Chet's expert and your expert differ about the 10 probability, and you haven't set any minimum probability 11 that is acceptable, then you are arguing about a number. I la But it's a number which has no role in the ( 13 licensing process, because you're not licensing according to 14 probability. 15 I don't see how you can win that one. Or am I 16 missing the point? 17 MR. MURPHY: Here we're talking -- you're talking 18 about a deterministic case. 'I'm talking -- I think, in 19 answering Chet's question -- at least as I understood it, he 20 was saying okay, now, we've given you the use of the l 21 probabilistic analysis. What effect is it going to have on 22 this particular case? 23 MR. LEWIS: No. I'm talking the probabilistic 24 case. Chet is saying that when you argue about magnitude, 25 it's the same as arguing about probability. l l
.._ _- - ..- . . - .. . . . - . .. - - .- -- . . . . . . ~
i i 122 l l( t i L 1 .When you say " reasonably" expect. 2- MR. SIESS: He's right. He was trying to-answer 3 my question'as to how the probability approach would get him ; r
~
4 out of a deterministic problem. 5 MR. LEWIS: I understand. I understood that. You i - - l 6 were saying that the -- let's see if we understand each l l 7 other -- if I understand you. 8 You are saying that in the deterministic 9 approach, the question of expected maximum earthquake is 10 open to Anterpretation. Someone will come in, and they 11 inevitably will, saying that.2 magnitude 35 is possible on 1 12 the site. Just to be whimsical about it. 13 And they will say, oh, no That has a very low ( 14 p'robability, and then the intervenor is going t'o say, ah, 35 but my expert say-it has a high probability. 16 .That's what the argument you posed is. ! 17- And then what I'm saying is that it isn't going to ! 18 do him any good to say my expert is'better than your expert, 19 because there-is no standard for the acceptable probability, l i !- 20
- or going back three steps for what is reasonable.
21 And that's the pickle they are putting themselves 22 into. 23 M9. . MURPHY: But there -- excuse me. lk -
- 24. MR. SIESS: Only three people at once. She has "I 25 trouble with four!
i i
!L l
'I 123 1 [ Laughter.] 2 MR. MURPHY: But there is -- and this is pushing l 3 the process. There is nome level of acceptance of the 4 criteria that went into the probabilistic analysis, okay? 5 And I'll hark bank to a study carried out by the 6 National Academy of Scione , and this does not prove 7 anything -- but, if you want, it bears to the weight of the 8 evidence. That kind of talk. 9 MR. LEWIS: They, of course, have no regulatory 10 responsibility. 11 MR. MURPHY: That's right. They have no 12 regulatory responsibility. They have examined -- I'll i i ( 13 say -- the two seismic hazard methodolog es that are 14 available e and have indicated that there are what they 15 have indicated -- they'll call characteristics of a good 16 seismic probability study. 17 And the.best thing that has come out of this 18 organi .lon is that the two studies that Were carried out.' la The EPRI and the Livermore study have these characteristics. 20 They do not explicitly say these are good seismic ! 21 hazard studies. They simply indicated that these studies l 22 has the characteristics that they said went along with a 23 good one. 24 MR. LEWIS: Well, these -- I understand that. 25 But --
Y 0 l l( 124 1 MR. CHOKSHI: May I answer? 2 MR. MURPHY: Go on. 3 MR.-CHOKSHI: There are two things I would like to L 4 say, as Andy mentioned. l l 5 The first thing, in the current deterministic l 6 process, one expert versus another expert is an issue. In l' 7 the probabilistic process, the way I think we see it is that 8 you are going to not only have two opinions, if there are 9 other opinions, they are there. l 10 And if it turns out that one opinion is l 11 overwhelming these, then you.have to assess that. I i 12 So I think that way, probabilistic puts everything 13 in perspective. And also not one parameter. You are (x- . . 14 dealing -- looking at the distribution of all parameters. 15 In this case -- 16 MR. SIESS: As I recall, there was a problem 17 deciding weights.
'18 MR. CHOKSHI: Right.
i 19 MR. SIESS: I don't care who your experts are, if 30 you let me assign weights to their opinions. It seems to me 21 you had that problem, didn't you? Of assigning -- whether 22 you assign weights? 23 MR. CHOKSHI: Yes. I did. But it becomes a j 24 little bit more or less diffused, I think, when you have all j 25 opinions into consideration, rather than one or two. l l __
a l( 125 1 Then you have to guess -- no type of situation. 2 MR. LEWIS: You are preaching to the choir on this 3 one, because I believe in putting all the evidence together 4 with its credibility. 5 And I once reviewed a German study in which they 6 consulted a group of experts, and in t!a questionnaire they 7 gave the experts, they gave a list of the other experts and 8 they said how would you assess the credibility of this other 9 expert. 10 And then they devised a weighting system according 11 to the consensus. You know,.you can go several levels with 12 this kind of thing. ( 13 But in the end, and I think this is the point you
- le were making -- in the end, the opinion of every expert is a l
15 piece of information. 16 MR. CHOKSHI: Right. And also you have other 17 uncertain parameters. In deterministic, you tend to focus l 18 on one. 19 And so, I think that probabilistic is a big help, j 20 looking at the total picture that way. i ( 21 One second point about the probability criterion. I
- 22 I think we have -- suppose you had been given a probability.
23 If you go past the debate and say your probabilistic models 24 are right, then the probability of exceedance defines the 25 level of motion.
1 t 126 t ( I 1- And if then the recurrence period and the ; 1 2 probability of an event happening are built into that. , 3 So I think that we have -- l MR. LEWIS: I understand that. I think the place l 4 1 5 where Chet and I may part company -- the place where we may , i 6 agree is, it's hard to see the justification for doing both. l l 7 And the place where we may part company is you 8 have always been asking your question in terms of why do the 9 probabilistic, and I ask why do the deterministic. l 10 -MR. SIESS: I think that there are advantages in l l' reaching a good decision in bringing to. bear all of the 11 12 information we have, and that includes the probabilistic 13 information. (* ' 14 One of the disadvantages'with 'that in the past has 15 been with the two sets of probabilistic approaches being j j 16 different by two or three orders of magnitude, it's been ! 17 rather awkward to bring it in. 18 You have found a way around that, I.think. . I'm not sure I disagree with Hal. I'm not sure 19 these are two different things. I think what may be ! 30 i I 21 muddying the water is calling them two separate things that 22 only come together down here. 23 I'll reserve an opinion on that until I have gone i 24 through the probabilistic approach. Now that I have this i This is the old way we f 25 one, .T think I understand this one. l a i i
127 ( 1 used to do it, and it did lead to arguments. p 2 I am not convinced that doing it probabilistically 3 is going to get away from arguments, but -- L 4 MR. MURPHY: We are only trying to get fewer 5 arguments and less time-consuming arguments. What we are 6 trying to do is control that, bring that into a more 7 reasonable arena. 8 MR. SIESS: Well, I think we are going to have to 9 hear more about the probabilistic approach. 10 When I read this the first time, I had a note on 11 it that there is no way anybody is going to be able to come 12 in with a number and have it accepted. l( 13 I could see the iteration, built into this, just s . 14 like it was before. Now ', hat was because I didn't see it 15 :very clearly. 16 And I'm hoping that once I see it clearly, I won't 17 see that iteration, that you and the licensees are going to 18 come together real quick,'lik~e that, and especially if it's 19 a 0.3g plant, you know. 20 Okay. 21 MR. MURPHY: Okay? 32 MR. SIESS: Where are we? 23 MR. MURPHY: I think we just finished geograph 4. l 24 MR. SIESS: Okay. Now, I'm just debating whether l 25 to let you go on for the next 20 minutes and stop you f i l
]
l 128 [ 1 there -- do you think you'd have a good stopping point in i 2 -about 20 minutes? 3 MR. MURPHY: I think I do, but in part it depends 4 upon the level of questioning. 5 MR. SIESS: Okay. Go ahead. We'll take care of 6 the level of questioning. 7 MR. LEWIS: The level of questioning will diminish 8 because I have to go. 9 (Laughter.] i 10 MR. SIESS: I think the next slide is fairly 11 straightforward. , 12 (Slide.] ( 13 MR. MURPHY: Okay.. I hope the next three slides 14 are fairly straightforward. 15 And all I'm intending -- 16 MR. SIESS: I wouldn't go that far! 17 [ Laughter.] 18 MR.' MURPHY: I said "I hope"! 19 [ Laughter.] 20 MR. MURPHY: This viewgraph is nearly a classical al viewgraph, depicting the four steps that are involved in I 22 carrying out a probabilistic seismic hazard calculation. 23 The first is simply defining the seismic sources
- 34 and characterizing the seismic sources in the second bullet !
i 25 of its recurrence rates. You've got the magnitudes here. t __ _ _
t 129 ( 1 And then talk about its recurrence in the second 3 panel. 3 The third panel is the continuation 4 characteristics that you are putting between the seismic 5- sources and the site. 6 These are manipulated to obtain a seismic hazard 7 curve, which in effect is simply the probability of the 8 exceedance of a particular ground level. There is an extra 9 decimal point in one of those accelerations. They do vary. 10 MR. SIESS: It does increase to the right? 11 MR. MURPHY: It does increase to the right. 12 MR. SIESS: Now, from my site near Indianapolis, ( 13 the information in (a) comes from where? 14 MR. MURPHY: The information about'(a) comes 15 from -- and this will be a single case -- it will come from 16 expert 1. 17 You will also have information about cite a if he
.18 chooses from expert 2.
19 MR. SIESS: No, I'm talking about little (a). That whole upper left end figure. That's what you -- that's 20 21 one -- that the geometry of homogeneous source zones 22 seismicity. 23 Where do I get that information for my site? 24 MR. MURPHY: For the probabilistic analysis, that ! 25' information already exists. I l
130 l( 1 MR. SIESS: Where? 2 MR. MURPHY: That exists in -- 3 MR. SIESS: In a computer? 4 MR. MURPHY: In Don Bernreuter's archives and in l 5 the EPRI archives. 6 MR. SIESS: And I could get hold of it? j l 7 MR. MURPHY: You could get hold of it. l 8 MR. SIESS: My contractor, my consultant -- . 9 MR. MURPHY: The contractor would get ahold of it. l l 10 MR. SIESS: The same for the next item? : i j 11 MR. MURPHY: The same for the next item. The same l 12 for all three items. All of that's available? For any ( ( MR. SIESS: , 13 i
)
l 14 site -- i MR. MURPHY: All of that's available. With one ; 15 \ 16 exception. You have to come up with your coordinates of the i l l 17 site. !
~
18 MR. SIESS: Okay. 19 MR. MURPHY: And you have to come up with the 1 20 geotechnical specifications for that site. ! 21 MR. SIESS: Yes. 22 MR. LINDBLAD: Andy, there seems to be a polygon . 23 around the site. Is that a source (c) as well? l MR. MURPHY: No. Well, it could be. But it has l, 24 not been specifically called. It virtually all cases -- I
! 25 1
l 131 [ 1 think in all cases in these United States, there is a 3 background zone around that site. 3 Some of those may be' specific in that they may 4 be -- in the Charleston area, this plant -- we're at 5 Charleston -- there would be a Charleston seismic zone in 6 -area (c) of the polygon. 7 But this is simply a generic one, at this stage. MR. MICHELSON: But those are not comparable to 8 9 the deterministic site sources like you had on your previous You know -- source 1, 2, 3, 4. , 10 slide. 11 Source 1 is -- could be in there, but I -- (a) and 12 (b) are comparable to some other sources like 2 and 3 in th,e ( 13 deterministic examination. - MR. MURPHY: In this figure -- sources (a) and (b)
^ .
14 15 are comparable to 2 and 3. 16 MR. MICHELSON: Okay. 17 MR. MURPHY: The source that Bill Lindblad was
' talking about"-- this polygon around them, would be 18 19 equivalent to seismic source 1 in the other viewgraph.
Right. That's what I thought, 30 MR. SIESS: Yes. al except that these would now be characterized not just by a 22 magnitude, but by a probability. . 23 MR. MICHELSON: Uh-huh. t 24 MR. MURPHY: That's correct. Somebody's probability. e ! 25 MR. MICHELSON: i' t , . . _
2 .. -. . .. .._._ . _ . . _ _ _.. . _ . _ _ _ . . . . _ _ . . . . . _ . . _ . . . _ _ . . _ _ . _ _ . _ _ . . l l-132 [ 1 MR. SIESS: Livermore. 2- MR. MURPHY: Whoever. drew the figure (a) -- little l 3 (a), would.also have in the past' assigned the recurrence 4 parameters over in little (b). l 5 MR. SIESS: Wait: a minute. On that the upper 6 righthand figure, the magnitude is plotted horizontally. Is l 7 that right? S MR. MURPHY: That's correct. And its cumulative 9 number of events at that magnitude are plotted vertically. l 10
- MR. SIESS: I've got it. It says recurrence, and j'
I was trying to figure out what "M" was in terms of i 11 ( 12 recurrence. I MR. MURPHY: No. This is the way the '( 13 14 seismologists work. 15 MR. SIESS: "N" is the number of events since 16 what? 17 MR. MURPHY: ,It is the. number of events with -- 18 greater than the "M". It's a cumulative number. 19 MR. CHOKSHI: It is the recurrence frequency. 20 Right. l 21 MR. SIESS: So it's not recurrence. The slope of its recurrence. I mean, 22 MR. MURPHY:
- 23 you derive it.
~ 24 Any other questions on that one? . 25 [No response.] I i 4 4
I 133 [ 1 MR. SIESS: I always understood the plot on that l 2 one, but I just didn't understand what it was called. 3 (Slide.] 4 MR. MURPHY: Okay, then.. The next two viewgraphs 5 will display what I call two of the principal products of 6 this probabilistic analysis. 7 This one here is simply a larger version of Panel 8- (d). Little panel (d). Panel little (d), where we have 9 plotted peak acceleration versus the annual probability of 10 exceedance. 11 MR. SIESS: This is for -- 12 MR. CARROLL: And to avoid mistakes, you'didn't l 13 label the horizontal curve, right? !
.(~
14 MR. MURPHY: Yes. 15 MR. SIESS: Now, this is for my. site? ; MR. MURPHY: This is for your site. This already-16 If you are going to build an advance reactor at an
~
17 exists. , , 18 existing site. 19 MR. SIESS: And I am now using PGA? 20 MR. MURPHY: You are now using PGA, in this l al particular plot. l 32 MR. SIESS: I've not no spectrum or anything to l 23 worry about yet? ! 24 MR. MURPHY: Not yet. I 25 MR. SIESS: Okay. f I l
. . . . ~ _ _ _ . - _ . - - . _ . _ _ - . - - - .-
i I I i 1 i 134 ( 1 MR. MURPHY: Okay? l 2 MR. SIESS: Now, wait a minute. Go back to that f 3 four-part one a minute. 4 MR. MURPHY: Okay. l 5 [ Slide.] 6 MR. SIESS: Item (b). Upper righthand corner. , l 7 That's for the site? l 8 MR. MURPHY: No. T, hat would be associated -- 1 l 9 there would be one of these plots for each of the seismic 10 sources (a) and (b). l 13 MR. SIESS: Okay. ; 1 l 12 MR. SIESS: And item (c) ? Attenuation. l 13 MR. MURPHY: That would normally be available for ( 14 the area between'(b) and the site. 15 MR. SIESS: It would be different for each source, 16 though? 17 MR. MURPHY:' It might be.
$8 MR. SIESS: It might be.
19 MR. MURPHY: It may be such that we are talking 30 about a regional model. ( 21- MR. SIESS: Oh, yes. I know. i l 22 MR. LINDBLAD: Is this accommodation specific for i 23 the site, rock or soil or something?
; 24 MR. MURPHY: If that information was available and 25 was available for the existing calculations, yes, it would
135 ! (- i l 1 be specific. l 2 Very general, but it would be specific. 3 MR. SIESS: Now, there's a little simplification 4 involved in an area there that says "by integration". What 5 did we integrate? How did we integrate it? 6- MR. MURPHY: I wco.1d say -- 7 MR. SIESS: I've got -- 8 MR. CHOKSHI: I think that j ust -- 9 MR. SIESS: I've got something called "Y". l i l 10 MR. MURPHY: "Y". l l 11 MR. CHOKSHI: Yes. Because it shows one curve is 12 integration of all the sources and all the distances. 13 MR. SIESS: I've got -- on (c) I've got a distance (~" . ? . and a ground motion intensity, and that's simply an _14 15 attenuation curve? And I've got them for three different 16 magnitudes? Are those three different magnitudes any ( 17 relation at all to sites at sources (a) and (b)?. MR. MURPHY: Not specifically.. If in the specific 18' 19 site (a), the maximum magnitude associated in there was 20 associated with a magnitude 7, we would be using the 21 attenuation curves for magnitude 7 to carry out the 22 calculations. 23 If it were magnitude 5', and you had the 4-24 information, you would be using the attenuation curves t 25 characteristic of magnitude 5 earthquake.
i 136 f( 1 MR. SIESS: But you just said, I thought, that the , 2 attenuation curves would be different for (a) and (b). 3 MR. MURPHY: Could be different for (a) and (b).
- 4. MR. SIESS: Could be different. Let's say they 5 are different. Now, there's only one magnitude associated 6 with (a), right?
7 MR. MURPHY: Let's put it this way. If we are 8 talking -- if figure (c) represents a regional attenuation , 9 curve, and the two sites had the same magnitude -- maximum 10 magnitude earthquake, they would be the same. 11 MR. SIESS: Let's don't talk about what's the 12 same. 13 MR. M U R P!I Y : All right. If the two -- ( 14 MR. SIESS: Let's try it my way. 15 MR. MURPHY: Okay. 16 MR. SIESS: Let's take site (a). And site (a) has 17 a magnitude 5. And there's some attenuation curve that relates site (a) -- sector A to the site from the nearest 18* 19 point. 20 MR. MURPHY: Uh-huh. 21 MR. SIESS: Site (b), location (b) of zone B has a 22 magnitude 7. 23 MR. MURPHY: Right. ! 24 MR. SIESS: And there's a different attenuation l 25 curve? I l l [
l-137 l(- 1 MR. MURPHY: That's right. 2 MR. SIESS: Now, what will your (c) look like for 3 those -- for this case? I don't expect to see a "6" on 4 there because I don't have a "6". ! '5 MR. MURPHY: For this case, for the site -- the 6 source zone with the magnitude 7 in it, you would use 7 attenuation curves appropriate for magnitude 7 in that l 8- region. 9 MR. SIESS: Okay. 10 MR. MURPHY: So that would be 7 -- the one labeled 11 "7". 12 MR. STESS: That would be one particular curve? l,( 13 MR. MURPHY: That's,right. 14 MR. SIESS: That I could label (b)?
- 15 MR. BERNREUTER
- Don Bernreuter, Lawrence 16 Livermore National Laboratories.
l 17 I think when you do the integration of the 18 probabilistic analys~is, you have' to use all of the 19 earthquakes. 20 MR. SIESS: I haven't got to an integration yet. 21 I'm trying to trace through -- I'm trying to take an example t 22 of one site -- a real site that I just named for you -- and 23 to take me through the steps.
, 24 And I don't want to go through steps that say, go 25 from one curve with three on it by integration to a curve i
i l t i I. i g 138 1 with one on it, where I don't have the slightest idea -- 2 MR. CHOKSHI: I think what -- 3 MR. SIESS: I want you to take some examples 4 through here and show me what they look like. 5 MR. MURPHY: Okay. )
)
6 MR. SIESS: I want to know how much trouble this ; 7 is. 8 MR. MURPHY: Okay. When it was originally done, 9 it was a lot of trouble. It was a lot of work. J So what happens 10 But now it's already been done. 11' is, you run a series of -- 12 MR. SIESS: It's already been done for my site at 13 India.napolis?
.( .
14 MR. MURPHY: The development of the process for 15 the existing sites has been done. Now, that development j 16 does not have to be repeated for your site in Indianapolis. 17 .So what happens is you have a site (a) -- excuse 18
~
me. A source A, that has a maximum magnitude earthquake 19 associated with it, and there is a distribution associated 20 with that maximum magnitude earthquake. i l al Also associated with that source is a recurrence 22 relationship. Okay? 23 MR. SIESS: That's difference than a distribution? 24 MR. MURPHY: A distribution. Intended to be the 1
-25 same meaning. Okay?
l 1
I i 139 l( l 1 And then there is an attenuation relationship that l l 2 is magnitude -- it may be hagnitude-dependent. j 1 3 MR. SIESS: Now, wait a minute. Do you have that ! 4 attenuation relationship developed for any site in the 5 country? 6 MR. MURPHY: You have -- basically you have , ! 7 that -- 8 MR. SIESS: I'm not in source A, now. I'm outside 9 of source A. 10 MR. MURPHY: The -- in the east -- 11 MR. SIESS: I'm just attenuated from the edge of l l 12 source A over to that "X". In the eastern United States, for i [%. 13 MR.. MURPHY: 6 the'se two s'tudies, you h' ave attenuation relationships 14 15 specified for all locations in the eastern United States. 16 It may be a region -- a very large regional 17 attenuation relationship, or it may be a smaller 18 relationship. 19 MR. SIESS: I can take my site, and I can take a 30 source A, and I can find an attenuation relationship that l 21 when I bring it in to you, you are going to accept it? I l 22 MR. MURPHY: No. 23 MR. SIESS: No argument? 24 MR. MURPHY: No. That's not what -- 25 MR. SIESS: That's not known? I'm sorry -- 1 + 4 i
. .. _. . _ __ . _ . _ . . .-~ . . _ . _. __ _ _
b L 140 ( 1 MR. MURPHY: No! Excuse me, Chet, for a second. 2 What we are telling you is that for the probabilistic 3 analysis that we are requiring, that there is a database, 4 which includes maps of the source zones, it includes tho l 5 recurrence information, and it includes the attenuation l l 6 information. 7 That's already specified. 8 MR. SIESS: Yes. t 9 MR. SHAO: But the answer is yes. 10 MR. MURPHY: The answer is yes. The information l 11 is there. If you are doing EPRI or Livermore, there will be I 12 no argument on those points. i ! (; 13 MR. ROTHMAN: Could I -- 14' You don't have the option of changing' that i l l 15 information. I I 16 MR. SIESS: I didn't say that. 17 MR. ROTHMAN: Could I -- I 18 MR. SIESS: I said I need an' attenuation l 19 relationship from the nearest point on source A to wy site. 30 And I said how do I get it? And you said the inforea'. ion is 21 there. 22 MR. MURPHY: Yes, sir. 23 MR. SIESS: And I said, could I go to that 24- information and get a number, and bring it to you and you'll f 25 accept it? 1 k
l
/ - 141
( 1 MR. ROTHMAN: Can I say something, please? This ) 2 is Bob Rothman. 3- Right now, we have the capability -- if you were 4 to give me the latitude and longitude of the site you are 5 interested in, we can run the Livermore program, and I don't 6 know how long it takes -- 20 minutes or half an hour -- and f 7 give you the hazard curves for your site. 8- With the attenuation relationships that-are built 9 into that program, the people at EPRI could do the same 10 thing. You give them a latitude and a longitude, and they 11 will give you the hazard curve for that site -- within -- in l 12 ttus time it takes to run the computer program. 13 Without changing any of the data. ( 14 MR.'SIESS: 'I can get all the way around here -- r 15 MR. ROTHMAN: Chat's been done already, for the 16 eastern United States. 17 MR. SIESS: I can get all the way around here with 18 a computer program? 19 MR. ROTHMAN: The data gathering -- that's already I l 20 been done. This has been done. 21 MR. SIESS: I gather with two computer prograins? 22 MR. ROTEMAN: Two separate computer programs. 23 That's right. 24 So right now, we could -- you could give us I 25 Indianapolis, Indiana, and running the computer program, we 4
I i 142 j( 1 could give you the hazard curve for whatever return period 2 you want. 3 MR. SIESS: And EPRI can do the same thing? 4 MR. MURPHY: Exactly. 5 MR. SIESS: So to get to the hazard curve, that 6 curve down in the corner, all I've got to do is go to the 7 right person and get a computer program, right?
*8 MR. MURPHY: That's right.
9 MR. SIESS: Submit that to you along with the 10 information and that we've agreed on? l 11 MR. MURPHY: That we have agreed on. , 4 12 MR. SIESS: Okay. For my purposes right now, I I don't care. ( 13 don't need to know these steps.
^
At this 'atage? No. 14 MR. MURPHY: 15 MR. SIESS: I want to know how much trouble it is, 16 I now know. 17 MR. MURPHY: You don't care at.all. 18 MR. SIESS; Okay. With that good point to stop -- 19 there's something that I finally understand -- let's take an 20 hour. l 31 Let's be back at 1:00 o' clock. Okay? I 22 MR. MURPHY: Okay.
?,3 [Whereupon, at 12:00 p.m., the meeting was ! 24 adjourned to reconvene at 1:00 p.m. the same day.]
4 25 h
l 1 1 ! 1 l, 143 I 1 A FTERNO O N S ES S I O-N 2' [1:06 P.M.] 3 MR. SIESS: Andy, I think you've finished up slide 4 5, but the bottom lefthand figure there is a scismic hazard 5 curve, and then slide 6, which I assume is your next slide, 6 is a seismic hazard curve, right? 7 MR. MURPHY: Right. 8 MR. BERNREUTER: One is a log. I just want to 9 point that one is log linear. 10 MR. SIESS: That changes the shape, right? I 11 MR. BERNREUTER: Don Bernreuter.just pointed that 12 out. 13 MR. SIESS: I,was just going to ask you, why all, ( -
'I thoug'ht l 14 of a sudden it changed'the shape of'it.
15 everything came out a straight line in log-log. That's 16- close to it. 17 MR. BERNREUTER: That's a straight line, isn't it? 18 MR'. SIESS: Okay. Now, we've got - for this I 19 site, we now have arrived at a SSE spectrum, I guess. 20 MR. BERNREUTER: From the deterministic -- 21 MR. SIESS: From the deterministic so-called And that SSE spectrum in itself is now sufficient l 32 process. 23 to make a design, right? 124 MR. MURPHY: Yes, sir. )
! 25 MR. SIESS: If it's acceptable. Now we're in the 1
. _ . _ .__ _ . . . _ . . .. _ _ _ _ .-.m - _ . . . _ _ _ . _ . _ . ~ . . _ . - . . _ ._ _ . _ _ . _ _
l' 144 l[ probabilistic approach, and on your chart, we're still in 1
'l ;
i f 2 the first box under PA side. Conduct an EPRI-LL seismic -- 3 MR. MURPHY: You're in the first box. 1 l 4 MR. SIESS: It says conduct a seismic '.lazard ) l l 5 assessment, right? l 6 MR. MURPHY: Right. And --
~7 MR. SIESS: Now, is that the -- is the end product 8 of the seismic hazard assessment the seismic hazard curve or j 9 have we got more?
l 10 MR. MURPHY: We've got a little more to do. 11 MR. SIESS: Before.we compare? ! 12 MR. MURPHY: Before we compare. 13 MR. SIESS: Okay. I just wanted to get us up to l(x i
.i4 speed.
l 15 MR. MURPHY: Okay. 16 [ Slide.] 17 MR. MURPHY: We go from five now to 6 -- 6 just 18 being a~replot of what you saw in D-5. Okay? 19 MR. CARROLL: Well, tell me what exceedance means. 20 MR. MURPHY: What exceedance means? It means 21- simply that -- 22 MR. CARROLL: What happens if -- l 23 MR. MURPHY: If it happens? What happens to the 34 power plant? 35 MR. CARROLL: Yes. I t
, . - , -_ . _. , ~. . ._- -. . - .- - -. . ._. .- f t .. L i 145 1 MR. . MURPHY: Nothing yet. I It doesn't apply yet.
.2 MR. SIESS:
3 MR. MURPHY: We're not worried about it. All 4 we're saying'is -- in fact the spectral shape is higher for ] 5 the actual case that you are talking about where you have an 6 exceedance. It's simply higher than the one that's been 7 plotted. It doesn't say that. j 8' MR. SIESS: No. 9 MR. MURPHY: At a particular level. , 10 MR. SIESS: No. No. Right now we don't have.a 11 spectral shape. That's a peak ground acceleration. i I 12 MR. MURPHY: Peak ground acceleration -- s
.13 MR. SIESS: Is one point on a spectrum, (e- i ,
L 14 MR. MURPHY: Okay.- 15 MR. SIESS: It could be exceeded and no other 16 point exceeded. 17 MR. MURPHY: That's right. 18 MR. SIESS: Another point'could be exceeded and'it 19 wouldn't be. 30 MR. MURPHY: That's right. 21 MR. SIESS: So, all we know is that a measure of l l l 22 the seismic -- a measure of the ground motion that we refer 33 to as a big ground acceleration -- j 24 MR. MURPHY: In this figure, yes. 25 MR. SIESS: Would be greater. 4 4 ) i d' 6 9
l l 1 1 !( 146 1 1 MR. MURPHY: Would be greater, j 1 2 MR. SIESS: These particular values. 3 MR. MURPHY: Right. 4- MR. SIESS: Right? i I 5 MR. MURPHY: Exactly. j 6 MR. SIESS: Okay. ( l 7 MR. CHOKSHI: But I think this graph showed -- l l 8 MR. SIESS: In our -- the way we think about it, j I 9 we say the earthquake is greater. But it - - 10 MR. MURPHY: All right. l 11 MR. CHOKSHI: It's.one parameter. 4 12 MR. MURPHY: One piece of the earthquake may be ( - 13 greater. i4 MR. SIESS: Yes. 15 MR. MURPHY: Okay. 16 MR. LINDBLAD: Well, thank you, Andy. Hazard, 17 . then, is hazard just talking about exceedance? Or does the 18 reader get the feeling.'that' h'azard has something to do with 19 safety? 20 MR. SIESS: Good point. 21 MR. MURPHY: At this stage, let's just say hazard 22 has to do with hazard, and we are not specifically 23 addressing the safety issue at this stage. l 24 MR. CHOKSHI: Look at two sides. One side may l 25 have a higher hazard, but depending on what plant you put on i.
... .__..m_. _ _ . _ _ _ . _ . . . . _ . _ _ _ . _ . _ . . _ _ . _ - . - _ . _
l l; 147 L( ! 1 it, the safety may be different. i 2 MR. SIESS: The hazard is the input, and it -- 3- MR. MURPHY: The hazard is the input at this j 4 stage. l 5 MR. SIESS: Using the word " hazard" is -- just . l 6 like we used to have, hazard reports -- the licensing of l 7 plant, and then we changed it to make them safety reports. 8 (Laughter.) , ! 9- MR. SIESS: Well, don't laugh. It's being 10 reconsidered right now, isn't it? Tnere's something I saw l 11 that came out of the commissioner's meeting about whether we l 12 were referring to something -- probabilistic risk analysis. l ( 13 MR. CHOKSHI: ' Oh, yes. l . j 14 MR. MURPHY: Safety analysis. i 15 MR. CHOKSHI: Probabilistic safety analysis. 16 MR. SIESS: Somebody -- thank God Lewis isn't i f - 17 ' here -- somebody on the staff was trying to explain the - 1 I 18 difference in some convoluted way, and iL s just that --- 8 19 MR. CHOKSHI: There's no difference. 30 MR. SIESS: Risk and safety are the inverse of 21 each other. Just like we used to have hazards analysis, 22 when we now have a safety analysis. 23 All right, proceed. Okay. f 24 [ Slide.)- i 25 MR. MURPHY: Okay. Now we go to figure 7. And i t i 1 r-- ,- -
, n. , ,y y , - - ,.c -
- , . , , , - - - - - ~ ,
i 1 148 I ( l 1 figure 7 is just another way to plot the information that ) l 2 came out of the hazard study. 3 In this case it is something called a constant 4 percentile uniform hazard spectra. 5 MR. SIESS: Except that somewhere you have gotten 6 from a big ground acceleration to a velocity. 7 MR. MURPHY: To a velocity. Yes, sir.
.8 MR. SIESS: Yes.
9 MR. MURPHY: And we switched parameters on you for 10 a moment. 11 MR. SIESS: Now, what's the -- the PGA was arrived 12 at using the Livermore data. 13 MR. MURPHY: This was arrived at using the
~
'14 Livermore dat'a the same way. This is just another step.
15 MR. SIESS: You can get that out with velocities. 16 MR. CHOKSHI: Right. 17 MR. MURPHY: Yes, sir. Very definitely you can 18 .get this with velocities. 19 MR. CHOKSHI: The previous speaker -- 20 MR. SIESS: It takes more than one point on a 21 spectrum to go from acceleration to velocity, doesn't it? 22 MR. MURPHY: Yes, sir. To go to where we are now, l 23 yes, sir. l 1 24 MR, SIESS: Okay. 25 MR. MURPHY: Then again, this is simply another l l I l
l l 149 1 product that falls out of the calculations that are done by 2 the Livermore program. 3 MR. SIESS: Or the EPRI. 4 MR. MURPHY: Right. 5 MR. CHOKSHI: I think one thing that may be 6 important. This is just one way of showing that. There are 'T 7 a number of other ground motion parameters one can -- 8 velocity, spectral velocity, displacement. Whatever you 9 want. The average of sum. So that's -- 10 MR. SIESS: We are talking spectra now. Because 11 we've got periods on here. 12 MR. MURPHY: Right. Yes, sir. ( 13 MR. SIESS: And these are the periods of single 14 degree of freedom of the systems that we feed these 15 earthquakes into, you see. 16 MR. CHOKSHI: Yes, sir. ! i 17 MR. SIESS: So we have gone to a spectrum 18 representation of the earthquake. 19 MR. MURPHY: Right. 20 MR. SIESS: And that step has been left out. We 21 have now converted the spectrum -- from a spectrum series of 32 them you can get velocities, accelerations, displacements, 23 spectra -- anything you want. 4
! 34 MR. MURPHY: Right. Okay. Five, 6 and 7 were 25 intended simply to be a review of what happens by way of
t 150 !( 1 input and output on the probabilistic hazard assessments. 2 MR. SIESS: Now, suppose I wanted to use a purely 3 probabilistic approach? So, far I've got no basis. I've 4 just got a lot of information. Right. C 5 MR. MURPHY: 6 MR. SIESS: About hazards with different 7 probabilities. Right? 8 MR. MURPHY: Right. 9 MR. SIESS: And in fact, if I look at figure 7, I 10 have something that varies with the period. 11 MR. MURPHY: Right. 12 MR. SIESS: That the probability of exceeding a 13 certain velocity is different at one period than it is at ( , 14 another. 15 MR. MURPHY: Well, what you have plotted here are 16 constant percentile, meaning that if you follow a line over, 17 the return period on each of those lines is' constant for 18 that spectral shape. 19 So that one -- figure 1 is the 500-year return 20 spectra. Figure 2 is the 1,000 year. Then it goes 2,000, 21 3,000, 5,000. , 22 MR. SIESS: Okay. But at 10 hertz, I get one -- 23 MR. CHOKSHI: Velocities are different for the l l 24 same -- l 25 MR. SIESS: I get different velocities at that f
.- - . - - . - - . . . ~ - -... .~. -
1 l 151 l 1 probability depending on the frequency. ! 3 MR. CHOKSHI: Right. I 3 MR. MURPHY: Right. Okay? 4 MR. SIESS: I couldn't design for the same ( 5 probability of exceedance for all the parts of my systems? l- 6 MR. CHOKSHI: If you have everything single degree 1 { 7 of freedom, then that wouldn't be a problem, because you are l 8 designing everything to the s.ame probability. , 9 But if you have a multi-degree of freedom systems, i 10 which depends on the different parts of the spectra, and 11 things got -- beginning to get mixed up. la Pure probabilistic is maybe to select for example, .( 13 10,000 year period, and design to that spectra. 14 MR. SIESS: You want the same probability of 15 exceedance at all? For all parts of the structure? 16 MR. MURPHY: Well -- 17 MR. SIESS: For example, if relay. chatter is only 18 a problem.for frequencies above 20 hertz, and a-tank 19 overturning is only a problem for frequencies below 2 hertz, 20 do I want the same probability of failure for those two? 21 If not, why not? 22 (Pause.) 23 MR. CHOKSHI: If you had the same performance i 24 goal, you would want uniform probabilities for that -- so i i 35 that all parts of the plant are equal.
l l 152
- (
1 MR. SIESS: You really want that? l 3 MR. CHOKSHI: If you can ao that, you know -- it 3 will never happen. If you want to keep margin constant and 4 not have one thing more margin than other, then that will be 1 5 one -- l 6 MR. SIESS: How you're talking margins. Margins 7 aren't probabilistic. They're -- 8 MR. CHOKSHI: Because in the design standard.is 9 the same probability, and then you can -- 10 MR. SIESS: Well, that's true of anything. Not 11 just nuclear. 12 You see, what I am trying to get at, I guess, is ,( 13 it's so easy for us to think of probability of exceedance of-14 a pabticular ground motion, which is interesting, 15 particularly to a seismologist. 16 But it'n not really inportant. What we are 17 concerned with is the probability that there will be an
'18 earthquake that will cause enough damage to the plant to 19 cause a core melt to kill people, or hurt them.
20 And that's about four steps removed. 21 MR. CHOKSHI: Right. And I think if you go down, 22 you will see that when it comes to determining the actual 23 ground motion, we will resolve that. We want some ' 24 engineering perspective in selecting that-motion, because as 25 you said, there -- if you design a tank on this chart,
g 153 1 that's two different things. 2 MR. SIESS: Okay, now. This is my probabilistic 3 hazard analysis for my site, and interpret it in a bunch of 4 single degree of freedom oscillators. 5 MR. MURPHY: Right. 6 MR. SIESS: Which presumably is a reason for doing 7 it that way. 8 MR. MURPHY: Yes, sir. Which we, at this stage, 9 will not go into. 10 MR. SIESS: We will not build them, either. 11 MR. MUPPHY; Right. Now. la MR. SIESS: All right. ( 13 [ Slide.) 14 MR. MURPHY: The next viewgraph, as it says, is 15 the probability of exceedance of the design basis using the 16 median Livermore hazard estimates. 17 And what.we got'is a cumulative distribution 18 versus annual probabilit'y', and what we have here -- it'might 19 be better plotted with plant name or plant number, along the 20 vertical axis. It is simply the location of and the 21 probability of a median Livermore hazard curve noted down 22 here. 1 i i 23 And all we're doing is listing or ranking the 24 plants against each other.
~
25 MR. SIESS: Now -- if I want to know which plants l (
.- .- . . . - -. - .. .. - - . - - . - . - - . ~ _ _ . . . . . .- _ _ -._ _ . - -_ . ._ -
s l l i 154 1 will wear on that, is that in the literature somewhere? 3 MR. MURPHY: The information that went into this 3 thing is definitely in the literature. It can be gleaned 4 out of the Livermore reports or the EPRI if you want. . I 5 It's not nicely categorized. I just felt like j 6 doing it like this., 7 MR. SIESS: I'm just curious. If I went in and 8 looked at the plants that were lic,ensed 20 years ago versus 9 plants that were licenses 10 years ago -- I was going to say l 10 five -- would I 1 -,d the more recent plants -- 11 MR. CHOKSHI: Lower half. l 12 MR. MURPHY: At the lower.
,(
- 13 MR. CHOKSHI: What you find in the lower half is
'14 the sites like in the gulf coast of Florida and the plants 15 designed with reg guide 1.60 spectra, tend to be'more j 16 recent. !
17 MR. MURPHY: On the lower part. - l 18 MR. CHOKSHI: Tehd'to be'more on the lower h'alf. 19 MR. SIESS: Those where the minimum govern. They 20 would be pushed over there. They would be skewed because al they weren't in any system. 22 MR. CHOKSHI: Right. 23 MR. SIESS: And if you go back to the Housner ; e i l 24 spectra -- 25 MR. CHOKSHI: Those tend to come out and certain ) l 1
e l i 155 1 soil conditions tend to put them on the upper half. 2 MR. SIESS: But now these probabilities that are , 3 plotted here are the same kind'of probabilities we were 4 looking at on the previous slide. 5 MR. MURPHY: On slide 6, yes. It's actually the 6 same. 7 MR. SIESS: On slide 6 1 didn't have a spectrum, 8 did I? 9 MR. MURPHY: No. You don't have a spectrum here. 10 MR. SIESS: Well, why would these others differ 11 depending on whether they were Housner spectra or reg guide 12 1.60 spectra? l( 13 MR. MURPHY: That's coincidental. 14 MR. CHOKSHI: No. This graph.is based'on the 15 comparing spectra -- average of spectral response, with the 16 value in 10 hertz. When we say -- 17 MR. S I.E G S : This one? 18 MR. MURPHY: This one right here. 19 MR. CHOKSHI: So, the acceleration design basis 20 means average response -- spectral response at value at 10 21 hertz. It's not PGA, okay? 22 We have taken all plants. If you have a Housner 33 spectra, and it's a 5 and 10 -- t 24 MR. SIESS: In Housner spectra -- these are 25 velocity or acceleration?
__ __ . ._ . ~ _ _ _ _ _ _ __ _ . . . _ _ _ _ _ _ . . _ _ _ _ _ . _ _ _ . . _ __ _ ._~ _ _ l l l I l i 156 1 MR. MURPHY: Pardon? 2 MR. SIESS: Velocity of acceleration? f i l 3 MR. MURPHY: It's acceleration but you -- 1 l 4 MR. CHOKSHI: You can convert'it. 1 5 MR. MURPHY: You can convert it, so it doesn't I 6 make that much difference. 7 MR. SIESS: Can't you give me the relationship i 8 between the_5 hertz, 10 hertz acceleration and the PGA? For i a 1.60 spectrum, that's a fixed -- 1 9 l l 10 MR. CHOKSHI:. It's a fixed relationship, right. 11 MR. SIESS: And it's a different relationship for la the Housner, right? 13 MR. CHOKSHI: That's right. ( 14' 'MR. SIESS: 'You could relat'e one to the other very I 15 easily. If I gave you a PGA and asked you for the average ; 16 of the 5 and 10 for the two cases, there would be a-fixed 17 relationship between them. 18 MR. CHOKSHI: Yes. 19 MR. SIESS: What is it? I assume the Housner is ; 20 lower? 21 MR. ROTHMAN: Yes. The Housner is about 22 equivalent to a 50th-percentile spectrum, and the Reg. Guide , 23 1.60 is equivalent to about an 84th percentile spectrum. 24 MR. SIESS: The whole spectrum is moved up. 25 MR. ROTHMAN: Except at the PGA, where they are {
i l 157 il 1 both anchored at the same point, but at all frequencies 2 below 33 hertz, the Reg. Guide 1.60 spectrum is higher. 3 MR. SIESS: Thank you. That's good. Okay. But 4 this is not the same as the one on the previous page, then. 5 MR. MURPHY: No. I mean the information that we 6 got from the probabilistic hazard analysis, the Livermore 7 stuff, went into compiling the list that was used and 8 plotted here. 9 MR. SIESS: Okay. I mean the ordering is going to 10 go from 0 to 1. The probabilities will be essentially the 11 same, however you get it. These were based on average 5 and 12 10 hertz off of spectra, Reg. Guide 1.60 spectra? 'l 13 MR. CHOKSHI: . Whatever the design spectra was. 14 MR. MURPHY: Whatever the' design was -- 15 MR. SIESS: Oh, whatever the design was. 16 MR. MURPHY: -- was for the plant, i 17 MR. SIESS: Okay. Okay. 18 MR. MURPHY: Okay. So, we simply, at this stage, 19 plotted them out. We've entered this at the 50-percent level and selected the result, 1 times 10 to the minus 4, as 20 al a target probability that we would like to meet with the new 22 facilities that will be licensed under Appendix B. i l 23 MR. SIESS: Now, what would that be if I had gone 24 through the whole process using the EPRI? 25 MR. MURPHY: If we had done the whole thing using I l
- . . - -. . -. -- - - - - .- - . - ~ . ..
l 158 l( 1 the EPRI, that'would be here in the next viewgraph, 9. ! 2 [ Slide.] 3 MR. MURPHY: The number would be 3 to the minus 5. 4 MR. SIESS: Okay. Thank you. ! 5 MR. LINDBLAD: Andy, on the previous viewgraph, l l 6 you show cumulative distribution of 0.5. That's another way 7 of sofing median. Are you saying this is the median of 1 8 medians? 9 MR. SHAO: For the Livormore curve and the EPRI l u 10 curv e , the difference was mean value. The median value was 11 never that much different. l I 12 MR. LINDBLAD: I'm not now distinguishing with ( 13 EPRI. I'm just looking at this Livermore presentation. We looked'at the Livermore median 14 MR. CHOKSHI': I i l 15 hazard. Based on that, we calculate probability of accident l 16 for each site, and that's what is plotted here. 17 MR. LINDBLAD: That's what's plotted. l
'18 MR. CHOKSHI: Right.
19 MR. LINDBLAD: And then the output of this, in 20 terms of setting some kind of a goal for new plants, is 21 taking the median of that distribution. 22 MR. CHOKSHI: Right. We said is the population 23 wholly safe, but is it setting.a target that it should be 24 similar to the lower half of the population. I 25 MR. LINDBLAD: Yes. I guess I'm saying that you
l 159 l[ 1 carefully used the word " median" in terms of the data 2 points, but actually the product of the distribution, when 3 you say 0.5 distribution, is another way of saying median of 4 that, as well. 5 MR. CHOKSHI: Right. 6 MR. LINDBLAD: And you're going to say why you 7 chose that 0.5. I understand why you plotted median data, 8 but now you're going to tell us why you took the median of 9 that distribution? 10 MR. SIESS: Wait a minute, because in the Reg. l 11 Guide, they've got plots of median, means, and 85 la percentiles. ,( 13 , MR. MURPHY: Yes, sir. If you'll forgive me, that 14 is bhat is in the copy that you hav'e go't, and by our 15 agreement, the copy that you got was where we stood on the 16 day it was issued. 17 MR. SIESS: Okay. Good.
'18 MR. MURPHY: We have specifically changed some 19 things that, at this stage, you do not have. One of the 30 things that we specifically changed was going from some sort 21 of an option, if you want, or discussion of mean, median, l
j 23 and 85th, to a staff selection of median only. I That ought to 23 MR. SIESS: Oh, that's interesting. l 24 leave out about eight pages, shouldn't it? 25 MR. MURPHY: Somewhere around there. r
I a 160 ( 1 MR. SIESS: That eliminates one set of questions. 3 So, that's helpful. 3 MR. MURPHY: Okay. That was also the thing that 4 was -- 5 MR. SIESS: You promise not to put it back in now. 6 MR. MURPHY: No. 7 MR. SIESS: Okay. 8 MR. MURPHY: But also, let me say that we are also 9 generating a set of questions, have generated some of them, 10 of questions that we want to go along with the document as 11 it goes out for public comment, and one of those is a 12 request for comment on our decision to use the medians. ( 13 We would like to have, specifically, public l'4 comment on'whether that's a good idea, a bad idea, whethe'r 15 we should have been using means or 85ths. 16 MR. SIESS: Well, you'll get some of each. 17 MR. MURPHY: We will certainly get comment en it, 18 and that's what we're -- you know,'this is for public 19 comment. 20 MR. SIESS: Now, of course, I have no j'udgement on al that until I know what we're going to do with this 10 to the 22 minus 4 figure, which would obviously be a different figure 33 if it was an 85 percentile, and that's in the Reg. Guide l 34 example somewhere. 25 MR. MURPHY: Yes, sir. 1
( 161 1 We'll also say that -- I think, to get back to 2 some of your question, we are conducting and have some very 3 preliminary results on looking'at something besides the 50th l l 4 percentile, took a quick look at the 25th and 75th, meaning 5 here and here, to see what difference that would make at the 0 final end, at either the controlling earthquake or at the 7 spectral ground notion, and preliminarily, it does not make 8 much difference. l 9 There is low sensitivity as long as you're, 10 quote / unquote, staying in this very steep part of this 11 graph, and that is very preliminary at this stage. 12 We just took a couple -- looked at a. couple.of I( 13 cases, and part of our operation between now and when this. 1 14 goes out for final is to firm up-those numbers. 15 MR. LINDBLAD: Well, after you use this process to 16 qualify site one, do you compare site two to the eastern 17 plus site one? 18
'MR . MURPHY: At this time, t'he way we are setting 19 this up, that would not happen.
20 MR. LINDBLAD: So, the eastern sites is a group of 31 < a fixed in time. 22 MR. MURPHY: At this time, yes. 23 MR. CHOKSHI: And that goes in part of answering l 34 question 50. If you put it at 100 percent, theoretically, f 25 as you go along, your distribution will approach that limit. I
i 162 l( l 1 So, I think you want to keep something below that. 2 MR. LINDBLAD: Thank you. l 3 MR. MURPHY: On this' basis, we come out with the ] l 4 median target level for the EPRI methodology of 1E to ' the i ( 5 minus 4 -- I'm sorry -- Livermore. l J l You have the same plot, you've said, I 6 MR. SIESS: l 7 for EPRI. , i 8 MR. MURPHY: Yes, sir. l 9 MR. SIESS: And of course, you pick a median off l 10 of it. It will be at a different probability. But would I 11 find roughly the -- I'm going to say it two ways -- would I l 13 find roughly the same plants in the upper and lower halves? 13 MR. CHOKSHI: Yes. Not exactly the same ranking ( l 14 but generally the same group, right. 15 MR. SIESS: Are there any big differences, any 16 plants that are 60 percent on one and 40 percent on another 17 or -- that's not a big difference. , 18 MR. M 'URPHY: There is not a' great deal of 19 difference between the two of them. I mean this is the same 20 arguments that we went through in the IPEEE process of 31 binning and ranking them. 32 MR. SIESS: Okay. And then I remember O'Hara's 23 paper has got some plots, and I thought there he had direct 24 comparison to the plants.
! 25 MR. MURPHY: There is a direct comparison in Tom l
163 ( 1 O'Hara's. 2 MR. SIESS: Okay. Fine. Don't go into it now. 3 MR. MURPHY: Okay. 4 MR. SIESS: I just wanted to know roughly. j 5 MR. CHOKSHI: I can show you later. 6 MR. SIESS: What's the highest plant -- what's the ! l 7 lowest probability, lowest -- 8 MR. MURPHY: Lowest? 9 MR. SIESS: Yes, the worst plant on there, way up l 10 on the top righthand side. Who would that be? E to the 11 minus 3. i ! 12 MR. MURPHY: Looks like 8 or 9E to the minus 2. t 13 MR. SIESS: Yes. What plant would that be? l( \ 14 MR. CHOKSHI: The top three plants coming out of 15 both, looks like Oconee, certain soil conditions, Pilgrim, 1 16 Maine Yankee. t 17 MR. CARROLL: Do you get that with both? 18 MR. CHOKSHI: Both EPRI and Livermore. 19 MR. SIESS: That's because of new seismic evidence 30 since they were -- 21 MR. CHOKSHI: Well, I think the soil conditions on 22 some of them, on the Oconee, and the spectra they used for 33 the design. l 24 MR. SIESS: Okay. Okay. l 35 MR. MURPHY: Okay? 4 l l l .
l i 164 ) 1 MR. SHAO: Some of the older plants, even though 2 the spectrum may be lower, but their damping value is also l 3 lower. 4 MR. SIESS: Where is Big Rock? It wasn't designed 5 for seismic. I just wondered where it is. 6 MR. MURPHY: Is it on your list? 7 MR. CHOKSHI: I'm looking. 8 MR. SIESS: I don't think you have it. 9 Go ahead, Andy. 10 MR. MURPHY: Okay. 11 So, given the previous -- 12 MR.' CHOKSHI: It shows up on the EPRI as the ( 13 bottom list. 14 MR. SIESS: What? 15 MR. CHOKSHI: On the EPRI, it's on the lower list. 16 MR. SIESS: How can it be there? It wasn't 17 designed for any earthquake. How do you get the probability 18 -- you had to have the design-basis earthquake or'the SSE in 19 there somewhere. 20 MR. ROTHMAN: Big Rock Point was an SCP plant, and 21 I think, for this study, they used the SCP spectrum as the 22 design basis. 23 MR. SIESS: Okay. You cheated a little. 24 The point is that here was a plant that wasn't 25 designed for seismic at all, and it doesn't come out on the l
^
1 i r i 165 l ( 1 upper end. It comes out on the lower end. 2 MR. MURPHY: Okay. 3 MR. SIESS: It tells me something. I don't know . 4 what. f 5 Are you going to explain the SGX versus 4GX? l 6 MR. MURPHY: Yes. 7 The Livermore study made use of five ground-motion 8 experts. There was some controversy associated with one of 9 the ground-motion experts' opinion about what the ground-10 motion model should be. 11 Because of this, some sensitivity studies were 12 done with the Livermore program in which the seismic hazard ( 13 calculations were run without making use of this fifth i 14 ground'-motion experts' models. 15 So, we end up with Livermore, five ground-motion 16 experts, and Liver, six ground-motion experts. 17 MR. SIESS: Were.there any technical, geological, ll 8 - seismological reasons for eliminating that one expert,
~
or 19 was it simply that he was an outlier? MR. MURPHY: That expert was not eliminated. We 30 al simply made a special effort to understand what 22 contributions that his models made to the hazard 23 calculations. We did not eliminate him. 34 MR. SIESS: In the 4GX line up there? 25 MR. MURPHY: When we say we did not eliminate him,
.. . .-_ ~ . - . - - - .- . - ..-_. . _ . - - . . - .- . _ .. ~ ... - - . _ - . - . - ._ - - .
[ 166 1 the' calculations and so forth, we do have, if you'll forgive , 2 me, two sets of Livermore hazard curves, j 3 MR. ,SIESS: I didn't mean elimination in the Mafia ! 4' sense. I meant, you know -- l 5 MR. MURPHY: Fair enough. He was isolated in the I 6 Amish sense, okay? 7 MR. SIESS: Now, again, was that based on 8 technical evidence or purely probabilistic evidence, that he 9 was an outlier? What made him different? 10 MR. MURPHY: There were individuals that
't 11 questioned his ground-motion-models, and that was on a 12 technical basis.
( 13 He made use, in part, of conversions of magnitude 14 -- of densities to magnitudes and did some of his- , 15 calculations and ideas that may not be the most current ; 16 today. J !
, 17- Ten years ago, when this in. formation was 18' originally collected' he was'in,the' forefront of ground-19 motion modeling and expertise.
20 MR. SIESS: Okay. 21 MR. MURPHY: The bottom line is we end up with two 22 target probabilities. 23 MR. SIESS: Excuse me. It just struck me, with 24 his model out of there, there is no difference between the i k 25 Livermore and EPRI median medians. a t 1
l'
)
1 ( 167 1 MR. MURPHY: In this particular measure, that is 2 absolutely correct. 3 MR. SIESS: This is the average of 5 and 10 hertz. 4 It's the median median. 5 MR. MURPHY: Yes. I l 6 MR. SIESS: Fascinating. The means are going to l 7 be way different. l l 8 MR. CHOKSHI: We wanted to put 4GX here, and the 9 original study was on the hazard and changes, what is the 10 impact on the ground motions and controlling earthquake. 11 So, you can see, if you were to use different l l .12 methods, what would be the outcome. l 'l +. 13 MR. SIESS: Now, with the means -- in the Reg. 14 Guide, the draft guide I had, you didn't have the 4G's, did i 15 you? 16 MR. CHOKSHI: No, we did not. l 17 .:MR. MURPHY: No, we did not, i 18 MR. SIESS: So, I can't really'tell how it would 19 be. ) l l 20 MR. MURPHY: At this moment, in the one that is ; 31 going around for circulation that has the means and the 22 85ths eliminated, there is not mention of the 4G expert 23 category either. This is simply here for your information. 1 i 24 MR. SIESS: Okay. I like the 4G. It agrees with } 25 the EPRI, and then we're home-free. i { i l
l i ( 168 l 1 MR. CARROLL: I think I know the answer, but the 2 expert you're talking about with respect to Livermore did 3 not participate in'the EPRI work. Is that correct? 4 MR. MURPHY: That is correct. There were l L 5 different experts. 6 MR. SIESS: A lot of differences. 7 MR. MURPHY: Yes, but none of the five that j 8 participated in the Livermore participated in the EPRI, as j 9 far as the ground motion is concerned. 10 MR. LINDBLAD: How many experts were in the EPRI 11 panel? i 13 MR. MURPHY: One. (- 13 MR. SIESS: EPRI grouped'them differently. It was 14 a different' approach. l ! 15 MR. MURPHY: Initially, there was one model for l 16 .. ground motion used by the EPRI calculations. The NRC staff, 17 in reviewing it, said give us some more information, and then t'wo addiElonal models were used. 18. 19 The first model was weighted, I believe, at 50 20 percent, and the two additional models were weighted at 25 j 21 percent apiece, so that, in effect, there were then three l 22 models used, but again, still basically only from one 33 expert. s l 24 [ Slide.] i 35 MR. MURPHY: Next, we take that -- for the A m <+-,-r, -, ,------,e---. .-w - ,- , - , . , . ,a.- -e . - - ,
- .- - --,-,-,s ,,.,\,---,.m . - , - - - - . - -r,- , -
_. ..~ .....__ _ . _. _ . . . _ . . . - _ _ . . _ _ . _ . - . ~ . - _ . . _ . . . _ . . - . . . _._... _ ._ _-. _ _ 169 (- 1 Livermore case, we take that 1E to the minus 4, go back to , 2 an actual spectral plot -- spectral velocity versus ; 3 probability of the exceedance plot and enter that at the 1 4 'to the 10 to the minus 4, follow that over, pick out a 5 ' spectral velocity, and in the case that we're going to be 6 talking through, for Vogtle, this simply turns out to be 8 i l l 7 centimeters per second. , T L ! 8 MR. SIESS: That would be based on Figure 7, then, ) 9 to get that spectral velocity exceedance. 10 MR. CHOKSHI: Something similar to that. This is 11' for not Vogtle but -- la MR. SIESS: No, but I mean that type of data. 13 MR. CHOKSHI: That kind of plot. j ( . . . , 14 MR. MURPHY: That kind of plot.
.15 MR. SIESS: Okay. Page 10 is for Vogtle? l l
16 MR. MURPHY: Page 10 is for Vogtle. , 17 MR. CHOKSHI: On the 8-centimeter-per-second-l . 18 ~value, 19 MR. MURPHY: Well, the 8-centimeter-per-second
.30 value -- we'll point out that the --
al MR. SIESS: Since there is no scale. 22 MR. MURPHY: -- the horizontal axis has only one
~
33 point on the scale. } 34 MR. SIESS: He learned his lesson about i 35 mislabeling scales earlier. l t
. . _ - . - _ _ _ - _ , . - , . . . , , _ . . ~ . _ , - - - - __ . . _ _ , ~ , _ .
.._._._._. ._ . _ . _ _ . - . _ . . . _ . _ _ _ . . _ . _ . _ . . . . _ . . _ _ _ _ _ . . . . ~ . _ _ _ _ _ , _ . _ _ _ . - _ _ _ .
1 1 1 l l 170' ( ] 1- .The 10 to the minus 4 was our median -- it's 2 better than half the plants out there, and now we are 3 looking at that number, which we got from the probabilistic 4 thing, to see what we would have to design for to get that, 5 right? 6 MR. MURPHY: That's right. For the median between 7 -- l l 8 MR. SIESS: Now, for my site in Indianapolis, how j 9 does - ,what would Figure 10 look like? 10 MR. MURPHY: Same exact thing except the 8 11 centimeters per second might be a different number. 12 MR. SIESS: I would have comparable data to Figure 13 7 come out of the computer. ( . i
-14 MR. M U R'P H Y : Right. And you would have, thIen, 1.
i 15 comparable data as here in Figure 10 come out of the 16 computer, as well. 17 MR. SIESS: Okay.
- 18) MR. M'URPHY: And again, using.all the nicely-19 canned input databases. ')
i 20 [ Slide.) , 1 21 MR. MURPHY: Now, at this time, we have got a i 22 target probability which we have nominally converted into a .. 33 target spectral velocity that we want to meet. t 24 We have calculated the seismic hazards for this i l 25 site, and the next thing we would like to do, propose doing, i i i i l - - . - , _ . - - . . _ _ . , . _ , . _ _ _ _ , _ . , . _ _ _ __. _
)
i 171 [ l 1 is to de-aggregate the seismic hazard at this site, your 2 site near Indianapolis, our site Vogtle, into different 3 magnitude and distance bins. 4 We would like to see what contributions those make l 5 to the overall seismic hazard at the proposed site. 6 MR. SIESS: Now, why do I need to de-aggregate 1 7 these? When were they aggregated? Who did the aggregating? ! l. l 8 MR. MURPHY: They were aggregated back in that I 9 four-panel figure. That aggregation or integration put them ! 10 all.together and gave you the seismic hazard curve. l ' l I 11 MR. SIESS: Okay. ! l i 12 MR. MURPHY: Now, we're interested in finding out } 13 what the contribution from these different magnitude l (' , 14 distance bins would be. 15 MR. SIESS: Seismic hazard curve I've got back l 16 there. It was for my site, Vogtle, if you wish, and that 17 gave me acceleration versus probability, right? 18 MR. MURPHY: Yes, sir. That was one of the 19 measured it gave you. 20 MR. SIESS: And I now have a probability -- at ( al that point, I didn't have an SSE. 32 MR. MURPHY: That's correct. 23 MR. SIESS: I still don't have an SSE. 24 MR. MURPHY: That's right.
- 35 MR. SIESS
- All I've got is this curve.
i 4
6 i 0 l i l j; 172 l I 1 MR. MURPHY: We do not have a probabilistic SSE 2 yet. 3 MR. SIESS: And now I've got a target probability, 4 and I want to work back from that to get an SSE. 5 MR. MURPHY: That's right. That's what we're
\
L -6 doing now. I 7 MR. SIESS: And I can't do that in terms of a ! 8 single spectrum. I don't have a single spectrum. [ 9 MR. CHOKSHI: Do you mean single frequency? 10 MR. SIESS: Single PGA or something, single 1 11 anything. 12 MR. CHOKSHI: We keep the same average of 5 and 10 ( , 13 hertz. That just gives you the magnitude and distance. j
\
14' MR. S I E'S S : Did that 5 to 10 hertz averaging come 15 in back here on this -- . l 16 MR. CHOKSHI: Yes, on page 10. 17 MR. MURPHY: -Yes. 18 MR. SIESS: No, back on page 5. See, back'there, l 19 I had an acceleration versus a probability for this site, 30 and of course, I had no basis for choosing either an 31 acceleration or a probability. 22 MR. CHOKSHI: Right. 23 MR. SIESS: Now, I've come up with a 10 to the t . 24 minus 4 as an acceptable probability, but that is a 35- probability of exceeding something, which is not the same {
\
.w 173
- (
1 thing I've got the probability of exceeding back there. 1 2 MR. CHOKSHI: Right. But you can get output from l l 3 a probabilistic analysis in terms of velocity, spectral ) I 4 velocity, acceleration, anything you want. So, you can view I 5 this on -- 6 MR. SIESS: All right. Suppose, back on page 5, I 7 had a spectral velocity. 8 MR. CHOKSHI: Right. 9 MR. SIESS: Now, why can't I then go in at 10 to 10 the minus 4 and pick off that value and say okay, that's my 11 SSE? 12 MR. CHOKSHI: That gives you one point on the 13 spectrum an average of five and ten hertz. Yes, you can do ( 14 that. , l 15 MR. SIESS: I can do it on PGA. 16 MR. CHOKSHI: Right. Right, you can. 17 MR. SIESS: Suppose I did it on PGA -- 18 MR. CHOKSHI: Right. 19 MR. SIESS: -- and then used that spectra for 20 design. Now, what's the deficiency in that process? 21 MR. CHOKSHI: Okay. The -- 22 MR. SIESS: Before, it was fairly simple. 23 MR. CHOKSHI: Yes. l 24 MR. SIESS: It was de-aggregated. l i 25 MR. CHOKSHI: In fact, I can show you the results. l r
i I 174 , [ 1 We also did that, too. If you go to a different firequency, l 2 pick up again the.50th percentile correspondi: q responses, , 3 and you can plot-that spectra, That spectra from an l 4 engineering viewpoint doesn't look anything like ground 5 motion, which we will design to. 6 MR. SIESS: Well, why don't I just take the Reg i 7 Guide 1.60 spectra and put the five and ten hertz values at 8 the two I get off of that figure and say that's it? It
'9 won't look like: anything. It'll look like a Reg Guide 1.60 10- spectrum.
Il MR. CHOKSHI: Yes. I think that's one -- you 12 know, in selecting SSE, that's definitely one option. 13 MR. SIESS: But couldn't I do that and then go l( back and compare it with what I'got deterministically? 14 i 15 Wouldn't I have done everything else? What else do I need l Why can't I stop there? j 16 to know besides that? 17 MR. CHOKSHI: The. reason -- what this does is, for example,. answers questions like'a large distance, how much
'18 19 it contributes. If someone is worried about Charleston, i j
30 then that is explicitly answered.
'31 MR. SIESS: I'm going to get that out of the 33 probabilistic?
23 MR. CHOKSHI: Yes. 24 MR. SIESS: It seems to me the deterministic is
- 25 where I'd get that.
i I
j.__._.. . . _ . . . . - _ ._ _ _ _._.m. _ . . . _ . _ - . _ . _ _ _ . _ _ _ _ _ . _ _ . _ . _ _ r i i $ 175 ,( l 1 MR. CHOKSHI: Well, here'you will get the 2 contribution from a Charleston earthquake or, say, a large l 3 magnitude earthquake. At that' frequency, what is the 4 contribution, what is the probability of exceeding that ; 5 level compared to contribution from other zones? That was, ; 6 I think, the thrust of developing this procedure, that you i 7 can see the sources, which contributes most at what ' ^ 8 distance. That way -- 4 9 MR. SIESS: But the sources we're dealing with in 10 the hazard analysis aren't the same sources we had for the 11 deterministic analysis. l I 12 'MR. CHOKSHI: Right. But the probabilistic, you > , ( 13 are looking at all kinds of different sources. .You're 14 looking at'a number of different'models, postulations. 15 MR. SIESS: So it's not just that we're using a 2 16 different technique; we're using a different model. 17 MR. CHOKSHI: It could be a different model in the deterministic and probabilistic,-but the'probabilistic has
~18 19 more than one model. It takes all different models out 20 there into account.
21 MR. SIESS: The only reason a deterministic 22 doesn't is because you don't do it. 23 MR. CHOKSHI: Well, somebody has decided that, you 24 know, as you said earlier, going through the mental process 25 of selecting one'or the other, and this is sort of -- you , l l
I I 176 l[ 1 know, just allows you to look at that process. 2 MR. SIESS: I think I understand what you're 3 saying it does. What I guess I'm still not sure is why we i 4 need to do it, and that keeps -- l 5 MR. CHOKSHI: If you go -- 6 MR. SIESS: If I started out saying I want to know 7 what the probablistic approach is, and second, how it helps 8 me. 9 MR. CHOKSHI: I think Page 13 is the crux of the 10 -- you know, is probably the most important information you 11 are getting out of this. 12 MR. SIESS: Okay. Let's go ahead, then. 13 MR. MURPHY: Okay. Figu.re 11 simply says, in (. 14 fact, 'that we want to de-aggregate, we want to figure out
~
15 the contribution, okay? We have assigned the bins that we 16 would like to find out the contribution in. 17 MR. SIESS: How do I do this? Do you have a 18 computer program for this? 19 MR. MURPHY: Pardon? 20 MR. SIESS: Do you have a computer program for al this?
'22 MR. MURPHY: We've already got one that exists.
23 MR. CHOKSHI: The same program. 34 MR. MURPHY: Same program. j I 25 MR. SIESS: Okay. l l l
i 177 lI l 1 MR. MURPHY: Okay. l 2 [ Slide.] l 3 MR. MURPHY: The next viewgraph, in fact, tells l 4 you how to do it, and basically what you do is you rerun the 5 Livermore -- we'll say Livermore in this case -- you rerun the Livermore case, and here, instead of letting the l 6 7 parameters vary over magnitude from five to seven end a half 8 or whatever the maximum is, you restrict it to these 9 particular bins. 10 You run it once for the Bin 5 to 5 and 1/2 at 0 to >
;11 25. You run it once for the. Bin 5 to 5 and 1/2 25 to 50, 12 and so on for the 25 bins that are shown here.
( 13 A reasonably simple process; take you maybe a day . worth of' Sun workstation computer time. Not an overwh'lming e 14 15 task. 16 MR. SIESS: That sounds like the way it was done 17 back in the very first -- what did we call this program out 18 at Livermore when it start $d? 19 MR. MURPHY: The -- 20 MR. SIESS: Seismic safety? 21 MR. MURPHY: SSMRP. 22 .MR. MURPHY: The SSMRP. f 23 MR. SIESS: Yes. 24 MR. CHOKSHI: It's a Monte Carlo. l And that was binning the 35 MR. SIESS: Yes. l
i i l l t 178 1 earthquakes. Is that right? 2 MR. CHOKSHI: Well, they had binning to the 3 acceleration ranges, right. 4 MR. SIESS: That led somebody out there to the 5 argument that hazard wasn't a major uncertainty. I think 6 that was a fault of his reasoning rather than of the I 7 process. l 8 . . MR. MURPHY: Okay. So you run -- in this case, you , 9 run this computer code again 25 times, and you end up with l 10 25 seismic hazard curves, a probability of exceedance versus 1 1 11 some ground motion measure, and in this case, we'll stay 12 with spectral velocity because we took our target l( 13 probability, in effect converted it into a target spectral 14 velocity of'eight centimeters per second. 15 You re-enter this series of figures, our 25 16 seismic hazard curves, and you read off a probability of 17 exceedance at that eight centimeters per second, okay? 18 [ Slide.] 19 MR. MURPHY: Figure 13 is a table that contains 20 the probabilities of exceedance of that median of the I 21 average of 5 to lo hertz, so that we see in the box at 0 to l 22 25 5 to 5 and 1/2 a2 to the E, the minus 5 contribution. i l 23 We see over here at -- i 24 MR. SIESS: Hold it a minute. Let's look at one , 25 box at a time. l
i l 179 l
?
1 MR. MURPHY: Okay. I just wanted to explain where i J 2 the O came from. That simply means that it became E to the i ! 3 minus 10 or larger. ; 4 MR. SIESS: Okay. But let's take a --
-5 MR. MURPHY: Smaller.
i f 6 MR. SIESS: Take O to 25 miles -- I assume that's 1 i 7 miles, or is it kilometers now? 8 , MR . MURPHY:- Kilometers. 9 MR. SIESS: Kilotaters? It might help to say,
-10 wouldn't it, although in the probabilistic domain, it's 11 minor. In the 5 to 5 and 1/2, now that represents -- O to
.)
la 25 represents a circle, and the 5 to 5 and 1/2 could be'from i 13 what? Any of several different zones, right? l( l 14 MR. MURPHY: I'm not sure what'you mean by several. l l 15 different zones. :
- 16. MR. SIESS: Well, go back --
.17 MR. MURPHY: Any zone that is within 25 kilometers j 18 - - ;
19 MR. SIESS: Yes. I 30 MR. MURPHY: Yes. I 21 MR. CHOKSHI: There are quite a few zones. 22 .MR. SIESS: So we're not de-aggregating into , 23 zones; we're just de-aggregating it to what we have defined. l 24 MR. CHOKSHI: Magnitude. On the magnitude, but T 25 not on each. zone, right. 1
l 180 (-. l 1 MR. SIESS: And how do we determine these bins? l l
'2 Just arbitrarily.
3 MR. MURPHY: The initial selection of the bins you 4 could say is arbitrary. It's convenient rather than l 5 arbitrary.
- 6 MR. LINDBLAD
- And is the data constrained to add 7 up to 10 to the minus 4th?
l 8 MR. CHOKSHI: If you combine the distributions. 9 These are two medians. 10 MR. LINDBLAD: Okay. . 11 MR. CHOKSHI: They are medians from the 12 uncertainty analysis sampling or large number of in each ( 13 bin.
- l
~
l 14 MR. LINDBLAD: But the total -- 15 MR. CHOKSHI: Well, you combine -- 16 MR. LINDBLAD: The total hazard was 10 to the 17 minus 4.
~18 MR. CHOKSHI: 'Right. But you combine 19 distributions to get the final median. You can't just add 20 up median. One really important thing here is that the al medians are -- ranking is -- the related ranking is the l 32 right one. The modeling is important because you are going 33 to decide which bin contributes most. J t 24 MR. SIESS: I just tried to figure what this tells
, 25 me. This tells me that within 25 miles of the plant, there t
._I
h l i 181 (- , 1 are some seismic potentials ranging from 5 to 6 and 1/2 3 magnitude. l 3 MR. MURPHY: Right. l 4 MR. SIESS: And that the 5 to 5 and 1/2 actually 5 contribute.more than the 6 to 6 and 1/2. 6 MR. MURPHY: Right. ' 7 MR. SIESS: Because those 6, 6 and 1/2, were what? !
\
-8 In a zone that was rated as lower probability by the l 9 experts.
10 MR. CHOKSHI: Right. And'there are two things. 11 Rate of occurrence also may be lower on those things. 12 MR. SIESS: Maybe the what? ( 13 MR..CHOKSHI: The frequency of. occurrence was -- 14 there are several factors which could lead you to that l 15 result. 16 MR. SIESS: And of course, the 6, 6 and 1/2 could 17 actually be two zones out there? 18 MR. CHOKSHI: Right. 19 MR. MURPHY: Uh-huh. 20 MR. SIESS: Now, why do I want this information, j 21 or shall I wait until the next two or three slides to tell 22 me what you're going to do with it? 23 MR. MURPHY: In the next few slides, we'll tell
- 24' you what you're going to do with it. But this information
~25 -- this provides an indication of where the hazard is coming i
t l
.. - . . - . _ - - - _ - - . _ . . - - . . - . - . - . - - - . _ ~ . - . . _
1 182 l[ 1 from for your particular site. 2 MR. SIESS: Well, we knew that to begin with, 3 didn't we? 4 MR. MURPHY: We knew -- 5 MR. SIESS: We had to know where the hazards were 6 coming from to start with. 7 MR. MURPHY: We did not know the relative 8 contribution of the various sources before we went through 9 these calculations. 10 MR. SIESS: Why wouldn't you -- why wouldn't you 11 know that a 7 out there at 30 miles is going to be a greater 12 contribution than a 5 at 50 miles or a 5 at 30 miles?
- ( 13 MR. MURPHY
- There is --
14- MR. SIESS: You said relative contributions, and 15 that's what I'm talking about. 16 MR. CHOKSHI: What is, I think, not clear in 17 determi'nistic is the rate of occur.rence. If the 5 occurs ! 18 more frequently than 7, that's captured here.
- 19 MR. SIESS: Okay. So that if I simply took each 20 of those boxes on the map and put down the probability of l
21 exceedance on that or the return period or something else, l 22 then I could look at them and see the same thing. 23 MR. CHOKSHI: Yes. If it was a -- you know, if it 24 was two, three zones, you can do that, yes. 25 MR. SIESS: I mean, what am I trying to do now? 4
~
a 183 ( 1 I'm trying to find out what peak ground acceleration with 2 the Reg Guide 1.60 spectrum I have to use to design this 3 plant for an earthquake that won't be exceeded more than 4 once in 10,000 years. 5 MR. CHOKSHI: That's one way of -- 6 MR. MURPHY: That's one way. 7 MR. SIESS: That's what I'm looking for. 8 MR. CHOKSHI: Right. 9 MR. MURPHY: That's one thing you can, yes. 10 MR. SIESS: And I can't get that spectrum without 11 going through this process. I couldn't get that spectrum 12 from back where I was when I had the velocity at 5, 10 13 hertz. , ( 14 MR. LINDBLAD: Deterministically, but not 15 probabilistically. 16 MR. CHOKSHI: Yes. I think that's -- 17 MR. SIESS*: Well, I've got a figure that says 18 probabilistically, the average 5 and 10 hertz spectral ! 19 velocity of 8 centimeters per seconds, 10 to the minus 4. 20 I've got a 10 to the minus 4 and I've got the npectral 21 velocity, and I can certainly draw a spectrum that has those 22 two values for velocity. What's wrong if I stop there? 23 What's missing? 24 MR. MURPHY: What we are saying is missing at this 25 time is an understanding of where, probabilistically i
. . .___.,m . - _ . _ m. . - . - . . . _ _ . - . - . - . . _ . . . . _ _ _ _ _ . . . . . _ . -_. _ ___ _ _ _ . . .
184 i 1 - speaking, out there in the source zones this contribution is , l l 2 coming from, where this -- 3 MR. SIESS: Well, in the first place, I guess just
- 4 having the areas labelled with a magnitude and a return I l
5 period is almost what y ou're doing here. You're making a -- 6 I could draw a map from that plot up there. 7 MR. MURPHY: You could draw a map from this plot, l 8 but understand that from this plot, you're drawing one ! 9 without an azimuth. You're just drawing rings and assigning l 10 probabilities to those rings -- 11 MR. SIESS: But now go back to the question of why l 12 do you need to know that. Why do you need to know how you !{ 13 attribute th.is to the various sources? 14 MR. CHOKSHI: I think this goes to the question of 15 stability. You know, your deterministic selection of source 16 zone, for example -- there was no 6 to 6 and 1/2 at certain l l i 17 distances. 2 18 MR. SIESS: ~ Wait a minute You said that -- we're 19 in the probabilistic now. MR. CHOKSHI: Right. But the reason we're saying 20 l,, 21 do both is to answer some of the questions which are hidden 22 in deterministic judgment. 33- MR. SIESS: Yes, but right now, we haven't even l 24 looked at -- we've already done the deterministic and put it l 25 off to one side. I
. . - .- -. . . - . . . - - . _ . . . ~ . . - - _ . . . . . ~ _ - ._.-. .. - . - .-. - _ _ _ - -
l e
.185 i
1 MR. MURPHY: Right. 2 MR. SIESS: Now we're in the'probabilistic, and I, - 3 got up to Slide 10, where I thought I had something -- 4 MR. MURPHY: Right. l l 5 MR. SIESS: -- that I could compare with, and you 6 say no, that's not enough, I've got to keep going.
-7 MR. MURPHY: That is -- Slide 10 was one place ,
l i 8 where you could make a comparison or judgment. 9 MR. SIESS: My deterministic -- I got a spectrum. L 10 I had something to design with, and I've got a -- I can pick i ' l 11 out 5 and 10 hertz,-right? i i 12 MR. MURPHY: Right. l 13- MR. SIESS: And I average them and they come out
'( '- l l 14 to be 9' centimeters per second. i 15 MR. CHOKSHI: Yes. I think I --
t l 16 MR. MURPHY: Let me say this: What we are trying i 17 to get at with this process, the aggregation, the de- - 18 aggregation, the assigning the bins and so.forth, is we are 19 'trying to get to a point of comparison that goes well with j l 20 th'e comparing to the controlling earthquakes and the spectra 21 that we got on the deterministic process. 22' MR. SIESS: It is right there.on Slide 10. You've 23 got a probability of exceeding 8 centimeters per second 34 average spectral velocity of 5 and 10 hertz, or whatever it ! l 25 was. I.
l l i I l 1 b 186 i 1 MR. MURPHY: That -- l 2 MR. SIESS: I've got a design basis sitting over l l' 3 'in my righthand desk drawer'that I could como out, and I've-l 4' got 8 hertz -- I mean 8 centimeters per second and 5 and 10 l 5 centimeters per second at 10, and that averages out at 9, j 6 and, gee, if I'm designing for 9 centimeters per second, I
- 7
- ought-to be better off than this. Gee, I must'be in the l 8 right ball park.
9 MR. MURPHY: That is -- 10 MR. SIESS: Or do I want to say more?' l
- 11 MR. MURPHY
- What we're saying is that with a 12 little-bit more effort, that we can say more. We can make 13 another series of comparisons at very little expense that
\* .. 14 would provide us-with, we think, better information to make 15 our regulatory decisions. l 16 It's quite valid, the point that you're making. 17 You can -- 18' ' MR .'. SIES S : Okay. Let's go ahead: and see what -- ' l 19 MR. MURPHY: You can stop several places in this ; 30 process and -- L 21 MR. SIESS: Let's see what you come up with. ! 22 MR. CHOKSHI: I think maybe -- two things. If you 23 want to use site specific motion not attached to any, say, 24 1.60 or something, then you have to go further, develop a
- 25 site specific shape.
i That's one thing. 1 1 ! l
-, . _ _ - _ _ . _ - _ _~ _ .
- . . . - . , . . . . . - __ .~ .- - - . _ - - - . . . . . . -
l I l( 187 ' 1 The.secepd -- 2 MR. SIESS: Wait a minute. 3 MR. CHOKSHI: Okay. 4 MR. SIESS: Site specific what? 5 MR. CHOICSHI: Site specific spectra.
- 6 MR. SIESS
- Are we supposed to use site specific -
l l 7 spectra? Don't we still use Reg Guide 1.607 8 MR. CHOKSHI: Site specific -- 9 MR. SIESS: In Appendix B, we don't use 1.607 l 10 MR. CHOKSHI: That option is there to use site 11 specific spectra. 13 MR. SHAO: You can use site specific spectra in { 13 lieu of 1.60. 14 MR. SIESS: Well,'that's a nice option, but 15 suppose I choose not to? Can I then meet all these other 16 criteria? I have to -- and if I can use site specific 17 spectra -- l 18 'MR . CHOKSHI: Righ't. j 19 MR. SIESS: I don't know why. I guase if I come 20 in with a standard certified plant at a particularly lousy 21 site, I might find that using Reg Guide 1.60 -- no, I don't 22 see how, because the certified plant is designed for Reg 23 Guide 1.60, isn't it? 24 MR. CHOKSHI: That's 1.60. l 25 MR. SIESS: All right. And if I choose -- why
. ~ - - . _ . - _.
l 188 ( 1 would I want site specific spectra? l l 2 MR. MURPHY: Yes, 1.60 is currently one way that 3 is permissible to carry this out. We have -- , i -4 MR. SIESS: Then I can -- 5 MR. MURPHY: We have a standard review plan that 6 provides a hierarchy, and I don't remember the exact order 7 now, of what the staff's preference is by way of calculating 8 the site specific spectra, and I believe the hierarchy ( 9 starts with a specific, if you want, single or multiple time ( 10 histories from a site at the same magnitude as being i 11 postulated for the facility -- f f ! 12 MR. SIESS: Is this in the new standard review l( 13 plan? i 14 MR. ROTHMAN: Yes, it is. l l 15 MR. SIESS: And it wasn't in the old one? l 16 MR. ROTHMAN: In Reg Guide -- I mean it's SRP 17 Section 252 -- l l 18 MR. SIESS: Yes. l 19 MR. ROTHMAN: -- 1989 revision. We have a i 30 hierarchy for justifying the design of the plant. The ( 31 preferred way is to look at site specific spectra rather 22 than just calculating a PGA and anchoring your spectrum with ! 23 it. 24 MR. SIESS: But now, the question is not what l 25 you've got. You know, if the criterion is what we've got, i l l
._ . - . . - . . - . . . _ . - - - - . - _ . . . . - _ - . . - . - - . - ~ - - - . _ . . - ~ . . . _ . . .
P . i s i ( 189 1 we can drop the whole thing, because what we've got is what 3 we've got. We're changing things. ! l- 3 The question is, why'do we have that and what's 4 its contribution to safety and what's its effectiveness. 5 MR. ROTHMAN: What we have found over the years 6 when licensing plants is that in order -- what would happen 7 is somebody would pick a peak ground acceleration and anchor j 8 a standard spectral shape with it. t 9 It was very hard to unequivocally attribute a'PGA 10 to a certain. magnitude earthquake; so what we did is we got L 11 suites of records from earthquakes recorded at the 12 appropriate distance with the appropriate magnitude and l(s, 13 compared that to the design basis. If the design basis was ) 14 higher than the's'ite1 specific spectrum, we were able to 15 accept the design basis. 16 We felt that just selecting a PGA with what we 17 know about magnitude and PGA, we felt that was -- what we
.18 preferred to do is look at the whole spectrum and compare 19 _that to the design basis rather than just a PGA attributed 20 to a magnitude. So that became the-way that we reviewed 31 plants over the last twelve years or so.
22 MR. SIESS: Well, then whet happened to Reg Guide 23 1.60 spectrum? 34 MR. ROTHMAN: Well, that's the design -- that may 25 be the design spectrum, but we justify --
.- - . ._. .-. - -. .- .- . . . _ . - = . .. - . ~ - .
l l' 1
.( 190, 1 MR. SIESS: But I mean, why should somebody design 2 for a spectrum that the staff doesn't want? I don't 3 understand.
4 l MR. ROTHMAN: 'The structural engineer -- 5 MR. SIESS: If you're going to pick the spectrum, 6-why don't you pick the spectrum and tell them what to design 7 for? ~, 8 MR. ROTHMAN: Well, the structural engineer is 9 like a standard spectral shape that they have time histories l 10 that they can generate to do their analysis, 11 MR. SIESS: Only because they want to do time 1 12 history analyses. I'm just confused now. ( 13 MR. CHOKSHI: I think I understand your question, 1 i 14 Dr. Siess. Suppose you start with some kind of standard 15 spectra and you are fairly comfortable that that is going to . l 16 envelope whatever you've proven your site specific bases. l 17 There may not be a need to go much further than that as you 18 go along the proces's. 19 The way this is written is a general process, how 30 to come up with a site specific ground motion. It does not al say, " Suppose we use .3g Reg Guide 1.60?"
- 22 MR. SIESS
- 'But this doesn't separate out the 23 outlier, the oddball earthquake out there in Zone 6.-
j 24 MR. CHOKSHI: This would. 25 MR. SIESS: No, it won't, because it only -- it
l l l l 1 l( 191 l' just arbitrarily divides the thing up into geographical l 2 areas without regard to what's in those areas'. 3 MR. CHOKSHI: But what would -- 4 MR. SIESS: You can have three different 5 earthquake regions -- 6 MR. CHOKSHI: Right. I 7 MR. SIESS: -- three different seismic things, and I 8 any one of those -- l I 1 9 MR. CHOKSHI: No, but rate of occurrence will l 10 spike it. If you have a rate of occurrence higher, then it l l ; l 11 will show up. I i 12 MR. SIESS: Yes, but I mean this doesn't allow you I
\
'( 13 to look at this earthquake over here, Zone 3 or whatever you 14 have called it, versus Zone 5 or versus Zone 6. I could 15 visualize where I had all three of those zones out there in 16 this 50 to 100 kilometer annular space, and all I've got is 1 17 one lumped probability. i t 18 l l MR. CHOKSHI: Yes. 19 MR. SIESS: Which one comes from which one is ! 20 lost. 21 MR. CHOKSHI: No, but that information is not 22 lost. That information is always there. Once you see l 23 something'like that, you can always go back and see that 24 information? < l 25 MR. SIESS: But why?
^
l l I l 1 I i ( 192 j 1 MR. CHOKSHI: For example -- ! l 2 MR. SIESS: I mean, what's the object of the game? 3 What are we trying to do? Understand everything about the j 4 seismic potential of that site based on a bunch of stuff l 5 that was averaged from four experts, and we're going to 6 learn all of this detail from it? Isn't the object of the l 7 game to get a seismic design basis for designing the plant l 8 and the components and the systems so that they will l 9 withstand an earthquake with a reasonable probability? l l 10 MR. CHOKSHI: Yes. 11 MR. SIESS: Are we losing sight of that, the idea i 12 of getting all this understanding about the geology, the l ( (, 13 seismology, starting from a bunch of assumptions out there 14 by six experts, and going from there down to three decimal 15 points, it looks like. 16 I am completely lost as to what we're trying to 17 do. 18 MR. CHOKSHI: Well, as I understand the past, the 19 problem has been that six experts -- that your six opinions 20 are there. This is one way of the effect of the six 21 experts. 22 MR. SIESS: It was a great system. The seismic l 23 hazard study was a great system, just like NUREG-1150, for 24 bringing out the uncertainties, which everybody knew was 25 there beforehand but couldn't quantify. i l
l l ( 193 1 Now, presumably, we can now quantify the 2 uncertainties, but we haven't got a procedure here based on l 3 quantified uncertainties. You're still back to using ! 4 medians. What are we gaining? Somewhere I'm missing 5 something. 6 I'm not saying you're wrong. It's just that I'm 7 having less and less trouble understanding what you did and 8 more and more trouble understanding why you did it. What's 9 missing? 10 MR. LINDBLAD: Chester, your point is supported by 11 the fact that the processes we're using are processes that 12 we used on plants already operated, already designed and '( 13 constructed and operated, and not for setting SSEs, but now 14 'we're proposing to use that evaluation tool'as a licensing - 15 - initial licensing tool, and I think the point he's making 16 is why? 17 MR. SIESS: I think we've just gotten this whole ! 1 I 18 seismic design thing so complicated 'that -- so refine'd that 19 it just bothers me. 20 I don't have that much confidence in the answers. 21 I don't care what you come up with, and you feed them into a 22 finite element program for piping and cracked concrete walls 23 and all of that stuff and think you've got stresses good out j l ! l 24 to three significant figures. j 35 I think that's ridiculous. I don't think the l i l 1
1 i i-l l( ! 194 1 method of analysis is that good. I don't think your 2 assumptions are that good. Your damping factors, you don't 3 know; the inelastic behavior, you don't really know. 4 When I get that kind of-feeling, I remember 5 something one of my great professors told me, and he says,
~
6~ if you've got to be wrong, be wrong by the simplest possible l 7 procedure. l l 8 He said, if you've got to be approximate, be l j 9 approximate by the simplest possible procedure, but I think,. 10 at the best, we are being approximate; at the worst, we are 11 being wrong, and we're making the procedure so complicated - 12 that I am. lost in knowing where I am in this thing. i 13 .I have no physical feeling for this, and every (l 14 time I try to get one, well the computer will do it. l-15 I am not particularly happy'at putting numbers in
. 16 one end of a computer, putting the X-Y latitude-longitude in l 17 my plant in one end and getting a.n average of a 5.and'10
- 18 hertz velocity, spectral velocity, out the other.
19 I think I could probably guess at it just as good. 20 If Nate Newmark was around, I know he could guess at it just 21 as good. I'm really having a problem here, and maybe'
- 22 there's more people driving this than the structural 23- engineers.
l 24 MR. LINDBLAD: I can picture that there is an l l 25 attraction to be able to say to the public that, without L- . ._ _ _._ .. . _ _ - - - _ .. . -
l f l I I ( 195 1 digging a trench or making a map, we have evaluated all , l 2 sources within 200 kilometers of the site, but that really j i , 3 doesn't add a great deal to knowledge. 4 MR. SIESS: We can do that now. The deterministic 5 is going to require more than that. 6 MR. MURPHY: Yes. That's one very specific point. 7 In our minds, the deterministic process is more 8 time-consuming and more costly than what we are proposing l l 9 here, and what we're proposing here, if we take it a few l 10 steps further -- and I'm not sure, after you had a good 11 opportunity to sit down and follow this through, that in the l 12 end it will be as complicated as you have the feeling now, 1 13 because what we are trying to do is to provide some physical ( 1 14 feel for what we have found out through the probabilistic ) l 15 analysis. ( 16 The objective here is to come up with something 17 equivalent to the controlling earthquakes that we talked l 18 about in the first few slides,'the earthquakes that control ! 19 the spectra against which the facility should be designed, i 30 and what we have here are some steps. al Apparently they appear complicated at this stage, l 22 and I know they were to me when I first looked at them, but l l 33 I think we take a process that gets us to a point where we , 34 again have something physical that we can talk about, a 25 magnitude and a distance for maybe one or more earthquakes
I i[ 196 l 1 that would control the design of the facility. 2 MR, SIESS: But that's where you were when you l 3 finished you deterministje approach. 4 MR. MURPHY: Exactly. That's where we were when 5 we finished our -- 6 MR. SIESS: Not even have to run it through a 7 computer and lose all the feel for it. I'm trying to figure 8 out what's better. 9 MR. MURPHY: Because getting to that deterministic 10 process, we did not track all of the uncertainties that were 11 involved in that process. 12 The probabilistic is attempting to get us to the 13 same place, a series of controlling earthquakes, by using a l(. 14 computer code or several computer codes to track that 15 uncertainty from our deterministic experts' opinions thrcugh 16 an analysis process and come up with something that is very
.17- close to what we had from the deterministic process, in 18 effect a controlling earthquake's magnitude and distance 19 from the facility.
20 MR. SIESS: Where are the uncertainties coming in? 21 1 haven't seen anything on uncertainties. I've seen 22 medians, I see single curves. l 23 MR. CHOKSHI: Because of the Monte Carlo process, 24 all the parameters are in there. 25 MR. SIESS: Well, the uncertainties were put in
- n. - ~ . ~ - - .~- - - . - . . . - ~ . . - - . - . . . - . .- _ . . . - -
i h' 4 l' ' j(- 197 , 1 there by the experts. 2 MR. CHOKSHI: Yes. t 3 MR. MURPHY: That's where they'd have to come from 4- anyway. 5 MR. SIESS: They're not coming out of this
~
l 6 analysis.
~7- MR. CHOKSHI: No. They are put into the analysis.
8 ;MR. MURPHY: They go into the analysis. l .9- M
.R. SIESS: When you had some experts do the 10 deterministic thing, they put in some uncertainties, put in-l 11 some margins for uncertainties. Sure they did.
12 MR. CHOKSHI: No, they do, but nobody else knows ((' 13 what it is. That has been the contention in the process. l~. . l 14 MR. SIESS: You're not going to.know when you get i I 15 through what it is. l 16 ' MR. MURPHY: When we get through the probabilistic l l 17. process, you can go back and you can look at the 18 uncertainties that were' involved'in the maximum magnitude ! 19 associated with a particular earthquake, a particular source l 20 zone, and you can look at the uncertainties that were
- i. 21 associated with the recurrence estimates that were provided 22 for that source zone.
23 <You can look at the uncertainties that were
- 24 involved in the ground motion.
25 MR. SIESS: What quantified those uncertainties? l I
- l. .
l l ( 198 1 MR. MURPHY: The experts' opinion provided the 3 initial quantification of those uncertainties. 3 MR. SIESS: Why wouldn't you just get the same 4 experts to quantify the uncertainties in the deterministic 5 method, just to say, you know, 6 plus or minus a half l 6 magnitude or -- l L 7 MR. MURPHY: As the deterministic process is set 8 up today, all that simply would enter into a table 1 9 someplace, and it says the applicant's expert says the 10 uncertainty on the maximum magnitude in this area, this 11 source zone, is plus or minus 2. 1 I l 13 MR. SIESS: What is wrong with that? What do you l 13 want to do with it here? [ , 14 MR. MURPHY: Here we have tracked that uncertainty 15 and we are in a position to answer the rebuttal questions 16 associated with alternate hypothesis, intervenor hypothesis, 17 staff hypotheses that the applicant did not consider this 18 pa'rticular earthquake in his analysis, in his deterministic 19 analysis. And Dave Korent would always ask the question 20 what difference does it make? And'what we are doing here is f 21 quantifying, trying to quantify and put some perspective on l 22 what difference it does make. 23 MR. SIESS: If then whoever is adjudicating this, 24 the Board or the Court, accepts your interpretations, then 25 you are rebutting. Actually, I suspect that if somebody l t 4
- r . . - .
( 199 l 1 wants to understand this process, he won't challenge you on j 2 this, earthquake out here, he will challenge you on about 14 3 steps in the process here where you could have done 4 something different. There are an awful lot of curves in 5 here to be challenged. 6 MR. MURPHY: There are an awful lot of curves in l 7 there to be challenged. We'said that this has been l 8 examined, if you want, by the staff and at this stage we 9 have taken the position that this is acceptable in carrying 10 out these kinds of calculations and we understand what the j , 11 results are like. l 12 MR. SIESS: All we have to do really looking at 13 { s.ome of this is to go find Expert No. 5. Obviously it was ! 14 credible enough that you are still using his numbers. You i 15 said you didn't eliminate him, so he must still be around. l 16 MR. MURPHY: I think he was the last time I heard, 17 yes. l 18 MR. SIESS: Okay, go ahead. Let's dig thr'ough the 19 rest of this. I am understanding, but I am still working on l l 30 why. 21 MR. MURPHY: Okay. So, the values that are listed l 22 in this table are something that we're calling hazard H md; 23 okay. 24 [ Slide.) 25 MR. MURPHY: The next viewgraph is simply a way of
. - = - - - ,
I ( 200 1 summing these up to obtain the magnitude and the distance 2 for what we are calling the controlling earthquakes for this 3 probabilistic scenario. 4 MR. SIESS: Now, is that a real earthquake out 5 there or is that just some artifact? 6 MR. MURPHY: It's an artifact. 7 MR. SIESS: It will give the same result, the same
.8 probability of exceedance of a median special. velocity as l 9 the collection of earthquakes you've got there?
l 10 MR. MURPHY: That's correct. This is an artifact. l 11 MR. SIESS: This is some fictitious thing that you 12 are going to put at some distance? Okay. 13 MR. MURPHY: Yes, sir. (. . l This is not a real 14 earthqu'ake in the sense that'it is not in the USGS catalogs. 15 MR. SIESS: And it might never be. l l 16 MR. MURPHY: Probably will never be.
~17 MR, SIESS: Are you sure it's possible that it can' 18 have these characteristics? There is no characteristic l 19 defined except the magnitude and the distance; right?
20 MR. MURPHY: This is just the magnitude and the l 21 distance. It does not even include an as azimuth from the ~ 22 facility. Now, that magnitude would be -- 23 MR. SIESS: the data on the previous page, that 24 magnitude would be less than-seven; right? 25 MR. MURPHY: Yes.
( 201
, 1 MR. SIESS: It is an averaging; isn't it?
? 2 MR. MURPHY: Yes, sir. So, it would be less than 3 3 -- as you see, there is no contribution in any of.the-bends ) , 4 greater than seven. 5 MR. SIESS: Take the magnitudes by distances and l 6 the distances by magnitudes and -- 7 I've looked at that and I've tried to get a 1 It is just" numbers.
~
8 physical feel for it and I can't.
- 9 MR. MURPHY: Okay.
10 MR. CHOKSHI: If you had a dominant source, there 11 was one earthquake out there that controlled the 12 probability, then you would get the same thing back. If you { 13 had. one earthquake which contributed to the hazard, then you 14 would get an earthquake which is like, you know, more real. 15 MR. MURPHY: In other words, if you were building 16 a power plant 15 miles from the epicenter of the Charleston 17 earthquake, basically you would end up with that. epicenter
- 18 and that Charleston earthquake as being'the controlling or 19 dominant earthquake for that facility.
20 MR. SIESS: It would be possible that this could
'21 be a single earthquake?
22 MR. MURPHY: Yes. It is possible that it could be 23 a single earthquake. 24 MR. CHOKSHI: Like Hosgri Piablo; right? 25 MR. MURPHY: Again, we would offer the caveat that I
)
i i l J i l 1 l g 202 l 1 1 right now what you've seen here has been carried out for a I 1 2 particular frequency band, five to ten hz. We are still I 3 invest,igating whether or not we shouldn't specify some 4 additional frequency bands for this controlling earthquake. 5 MR. SIESS: Why? l 6 MR. MURPHY: Same reason for the deterministic 7 one. It may be that the five to ten hz is controlled by a I 8 moderate local earthquake and the higher longer periods, low 9 frequencies would be controlled by a more distant 10 earthquake. 11 MR. SIESS: And this wouldn't bring that out? 12 MR. MURPHY: That wouldn't bring that.out if we [ 13 were simply using the five to ten hz. 14 MR. LINDBLAD: Are these distances epicentral or 15 hypocentral? 16 MR. MURPHY: For the most part it doesn't make any 17 difference. We are not *. hat refined at this stage. 18 .' MR. BERNREUTZR: Don Bernreuter. Strictly 19 speaking, the distance would be hypocentral because that is 20 the calculation of --
'21 MR. SIESS: Once you get beyond 25 kilometers --
'22 MR. BERNREUTER: Once you get beyond ten 23 kilometers then it doesn't make a difference. l l
24 (Slide.] I l 25 MR. MURPHY: These numbers that are listed here
]
l l 1 I
_ _ . . . . - . . - . . , _ . . _ - . - ~ . - . . - . . - _ _ _ - . - . - . ~ . - . . . - . . ---. ... .- ~ . ._ ._ t 203 f 1 are from theJde-aggregation for the Vogtle site. Those are . 2 specifically Vogtle de-aggregation. 3 MR. SIESS: All right. i 4 Now, this table shows five experts and four l 5 experts, but the Figure 13 didn't show both. Which is it? : 6 MR. MURPHY: The Figure 13 was for the five-expert , 7 case. That's the one we're proposing to use at this time. 8 MR. SIESS: All right. 9 So, I've got a 5.8 at 32 kilometers. 10 MR. MURPHY: That's right. And the actual' design 11 that we're representing -- t 12 MR. SIESS: That puts me in a particular bin, , I ( 13 right? 14 MR. MURPHY: Pardon? 15 MR. SIESS: That puts me in a bin that -- 16 MR. MURPHY: No. This doesn't -- 17 MR. SIESS: It does. 18 MR. MURPHY: This -- 19 MR. SIESS: If I took that earthquake now and put 30 it at 10 to the minus 4, am I right? 21 MR. MURPHY: Yes, sir. That's one way of doing
-22 it.
23 MR. SIESS: So, if I had only that one earthquake, 24 it would go in one bin here, but instead of being 8.9 times 25 10 to the minus 6, it would be 1 times 10 to the minus 4. I r -
-m . .___ .___ ._ . . . _ _ . _ _ . _ _ . _ . _ _ _ . _ _ _ . _ _ . . . _ . . - , 204 1 could forget about all the other earthquakes and bins.
2 MR. CHOKSHI: Yes. This is -- if you look at the 3 bins, as you see, the first two bins, 5, 5 1/2, 5.5, and 6 4 alo the dominant bins. 1 5 MR. SIESS: Now, what I've got is an earthquake, 6 5.8 at 32 kilometers or 5.6 at 27 kilometers, I guess, 7 right? 8 MR. CHOKSHI: For the four experts. 9 MR. SIESS: If I replace this whole chart with a 10 single bin in that area, 10 to the minus 4 -- 11 MR. CHOKSHI: Right. 12 MR. SIESS: -- I'd have the equivalent risk. ; 13 MR. CHOKSHI: Yes. This 5.8 -- 14 MR. MURPHY: Right-. 15 MR. SIESS: Equivalent risk. 16 MR. CHOKSHI: This will put you at the level of 17 ground motion comparable with 10 to.the minus 4 probability 18 of accidents at*5 and 10 hertz. s 19 MR. SIESS: And if I designed my plant to resist 20 that earthquake, I'd have 10 to the minus 4 probability of 21 exceedance. 22 MR. CHOKSHI: At that frequency. 23 MR. MURPHY: At that frequency, right. 1
- 24 MR. SIESS: And the equivalent, then, of all this 25 whole array of earthquakes have different probabilities,
( 205 1 different distances, and different magnitudes. Okay. 2 Shall I call it a su'rrogate earthquake? 3 MR. CHOKSHI: Scenario earthquake, I.think. 4 MR. MURPHY: Yes. 5 It's one of a series of terms. At this stage, 6 we're calling that a probabilistic controlling earthquake, 7 which we are specifically, at this point, comparing with the r 8 deterministic controlling earthquake that was actually used ; 9 for Vogtle, which was a 5.3 at 15 kilometers. 10 MR. LINDBLAD: So, the man on the street saying 11 that says you haven't considered a large-enough earthquake. 12 MR. CHOKSHI: No, because the distance is there, -( 13 too. 14 MR. MURPHY: 'Which' one? This is step one of the 15 comparison. 16 MR. SIESS: Right.now, it doesn't tell me 17 anything. I look at that and say, well, they designed for 18 5.3 at 15. They designed for a smaller-magnitude earthquake I 19 but a lot closer. , 20 MR. CHOKSHI: Right. 21 MR. SIESS: That doesn't tell me anything. 22 MR. MURPHY: That's right. That's why, in this 23 case, we're not finished yet. t
- 24 MR. SIESS
- Okay.
25 MR. MURPHY: Okay.
I ( 206 1 [ Slide.) 1 2 MR. MURPHY: Let me skip over a few viewgraphs to l 1 3 Figure 18. 4 MR. SIESS: Okay? ) 1 5 MR. MURPHY: Okay? We should have brought those 6 other ones. 7 Now, Figure 18 has got four curves on it. The 8 easiest one is the Reg. Guide 1.60 spectra anchored at 0.3g. 9 That's "R". I 10 MR. SIESS: We can forget about that right now. l 11 MR. MURPHY: Right. 12 The next one is the actual design spectra anchored
- / 13 correctly for Vogtle, 0.2g, 1 . 6.0 .-
14 MR. S I ES S': What was that? 15 MR. ROTHMAN: Vogtle's SSE is 0.29, Reg. Guide 16 1.60. 17 MR. SIESS: Okay. That's 0.2g, Reg. Guide 1.60. 18 Okay. 19 MR. MURPHY: Okay? 20 MR. SIESS: That helps me, because -- that's 21 velocity. No, that's acceleration. 22 MR. MURPHY: No, this is acceleration. 23 MR. SIESS: Here's 0.2. I wish we'd get rid of 24 computers for plotting figures. They always want to do 25 everything. l t
_._., __ _ . . . _ . _ . - _ _ . . . _ . _ _ . . _ . . _ _ _ . _ _ _ . _ _ _ . _ . . _.__..-.m__ . . . . _ _ _ _ .m..
'i ;
( 207 l 1 It's interesting -- 10 to 100 centimeters per l' l 2. second. Okay. And that's g, so E plus 3 is about ag,- l ! 3 right? l t 4 MR. CHOKSHI: Right.
- S MR. SIESS: Okay.
l 6 Have you got somebody in Research that decides how 7 to draw graphs to confuse people? 8 [ Laughter.] 9 MR. MURPHY: We don't have them intentionally l 10 employed. i 11 MR. SIESS: Have you officially changed from g's - 12 'to centimeters? '(. 13 MR. MURPHY: It's in the works. There is'a 14- statement that we're. supposed to be d'oing both of them in 15 the interim. I 16 MR. SIESS: I guess a gal isn't SI. So, you can't l 17 use it. 1 l
.18 Incidentally, you can't use centimeters in SI, 19 millimeters and meters. You don't deal with anything except l i 30 factors of 1,000. So, if you're really going to go SI, l 21 centimeters is nothing better than --
l 22 MR. MURPHY: Blew that one. I'm glad that was the l l 23 only mistake today. l , 24 MR. SIESS: There are no liters in SI. I I 35 MR. MURPHY: Just for completeness, the two other l i l - i t i I i
-, , .- .- i
4 t l l 208 1 figures on there, "C" is the spectra from the controlling ! 2 earthquake with five ground-motion experts using the five 3 ground-motion models that we had from those experts. 4 "S" is the magnitude and distance for the actual i 5 Vogtle, again using the same ground-motion model that was I l l 6 used for "S", "C" and "S", same ground-motion model, so that I ! 7 they can be compared. f l 8 MR. SIESS: "D" is what they designed for. ' 9 MR. CHOKSHI: Correct. 10 MR. MURPHY: Yes.
- 11 MR. SIESS
- And now, which of these is what we got 12 of all this previous stuff?
l 13 MR. MURPHY: "C" is what came out of the 1 14 probabilistic -- 15 MR. CHOKSHI: "C" is 5.8, 32, right? 16 MR. MURPHY: Right, 5.8, 32, the five ground-17 motion experts. 18 MR. SIESS: Back to the'five experts. 19 MR. MURPHY: Right. 20 MR. LINDBLAD: "C" stands for controlling 21 earthquake? 22 MR. MURPHY: Yes. 23 MR. SIESS: "S" is what? i 24 MR. CHOKSHI: 5.3 at 15. 35 MR. SIESS: "S" is 5.3 at 15. Okay. Now I see, f i
. ,,g_.,- , ,, . -
l I i l \ t l t I ( 209 1 1 MR. CHOKSHI: And it's the same ground-motion 2 models. They are comparable, one to one. 3 MR. SIESS: So, now we can see the difference . 4 between.,the design and this for the five experts. : 5 MR. CHOKSHI: Correct. 1 L 6 MR. MURPHY: Right. 7 MR. SIESS: And the five experts are a-little ; e 8 higher than the four experts. i f 9 MR. MURPHi: And that's what we're proposing to i 10 use. ,
-1 11 MR. CHOKSHI: No. This is --
t 12 MR. MURPHY: Five experts -- what we're proposing ' ,1 13 to you -- five experts is all that's plotted here. That's 14 "C". i 15 MR. SIESS: You're going to use the five 16 everywhere. 17 MR. MURPHY: It.is our intent at this time to use 18 the five. 19 MR. CHOKSHI: I think the point here is that, if 20 you use the same ground-motion model, a smaller close-in and al a larger far-in, what is the impact on the ground motion, 22 and it's not much. l 23 MR. SIESS: I thought you had said earlier you l
- 24 were just going to use the four.
I !' 25 MR. MURPHY: No. That was a misstatement if I l t l l
1 ( 210 1 said it. 2 MR. SIESS: You had a figure that didn't have both 3 of them. , 4 Now, "C" versus "S" -- 5 MR. MURPHY: i Is the comparison that we're a34 Lng i 1 6 to be made. 7 MR. SIESS: Using a model. Of course, "D" is 10 i i 8 miles off of that, because that was probably a Reg. Guide 9 1.60 spectrum, which says to me, gee, have we been doing i 10 things that wrong all these years? i 11 MR. CHOKSHI: We're doing better. 12 MR, SIESS: Why does Reg. Guide 1.60 call for ( 13 almost twice the acceleration that you would get using your ! 14 model? ' j 15 MR. ROTHMAN: Can I make a statement about Reg. 16 Guide 1.607 l 17 First of all, Reg. Guide 1.60 was developed using 18 spectra, about 30 spectra I think, from large distance 19 earthquakes, so that they were very rich in long-period 20 ground motion, but they were all normalized at the high ! 21 frequency. l 22 So, the high-frequency end is controlled by 23 whatever peak ground acceleration you pick, but they are 24 very rich in the longer periods, as compared to nearby 35 recorded earthquakes, because they were large distance
l 'l 211 1 earthquake recordings. 2 So, the Reg. Guide spectrum tends to be_very rich 3 in longer periods. 4 Also, it's a mean plus 1 sigma spectrum at those . S long periods, and I think that the medians are what we use 6 for the "C" and the "S". 7 MR. CHOKSHI: No. 8 MR. ROTHMAN: No? Median ground motion was? r 9 MR. CHOKSHI: Mean plus 1 standard deviation. 10 MR. ROTHMAN: Oh, you use mean plus 1 st8ndard -- 11 okay. Well, anyway, the Reg. Guide spectrum is very rich in la long periods. ]
- 4 13 MR. SIESS
- What does that mean to me in terms of 14 whether we did things right or not?
15 MR. ROTHMAN: It means that we're very 16 conservative with using the Reg. Guide spectrum at the 17 nearby small earthquake. ! 18 MR. SIESS: Now, do we really want to remove that 19 much conservatism? If I like what I have been doing and I'm i 20 not going to change it, why did we just go through this al whole thing? 22 MR. CHOKSHI: I think the next example will show 23 you what happens if we do what we are doing'. i 24 [ Slide.] 25 , MR. MURPHY: If we go back and look at this one,
. . _ _ _ _ . _ _ _ - _ _-.m.._ __ _ _ . . .. . - - ~ . _ . . - _ . . . . _ . _ . _ _ _ . _ _ _
E ( 212 l' which is.viewgraph 16, and concentrate on Perry, the five
~
a ground-motion experts, at'5.5 and 25 kilometers, compare t 3 that with the 5.3 at 15 kilometers for the actual design, l 4 I'll show you another viewgraph, Figure 22 -- l 5 [ Slide.] l 4 6 MR. MURPHY: -- in which we compare -- for a Reg. l 7 Guide 1.60 spectra, we compare our reference, our "R", the l 8 "D" is the actual design, and then "C", in this case, is a i 9 probabilistic controlling earthquake, again anchored as we l 10 would. 11 MR. SIESS: You still don't pick up anymore high-r j 12 frequency stuff. You just raised the whole thing a little 1 ( 13 bit.
~
14 The oddball at Perry was it was a high-frequency, ( 15 had a fairly large component high-frequency, and your j 16 spectrum doesn't have anymore at the high frequency than the ( 17 original did. 18 MR. CHOKSHI: The way this spectra was derived was i ! 19 using this -- you have magnitude and distance from ! 20 probabilistic, 5.5, 24, then work that into the PGA type of i 21 quantity, and then anchor it to the 1.60, the way it was 22 done initially in the Perry . 23 MR. SIESS: The difference between "C" and "D" l 24 here is negligible. f i 25 MR. CHOKSHI: Right. I J
. - . . - . - . . . _ -_. - -- . ~ _ . . . . - . . . . -.-
i l !( 213 1 MR. MURPHY: Right. 2 MR. SIESS: I hope. 3 MR. MURPHY: Yes. 4 MR. SIESS: I mean if we've got plants out there 5 where that'much is going to make a difference, that's 6 negligible. So, what's the point about Perry? The shape is l 7 the same. 8 MR. MURPHY: The point about Perry is just to put 9 the shape problem into perspective. 10 MR. SIESS:- Why are the shapes the same for Perry? , 11 MR. CHOKSHI: We assumed, i 12 MR. MURPHY: We took and assumed and use a Reg. :
~
i ( 13 Guide spectra anchored for Perry. , ; 14 MR. SIESS: Okay. 15 [ Slide.) 16 MR. MURPHY: If we look at Figure 19, which is 17 done the same way as we did the figure for Vogtle, we have l 18 that richness in'the high frequency and the lack'of input in 19 the long period part of it. 20 -MR. LINDBLAD: I'm having trouble. Tell me what 21 the peak ground acceleration will be for curve "C". 23 MR. MURPHY: The high frequency for "C" at Perry, 23 I think, was 15. l 24 Phyllis, do you remember? , 35 MR. LINDBLAD: It seems to plot at 30 percent. I
.. - --- - .. _ = - . . - . . _ -
( 214 1 Doesn't the ZPA show 30-percent ground motion? 2 MR. BERNREUTER: Let me make a comment. Don l 3 Bernreuter, Lawrence Livermore National Lab. l 4 You'are tending to equate from the old Reg. Guide 5 that you reached PGA, effectively ZPA, at 33 hertz, so that 6 you could go out at the end of the spectra and read off your l 7 PGA. 8 In the eastern U.S., that's not the case. You 9 really don't come back down to PGA until you're about at 10 least 50 to 100 hertz, and at 30 hertz, or actually it's 25 L 11 hertz, you're still in the amplified region of the spectra. ( 12 So, the actual ground -- there's is enough high-( 13 frequency energy in there that you're still in the amplified
~
14 region of the spectra. 15 So, that's about a factor of 1/2 or 2 above the 16 actual PGA. 17 MR. LINDBLAD: Thank you. 18 :MR. SIESS: If you look at F'igure 22, which is the 19 Reg. Guide spectra, you can pick off the PGA. It's looks
-20 like about -- it looks pretty darn small to me.
'2 1 MR. CHOKSHI: Almost the same as what the design 22 earthquake was.
23 MR. SIESS: It looks like 0.1. 24 MR. MURPHY: Do you remember what the design -- 25 MS. SOBEL: 0.15.
_ . . _ . _ - . . __ __ _ _ . _ _ . . _ _ _ . . _ _ _ . _ _ _ _ _ _ _ . ~ . .__.--__m _m_ ._m . i l ( 215 i 1 1 MR.'SIESS: It could be 0.15, I guess. I'm i 2 picking it off at the end point there, which is the 33 3 hertz, I think. ( 4 Now, what am I going to do for my site near ] 5 Indianapolis, or whatever site we've already done? What do ; 6 we take for the deterministic? I've done a deterministic on
- 1 l 7 my site in Indianapolis. We haven't done this so I don't 1 l
8 know, so what do you call the deterministic for Vogtle? 9 MR. MURPHY: What we call the deterministic for 10 Vogtle would be a 3.5. I 11 MR. SIESS: Would be the design earthquake. 12 MR. MURPHY: The design earthquake. t 13 MR. SIESS: And-we get a synthetic earthquake from 14 our experts, bearing somewhat -- if we take the'most --
- 15. let's see, are we being consistent if we take 5 all the way l 16 through? If I was doing EPRI, I would get the same sort of 17 thing?
18 MR. MURPHY: Yes. 19 MR. SIESS: If I were doing this whole thing by 20 EPRI, what.would I come up with for an M bar and D bar? 21 MR. MURPHY: We have not done that calculation 22 yet. 23 MR. SIESS: What would it -- but you say in the 24 proposed revisions here that you can use either method? 25 MR. MURPHY: That's correct. l l l
._ _ . .. _ _._.- -.._.--_._ _ ___.._ .- _- _ ... _. m--_ - - ._..- -
[ 216 1 MR. SIESS: But how do you know they.give the same 2 answer? 3~ MR. CHOKSHI: One of the reasons was doing 5 and 4 4, is to test that, okay, in part. 5 MR. MURPHY: The four was an' initial surrogate for 6 doing the EPRI. . The four experts, as you noted in that
.7 first table --
8- MR. SIESS: Yes, they gave the same overall. 9 MR. MURPHY: They gave close numbers.
'10 RMR . SIESS: All right , so now I've done my l
11- deterministic for Vogtle and.I get 5.3 at 1.5. I go through 12 'this.and it turns out that to get 10 to the minus 4, I come t 13 up with a synthetic earthquake of, say, 5.8 at 32. i h
)
l 14 Now, my next step then is to' plot the spectra and I i 15 look a them?
.16 MR. MURPHY: In that case, the next one would be 17 to plot that spectrum and look at them.
18' MR. SIESS: So that's page 18? ! 19 MR. MURPHY: Yes. 30 MR. SIESS: I look at that and I come in to you 21 and say, okay, I want to use that C to design my plan. Is 22 that okay, Goutam? 23 MR. BAGCHI: No. ! 24 MR. SIESS: Why not? Which one do I have to use. 1 i 25 MR. BAGCHI: Goutam Bagchi. I believe that the I l
L 4 I i l i 217 f. 1 important frequency range for engineering significance-is 2 substantially lower. I wouldn't believe we would accept 3 this for design. l 4 MR. SIESS: But according to this, -- 5 MR. MURPHY: Now, let's back up just a couple 6 steps. 7 MR. SIESS: Back up more than a couple. Back up 8 into DG-1015 and point to the portion in that where it tells 9 me what to do. I'm not asking you anymore. I'm asking you I 10 to tell me where in the Reg Guide or in Appendix A of Part 11 100 does it tell me what I do at this stage, okay? 12 Now the thing is that you've got a copy.of the Reg l ( 13 14 Guide and I haven't, but we'll -- MR. MURPHY: We're working with the same copy that l 15 you've got. 16 MR. SIESS:' Oh, okay. I 17
- MR. MURPHY: We don't have,the amended, abridged l l- *
.i l'8 version. ,Now, where you want to look at'this stage is, I. l l
19 believe, on page 26 of the Reg Guide, Step No. 5. i 20 MR. SIESS: Okay, the ground motion which l 21 corresponds to the controlling earthquake is determined -- 22 is outlined in Section 2.5.2 of the standard. 23 Step Four, I have now computed the controlling l 24 earthquake for the probabilistic, right? l 25 MR. MURPHY: That's correct.
I l( 218 1 MR. SIESS: And I'm in Appendix C. 2 MR. MURPHY: Yes, Appendix" C. 3 MR. SIESS: Which is determining the controlling l 4 earthquake from the probabilistic. 5 MR. MURPHY: Right. 6 MR. SIESS: So I've computed the M and D. Now, 7 Step Five says I get the ground motion. Now, what does - , ) 8 MR. MURPHY: What we did in Figure 18 is not 9 exactly what we described in Step Five, because in Step 10 Five, our preference, as stated here, would be either to l 11 find a specific time history for this 5.8 earthquake at 32 12 kilometers and look to see what spectra we would get out of l(' 13 that. !s . 14 The second would be to collect a series of time 15 histories at about that distance, at about that magnitude. 16 We did not do that here. 17 MR. SIESS: Now you have gone off outside the 18 physical world, because you've come up with some sort of an 19 artificial surrogate earthquake, playing with Ms and Ds, 30 mathematically, and now you want to assign a time history to l 21 it. That earthquake never existed, except as two numoers, 23 so how in the devil are you going to assign a time history 23 to a fictitious earthquake? 1 24 MR. MURPHY: The same way we assign it to a real 25 earthquake.
l t ( 219 1 .MR. SIESS: Oh, no we don't. On real earthquakes, 2 what we do is either take a whole lot of time histories -- 3 for real earthquakes, all we do is take a standard design 1 4 spectrum and get.some synthetic time histories that will 5 give you the same spectrum so you can stick them into a t 1 ! 6 computer. 7 MR. MURPHY: Or, if we were doing Vogtle today, -- l 8 if we were doing Vogtle for the deterministic today, we ! l l 9 would ask them to come up with a time history of series of l l 10 time histories represented by the 5.3 at 15 kilometers, and i 11 then we would normally take the mean plus 1 sigma of the 12 envelope of the spectra that we get from those time l( 13 histories.- Okay,.that what we would do today to come-up l l 14 withLthe deterministic, i 15 We're saying, do the same kind of process with the l 16 probabilistic controlling earthquake, i.e., go get a set of l 17 real time histories that are approximately 32 kilometers l ' l i 18 from the recorder and approximately of the magnitude of 5.8. I 19 MR. SIESS: Why wouldn't you just look at this 20 thing and say that you've come out with 5.3 at 15 in one 21 case and 5.88 at 32 in the other, and there ought to be some 22 way at arriving at roughly what a PGA is for that magnitude l
- 23. at that distance? There are some plots for that; aren't l 24- there?
25' The PGAs come out about the same, go ahead with
J l 1 l ( 220 1 your deterministic analysis and design the plant. l 2 MR. CHOKSHI: Look at page 8, Dr. Siess, on the l 3 draft guide.
'4 MR. MURPHY: Page 8. l l
5 MR. SIESS: Got it. 1 6 MR. MURPHY: The last paragraph. 7 MR. CHOKSHI: I think it tries -- l 8 MR. SIESS: Oh, yes, that's a great one. That's 9 what I thought you were going to point to. Now, you look at 10 them to see if they're similar. 11 MR. CHOKSHI: Right. 12 MR. SIESS: Okay, and also to determine if the l(, 13 controlling earthquake is easily identifiable. 14 I assumed that this means that if that matrix you 15 had up there before has got one box with ten to the minus -- 16 2(10-4) that is pretty darn likely to be the --
~
17 MR. CHOKSHI: The M&D would be the same thing 18 though. 19 MR. SIESS: Okay. Now, if the dominant 20 controlling earthquake can be identified, ground motion to 21 be determined only for this one -- that's if you put the 22 Charleston dominated which we're not going to get that close l 23 to those very often. The controlling earthquakes and the 24 two approaches are significantly dissimilar. Those are real 25 good words, "significantly dissimilar" -- then the ground i
- ~ . . - . , _ -
1
\
I 1 l lI 221 l l l 1 motion estimates may have to be determined for various ; l 2 controlling earthquakes and compared to derive the final ]
)
3 ground motion estimates for use in establishing or comparing ; l . i l 4 with the SSE ground motion. l 5 Now, somewhere in here you have got three 6 different earthquakes. You have got a controlling -- you 7 have got this thing, you have the controlling earthquakes, 1 1 \' 8 that you enveloped before and then you've got the -- various 9 controlling earthquakes. Now, that is a reference to the ! 10 deterministic approach; right? Is a controlling earthquake i 11 something that comes out of the'probabilistic? 12 MR. MURPHY: You get controlling earthquakes;out ( 13 of both the probabilistic and the deterministic. 14 MR. SIESS: Right. But when you say various t 15' controlling earthquakes, that means I can have three out of i 16 one and two out of the other? There'is something lacking in l 17 the Reg Guide. After,I compare them, what do I do? After 18 all, I have gone through this long involved process to get a 19 second value that I compare with my first value. l 20 MR. MURPHY: That's right. l 21 MR. SIESS: Now, what do I do? 22 [ Slide.) l 23 MR. MURPHY: If you can in this particular case -- 24 .1 you want to intuitively say that a magnitude 5.8 at 32 4 25 Kilometers will produce the same ground motion as a 4 1
! 222 ,
1 magnitude 5.3 at 15 kilometers, you say I'm done, I'm j 2 finished, we've got controlling earthquakes to use for the l 3 design. If you want to intuitively say that the 5.8 at 32 1 ! 4 is equivalent to the 5.3 at 15, you go to the next step, S which is Step 5 and you calculate ground motions for those 1 6 two earthquakes. I 7 MR. SIESS: Now, let's take that first statement 8 you've made. I think they're the same. I can do certain 9 things. I'd like to hear NRR tell me that. They are the 10 ones that have to accept it when I bring it in. , I 11 MR. MURPHY: I would have to look at the response 12 spectra from those. I couldn't offhand say that is what I '(. 13' am s,aying. l 1 14 MR. SIESS: So, I am now back to Page 18. I have' 15 got two response spectra. I have got one called C and one
]
16 called D. Now, looking at those two numbers up there, I've
, 17 done whatever is necessary to get them out to the site, two 18- response spectra; right? Is that wha't you meant?
19 I've given you the response spectra, C and -- 20 MR. MURPHY: I'd say essentially within the 21 variability and ground motion I'd say that those are 22 essentially the same. 23 MR. SIESS: I said, okay, I'd like a design for S. 24 Goutam wouldn't let me. 25 MR. ROTHMAN: What I would say is those are a l L
. . . . - . .- -- . .._. - .-. - - , - . - . - . ~ . _ . . . ~ . . - - , ._
I' l l i. [ 223 l' seismological characterization of those earthquakes that l '2 could be expected. Now, structural engineers have to decide 3 what they want.to design. 4 MR. SIESS: Okay, what you're willing to say is I 5 that the difference between D and S -- 6 MR. MURPHY: C and S. l l 7 MR. SIESS: Or D and C, it doesn't make any l 8 difference. You said you wouldn't be able to look at the l 9 table, the 5.8 at 32 and the 5.3 at 15 and say they're the q 10 same. You said-you would like to see the spectra. l 11 MR. ROTHMAN: Yes. l l 12 MR. SIESS: Now, here I've got the spectra. The i [ 13 one labeled D is the 5.3 at 15 -- 14 MR. MURPHY: No, C and S. One labeled S -- 1 l 15 MR. SIESS: I'm sorry, I'm sorry. C and S; is l l 16? that right? l l 17 MR. MURPHY: C and S. C is the pr.obabilistic, S 18' is the deterministic. 19 MR. SIESS: I have got it. I apologize, I was 20 confused here. l l 1
.21 Now,rso most of us would say that's not bad. It l
23 doesn't make any difference which one I use. l 33 MR. ROTHMAN: Dr. Seiss? i l 24 MR. SIESS: Yes? i ,
. 35 MR. ROTHMAN: For Vogele also what we would do is
i (' 224 , i 1 we would look at the magnitude seven say in Charleston -- ! 2 MR. SIESS: This isn't Vogele, this is myisite in ' 3 Indianapolis. ; 4 MR. ROTHMAN: But I would be looking over at the 5 Wabash Valley. l 6 MR. SIESS: I don't want you to know anything more l 7 than what the Reg Guide calls for. ; 8 MR. ROTHMAN: Except in my deterministic analysis l 9 I would be looking for magnitude six and a half -- l 10 MR. SIESS: You've already done your deterministic L l 11 analysis. ; i . l 12 MR. ROTHMAN: Yes, but that's not the spectra for l ( 13 that. What I am saying is this is from 5.3 at 15 14 kilometers. For Indianapolis we would also have to consider 15 the Wabash Valley zone which we.didn't consider in here in 16 our deterministic analysis.
. 17 MR. SIESS: I am sorry, we have already done the l
18' deterministic analysis at'Vogele, and that's what gave us - 19 the design basis. 20 MR. ROTHMAN: No, no, no. 21 MR. SIESS: It wasn't done deterministic? 22 MR. ROTHMAN: No. What I am saying is that that S 23 is not -- the design basis for Vogele is .2g Reg Guide 160, ( 24 not that S. 4 25 MR. SIESS: I am sorry. I am talking about the 2 l i I <
-_. ~_ . .. . _ _ - . _ ___ ._ __ __ _ . - _ _
j( 225 1 earthquakes, and somebody is confusing the hell out of me. 2 We have two different designs, we have two different bases. 3 We have 5.8 at 32 and we have 5.3 at 15. 4 MR. ROTHMAN: But also in the Vogele review the 5 deterministic earthquake -- magnitude seven earthquake at 6 Charleston was also considered and the long period ground 7 motion was looked for that. 8 MR. SIESS: Then Figure 16 is wrong. l 9 MR. ROTHMAN: Figure 18 I think it is; right? 10 MR. SIESS: I think you people should get
- 11 together. Look, at Figure 15 there is a table labeled l 12 " Design Earthquake".
13 MR. ROTHMAN: Right. 14 MR. SIESS: And that is not right. 15 MR. ROTHMAN: What that is is 5.3 at 15 kilometers 16 was considered, but also -- 17 MR. SIESS: It's not right. It's not the design 18' , earthquake. Thank you. 19 Now, then, this example then does not give us the 20 whole story. 21 MR. CHOKSHI: Your Honor, I think the way it was 22 done, this is 5.3 at 15; okay? And you compute PGA and then 23 anchor to 160, okay -- 24 MR. SIESS: Now, that's not what he says. I will 25 give you guys ten minutes. Just a minute, sit down. I
_ _ .._ . ~ ~ . l 1 ( 226 1 MR. ROTHMAN: I don't know how they got the .2g 2 PGA for the anchor at 160. l 1 But what we did is that was the 3 design, .2g PGA for 160. We then looked at a 5.3 at 15 4 kilometers and also the charleston which was -- what was the 5 magnitude? -- 80 miles, magnitude seven earthquake to l 6 estimate the ground motion. And we made sure that that did 7 not go above the design basis. 8 MR. SHAO: The earthquake you are talking about l 9 was -- 10 MR. ROTHMAN: .2g Reg Guide 160 enveloped. 11 MR. SIESS: Anybody that started this off saying ! 12 this was going to make it easier to license the plants and 13 easier to reach agreement -- you guys can't. agree with each R 14 other, much less a poor licensee coming in and trying to get ! 15 somewhere. But let's accept for a fact -- a hypothetical -- 16 I was trying to get away from Vogele -- I have got l 17 something that was designed for 5.3 -- or whatever the two 18 numbers were -- l 19 MR. MURPHY: 5.3 at 15 and 5.8 at 32. 30 MR. SIESS: I didn't design for it, I got that by l 21 the deterministic method. I get 5.8 at 32. Some people at i 22 least on the staff might not be able to look at those and 23 say they are the same so they can plot to calculate spectra, 24 and you have done some spectra just for those two 25 earthquakes and you look at those and you certainly don't 1 l
- . _ .___ _ .._ .. _ . m _ .. . . _ . . _ . . _ . . . _ . . - . . m_.._ _
t i I l (. 227 l' see'much difference. Without regard to those spectra you 2 'can say, well, I will pick one of them. And if you ask some 3 of the staff which one do you want, they will take the one , 4 which gives the highest spectra probably. 5 MR. MURPHY: Probably. 6 MR. SIESS: Now, you might not design for that 7 spectra, you might go back and develop some other. But 8 again, basing on that magnitude and distance as a start. 9 MR. MURPHY: Right. l l 10 MR. CHOKSHI: Right. 11 MR. SIESS: What happens if the probabilistic L 12 comes out, instead of 5.8 at 32, it comes out 6.5 at 35? l( 13 MR. MURPHY: That's where we fall back on the de-l t 14 aggregated information that we've got. 15 MR. SIESS: And you got back through the whole l 16 thing, and it says, yes, that's true. This thing just l 17 doesn't -- it's not better than half the plants out there. 18' If I design it for what I got by the deterministic i 19 method, my probability of exceedance is going to be 10 to 20 the minus 3 instead of 10 to the minus 4. Isn't that what 21 it means? 22 MR. CHOKSHI: If you find those reasons and if you 23 go back and convince that your probabilistic method was --
- 34 there was nothing anomalous in it --
l 25 MR. SIESS: I'm assuming it's right. After all,
(. 228 1 it ~ gave you a lot of plants at 10 to the minus 3, 3 times 10 2 to the minus 3. It's possible to have another one. 3 MR. CHOKSHI: But there's the point that, in some 4 deterministic models, you did not -- 5 MR. SIESS: Okay. Now what I do? I am now -- I l 6 have gone through my deterministic method, and I have gone 7 through the probabilistic, and what the probabilistic says 8 is that, if I want 10 the minus 4 probability of exceedance,
)
9 I'm going to have to design for a lot bigger earthquake. 10 What am I going to do? What do I design for now? 11 Half the plants out there have a probability greater than 10 12 to the minus 4. You haven't made a requirement that my ( 13 probability must be less than 10 to the minus 4. l 14 MR. MURPHY: No. Among other things', it's called 15 a target. 16 MR. SIESS: So, what am I going to do? 17 MR. MURPHY: I would say the first thing that 18 you're going to'do is that science thing and that's to go 19 back and to figure out and to understand why the 20 probabilistic came up with that particular earthquake 21 magnitude / distance bin and showed you an earthquake -- in 22 effect, it's showing you an earthquake that had not been 23 considered in the deterministic case, and either you have a 24 glitch in your analysis, your probabilistic analysis, or I - 25 there is something specifically out there that should have i
l i
- ( 229 l 1 been considered deterministically.
2 MR. SIESS: All right. 3 Now, what you're saying, then, is, at this point, 4 I have two choices. I can choose the one that gives me the l 5 greatest value of whatever, or I can go back until I get 6 them to agree. 7 I can redo both of them, except, you see, I can't l l 8 do much redoing on the probabilistic, because it's in the i 9 computer. l l 10 MR. MURPHY: That's correct. I 11 MR. SIESS: As long as I put those coordinates in, l 12 I'm going to get the same answer every time. 13 So, now I have to say the probabilistic is right. ( l 14 I've got to go back and figure out why my det'erministic is 15 wrong? Okay. Then why did I even bother to do the 16 deterministic in the first place? There's no chance of l 17 changing the probabilistic one.
'18 MR. CHOKSHI: That's right, but you may not agree l 19 with the probabilistic. You're not changing it to believe 20 it, but when you go back and you find --
l 21 MR. SIESS: I don't agree with it. I think my l 22 deterministic was right, but I came in to you to get a 23 license. 24 MR. MURPHY: Yes, but if you come and say that 25 there was one expert which assigned to this zone a rate of
I 230 1 occurrence which is -- ! 2 MR. SIESS: You want me to come argue it. 3 MR. CHOKSHI: Sure. 7 4 MR. SIESS: Oh, heck, we can argue without going ) 5 through two procedures like this. j 6 MR. CHOKSHI: No. It's not argument, but I think l !
- 7 that the applicant will have to go through that process !
L 8 itself to reconcile his findings. 9 MR. SIESS: I've got a fat chance of challenging ! ! 10 the probabilistic thing. 11 MR. MURPHY: You do. If this scenario that you've l 12 outlined happens, and my initial suspicion is that it's not 13 -- that's just an educated guess at the moment. j(' 14 MR. SIESS: I mean you've got to assume it will. 15 Otherwise, you wouldn't be wanting all this. 16 We've started out with the assumption that we're 17 going to reduce the uncertainties, reduce the basis for 18 argument, make it easier to come up with a number with i 19 confidence, but that's true only if we get the same answer, ! 20 and if we get the same answer, we don't need two methods, i i 21 and if we can't work it when we get a different answer, two i 1 22 methods isn't helping us, unless we always know which one is 23 going to give the safer value. 24 HR. MURPHY: Okay. And to date, we have done four ' 25 trial calculations, and to some level, we'll say that we l
i : ( 231 1 think we've gotten basically the same answer when we look at l 2 the ground motion for those four cases. l ! 3 There may be argument about that. Wo won't get 4 into that at the moment. 5 MR. SIESS: Well, there's nothing in here as far 6 off as I mentioned. 7 MR. MURPHY: And what we are intending to do while i j 8 this document is out for public comment is to run the other l 9 65 cases to confirm that that is, indeed, the case. l [ l 10 That will not answer the question for every I l 11 conceivable site in the eastern United States, but it 12 certainly will very significantly blanket the cases for the ( 13 eastern United States and put us.into a very good regulatory 14 position for the final rule. 1 15 MR. SIESS: But now, in the first place, it's not 16 going to be this good for all of them, because we've got 17 minimums _and we've got some arbitrary things and, you know, 18 Diablo wouldn't -- well, west coast we can leave out. 19 MR. MURPHY: Yes. West coast doesn't count. 20 MR. SIESS: Now, if what you're saying is, if l 21 we've done 69 sites and they all come out as good as these, 22 which I think is darn good, does that mean that we could -- 23 gee, what we've been doing was just fine. Let's keep on 24 doing it that way. 35 MR. MURPHY: Okay. J f I i
. - .. -.- _ . - .. - - .-- ... .-. _ . ... _ . - . _ .. . - . . . - . ~ - - - - - . . .- -
l l i ( 232 l l 1 MR. SIESS: Or are you saying that we've just been ; I
'2 lucky that what we've been doing was fine; we.need this 1 3 probabilistic check to see if we get the same answer?
4 How did we manage to do such a good job in the 5 past? By the arguing process? 6 MR. MURPHY: It certainly contributed to it. 7 MR. SIESS: Now, what, then, are we accomplishing? 8 See, the old process was the licensee came in, the 9 applicant came in, he had done his geology and seismology, 10 and I never saw anybody yet that came in with a g value that i 11 the staff accepted the first time, and sometimes, even 12 they'd up theirs, and it wouldn't get accepted, and then ] 13 there was a back and a forth, and they usually ended up a 14 little higher than they were'before. 15 Now, is that the way we're going to continue to do
- 16. it on the deterministic, or are you going to tell them to l
l 17 say, okay, you do the deterministic, you do the l !' l 18 probabilistic, bring them both to us, and if they are close, 19 we won't argue with you; we'll assume that you probably did
~
1 20 the deterministic pretty well; we won't go out and review ; y 21 every bit of your geology and seismology and challenge, like 1 22 was done in the past. i 23 You know what I'm talking about. You know how you e 24 did it. You reviewed their exploration and made them dig
- 25 some more trenches and had your own experts. The USGS would i
[ 233 l' argue. o 2 But you're going to say, if you do it both ways 3 and they agree this well, we won't check you? l I 4 MR. MURPHY: We'll ask for comment from Bob. 5 MR. ROTHMAN: First of all, this is for the 6 vibratory ground motion. We still have to worry about other 7 site conditions. 8 MR. SIESS: Right. I'll concede that. 9 MR. ROTHMAN: We still have to review the l l 10 liquefaction, displacement, amplification. We have to look 11 for seismic sources that may be closer to the site than 12 those that have been considered in the past, things like l ( 13 that. 14 MR. SIESS: Oh? 15 MR. ROTHMAN: So, we're not going to write off on l 16 -- L 17 MR. SIESS: You just sold me now. , l 18 S o', I don't get -- I'm an applicant, a license 19 applicant for a site permit. I don't get a thing out of 20 doing it both ways. If I get more-confidence, unless you 21 get more confidence, you know, we haven't accomplished much. l 22 1ha both have to be confident before we get this plant built. l 23 MR. MURPHY: What Bob has told is that there are
- 24 considerations other than vibratory ground motion.
35: MR. SIESS: There might be, but 90 percent of the l l I
I 8 1 l l 234 1 extra work that was done in the past was on vibratory ground 2- motion. People got away from liquefaction as soon as they 3 pointed -- got off the flood plane, because they had to put 4 cooling towers in. 5 We haven't had liquefaction problems since the 6 very early days when we were building plants on the flood 7 planes. 8 But again, I'm just taking you at your word as to 9 what we think we're going to accomplish by going to a dual 10 procedure, even if the dual procedure is something you can 11 stick in one end of the computer and get the answer out the 12 other. 13 MR. MURPHY: Like I said, we.are anticipating ( 14 putting this document together to simplify the' process of 15 licensing.the plant, and we feel that, indeed, by doing the 16 dual procedure, that the review that will have to be done 17 and the discussion of that with the applicant will be
'18 simplified, streamlined, by doing'this process.
19 MR. SIESS: Well, you're saying that, but you're 20 not doing that review. 21 MR. MURPHY: No. I have only done one review, 22 only participated in one review. l 23 MR. CHOKSHI: But Dr. Siess, you know, pointing 24 back to that '86 symposium and a lot of various other 25 discussion with others, that seemJ to be the major concern l
_ , ~ . _ . - . - - . - . - - - - - . _ _ . . . - . . - - - - . - . - . - - - . -. . - - . . . . , . - - . - l l 235 {' . 1 of a lot of people, the deterministic process is not unique. ! l 2 Different analysts will come up with different things. i i 3~ MR. SIESS: Don't tell'me the probabilistic one is , I 4 unique. 1 5 MR. CHOKSHI: No. 6' MR. SIESS: No, of course it isn't. All you can ! 7 do is compare, relatively. speaking. 8 MR. CHOKSHI: Yes. 9 MR. SIESS: You've admitted that the probabilistic J 10 is not unique. You've got two of them. You're just saying 11 each one came out with this value. 12 MR. MURPHY: Individually, we're saying they are l(\~ 13 unique. 14 ..MR. SIESS: Deterministic is not unique. What do 15 you gain by this? If you are going to have USGS, as it used 116 to be, check everything the licensee does, read his 17 proposal, check his geology and so forth, that was the way 18 ws avoided onder conservatisms, pure mistak'es or whatever. l 19 .Now, if you're claiming that the probabilistic method is a I . I 30 way of checking on that and we're still going to have 31' consultants checking everything they've done, we're not 23 going to be any better off. l
'23 I'm-looking for a way to be better off in some !
l 34 way. ) ! 25 MR. LINDBLAD: It seems to me that when people in 4
'I
- ,...._. .s - _ _ _ _ - - - _ ,
I 236 1 the past have talked about the probabilistic approach, 2 they're also talking about the contribution seismic hazard 3 had to safety. And in terms of the seismic margins and the 4 other way to the final result and how it affects health and
, S safety, probabilistically, the seismic hazard is thought to 6 be rather small.
7 I think that's why there is so much commotion in 8 years past about the attention paid to deterministic 9 resolution of geologic hazards when it had a tenuous 10 connection to safety of the public. 11 MR. CHOKSHI: As I understand it, one of the 12 concerns is things like the maximum credible earthquake. ( 13 It's just an open-ended question, and the probabilistic at 14 least allows you a way of incorporating different views on 15 that. 16 MR. SIESS: But you still don't do the right 17 things. Can you tell me that if I've got a design for a 18 PGA of 3/10ths, and that's say, a probability of 10 to minus 19 4 of exceedance, and really, there's a probability of 10 to 20 the minus 4 that the PGA will be not 3/10ths but 4/10ths. 21 How much have I increased my risk? 22 I've got a plant designed for a PGA of 3/10ths 23 with code values, conservative assumptions, conservative 24 damping, assuming pipes are going to fall in a way they 25 can't possibly fail, and I up my peak acceleration 33 l l I
- . .. . . . - . _ . - - - ~ . - = - - . . - . _. ,
l l f 237 1 percent; do you think I'm going to increase my risk by 2 anything like that? 3 The probabilities aren't the big factors here. 4 Whether the probability of exceedance is 4/10ths or 10 to 1 5 the minus 4, 10 to the minus 5, that doesn't mean it's going 6 to fail just because it's exceeded it. Margins tell us that 7 we've got a lot of -- what do the seismic margin studies 8 show us, leaving out liquefaction? It takes, what, two or ) l 9 three times the SSE to start to get any contribution to ' i 10 risk. l 11 Nothing we're going to change is going to reduce 12 those margins that much. In fact, as we go along, we're l { 13 getting better about bracing things and anchoring stuff down 14 and electrical cabinets that won't fall over. We're ! l 15 learning things. i 16 But again, I'm getting the impression that if I { i 17 were a 1icensee, I could probably get this probabilistic i 18 analysis done in a couple of days with 2 or 3 man-days, plus l 19 Whatever I pay somebody for computer time, so it's not a l 20 great big burden. 21 But what do I get out of it? You see, that's my l 22 problem. I don't see that we've gained a lot in the 23 process, because when I get down to the bottom line, if 34 these are the same, what do we do about it? If they're l 25 different, what do we do about it? i l
238 l 1 I haven't gotten an answer. 2 MR. SHAO: To answer all the future questions, say 3 this site may be subjected to a larger earthquake, we'd have 4 to look at all the earthquake. The policy is that we look 5 .at all the possibilities. 6 MR. SIESS: We're going to take the -- 7 MR. SHAO: There won't be any more questions like, 8 say, suppose the earthquake is twice as large; what happens 9 to the site? We have looked at all these earthquakes. 10 MR. SIESS: Because we're not up to twice as 11 large. We may be just twice as probable, which isn't the 12 same thing. But we'have looked at what happens if it's 13 twice as large. That's the margins program. 14 That's the question we ask; what happens -- 15 MR. SHAO: In Maine Yankee, you look at the margin 16 and it's there to fix a lot of things. In order to meet 17 this,.they had to fix a lot of tank, ranchor the bolts,
~18 anchor the foundation.
19 MR. SIESS: But the ACRS asked the simple 20 question; what happens when you exceed the SSE? That's what 21 worried us. I mean, the -- back when we started in this 22 business, most of us thought that the SSE was the maximum 23 possible earthquake at that site. 24 MR. SHAO: Yes, but that's not so. 25 MR. SIESS: Then we began to get somebody's
f 239 1 probabilistic studies, geologists, of all people. They come i 2 in and said, gee, the return period on your SSE is more 1 l 3. likely to be 10 to the minus 3, 10 to the minus 4 years. i ! 4 This isn't even a 10 to the minus 6 thing you're designing 5 for. 6 Then the thinking began, it's still a low l 7 probability event, but how bad is it if it happens? Suppose l l 8 we do exceed the SSE? Is the world going to end? Is it 9 going to come down on top of us? How much margins do we l l 10- have? l 11 And'you know how many letters we wrote about 12 margins, and we had a program that was done and it was a ! 13 very enlightening. program. We learned a lot about these 14- plants. -We learned where the margins were and how big they 15 were. 16 And we maybe even learned a little bit about why , 17 they were there,,because all of the assumptions got piled 18 up. But we've got margins, we know.that. If -- you know, 19 if an earthquake bigger than the SSE occurs, the margin 20 studies aren't even going to tell you what's going first, 21 unless it's a real big one and you've got a tank that you 22 know is just a shade above. 23 What's going to fail that's going to get you in I i 24 trouble is a transformer that falls over or an electrical ! 25 cabinet that falls over or a pump that wasn't bolted down or ; l
- ~
t l t l l( 240 1 steam generator where the snub has got locked up. You know l 2- what I'm talking about. l 3 It's not going to be anything that you can relate l 4 back to the design value; it's going to'be something that 5 just wasn't done right. The extra assurance that I'd get 6 from going from 2/10ths.to 3/10ths G isn't that -- isn't , 7 very much. .You can make the mistakes just as easily at i l 8 3/10ths G as you can at 2/10ths.
- 9 The big increment in seismic-safety for nuclear
-10 plants was going from zero to X. Once you started thinking 11 seismic safety'at an nuclear. plant and looked at batteries ]
l 12 that could fall over; once you made any kind of an analysis, l 13 you made the-biggest possible step. ( 14 From there on out, it's'asyntotic. I'm still 15 looking for where's the advantage. Maybe this isn't much of 16 a complication. 17 MR. CHOKSHI: I think the intent is that -- we're I 18 ' talking ab'out now-siting.-- our administrative process seems 19 to have worked, but it's always open to question. If I 20 assume this, if somebody's expert says this, here we'll have l 21 'a' body of information-which will already have that 22 information accounted for, and I think as Larry said, bring ; 23 the hearing process or bring that review process. j 24 If somebody raises that question, in the past, you i
- ;25 had to go back and redo some things; here, in many cases, l.
i
- f 241 1 you will have an answer. That is the biggest advantage.
2- That's stability in that sense.
~
3 MR. MURPHY: It does the same thing that the 4 margins program did for the facility itself; that it 5 provides a margin to accept new ideas, whether they develop 6 during the license review process, or after the plant has 7 been licensed. It provides a margin for the consideration 8 of a new hypothesis or a newly discovered earthquake or 9 newly postulated earthquake that we don't have with the 10 deterministic process. 11 If we have that done specifically, that's where we la are. l( 13 MR. SIESS: Now, let's suppose we go to a hearing 14 and somebody says, good' gosh, you guys didnt count this j 15 earthquake'out here. You know, several of us could agree
- 16 that it's at least a magnitude 6 and it's roughly in that !
17 area. Now, if you haven't got that built into your . 4 18 compu'ter program, youre probabilistic.iaethod doesn't help 19 you. It's not in there. It wasn't in there to begin with. L 20 MR. MURPHY: That's the first case. If it is not j 21 in there, you can take and turn around and do a sensitivity 23 study. You can run the calculation with that model in there l ! 23 and see what effect that does have on the output end and see 24 if that changes the spectruin. 25 MR. SIESS: It is not at the deterministic either. l l l l . .
l i l I a ( 242 J l 1 How much trouble is it to put it in there? l 2 MR. CHOKSHI: Suppose the deterministic does give 3 you a different ground motion that you don't have-any i ! 4 choice? 5 MR. SIESS: You will argue that the probability is 6 low. 7 MR. CHOKSHI: Right. 8 MR. MURPH,Y: One of the basis today for arguing-l 9 the probability is not low. 10 MR. SIESS: The probabilistic will throw it in l 11 there, too. So, you are going to argue the probability is I 12 low. And what makes you a better expert on probability than 13 the other guy? I think you are kidding yourself as to what l( . i 14 you are going to save in the hearing process. I guess 1 l 1
- 15 have heard nobody arguing that you are really going to l 1 l 16 simplify the review process. Every time I point out that l
17 you are not simplifying the review process, you say, well,
- 18. but it will help at the hearing. I am willing to accept L 19 that argument.
20 Well, I tell'you, I understand that process very 21 well now. Let's get down to the last bit, and I think that l 22 that is a weakness in the Reg Guide'though. I 23 MR. MURPHY: Okay. It is a point we can consider. 24 MR. SIESS: I am not even sure that Part 100 does 25 more than mention probabilistic. At one place you refer to l l
5 ! a i I l( > 243 1 this -- 1 1 2 When I went through the Reg Guide -- 3 Incidentally, has this been through the technical l' 4 editors? ! l 5 MR. MURPHY: I think the first five pages were. I 6 She got distracted to other assignments. 7 MR. SIESS: I am sorry. I have a couple of I 8 problems that Item 2 is expected maximum earthquake, but L 93 Item 3 is probabilistic seismic analysis, which isn't 10 exactly a parallel structure-to deterministic and 11 probabilistic. 12 There is a discussion on Page 6 about 13 probabilistic. seismic hazard analysis., It says the PSHA can 14 'be used to -- and it lists all of these' nice things that it 15 can be use'd for -- I am serious. Except when it gets down 16 to E, it says it can be used to demonstrate the probability 17 of exceeding the SSE and compares favorably -- and must it . 18 b5 -- I couldn't figure out if that is an option.
- 19 Regulatory position C-9 says that you shall l
30 perform a probabilistic seismic hazard analysis and you 21 shall demonstrate that the probability of exceeding the SSE 22 compares favorably, et cetera. That is a requirement. I
- 23 think you sort of mixed up on -- some of the things that 24 you've got listed there are nice things and some of the 1
l 25 things that are covered by this Reg Guide. You might want i 1 i l l
l L l > l i- l
\
l. l L !( 244 1 to think about~ eliminating tid things that aren't used here. l 2 At the top of Page 7, you slipped up again on the l l 3 language. Lines five through seven, those are mandatory
'4 words which don't belong in the Reg Guide; am I right? !
5 MR. CHOKSHI: Yes, that is coming out. j 6 MR. SIESS: I:still say that when I get through j l 7 and read-Regulatory Position C-9, it says demonstrate that
^
L 8 the-probability compares favorably at'the currently . l 1 9 operating plans. It refers to Appendix B. 1 10 The last paragraph on Page 8 -- it doesn't help me l l 11 know what you-are going to do when they're different. That l l 12 is what a Reg Guide is supposed to be. l 13 Now, th.ere was.a time when Reg Guides codified o .... 24 . procedures that had been arrived'at over the last ten' years 15' of resolving issues. 'Now, we are breaking new ground with l 16 Reg Guides. l l l 17 .And Appendix-B is simply -- you make.the l l, , . - 1 1 18 comparison; right? i 19 MR. MURPHY: Right. 20 MR. SIESS: It doesn't tell you what to do. You l
- 21 could have an example, but it really doesn't carry you far 23 enough there. Appendix B is stated probabilistic comparison-23- of safe shut-down earthquake for two operating plants. That L
24 just tells you how you get to that first stage. It doesn't 1 ! 25 tell you how to compare it with -- l i
l i e : l[. 245 : 1 1 You have eliminated the mean in the 85 percentile, i 2 .so that -- i 3 MR. MURPHY: That is right.- l 4 MR. SIESS: Some of these things you may have i 5 caught. B-1 and B-2 on Pages 21 and 22 -- B-1 these return 6, periods.- And then when I go to the next page I have got 7 probabilities. I will admit I could make the transition ;
.8 L when I*always feel like I shouldn't have to. ,
- .5F Now, B-3 will be limited to just the median
! 10l figure; correct? ; i 11 MR. MURPHY: Yes. 1 12 MR. SIESS: In Appendix C, page 25, line 17, it ; l(. l 13 says, " discussed in Section C-3." There are two Section l 14 'C.3's, Section C.3 of the Reg Guide which isn't numbered ) l l 15 3.C, but Section 3 and C. Some of my questions got ! l 1 16 answered. j l' 17 On page 27 in line 15, I got. lost when it said l
)
i u 18 extrapolating. I couldn't quite figure what extrapolation l l 19 was on the 8 centimeter per second values that.come in 20 between the 5 and 10 values that are in the table. Is some l
. 21. other extrapolation involved?
l 32 MR. MURPHY: No, that was it. 23 MR. SIESS: That's in interpolation. !. I l1 24' MR. MURPHY: Interpolation. Indeed, for 25 clarification, we may substitute figures in there rather l l l
l l-I I i l 1 246 l l 1 than the tables. L l 2 MR. SIESS: Okay. Those other things look like 3 typos and I'm not going to bother with them. But speaking l l 4 for myself, I understand the process now. I understand what l , 5 you're doing. , I l 6 I think I understand why you're doing it. I don't l l 7 think the Reg Guide goes quite far enough in telling you 8 what to do or what you will do at a certain point. It 9 describes a lot of the intermediate steps, but it doesn't l 10 even say that the staff will advise you what to do when you I 11 get different answers. i 12 MR. MURPHY: Right. I 13 MR. SIESS: You see, I think that's the.part
~. .
14 that's 'alweaknes's. Anybody that looks at this and tries to l 15 carry it down to the final point -- I personally don't think l 16 that what we call the deterministic method will turn out to i l- 17 be as good as it is for these few plants as for all the l l ~ l 18 plants. 19 I'm not sure it makes any difference. I think 20 we've ended up with G values in the right range for most 21 plants, even the West Coast, but that's just a personal 22 feeling about'how we handle this whole reismic thing. 23 What I'd like to -- we have ci.' er things scheduled 24 today, but before we ge to them, -- and this is my personal 25 feeling, for the benefit of the Subcommittee members as much l l l i l l
l i
, i 247
!(. 1 as anything else -- I think I'm willing to tell the staff, ) 2 tell the Commission, that as far as we're concerned, they 3 could and should put this stuff out for public comment, and 4 there's a lot-that needs to be done before they finish it 5 up, which everybody admits is true, because they -- and we 6 might add a comment to the effect that we're not -- that I'm 7 not too' happy about the reasons for having the dual 8 methodology. I don't see that many real. advantages to it. 9 On the other hand, I don't see a whole lot of 10 costs to it. I. understand that the industry has been 11 working on the same problems of seismic siting and have some 12 proposals coming up in a year or so. They don't work a heck ( 13 of a lot faster than the NRC, I've found, but as long as the 14 *s t n.f f is in the' Reg Guide, it's fairly - .well, not fairly'- 15 - it's a little more flexible than if it's in the rules.- ; 16' We always say it's easier to revise a Reg Guide j 17, than it is a rule, but I've never seen the statistic's t at
,18- prove it. 'I think'it takes three years to do anything, and l l
19 we're taking ten years to do a Reg Guide here, but that's i l 20 beside the point. l al So, I'm figuring, if we were going to recommend to 22 the-Full Committee, just to say, okay, send it out, but tell 23 the Commission, please don't take any action on this till 24 you're really had a chance to air it publicly, if that's
- 25 what we were going to say, frankly, I don't see any reason I
l
l 248 j 1 to have them in Friday for the Full Committee, because for i 2 that kind of a letter, we don't really -- they'll take a 1 i 3 recommandation from the Subcommittee. 4 If I-wanted the Full Committee to write a letter, l l 5 really questioning the'need for the two, or the desirability l 6 for the two. procedures, then I think the Committee should l 7 have a chance to hear the arguments and so forth. I have a 8 feeling that. those arguments will.come out once it's 9 published. 1 10- I would hope that the industry -- and we may hear
~
11 something today -- wold have some arguments against it with 12 some pretty good reasons for not doing it that way, or maybe 3 13 they'll see the light.and decide it's not a bad way to go. 1 j 14 .I don't know. But we'll decide on that before we close l l l- -15 today. I have a feeling you're going to come in Friday 1 l l 16 because we' don't want to change the agenda. ! {' 17 Unfortunately, I won't be around to close this 18 issue out 'after publication, but I'm really not aiming at
- 19 .getting my licks in before I leave. So, I have no problem i-U 20 .with telling you to go on out and let the public beat on you 21 for a while.
22 We did not let you get to your last page, Andy, 23 your next to the last page, the rationale that Dr. Lewis was 24 particularly unhappy about. I think he told you why he was
- 25. unhappy. He said that you listed the principal limitations
{. G
._ )
l > i 1 i
- 249 i
1 of the probabilistic approach and its tendency to allow its l 2 results to be dominated by the tails rather that the central 3 tendency of distributions. i 4 I guess that limitation can be assessed by ( i 5 comparing its results with one or more deterministic -- his ' 6 concern was that what you call a limitation'is what he ; L 7 considers a virtue. i 8 MR. MURPHY: [ I understand his point of view. 9' MR. SIESS: The fact that the results from the , 10- tails dominate is not the fault of the probabilistic l 11 analysis; it's the fault of the people that use it. ) 12 MR. MURPHY: Right. t ~. T 13 1 MR. SIESS: You don't reduce the uncertainties by 14 not presenting them. They're there, and they always have 15 been there and unfortunately, I'm afraid they.always will be 16 there. But that's what engineers learn to do, is live with 17 uncertainty and design things.without knowing everything we 18 need to know about them or everything we'd like to know 19 about them. 20 Now, except for not letting you review that, did 21 you finish essentially what you had? 22 MR. MURPHY: Essentially what we had as far as the l 23 technical part was concerned. 24 MR. SIESS: Yes. J 25 MR. MURPHY: We wanted to make a couple of extra 4 4 w .
~ ^ ~
- .L.
, I i l 250 l l 1 points. Specifically, one of them was, we have carried out 2 the four calculations, the four trial plants at this stage. 3 It is our intent to go out for public comment. We'll call l 4 it the comment period to carry out additional calculations 5 for nominally all of the rest of the sites to confirm what 6 we have gotten so far. 7 The other thing is that not in the package that 8 you have reviewed, we are putting together a number of 9 questions that we would like to propose to the public during 10 the comment period. One of those, as we mentioned, was 11 associated with the use of the mean or median in the 85th l 12 percentile. l ! l 13 There were several other comments requesting i j 14 information or comment about the target levels that we are l 15 using, that we're doing that in the right sense and a few l l 16 other items, and potentially any that we develop in, let's ! ! 17 say, the next few weeks in conjunction with our colleagues 1 18 at NRR. l 19 MR. CHOKSHI: I think one of the most important I L 20 questions we are asking is what are the views on the dual l l 21 approach? 22 MR. SIESS: You haven't considered trying to 23 separate this? I am just thinking about the totality of the 24 changes on siting. The source being one part, the seismic l 25 being another part. l
251 % 1 The source term part seemed to be a lot further 2 advanced, firmed up, than what we have here as far as the 3~ proposed r.ule. Do you think so? 4 MR. MURPHY: I have that impression as well. I 5 would hope that by the time that;we come for final on this 6 rule, the firmness of the two proposals will be more than 7 adequate, they'll be excellent. 8 MR. SIESS: I don't-think there is going to be 9 Ltime to come back to the ACRS after this thing has been 10 through CRGR and all of that. That could add a month or two 11 to your schedule. And since you're really not going final, 12 'you really are. going to be publishing a proposed rule -- i 13 when you publish a change with questions, to me that clearly 14 indicates you still are looking for positions on it. That h 15 doesn't bother me. 16 I might mention one little issue that i understand
-17 is kicking around that I might point out. The other people 18 that are working on the source term thing are rather proud .19 of the fact that they were taking stuff out of the Reg Guide 20 and putting it into the rules.
21 MR. MURPHY: Yes, sir. 22 MR. SIESS: You were equally proud of the fact 23 that you were taking stuff out of the rules and putting it 24 into a Reg Guide. I know your reasons, there are things 25 that we wanted to change a long time ago and I think
1. 252 l 1 everybody wanted it changed and it wasn't easy to do, so we 2 just didn't do it. I don't know what their reason is. I am 3 thinking it is good to go the other way. I don't know 4 whether it is going to get called to the Commissioner'.s ! 5 attention or not. And if it is, I would like to have a i 1 6 chance to put in our two cents worth. j l j 7 I might be able to say something in a letter that 8 we think some of these changes -- the direction of the l 9 changes, and why we think they are desirable. I definitely i ! 10 think that the more stuff we get into, the kind of document 11 we can write as a Reg Guide, the better off we are. I 12 Do you know of that concern? l 13 MR. MURPHY: I know of that concern. l l 14 MR. SIESS: It hasn't gone outside of research; 15 has it? 16 MR. MURPHY: I honestly don't know, sir. I know 17 it was kicked around a good bit. I 18 MR. SIESS: I don't think it tus ever brought up 19 as an issue except I happen to know both of them. That is I a l 20 one of the things that ACRS does is see a little more of the l 21 picture. l 22 Well, new what I would like to do now, I think, is 23 close the book on the staff at this stage and see what 24 comments we might want to get from representatives of the 25 industry. a i (
253 1 Now, you had a meeting yesterday with IJUMARC, I l 2 understand? 3 MR. MURPHY: That is correct, we did have a 4 meeting with NUMARC. l 5 MR. SIESS: Would either you or NUMARC like to 6 tell us what you discussed? 7 MR.' MURPHY: At this stage I'll say I wouldn't i 8 presume to tell you what NUMARC's position is. I am not 9 certain whether or not there has been a -- what we had is 10 the same thing that we do with a lot of these things. It is l l 11 the staff -- individuals express their own opinions. I 12 don't believe yet there is a NUMARC position on this. 13 MR. NG: This is Ray Ng of NUMARC. 14 We received a packet some time last week and at I i 15 this time we are currently reviewing the document in some ) 16 detail. We have not been able to canvass our utility 17 numbers to generate a consensus in this reposition at this 18 time. 19 MR. SIESS: You realize from what you heard today 20 that the package you are reviewing may not be the final one, 21 too? l 22 MR. NG: Yes, sir. It makes it a little bit more l 23 challenging in the sense of reviewing and generating a set 24 of comments that we would submit. 25 I think I would say at this time, at least based
254 l 1 upon some of the prior discussions that we have had at 2 NUMARC last year in anticipation of this proposed rule-3 making that we at NUMARC are also very concerned about the l i 4 requirements for doing both deterministic and probabilistic ' l 5 approaches. And we are looking at that much more closely 6 now in the sense of the value certainly that that brings, 7 and.we certainly will look at it not only within the sense 8 of interacting with the staff and trying to come to a mutual 9 understanding with regards to the site, but also we are 10 concerned very much with regards to the entire licensing 11 process, not only gaining the agreement of the staff with 12 regards to the design demands for the site, but also as you l 13 proceed through the hearing processes and the information l 14 that may be generated as a result of the interactions. And 15 one of the major concerns, certainly if there is to be no 16 plant order, is a predictability of the licensing process l 17 and the stability related to the regulations. 18 And certainly, one of the things that we have r 19 experienced at NUMARC and we have certainly had substantial 20 interactions with the Commission. I am involved on the Part 21 52 process and the level design detail has certainly 22 consumed quite a bit of resources time on the part of both j 23 parties, NRC and the industry. Certainly in that endeavor l 24 there was a question about information -- how much 25 information would be necessary for the staff to reach a
l l l 1 1 i 255 1 determination. And certainly there was, I think, a SECY 2 paper that was proposed at one point that suggested that a ( 3 substantial amount of information be submitted. I think 4 what it was referred to, was Tier 3. And we certainly l 5 argued from an industry standpoint that the information that ( l 6 should be provided should be that which is necessary for the 7 staff to reach a conclusion. And with other information 8 that does not have a regulatory significance, there should 9 not be -- would be available on file should it be requested 10 and necessary in the staff's determination. 11 The Commissioner, I think, responded to NASRM I 12 coming back and saying that with regards to each system, 13 structure or component the level of detail will vary as the i 14 safety significance of the component. It would urge, in 15 this endeavor also, with regards to siting, that there be a 16 purpose in the sense of the information that is requested, a 17 very obvious purpose. Otherwise, we feel very strongly that 18 in a licensing process which takes you through the hearing 19 process, that it is very difficult to explain the use of 20 every piece of information that is generated. And, 21 certainly, if a piece of information is not generated, there 22' m;tet be a reason why not. 23 So, from that perspective we certainly have urged , 24 from a reliability and predictable licensing process, that 25 the information that is being requested be thought through i
l l 256 1 very carefully and not just have information for the sake of 2 the knowledge without a purpose. 3 MR. SIESS: Thank you. We got a hold of last 4 month of a paper published at the last Water Reactor Safety 5 Information Meeting on a Basis for Standardized Seismic j 6 Design for Nuclear Power Plants. And the author of that 7 paper, Mr. O'Hara, is here, I believe. Is he still here? 8 MR. O'HARA: I am still around. 9 MR. SIESS: He has an approach which is not the 10 same as what the staff is proposing, but it is similar 11 enough in some respects. I thought you'd find it 12 interesting. He was looking at the probabilistic approach ! i 13 for all of the existing plants, as a basis for comparing. ! 14 Much of what he has is'really related to the .3g standard I 15 plant design item; but he also has something in it about
'16 picking a site-specific earthquake, less than .3g for the 17 site-specific portions of the plant, which I assume is 18 permitted under the certification, in fact, I know it is.
19 Plants are certified for three-tenths, but when -- anything 20 else you build at the site can be designed to the site-21 specific value. 22 So, we've invited Mr. O'Hara to present this to 23 us, if you'd like to hear it. I think you should.- And I'd 24 like to take a short break before he starts. It's 4:00 25 o' clock, so come back in 10 minutes, and we'll let him set i
. . ~ . - - - - - - - . . . ~ - _ . ~ - - - . _ _ - . - . - . - .. -- .
l l l 257 1 up. 2 (Brief recess.] 3 MR. SIESS: Mr. O'Hara, you have the floor. 4 Introduce yourself appropriately. l 5 [ Slide.) 6 MR. O'HARA: My name is Tom O'Hara from the Yankee 7~ Atomic Electric Company. And I'm making some comments on 8' the draft Regulatory Guide DG-1015, Seismic Sources. It was I 9 a copy I received last week. 10 The essence of my comments comes from two papers. 11 One was a 1990 paper on nuclear engineering and design. , 12 We've been following this issue for quite some time. And,
' 13 secondly, it's a paper you've all got. It's a basis for 14 standardized seismic design for nuclear power plants. It's l 15; one that we gave at the 19th Water Reactor Safety Meeting l
L 16 last year. 17 I want to let you know that I've been involved in 18 the deterministic Appendix A process, the classic Appendix A , 19 process, and have also been involved with probabilistic 20 . processes for many years now. I've followed the Lawrence 21 Livermore studies again in 1978 or '79 for the SEP, 22 Systematic Evaluation Program. I think I'm fairly familiar 23 with the Livermore Methodology. And I've also been involved i 24 with the EPRI methodology. I've assisted in developing the 25 methodology, and I've also performed all of the seismic i l I_. , _ _ . -- _ -
1 258 1 hazard calculations for all of the Eastern U.S. sites. 2 (Slide.] 3 MR. O'HARA: The presentation will go as such. 4 We'll give you an introduction on a few points of view that l 5 we have, then an overview of the proposed NRC approach, 6 comments on the deterministic approach, comments on the l 7 probabilistic approach, and some conclusions. And I will 1 l i
- 8 try to speed this along as much as I can, since it's been a 9 long day for everybody here.
10 (Slide.] l 11 MR. O'HARA: Some general comments. First off, we 12 certainly support this revision of Appendix A. We applaud 13 you for some of the things that you've tried to do. We ! l 14 believe it's necessary. We support the use of probabilistic ! 15 and deterministic methods. It's a question of how you apply 16 these things and the emphasis you give where. l 17 One could certainly see where one might consider i 18 giving a heavy degree of emphasis on deterministic methods 19 in the Western U.S., and a heavy degree and heavy emphasis 20 on probabilistic methods in the Eastern U.S. 21 My comments apply primarily to the application of
- 22 the proposed Dg that you've all been talking about today.
23 (Slide.) 24 MR. O'HARA: The basis or my criterion for making 25 decisions as to whether or not I think that this approach is i
l 1
)
l 259 1 acceptable or one that will work, is basically these NPOC 2 goals. And this comes from the strategic plan for building 3 new nuclear power plants, Nuclear Power Oversight Committee. 4 And we have a couple of goals. I would ask the question, 5 will this process that you're proposing assure that 6 regulatory processes are in place for predictab.e licensing, j J 7 including site approval, and two, will they assure a high-8 quality, stable, regulatory process to assure safety and 9 environmental protection? And the two key words are, of 10 course, predictability and a stable process. 11 [ Slide.) 12 MR. O'HARA: My next slide is the proposed NRC 13 approach. It's a parallel pathway approach. I agree with : 14 Ray Ng, if he's still here, that this ought to be an or 15 gate, we should have a choice what we do. This parallel l 16 pathway entails these geological, seismological and j 17 geophysical investigations. We then do our deterministic, 18 classic, good old Appendix A approach, which is identify the 19 seismic sources, determine expected maximum earthquakes, l 20 determine the controlling magnitudes and distances. And it I 21 all sounds so easy. l l 22 The other pathway is this probabilistic approach, I l 33 which I feel familiar with. And out of this probabilistic l 4 24 approach we do a Livermore or EPRI study, we compare their 25 operating plants and somehow or another out of that box, as 1 l
I' i E , l 1 i i 260 1 I. understood it, we came up with acceptable ground motions. I
'2 The ground motion was then used as input to determine the !
Li l 3 mean magnitudes and distances. We then compare -- and I t l 4 don't mean.to make-light of this topic but, in my idea, this ' 5 whole process right now would look something like this.
- 6 [ Slide.)
! 7 MR. O'HARA: I.think we ought to replace this
- 8 compare within a miracle occurs. Because it's pretty fuzzy, i
l 9 in my opinion, at'this time. l 10 (Slide.)
-11 MR. O'HARA: I want to go down the deterministic 12 pathway, because it all relates to the probabilistic 13 approaches.
14 This deterministic pathway, and I'll say it again, 15 is the-good old deterministic pathway, one that we all feel 1 16 very comfortable with, and basically, what you see is you ! 17 would have seismic source zones, ano there is no uncertainty 18 there, and you would deternine the expected maximum 19 earthquakes or whatever terminology you want to use in the 20 future. i ! 21 You'd attenuate them to your site, and you get I l 22 these controlling earthquakes, and basically, you would i 23 envelope these. It sounds very. simple. I L 24 [ Slide.) ! 25 MR. O'HARA: I want to note that no place -- i I
-. --_. _.. =- . . . . . -
l l I 261 l 1 unless things have changed, no place in this draft guide is 2 there a differentiation between probabilistic seismic source 3 zones and deterministic seismic source zones. Correct me if 4 I am wrong on that. { 5 Anyway, in this Reg. Guide, we have a seismogenic 6 source as a portion of the earth which is considered to have 7 uniform seismicity, etcetera. ! 8 I then have below that the definition that's used 9 in the NUREG/CR-5250. 10 This is the reports that Don Bernreuter did which 11 gives the probabilistic results, the Livermore results, and 12 I would contend that both definitions for seismic source
) 13 zones are essentially about the same, and I think this opens 14 up a whole bag of worms for us.
15 (Slide.) 16 MR. O'HARA: The expected maximum earthquake, and 17 I know that it's changed, but I will just use it for the 18 time being, that what I would contend, the prior draft that 19 you had, was that there was no difference between the 20 expected maximum earthquake in the probabilistic approach 21' and the deterministic approach, and this presents -- would 22 have presented significant problems. 23 I understand that you intend to change that to 4 24 some sort of deterministic earthquake -- I forget what you 25 called it -- but what I would ask is that this term should l l \ l
- . _ - _ _ . _ _ _ _ _ _ _ . _ _ ..._m..- . ~ . ... _ _ _ - . . _ _ _ _ ._ _ _
i 262 I 1 be unnabiguous, and I would also challenge you to take a few ,. 2 seismic source zones and test this definition out with a few 3 or'several experts and see if you get back the same value ! l 4 I think this would be a very, very difficult task l l 5 to do. 6 What I want to do at this point is highlight a 7 problem that I see with respect to this whole deterministic 8 process. 9 [ Slide.) 10 MR. O'HARA: And I believe in the deterministic 11 process. It certainly has worked. I gave a presentation 12 back in '86 to you all, and I said that, hey, don't fix it, ' i 13 it ain't broke. I also said I much prefer dea; .;g with the 14 devil that.I know. 15 However, what I would contend is that what exists 16 out there are all of these seismic source zones, and here is 17 a picture of what the Woodward-clyde seismic team has 18 determined to be their seismic source zones for the eastern 19 U.S. 20 [ Slide.] l 21 MR. O'HARA: Let me show you a few more of these ! 22 seismic source zone interpretations, and I would suggest you l 23 focus in on New England or Washington, D.C., and if you find i 24 any similarity between these seismic source zones, let me 25 know, because I don't see any. i 6
.. . . _~ _ - .. - - . -. =
i 263 1 [ Slide.] 2 MR. O'HARA: This one here is for Western 3 Geophysical, and bear in mind, associated with each one of 4 these different interpretations of seismic source zones are, 5 I'll call them, expected maximum earthquakes, and I contend 6 that I don't see any way that you can keep these 7 probabilistic inputs from interacting with the deterministic 8 analysis. 9 [ Slide.] 10 MR. O'HARA: That's just a few. I want to throw l 11 up just a couple more, and these will show some of the 12 Livermore inputs, and what I would note to you is that what 1 i 13 we are talking about here -- if I pick this site 31 in New 14 England, what are we talking about in terms of expected 15 maximum earthquakes? 16 Typically, you have a distribution or.an upper 17 bound magnitude. If I found the expected value, what would 18 I be talking about? And we're talking about 6.5. This 19 would be very tough to handle in a deterministic process. 30 I want to add another point here. 21 When we are talking about these seismic source 22 zones, they aren't developed by graduate students, they 1 23 aren't developed by your -- they are developed by recognized 24 experts. l 25 [ Slide.] l
l 264 I
'l MR. O'HARA: This is from the Lawrence Livermore 2 study, and we see here Dr. Pomeroy. We see here Dr. Carl 3 Stepp. These are recognized experts. l l
4 How are we going to handle this in the licensing 5 process with respect to Appendix A, the simple deterministic 6 Appendix A process, where we define a set of seismic source 7 zones? How are we going to do it? How are we going to 8 handle that? 9 [ Slide.] ' 10 MR. O'HARA: What I would contend, at this point, 11 is that I don't think -- I don't think it's workable, and 12 what I challenge you to do -- you say you're going to do i 13 these four sites and you're going to do the rest of the 14 sites. 15 I challenge you to try it, deterministically, for 16 one site and show us something that will work. Show us. 17 If the NRC wants a deterministic pathway that will 18 not result in endless opportunities for intervention and 19 delay, then I contend that what you must do is the same 20 request that was made back in the mid '70s by the ACRS to 21 the staff, come up with a set of tectonic provinces that are 22 unambiguous and also give us the expected maximum 23- earthquakes, i
- 24 If you don't do that, it's going to be endless '
25 litigation, and the reason is -- the reason I say this is { l l
i l 265 1 because what we have got right now is the current-generation 2 of plants -- I agree they are safe, I accept the seismic ! 3 design bases are okay. 4' All of that was done pre-Livermore and -EPRI, pre-l 5 Livermore and -EPRI. Livermore and EPRI has -- there has 6 been an explosion in terms of seismic source zones, an 7 explosion with the notion of upper bound magnitudes. How do 1 l l 8 we divorce that from the deterministic process? 9 So, therefore, I feel very queasy about any 10 conclusions that we will be able to simply apply -- and I'll ! 11 say it this way -- the good old Appendix A process in light 12 of the information that's out there. 13 [ Slide.] 14 MR. O'HARA: What I want to do is change gears at 15 this time and address the probabilistic approach, and I have 16 highlighted what I consider to be significant problems. 17 MR. SHAO: Jim, can we ask questions? 18 MR. O'HARA: Sure. Go ahead, but not too long. 19- MR. SIESS: If you want a debate, you ask 20 questions that can be answered, Larry. 21 MR. SHAO: Okay. I just had a simple question. 22 If I believe what you say, you say the 23 deterministic approach will not work now. l l 24 MR. O'HARA: I contend -- it's just my point of ; ! 25 view -- that it is unworkable in the eastern U.S. at this 1
.. - .. . _.. . . _ . _~. .- -
l l ! l 266 1 point, at least in terms of -- we will end up with endless 2 litigation. 3 MR. SHAO: Yes, but I thought Andy just said we 4 finish four plants. 5 MR. O'HARA: No, no, no, probabilistically. l l 6 MR. SHAO: Okay. You used the old deterministic. 7 MR. O'HARA: Yes, and I challenge you to do it for 8 one site, one site. 9 MR. SHAO: Because there is always a potential 10 that the deterministic approach didn't -- 11 ! MR. O'HARA: I challenge you, one site, do it ' 12 deterministically.
\ 13 MR. SHAO: Why don't we do that?
14 MR. O'HARA: Make it four sites, then the 15 Northeast, the Southwest Central, Southwest. Very simple. 16 I would appreciate that immensely, and I will be 17 in your corner. 18 MR. SHAO: Let me summarize your comment, if I 19 understand you right. You say in view of the Livermore work 20 and EPRI work, why the deterministic approach will not work 21 because there is always one earthquake bigger. 22 MR. O'HARA: Not bigger. The problem is how do I 23 arrive at appropriate seismic source zones? How do I debate 24 Dr. Pomeroy? How do I debate Dr. Stepp? Why am I -- why is 25 my opinion better than theirs? l
.- .. - - . ~ . . . .. - ~ . . - . . . ~ ~ _ . ~. . - - - .... .
I l I l 267 1 It's very difficult. 2: MR. SHAO: Thank you. 3 MR. O'HARA: You're welcome. The probabilistic 1 4' approach. I want to add another comment. I forgot to add l l 5 it, let me bring it up. I 6 When you talk about these seismogenic source zones
- 7 in your statement -- your statement was that it's 8 homogeneous and the frequency of occurrence was uniform i 9 throughout the seismic source zones.
10 I want to point it out -- and I think you may be 1 11 aware of it -- in the EPRI methodology we allow for the " l 12 seismicity to vary from one degree cell to one degree cell. I 13 I presume, if you still accept these EPRI seismic 14 source zones. Is that correct? 15 MR. MURPHY: We're not d'oing -- in our proposed 16 reg guide, we are doing nothing to change either the EPRI or 17 the Livermore programs. 18 We are not redefining terms. We are not changing 19 database. 20 MR. O'HARA: I understand that. But the 21 definition is, seismic source zones have uniform seismicity. 22 That's the definition in the draft reg guide. l 23 MR. MURPHY: That's the definition in the -- l l 24 thac's the definition in the draft reg guide. Right. l 25 MR. O'HARA: In the EPRI source zones, the f-l l
268 1 seismicity varies within the source zone. 2 MR. MURPHY: We are not claiming or requiring 3 equivalency between the three definitions in effect. The 4 one that would be used for EPRI, the one that we use in 5 Livermore, or the one that would be used in the 6 deterministic process. 7 That's not a requirement at this stage. 8 MR. O'HARA: Okay. I just wanted to bring it up, 9 to make you aware of this situation. 10 MR. MURPHY: We are aware of it. 11 MR. O'HARA: Okay. The probabilistic process. 12 There has been a great deal of debate about this, i 13 And I want to add one other thing with respect to 14 the deterministic process. My conclusions as to whether 15 something is acceptable or not in general is, is it -- can I ' 16 do a calculation? Can you reproduce it? 17 Am I predictable and am I reproducible? 18 And I get nervous about the deterministic pathway 19 in terms of those criteria. 20 The probabilistic pathway -- it is carved in 21 granite. All of the inputs are essentially frozen at this 22 point. 23 All of that information that we talk about -- the 24 geological, seismological, geophysical investigations -- 25 that doesn't go into this seismic hazard analysis other than
i l 1 I i i 269 1 at your site. 2 And we argue -- or I argue -- very strongly that 3 there be detailed site investigations -- deterministically. 4 And when I talk about having probabilistic and l 5 deterministic, that's the blend. Detailed site 6 investigations deterministically. l 7 Then you handle all this uncertainty that exists ! 8 out there using probabilistic approaches. - 9 MR. BAGCHI: That is the strong point we made, ! 10 right, Tom? 11 MR. O'HARA: Yes, sir. When we conduct the EPRI 12 or Livermore seismic hazard calculationt you asked how.long l i i ! ' 13 it. takes? I've done them all. It takes awhile, but it's 14 very straightforward. 15 And the key is, I can get predictability out of 1 L 16 it. And keep in mind one other thought. We've got this -- l 17 we talk about Livermore and EPRI -- there is also this 18 notion of a resolution study. l 19 I'm concerned that the conclusions I arrive at 20 using Livermore and EPRI, which you think are okay right 21 now, I want to believe I would get the same conclus3aos if 22 we get this resolution study. 23 The resolution study has to do with the fact that
'24 there are differences between these methodologies. And the 25 reason we go this relative approach is because of these l
l l _
.-.-.__.._.._m.._____.. _ _ . . _ . _ . . _ . . _ . . _ _ _ _ _ _ . . _ . . _ . _ . _ _ . , _ _ . _ _ _ . .
y r l l 270 1 absolute differences. t i-2 You have a next step here which is compared to 3 operating plants to set probability of exceedance level. ! 4 Really that's not what we do there. I don't , l 5 believe. What you do is you determine -- you do your 1 6 seismic hazard analysis at your site, and then you have > 7 these " acceptable" probabilities. 8~ You go in and you determine the appropriate or I t
- 9 acceptable ground motion. Then given the ground motion, you L
10 do another hazard calculation and you calculate the mean ' i l 11 magnitudes and distances, and you compare those. l 12 That's correct. Okay. l 13 Let me move this guy out for a second. l 14 (Slide.) l 15 MR. O'HARA: Typically, this is what you would see 16 in a seismic hazard analysis. You define seismic sources. l 17 They're already done, so there's no work there. , 18 You define'-- determine seismicity parameters. i I 19 Already done. Determine the expected maximum earthquakes. l l 20 Already done. 21 It's just a matter of pulling the trigger, 22 basically. Although you do input this detailed site 23 information. 24 I want to add a couple of comments here. One of 4 25 them is from one of your former branch chiefs, Leon Reiter, l
271 1 a man I have a great deal of respect for. Leon wrote a 2 book, which I recommend everybody should read. I'm plugging 3 his book. It's Earthoucke Hazard Analysis: Issues and 4 Insichts (1990). 5 And what Leon says is that these probabilistic -- 6 and he went through the Appendix A process. He has been 7 through Livermore. He's been through EPRI. And I think 8 he's -- I'm just sorry he isn't here to help us with this 9 whole process. 10 He says these probabilistic methods allow the l 11 incorporation of uncertainty and frequency of occurrence. 12 You can accurately reflect the true state of knowledge
- 13 thereof.
14 He also says it's capable of absorbing and !
- 15 integrating a wide range of information and judgement and l
)
16 t'.leir associated uncertainties into a flexible framework. j l 17 I couldn't agree more. 18 What I really think Leon meant to say was that 19 these probabilistic methods, what they offer is a rational l ! 20 framework to incorporate this abundance of diversity that l 21 exists our there, in terms of the seismic source zone, l 22 seismicity parameters, magnitudes, attenuation models, and ( 23 we can go on. 24 [ Slide.] 23 MR. O'HARA: That's why we are using probabilistic f
1 l l l l I 272 1 methods, because we need some way, some framework to l 2 incorporate all this stuff. 3 This, I thought, was by far and away the most 4 ! outstanding statement I have ever read in a regulatory l 5 guideline. I think this took real courage to put this thing 6 down on paper. Has it changed? Is it still the same? 7 MR. MURPHY: Yes. 8 MR. O'HARA: I applaud you for this. You have 9 come up with a quantitative standard that we can compare i 10 against to decide whether or not a site-designed basis is 11 acceptable or not. I think this is incredible. I think 12 it's outstanding. l 13 I do have some minor concerns here, and one of 14 them is, and you've already started to change, you're saying 15 well, maybe we shouldn't use the median, maybe we should use l 16 the 25th percentile, or maybe we should use the 75th. This 17 gets me concerned. 18 The problem with using comething like the median 19 probability is that what is your basis, other than it's the
- 20 middle of the plant, something to compare with. And all 21 I've argued for in this other paper that I wrote was that 22 you should use some deterministic standard, a common 23 standard, between Livermore and EPRI, such that I get 24 consistent conclusions between the methodologies.
25 And what I did was, I said let's assume that a .3 i i
I l \ l j l 1 l 273 1 Reg. Guide 1.60 existed at all current sites, as if there 1
}
2 ! was no design basis there. And I ran the numbers, and I i 3 found that, using your criterion, a .3g Reg. Guide 1.60 was l I 4 acceptable at all sites. 5 I l All I'm saying to you is, if you want to define -- 6 and the reason we go to this statement here, the median l 7 probabilities of exceeding the current population of plants, 1 8 the reason you do that is to define acceptable 9 probabilities. You now call them target probabilities. 10 I would prefer to use some objective, consistent 11 standard, such that when we go to a resolution methodology, I i l 12 I don't find out that all of a sudden .3g at existing sites i
' 13 is no longer acceptable. t
! 14 [ slide.) i 15 MR. O'HARA: Has this slide changed? This has ! 16 changed, hasn't it? In other words, what you had before was 17 that this table -- 18 MR. CHOKSHI: May I ask a question? 19 l If this worked for Livermore and EPRI hazards, .3g L 20 at 1.60, then it will not work for .3g if you are coming out 21 with new hazard estimates higher than Livermore? Am I ! 22 understanding right? The only reason it will not work for l 23 .3g Reg. Guide 1.60 if your operating hazards are higher
, 24 than either of these.
4 25 MR. O'HARA: Let's just say I went to a site like ! i l l l {
l l l l l 274 1- Charleston, and I calculated the hazard. l I will not be able l 2 to site a .3g plant there. l l l 3 MR. CHOKSHI: No, but the comparison in this table l 4 -- j 5 MR. O'HARA: Oh,'in the table. Oh, okay. l 6 MR. CHOKSHI: -- for the existing site, l 1 7 MR. O'HARA: That's correct, yes. 8 MR. CHOKSHI: If I repeat that process for the l 9 existing site with-the new hazard curve, unless my hazard l- 1 l 10 curve is really above Livermore -- i 11 MR. O'HARA: When you say a new hazard curve, what 12 do you mean? Do you mean the resolution study? l i 13 MR. CHOKSHI: Yes. Otherwise, you are guaranteed l 14 that you will meet for the existing site that requirement. 15 MR. O'HARA: That's right. That's what concerns - l 16 me, is we need to stabilize this. Something must be i i 17 acceptable. And all I've said is, to calibrate these l 18- methodologies, let's assume .3g at existing sites and take ! 19 the envelope probability of exceeding these design bases. , 20 And that is what is the essence of the paper. l 21 Go ahead, Bob. 22 MR. ROTHMAN: Are you surprised that .3g at l 23 existing sites was accessible since all of the sites are 24 acceptable and the highest one is .25g? ' 25 MR. O'HARA: Excuse me, Bob. I would ask you one l r i
_ _ . . . . . . - . - - - . ._ . . _ . _ . _ . _ _ . ~ _ . . . . _ _ _ . _ . . _ _ _ _ . _ _ . _ _ _ _ _ . . . . _ _ _ . . 275 L 1 question. l 2 Do you believe that .39 at existing sites is 3 acceptable?
- 4 MR. ROTHMAN
- Yes.
5 MR. O'HARA: All right. 6 MR , ROTHMAN: But, I mean you're asking like it l 7 was a big surprise. 8 MR. O'HARA: Well, you know what? I ask that 9 sometimes of our industry people, and they say well, well, 10 you know, you know, you know, we're not sure about that. l 11 Thank you very much. I appreciate that. We'd like to nail 12 it down. l 13 I'm not going to go into some of the -- you know, i 14 we're very close. And this is a nice feeling. I'm not L 15 going to go into some of the reasons why .3g.is acceptable. I i 16 I want to give you some conclusions. But what I also want l 17 to do is'-- what this plan was earlier was to have John 18 Sutton over here was going to give you some of our 19 philosophical points of view. 20 MR. SIESS: Excuse me. l 21 MR. O'HARA: Okay.. Go ahead. l 22 MR. SIESS: It's just dawning on me what was said 23 about .3g at existing sites. I l 24 I thought that at the December meeting I asked the
~
I 25 staff if somebody applied to put a certified design at l l
l l 276 1- Millstone, would they accept it? And they said oh, no, l 2 they'd have to start off and rctiow the site. l 3 Am I wrong? Does anyPody remember that? I asked l i 4 somebody that. 5 MR. ROTHMAN: Someone from NRR? 6 MR. SIESS: I don't know. 7 [ Laughter.] 8 MR. SIESS: I'm going to make you. guys wear 9 different colored hats, or something. 10 It was somebody from NRC, because I, you know, 11 flatout said somebody wants to put a certified plant at one 12 of the existing sites. I might have picked Braidwood. I
\ 13 don't know. I thought I'd picked Millstone because there l
14 are three different designs there now, I think, for three 1 15 different g-values. I 16 MR. BAGCHI: It should be acceptable. l 17 MR. SIESS: No problem. { j 18 MR. BAGCHI: It should be acceptable. 19 MR. SIESS: Okay. Thank you. 20 MR. MURPHY: It should be acceptable, but I'm 21 certain that the staff would require some documentation on l 22 that. 23 MR. SIESS: You send them three FSARS. I 24 [ Slide.) 25 MR.~O'HARA: John Sutton was going to discuss, i l l {
J i i i i i f l 277 J 1 basically, our philosophy. l We're going to'get back to you l l
\
l 2 with detailed comments on this whole Reg Guide, what you're 3 proposing here, and -- but I just want to point them out to i 4 you, that what we're saying is -- and you are trying to do 5 it -- is the content of the rule should be significantly 6 simplified. t 7 The model for simplification -- and we went to -- 8 .I don't want to say extremes, because I believe what we did 9 in our paper, but the model for simplification should be 10 current treatment of hydrology and meteorology -- you've 11 heard this from me before -- and details of how one 12 demonstrates compliance with GDC 2 should be addressed only 13 in regulatory guidance documents. l 14 That's a philosophy. 15 [ Slide.] 16 MR. O'HARA: What I.want to do is just give you a 17 couple of conclusions and show you what I would do, if you 18 ant to simplify this process and ensure predictable and 19 stable regulations. 20 First, my conclusions: I propose that the NRC -- 21 I contend this NRC deterministic analysis will prove to be l 22 unworkable in the eastern U.S. It's j ust my point of view, 23 Bob, and it's based because EPRI and Livermore are on the 24 table now. The times have changed. 25 Acceptable site probability should be defined
- . . _ . _ . _ . . . _ _ _ . . _ . - _ . _ _ _ _ _ . . _ - ~ _ _ - . . __. _ __ ..
278 1- using a common deterministic standard. You don't calibrate 2-a machir.a relative to some numbers within itself; you have a 3 standard. 4 Propose tha NRO probabilistic analysis will work.
,5 I am convinced that the proposed NRC approach, using 4 6
Livermore, using EPRI, will work, and it,will simplify this
- 7 licensing process.
8 Evaluation to determine mean magnitude and 9 aistance, I see no need for it. I.see no reason why one 10 would derive any comfort in the fact that you go through 11 this whole process, you come up with two values and they I 12 compare favorably. 13 Why should that give you an comfort? They are two l 14 different animals. One is probabilistic, one is 15 deterministic. You should derive no comfort from that, t 16 (Slide.) 17 MR. O'HARA: What I would contend is that we do 18 get rid of all this and tPis, and out of this step right 19 here comes your SSE with detailed site investigations. I am
-20 convinced it can work, and we are calibrating this relative l
21 t'o existing sites. It's conservative. 22 And that, in essence, is my presentation, and I 23 thank you for giving us the opportunity to bring this up.
- 24 MR. LINDBLAD
- That refers to eastern United 25 States, a
t l I
.- . . , . - _ . - . - . - ~ . - - . . - ~ . . - . - . --. . . - , . ._. _ . . . - - . - .
279 1 MR ~. O'HARA: That is correct. I am focused l-2 primarily on the eastern U.S., and I just am not convinced l 3 that you can do this deterministic process in the eastern 4 U.S., and I'd appreciate it if you'd do a couple of l 5 examples. That would make me feel better. 6 MR. SIESS: You're not convinced they can do the 7 deterministic. t 8 MR. O'HARA: I am not convinced you can do it, and ; 9 I have been through the Seabrook deterministic process. j 10 - MR. SIESS: Oh, you mean and get away with it. l 11 MR. O'HARA: Get it done. 12 MR. SIESS: I mean it's been done. 13 MR. O'HARA: No, no, no. That was all pre-14 Livermore and -EPRI. 15 MR. SIESS: Yes. Okay. But I mean the process 16 has been done. You mean get it accepted and through a l 17 hearing. 18 MR. O'HARA: Yes. Yes. : 19 MR. SIESS: Yes. Okay.
.20 MR. O'HARA: Exactly.
21 MR. SIESS: I see your point. 22 MR. MURPHY: And the real crux of what you're 23 saying is to get it through a hearing. Someone or some l 24 company can do the process, and then what you're saying is , 25 the actual process of getting that through a hearing board 1 I w ,- -- y - r7
i l 280 1 and intervenors will not work. 2 MR. O'HARA: I'think it will be unworkable, and 3 that's why I challenge you to try it and do it, l i 4 MR. SHAO: But that's why we asked for both. 5 MR. O'HARA: I say give us the option, and I know 6 which way I'm going. I say try it. Let's just see. 7 Thank you very much for letting us give this 8 presentation. 9 MR. SIESS: Thank you.
'10 Does the staff get copies of our official minutes 11 of these meetings? Do yet ever look at them to see if you 12 were quoted?
13 Because on page 3 of the minutes for the December i j 14 loth meeting is a. sentence that says, "In response to a i - 15 question regarding the suitability of building another plant L 16- on an existing site, they indicated that automatic approval 17 would not be given but that the original information
- 18 generated for the site would have to be reviewed."
l ! 19 Now, that's not quite the same point of where I t 20 specified a certified design, but at least my memory is not 21 failing me. Maybe yours is. I don't know who answered 22 that. 23 MR. MURPHY: Probably me. I
- 24 MR. ROTHMAN
- I wouldn't argue with that 25 statement. I think we'd have to see if there was any new l
l
281 i 1 information that became available since the last SER was l 2 reviewed. 3- MR. SIESS: Well, you know, this bothers me, l 4 because if there is new information that became available, 5 that plant on that existing site needs to be looked at. 1 Why 6 would you wait until somebody applied to put another plant t l 7 there before you would look for new information? i 8 MR. ROTHMAN: Because I don't know about 9 overything that's become available unless it's brought to 1 10 our attention. [ 11 MR. SIESS: So, an existing plant could be in 12 jeopardy simply because nobody else has applied to put l 13 something else at that site, huh? That doesn't make a lot l 14 of sense. It doesn't speak well for the vigilance of the j 15 staff. 16 Isn't there something in the license that says -- 17 you know, we tell the licensee -- frequently we have a 18 license condition that says he's got to keep up to date on 19 population changes, he's got to keep up to date on man-made 20 hazards like how many shipments go by, and transportation, ' 21 somebody puts a pipeline in there close, so he has to keep t l 22 up to date. l l 23 Is there anything in the license that says he has 24 'o -- t 25 MR. BAGCHI: Oh, you're absolutely right, Dr. I l l l
282 1 Siess. We're doing that with Trojan site. We're asking the 2 licensee to develop the information from the subduction zone 3 earthquake potential. 4 MR. SIESS: I mean if you haven't heard about it 5 or Bob hasn't heard about it, is there something in the 6 license that says the licensee should keep aware of what's 7 going on geologically? 8 MR. BAGCHI: It's not automatically triggered, but 9 something that's going to be so significant that the 10 Millstone site will have to be designed for 0.3g, we would 11 know about it. 12 MR. SIESS: It seems to me that Salem has to keep 13 track of the barge traffic on the Chesapeake Bay to be sure 14 that it doesn't get worse that it was when they got their 15 license. 16 Is that all? Anybody? 17 [No response.] 18 MR. SIESS: I've been thinking a little bit about 19 full Committee, and I think the full Committee should get 20 this thing. 21 They should get a fairly complete overview, and 22 what I have done is look through the handouts from the 23 presentation in December, which, of course, was the broader 24 picture, and if you will take a few minutes after we 25 adjourn, if you've still got time, I'll show you from that
,_ _ . _ . _ _ . _ . _ . . . _ . . . . _ . . _ _ . _ . ~ . _ . _ - . _ . _ . - - . . - _ _ _ . . _ _ _ . _ - . _ - _ _ . _
t I l 283 1 and-eliminate.a few things that I don't-think we need to' 2 talk about,.and then, I'm trying to think whattwe want to L 3 say about DG-1015, since it's still in the process of being 4 revised, but the general approach of the dual method I think 5 we have to discuss at some level, and we'll talk a little-6 bit about that. 7 Although we've spent all of this meeting on draft i D 8 guide 1015, there was a lot more in the past, and Mr. O'Hara l ! 1 l 9 has commented on a few other things that' relate to that l 10 overall picture that we can expect to hear from the industry 11 on. l 12 Unless members or consultants have any questions, l l l
- 13 I'm going to adjourn the meeting and sit down with the staff I i i 14 and run through- some of your previous handouts and suggest - i l 15 - it's not like you've got a month to redo everything.
16 The meeting is adjourned. 17 [Whereupon, at 4:48 p.m., the meeting was , 18 adjourned.) i 19 l 20 i l 21 22 l 23 l 24 25 f
1 l 1 l
- REPORTER'S CERTIFICATE l
This is to certify that the attached proceed-ings before the United States Nuclear i Regulatory Commission j in the matter of: NAME OF PROCEEDING: Extreme External Phenomena l DOCKET NUMBER: , PLACE OF PROCEEDING: Bethesda, Maryland were held as herein appears, and that this it-L the original transcript thereof for the file of j the United States Nuclear Regulatory Commission taken by me and thereafter reduced to typewriting by me or under the direction of the court report-l ing company, and that the transcript.is a true and accurate record of the foregoing proceedings. a 1 J ' ( Official Reporter , l Ann Riley & Associates, L:d. l ? l l l-i i \
l l l l AN UPDATE ON THE REGULATIONS AND REGULATORY GUIDES
- ASSOCIATED WITH THE REVISION OF l l 10 CFR PART 100, APPENDIX A
l l l l i PRESENTED TO THE I ADVISORY COMMITTEE ON REACTOR SAFEGUARDS EXTREME EXTERNAL PHENOMENA SUBCOMMITTEE 3 FEBRUARY 5,1992 ; l l l l ROGER M. KENNEALLY U.S. NUCLEAR REGULATORY COMMISSION DIVISION OF ENGINEERING OFFICE OF NUCLEAR REGULATORY RESEARCH (301) 492-3893 i
\
i
i { ON DECEMBER 10,1991 THE SUBCOMMITTEE PROVIDED l DETAILED COMMENTS TO THE STAFF ON THE FOLLOWING i DOCUMENTS:
- REGULATIONS APPENDIX B, CP.!TERIA FOR THE SEISMIC AND GEOLOGIC
] SITING OF NUCLEAR POWER PLANTS AFTER [ EFFECTIVE DATE { OF THE REGULATION]," TO 10 CFR PART 100 i i l APPENDIX S, " EARTHQUAKE ENGINEERING CRITERIA FOR i NUCLEAR POWER PLANTS," TO 10 CFR PART 50 i i I j REGULATORY GUIDES DRAFT REGULATORY GUIDE DG-1016, SECOND PROPOSED REVISION 2 TO REGULATORY GUIDE 1.12," NUCLEAR POWER 2 PLANT INSTRUMENTATION FOR EARTHQUAKES"
- DRAFT REGULATORY GUIDE DG-1017, " PRE-EARTHQUAKE j PLANNING AND IMMEDIATE NUCLEAR POWER PLANT t
OPERATOR POST-EARTHQUAKE ACTIONS" i ~ DRAFT REGULATORY GUIDE DG-1018, " RESTART OF A NUCLEAR POWER PLANT SHUT DOWN BY A SEISMIC EVENT" i 4
l THESE DOCUMENTS HAVE BEEN REVISED TO INCORPORATE COMMENTS RECEIVED FROM: l l OFFICE OF NUCLEAR REACTOR REGULATION (NRR) l OFFICE OF THE GENERAL COUNSEL (OGC) l ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS) l l l I l l l l i
.. . . . . - - .. . -. - . . . - . - ~ . . . - . - _-..- - . -.. . . - . - - . . - -
SilMMARY OF CHANGES TO THE REGULATIONS NO TECHNICAL CHANGES SUBSTANTIVE EDITORIAL CHANGES REMOVING THE "ZERO PERIOD ACCELERATION" DEFINITION, l REPLACED THE PHRASE IN THE TEXT WITH " PEAK GROUND l ACCELERATION" (UNDEFINED) l l ADDED A DEFINITION FOR " STRUCTURES, SYSTEMS, AND COMPONENTS REQUIRED TO WITHSTAND THE EFFECTS OF l THE SAFE SHUTDOWN EARTHQUAKE GROUND MOTION AND SURFACE DEFORMATION," REMOVED THE REPETITION FROM THE TEXT. ADDED THE PHRASE " GROUND MOTION" AFTER " OPERATING BASIS EARTHQUAKE" l l COMMENTS FROM THE TECHNICAL EDITOR HAVE NOT BEEN i INCORPORATED l I i l
4 i i
SUMMARY
OF CHANGES TO THE REGULATORY GUIDES TECHNICAL CHANGES i REMOVED THE EXCEPTION TO THE DEFINITION OF " FELT
- EARTHQUAKE" IN EPRI NP-6695 '
l NOW: INSTRUMENT ACTIVATION OR CONSENSUS i OF CONTROL ROOM OPERATORS ! WAS: INSTRUMENT ACTIVATION AND CONSENSUS q OF CONTROL ROOM OPERATORS 4 i l AFTER AN EARTHQUAKE, THE RESPONSE SPECTRUM AND i CAV SHOULD BE CALCULATED USING A CALIBRATION STANDARD TO DEMONSTRATE THAT THE FREE-FIELD l INSTRUMENTATION FUNCTIONED PROPERLY l EDITORIAL CHANGES i l REMOVED THE COMMENTARY OR DISCUSSION FROM THE
" REGULATORY POSITION" SECTION i
a l' REVISED THE " DISCUSSION" SECTION TO PROVIDE A COMMENTARY TO THE REGULATORY POSITION 1 i i I COMMENTS FROM THE TECHNICAL EDITOR HAVE NOT BEEN
. INCORPORATED
i APPENDIX A REVISION DRAFT REGULATORY GUIDE ON SEISMIC SOURCES [ PRESENTED TO THE ' s EXTREME EXTERNAL PHENOMENA SUBCOMMITTEE i ADVISORY COMMITTEE ON REACTOR SAFETY ! t,
,p* "*% i
**,e i
February 5,1992 Andrew J. Murphy, Chief , Structural and Seismic Engineering Branch Division of Engineering Office of Nuclear Regulatory Research ! l 301-492-3860 _ _ _ - _ .-- . - _ _ _ - _ _ _ _ _ _ _ . .__________________-__I
i I OUTLINE OF PRESENTATION e OUTLINE OF REQUIRED DUAL ANALYSIS TYPICAL STEPS OF PROBABILISTIC SEISMIC HAZARD ANALYSIS STEPS TO DETERMINE CEs FOR SEMI-PROBABILISTIC ANALYSIS
- INITIAL RESULTS
* \
1 COMPARISON FOR FOUR TRIAL SITES
- OUTSTANDING CONFIRMATORY STUDIES
- RATIONALE FOR REQUIREMENT OF DUAL ANALYSIS 2 l
i SITE DETERMINISTIC , PROBABILISTIC ANALYSIS ANALYSIS l T Geological, Seismological and GeophysicalInvestigations [ l I Conduct an EPRI or LLNL Identify Seismic Seismic Hazard Assessment Sources i Compare to Operating Determine Expected Maximum Plants to Set Probability Earthquake for Each Source i of Exceedance Level ' Determine Controlling Determine Controlhng i Earthquakes Ms & Ds Earthquakes Ms & Ds ! I , Compare Compute SSE Ground Motion 3
. _ . _ ._ _m . ~ . _ - . _ . _ . _ _ . _ . _ _ . _ _ _ . . . . _ . . . _ _ _ _ _ , _ . - . , _ _ _ _
l l l kisate Souner 2 Y kasaic som:r 3 0 8PP sart
$(15ast somer 3
$(1$MIC SousCE 4 I
I l sgrst YNEss4 N ,... . * * . . - 3 / ,D\
- i e
/ i-3 %
/ ,: *
/ e 'f
- d's N Kl9dC ~
7 /- 4
/ t < \
s g
'k , ug, \uh, \
h l l i 0.01 0.1 i 1 to Period (sec) l SCHEMATIC DEFINITION OF CONTROLLING EARTHQUAKES
l I l l I LO Log N h B a - 1 l ! 1 A I O Site -- l b, Slope 1
-1 I
I 1 1 l l
' =M 5 0 "I i
a) Geometry of homogenous source (b) Magnitude recurrence model zones (seismicity / tectonics) (frequency vs size) Log P h Log Y d . l 10~1 - Ground motion i intensity a, Dispersion 10-2 - By integration > *b
- 7 10-3 - m b =6 l mb=5 d I T0 -
i I I = WR 0.1 g 0.1 g Log Y Log distance l d) Seismic hazard curve (c) Ground motion prediction model (attenuation) FOUR STEPS INVOLVED IN A PROBABILISTIC SEISMIC HAZARD ANALYSIS ( l i
PRINCIPAL PRODUCTS OF PROBABILISTIC SEISMIC HAZARD ANALYSIS i A 10 2_ w t o z 4 10 3 o m m o X 10 4-w ' u. O f 10.5 i o oc Q. J 10 6 3 z 2 q 10 7, l
> t PEAK GROUND ACCELERATION, cm/sec 2 ,
Typical seismic hazar d curve. l
t i
~
PRINCIPAL PRODUCTS OF PROBABILISTIC SEISMIC HAZARD ANALYSIS ' SO-TH PCRCENTILE SPECTRA FOR ALL RCIURN PERIOOS 3 10 2
- to S
10 - C - - - g -
^
d p 10 RETURN P[RIODS : CURvt 5 = 10000. YEAR $ CURv[ 4 = 5000. YEARS CURVE 3 = 2000. YEARS ' CURvt 2 a 1000. YCARS
.g CURyt 1 = 500. TEARS '
- 'O ,
n - < on u n .neo ~ n . .n ao '
"3 Te PtRioo (stc) 2 ~2 L
SEABROOK Comparison of the 50th percentile CPtMS for return periods of ' 500, 1000, 2000, 5000 and 10000 years for the Seabrook site. t f 7 ,
P R O o32>a<Hgm
- e O nFcgDpg c B A .
O . . - - . . 1 B I
. I 2 3 4 5 L
~ :
~~ . . .. .
y:- 6 7 a 9 c H I A T A _ - _ N Z Y N ~' . A O U n' R F _ D E A L - ' E X _ P o S C _ R r T E I d O , ' e . M E B A r e A D I B d T N _ E G u . I u L s S I T n ' i . F E D _ n Y n g . O S _ R G
. I _
O m e F d E N E , " I
. . A B .
X S T A l C u' g o E S E . R S I E D n' h . N - I a U _ N - n' z . U S G . .
.S N I _
S S , ' S G I E TM _
~ - : : - : . _ -
. : 7: - - - : _ : 5- - : ..
E E . S D I . A N L L N L . 8 _
' j ', * ' j '
TARGET PROBABILITY OF EXCEEDANCE. Ps , (FOR SPECTRAL RESPONSE AVERAGE OF 5 AND 10 HZ.) METHOD PROBABILITY OF EXCEEDANCE FOR MEDIAN HAZARD ESTIMATES Ps LLNL, SGX 1 E-4 EPRI 3E-5 LLNL, 4GX 3E-5 i
- I l
t I I 9 ! f'
t b g 10 2_ o [ z ' f O m 10_3 , i o X ; w !
- u. 10 4 :
O O ' [ 10 5 _ a D 2 2 10 6
' 8 10 7-4 E
o m 4 5 a a 6 L a s SPECTRAL VELOCITY, Sy , cm/sec. PROB. OF EXCEEDANCE VS. SPECTRAL VELOCITY (AVG. OF 5 AND 10 HZ) 10 ,
e DEAGGREGATE THE PROBABILISTIC SEISMIC HAZARD CURVES BY MAGNITUDE AND DISTANCE GRIDS, i.e., DEVELOP THE CONTRIBUTIO'N TO THE SEISMIC HAZARD FOR THE BINS INDICATED BELOW. Magnitudes and Distance Bins Used in Example Distance Magnitude Range of Bin Range of Bin (km) 5 - 5.5 5.5-6 6 - 6.5 6.5 - 7 7 - 7.5 > 7.5 0-25 25-50 t 50-100 ' 100-150 I 150-200 ' t
> 200 g l 4
11
i i COMPUTE SEISMIC HAZARD CURVES FOR THE INDIVIDUAL BINS.
- DETERMINE THE RELATIVE CONTRIBUTION, THE PROBABILITY OF EXCEEDANCE, THAT A SPECIFIC BIN MAKES TO THE CONTROLLING t EARTHQUAKES. THIS PROBABILITY IS OBTAIN AT THE GROUND MOTION PREVIOUSLY DETERMINED. I i
l h :
- \
~
\
\h
\\ N j g; < Bin 1 p
j Bin 2 2 Bin 3 n. Bin 4 t I
& i Ground Motion !
t i 12 ; f i
DEAGGREGATED RESULTS Hu Values for All Bins Based on the Median Hazard (Note: If H a s 1.E-10, it is listed as O) i Distance Magnitude Range of Bin i Range Bin 5 - 5.5 5.5 - 6 6 - 6.5 6.5 - 7 7 -7.5 > 7.5 0-25 2.0E-5 1.1 E-5 2.4E-6 O O O 25-50 6.2E-6 8.9E-6 5.OE-6 6.5E-9 0 0 50-100 6.0E-7 2.3E-6 6.8E-6 8.4E-7 0 0 100-150 1.6E-9 1.6E-7 1.5E-6 2.8E-6 0 0 150-200 0 1.1 E-9 2.1 E-8 4.6E-7 0 0
> 200 O O O 6.OE-9 0 0
. 13
6 t e USE THESE RELATIVE CONTRIBUTIONS, PROBABILITIES, TO WElGHT THE COMPUTATION OF THE MAGNITUDE AND THE DISTANCE FOR THE CONTROLLING EARTHQUAKES. ! f l M = I I mHma/Z Z H mo md md i D = Z Z dHmo/Z Z H mo , md md m = the midpoint of the magnitude bin d = the centroid of the distance ring .
i i i COMPARISON OF CONTROLLING EARTHQUAKES WITH DESIGN EARTHQUAKE RESULTS FOR VOGTLE SITE ! i V SITE CONTROLLING EARTHQUAKES FROM i PROBABILISTIC ANALYSIS DESIGN EARTHQUAKE 5 - EXPTS. 4 - EXPTS. i M D M D M D i i VOGTLE 5.8 32 5.6 27 5.3 15 l f I I r 15
s INITIAL RESULTS FOR FOUR TRIAL SITES CONTROLLING EARTHQUAKES FROM ' SITE PROBABILISTIC ANALYSIS DESIGN EARTHQUAKE . 5 - EXPTS. 4 - EXPTS. t t M D M D M D VOGTLE 5.8 32 5.6 27 5.3 15 SEABROOK 5.7 19 5.7 18 6 15 E PERRY 5.5 24 5.5 21 5.3 15 l BELLEFONTE 5.7 20 5.6 19 5.8 15 16
i i COMPARISON OF FOUR TRIAL SITES j: t
\
D = THE ACTUAL DESIGN SSE SPECTRUM ! R = REFERENCE SPECTRUM - R.G.1.60 SPECTRUM ANCHORED AT 0.3 g l C = GROUND MOTION SPECTRUM FOR CE USING LLNL SGX MODEL FOR ATTENUATION i S = GROUND MOTION SPECTRUM FOR MAGNITUDE AN7 ! DISTANCE OF ACTUAL DESIGN SSE USING LLNL SGX MODEL FOR ATTENUATION I e r I 17
l l E+4 . , , , , . , , , 6- - l 4 - 3- - l t - t l l l E+3 - i l . 5 -
. l 4 -
" C' 3 - -
. 2 -
CE ' u E+2 - 3 o 5- - 1 4 - 3 - 2 - i E+1- ce n' ~~ o n' n' I E $ $ $ Irequency~ cps , COMPARISON OF GROUND MOTIONS FOR VOGTLE SITE , 18
~ ~
E+4 . . . . , . . . . i . . . . 4 - 3- , 2-E+3 - - 5- - 4 - . 7
~
3 - U - 2- - C 3 E+2 - -
%O o 5- -
4 3- - 2- - n e' .' E + 1_ ~ n o n' n' -- ~ n' n- c, E $ $ $ frequency-cps COMPARISON OF GROUND MOTIONS FOR PERRY SITE 19 l
- . - - - - - . - - - - - . - - . - - - - - - - - - - - . , s ,,
_- - - - - - - . - - - . - - . - - - - - - - - - - - - - - - - - - - - - - - - - - . - - - - - - - - - - . - - -. - - - - - - - - - - - - - - - - . . - - - - , - - - - _ _ _ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - . - - - - - - - - - - - - - - - - - ---.-,,--__,---------------------.-.----.--.-.c-- - - - - E+4 . . . . , , . , 5- . 4 . 2- . E+3 - e ?c e s -
- r:
2 C 3 3 . S E+2 -
!o O $- .,
2- - t E EEE E EOE a : : : I frequency ~ cps f f COMPARISON OF GROUND MOTIONS FOR SEABROOK SITE 20 :
~
E+4 . . ., , . . . . , , , ,, 5- - 4 - 3 - - 2- - E+3 - - C e 5- \ p _ 4 - u _ 3-g
- 2 -
7o E+2 - U o 5 - 4 - 3- . 2- _ E + 1_, A A;A ' ; ; ;; I E $ $ $ Irequency~ cps COMPARISON OF GROUND MOTIONS FOR BELLEFONTE SITE 21
E+4 . . .. , . . .i g . . .. 4 . - 3 - 2 . - E+3 - - 5 N-
? 3 -
R - 2 - 4 I 5 E+2 - - 3 l o 5 - g . . 3 - 2 - - E+1 , 4 A 4i 1 A A ii 1 A A ?I m i i . + l ena M M M 3 frequency ~ cps COMPARISON OF GROUND MOTIONS USING RG 1.60 SHAPE (PERRY SITE) 22
i i CURRENT STATUS OF SEMI-PROBABILISTIC/ DETERMINISTIC APPROACH
.l o GENERAL CONSENSUS THNT THE PROBABILISTIC METHODS ARE NEEDED, ALSO GENERAL CONSENSUS THAT THE CAllBRATION WITH THE PAST PRACTICES IS A SUITABLE APPROACH. HOWEVER, SOME ON THE EXPERT PANEL WOULD PREFER j MORE PROBABILISTIC WElGHTED APPROACH, WHILE OTHERS PREFER DUAL APPROACH AS SUGGESTED. SOME MEMBERS OF THE STAFF WOULD PREFER MORE DETERMINISTIC WElGHTED APPROACH.
o SEVERAL ADDITIONAL STUDIES NEED TO BE COMPLETED PRIOR TO DEVELOPMENT OF THE FINAL CRITERIA AND PROCEDURES: HOWEVER, INPUT ON CONCEPTS NECESSARY AT THIS TIME. SPECIFIC QUESTIONS WILL BE INCLUDED FOR THE i PUBLIC COMMENTS. l o GROUND MOTIONS TO BE ESTIMATED USING DETERMINISTIC APPROACH TO ASSURE !' THAT IMPORTANT ENGINEERING CHARACTERISTICS ARE MAINTAINED. o LIMITED STUDIES INDICATE THAT THE APPROACH IS WORKABLE AND WOULD LEAD ' TO ALLEVIATING SOME PAST DIFFICULTIES. t i i 23 j i
l i-OUTSTANDING CONFIRMATORY STUDIES - j i o PERFORM STUDIES FOR SIMPLE TEST CASES (i.E., ONLY ONE SOURCE ZONE ! SURROUNDING THE SITE OR A FEW SOURCE ZONES) TO BENCHMARK THE. . [ PROCEDURE. t o TARGET PROBABILITY OF EXCEEDANCE COMPARABLE TO NEWER PLANTS. , o SCENARIO EARTHQUAKES AT DIFFERENT SI'ECTRAL FREQUENCIES. - o PERFORM STUDIES USING EPRI HAZARD ANALYSIS. i o PERFORM STUDIES FOR ALL EASTERN SITES (5GX AND 4GX) AND COMPARE WITH ACTUAL DESIGN EARTHQUAKES. l o PERFORM STUDIES FOR FEW CASES USING MORE RECENT GROUND MOTION MODELS. ! l f i 24 r
RATIONALE FOR REQUIRING DUAL APPROACH THE STAFF PROPOSES TO USE BOTH THE DETERMINISTIC AND THE PROBABILISTIC APPROACHES TOGETHER, AND TO COMPARE THE RESULTS OF EACH TO PROVIDE INSIGHTS UNAVAILABLE IF EITHER WERE USED ALONE. THE PRINCIPAL LIMITATION OF THE DETERMINISTIC APPROACH - ITS ABILITY. TO INCORPORATE ONLY ONE MODEL AND ONE DATA SET AT A TIME AND ITS INABILITY ! TO ALLOW WElGHTED INCORPORATION OF NUMEROUS MODELS - CAN BE ASSESSED BY COMPARING ITS RESULTS WITH THE RESULTS OF A PROBABILISTIC ANALYSIS j ACCOMPLISHED IN PARALLEL. i SIMILARLY, THE PRINCIPAL LIMITATION OF THE PROBABILISTIC APPROACH - ITS TENDENCY TO ALLOW ITS RESULTS TO BE DOMINATED BY THE TAILS RATHER THAN THE CENTRAL TENDENCY OF DISTRIBUTIONS OF UNCERTAIN KNOWLEDGE OR EXPERT OPINION - CAN BE ASSESSED BY COMPARING ITS RESULTS WITH THE i
- RESULTS OF ONE OR MORE DETERMINISTIC ANALYSES.
l NATIONAL RESEARCH COUNCIL PSHA STUDY (1988) RECOMMENDS IDENTIFYlNG "THE COMBINATIONS OF MAGNITUDE AND DIST/sNCE THAT CONTRIBUTE MOST TO ! EARTHQUAKE HAZARD" (P. 51, 52, 58).
- l#
1 i 25 l t i
PROBABILISTIC ANALYSIS l STRENGTHS
- o EXPLICIT RECOGNITION OF VARIOUS HYPOTHESIS, MODELS, OPINIONS, AND
~ UNCERTAINTIES IN PARAMETERS - QUANTITATIVE IMPACT ON RESULTS EVALU-ATED. o RECURRENCE PERIOD AND SElSMICITY CONSIDERED IN EXPLICIT QUANTITATIVE FASHION. o IMPACT OF NEW FINDINGS AND NEW HYPOTHESIS EASILY EVALUATED. o EVENTUALLY, CAN BE USED TO ESTABLISH OR MEET A PERFORMANCE GOAL. o PROBABILISTIC STUDIES USED AT MANY SITES IN PAST TO ADDRESS VARIOUS SEISMOLOGICAL ISSUES (E.G., DIABLO CANYON AND EASTERN SEISMICITY ISSUE). WEAKNESSES o " BOTTOM-LINE" RESULTS TEND TO BE DOMINATED BY TAILS AND OUTLIER I OPINIONS. i o LACK OF EXPERIENCE IN DIRECT LICENSING APPLICATION. i i 2.6
DETERMINISTIC ANALYSIS STRENGTHS o EXTENSIVE PAST APPLICATIONS MAKING APPROACH WORKABLE, AND ALSO JUDGED TO BE SUITABLY CONSERVATIVE. i o SITE-SPECIFIC INFCRMATION AND RESULTS OF INVESTIGATIONS EASILY FOLDED INTO CONSIDERATIONS. WEAKNESSES o LEADS TO CONTROVERSIES AS DIVERSE OPINIONS ARE NOT EXPLICITLY TREATED. o PAST EXPERIENCE INDICATES DIFFICULTIES OF INTERPRETATIONS. 27
V I t i YANKEE ATOMIC ELECTRIC COMPANY i i PROPOSED REVISIONS TO 10 CFR 100 SEISMIC SITING CRITERIA AND i RELATED IMPLEMENTATION GUIDANCE i L Thomas O'Hara j John Sutton l l l i I l
Sii h hVALUAl'tuti t sa a vow , 5 100.10 Factors in be' consider:d when ; cvalenting sites. (c) l'hysical characteristics of the site, ; including seismology, meteorology, 4 Factors considered in the evaluation of ; y sites include those relating both to the geology and hydrology. , s' proposed reactor design and the charac- - A teristics peculiar to the site. It is expected (I) A p pe n dix A. " Seismic and j that reactors will reflect . through their Geologic Siting Caiteria for. Nuclear i design, construction and operation an l'ower Plants," describes the nature of j extremely low probability for accidents investigations respeised to obtain the ; geologie and sessimc unta necessary to - r that could result in release of significant determine site suitability and to p: ovide ! quantities of radioactive fission products.
- reasonable assurance that a nuclear power . I in addition, the site location and the a plant can be constructed and opciated at - 'f engineered features included as safeguards l a proposed site without undue risk to the -
against the hazardous consequences of an -'
; health "and safety of the pu blic. It [
accident, should one occur, should insure . ~ describes.procedmes for determining the ! a low risk of public exposure. In particu- quantitative vibratory ground motion lar, the Commission will take the ' f design basis at a site due to canthitnakes ! following factors into consideration in and describes information needeil to [ determining the acceptability of a site for determine whether and to'what extent a ! a power or testing reactor: nuclear power plant need be designed to ; (a) Characteristics of reactor tiesign withstand the effects of surface Ianiling. I and proposed operation includmg: (1) Intended use of the reactor [ (2) hieteorological conditions at the l including the proposed maximum power _ j site and in the sunounding area should be } level and the nature and inventory of er considered-4 contained radioactive materials: I (3) Geological and bydiological . (2) The extent to which generally " characteristics of the proposed site may [ accepted engineering standards are have a bearing on the conscipiences of an f applied to the design of the reactor; . escape of ' radioactive material faom the ! (3) The extent to which the reactor- facility. Special precautions should bc ; incorporates unique or unusual features planned if a reactor is to be located at a
- having a significant bearing on the- site where a significant quantity of radio- ,
probability or consequences of accidental active cIllnent might accidentally flow l release of radioactive materials; . into nearby streams or riveis or might (4) The safety features that are to be find ready access to underground water . engineered into the facility and those tables. ; barriers that must be breached as a result (d) Whese unfavorable physical char- . of an accident before a release of radio- acteristics of the site exist, the proposed [ active material" to the environment can site may nevertheless be found to be occur. acceptable if the design of tiie facility ; (b) Population density and use charac- includes a ppropriate anil adequate q teristics of the site environs, including the conypensating engineering safeguards. i exchision area, low population zone, and _._____.._._._mm_m_ _ _ ____________mo_. _ _ _ _ _ _ _ _ _ -____________m.-,..#.~-r_ -,_._.,.-#.,,, .-m .u_.___.~_ # - _ _ _ _ _ __ , - ,.-..%. , . , - . . , .wm.,,., . # m.. m_~_.,
(1) Appendix A, " Seismic and Geonogic Siting Criteria for Nuclear Power Plants," describes the nature of investigations ' required to obtain the geologic and seismic data necessary to determine site suitability and to provide reasonable assurance that a nuclear power plant can be constructed and operated at a proposed site without undue risk to the health and safety of the public. It describes procedures for determining the i quantitative vibratory ground motion design basis at a site due ' i to earthquakes and describes information needed to determme
- whether and to what extent a nuclear power plant need be j designed to withstand the effects of surface faulting.
PROPOSED PART 100.10 (C)(1) REVISION
" Consistent with General Design Criteria 2 of Appendix A to Part 50 of this chapter, geologic and seismic data analyses sufficient to determine site suitability and to provide reasonable assurance that a nuclear power plant can be constructed and operated at a proposed site without undue risk to the health and safety of the public shall be developed."
i
- -- - - - - - - - - - _ - - _ - _ - - _ _ _ - - -_ a
SIMPLIFIED RULE APPROACH
- Delete Existing Appendix A
- Confine All Details (Including Defm' itions, Nature Of Investigations, Endorsed Methodologies, Etc.) To
- Appropriate Guidance Documents
l l l i ADVA~NTAGES OF THE SIMPLIFIED RULE APPROACH ! i e Consistent With Existing Regulatory Treatment Of Other Significant Siting Concerns e Emphasizes Technical Rather Than " Legalistic" Treatment Of Issues e Provides For Straight Forward Treatment Of The Existing Population Of Plants . e More Readily Accommodates Evolutions In Our Collective Understanding Of The Issue e Can Sustain Possible Court Challenges ^
(2) Meteorological conditions at the site and in the I surrounding area should be considered. ' F i n i I
- 4 k
TOM O'HARA i YANKEE ATOMIC ELECTRIC COMPANY COMMENTS ON DRAFT REGULATORY GUIDE DG-1015 SEISMIC SOURCES i [ 2/5/92 i l i
t OVERVIEW OF PRESENTATION '
- 1. INTRODUCTION
- 2. OVERVIEW OF PROPOSED NRC APPROACH
- 3. COMMENTS ON DEIERMINISTIC APPROACH-i
- 4. COMMENTS ON PROBABILISTIC APPROACH
- 5. CONCLUSIONS
~
GENERAL COMMENT
S i
- 1. SUPPORT REVISION OF SEISMIC SITING CRITERIA
- 2. SUPPORT THE USE OF PROBABILISTIC AND Dt-lERMINISTIC METHODS
- EUS e WUS
- 3. COMMENTS APPLY TO APPLICATION OF PROPOSED DG-1015 IN EUS
- 4. NPOC GOALS
NUCLEAR POWER OVERSIGHT COMMil I EE ; GOALS: . l
- 1. ASSURE THAT REGULATORY PROCESSES ARE IN i PLACE FOR PREDICTABLE LICENSING, ,
INCLUDING SITE APPROVAL... l i
- 2. ASSURE TFIAT A HIGH QUALITY, STABLE ,
REGULATORY PROCESS IS IN PLACE TO ASSURE i SAFETY AND ENVIRONMENTAL PROTECTION... .
._.._..-._.m... . . . _ _ _ - - - - - . - - - - - - - - - -
a . . . . . . . . . . u u u u u u i_ -- -uunauuuuu oem-sosa4.uuroe=ac.m r o e = w m en uuueoe=wam u u u r e r r r r r r r r u u e o e a a e, a . u o e #
)
i 2 T r 2 l r t 3 i PROBABILISTIC SITE DETERMINISTIC i a ANALYSIS ANALYSIS '
'+
Geological. Selsmological g and Geophysicalinvestigations o ,. 8 E i e o I I ' o . e . u Conduct an EPRI or LLNL Identify Seismic i
- g i Seismic Hazard Assessment Sources
" i a
% I j Compare to Operating Determine Expected Maximum o Plants to Set Probability Earthquake for Each Source i of Exceedance Level i t
-a O l
,+ Determine Controlling Determine Controlling I
l i Earthquakes Ms & Ds Earthquakes Ms & Ds m i M ,
' I Compare .
- I Compute SSE Ground Motion
. r .-
.. t
'd .
I 9 _ - . - _ c . - - . _ - _. - _ _ _ _ _ , _ . _ - - _-_,-_-.--_..__.----_---_--_---_-__.______-__.._-_______-______-__--_____-_._--__-c.__-___ _
. - . . . - - _. . _ - ~ - . _ - . . , - . ~ . . . - . . _ . _ . - . . - . . . . . - . _ .
. . . l
/ pg..-
p, . - - I smance 2 U11: . klaste 5 emes 3 0 m uri kipec l Semce 1 I l klonec
$seca e i
I i j M f E sa m. N .... '....' . y # 7.4, *, e g
/. e *
/ l \
- e / \
/ :t \ ,E!SK SIEG i 4
/ *
. \
/ / : \
/k.
' ~
Eg \ "Elr k I f k 0.01 0.1 1' 10 Period (set) l Figurt 2. Schematic representation of the determination of tne
- Contrc,lling earthquakes for the deterministic analysis patn.
i OG-1015-13 l t
i SEISMIC SOURCES DG-1015 A "SEISMOGENIC SOURCE"IS A PORTION OF THE EARTH WHICH IS CONSIDERED TO HAVE UNIFORM SEISMICITY (SAME EXPECTED MAXIMUM MAGNITUDE AND FREQUENCY OF OCCURRENCE). NUREG/CR-5250 VOL 7 s. u. A <. w i A SEISMIC SOURCE ZONE IS A REGION WHICH HAS HOMOGENEOLIS SE!SMIC CHARACTERISTICS IN TERMS l OF RATE OF ACTIVITY, MAGNITUDE DISTRIBUTION l l AND UPPER MAGNITUDE CUT-OFF. l l
EXPECTED MAXIMUM EARTHQUAKES 5 DG-1015 EXPECTED MAXIMUM EARTHQUAKES ARE THE LARGEST i EARTHQUAKES THAT CAN REASONABLY BE EXPECTED TO OCCUR IN A GIVEN SEISMIC SOURCE IN THE CURRENT TECTONIC REGIME. - NUREG/CR-5250 VOL. 7 THE UPPER MAGNITUDE CUTOFF CORRESPONDS TO THE j LARGEST MAGNITUDE THAT WILL OCCUR GIVEN THE CURRENT GEOLOGIC AND TECTONIC CONDITIONS WITHIN A SEISMIC SOURCE.
]
r
i 3 r , i t 1 TABLE 3-1 EUS ZONATION AND SEISMICITY PANEL MEMBERS (S-Panel) Dr. Peter W. Basham(2) ,
- Professor Gilbert A. BollingerII)
- Mr. Richard J. Holt (1)
- Professor Arch C. Johnston l
- Dr. Alan L. Kafka
- Professor James E. Lawson
- Professor L. Tim Long(5)
- Professor Otto W. Nuttli(1)&(4)
- Dr. Paul W. Pomeroy(1)
- Dr. J. Carl Stepp Dr. Anne E. Stevens(3)
- Professor Ronald L. Street (1)
- Professor M. Nafi Toksoz Dr. Carl M. Wentworth (3) ,
Notes: (1) Also participated in the SEP Panels (2) Only provided zones and seismicity parameters for Canada l (3) Only provided zonation--no seismicity parameters (4) Also member of the Ground Motion Panel (Table 3-2) l l.s (*) Final member of the S-Panel b
. i = .
I'. , X :
CONCLUSIONS ON THE Dt-I ERMINISTIC PATHWAY l 1F THE NRC WANTS A Dt-IERMINISTIC PATHWAY THAT WILL NOT RESULT IN ENDLESS OPPORTUNITIES FOR INTERVENTION AND DELAY, THEN THE NRC MUST
- 1. PRODUCE A SEISMIC SOURCE MAP FOR THE EUS SIMILAR TO THAT REQUESTED BY THE ACRS OF THE NRC DURING THE INDIAN POINT HEARINGS (MID 70S), AND
- 2. THE NRC MUST ALSO DEFINE THE EME FOR EACH OF THESE SEISMIC SOURCE ZONES.
_1 PROBABILISTIC SEISMIC H.AZARD ANALYSIS ,
- 1. DEFINE SEISMIC SOURCE OR SOURCES t
- 2. DETERMINE SEISMICITY PARAMETERS FOR EACH SOURCE ZONE i
- 3. Dt-IERMINE EME FOR EACH SOURCE
- 4. INTEGRATE OVER MAGNITUDE AND DISTANCE TO DETERMINE EFFECT AT THE SITE t
I L
1 PROBABILISTIC SEISMIC HAZARD ANALYSIS (PSHA) i PSHA ALLOWS THE INCORPORATION OF UNCERTAINTY AND FREQUENCY OF EARTHQUAKE OCCURRENCE. IT CAN ACCURATELY REFLECT THE TRUE STATE OF KNOWLEDGE AND LACK THEREOF. PSHA IS CAPABLE OF ABSORBING AND INTEGRATING A WIDE RANGE OF INFORMATION AND JUDGEMENT l AND THEIR ASSOCIATED UNCERTAINTIES INTO A FLEXIBLE FRAMEWORK... l LEON REITER - EARTHQUAKE HAZARD ANALYSIS ISSUES AND INSIGHTS (1990) I
I l APPENDIX B Dt-IERMINATION OF SAFE SHUTDOWN EARTHQUAKE , , GROUND MOTION THE PROBABILITY OF EXCEEDING THE SAFE SHUTDOWN EARTHQUAKE GROUND MOTION IS CONSIDERED ACCEPTABLY LOW IF IT IS LESS THAN THE MEDIAN PROBABILITY COMPUTED FROM THE CURRENT ,. POPULATION OF NUCLEAR POWER PLANTS.
THE SSE IS ADEQUATE WHEN THE PROBABILITY OF EXCEEDING THE SSE COMPARES FAVORABLY TO THESE LIMITS , 1 TABLE B.1 DG-1015 . ESTIMATOR METHOD MEDIAN MEAN 85% LLNL 1 E-4 (ee-s> 2E-3 o.8e-33 2E-3 (i.2e-33 I EPRI 3E-5 o se-s) 6E-5 (6 oe-s) 1 E-4 (9 ee-s> VALUES IN () = HIGHEST COMPOSITE PROBABILITY OF EXCEEDING 0.3G R.G.1.60 AT EXISTING SITES
1 DETERMINATION OF ACCEPTABLE PROBABILITIES FOR LLNL AND EPRI PSHA METHODOLOGIES
- 1. CALIBRATE LLNL/EPRI RELATIVE TO A COMMON DETERMINISTIC STANDARD ASSUMED AT EXISTING EUS SITES '
i
- 2. BASIS FOR DETERMINISTIC STANDARD BECHTEL (1986) Dt- I ERMINISTICALLY EVALUATED 21 POTENTIAL FUTURE SITES - 0.3G ENVELOPED 4
DETERMINISTICALLY DERIVED SSE 0.25G HIGHEST DESIGN LEVEL AT EXISTING EUS SITES 0.3G ACCEPTABLE AT ANY EXISTING EUS SITE CONSISTENT IN PROBABILITY WITH DG-1015
CONCLUSIONS
- 1. CONTENT OF RULE SHOULD BE SIGNIFICANTLY SIMPLIFIED
- 2. MODEL FOR SIMPLIFICATION SHOULD BE CURRENT TREATMENT OF HYDROLOGY AND Mb FEOROLOGY
- 3. DETAILS OF HOW ONE DEMONSTRATES COMPLIANCE WITH GDC 2 SHOULD BE ADDRESSED ONLY IN REGULATORY GUIDANCE DOCUMENTS
CONCLUSIONS
- 4. PROPOSED NRC Dt-It-RMINISTIC ANALYSIS WILL PROVE TO '
BE UNWORKABLE IN EUS , i i i
- 5. ACCEPTABLE SITE PROBABILITIES SHOULD BE DEFINED USING A COMMON DETERMINISTIC STANDARD (0.3G R.G.1.60) i i
- 6. PROPOSED NRC PROBABILISTIC ANALYSIS WILL WORK FOR DETERMINING A FUTURE SITE SSE IN THE EUS :
i
- 7. EVALUATION TO DETERMINE MEAN MAGNITUDE AND DISTANCE NOT NECESSARY i
6
s
..~....
BZ7 : i~~---------......,,,_,. . . . . .
" "'.- l BZ3 ./
/
i
/
, p, Z
}n j
/
V *-. ( 3.:' ( .
. .t t
a t \ g : .
,, ( a '[pg,19,,... I !
I f , - G **l,,t,
\ ,,
t -m .
' ~
* .; - .-; h ,
..y ,
=u . e
- f : 1 i
,- gL ; j _g f y- i SOURCE ZONE
-
- 7-.E,, g'y -
- 8ACKGROUND ZONE ;
I ** M '! . .\.' , j l .- t
/'
BZ2 A t c._ p
, ~$
j b i t I i [ Figure C-5. Ilechtel Sa.6imic Source Zones
<~~ . .
. . . . . - - - . ~ . . . . . . . . . . _
+
/ --}
/.j .. -
-e
< t
,! .c. .. ....
. -. 1 '
..... ? .... . ..........--....._:. -~
.:0 g g ,
y s ...
- e : <
I : .' . '~ . .f . ' i.. .., .
..s f .. !
\:
l ..
.' / ...a.
c
,.- te' l : :
- n
.......bs ej i o .
9' c . ,.g.. . . . .r.<
,s' %!'
. , . ~..... r ,. ,.; .N.%._ _ _ . . . , . '%'
... ..f
- l. .,?:';ii .cl ~
- .'t} :.' . ~t.
. . - %, L.
~' .j.
? . . -
~~
l . -
~
.. ~
4~3
,, ~ ~ - . . . . -
e
- r '\ .
c 4 , *'- . < :' ; ,'
*s .i
..j.};
1
! ? '. i. * .
. .k '.?.*.. -( * " ~ b , ,.
) \., ..
*~~- ~ . .
f. g
- j , { ,.
- . s,..
p
\ . - v* j s'
[ ~- -
'=
. . . i '. . ...* ;
I ;
....... . ~ ' '- .- i
*.l .~y..... .
I
. . . / . .~. . . - .
.: .I
...a t I i ,,
.m .
~
; ~: ...,e . . . ..,- . - .. .
~
/ .
g
.'v.. .
- y,
. .. . s t.%.. 1.
- i. .. .
- a. t.
t. f a
- .a.c-r'
,~~:';.l...
. SOURCE ZONE
. r ' - ' ' ' '
BACKGROUNozoyg e
' 'l h 'y , %. .* . .
\,
.- . n. .'l' .
~ .
.. :. e t t-8 sr t.
m s.
. 3., *.s .k.
/ .... ....k lj
$-* ~
Figure C-2' # " Geolhysical segg,gg Source 40nes
9 r .
-,__ =
1
.7
T M& / "s d T
,z _
1 m' . 4 . . - y
/
2, f; 'n
\
A. ..- l
,, j. ...t n4 l .-~r.
/
c - .:. _I'2 a.I'.9.
' to I -
l ' l3 .- m; f no y
~ > , ,t y 'alc ~.y N
I (
~
G d[ g__)* _ s _
'!L . c 5
e
/.
NS ,
.a . .~.n
,, c ., >
/- ..
( : _ ::.~
-t. f. -
\. ~ ~I , *'., ,' , , ' _ 80 8
~~
h \ E1 N .; 12 4 = i
' ~;?
126 g 'fif}~. sm~~~ .
,.~~~~s"**'
\
\/
-...... BACKGROUND zg PseuAny >
r s Figure C-3. I.aw EH9f neerin9 Seismic Sourn. g , l
i L
~
s
.,\
s
.* +
,, \ >
s.
~~ h.
' \
s
--.......------~"""~'"?
1
""**~~~~~~~'~-~~-
-.- , ; / \
/
** \ .,
v
- \.g
[
-r
= 49 P< ' . s. ;-
l , p~ s *
~,=
" .'s f.
.* w f
w
'}
*.. s *-
g K _. 4'Y T l as
=
r i
".b .
-' .f
( '
}, [+ $ ,, . G ,
(
- s?
e r- ,. L 'l 3i J!
~
a I Il .
't It 3_
g ' $2 g i
//
St- 5 SOURCE ZONE O +
- n '
~~- B ACKGROUND ZONE
-N ' 9 r
------~.-..
. L
.\
. sf .
51
\. l j j
j 4 Figure C-6. Rondout Seismic Source Zones.
. . + . . , , - - ~ - - 2 a-a.-. > s a.us - - .s- .-=----xunas.u-- -..---na.n- ~ ,..u.u., - s, su..>.-.1-.- -.n.
l I W 5
~
4
~~ W y '
< , q'~ @x }
\ /
n
- . ~ .
a 1 e e
, ' a
. ,,' u i
1 g\hg . t e, 3, -
/
wp f'i a
, , y ,
~ N i: \v'
\ -\
1 1 ;
', \, .
g , 9 i s M [i s . O.! ' e 5 { c i
\
,, ' t s
*1
) ( t s
,s G
' e /- 2 [. ** %.
k 1 = . ( * /'\ i '
\
e- , s [ *
,t ,
, .I I eb" sum g f 8 t
t i e r""7 aY\ "
\ 4' \ \
"b ,r ,. s I e Y/ T... d t ' \ >
s\
- s J 1
1 s
- I i . \
s ,
)
I i , c- . , I g t i s -- I L i ..--4t e d st I
- y
-4 t'
b.t... ___ l d* I
- e i
l i .
\! '! '
./ * ,
l t g t~ ~ 1 d i l
n t i G 1 I i W 3 C N
. b a
3
- 44
~
s m e
= \
i \ t
$ a u
e
- 1 t
I
$g \
s. e < S % A $ - e i
' V @
;/ k>
\
\ / -
- g
^
?g o
t - c5 G y
)
/~
V. ~ y E, W n e D
- in
\
/
/ -
\
\ G'- 37- ~ 11 )
i l Appendix B i Maps of the Seismic Zonation for Each of the 11 Seismicity Experts i f B-1 1
_ aA ., __A_s4 as 4. 3a - ==e- 4 A.. == - ' \' \ ' l
/
\
\
;- i
(
\
i
) ,
\
\
bli kkk$ s
) v ,
N .
, o.
g s N ' 5 b s s I
, x s e
0 n
- 1 g y ca
- r. t k.
1 1 B-2
- u. >.a .as-..m J 1 1 .a.x n 2 . . , _ , . - a- --
< e s -
Sk?% l 4 .s.1) X A i e ( 8 M N v i n - N - 5 g - V R - 1 .: C
=f s a a n ( : O N r h
g I 2 [ L& .' l l } B-3 l o
, .t
,w
. /
"d '
2)x \ h -%bN _7 N w - y ( gs Ny - s u m X 1 p, 1 i~ , .
~
.t x 5 's, 1 ej g
9gx @N s
's (s .
1 \% NN ' s e .
\ =
%j9 \/ -
/
i,
, o
= c? ; d'swh v[
2 I =
)
1 *
, g aw -
1 og 1 ,, -
$$f " -
T e -3 d
i t ' 8
\,
\s 3
+
5 B-4
A 4 9 ~~-,-a. :p A 4 -- -- 43.-- 4., Ake - r, a a ena M -Lu_,,a, a & Lway4 u--Je 4 sb -wM--ana,.M,,.s - k.- apAa a 2 t 6 / t \
~ l 1~ -
N x .
- =
\w i 5
- N e g U N - &
a
\ :
I i /
- i e -
s.
\ 4=
i
+
o
- f5e
- u.
~
/
I _ - L l B-5
,,-**~__,'"\
1 l l
'i
\
/
\
\s
%} '
~
YA m
\
a
.\
a \ v \ w R ; O E
- N n . \
n * # C
.G N
k e=* i O 7:a
~
I B-6
l 8 s 3
\m t .
,~1 ,
s
- u.e- ~ )~ f
. , Q N .
/ - s
/ s (5 g*
\
1 j [9 S a h x e 1
. a -
Q #. Is
\
\
- g.7
.w ._ - AA s m. ,. = .x -.s.s.-an --,s ,
s... a.u a .I ~ s-.. es. - - .- -n f I l \ [ 8 I 7-kz e q L i Q.i[k~ sA% 2 i; 1 / = s C N\\
/g N,
\
s N 5 x 4
/ 1 1
w* S y i.
\
-zt 4 I i N
E, e W 4 B-B
\
- _ - - , _ = , .. , - - . . , . , ,_ _ .,
i [a/ , q '#-
\
~
i e e
= ,
- o
/ a p.
4 N - N . a. d
~ a a -
O".\)
\
s s g.9 l
f, aN
\'~[.N
/
. -s
.O /-
m-
- e ss3 , .
$ a *
./\
/
3 geq
=
/ t
,[ 4]9 <
e 4 B-10
,,nW rs _y_c
J Y
-- H" " __,. M
'r, 1
8 a Q
%g-1 .
}., y i:
\ C* e -
+
xa 1 o 9' t4 a C t S
/ $.
k 4 % k
\
a ( S S N O. S
?
1A.
\ \
B-11
,_p.
N J
? 4 ,-7' i
i I
\
J
+
R. %i c 4 % , N *
\
=
o 1 a - \ c ,
\
e u hk /g s
@ a >
4' $ i s - o f,
\ < a a -
%.-.* y 2
~
~
^ O 11-n JN s
m B-12
i L W
- "B .
.m i
b y _raxs v g
.N?
. A\
*\ g
_p a
-O x
. Y.
\" ,4 ss; ,
r, c . t i A i:
/
C-e e o
\
B h v's7 5 . S @
'N f
g.T3
A e 4<=u w -w. - 2- -4 ma 2 - .-a La s/ ( ~(b i {% s *% =
.3 D'
x
~$
-1 2 l l w a
s O N 2e E O
$ $ 3 e.O 4J QN 3
'N o
2 l l 5
- u. l l
B-14
m w ,, n 4 is4.a= & ""^3 A -
\
\
.c:r \
f - \ [ 7 \
\
y ) \ T
& V, ' + . '
O \ e \ o
- 4 . *
/ a g
n
\
c e s e
\
Y, - ' 3, k
. \
Y p \ W h **
,6 C H
o
\
\ \ \ \ o
m a _m.. 4 _2- . . m 4..__2,. _ -
- ma#+ . _ _ - -# __
\ a \ e m
% ~
N - O m - e x e
/~
O cu 8. 5 N y
\
)3o N
Y N ,
# N *
$ \
a # \ 9 \ f O C 7 k s N. 0 . p
\
O e e
/[a
/ E w \
t g N M
' w
- O e #
-~j C, g.16
g ., - - _h y A. 4 a m ..cs Aa _w< m A - 4_AF ,.e a i
'9' n#
\
N 4 e e h s g
# s :
i 3 s* k #
/ .
g a 9 C x O ha.
$ d O
~
o e s a
. ,Y O
i 3 -
- \
q . ,a @ G g l, \ i \ h e
,_-'#' .s' _j,W' l
(
=
L- . d. _ s g
.O
+
l - g i.
h c :-
O
) /E. C ed 9 / 3
, s f 1
, r
.P , Y q .
, t ;
5 x *
\
o f v5 e4
! S u.
/
1 i I B-18
svna p(.fSP/890O R \l A G _ esce O ,
~
. OFFICIAL TRANSCRIPT OF PROCEEDINGS
\'
- a. #
m;;;,.; ?
+ j. g'j.'
L '
'jj 4,
~SMN s % ,,;;; ' ,, * +
,geq1f,1 y^O p' r ' ( . ,'.j l + , , q; f}.R [.;["';q, ' *I, s j, jf?;',j it;[ ;qfl1 kN M.M?k ., "
* , ,i Y
' k; , 4.+'" , , ';? y' yI !< y.c.rJ! W >
@,',d I
*j_',, .
., +\ 't. - .. y Q'.
l g? 'f,i[ .. . . -
. A_ RT " \[ ; g . i
+
c, y. , K ~ :. , y y) ~ll 1
.f kL pEB t 01991i *f G*:i! i l~N/51%pl) j.:1g.%', -- -l ,.f I
^
t 4 5 c . i o ' .v "g3".. 42 , ; . ,; .
,' , g amcTOM.
',y.,
l ,
~
1 0.C. d '
'p.,
+
g
,e g s . - i
^
V',
^
MCU@ U.S. Nuclear Regulat O sion ,.
,['- '
[' ; Advisory Committee On Reactor Safeguards ,' ; t . < ' Tide:- 382nd neeting , 1
\
v . Docket No. U ( -
,x , . . .
(, ,0 l w I MUott Bethesda, !!a ryla nd V OATE Friday, February 7, 1992 PACD: - 291 533 :
- n , , ,
)
.g
.8 -
n ._...
, e u Vr...,Ce ony
- a,_,:am_ e
~0' *.I oo 1;N c ^f'h^ U1 [uc j0lii;lili.te.
- - - - 'u 1200.;0 ;
2 Y ANN RIIF(& ASSOCIATES,I.TD. )
'- A 1612 K st MW,Suke M0
% DC 20306 l ,
(202) 295-3950 g l NPDR' O192-920207' ACRS I
.~'" T-1R90 png l
l
>. f 3l'
, J m ,
'M ]
A G M A am ,
~
\ y1 ' I
.q 1): , -
+
.4 ' :. , , . . .~. - .. ,, ,
- n. .
W;g sM,, #y Lw#.Ady.,M.&vOFFICIAIR,IRANSCRIPT w M KWi: lv W -u 9 > M e ~
- g% W s a p
, m Q nlW 's, t ? MDJ)N.. mwpp OEfROCEEDINGSM:g<gsp a , : ., 99 w?g MM nM .Y'g a.~p:QQ~;&, .QT
! c M;W - Wd Mw "h,yy M-.xg i:(e.nuh%l{W ,a w vfN% u dp '
, :qW rn ,, s ! M 95 e ' e um A m ca
% b.e:s x g&c ny44 .- A Uw pp eyy)%wyyy%m@ +f4 j Q; Q Q $m p w/w?w p gf$
g&&w,Q.q;llr&Q?pgf QA gy 4 Q QQ pgky l
%W p%59y RE:H M A/a- r LM.W .
- ghhn4p hyast d.w]#iv e@hh;.gvk e
's a p Mm'qn w ~
v r
@;.y@ux< Nk,#g m p m v~eedt -
- t-dc" g.
V
-w 3,f h h@h d
~
r wlll.M.NhbNh':s"h c ~; w.
=4&Jw w m w hh w vc % ed n% VQ o m, i
I h yn , a WhNY ifl &. L N ' A # U M F #W
' $ ~.! c bNk_Eh?N&. Aq;&Mg &ONWWW$ .+W+$ Whg M@fW@MW{P.5$@( MM &b$0N.;q
? i ?NMy&fyQlfK v mWw.s M'hl, b v.
- s. %
.. . . '*,. Y3%MWh)?
, w ~ &g% !
OW: y+ d$sMQyyfMk EQVh$
$&%HgkMy m%Q; y :
q&,k- % Q-wm . Q Q.'/Mk ~ g +4VMW9 WhF
\
l mf' - ??.? &,, . f@.ca Q $$hk l k$ b h l,78 C Q .[,W [,WN.$USS.<NEclear.!Regulat ". ? ,, ~ Y h l si,on r g& jhM k h f fh% f'$h ac tor : Sa f M l 8Jg%@a 1
$.x:w;F? : W s s. Advis o ry;! Commi t t e .,,c. e ',Onvt ,Requgn e gua pj m ,.. .
p.. s,;cn y ,
.x xw m y; grr ; ;s e em: . ,
, ~p .~c x- w ,1 1, , = .
- m. q$m e s; mna synQ.m- Q 4,qw Wp 2 M. m. . s
.\ e\'
'u ' ~ ' ,M* '
?*q' "$ ',yl
%p.,: p k 1 C:W* G' j9' e$ >> ' &)?n , wy %w&h(v &:.%m e
i
;&q$W p cA a '
", q? 382nd: Meeting 4:q aa,a d, .' Ac WWM
&ys w w;Ns4 M
%,.u;~& .v
@,4yp:sp m;w m, ,4, W .,
s
~
.g .n N ,v,a,p ,4.,d fo. a m%;p w ,:,-
y 6 ..'I. 3 I m , . ..nHV
. ^, e s-- ; ij. M \ '-ik ),iy,f;soa .yfw.o .3q ic Q .WlH
@ %c <~ n 'l qJc. n <+ # g. f r f g._. - 3 -
p 4 .f ,. .--
- c .,h
. h&sc.q:h.,> DOC 0,b,';e M. W . '
o .dk.s v(,7?h,hWk.h , k,3 .Y p % d l, Mr;f, $ f N> ' Onw n . q u.-t,N. c O,Y . . N - s m s yo e o m .
.p:
pg e.smp.o r my#[t - t
%yh,[
4 G $@2[$ ,mn, ." Q .,g l @s. w;[g M.MaO M;gh m Ny G n..r [ .n l 3;vn 3 a. . a , m, e y1 gg y e e
; gJS,yyg,4t;e'? g
, ln,y .
; .;4 .7 j
. g y n1:<g.n .r7 n,,l - g, ,
, y~. w vg*p 9, ? a
. , g ,%, v.c,1Qp;gn.g,l y . ,
~
,' . v' : yn.,w wf 4e , ,
, . t
$ ;n y....d q f w x;, : %n. ,e w m u.. .. ,
' / w. .y a' .9 , % , V,: - Qi! b s ..v .
;y e Jp y q ~,
.cm; 1 n
;~
s ig-r
; il ' .,
. > ,w>+h \.
u.jyg , 4, u , ,r,p ' < f . , 10 CAT 10& ;
> @ Bethesda, Maryland s
- y4
- a o , m. ,
.m yQ, ym 3; s ;v .rs m%w,a tm s < ( 3.:
1 +:V s,. , ,
- 5- , ,;
v .. m .,. s . . .c. ,, v 2
., ; +. a. ga.
j h:: gy ws.[h(//2h ,DA K n,x gn h. N ,',-]9 4-h,F$iday,fFebruaryR7',;1992:( [
> . ; $y291$.5'33[g-< %
W, . m y ,m;4p ,s
&* ' 7(d Q,6,
.t .
.c m
~ .g 5.9
'.p > :m' 4 ' l ?* y , dit.i-G
< , p ,,, a * ( , ' : # #' .s
], *
, . . ' ^ ' t q )" " . [: ,
" 6 M '] W '[n 4 i [. ?( - ;
. %. . m.4
- sea; .- ;;
,y;'. t a. .> 4 a n,o -- e,, ,
h4 ^ } i g:l7s . ~,
- .m , ,
+
. ?,r . , , ,y',
v.; g ,
.( .
,,h-
*,.cw,
.I g g' f}.;.
q - M .#n * ,-
*- ,,,,;4
, g. 6 s( p 4
,h,I,If$
n . '-1Q, % h f 90 * .===.
** "J* '~
s Y 3.s Yh , ()
. 7ey: . 1, e -*
,e i k ,.,,
A e L,.,, q ;A t+ I i 4 4 , . . *' j -= =..
*tt S I*3 r + 1
,. y
'v . I b l h W]*
a 7m; y .
. )
Y s'
'% *L,
# '
- 4
;$*f[]m
, ,t w- r . :
'l -
c As s s 4 d ,g h ' g te 7 (
} l p; 1.f f kb $ , [ k h ,
l : 1612Kt.MW,$de300 M@ MpMG.W g fo w, p ltshinpan,DC 20006 , n e
%WSFC an t m =
1 (202) 295-3950
/y j i
tPD .
, , l, ,
W_y dQ .JPg#5dCity', ,920207. ~> yn b.T-te u@ %R ' ' :
- N '3,
', :P M'~,; "- ' t l
I m3m e w ,ma ; ..
, , v' ' >
+,
7 r - Ni% MO M.. * ! I
I ( i ..-- 453 1 1 [Brief recess.] ! 2 MR. WARD: Our next item is a discussion of a 3~ proposed revision to Part 100, Appendix A. Dr. Siess will l 4 lead the discussion. 5 MR. SIESS: I am going to make a very brief j
- 6 introduction, and then we will turn it over to staff. .If I 7 you will recall, the staff has underway a proposal to 8 separate siting criteria fEom plant design criteria. That 9 project is divided into two parts. One was the reactor 10 sicing criteria non-seismic, and the other was seismic 11 siting and earthquake engineering criteria.
12 The reactor siting criteria non-seismic included 13 such things as source term doses, population density, 14 meteorology, et cetera. We review'ed that.last month under i 15 the aegis of Dave Ward, and we wrote a. letter dated January 16 15 on that part of it. This month we are looking at the 17 portion called seismic siting and earthquake engineering 18 criteria which is the rest of the package. They were both 19 included in a draft SECY that we had dated October 10. That 20 has been changed or is being changed. We have never 21 received an update on it but it has been changed, we know. 22 We had a Subcommittee meeting in December at which 23 we looked at a lot of this stuff and gave the staff some 24 good advice, and they went back and changed some of it and 25 brought it back in. A lot of the advice was editorial, I f i I
454 (" } 1 will admit. We had another Subcommittee meeting on 2 Wednesday of this week. 3 At the first Subcommittee meeting Carlyle 4 Michelson and Charlie Wylie were there in addition to me, 5 and on Wednesday they were both there.- Jay Carroll was . 6 there for a good part of the day. Dave Ward wasn't there. 7 Hal Lewis was there in the morning and had to leave in the . 8 afternoon. We had Bill Linblad as -- I guess we had to call 9 him an expert rather than a' consultant there. 10 MR. CARROLL: It's probably the first time he's 11 ever been called an expert. He's been called a lot of other la things. 13 MR. WILKINS: At least he didn't react violently. 14 MR. SIESS: Consultants' don't have to be experts. 15 Anyway, that's the background. Although I could very easily 16 explain the whole thing to you, Andy Murphy has a good set 17 of slides and he can do it at least as well as I could. 18 Andy, I will turn it over to you. 19 He is going to give us an overall picture. and 20 then we are going to hear from Roger Kenneally on the 21- engineering part and then back to Andy on the seismic part. 22 One other item before you start. I have prepared a draft of 23 a-letter. When Dean finishes what he's doing, he is going 24 to pass out copies of the draft that you might just read and 25 glance at. It's not very long. It is in shape now to be i
e , 1
- 455 P ,
i
1 what we will pass out tomorrow unless we have a discussion ,
2 time after this, which I doubt. 3 (Slides.] 4 MR. MURPHY: We will start with the first 5 viewgraph quickly. That's what the presentation is going to l 6 be. There's my name, and there's the phone number. ; 7 MR. CARROLL: You passed the test on the' title 8 .page but you didn't number the pages of.your slides. I
-9 MR. MURPHY: Thats was a deliberate screw up, 10 because.they were still being made about an hour ago. Page 1 11 numbers'are available on some of them. Don't try to follow la them, though. ;
13 MR. .SIESS: They were all numbered at the ! l 14 Subcommittee meeting but he'Js only'had two days to get this , 15 thing reorganized. 16 [ Slides.] .; 17 MR. MURPHY: What we have on this slide are the i 18 issues that we'are trying to address with the revision that l 19 we have started. We want to start out with a point that we j 20 made at the'first Subcommittee meeting and was well made 21 again at the second Subcommittee meeting, and that is that 22 the current Appendix A has generally been. successful for a j 33 majority of the nuclear power. plants that were licensed 24' using it. 25 There have been difficulties. It is these
J. 't 456 s
)
1 difficulties that led in our minds for the need to revise i 2 this document. The principal ones that-we are addressing is 3 -that Appendix A was too proscriptive. A lot of the details 4 that today are prepared and put into regulatory guides were 5 in'that document itself, in the regulation. That made , l I 6 thir.gs on the inflexible side, and because of the level of l 7 detail there is also some difficulty with the clarity of the - 1 8 definitions that were used, the technical definitiens that 9 were used in that document. ; 10 MR. WARD: Despite the fact that there's a lot of , i 11 detail, there was lack of clarity. Usually, those things go 12 --no, not necessarily. 13 MR. MURPHY Not in this case. That detail led to 24 very specifically difficultit wit'hin the interpretation I 15 that was in there, zus we say , actually conflicting 16 interpretations. 17 MR. SIESS: There were some places you could j l 18 almost say the less we knew the more detailed it got. 19 MR. MURPHY: The next point is that the regulation i 20 is an old document. It has been in place since 1973. It - 21 took a number of years to develop it so that much of the 22 science that was in the document is on the order of 25 to 30 23 years old. Seismology and geology have made significant 24 advances since that time, so we are talking about a piece of 25 documentation that is reasonably well out of date. I? I
4 457 f. 1 Another difficulty that was noted in this thing 2 was the definition of the operating basis earthquake. The 3 document had three definitions; one associated with 4 functionality, the one with recurrence rate, and then a 5 ratio between the OBE and SSE. These conflicting 6 definitions, again, led to difficulties. 7 MR. SIESS: Andy, I think one thing I could 8 mention and maybe you wouldn't want to, another problem with 9 Appendix A was that we had almost quit using it in some 10 places. There have been enough exceptions made, OBE 11 exceptions that have been made repeatedly, and those weren't 12 really permitted. I don't know how we argued it. Half of 13 the SSE, we had a number of exceptions to that on another-14 basis. It was sort of -- it took an interpretation. 15 MR. MURPHY: That's correct. 16 MR. SIESS: A lot of the obsolete stuff just got 17 ignored. 18 MR. MURPHY: The next slide simply provides a list 19 of the documents that have been prepared over the year, 20 officially pointing to the need to revise this document. We 21 note that the first official indication in the SECY paper 22 that this thing needed to be revised or worked on came out 23 four years after it went into effect, and actually there's 24 probably considerable discussion before it actually came to 25 this document about difficulties with it.
l i II 458 h, + 1 MR. SIESS: I think it's misleading, because 1973 l 2 'is when it finally got approved. Work must have started on 3 it -- 4 MR. SHAO: Five years ago -- 5 MR. MURPHY: At least five years before that. , 6 MR. SIESS: I came on the committee in 1968, and I 7 think we were working on it. - 8 MR. MURPHY: It is an old document as far as the 9 technology is concerned. Again, we simply list three of the 10 indicators as to the difficulties with the document. The 11 next few items simply list the bureaucratic steps that we la went through to get permission to begin the process of 13 revising the document. 14 The next-slide provides'the objectives of the 15 revision. They go along with the difficulties that we 16- noticed with the document. First, we want to remove the 17 sources of misinterpretation. We want to increase the ease 18 of updating the technical guidance. We don't want to have ! 19 to wait another 20 years if science is changed to revise 20 this document. We want to provide stability in the - al licensing process. The final one is more a scheduling 33 objective than a technical objective. We want to have the 33 revision in place so that it may be used for the early site ; 34 reviews. 25 MR. SIESS: Do you have any indication of when you f
1 j- 459 1 might get an-early site review? 3 MR. MURPHY:' My understanding is that -- 3 MR. SHAO: Sometime next year. ! 4 MR..SIESS: Sometime next year, okay. . f 5 MR. MURPHY: I.would disagree with that a little . 6 bit. I-believe' DOE has just let a contract to three utility 7 organizations to begin an early site review process. I 8 believe the first two years of that contract were to do i i 9 preliminary studies, and that possibly in the second and I 10 third phase.if comebody would actually identify a site for 11 this early review. I believe that makes it at least two ! 12 years off from the time of siting the contract. ; 13- MR. SIESS: Are they thinking about a site for 14 custom plant or site for a certified'three-tenths G plant? 15 MR. MURPHY: I honestly don't know. I would 16 speculate that it is for a standard plant. 17 MR. CARRO'LL: These are all Eastern" utilities? 18 MR. MURPHY: Yes. I think it was Southern 19 Services, New Jersey Public Service and Commonwealth. 20 MR. SIESS: I have a handout here. 21 MR. SHAO: Yes, this just came out. 22 MR. MURPHY: Here we have a summary of what we 23 will call the scope of the revision that we are undertaking. 24 We are revising Appendix A, is what we call this thing. 25 Actually what is happening is, Appendix A will not be i
{ 460 l' revised. It will be left as is. On the advice of'the 2 lawyers we have to have some regu3ation in place to cover 3 the existing and operating facilities, so that the revised 4 regulation will basically be for new facilities. S We are revising Appendix A which is Part 100 - 6 MR. SIESS: Excuse me. Everything you are going 7 to talk about today is a forward fit, right? - 8 MR. SHAO: Forward fit, yes. 9 MR. MURPHY: That's correct. In the next slide I i 10 will mention a little circumstance where it might be used 11 for operating facilities, but it's intended as a forward la fit. We J ,ee created a new Appendix S for Part 50, which is 13 intended to decouple the design from the siting. This is 14 taking the engineering that was in' Appendix A and putting it 15 into Appendix of Part 40, the design section. 16 We'are creating some new guides on the geological 17 side, the seismic source term. We are revising -- and this 18 is actually a minor revision at this time -- the standard 19 review plan section 2.53 on vibratory ground motion. We are 20 revising the nuclear power plant instrumentation guide. We . 21 are developing two new guides, one on pre-earthquake 22 planning and another on restart following an earthquake. 23 We will have to take and revisit a number of other 24 guides to update them editorially to fit what we are 25 changing within the new regulation and the new guiden. e x_
I i 7 461 l 4 1 MR. SIESS: The revised guide on instrumentation, i 2- .that is not essential to this relocation process, is it? 3 You are-just taking the opportunity _to. toss that into the l 4' pot, aren't you? l 5 MR. MURPHY: Yes, that would be one way to put it. 6 We feel it's a little' bit more important than that. 7 MR. SIESS: I am thinking that normally guides 8 don't go to the Commission. Three of these guides should go I 9 to the Commission because they are tightly related -- 10 MR. MURPHY: Because they are new. 11 MR. SIESS: Even new guides don't go to -- 12 MR. MURPHY: They are a new concept -- ) 13 MR. SIESS: They are related to this rule change. 1 14 MR. MURPHY: Right. 15 MR. SIESS: But you could easily leave that one 16 out, I think. 17 MR. MURPHY: Yes, sir. 18 MR. SIESS: Okay. 19 MR. MURPHY: Just to summarize for the earth , 1 20 sciences, we are talking about a new Reg Guide, revising the i 1 l 21 standard review plan section. Under the engineering, we are 22 revising the instrumentation guide and developing two new 23 Reg Guides. 24 Here, we are talking about the applicability of 25 this new regulation. As Dr. Siess has mentioned, we are i i l l 1
)
[ 462 A. i 1 intending this thing to be a forward fit only, applying it 2 to new reactors. We do note here that there is the ! 3 potential of using this to resolve seismic issues with the i 4 current generation of facilities. There, it will be 5 incumbent upon the operating reactor to come in and request , 6 an exemption to go ahead and make use of the provisions that 7 are in the new regulation. - 8 It is not intended in any sense for this stuff to 9 be backfitted onto the power plants. It will be a voluntary 10 act on their part, if they want to try to make use of the 11 new provisions in the regulation. 12 (Slides.] I I 13 MR. MURPHY: This is a list of an expert panel l 14 that we had put together and has b'een working with us. We 15 have been using them as a~ sounding board for what we have 16 put together. We have not requested them to write a report 17' or whatnot, but to simply use them.as a sounding board. 18 Some of the material that you will see today they agree with 19 and some of it they disagree with. It's a panel of well 20 known folks that have been working with the old regulation 21 and know its difficulties, and have sound advice to be given 22 to us. 23 MR. SIESS: Did you work with them collectively? 24 Could they meet together? 25 MR. MURPHY: Yes, we have met with them at least
1 l l 1 l; i ', 463 a 1 three times. The first two-times was with the full group l 2 and the last time the two members, Jim Devine and Walt Hays 3 from U.S. Geological Survey were not able to make it to that 4 meeting. Generally, we have worked with them as a group. 5 MR. SIESS: If we wanted to know what parts of I l 6 this they don't agree with, how could we find out? 7 MR. MURPHY The easiest way would be to ask them. 8 MR. SIESS: We could write a letter to them and j 9 ask them to give us their views? Since this is going out j l 10 for public comment, could we expect them to comment during l 1 11 that period if they disagree? 12 MR. MURPHY: Yes. At the last meeting they were i 13 invited to take that opportunity. One of the experts -- and i 14 I don't want to name names because I may slightly 15 misrepresent what they actually feel. One of the experts 16 definitely disagreed with our use of the deterministic and 1Y- probabilistic. That individual felt that only probabilistic 18 techniques should be used. 19 MR. SIESS: Only one of them felt that way? 20 MR. MURPHY: It is my understanding and my 21 recollection was that only one of them specifically said l 22 probabilistic only. All the rest of them felt a combined 23 methodology would be appropriate. 24 MR. SIESS: Good. 25 MR. MURPHY: I was going to end my presentation at l B
)
464 f 1 this time and will ask Roger Kenneally to give you a 2 presentation on the engineering material. When he is 3 finished, I will come back and talk-about the geology and t 4 . seismology again. 5 MR. SIESS: Thank-you, Andy. Are there any , 6 questions for Andy before he sits down this time? 7 [No response.] , 8 MR. SIESS: We do have this division between the 9 engineering part and geologic and seismic part. That's the 10 Eway it was divided with the staff. , 31 (Slides.] 12 MR. KENNEALLY: The staff is proposing a new 13 appendix to Part 50, the designation is Appendix S. That 14 was the next available letter. It'would be the earthquake 15 engineering criteria for nuclear power plants. 16 MR. SIESS: What do you do when you get to Z? l 17 MR. KENNEALLY: Double A. Within the current. g 18 regulation, earthquake engineering criteria are currently ! t l 19 located in Appendix A.to Part 100, the seismic and geologic *
?. 0 siting criteria. It is actually Roman Number VI of that .
21 Appendix.
~
l 22 I have highlighted on this viewgraph a few of the l 23 things that are important that have come'out of that 24 particular Appendix, and also the section to the regulation. 25 It defined the safe shutdown earthquake and the operating l
- .. - _. - . ~- . . . $ 465 ,
1 basis-earthquake. Next to it, I have provided places within 3 Appendix A to Part 100 where these particular topics are. I 3 will not repeat that every time, but sometimes it's awful , 4 hard to find things in that particular one. You say I know 5' it's in there and cannot recall where. 6 It defines safety related structure systems in 7 components in a few places. The minimum value of the SSE 8 was defined, a ratio between the OBE and SSE. Also, a 9 requirement that plant shutdown would take place if the OBE 10 was exceeded. Vibratory ground motion was defined by 11 response spectra at.the foundation elevation. This may be la one of the cases where we have had some problems. It's 13 inconsistent with what engineering practice would deem as a l 14 good measure, and the 1975 Standard Review Plan had 15 reflected something different than that. 16 It identified acceptable analytical methods, the 17 design for surface faulting in other design for seismically ! 18 induced floods waterways, and also a section at the very end L l 19 of it on soil' considerations. The bottom line of what the l 20 staff is trying to accomplish is incorporating within the i 21 regulation the Appendix S to Part 50. As the lawyers have I 22 advised us, the Appendix has to be cited other places within 33 the regulation to be effective. So, we have taken some 24 ch'nges a to 50.34 which are the contents for the application, 25 the technical information, ( A) (12 ) would be your PSAR type
466 x[-
-1. of submittal or (B)(10) the FSAR. type submittal, simply to 2 identify that after a'certain date the Appendix S to Part 50 3 would be applicable.
4 The staff'has also taken a change to 50.54, 5 condition of a license, that would require an exceedance of , 6 the OBE or damage,.would be a condition of a shutdown for 7 the plant. We have. proposed based on industry work,- - 8~ criterion and direction as to how to accomplish these. . 9 MR. WILKINS: Is that any different than what is a 10 already in Appendix A in Part 100? 11 MR. KENNEALLY: Different, from the standpoint 12 that at one time staff and industry thought that that'may be I 13 licensee invoked action. A review has indicated'that-that 14 may not be the case. So this way, by putting it in -- 15 MR. SIESS: Excuse me. Were you talking about the 16 OBE? 17 MR. KENNEALLY: Shutdown, if you have exceeded the 3 18- OBE. 19 MR. WILKINS: Yes. I saw exactly the same words on
- 20. the previous slide. -
21' MR. SIESS: Yes, but they are being moved to , 1 22 . Appendix S. j 23 MR. KENNEALLY: They are being moved to Appendix 24 -S, and the requirement is put as 50.54 which would j 1 25 reference Appendix S. 4 i
"- T
- y
I 4 467
- i 1 MR. WILKINS: I am with'you'now, thank you. :
3- MR. SIESS: Everything on the previous slide is l,
- 3. 'being moved. ,
'4. MR. KENNEALLY: Yes, that's correct.
'5 MR..SIESS: Appendix A is being changed --
6 Appendix B will apply to future plants, and all of this is 7: taken out of Appendix A. t i i 8 MR. WILKINS: Yes. , i.
'9 MR. KENNEALLY: It will say in A, but won't be t
10 . applicable for the.-- + 11 MR. SIESS: Out of Appendix B. Is that. clear? 12 Appendix A is just what is.in there now, and Subpart A in ; i 13 Part 100 is what is there now. There's a new Subpart B, a 14 new Appendix B for future plants. Appendix B -- if you saw , 15 the marked up version it's a fraction of the old Appendix A. 16 MR. KENNEALLY: The proposed Appendix, as we have 17 stated a few times before, will not be applied to the { 18 existing plants. The criteria will apply to Part 52 or Part ; 19 50 applicants on or after the effective date of the 20 regulation. That'was one of the iterations we have had with 21 OGC to make sure'that the regulation will allow either type 23 of an application. 23 Part 52 applicants, as you are very familiar with I am sure relative to engineering, would be for design j 34-25 certification or the combined license. That would be a i I i 1
f 468 ., R ., 1 combined operating -- construction permit and operating , 2 license. The Part 50 applicants would be the traditional 3 one that we have had,'the construction permit or the 4 operating license. 5 As part of this revision, there have been some . 6 clarifications to the text. One of them would be that the 7 designation in the Appendix A safe shutdown earthquake isn't . 8 an earthquake, it is within the regulation indicated that it 9 is a response spectra. So, we are trying to make that known 10 by saying it is a SSE ground motion safe shutdown earthquake 11 ground motion. However, we are still recommending that the 12 acronym would remain the same, SSE. 13 Rather than trying to get in and come up with a , 14 more definitive title or different title, it's in so many 15 code standards that the industry would be ten years before 16 we.could chase that one though. 17 Some of the ambiguity in the regulation has been 18 removed. Within the SSE in Part 100 is termo like that 19 earthquake, maximum earthquake potential,. The SSE is 20 characterized by a spectra, it is not in earthquake one. . 21 That type of clarification has gone. The control point 22 motion has been changed from the elevations of the 23 foundations to the free ground surface or hypothetical rock 24 outcrop as appropriate. There is still the concept that is 25 within the regulation of defining a minimum value of the
1 (, . 469 L l 1 SSE, of 0.1 G, and that one would be at the foundation 2 level. , 3 With regard to'the operating basis earthquake, it 4 was mentioned earlier by Andy that there are a lot of 5 different facets that are in the current regulation; the 6 functionality of structures, equipment and components, the , 7 likelihood of cccurrence. It's actually stated as affect 8 the plant site'during the operating life. It is defined as 9 a percentage of the SSE, and it is also associated with 10 plant shutdown. 11 In the proposed regulation the emphasis will be on 12 the functionality of the structures, equipment and 13 components, and also with shutdown if you have exceeded 14 criterion or have seen damage. 15 A significant change in the proposed regulation is 16 the applicant would choose the value of the OBE, at a third 17 of the SSE or a value greater. There are some thoughts that 18 the applicant would be doing relative to this decision -- 19 and on the plus side of one-third of the SSE, that the 1 20 amount of response analysis requirements associated with 21 that would be eliminated. However, they are also keeping in 22 mind of what one-third of the SSE would be and a plant 23 shutdown if a value were exceeded. So, they may elect to go 24 to a value higher based on economics. ] 25 They may say that I don't want to subject myself i,
[ 470 t
' Therefore, I would l to walk down inspections too frequently. <
2 go to a higher value or the OBE, which they would
.3 voluntarily select. Associated with that higher value would 4 be specific analysis and design requirements.
5 MR. SIESS: Roger, we have quite a few years of , 6 reactor operation now. Has any plant other than possibly Perry been subjected to something as large as the OBE?
~
7 - 8 MR. WARD: Humboldt -- 9 MR. MURPHY: Humboldt III. 10 MR. SIESS: That's right. We shut it down, and 11 kept it shut down. 12 MR. CARROLL: It was shut down. i 13 MR. SIESS: It was at the time, okay. 14 MR. CARROLL: It was before the -- 1 15 MR. SIESS: I remember. It was used to check off 16 some of the analysis -- 17- MR. WARD: Right. 18 MR. SIESS: -- soil structure. Thank you. Go 19 ahead. 20 MR. WARD: No one has ever had to go through this . 21 shutdown -- 22 MR. SIESS: Perry did. 23 MR. KENNEALLY: Perry wasn't operating at the 24 time, and it certainly delayed what the process would have 25 been of loading fuel. There really wasn't a lot of clear I
( I l I , 471 )
~ ll i 1 .
guidance of what to do during this period. j 2 MR. SIESS: hit got something very close to the l 3 SSE. IT.has a particular -- 4 MR. CARROLL: Spike kind of earthquake without 5 much energy. 6 MR. SHAO: The Japanese plant had been subjected ,
'7' to a couple of these.
8 MR. SIESS: That, I would expect. "Go ahead. , l 9 MR. KENNEALLY: There are certain places -- 10 MR. WARD: Is the design of the plant ---under the 11 new rules -- really governed by the OBE or by the SSE? j 12 MR. SHAO: It will'be governed by the SSE now. l 13 MR. WARD: It will be governed by the SSE now.- ]
.14 MR. SIESS: You pick one-third and you don't even 15 have to check it.
- 16. MR. CARROLL: That just becomes your trigger for -
l 17 - 18 MR. SIESS: That just becomes a trigger. 19 MR. SHAO: One of the criticisms of the Appendix A 20 in the past was because OBE equal to one and one-half SSE, 21 in many cases OBE come in the design rather than the SSE. 22 MR. SIESS:- If you go above one-third you are 23 going to have to check it. You will use different damping. 24 25 MR. SHAO: Yes. We use 161 lower damping -- k
- . .. _- - - . - . _ . _ _ . _ . ~ - __ _ . . . _ . _ -
I l l [ 472 .
)
1- MR. S1ESS: .That's where the economics comes in. 2~ You might save a. walk down, but you-may also have to change 3 your design. 4 MR. KENNEALLY: A position on a. single design 5 basis earthquake is also being done in Germany. They have , 6- an inspection level earthquake for shutdown. set.at .4 of the ] 7 equivalent of our SSE. The level was set so that the - 8 corresponding vibration would not induce stresses exceeding 9 the allowable limits that were originally required for the , 10 OBE. 11 MR. SIESS: If you think metric you wouldn't come 12 up with one-third, you would come up with four tenths. , J.' 13 [ Slides.] ; 14 MR. KENNEALLY: In support,of the staff's position 15 of one-third, we have considerable amount of ongoing 16 studies, one being in the area of risk assessment or margin 17 study, where-there was a survey of the architect / engineers 18 and designers and they have-identified components that might 19 be affected by a change in the OBE; concrete' penetrations, 20 piping and supports. There was also an evaluation of the - l 21 existing seismic PRA's in general categories of components 22 which were potential seismic risk contributors were i 23 identified. The motor control centers and the usual actors, 24 the pumps and valves. l 2! A subset of the above components were selected and
?
l L y 473 ' i .T 1 they were reviewed, and after the review in general the OBE 2 tended to govern that the earthquake contr'ibutions to the I 3 total loads were small and the SSE tended to govern if the , j 4 contributions were large. This is still an ongoing study ) i i 5- and we are still doing more in that particular area, and 6- will be up through the public comment period in obtaining a 7 lot more information that would support the staff's 8 . position. 9 MR. SIESS: This study is interesting, because it , l l l 10 is based on the assumption -- and I think.we all tend to ! i l 11 make it -- that what we were doing before must have been ; 1 12 right. If the OBE governed and now we change it and it
- 13 won't govern, we are going to be wrong. It is really a lot j 14 more logical, I think, if you sat 'down that what we were 15 doing before was wrong. If it turns out that there are some 16 places where it won't be quite as much margin with an SSE 17 design as they had before than OBE design, then so be it.
18 MR. KENNEALLY: We have also had a survey of the 19 utilities and design. The impact of the OBE depended to a 20 large extent on the philosophy that the design is used -- 21 some of them used different damping to generate the OBE and l 22 the SSE spectrum. They did two analyses. Another one may
- 23 have used OBE damping for both the OBE and an SSE, and it 24 allowed them to do scaling without doing a separate 25 analysis. There were different amounts of short cuts that
'i a
4 f
j -
;< 474
)
'l- were taken.in their; analysis.- That's why, when you talk to l 2 people about what control-and the importance of it,;you ,
3 really have to look back a considerable amount to find out- l 4 all that went into their work. l 5 That was' pretty much a summary of what was:in the
- i
'6 regulation. Now, a little. bit about how the staff would +
7 attempt to apply it through regulatory guides and the like. - l i 8 There would be recommended in the regulatory
.l 9 guides some pre-earthquakeLplanning and the like. First of 10 all, operable seismic' instrumentation was something that 11' would be strongly recommended. In the past the operability 12 of it has beenl questionable. It seems to be out for 13 maintenance or~just inoperable at the most-inopportune time.
14 In addition, there would be-a sele'ction and inspection of l r 15 structures and equipment as a baseline. I will be getting
'i 16 into this a lot more in a few minutes. ,
i 17 Following an earthquake"there are procedures i i 18 identified for a. walk down, evaluation of ground motion , 19 records determinate of the OBE exceedance or not, and then 20 going into other evaluation modes as appropriate. - 21 (Slides.) l 23 MR. KENNEALLY: The first draft guide that would j 23 help implement this is the second proposed revision to 24 regulatory guide 1.12, nuclear power plant instrumentation l 25 for earthquakes. A major change in this one over the k, i ' t f
l. L -l { 475 l' existing one'is that the staff is recommending solid-state,. l
'2 state of the art, digital time history accelerometers. The f3 instrumentation sensors would be placed at freaffields, . ,
! r 4' foundations, and-elevations within the buildings.. l l 4 5 Other maintenance and requirements associated with l l
'6 - it are very similar'to what is in the existing standards, ,
7 also very similar to what is within the ANS 2.2 standard f t 8 which'the staff is_ currently not endorsing. 9 MR. .SIESS: You are no longer calling for j 10 instrumentation on equipment. i 11' MR. KENNEALLY: No,.we are not. We are l
~
l
-13 recommending that the instrumentation be put at the slab 13 level,_ foundation level in free field rather than equipment' >
l 14 supports or whatever. I have talked a considerable amount i 15 with the peoplo that have experienced the earthquakes like 16 the. people at the Perry plant and.also the ANS working 17 . group, and it seems that when one starts.to resolve the -- r 18 evaluate'the data from instrumentation that has been placed 19 on supports,.you have operating loads combined with l 20 earthquake loads and sorting it out is a horrendous problem. i 21 MR. SIESS: You have background -- 22 MR. KENNEALLY: Yes, sir. 33 MR. SIESS: If water is flowing through the pipes 34 that must be doing something to you.
.25 MR. KENNEALLY: That's doing something to it.
,-wm.,.
_ w W - -
476 f,
~
1 'Some of.the current instrumentation, the peak accellograph, 2 it takes three or four outages before they can even 3 determine an appropriate place to put it. EPRI had a 4 workshop on seismic instrumentation in 1988, I believe it 5 was, and they are doing some additional studies on that, as , 6 the type of instrumentation and its location. We are 7 keeping abreast of any work that they are doing in that area - 8 as well. , 9 Where the staff is really at a shortcoming in its 10 guidance is, what do you do after an earthquake. Industry 11 had assembled a considerable amount of effort on this to 12 prepare a few documents that would identify how a plant 13 shculd be evaluated after an earthquake, suggested criterion' 14 that would determine if you-have e'xceeded an OBE. This has 15 been submitted to the staff for operating plants and is 16 undergoing a review, and we are also using the concepts from 7 17 it for the future reactors. 18 one particular_ document on that is EPRI NP-6695, 19 guidelines for nuclear plant response to an earthquake. The 20 staff is endorsing that particular one. It recommends a . 21 baselining of your plant before the earthquake so that you 22 have an idea of the physical condition of it, so that after 23 an earthquake you can go through and do a very quick 24 evaluation to determine the severity of the earthquake. 25 This is also coupled with an evaluation to the
{ 477
-1 instrumentation data. That is based in part on two EPRI 2 reports. One of them, EPRI NO-5930 and a second study which I
3 is EPRI report TR-100082, an update of the 5930 report. 1 4 What is being endorsed by the staff is an 5 evaluation of OBE exceedance based on a check of the j 6 response spectrum, a new method, a new check which was a i l 7 cumulative absolute' velocity. This is an empirical value 8 that is obtained by taking the time history and breaking it 9 into one second increments. If any part of that one second 10 has an acceleration into excess of .025 G, you would sum all 11 of the accelerations within that one second, go through the
--' l entire earthquake record.
12 If the total sum for the duration 13 of the earthquake having accelerations in excess of .025 j 14 were greater than .16 G seconds, that would be a parameter 15 that would indicate you have exceeded the OBE. 16 MR. SIESS: Is that the product of a single l
)
17 individual or Committee? 18 MR. KENNEALLY: That's a product of a Committee.
- 1. 9 The EPRI work was done in part by Dr. Kennedy, Jack Benjamin 20 Associates through John Reed. The report up here, 6695, was 21 done with the efforts of Dr. Stevenson, Kennedy, MPR in 22 Washington.
23 MR. SIESS: That one, I have. Do you have to do a 24 walkdown -- you are going to get to that, aren't you? 25 That's the next item. j i
i l(, 478 /m 1 MR. KENh?'71Yi Right. The staff is recommending 2 a shutdown criteria c. 'xceedance to the OBE which would be l 3 you have exceeded a respc se spectrum and the CAV, the 4 cumulative absolute velocit), or after a felt earthquake the 5 operators will perform a wa.'.kJown. This is a walk down of , 6 equipment and structures that are normally looked at during
-7 shift operation changes. The status of the plant is .
8 identified. They would discover if there were any changes 9 within that. It consists-of safety related structures. Non-10 safety related structures would serve as a barometer of the 31 severity of the earthquake. 13 MR. SIESS: What triggers this? 13 MR. KENNEALLY: This particular walk down is -- 14 MR. SIESS: You said felt earthquake. 15- MR. KENNEALLY: A felt earthquake, which is either 16 activation of the instruments or a consensus of the 17 operators within the control room. ' 18 MR. SIESS: If your instruments actually operate, 19 you should do some of this. ; 1 20 MR. KENNEALLY: That's correct. - 21 MR. SIESS: Although the trigger that you have is 22 well below the -- 23 MR. KENNEALLY: The triggers right now are 24 typically in the range-of .01 to .02 G. 25 MR. SIESS: Why wouldn't you want to say set the l
'/ 479 1 trigger up a lot closer to the OBE?
2 MR. KENNEALLY: If the trigger itself were set up 3 a lot closer to the OBE it would only actuate the-sensors i 4 during the period where the trigger were exceed and you
- 5 would miss a lot of your time history. 1 6 MR. SIESS: I mean, we think now that unless you l
7 have something greater than the OBE you are not likely to 8 have any problem, right? We have a trigger set down at .02 9 G or something. 30 MR. KERR: What is the trigger for? Is it to turn 11 on the -- R 12 MR. SIESS: Turn on the instrumentation. l l 13 MR. WARD: He says they want to record the whole l i ; 14 thing, so they want to catch it early. l l 15 MR. SIESS: If you have the instrumentation they ! i I 16 call for i, Reg Guide 10.16, it will' store the five seconds 3 17 constantly. When you trigger it, it will have the five 18 seconds before the trigger and quite a bit of time 19 afterwards.
?O My point is, we set the trigger low simply to j
1 accumulate data, right?
22 MR. KENNEALLY: Yes.
?3 MR. SIESS: But, do we really have to worry about 24 exceeding the OBE or worry about damage simply because we
?S have hit the trigger level?
1 l
fs> 480 1 MR. KENNEALLY: The hitting of the trigger level 2 would start a walk down. 3 MR. SIESS: Why? Why, if it's that low? Do we 4 really honestly feel as structural people and mechanical 5 people -- what is the trigger level? , 6 MR. KENNEALLY: It's .02 G. 7 MR. SIESS: It's .02. Does anybody really think . 8 that a .02, .03 or .04 is going to damage a plant? 9 MR. KENNEALLY: No. In fact, the walk down 10 associated with that is not a significant deal like T 6.ift 11 . change, and would be done very rapidly and you would not - la involve a considerable amount of time. i 13 MR. SIESS: I think the first thing I would do is 14 look at the instrument and see whether it was kicked. I get ! 15 the point. Go ahead. 16 MR. KENNEALLY: Specifically, to lay to rest the l 17 indicator, the operators will know there was an earthquake, ; 10 I have been told at .02 G. You will have some concern about 19 what might have transpired -- 20 HR. SHEWMON: Without looking at their instruments . 21 they will know it. < 22 MR. SIESS: It would be felt. 23 MR. KENNEALLY: It would be felt. 24 MR. SHEWMON: Felt by them, not by their ! 25 instruments. MR. KENNEALLY: At .02 G the instrument
)
g 481
)
l' should have been activated as well. ! l 2 MR. SHEWMON: I don't care about the instruments, 3 I am try'ing to understand what you meant when you said they 4 will know. 5 MR. KENNEALLY: It will be felt by them.
~
i 6 MR. SHEWMON: Thank you. 7 MR. WILKINS: If they were seated or they were 8 standing, or in either case. 9 MR. KENNEALLY: Either case. 10 MR. S I'E S S : There will be some of each in most 11 control rooms. l 12 MR. WILKINS: The reason I asked that question is 13 that I went through an earthquake in Idaho and I didn't know l 14 it, and I got-inside and the building-was being evacuated. . l 15 I was driving an automobile with four nice cushions. 16 MR. SHAO: You must have earthquake in San Diego
- 17 when you lived there.
18 MR. WILKINS: That was my first earthquake Larry, 19 and I didn't know what the hell was going on. 20 MR. SIESS: How long after a trigger would it take 21 to get the instrumentation reading to know whether -- do you l 32 have to send them off? 23 MR. KENNEALLY: No. These are recommending that l 24 you would have on site processing and evaluation data, and ! ! l l 25 it's being estimated within four hours. j 1 7-d l s l I l 1
)
1
482 , -l'
) ,
1 MR. SIESS: In the meantine, you just go ahead - 2 about your business. 3 MR. KENNEALLY: If you have tripped off line -- 4 the first thing you do is stabilizing the plant. You are 5 responding to -- , 6 MR. SIESS: If something shook you know it. You 7 have done the shift change walk down and somebody is looking - 8 at the data. 9 MR. KENNEALLY: Yes. 10 MR. SIESS: If the data don't come out over the 11 OBE, you go on about your business. 12 MR. KENNEALLY: Your shift walk down comes clear, j 13 you are back. 14 MR. SHEWMON: Can-they keep operating during this? 15 MR. SIESS: Oh, yes, unless you tripped off.. 4 16 MR. KENNEALLY: Yes. If you have tripped off line. ) 17 --your turbine or some other reason. 18 MR. SIESS: If they did trip off, would you -- it 19 probably shouldn't start up again until they have looked at 20 the record. . 21 MR. KENNEALLY: That's correct.
~
22 MR. SIESS: That's covered, which is logical. 23 MR. KENNEALLY: The last draft reg guide, 10.18, 24 the restarting of a nuclear plant that was shut down by a 25 seismic event also endorses portions of the EPRI 6695 4,
l l i l i l 483 1 r. 1 report.- That document provides the_ guidance for performing tw d e ak ng pl pr or to he e tar o a p an i 4 In addition to that, there is guidelines for performing long 5 term evaluations to determine the affects of the earthquake l 6 to the plant. This would be evaluating ' the data from the n 7 other recorders and doing other type of analysis to see the 1 8 correlations between the recorded data and the mathematical 9 models and the like. 10 Except in the most severe of all of the l 11 earthquakes that is identified in that report in EPRI 3 12 damage scale, this last one would be done after operation. > j 13 That concludes what I wanted to discuss relative i 14 to the engineering portion of it. The staff feels that wo 15 have filled in a lot of the voids on just how to respond l 16 after an earthquake. We have benefitted tremendously from 17 some of the effort that industry has gone through on this 18 and had a lot of dialogue with them, to understand what they 19 have done. We feel that we have come a long way towards 20 making the operation in the future, taking some of the l 21 questions and add it to the stability of the licensing.
- 22 MR. SIESS
- Thank you. I would like to just go a 23 little bit further. Appendix S has a section headed l
j 24 required plant shutdown and it's very brief. It says a l 25 vibratory ground motion exceeding that of the OBE or t
- ... .. . . -. - - . - - - - - - - - . _ . - - _ - . _ _ ~ _ - .- -. .-- ..
R j - 484 l 1 significant plant damage occurs, shutdown is required. 2 Prior to resuming operation the licensee shall demonstrate ; 3 to the Commission that no functional damage'has occurred to . 4 those features necessary for continued operation without ! 5- undue risk to the health and safety of the public. , 6 That's all the regulations will say. The rest of 7 it is in the reg guides. - 8 MR. KENNEALLY: Yes. t 9 MR. SIESS: Also it says it's a required seismic 10 instrumentation, and that's in the reg guide. It said it l 11 shall be suitable. The reg guide defines what is suitable, la and I just wanted to make it clear. Even so, this is more t i , 13 in the reg guides than we have ever had before. Now, we 14 have it in a different place and have got the guidance that 15 I am sure EPRI wanted. They did a lot of work on this and 16 staff has used it. 17 Andy is going to go back and what he will be 18 covering is what is in reg guide draft Guide 10.15, which is i 19 invoked by essentially a single sentence in Appendix B. 20 MR. MURPHY: Yes, sir, that's basically it. . i 21 MR. SIESS: That says in selecting the SSE ground 22 motion both probabilistic and deterministic approaches shall , 23 be-used. 34 MR. MURPHY: That's right. 25 MR. SIESS: I think that's the exact quote. If it l I l I
- - ~ ,
i j. i 485
~1 isn't -- that is all it says. It says using both the i 2 probabilistic and deterministic approaches, and he is going 3 to tell us what that means. It is all covered in a draft 4 guide with a few appendixes of'its own.
5 [ Slides.) 6' MR. MURPHY: We will start with viewgraph number ;
-7 three, which is about the seventh or eighth.viewgraph back 8_ in your package. This viewgraph is a schematic of what we ;
9 proposed by way of.what we are calling the dual approach, ! 10 the-two requirements. We are requiring a deterministic path i 11- which is very close to what we have been requiring through 12 Appendix A,-the current regulation. We use the 13 deterministic path as our primary vehicle for developing - 14 site specific information about the proposed nuclear power 13 plant site. I 16 The second path that we are requiring we called a 17 probabilistic analysis path. We have run into a number of I 18 difficulties using that term, probabilistic analysis, ano we 19 recognize that difficulty. What we are basically proposing i 20 here is that the utility carry out a --- 21 MR. SIESS: Excuse me. Andy, the Appendix B refers 22- to an approach, not an analysis. Are they the same? 23 MR. MURPHY: Basically, I will be using them j 24 interchangeably. Technically, there is a little bit of i l 25 difference between them, yes, sir. l r l
\
.- ~. . . . - - - . .
3 486 A 1 Upon selecting a site for'a new facility the 's -) 2 applicant would go to the EPRI or to the National Computer 3 -Code. Center or to a consultant, and they would get a copy of - 4 either one of these two codes and they would operate them. 1 5 They would put in the coordinates of the site and a little , 6 -bit of geotechnical information about that site. They would 7 turn the crank, and approximately one hour later they would = 8 have the seismic hazard curves for that site.- ]
.9 We are asking for some additional calculations to l
10 be carried out. Basically those calculations would be- ; 1 11 through an accepted methodology that has been reviewed j 12 already by the NRC. ', 4 13 MR. MURPHY: Excuse me, Andy. Let me back up just 14 a minute. i 15 MR. MURPHY: Okay. 16 MR. SIESS: The top box, the necessary geological 17 seismological and geophysical investigation formerly were in 18 Appendix A. 19 MR. MURPHY: That's correct. , 20 MR. SIESS: They are now in Reg Guide 10.15. , 21 MR. MURPHY: The guidance on how to carry these 22 out is in 10.15. The statement that you do have to, carry 23 out -- 34 MR. SIESS: I want to make it clear. That is one 25 of the items that has been taken from the omnibus appendix h [. 1 - -, . .. - . - - _ . _ .
I 487 1 and put in a reg guide, and that had some of the most 3 obsolete stuff in it. , 3 MR. MURPHY: That's correct. 4 MR. WILKINS: You were about to say though that 5 the rule still says you have to do them. 6 MR. MURPHY: That's correct. 7 MR. SIESS: You have to do an investigation. 8 MR. MURPHY: You have to carry out these 9 investigations. That is still required. That's the what -- 10 that's what you have to do and the guides tell you how to 11 carry it out. 12 MR. WILKINS: Give you one acceptable means for I 13 carrying them out. 14 MR. MURPHY: You are correct. 15 MR. WARD: The tool for doing the first box where 16 is that tool? What is that tool? 17 MR. MURPHY: That tool is, ona part of it is the 18 EPRI methodology which has been documenh e arid submitted to 19 the NRC. That is -- 20 MR. WARD: This is a handbook or something? 21 MR. MURPHY: This would be a handbook and a series 22 of reports documenting a code, a computer code, and a series 23 of databases that are used in that computer code. The same 24 thing exists for -- 25 MR. SIESS: Seismic hazard assessment.
- . - -. . - _ = . .
- 488
- 8. '
\
1 MR. MURPHY: The same thing exists for Livermore. 2 It is a code with a set of databases that go along with it. 3 4 MR. WARD: The utility or the applicant, whoever 5 this is, can use either one. . 6 MR. MURPHY: They can use either one. But there 7 are provisions that once they have decided to use one they 8 have to carry their analysis through using that one all the 9 time. 10 MR. WILKINS: What is to prevent them from using
'11 both before they talk to you? '
12 MR. MURPHY: Nothing. That's fantastic. If .. 13 that's what they want to do and they can do it and pick the 14 best answer for them, so be-it. , 15 MR. SIESS: The thing is, don't -- 16 MR. MURPHY: Based on our discussion on Wednesday l 17 with the Subcommittee, we would revise this figure somewhat. 18 Sasically, all I want to do is show you that we do have two 19 paths that we are following and that we are, at this time,
?O we are requiring the applicant to follow both paths. .
21 MR. WILKINS: I wasn't at the meeting Wednesday. 22 What is earthquakes MS and DS.
?3 MR. MURPHY: Earthquakes -- I will talk to you
4 about controlling earthquakes in a minute in the next l
35 viewgraph. MS and DS are magnitudes and distances. l I j
l il 489 r. 1 MR. WILKINS: Thank you. 2 MR. SIESS: This is all -- the boxes are going to 3 be enlarged. 4 MR. WILKINS: All right. 5 MR. MURPHY: The boxes are going to be enlarged 6 and the contents are going to be' improved. 7 MR. WARD: I am still puzzling over the-freedom to ) l l 8 use either -- 9 MR. SIESS: You will understand. You have to wait i
- D just a minute, i
11 MR. SHEWMON: Before you leave that, when you get l 12 down to the bottom you are going to tell us how they can ; 13 compare them and take the best answer or they take the worst
- 4 answer and add them and divide by 'wo t --
;5 MR. MURPHY: I will tell you that at the end.
16 Basically at the bottom they take the best answer. 17 MR. SHEWMON: Next to the bottom box. 19 MR. SIESS: Comparing the two. That's the $64.00 question that he damn well better answer.
'O MR. MURPHY: Correctly.
t' [ Slides.]
~2 MR. MURPHY: Next, I will try to tell you what
?3 controlling earthquakes are all about.
l ?4 MR. SIESS: This is deterministic. I 25 MR. MURPHY: This is easiest to think about this l.
. . . . - - . _ - _ .. -~ -- . - _ - . - _-- . . .
(: , 490-5 ') 1 in the deterministic -- 2 MR. SIESS: This is the right hand side. !
- 3 MR. . MURPHY: This is the right hand side. This is t
4 basically what we would.be doing today. You have a nuclear 5
~
power plant site here. You have it located in a background , 6 zone or a province?in and of itself. Here, we have shown .i 7 three other seismic source zones around-it. The standard - 1 l 8 review plan.section tells you how to figure out the ground ! 9 motion for-each of these sources. You have, in effect, 10 moved what we are now calling the deterministic source i 11 earthquake to the vicinity of.the power plant'and 12 calculating a spectra associated with earthquake and.its i
'{ ,
13 ground motions.
- 14. You will do the same sort of thing here for each l 15 of t.hese source zones. You will take and move the D .
16 deterministic source earthquake as we are calling it to.the , closest approach on the boundary, you will let that ground ! 17 18 motion attenuate to the site and then you will calculate a ; 19 spectra for that ground motion. ; 20 What we are showing here -- i i al MR. SIESS: That is the way we have done it in the
~
'22 past.
23 MR. MURPHY: That's the way we'have done it in the f 24 past.
-35 MR. SIESS: The spectra are set in the standard l 5
7 i
l 491 (f. 1 review plan though, not anywhere in the regulations at all. 2 MR. MURPHY: That's correct. What we are saying - 3 - I 4 MR. SIESS: And won't be. 5 MR. MURPHY: Right. What we are saying is that 6 these two spectra are from what we are now calling the 7 controlling earthquakes. A very close envelope or actually 8 these two spectra can be used for.the design spectra for 9 that facility. 10 MR. WILKINS: Is that an envelope?
'll MR. MURPHY: It is shown as an envelope there. We 12 are not requiring them to use an envelope. If they want to 13 use the two spectra individually they are free to do so.
14 Technically for a structural engin~eer that would be a very 15 difficult task. 16 MR. WILKINS: Well, let me ask you this. If you 17 draw this thing with a log scale over there you could add 18 those two, and you get pretty close to that envelope -- the 19 curve that I call an envelope. On the other hand if that's 20 a linear scale, then what I said is false. I can't tell by 21 looking at it. 22 MR. SIESS: You wouldn't. Somewhere at that 23 period you would use whichever -- 24 MR. WILKINS: Whichever one was controlled. 25 MR. MURPHY: Whichever one was controlled. i
[- 492 fg 1 MR. SIESS: If you didn't want to draw that ( >) 2 controlling envelope you would probably'not make any 3 difference, because if you have to do peak grounding you 4 really couldn't miss anything in that range,anyway. 5 MR. MURPHY: Yes. , 6 MR. SIESS: You don't do it that way anyway. 7 MR. MURPHY: After defining that term, controlling . 8 earthquakes, we want to give you a little background on what 9 happens with the two probabilistic hazard studies and how
'10 they actually --
11 -MR. SIESS: Now, we are on the left-hand side. 12 MR. MURPHY: Now, we are on the left-hand side. 13 MR. SIESS: The right hand side is -- 14 ' MR. MURPHY: Is determin'istic. The left is 15 probabilistic. 16 MR. SIESS: Once you get an SSE enveloping those 17 controlling earthquakes -- 18 MR. WARD: That's why it's an SSE GM, because it's 19 just -- 20 MR. MURPHY: Yes. Just to make clear things, we - 21 will be using SSE and not SSE GM. It is the safe shutdown 32 earthquake ground motion, but we will stay with the old 23 nomenclature of SSE. Here, we have~a power plant site. We
.24- have postulated in this case two seismic source zones around 25 it. You use the seismicity; i.e., the seismic source zones.
T
q- 493 4, 1 We have another data set which is the magnitude recurrence 2 models for each of these individually. There's one for this 3 source zone and there's another one for this source zone. 4 We also then have a series of ground motion models 5 that tell us how to attenuate the ground motion from this 6 source zone to the site or regionally. Then, we take these l 7 two data sets and in effect integrate them in the Livermore 8 case through a Monte Carlo technique, and we arrive at a 9 seismic hazard curve which is simply a plot of probability l 10 of exceedance versus acceleration. 11 The handout has this corrected. The viewgraph 12 does not. 13 MR. SIESS: To two one-tenth G's. 14 MR. MURPHY: The two one-tenth G's. The thing to 15 understand now, the basic point is that the data sets that 16 we are talking about, A, B and C exist today. They have l 17 been reviewed by the staff for both EPRI and Livermore, and 18 what we.are saying at this time is that we are freezing them 19 as they are today or as the regulation goes out. This data 20 set is fixed, the computer codes are fixed and are available 21 to the applicant. 22 MR. WILKINS: What you are telling me is something 23 which I guess I would have believed, but never quite knew l i 24 before. 25 MR. SIESS: I can't hear you. i l
i { 494 ,, 1 1 MR. WILKINS: I'm sorry. What you are telling me l 4 2 is something that I guess I would have believed but I didn't 3- realize I knew; that, the entire United States is mapped and 4 people know where these homogeneous source zones are and 1 5 they know the frequency and size of the events that can , 6 occur for each of those, and they know how the waves are 7 attenuated as they go through the very heterogeneous earth . , 8 that lies between any site and any source. 9 MR. MURPHY: No is probably too strong a verb. 10 They have an opinion about -- in this case for the Livermore l 11 methodology there are 11 experts who have at least 11 models 12 for the Eastr~.n United States, some overlap a bit and some ! . 13 don't overlap hardly at all. Some of those experts have i i 14 even provided second opinions about these source zones on 15 their own. 16 They have provided information, each again 17 individually about these source zones, with distributions _on ) 18 the recurrence information. We had for Livermore, we had 11 l 19 experts providing these two data sets and for ground motion l l 20 we had an additional five experts providing the ground . 1 21 motion models. I have forgotten the number of ground motion i 22 models, but I would say it was on the order of ten. They ) 1 23 were basically weighted according to the expert opinion. l 24 MR. WILKINS: That was a Delphi technique of some 25 sort? i
_. . - --- . .. - .- . . ~_ l l f. 495 l. t ! 1 MR. MURPHY: Yes, sir. 2 MR. SIESS: Ernest, what I think you have to 3 realize here is that that project had objectives somewhat 4 similar to NUREG 1150 on reactors, and.that it was designed 5 in part to come up with a probabilistic estimate of ground 6 motion at any site, and partly to bring out all the 7 uncertainties in that estimate. 8 MR. WILKINS: Yes. 9 MR. SIESS: A lot of it was done with expert 10 opinion, and no two of them ever agreed to anything. , l 11 There's another set that comes out of the EPRI study, and ; i la for a given plant the probability of exceeding a design , 13 basis earthquake for that plant differs by at least two 14 orders of magnitude for the two me'thods, one to two. 15 MR. WILKINS: A factor of one or two, or orders of 16 magnitude? l 17 MR. MURPHY: It would probably be better 18 characterized from a factor of two to two orders of 19 magnitude. Some of them are reasonably close and some of 20 them are at least two orders of magnitude -- 21 MR. WILKINS: Depending on the sites. 22 MR. SIESS: That's taking the existing plants and 23 a specific SSE, what they were designed for. If you use the 24 two methods to come up with an SSE you then have to start 25 with a specified probability of exceedance, which is what he
_ __ . . . __ _ . ___ . ~ . 4 1 '( 496
/1 '
1 is going to do next I think. 2 MR. MURPHY: The th'ing to-understand here is that 3 for each of the' nuclear power plant sites in the Eastern 4 United States there is one of these seismic hazard curves 5 available, one from the EPRI methodology and one from the , 6 Livermore methodology. With a few exceptions on the EPRI 7 some of the utilities opted not to buy into that program, - 8- , and I don't think those have been calculated yet for the 9 EPRI methodology. 10 But there is a canned set of seismic hazard curves 11 available-basica,lly for all the plants in the Eastern United ) 12~ States. ,
)
13 MR. SHAO: East of the Rockies. 14 MR. MURPHY: Yes, East o'f the Rockies is what we 15 are talking about by the Eastern United States. Right now, 16 I will show you a figure that depends upon a tabulation of 17 those seismic hazard curves for the plants in the Eastern 18 United States. ) l 19 MR. WARD: Is that just sketched in on the j 4 ! 20 previous one? That's a pretty fierce curve there, isn't it? ! 21 y MR. SIESS: Go back to that figure, j B 22' MR. WARD: I guess I don't even know what the Log _23 P is. , 34 MR. SIESS: That's a log -- 25 MR. MURPHY: Log of probability. Log of 1
.. ~.
.... . - ..., . .. - ... - - - - - ---..- ~ . . _ _ ...-. . .-.-
l h f. 497 i i probability' exceeding the SSE for that facility. l 2 MR. WARD:~ For year -- ; i L 3 MR. SIESS: We have to switch gears a minute. We ! 1 l 4 were talking -- he just said we have this curve for every i 5 , plant.- Forget about that, unless you are going to put l 6 another plant at that site. For a new. site you would go ; 7 through with'this procedure and come out with a curve like ;
?
8 that for the site. i l 9 MR. MURPHY: That's right. 10 MR. SIESS: We have not yet put a plant there. 11 Before we put a plant there, we have to decide on what 12 earthquake we are going to design it for. ;
- 13 M
.'R. WARD: Okay.
14 MR. SIESS: I could pick'any earthquake I want. I 15 can pick one of a probability that exceeds ten to the minus 16 one, ten to the minus four, ten to the minus six. What am I ;
'17 going to use. Now, he's going to answer that question.
18 ' MR. WARD: Could I still stay with this a minute? l 19 MR. MURPHY: Sure. 20 MR.; WARD: It says Log P. You really just mean i l 21 the probability, don't you? I mean, ten to the minus two is
- 22. --
23 MR. MURPHY: Yes. It's the log scale. 24 MR. WARD: Okay. Now, if that means that a return 25 period of 1,000 years the number is over 1-G, is that really
.;. _.-. _ . _ . . , ~ . _ _ _ _ _ _ _ _ - . _ _ . . _ . _.. __ ,_ _ __ _ _ . . . _ .
{) s' i
'498
{~- .
- \
1 what -- is'that a realistic curve or is.that-just drawn in. 2 there? 3 MR. MURPHY: This is just schematic. l j 4 MR. SIESS: There's nothing realistic about it. ] 1 5 MR. MURPHY: As we indicated down here, we-have an - 1 6 - escape clause. The one'that's up here on the board -- not
-7 . the one that.you'have because the error was not corrected -- .
I 8- we have-two points at -- l 9 MR. WARD: I know. I thought it might be. J 10 -- realistic. 11 MR. MURPHY: It's not. , 12' MR. SIESS: It wouldn't be realistic for any site
- _ 13 in the U.S. -I am not even sure it would be realistic'--
14 MR. MURPHY: It's just a'. figure that,was pui l 15 together to show what the process was. ! 16 MR. WARD: Okay, I don't want to slow you-down. 17 (Slides.) f l
.18 MR. MURPHY: In this figure we show a cumulative 19 distribution-of the plants for median Livermore hazard 20 estimates for the Eastern United States. Horizontal access t 21 we. plotted, the annual probability of exceeding the SSE,-and l 23 here we' simply have accumulated the distribution. The plant (
23 - number and the plant name'we plotted around there, and all 24 we are.doing is accumulating the -- ; 4 25 MR. SIESS: How many plants on that plot, how many , k. F i
,w- , . - - - - - - , -
l i j e 499 i 1 sites, 89? t 2 MR. MURPHY: Sixty-nine for right now. We have
.3 accumulated 69 plant. sites along that axis. What we have l 4 done, based upon the Commission policy statement that the j i
5 current nuclear power plants are safe enough, we have gone 6 in here and selected a 50 percent level, drawn a line and t l 7 come down here and read off an annual probability of l' 8 exceedance. We have termed this our target annual l j 9 precability exceedance that we are going to have the next 10 generation of nuclear power plants meet. 1
- 11 We have done this for both. Here we have the l l 12 Livermore case, and we have done it also for the EPRI case. i I
!; 13 MR. WILKINS: This will be a situation where, in 14- the future, all the plants will be~better than average. l 15 MR. MURPHY: Yes. 16 MR. SIESS: No, half of them.. I ! l 17 MR. MURPHY: All the new:ones will be better -- I 18 MR. WILKINS: Will be better than this median, 19 anyway. l 20 MR. MURPHY: Right. l 21 MR. SIESS: The ten to the minus four is for I i 22 Livermore. l l- 23 MR. MURPHY: The ten to the minus four -- l l
- 24 MR. SIESS
- The same thing for EPRI, it wouldn't 25 be ten to the minus four.
l .i s l
- l. .-. .-. . - _ l
i 1 f 500 !
).
.1 _ MR. MURPHY: No, it. definitely.would not. If we ,
i 2- .did-the'same thing for EPRI, and we did, your number of
'3 three to the sinus five. That is the target probability. !
[ 4- MR. WARD: The EPRI curve is way over -- 5- M R . -- S I E S S : - That's what I said. You have to stick , , 6 to whatever one you pick. ! 7 MR. WARD: Okay. - 4 E
.8 MR. MURPHY: Right. So that, if you are going to .
J 9 .use EPRI the target probability.you use is three to the i 10 minus five. If you are-going to use Livermore, the target 11 probability you use is one E to the minus: four. 12 MR. SIESS: You'are going to pick SSE for your i 13 site that will have a probability of exceedance smaller than -] 14 that.of half the plants that have"been built. l l 15 MR. WARD: The way you got this number, you said 16 based on the Commission's policy statement.- I can't help 1 17 smiling that existing plants are safe enough. You went into I 18 this as .5. 19 MR. MURPHY: That's right. 20^ MR. SIESS: Good question. 21 MR. WILKINS: In theory, you could go into it at 22 one. 23 MR. WARD: That's just -- 24 MR. SIESS: You could go into it at one. 1 25 MR. WARD: Or, you could say that's sort of a I
- . . - - . -~ . . - . . -. --..- --.. -.. -.- .- . . . . . . - - . - . . . . ~
L/ 501 , 1 senseless way.to.make this evaluation, that's another l 2 - choice.. 3 MR. SIESS: Yes. l 4 MR.'WILKINS: It's arbitrary, isn't it. 5 MR. SIESS: If you are going to -- 6 MR. WARD: I would say it is. 7 MR. SIESS: If you are going to.use a 8 probabilistic basis you have to start with a desired l i 9 probabiljty. l 10 MR. WARD: Yes, but I would rather hear what some 11 expert said about it or something -- 12 MR. SIESS: Than the Commission? D 13 MR. WARD: Yes. l 14 MR. SIESS: We will poll the ACRS.
'15 MR. KERR: Listen to the expert.
I 16 MR. SIESS: Look, Dave, you could pick 85 17 percentile and you would come out at -- ] 1 f 18 MR. WARD: What's the difference -- l 19 MR. SIESS: -- at three times ten to the minus 20 four. l 21 MR. MURPHY: We have done a few preliminary 22 sensitivity studies where we picked the 25 percent and the 23 75 percent, and then when you carry that through to the l' 24 ground motion that you get out at the other end the i 25 differences are reasonably close to insignificant. ; h i i
l
'f 502 N' ,,,)
1 MR.-WARD: I am being picky, but- I guess I would
.2 rather have you say you have plotted these and you have 3 stared at that curve and you picked this number rather than 4 --
1 5' MR. SIESS: 'This is the first time we heard him , 6 attribute it to the Commissioners. l 7 MR. WARD: Okay.- - l 8 MR. KRESS: You want him to run a PRA Dave, and i l 9 see what criteria gets him below'the safety goals?
-10 MR. KERR: Half of the Commissioners wanted to be 11 bigger than the median and half wanted. to be smal'.er, so --
12 MR. KRESS: That's a logical --
, - 13 MR..SHAO: One thing you should notice --
14 .MR. KRESS: That doesn't'please anybody. j l 15 MR. SHAO: These are all the plants,;all the new !
- 16. plants for below --
17 MR. SIESS: The new plants ~are-up a level and the 18 lower ones down~here -- 19 MR. SHAO:- The old plants, yes. 20 MR. SIESS: -- are some one -- - l 21 MR. SHAO: Above average.
- 23 MR. .SIESS: Let's face it, we are not talking
. 23 about-the probability of core melt. We are not' talking
- 24 about'the probability of damage. All we are talking about 25' is the probability of the SSE occurring. Now, we design a
- 9 i
9 k 503 . 1 plant for the SSE with margins and all this other stuff -- 2 MR. WARD: We haven't talked about that yet. 3 MR. MURPHY: We aren't going to. 4 MR. SIESS: This is the probabilistic approach. 5 You are going to start somewhere with a probability. What 6 they haven't gone through -- I am not sure the design of the 7 plant will look a whole lot different whether you took this 8 at the median or the 85 percentile or somewhere else, you 9 see. Go ahead. 10 MR. MURPHY: The point that -- 11 MR. SIESS: We understand it.
- 12 MR. MURPHY
- Okay. We have that. We whipped it 13 out for EPRI as well as Livermore. ,
4 14 (Slides.] 15 MR. MURPHY: Here, we are going to run through -- s 16 as I go through the rest of this discussion, we are going to 17 take Vogtle as an example of how this works. One of the 18 other. things that you get out of the Livermore or EPRI 19 programs is a spectral velocity'versus annual probability of 20 exceedance -- al MR. SIESS: Wait a minute, Andy, I have to correct 22 you. This is not simply an example of how it works. It is 23 going to be two things. The deterministic analysis has 24 already been done for Vogtle. 25 MR. MURPHY: That's right.
. .__ _ . .. . . . _ . _ . . _ . _ _ _ _ _ . _ _ _ . . . _ . . _ . _ . . . . _ . _ - . _ . _ .__ m. ..... _ _ _ .,
W .(- 504 ,N' 3.' ) r' ,
) .
1 l' MR. SIESS: 'That's how its SSE was' arrived at.
- 2 So, what he is now going to do is arrive at an SSE using the.
t 3 .probabilistic approach -- 4 MR. WARD: To see if it's different. 5 MR. MURPHY: To see if it's different. . l 6 MR.'SIESS: -- and to compare, and get down there. 7 , 8 MR. WARD: So, this is Vogtle and these numbers 9 are accurate. 10 MR. MURPHY:. These numbers are for.Vogtle. 11 MR. SIESS: Now, we are dealing with Vogtle. _ i 13 MR. SHEWMON: The dimensions are right --
,- 13 MR. CHOKSHI: This is the response spectra of
[-
~
i l 14 velocity and the response spectra of a response of a single 15 system mea.sured in velocities. 16 MR. SHEWMON: This is the speed of sound for a i 17 'particular frequency? 18 MR. CHOKSHI: .No. This is the response for a i 19 . single degree -- 1 20 MR. SIESS: The entire history of the earthquake on a series of single degree of freedom oscillators at
. 22 :[different) frequencies in determining the maximum velocity. :
23 It's just like you do acce,leration spectra, except this is 24 velocity. 1 l 25 MR. WARD: What is the numbers? ' b; [ . 4. 4 4
> m- 7 -- - , _ _ _ , _
f I I If-li 1 505 1- MR. WILKINS: I see'where'eight is. l '2 MR. MURPHY:- That's eight. 3 MR. WARD: More or less. Eight-is where the'line i 4 -is? 5 MR. MURPHY: Eight is where the line is. 6' MR. WARD: 'Over here is~seven or zero, or.what?
-7 MR. MURPHY: Over-here is ten, --
8 MR. SIESS: That's not log, so-it's zero. 9 MR. MURPHY: -- ten, and. someplace over here is - 10 zero. This'is simply to show you that we do..have -- l 11 MR. WARD: Another one of those curves. 13 MR. MURPHY:. Yes, sir. This is a product from the 13 Livermore or EPRI program, where.you get a plot'of these two , 14 items -- 15 ~ MR. SIESS: You left out a step. You went back
~
-16 now to the Vbgtle site and gave them a latitude and 17 longitude of Vogtle and said come'out and give me this.
18 MR. MURPHY: Right. 19 MR. SIESS: That comes out of the computer
.30 program. This takes all the seismic sources'that Livermore 21 had around the Vogtle site which-may or may not be the same 22 seismic sources that were used in the deterministic method 23 when'We got the SSE for Vogtle. That was done. That was 24 not only a dif ferent method but at a dif t'erent point in j 25 time.
- - - ~. - __
e 506 r ..
/
) ,
1 But now he's gone'back to the Livermore and come ' ' 2 up with this curve. 3 MR. MURPHY: This plot. 4 MR. SIESS: He says all right, which do I use for i S defining a spectrum in terms of the spectral velocity, , 6 average it five and ten hertz, and ten to the minus four 7 which he said now that -- - 8 MR. MURPHY: That's the target probability. 9 MR. SIESS: That's the target probability, the , 10 median of all the other plants out there using this 11- technique. 12 MR. KERR: Where did you get five and ten hertz > [ , 13 out of thic? 14 MR. .SIESS: The spectrum is a curve. Velocity is l 15 a function of frequency. What we used to do on acceleration 16 spectra was anchor everything out at the zero period 17 acceleration except we called it the 33 hertz, and then drew 18 from that. Now, using spectra velocity as a better measure 19 of the damage potential. Since you have to have a number 20 you can't plot a curve, you take an average of five and ten 21 hertz values. Sort of the same thing we did on Diablo. 22 MR. WARD: Okay, the average of the five and ten 23 hertz. ! 24 MR. SIESS: Yes. 1 25 MR. WARD: Okay, right.
,4 l 1
4
i f
~4 507 1 MR. SIESS: You pick a value off five and pick a 2 value off ten. The peak may be somewhere in between, it 3 usually is --
4 MR. MURPHY: Right. 5 MR. SIESS: Just like it is on acceleration. l 6 MR. WILKINS: That's the key point -- that's the 1 7 -geology that I didn't know why you picked five and ten. l l 8 Someplace in that range is where all the action is. l 9 MR. SIESS: I guess if you could integrate the 10 thing you would be better off. In fact, they do use -- 11 MR. MURPHY: We picked five and ten as an example 12 to work through this. One of the questions that we will be l{ 13 asking when we go out for public comment again, is whether 14 or not we shouldn't be using more frequency bands and what 15 those frequency bands are. 16 MR. WILKINS: And how you should use them. 17 MR. MURPHY: Okay, and how we should use them. 18 MR. SIESS: There is one approach where they use 19 the spectral density and you integrate a whole bunch of -- 20 MR. MURPHY: Right. 21 MR. WILKINS: The area under this curve -- 22 MR. SIESS: But it never is that simple. What 23 damages a structure or a system is likely to be integrated. ! 24 Most buildings for example and structures end up between two l l 25 and ten hertz. You want to just average that? u
. . ~ . _ _ .- _ _ _ . . _ .
508
- ( . What I want is a weighting function 1 MR. WILKINS:
2 which sort of says it's zero below two and it's zero above 3 ten, and it fluctuates between two and ten. 4 MR. SIESS: It's something like this, and here's 5 two and here's ten probably. If you talk to a structural , 6 engineer, he's only interested between two and ten. 7 MR. WILKINS: Because he knows that's the weight - 8 that does the damage. The building can ignore the one hertz 9 motion. 10 MR. WARD: That's wny you are using five -- right. 11 12 MR. MURPHY: That's why we are using five and ten [ 13 in the example. i l 14 MR. SIESS: He couldn't use any set of numbers -- l 15 you can't use any set, you will get different answers. The. 16 ching is, what you are aiming at is whatever I get I use for 17 design in such a way that it gives me a certain design. 18 It's the robustness of the design that I am interested in.
- 19. I think the guys that have been designing them probably i
20 decided that five to ten hertz is a pretty good measure. l 21 MR. WILKINS: And probably decided it a long time 22 ago, and there are lots of experience for it. 23 MR. SIESS: So far we have numbers. We are going 24 to use these numbers somewhere to get structures. It's the 25 structures that we are interested in. I can take different
~ .- . . _ _ . . _ _ _ _ _ _ . . - . . . _ . _ . . . _ _ . . _ . _ . _ _ _ . _ . _ . _ _ . _ . _ _ . _ . _ _ _ . _ . . _ . _ _
1 1 l i { 509 1- numbers and use them in'different ways and get the same j ~ 2 answer.and sometimes I can get the wrong answer.
' j
\
- 3 MR. MURPHY
- For the Vogtle site _you have a hazard I 4 curve'in spectral velocity. We enter that at one E to ten )
i
~
5 to the minus four, ten to the minus_four here. We read off 1 i 6 a spectral velocity of eight centimeters per second. Let's )- 7 put that piece of information in memory for a moment. 8 MR. SIESS: I am going to ask a question here 9 before somebody else does. At this point I.have. picked off 10 an eight centimeter per second average spectral velocity. I 11 can certainly create a velocity spectrum correspondingly. 12 Why don't I use that for design? 13 MR. MURPHY: You could use that for design,-but
, 14 -what we are interested in at this stage is taking just a few 15 steps further and making a comparison of-the controlling 16 earthquakes and the controlling earthquake ground motions 17 -that we get from the probabilistic side with what we have 18 from the deterministic side.
19 MR. SIESS: I have done my deterministic for 20 Vogtle and I came up'with an SSE -- what was-it, two-tenths 21 G -- and I took Reg Guide spectral for Reg Guide 1.60. I 22 anchored it at .2 G .- That gave me an acceleration spectrum. 23 I can get a velocity spectrum from that. Why can't I take 24 that one and compare it with this one? I know what's coming 35 next, and I am still trying to figure -- 4 4
- . , w w , - - , , -+ -,w ., . - - - - , , - - y e - .-- , , - = - ~
510 -, { ., > ll MR. MURPHY: I understand. We are saying that is 2 one step that could be taken to make that comparison. 3' MR. SIESS: In other words, coming down your 4 parallel charts I could stop at that point and compare. 9 MR. MURPHY: You could~. . 6 MR. CHOKSHI: This is a very general procedure on 7 coming up with the ground motion at a site you don't know - 8 anything about. If you had that information I think you 9 could compare at intermediate points and may,not have to go 10 much further. , 11 MR.'SIESS: Vogtle, we didn't know anything about 13 when we. started. j- 13 MR. CHOKSHI: Yes. , 14 MR. SIESS: Right? We came up with an SSE by what - 15 we now call the deterministic method. 16 MR. CHOKSHI: Yes. 17 MR. SIESS: You didn't come.in the next day, it's 18 a few years later and you are using the Livermore approach 19- probabilistically with a criterion you have selected of four 20 times ten to the minus four probability of exceedance and 21 gotten an eight centimeter per second. Can't I now compare 22 design basis by the two methods. I know what it was 33 designed for, and can't I express this one in the same ! 24 terms, or can't I? j I 25 MR. CHOKSHI: I believe so, you can. b 1 4
t 4 511 4 1 MR. MURPHY:- You can. What happens is, there are 2 a number of steps along here where you can make comparisons 3 and make decisions based on those. At this time we are 4 going on two recommendations. The first is an overall 5 recommendation from our panel to go on to the next series of 6 steps. 7 MR. SIESS: You are going to go on with both 8 numbers now. You have the deterministic design basis and i 9 you have this one, and you are going to do something else , 10 with both of them and compare something else. 11 MR. MURPHY: That's right. l 12 MR. SIESS: I just wanted you to see where we are. i 13 MR. MURPHY: Okay. At this time for the Vogtle 14 site, we have done, our calculations. We have our seismic 15 hazard curves, we have calculated a target probability, we 16 have arrived at a target probability and a target ground 17 motion. Now, we are going to take the seismic. hazard i l 18 information and we are going to deaggregate it into 19 magnitude and distance bins to arrive at what we will call a l 20 probabilistic controlling earthquake. Here, we simply have 21 selected a series of bins conveniently to carry out this 22 deaggregation; i.e., finding out what the contribution in 23 each of these bins are to the overall hazard at the Vogtle 24 site. 25 [ Slides.] i
\
) ( 512 k. 1 MR. MURPHY: The next viewgraph tells us how to do-
..)
2 this. 3 MR. SIESS: I am going to suggest that you go
~
4 through this as fast as you can, before somebody stops you. 5 I really mean that. . l 6 MR. MURPHY: I understand that. That is my ; 7 intent. 8 MR. SIESS: Get down to the MD's where we can look 9 at those and then to the spectra -- 1 10 MR. WILKINS: I would suggest that you do that. I 11 have looked ahead, and you are 9 tting into far more detail ; 12 in how you did it th:n I think is appropriate -- i - 13 MR. ?10RPHY : Okay. We just wanted to make certain la ther the detail was available to you if you were interested. 15 We can skip it, as you feel appropriate. ] 16 MR. SIESS: We can raise questions about them. 17 MR. MURPHY: We have had a lot of discussion about 18 what happens next. 19 MR. SIESS: The thing is, he is going through a 20 process to convert everythlug that he knows at the Vogtle 21 site from two sources, one from the deterministic and the 22 other from the Livermore, to convert it into an earthquake ; 23 of some magnitude at some' distance. He comes out with two l l 24 different answers, which he then goes back and gets spectra 35 for to compare. b i i
.. l
513 1 We are going to end back up with the spectra that 2 we skipped over before, and how they are different I don't 3 know, but that's what I want you to get through so that we 4 can get to that point. 5 MR. MURPHY: I will ask you to some extent to take 6 my word for it, that we are able to recompute scismic hazard 7 curves to fill in each one of those blocks. We get those 8 seismic hazard curves, and they are represented as being bin 9 one through three seismic hazard curves in the lower portion 10 of thic figure. Those are simply, again, arbitrary numbers 11 to go along with the maximum contribution to the fourth 12 highest contribution.
- 13 We take this eight centimeters per second that we 14 picked off for the Vogtle site, come back up the hazard iL curve and read off the probability exceedance for that bin 16 as we have called it. This table is a list of those 17 contributions at the various bins, having a contribution 18 here at two E to '.na minus five. The zero represents 19 contributions smaller than E to the minus ten. We have 30 called these values in here H of Md.
al We turn around and we sum these by this 22 formulation to enme up with what we are calling the 33 controlling magnitude and distance for that probabilistic 24 hazard assessment. 35 MR. SIESS: I called it a surrogate -- i 1
. - .. , . .. ,.- - - . . ~ . . . ~ ... ~ .- . - . - ._. . _ . - .. - . . .- . - - . ~ . ..- . .
i e I ( 514 /N ! . k. ) { l' ' M1R . SHEWMON: The distance of kilometers? ; 2- MR. SIESS: Yes, kilometers in this case. It i 3 could be as well done in miles. l 4 MR. SHEWMON: I am'sure it could be done in i 5~ furlongs too,:but here it's kilometers. . 6 MR. MURPHY: Here, it's kilometers. l 7l MR. SHEWMON: Thank you. i 4 8 MR. SIESS: This is a fictitious earthquake. [ 9 MR. WARD: How can you tell? E 10 MR. MURPHY: You cannot. You'have to believe me , s 11 on that one. 12 . M R . SIESS: This is a fictitious earthquake. l l {- 13 <MR . IURPHY:' This is a fictitious earthquake. ! i 14' MR. SIESS: Which, if you put it at--- I guess you ?
~
15 have;to get to the next one. If you put M bar at distance D 16 'bar from the site and assume that that's all there was i 17 there, and used some sort of attenuation -- t 18 MR. MURPHY: This would give you the same hazard - 19 as.we got from the 11 models -- 11 seismicity experts and , p 30 the five ground motion experts. This is a surrogate to that ' 21 hazard curve.' : l . 22 MR. SIESS: And then, you~are going to compare , t 33 that with a surregate earthquake for the deterministic
.24 method where yum n ad two different controlling earthquakes, l 35 right? ;
I r l
_. _ . . _ _ _ _ _ _ _ . _ __ _ - _ _ _ . _ . . . _ - . _ . _ _ . ...m__m. . _ _ _ . . . l e 7 515 ; a b 1 MR. SIESS: Right there. i 2 MR. SIESS: Wait a minute. At Vogtle you only ! i ! 3 ~have one controlling earthquake. l L ( 4 }!R . MURPHY: We had one controlling earthquake. j 3 There was-a contribution from a more distance earthquake -- l 6 MR. SIESS: , What if you had two earthquakes, one 7 on this side and one on that side. l 8 MR. MURPHY: Pardon? 9 MR. SIESS: Suppose yot: had two earthquakes, one 10 in the five to ten hertz range and the other in the ten to ! 11 15 hertz range. 12 MR. MURPHY: You mean, back there when we were t 3 13 binning it? 1 14 MR. SIESS: No, we are n'ot binning it, we are in lb the deterministic sida n.cw. l 16 MR. MURPHY: If we are in the deterministic side li and we had two controlling earthquakes -- I 18 MR. SIESS: You had two in your example, remember? l
)
2 ') MR. MURPHY: We had some discussion about two in . l 20 our example for Vogtle. The concern there was a 21 contribution -- l 22 MR. SIESS: The example wasn't for Vogtle t;as it? 23 MR. MURPHY: Yes, sir, I believe the example we 24 went through on Wednesday was for Vogtle. i 25 MR. SIESS: No, I am talking about -- F I l
- - . - -- . . . . . _ - _ - - . - - ~ . - - -- - _ _ - -
(, 516 1 MR. MURPHY: Remember, there was some discussion - 2 - 3 MR. SIESS: I am talking about this' example. 4' MR. MURPHY: Right. The schematic itself is not 5 laid out for Vogtle. , l l 6 MR. SIESS: That had two controlling earthquakes. ! 1 7 You only have one controlling earthquake'there. - 8 MR. MURPHY: That's right. j 9 MR. SIESS: What do I do if I have two; do I get 10 two more' numbers up there? ] 11 MR. MURPHY: Yes, sir. L 12 MR. SIESS: You have to compare each one of them i- 13 with the probabilistic controlling earthquake. 14 MR. MURPHY: That's right. As we have just said, 15 we end up with a controlling probabilistic earthquake and a l 16 controlling deterministic earthquake. In this case, because 17 we are using a preexisting plant, we have called this the 18 -design earthquake. At this stage we take and look at the 19 magnitude and the distances, and the first question here 20 would be to try to decide whether or not we were satisfied - 21 that these two earthquakes would cause the same ground 22 motion at the Vogtle site. 23 MR. SIESS: How do you compare them; that's what 24 Dr. Shewmon asked you. 25 MR. MURPHY: At this stage I think we make the 1
1 517 1 determination that it's not obvious that these two 2 earthquakes would produce the equivalent ground motion at 1 l 3 the site. So, we take and we carry this another step l 4 further. I will ask you to skip over slide 16 and we will 5 take a quick look at a legend on slide 17. 6 [ Slides.) 7 MR. MURPHY: Here, we have our three letters. The 8 actual design earthquake, a reference earthquake to the Reg 9 Guide 160 anchored at .3 G. l 10 MR. SIESS: Skip it. l l 11 MR. MURPHY: A probabilistic controlling 12 earthquake using the Livermore ground motion models for five 13 experts. We take the magnitude and distance on the 14 deterministic side and we used the same ground motion models 15 so that we were able to compare these. 16 This is the comparison for Vogtle. Our reference 17 earthquake, the actual design spectra they used for Vogtle, 18 C is our probabilistic earthquake and S is from our actual 19 earthquake. Basically, we make the determination at this 20 stage that the spectra from these two earthquakes are 21 indistinguishable. Given the way the NRC operates, we 22 probably actually will pick the controlling -- the 23 probabilistic controlling earthquake for the -- 24 MR. SIESS: What's confusing about this particular l 25 thing is that the two upper curves you should forget about. 'i.
..~ .. - - . - ~ ~ . - . - . - . . - - . , - . . . _ . . . . - . - . . . - . . . . . . . - . . - . . . - . - - . - r i t 518 -. 1 MR..WILKINS: They are irrelevant. \ .' , f 2 MR..SIESS: They have the same anchor as the two i 3 lowericurves=but the spectrum is derived differently. 4 11R. MURPHY: That's correct.- [ 5 MR. SIESS: If you are actually doing this in , 6 accordance with the regulations you would convert the ; 7- Livermore lower curve to.a spectrum like the design - 8 earthquake,.because it says use standard review plan -- , 9- what's the number? 10 MR. MURPHY: Two, five, two.
'll MR. SIESS: And, 2.52 would give you a spectrum ,
la that looks like the upper one, which points out an j j ~ 13 interesting. characteristic of what we have been doing all.of : l 14' these, years. The' lower curves essentially have the~same. ; 15 probability of exceedance at every value of frequency,. 16 right? 17 MR. WARD: The lower curves.- l 18 MR. SIESS: .The lower. curves, am I right? 19 'MR.. MURPHY: Go ahead.
- 20 MR. CHOKSHI: It has the probability of at average 21 of five and ten hertz, ten to the minus four.
22 MR. SIESS: The curve we would normally use which 23 is the upper curve, the Reg Guide 160 type curve,.has a much 1 24 . lower probability of exceedance at the low frequencies than 25 .it does at the high frequencies. If you start breaking it f k. e l e
I
\
l 519 1 .down by frequencies, what is the probability of exceeding so 2 many G's at two hertz versus ten hertz? You would not be l l 3 permitted to use the --
.4 MR. WARD: I don't see how you know that. You 5 don't know what you are seeing from.those curves. f j 6 MR. SIESS: The reason the C&S curves are below.
l 7 the others is because they have been using the actual' ground 8 model from.the Livermore studies, and that tends to give you ! 9 the~same'results at different frequencies; right? l 10 MR. MURPHY: Right.
- 11 MR..SIESS
- That's the two lower ones. They have 12 both been done using the same ground model. C&S has been 13 done using the same ground motion model for the two 14 different M's and D's. The D curve or the R curve --
15 MR. MURPHY: The R curve is just a reference. 16 That's the -- 17 MR. SIESS: That's the three-tenths G. 18 MR ., MURPHY: That's the three-tenths G. 19 MR. E1ESS: That has been done using the Reg Guide 20 160 spectrum. That is one option that the standard review 21 plan allows you. 22 MR. CHOKSHI: Yes. 23 MR. MURPHY: Yes. 24 MR. SIESS: There are other options of doing it, 25 but not many of them will get you down as low as those other I f l
_ _ _ ~..-.~.... _ _ ._.- . ._.._._ . .__ _ _ .. _ _ . _ _ _ . _ . _ _ _ _ ._ . _ . _ . . i i 520 + q. 1 two. ; 2 MR. CHOKSHI: That's correct.' 3 .MR. SIESS: If-you use the Reg Guide.160 spectrum s 4 on the Livermore peak ground acceleration.it will be very l 5 close to the design curve. - 6 MR. MURPHY: To the D curve, yes. That's right. l At this point they are very.similar, l 7 MR. SIESS: i 8 and the answer to what do you do when you compare them is j 9 simple. They are close -- I will'take the' bigger one. It i 10 doesn't make much difference. What if they are not close? 11- MR. CARROLL: He has some more examples. 12 MR. SIESS: The question has been asked. I will i I ask' the staff to answer it again. 13 14 MR. MURPHY: Let's to --
- 15 MR. SIESS: I want to point out something. The
- 16 ' rules say that thou shalt do it both ways. But the' rules. l 17 don't tell you what toido when you get two different 18 answers. Absent that information, I think a prudent 19 engineering and any kind of a lawyer will say you take the !
20 worst one. That was not the answer the staff gave me. ; 21 MR. MURPHY: No, it was'not. Here, we back up a 22 little bit and we list four trial sites that we carried this ; 23 out for so far. The first one is Vogtle, Seabrook, Pilgrim, l . 24 Perry and Bellefonte. -Here, we would say okay, probably on
. 25 the first order that we would potentially take these bottom 't- ,
6 J' l 521 1 two, Bellefonte at 5.7 at 20 kilometers and 5.8 at 15 2 kilometers and basically say that they were equivalent. At j 1 3 this stage I think we could make that determination. l 4 MR. SIESS: We are sure not smart enough to say l l 5 they aren't. We don't know either one of them that closely. l 6 MR. MURPHY: Right. Then, for Vogtle and Seabrook 1 7 we would say probably you cannot easily make that l 8 determination, and that you need to go to that next step 9 that we saw in the previous viewgraph and plot the spectra. i l 10 (Slides.) 11 MR. MURPHY: We will go back to figure 19. Of i la course, this is the one that we say is the typical case. I 13 The probabilistic and the deterministic overlay each other.
)
14 MR. SIESS: That always happens and we are home 1 15 free, aren't we? l 16 MR. MURPHY: Yes, sir. That's the casf;st case. 17 This is what we end up in this example for Seabrook. We 18 have the deterministic up here and probabilistic down here. 19 Probably the first thing that would happen that we would ask 20 the staff to take a careful look at this and to see whether l 21 or not there is something in the deterministic that we could 22 indicate was the source of this difference. 23 MR. SIESS: But you have two differences. You i i 24 have a difference essentially in magnitude and you have a 25 difference in the distribution there. .If I used the l l
3 4< - 522
} !
1- Livermore ground motion model on the original design 2 earthquake 1[ would get the C, wouldn't I -- or the S,
-3 -whichever one --
4 MR. MURPHY: Yes. 5 MR. SIESS: I didn't use that ground motion model. - 6 I used the Reg Guide 160 one, and I am way off up there at
-7 20 hertz.
8 MR. MURPHY: Yes, sir. Some of that is a ground 9 motion question and is involved in the source of'the .
'10 earthquakes that were used to composite the Reg Guide 160 11- spectra.
~12 MR. SIESS: You see, if that's wrong, we have been
- 13 wrong long before we did a probabilistic analysis. We can 14 fix that-without> going probabilistic, can't we?
15' MR. MURPHY: We could fix it, and we need to fix 16 it. It will be one of the tasks down the road.- 17 MR. SHEWMON: Seabrook is weaker in magnitude and 18 farther away than your design earthquake, yet you'say it's a 19 close horse race. 20 MR. MURPHY: No. I thought I had indicated that 21 the two, Vogtle and Seabrook would be the ones that we would 22 ask for a spectral -- we would ask for the ground motion to 23 be taken into consideration and the spectra actually f 24 plotted.
. 25 MR. SHEWMON: Which I took to mean that you i
-I
l 1 2 L l l l ! .h , 523
' 1. thought that it was'close, which one you should choose. I 2 . M]R . MURPHY: No. If they were close as in the l l
3 case of Perry,~it doesn't matter which one you chose. Here, I 4 we cannot tell simply by looking at -- l l 5 MR. SHEWMON: I understand now, and I didn't l l I 6 before. t 1 7 MR. SIESS: C is'Livermore, S is design; right? l 8 MR. MURPHY: C is Livermore probabilistic, S is { 9 design. L 10 MR. SIESS: You are. starting from scratch and you l 11 would have those two -- The point is, I look at your table
- lL2 and 5.7 at 19 kilometers versus six at 15, obviously the 5.7 13 at 19 is going to have less effect.
14- MR. CHOKSHI: Right. l 15 MR. SIESS: I don't have to draw the spectra to 16 see that they are different. How am I going to decide which 17 one to use? l 18 MR. CHOKSHI: I think in this case as you pointed i 19 out, six at 15 is going to give you a higher spectra. I , 20 think as Andy mentioned, the first thing you question -- in l 21 this case I don't think these are too dissimilar. The 22 reasons they are drastically different you would look at 23 what is the difference. If you understand this difference i l 24 and you are convinced that although the probability of a - 25 deterministic earthquake is low but for its in all sense a t l P
--c.- .. - - ,- - - . . , ._ e. - --
I 524 ) k. credible earthquake then you will'haveito pick which is the
~1 ,
2 worst one, if you convince'yourself that'the.both are coming 3- out from a-proper analysis. Ifoyou go back and find out that one was different 4-5 by some assumption which'is'not--- 6 MR. SIESS: I get you. Let's take a site and do a 7- deterministic, and I come up with .25 G for my SSE ground 8 motion. I do the probabilistic and I come up with .15. 9 Obviously, I should be lower on the probabilistic because I 10 am actually going to assign probabilities to those 4 11 earthquakes. When I do the deterministic I only 12 subconsciously assign probabilities by picking aul 13 earthquake. I have some cut off level.in my mind as.to what j 14 makes that reasonably maximum earthquake but I don't know l 15 what it is. 5 l 16 It's probably going to be lower than the one that 17 is really there. Are you going to debate that? Are you 18 going to go from .25 deterministic and accept .15 19 probabilistic. It would be worthwhile doing the 20 probabilistic study if I could gain that much. I 21 MR. CHOKSHI: I think one of the difficulties is i 22 always the question of what was assumed in the deterministic L 23 is always a question toward the judgment. This is the l 24 probabilistic, is provides that what is the other judgment - 25 - what is the effect -- l l'
~._-, . . - , ~ - . . ~.---- ~ . - _ - . . . . - . . - - .. - ... .
k
)
j' " ' '. 525 - 1 MR. SIESS: What-is the. probability that on a new 2 site you would actually come out larger with the 3 probabilistic approach than with the deterministic approach? l 4 The new site --keep in mind that you certainly know-about 5 the Livermore and EPRI source characterizations. 6 MR. MURPHY: To come out significantly different, i 7 I think, would have a reasonably low probability. . 8 MR. SHAO: Unless you include earthquake which has 9 very low probability. In the deterministic analysis you may , 10- include a very large earthquake which has very low 11 probability, but you include-in your deterministic study. 12 MR. SIESS: You are much more likely to include , 13 that large earthquake when you didn't need to in 14 deterministic. 15 MR. SHAO: In deterministic, yes. i 16 MP.. SIESS: I am saying -- let me speak like a i 17 regulator for a moment and say I don't care if I am too 18 high. Can I ever get in trouble just using the 19 deterministic method? I mean, trouble meaning that I have 20 designed it for too small a value. j .21 MR. SHAO: You say that earthquake I should 22 exclude but you shouldn't exclude. In that case 23 deterministic may be too low. ;
'24 MR. SIESS: I find that inconceivable, that any c 25 process that the staff was involved in something like that l
d
., ,. , - --s-
.g- 526 1 could happen. You might not know about it, but if you know 2 about it I don't think anything is going to get thrown out.
-3 MR. MURPHY: Right. You have at least 69 cases i
4 where we have tried it in the past, and that sort of thing 5 does not seem to have happened. - 6 MR. SIESS: Can I get in trouble if I went all 7 probabilistic? I don't know how to define trouble, because 8 the difference between two-tenths and .25 G, the risk of a I 9 core melt due to an earthquake for that difference in design I 1 10 G value, I don't think you will find. I don't know if j 11 anybody has made that kind of a study. By the time I get my l l 12 margins in there and I have to get up to the 1 G to start l l
\ 13 causing core melts whether I design for .2 or .25, most of I 14 the equipment doesn't come in that kind of increments.
15 MR. MURPHY: I would say if we are -- the ! i 16 potential, I think, is larger to get in trouble on the i 17 probabilistic side at this time because we just simply don't 18 have as much experience with making use of that technique . l 19 than we do with the deterministic. I would say I would hope 20 that we would not get into any kind of trouble like that. l 21 Right now, we are being a bit on the cautious side in saying 22 we haven't tried this as a licensing procedure and are a 1 23 little bit leery about turning it over into an entirely i 1 24 probabilistic approach. 25 MR. KERR: Why turn it over to a probabilistic
l. l l ff 527 l 1 approach at all if you are afraid you will get in trouble. l 2 There is no evidence that you would get in trouble if you l-3 used a deterministic. 4 MR. MURPHY: There, we ask for you to redefine 5 trouble and saying okay, we are probably not going to get 6 into trouble bccause we probably have not gotten into any l 7 trouble in the past using the deterministic but we have had 8 difficulties associated with getting these licenses through 9 using the deterministic process. 10 MR. SIESS: Andy, I think the Chairman said I have 11 ten more minutes. You want to just go into your arguments 12 on rationale. 13 MR. SHAO: Yes, we will skip -- 14 [ Slides.) 15 MR. MURPHY: We will skip back to after slide 22. 16 What we are saying is the staff proposes to require the use 17 of both the deterministic and the probabilistic, and to 18 compare the results to get the safe shut down earthquake 19 ground motion. We will skip the next slide and look at the 20 two following viewgraphs. We have enumerated what the staff 21 has considered the strengths and weaknesses of both 22 approaches. 23 MR. SIESS: Your last item says the staff has used 24 two approaches in license reviews. ( !. 25 MR. MURPHY: Yes, sir. u
._ _ . - . . _ __.. _ _ ., _. _ __ _ _ . _ . ~ . _ _ _ . . _ . _ . . . _ . . . _ . . _ _ _ . . _ - .
?
t
' (r 528 ' ]-
~
. 1 MR. SIESS: But you had no authorization to do 2 that in the regulations.
3 MR. MURPHY: Yes, sir. t 4 MR. SIESS: Even with Appendix'A, with all that i 5 stuff in there, now that you have it:in a Reg Guide-can't . 6 'you continue to'use it when,it's needed, even if it isn't in l 7 the rules? 8 MR. MURPHY: The last part of.your question again, 9 please? 10 MR. SIESS: Suppose I took the phrase out of the : 11 Appendix B that-says thou shalt use both approaches. You i 12 .mean, you couldn't use both of them? You say that you have ;
~
13 -been using the two approaches. 3 14 -MR.-MURPHY: What'I will say in the next viewgraph
)
15 --let me slide that up for a few seconds. -l 16 MR. KERR: - Before.you take that first viewgraph 17 'off, In the second paragraph do you really mean compliment .) i 18 or do you mean complement? 19 MR. MURPHY: Complement. These were banged out
'2 0 ' about 15 minutes before we arrived here. If you will, .
l 21 excuse us on a little bit of grammar. j 22 MR. KERR: That's the only thing I understood i 23 about the slides. 24 [ Laughter.) j 25 MR. SIESS: Go ahead, Andy. i I a w w * +er
l i l l-ll . l 'g 529 11 [ Slides.) i. 2 MR. SIESS: That's the answer to my, question, L 3 right? 4 MR. MURPHY: That's-the answer to your. question.
.5 For these two cases which is one that I tried chasing down 6 since we last met, for the Midland site the. staff had
.7 proposed using probabilistic analysis to justify dividing'
'8 the site up. The' Tectonic Province that had the Midland 9 sitenin it into two pieces. .One of the judge's on the 10 licensing board wrote an opinion that the use of the l
'll probabilistic analysis was not permitted within the j 12 regulatory framework. It didn't matter, because nothing 13 ever happened with Midland.
14 On Seabrook the use of probabilistic analysis -- 15 and I don't know the exact details -- was allowed to be 16 used. , l 17 MR. SIES.S: What'do you think a board is going.to 18- do if you go through the present proposed procedure and come L 19 up with a significant difference. Are they going to let you 20 argue that I don't have to use both of them, I don't have to 21 use the worst one? 22 MR. MURPHY: I think they will do to some extent 23 the same thing that the staff does, and we will first try to 24 understand why those differences are there. Then, probably ! 25 you are right, if we cannot understand those differences and L t
530 ^s
- ( )
1 cannot justify them and we have gone as far as the review 2 licensing board, they are probably going to say to use the 3 upper one. 4 MR. SIESS: Of course, that is what you were doing 5 before. You were using probabilistic in an attempt to 6 understand some differences and the board wouldn't let you, 7 you said. 8 MR. MURPHY: That's right. One of the differences 9 here would be that we would have this in the regulation -- 10 MR. SIESS: Suppose the regulation said that both 11 may be used, which I think is the same thing you are saying, 12 either or both may be used.
- ( .
13 MR. MURPHY: That's right. At this time what is 14 happening is, the staff is proposing to require both of them 15 to be used. 16 MR. SIESS: I know it. 17 MR. MURPHY: One of the things that we are doing , 18 with this document for public comment la requesting specific 19 comment on that point; should both be used, should either be 20 used, or should one or the other of them be excluded. We i 21 are interested -- the staff has an opinion and we have 22 expressed it and what we have put in this document, but we 23 are soliciting comment on this topic from the outside. 24 MR. SIESS: The thing is, you might be requiring f 25 both of them to be used but -- you say you are requiring I
I 4 I' 531 3 both of them to be used and you are not really, because I
~
1 2 don't know what'I would do when both are used and they~ don't
- 3. give the same answer. I don't see -- you say you are. going 4 to make them do both of them to show up these differences, 5 then what are you going to decide? Are you going to try to l 6 resolve the differences. ]
!- 7 I kept asking you can I get into trouble using j I l 8 just one of them. 9 MR. MURPHY: Can we get into trouble with the l 10 structure'using just one of them, and for deterministic we j 11' said probably not. We have a good bunch of experience on 12 that. P 13 MR..SIESS: Then you can get in trouble with the-14 board. 15 MR. MURPHY: We could get into difficulty with the ; L i i 16 board or with the intervenor -- , 4 17 MR. SHAO: Intervenor. With all the work that has , l '18 been done by Livermore and EPRI in the last ten years it is
~
19 very easy to pick about something and say'why didn't you use 20 that which was mentioned by Livermore or EPRI, magnitude 6.5 1 21 how can you exclude that -- 22 MR. SIESS: Larry, you are right, and I am going 23 to tell you what I told you before, I think -- at least I 24 told somebody. You have given some darn good arguments for 1, 25 using probabilistic approach. fj-l t
. .- - ~ - . - . _-.
I 1
'~
532 .f 1 MR. MURPHY: Right. l 2 MR. SIESS: None of them are good arguments for 3 using both. I am looking for the good arguments for using 4 both. I read what Goutam said four years ago as to why he 5 wanted this probab'listic approach opened up, and they were - 6 good arguments of things he could do with the probabilistic
~
7 approach that he couldn't do with the deterministic. That's 8 not an argument for using both. I haven't heard any 9 argument that says I am better off if I have both of them, , 10 and I think that's the place to leave it. There is no point 11 in giving us an answer now. 12 MR. ROTHMAN: I might point out that Andy hasn't 13 mentioned that they have an ongoing sensitivity study going 14 on, where they are going to do these same calculations for 15 all the plants in the Eastern United States and see if there 16 are any problems for the existing plants right now as part 17 of the program.. 18 MR. SIESS: I will leave it up to one of our 19 probabilistic experts to tell me that if 63 plants that we 20 have already used show no differences, that the probability 21 that the 64th one won't either. 22 Are there any more questions for Andy? Does 23 anybody else have any? 24 MR. KERR: I want to compliment the presentation. , 25 I thought I half way understood it. is
i i 1 533 1 MR. SIESS: I will guarantee you, if that's true 2 Bill, it was a damn good presentation. I had to go through ! 3 two different Subcommittee meetings to get to where I am 4 now, and you are not nearly as smart as I am. l 5 [ Laughter.] : l l 6 MR. KERR: But I had your guidance. ' l 7 MR. SIESS: Every time I hear it I understand it a l 8 little bit better, but I don't like it a little bit better. l 9 MR. WARD: Is that all, Dr. Siess? 10 MR. SIESS: That's all. I have a letter drafted, 11 and I have heard nothing to change my opinion. 12 MR. KERR: Are you having the Reporter record any ; 13 more? 14 MR. WARD: No, thank you very much. Let's take 15 five minutes. That's the end of the record. 16 (Whereupon, at 5:00 p.m., the transcribed portion 17 of the meeting concluded.] 18 19 20 21 22 23 24 25 l l J
. . __ . . _ . . - _ _ m._ . ._. . . _ _ _ . _ __ .
l l REPORTER'S CERTIFICATE 1. l l This is to certify that the attached proceed-ings before the United States Nuclear ! Regulatory Commission ; i in the matter of: j i NAME OF PROCEEDING: 382nd M.eeting i
- 1 DOCKET NUMBER:
l 1 PLACE OF PROCEEDING: Bethesda, Maryland were held as herein appears, and that this is i che original transcript thereof for the file of I I the United States Nuclear Regulatory Commission taken by me and thereafter reduced to typewriting by me or under the direction of the court report-ing company, and that the transcript is a true and accurate record of the foregoing proceedings. ! f, M '- d I Official Reporter Ann Riley & Associates, Ltd. l l l l l l l
1 i REPORTER'S CERTIFICATE This is to certify that the attached proceedings before the United States Nuclear Regulatory Commission l in the matter of: NAME OF PROCEEDING: 382nd Meeting DOCKET NUMBER: PLACE OF PROCEEDING: Be thesda , Maryland ) were held as herein appears, and that this is the original transcript thereof for the file of the : ) United States Nuclear Regulatory Commission taken l by me and thereafter reduced to typewriting by me ; or under the direction of the court reporting company, and that the transcript is a true and 1 l accurate record of the foregoing proceedings. i i g -
/
Official Re' porter 1 Ann Riley & Associates, Ltd. l i l 4 l j i l' l l l
I i
~
APPENDIX A REVISION PRESENTED TO THE ! t ADVISORY COMMITTEE ON REACTOR SAFEGUARDS i
/SR REO(f"% :
February 7,1992 Andrew J. Murphy, Chief i Structural and Seismic Engineering Branch ! Division of Engineering Office of Nuclear Regulatory Research i 301-492-3860 l
.. . _ . _ _ . - . _ _ _ . . _ . _ . _ . . _ _ . _ . _ . . ~ . . _ . _ . . _ _ _ _ . _ . . . . . _ . _ . _ . _ . _ . _ _ _ _ _ . . _ _ . _ _ . _ . .
t REVISION OF APPENDIX A TO 10 CFR PART 100 SEISMIC AND GEOLOGIC SITING CRITERIA FOR NUCLEAR POWER PLANTS e ISSUE i e WHILE THE CURRENT APPENDIX A HAS GENERALLY BEEN SUCCESSFUL IN i THAT THE MAJORITY OF CURRENT NPPs WERE LICENSED USING IT, THERE [ HAVE BEEN DIFFICULTIES. e APPENDIX A TOO DETAILED, INFLEXIBLE, AND LACK OF CLARITY IN SOME i SECTIONS LEADING TO CONFLICTING INTERPRETATIONS. e CONFLICTING INTERPRETATIONS OF' APPENDIX A OFTEN LED TO TIME CONSUMING DISCUSSIONS AND ADJUDICATION BY THE LICENSING PANELS ; e CURRENT APPENDIX A ISSUED IN 1973 DOES NOT REFLECT ADVANCES IN THE SCIENCES OF SEISMOLOGY AND GEOLOGY (FOR EXAMPLE, l PALEOSEISMIC DISCOVERIES AND PROBABILISTIC HAZARD ANALYSIS), AND DOES NOT REFLECT EVOLUTION OF THE LICENSING PROCESS THROUGH ADJUDICATORY PROCESS. t e THE MULTIPLE DEFINITION OF THE OBE (FUNCTIONALITY, LIKEllHOOD OF OCCURRENCE, AND MINIMUM FRACTION OF SSE) HAVE,IN SOME CASES, RESULTED IN THE OBE CONTROLLING DESIGN.
. .. . . . _ _ . _ . _ _ . _ _ . . _ . . . . _ . . . _ _ _ _ _ _ . _ . . . _ _ . . _ _ . _ _ . _ . ~ _ _ . ~ _ . . ._ ._ . . _ _. _ . l t REVISION OF APPENDIX A (CONT'D) t e IDENTIFICATION OF NEED FOR REVISION OF APPENDIX A e " GEOLOGICAL AND SEISMIC SITING POLICY AND PRACTICE FOR NUCLEAR I POWER PLANTS", SECY-77-288A, AUGUST,1977. e "lDENTIFICATION OF ISSUES PERTAINING TO SEISMIC AND GEOLOGICAL SITING REGULATION, POLICY AND PRACTICE FOR NUCLEAR POWER PLANTS", SECY-79-300, APRIL,1979. , e "
SUMMARY
REPORT OF THE SYMPOSIUM ON SEISMIC AND GEOLOGIC SITING CRITERIA FOR NUCLEAR POWER PLANTS", NUREG/CP-0087, OCTOBER,1986. e USER REQUESTS e MEMORANDUM FROM NRR (GILLESPIE) TO RES (BECKJORD) DATED AUGUST 2,1990. e MEMORANDUM FROM EDO TO RES DATED SEPTEMBER 6,1990 APPROVING RULEMAKING ACTIVITIES. HIGH PRIORITY STATUS. REVISED REGULATION GENERAL IN NATURE, DETAILS IN REG. GUIDES. SITING / DESIGN DECOUPLING.
i i t REVISION OF APPENDIX A (CONT'D) , i e OBJECTIVES OF THE REVISION _ ,
- REMOVE SOURCES OF CURRENT MISINTERPRETATION
- INCREASE THE EASE FOR UPDATING THE TECHNICAL GUIDANCE ,
e PROVIDE STABILITY IN LICENSE REVIEWS e COMPLETE REVISION FOR EARLY SITE REVIEWS
- f REVISION OF APPENDIX A (CONT'D1 '
8 SCOPE -
- REVISE APPENDIX A (NEW APPENDIX B), NEW APPENDlX S TO PART 50 TO DECOUPLE DESIGN FROM SITING.
i e CREATE NEW GUIDE, IDENTIFICATION AND CHARACTERIZATION OF SEISMIC SOURCES WITH PROBABILISTIC ACCEPTANCE CRITERIA.
- REVISE SRP SECTION 2.5.2, VfBRATORY GROUND MOTION.
t. e REVISE GUIDE ON NUCLEAR POWER PLANT INSTRUMENTATION FOR I EARTHQUAKES. e DEVELOP TWO NEW GUIDES: PRE-EARTHQUAKE PLANNING AND RESTART FOLLOWING AN EARTHQUAKE. e REVISIT EXISTING REG. GUIDES; IDENTIFY THOSE REQUIRING TECHNICAL AND EDITORIAL CHANGES. SUMMARIZING: e EARTH SCIENCES - 1 REG. GUIDE & 1 S.R.P. SECTION e ENGINEERING - 1 REVISED REG. GUIDE & 2 NEW REG. GUIDES
- -w. -- , , e c -~ ,-
t i, i 1 1 . I t REVISION OF APPENDlX A (CONT'D) i t APPLICABIL!H 9 t l- i i e NEW NUCLEAR POWER PLANTS
~
e USE FOR RESOLUTION OF ISSUES AT CURRENTLY OPERATING PLANTS ON A 4 CASE-BY-CASE BASIS i
b i' t t EXPERT TECHNICAL PANEL i; I Robert Budnitz Future Resources Assoc. : Kevin Coppersmith Geomatrix Allin Cornell Stanford University ; James Devine U.S. Geological Survey Walter Hays U.S. Geological Survey ' Robert Kennedy SMA Paul Pomeroy ;iondout Assoc. i J
_ . . . _ . _ . . _ _ _ . _ . _ _ . . _ _ _ . _ . _ _ . . _ . . _ . ~ _ _ . _ . _ . _ _ _ - . _ . . . . . _ . _ . - - - . - _ _ -. _ __ .. ._ SITE . DETERMINISTIC PROB BILISTIC ANALYSIS t ANALYSIS Geugical, Seismological and Goophysicalinvestigations
= l r i Conduct an EPRI or LLNL Identify Seismic Seismic Hazard Assessment Sources Compare to Operating Determine Expected Maximum Plants to Set Probability Earthquake for Each Source of Exceedance Level Determine Controlling Determine Controlling Earthquakes Ms & Ds Earthquakes Ms & Ds I I i Compare Compute SSE Ground Motion 3
k nit
- C sounce 2
)
I { l i 5(ISMIC See:t 3 , 1 O wt sin i i 54:sase I' So wcr 1
)
i I 5(asaac Sowcc 4 l l 1 i l l l tateoult,tes f am 151 AnD 554 i 4 sggi 1 N ... .**.,
~% .
t / .
\
- 8 /
- a / ~\ .
e / ;l \' ~
/
/ ,
,. , Mg.
\s sinc q 4 e g
/ - < g N
< s s
,/ \ KI3MIC f k x:wc 50uMI 1 swz:
SElWC Mh t I 0.01 0.1 1 10 Period (stcl SCHEMATIC DEFINITION OF CONTROLLING EARTHQUAKES I
- - . . . _ . _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ .. _ . - ~ . _ _ _ . . _ - - . _ . _ _ _ _ _ . _ _ _ . _ . . _ _ _ _ _ _ _ .
LO , l Log N h B ._ l- 1 1 A t b, Slope Site I I l I I ! I I 1 l t g l l m0 "1 al Geometry of homogenous source (b) Magnitude recurrence model zones (seismicity / tectonics) (frequency vs size) 1 i Log Y d Log P h 10-3 - Ground motion intensity o, Dispersion 10-2 By integration "b " 7 m,=6 10~3 - mb=5 4 - 10 l 1 = rR 0.1 g 1g Log Y Log distance (c) Ground motion prediction d) Seismic hazard curve model (attenuation) FOUR STEPS INVOLVED IN A PROBASILISTIC SEISMIC HAZARD ANALYSIS
PROBABil_ITY OF EXCEEDING DESIGN BASIS USING MEDIAN I,% HAZARD ESTIMATES FOR EASTERN U.S. SITES i :0 _ .
- ordered using med, log. har. :
9 : T p- : . D 7 _ Z m . . oc : . : F _ . V) .6 - - Q : . : m -
.5 5 p - I .
a
~
4 3 . - lE - : 3 s O
.2 i I : T t . . . . i O. u n we e u n we , u n wa m O O O ANNUAL PROBABILITY OF EXCEEDING SSE 8
TARGET PROBABILITY OF EXCEEDANCE. P 3 (FOR SPECTRAL RESPONSE AVERAGE OF 5 AND 10 HZ.) METHOD PROBABILITY OF EXCEEDANCE . FOR MEDIAN HAZARD ESTIMATES P. LLNL 1 E-4 EPRI 3E-5 i 9 ; i
\ .
t A !
, 10.2 g ,
o R Z . 1 0 ,0 3 m - m i o , x m u 10 4 ,. O : m i o [ 10-{ a 3 . Z [ z 10 6 E - 10-( j i o m SPECTRAL VELOCITY, Sy , cm/sec. PROB. OF EXCEEDANCE VS. SPECTRAL VELOCITY (AVG. OF 5 AND 10 HZ) - 10
i t e DEAGGREGATE THE PROBABILISTIC SEISMIC HAZARD CURVES BY MAGNITUDE AND DISTANCE GRIDS, i.e., DEVELOP THE CONTRIBUTION TO ! THE SEISMIC HAZARD FOR THE BINS INDICATED BELOW. l Magnitudes and Distance Bins Used in Example ; i Distance Magnitude Range of Bin Range of 5 - 5.5 5.5 - 6 6 - 6.5 6.5 - 7 7 - 7.5 > 7.5 Bin (km) O-25 25-50 50-100 ! 100-150 150-200
> 200 11
Aa % 'g-i i e COMPUTE SEISMIC HAZARD CURVES FOR THE INDIVIDUAL BINS. !
- DETERMINE THE RELATIVE CONTRIBUTION, THE PROBABILITY OF l EXCEEDANCE, THAT A SPECIFIC BIN MAKES TO THE CONTROLLING EARTHQUAKES. THIS PROBABILITY IS OBTAIN AT THE GROUND MOTION PREVIOUSLY DETERMINED.
t I i ! h
\ :t
\
l g
,~
_ \
\t
\
~
Bin 1 2i A 3 Bin 2 E Bin 3 , A Bin 4 Ground Motion 12
- - - - - _ - - - _ - - _ - - - - _ _ - - - - - - - - - - - - - - - - - - - - --- - - _ - - - - - - - - - - - - _ _ o
-m s
DEAGGREGATED RESULTS ; i i i : l H, Values for All Bins Based on the Median Hazard
- (Note
- If Ha s 1.E-10, it is listed as 0)
Distance Magnitude Range of Bin Range Bin 5 - 5.5 5.5 - 6 6 - 6.5 6.5 - 7 . 7 -7.5 > 7.5 0-25 2.0E-5 1.1 E-5 2.4E-6 0 0 0 25-50 6.2E-6 8.9E-6 5.0E-6 6.5E-9 0 0 50-100 6.0E-7 2.3E-6 ~ 6.8E-6 8.4E-7 0 0 i 100-150 1.6E-9 1.6E-7 1.5E-6 2.8E-6 0 0 l 150-200 0 1.1 E-9 2.1 E-8 4.6E-7 0 0 i L
> 200 0 0 0 6.0E-9 0 0 i
t i i i f i 13 ! 1 s
l
- USE THESE RELATIVE CONTRIBUTIONS, PROBABILITIES, TO WElGHT THE ;
COMPUT'ATION OF THE MAGNITUDE AND THE DISTANCE FOR THE CONTROLLING EARTHQUAKES. i Ni = I I mH,,/Z Z H , j md md t D = Z Z d H ,o/ Z Z H ,o md md ; i m = the midpoint of the magnitude bin l the centroid of the distance ring l d
=
I ! 14
s i d t i i i i COMPARISON.OF CONTROLLING EARTHCUAKES WITH DESIGN EARTHQUAKE BESULTS FOR VOGTLE SITE CONTROLLING DESIGN SITE EARTHQUAKES FROM EARTHQUAKE i PROBABILISTIC ANALYSIS M D M D t l VOGTLE , 5.8 32 i 5.3 15 4 t I l i 15 ;
. - - . _ _ _ _ _ _ - - _ _ _ _ - - - _ - _ _ - - - - _ _ _ - - - - - _ - _ _ - - - - . _ - _ - _ _ _ - - _ - _ _ - _ _ - _ _ _ _ _ _ _ _ - - _ - _ _ _ _ _ _ - _ _ _ _ __---_---____i
w , INITIAL RESULTS FOR FOUR TRIAL SITES CONTROLLING SITE DESIGN EARTHQUAKES FROM ' EARTHQUAKE PROBABILISTIC ANALYSIS i M D M D-VOGTLE 5.8 32 5.3 15 SEABROOK 5.7 19 6 15 PERRY 5.5 24 5.3 15 BELLEFONTE 5.7 20 5.8 15 ; 16
i J COMPARISON OF FOUR TRIAL SITES : D = THE ACTUAL DESIGN SSE SPECTRUM R = REFERENCE SPECTRUM - R.G.1.60 SPECTRUM ANCHORED AT O.3 g C = GROUND MOTION SPECTRUM FOR CE USING LLNL SGX MODEL ' FOR ATTENUATION S = GROUND MOTION SPECTRUM FOR MAGNITUDE AND DISTANCE OF ACTUAL DESIGN SSE USING LLNL SGX MODEL FOR ATTENUATION i k 17
- , . - - - - - - - - - - - - - - ------ ,4 _a4.. -tma a---'-----------'-------'-- - - - - - - - - - - - - - - - - - ' - - ' ' - - - - - - ' - - - - - - - - - - - ' ' - - - - - - ' ' - - - - - - - - - - - - - - - ~ - - - - - - - - - - - - - - - - - - - - - - - - - r------- - ------- -- - - - -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - a. E+4 . . . . , . . . .. , . . . . 5- , , 4 -
~
3 . 2 - E+3 - 5- - 4 - 3 ;
- o. 2- D -
CG u - , E+2 - o E 5- - 4 _ 3-2-
,' .' ' n'
*
- I n'
E+ 1_. ~ n' c, a' .. _. n . . m
+ + +
s s.2 s-2 = s.2 i i frequency ~ cps t COMPARISON OF GROUND MOTIONS FOR VOGTLE SITE ! 18 , i
~ ~
E+4 . - - - g
- - - - i -
i 5- _ 4-3- . t - E+3 - s- - 4 - C - 7. 3-
% 2 -
G o - E+2 - U U - o 5- - , 4 - 3 .
~
2-s . . . . 1 . . . .
. . . . n ~e e ne ~ e, E t t_, a n no a a _.
+
+ +
a so s.a w s.2 Irequency~ cps COMPARISON OF GROUND MOTIONS FOR PERRY SITE 19
l l - l l E+4 . . . . i . . . . . . . . l 4 . 3- - 2- - E t3 - - c b c, G- p - 3
'o. 2- -
Yo , E*2 - - o
% s- -
4 3 - . 2 - . gg . . . . t . . . . t . . . . _ ~ m .. o o n .. _ o o .. m i frequency ~ cps COMPARISON OF GROUND MOTIONS FOR SEABROOK SITE 20
a s s a 4 s e a a g s a s a g 5-4 _ 3_ 2-E+3 - c
"\ X -
r s 2 2-
<a o
E+2 - E 5-4 . 3-2 - E+1_ 4 A 4a i+ a a ;a 1+ A a ;a m 4
+
so so na w frequency ~ cps e n COMPARISON OF GROUND MOTIONS FOR BELLEFONTE SITE 21
b e i t i RATIONALE FOR REQUIRING DUAL APPROACH THE STAFF PROPOSES TO REQUIRE THE USE OF BOTH THE DETERMINISTIC AND THE PROBABILISTIC APPROACHES TOGETHER, , AND TO COMPARE THE RESULTS TO OBTAIN THE SAFE : SHUTDOWN GROUND MOTION. ! i AS NOTED ON THE FOLLOWING TWO VUGRAPHS THERE ARE t STRENGTHS AND WEAKNESSES TO BOTH APPROACHES, BUT THE ! STAFF BELIEVES THAT THE TWO COMPLIMENT EACH OTHER. IN PAST ACTIONS THE STAFF HAS USED THE TWO APPROACHES , IN LICENSE REVIEWS IN RESOLUTION OF POST-OL ISSUES 4 i
- \ t IN LICENSING CASES, LICENSING BOARDS HAVE EXPRESSED j DIVERSE VIEWS ON ALLOWING PROBABILISTIC ANALYSIS. MIDLAND l SEABROOK ! t IN RESOLUTION OF POST-OL ISSUES, NRC HAS DE-FACTO . ACCEPTED THE USE OF THE DUAL APPROACH VIA THE SERs. CHARLESTON ISSUE l i SEP l DIABLO CANYON i I
PROBABILISTIC ANALYSIS STRENGTHS o EXPLICIT RECOGNITION OF VARIOUS HYPOTHESIS, MODELS, OPINIONS, AND UNCERTAINTIES IN PARAMETERS - QUANTITATIVE IMPACT ON RESULTS EVALU-ATED. o RECURRENCE PERIOD AND SEISMICITY CONSIDERED IN EXPLICIT QUANTITATIVE FASHION. o IMPACT OF NEW FINDING'S AND NEW HYPOTHESIS EASILY EVALUATED. o EVENTUALLY, CAN BE USED TO ESTABLISH OR MEET A PERFORMANCE GOAL. o PROBABILISTIC STUDIES USED' AT MANY SITES IN PAST TO ADDRESS VARIOUS SEISMGLOGICAL ISSUES (E.G., O!ABLO CANYON AND EASTERN SEISMICITY ISSUE). WEAKNESSES o " BOTTOM-LINE" RESULTS TEND TO BE DOMINATED BY TAILS AND OUTLIER OPINIONS. r o LACK OF EXPERIENCE IN DIRECT LICENSING APPLICATION. i J 26
- _ _ __l
t I l l t l t m STRENGTHS - o EXTENSIVE PAST APPLICATIONS MAKING APPROACH WORKABLE, AND_ ALSO JUDGED TO BE SUITABLY CONSERVATIVE. o SITE-SPECIFIC INFORMATION AND RESULTS OF INVESTIGATIONS EASILY FOLDED INTO CONSIDERATIONS. WEAKNESSES o LEADS TO CONTROVERSIES AS DIVERSE OPINIONS ARE NOT EXPLICITLY TREATED. .! o PAST EXPERIENCE INDICATES DIFFICULTIES OF INTERPRETATIONS.
+ !
27
-----_--mm___.. - . . . _ _ _ - - - - -. - - - - - _ - - - - - - _ _ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - _ - - - - - - - - - _ - - - - - - - - - - - _ - _ - - - - - - - - - - - - - - - - -
i s [ OUTSTANDING CONFIRMATORY STUDIES t i o PERFORM STUDIES FOR SIMPLE TEST CASES (l.E., ONLY ONE SOURCE ZONE SURROUNDING THE SITE OR A FEW SOURCE ZONES) TO BENCHMARK THE i i PROCEDURE. o TARGET PROBABILITY OF EXCEEDANCE COMPARABLE TO NEWER PLANTS. o SCENARIO EARTHQUAKES AT DIFFERENT SPECTRAL FREQUENCIES. I o PERFORM STUDIES USING EPRI HAZARD ANALYSIS. o PERFORM STUDIES FOR ALL EASTERN SITES (5GX AND 4GX) AND COMPARE WITH ACTUAL DESIGN EARTHOUAKES. o PERFORM STUDIES FOR FEW CASES USING MORE RECENT GROUND MOTION MODELS. i 24
1 4 I L l l { l j APPENDIX S TO 10 CFR PART 50 EARTHQUAKE ENGINEERING CRITERIA L FOR NUCLEAR POWER PLANTS i l i i d PRESENTED TO THE ADVISORY COMMITTEE ON REACTOR SAFEGUARDS j FEBRUARY 7,1992 1 1 f i ROGER M. KENNEALLY U.S. NUCLEAR REGULATORY COMMISSION DIVISION OF ENGINEERING OFFICE OF NUCLEAR REGULATORY RESEARCH (301) 492-3893 4 9 w _,. - ,, . .. -,,,
l l l l ENGINEERING CONTENT OF APPENDIX A TO PART 100 DEFINES THE SAFE SHUTDOWN EARTHQUAKE AND THE L OPERATING BASIS EARTHOUAKE - Ill(c) and Ill(d) l l l DEFINES SAFETY RELATED STRUCTURES, SYSTEMS AND l COMPONENTS - lil(c), Vl(a)(1)(i to iii), and Vi(a)(2) l l l l l DEFINES THE MINIMUM VALUE OF THE SSE - VI(a)(1)v ESTABLISHES THE OBE/SSE RATIO - V(a)(2) j REQUIRES PLANT SHUTDOWN IF THE OBE IS EXCEEDED j V(a)(2) i VIBRATORY GROUND MOTION DEFINED BY RESPONSE l SPECTRA AT ELEVATIONS OF THE FOUNDATIONS - Vi(a)(1) and Vi(a)(2) l I ! IDENTIFIES ACCEPTABLE ANALYTICAL METHODS - Vi(a)(1) and Vl(a)(2) DESIGN FOR SURFACE FAULTING - Vi(b) DESIGN FOR SEISMICALLY INDUCED FLOODS AND WATER WAVES - Vl(c) l SOIL CONSIDERATIONS - V(d)(1 to 4)
l 1
- PROPOSED CHANGES TO PART 50 l
l i 550.34 CONTENTS OF APPLICATIONS: TECHNICAL INFORMATION f ADD PARAGRAPH (A)(12) AND B(10) 4 STATES THAT EARTHQUAKE ENGINEERING CRITERIA ARE IN APPENDIX S TO PART 50 l 550.54 CONDITIONS OF LICENSES 1 ADD PARAGRAPH (EE) ) ! PLANT SHUTDOWN REQUIRED IF THE OPERATING BASIS l I EARTHQUAKE IS EXCEEDED 1 .l j- ; j PROPOSED APPENDIX S TO PART 50 GENERATED BY MOVING SECTION VI, APPLICATION TO ! ENGINEERING DESIGN, OF APPENDIX A TO 10 CFR PART 100 i TO PART 50 a
4 i
- PROPOSED APPENDIX S TO PART 50 -
j APPLICABILITY i REVISED CRITERIA WILL NOT BE APPLIED TO EXISTING PLANTS !. CRITERIA APPLIES TO PART 52 OR PART 50 APPLICANTS ON OR ,' AFTER THE EFFECTIVE DATE OF THE REGULATION PART 52 APPLICANTS j DESIGN CERTIFICATION i COMSINED LICENSE i i j PART 50 APPLICANTS i CONSTRUCTION PERMIT i, ! OPERATING LICENSE f
l PROPOSED APPENDIX S TO PART 50 SAFE SHUTDOWN EARTHQUAKE GROUND MOTION DESIGNATION CHANGED TO SAFE SHUTDOWN EARTHQUAKE GROUND MOTION (ACRONYM REMAINS SSE) AMBIGUITY IN THE CURRENT REGULATION HAS BEEN REMOVED. THE FOLLOWING PHRASES HAVE BEEN REMOVED FROM THE REGULATION:
"THAT EARTHQUAKE"
" MAXIMUM EARTHQUAKE POTENTIAL" THE CONTROL POINT MOTION HAS BEEN CHANGED FROM "THE ELEVATIONS OF THE FOUNDATIONS" TO "THE FREE GROUND SURFACE OR HYPOTHETICAL ROCK OUTCROP, AS APPROPRIATE."
THE HORIZONTAL SSE AT THE FOUNDATION LEVEL SHALL BE AN APPROPRIATE RESPONSE SPECTRUM WITH A PEAK GROUND ACCELERATION OF AT LEAST 0.1G l 1
~
4 h
- PROPOSED APPENDIX S TO PART 50 1 OPERATING BASIS EARTHQUAKE 1 1 ,
i I MULTI-FACETS OF THE OBE THAT ARE IN THE CURRENT l l i REGULATION: 5 i FUNCTIONALITY OF STRUCTURES, EQUIPMENT AND i i COMPONENTS ! LIKELlHOOD OF OCCURRENCE (AFFECT PLANT SITE j DURING OPERATING LIFE) PERCENTAGE OF THE SSE l l l PLANT SHUTDOWN 4 4 l ! IN THE PROPOSED REGULATION THE OBE WOULu BE I i ASSOCIATED WITH: l i ! FUNCTIONALITY OF STRUCTURES, EQUIPMENT AND 4 COMPONENTS. i PLANT SHUTDOWN i ; i
l PROPOSED APPENDIX S TO PART 50 DPERATING BASIS EARTHQUAKE (CONTINUED) i THE APPLICANT CHOOSES THE VALUE OF THE OBE: l ONE-THIRD OF THE SSE l A VALUE GREATER THAN ONE-THIRD i I RESPONSE ANALYSIS REQUIREMENTS ASSOCIATED WITH THE OBE MAY BE ELIMINATED IF THE APPLICANT ELECTS TO I SET THE OBE VALUE TO ONE-THIRD OF THE SSE I DESIGN PROVISIONS ASSOCIATED WITH, FOR : INSTANCE, FATlGUE, ARE DISCUSSED IN REGULATORY l GUIDES HIGHER OBE VALUES MAY BE SELECTED, HOWEVER, ANALYSES AND DESIGN SHALL BE PERFORMED. f
GERMANY HAS RECENTLY REVISED THEIR SEISMIC REGULATION f I THEIR NEW REGULATION REQUIRES ONLY ONE DESIGN BASIS I EARTHOUAKE ANALOGOUS TO OUR SSE , 1 l AN INSPECTION LEVEL EARTHQUAKE (FOR SHUTDOWN) IS j SET AT O.4 SSE. i i THIS LEVEL WAS SET SO THAT CORRESPONDING VIBRATION i SHOULD NOT INDUCE STRESSES EXCEEDING THE
- ALLOWABLE STRESS LIMITS ORIGINALLY REQUIRED FOR THE
! OBE 3 ,
)
i l i l 4
'l
'O
l l
' 1
) ONGOING STUDIES IN SUPPORT OF THE OBE POSITION i RISK / MARGIN STUDY ; i
- BASED ON A SURVEY OF AEs AND DESIGNERS i
- COMPONENTS THAT MIGHT BE AFFECTED BY A' CHANGE l
. IN THE OBE DESIGN CRITERIA WERE IDENTIFIED BASED ON AN EVALUATION OF SEISMIC PRAs GENERIC CATEGORIES OF COMPONENTS WHICH ARE POTENTIAL ) SElSMIC RISK CONTRIBUTORS WERE IDENTIFIED i i A SUBSET OF THE ABOVE COMPONENTS WERE SELECTED AND REVIEWED IN GENERAL, THE OBE TENDS TO GOVERN IF THE i EARTHQUAKE CONTRIBUTION TO THE TOTAL LOAD IS SMALL, AND THE SSE TENDS TO GOVERN IF THE EARTHQUAKE CONTRIBUTION IS LARGE SURVEY OF UTILITIES AND DESIGNERS IMPACT OF THE OBE DEPENDS TO A LARGE EXTENT ON THE PHILOSOPHY OF THE DESIGNERS 4 DIFFERENT DAMPING TO GENERATE OBE AND SSE SPECTRUM ONLY OBE DAMPING FOR BOTH OBE AND SSE TO ALLOW FOR SCALING
l OVERVIEW OF EARTHQUAKE RESPONSE i l 1 i
- PRE-EARTHQUAKE OPERABLE SEISMIC INSTRUMENTATION SELECT AND INSPECT STRUCTURES AND EQUlPMENT (BASE- '
LINE) l l 4 l POST-EARTHQUAKE WALKDOWN EVALUATION OF GROUND MOTION RECORDS DETERMINE OBE EXCEEDANCE l 1 s-i
- l
( QRAFT REGULATORY GUIDE DG-1016 l
)
i SECOND PROPOSED REVISION 2 TO REGULATORY GUIDE l 1.12, NUCLEAR POWER PLANT INSTRUMENTATION FOR EARTHQUAKES i t l
\
SOLID-STATE, STATE-OF-THE-ART DIGITAL TIME-HISTORY ACCELEROGRAPHS l INSTRUMENTATION SENSORS PLACED IN THE FREE-FIELD, BUILDING FOUNDATIONS, AND AT ELEVATION IN THE BUILDING l l l i l 1 1 i )
$ i i I 4 ;
3 5
l l i DRAFT REGULATORY GUIDE DG-1017 i PRE-EARTHQUAKE PLANNING AND IMMEDIATE NUCLEAR POWER PLANT OPERATOR POST-EARTHQUAKE ' ACTIONS i )" ENDORSES SPECIFIC SECTIONS OF EPRI NP-6695, j
" GUIDELINES FOR NUCLEAR PLANT RESPONSE TO i
AN EARTHOUAKE" ( - OBE EXCEEDANCE CRITERIA j POSITION BASED ON EPRI NP-5930 AND EPRI REPORT TR-100082 (UPDATE OF EPRI NP- ! ] 5930) a BASED ON A CHECK OF THE RESPONSE j SPECTRUM, CUMULATIVE ABSOLUTE ! VELOCITY LIMIT, AND OPERABILITY OF THE j INSTRUMENTATION l PLANT SHUTDOWN CRITERIA OBE EXCEEDANCE, OR I WALKDOWN INSPECTIONS DISCOVER l DAMAGE I l J l
l l DRAFT REGULATORY GQLQE DG-1018 ) ! RESTART OF A NUCLEAR POWER PLANT SHUT DOWN j BY A SEISMIC EVENT
- ENDORSES SPECIFIC SECTIONS OF EPRI NP-6695, i l " GUIDELINES FOR NUCLEAR PLANT RESPONSE TO i AN EARTHQUAKE" PROVIDES GUIDELINES FOR PERFORMING VISUAL l
lNSPECTIONS AND TESTS OF EQUIPMENT AND STRUCTURES PRIOR TO PLANT RESTART i l l PROVIDES GUIDELINES FOR PERFORMING LONG-l TERM EVALUATIONS TO DETERMINE THE EFFECTS l
.- OF AN EARTHQUAKE l j l
l I i 3 l l
cua IN d V E 0 0 s6-t1-20 I r *** i ,,,, i AP93 -\ \ j UNITED STATES j [ y, NUCLEAR REGULATORY COMMIS$loN j ADVISORY COMMITTEE ON REACTOR SAFEGUARo$ wasumarow. o. cm,a i i February 14, 1992 i l l l The Honorable Ivan Selin i Chairman j U.S. Nuclear Regulatory commission 1 Washingt.on, D.C. 20555 3 Dear Chairman salin 1
SUBJECT:
I PROPOSED REVISIONS TO 10 CFR PARTS 50 AND 100 AND {
- PROPOSED REGUIATORY GUIDE 8 REIATING TO SEISMIC SITING A EARTHQUAKE ENGINEERING CRITERIA During the 382nd Safeguards meeting of the Advisory Committee on Reactor February 6-8, 1992, j thoseseismology, with propo, sed revisions to reactor siting regulations that dealwe complet geology, and earthquake engineering.
These matters were considered also during meetings of our Extreme } External 5, 1992.Phenomena Subconsittee on December 10, 1991, During these meetings, we had the benefit of discussionsand February l with representatives of the NRC staff and the industry. The proposed revisions are to be submitted to the Commission as . part of a package -intended to decouple siting from plant design. ! { Our report of January 15, 1992, provided comments on those portions
- of the package relating to the nonseismic revisions to 10 CFR Parts 50 and 100 and to the source term. The specific revisions covered '
!. by this report are those referenced at the end of this report. < } The existing requirements in 10 CFR Part 100 and its Appendix A
- remain date in effect of the proposed for revisions.
all plants licensed prior to the effective > j coismic siting portion of Part For future plants and sites, the Subpart B 100 will be included in a new codition, a(the existing requirements will become subpart A) . In new appendixnew Appendix B will be referenced in subpart B. This will those in Appendix A.contain much less detailed requirements than Caismic sources and procedures The identification and characterization for the selection of of a Safe Shutdown Earthquake (SSE cnd the enginee) ring criteria for seismic design of structures,w cystems,
- 50. and components will be in a new Appendix s to 10 CFR Part i These relocations regulations and guidanceof various serverequirements two purposes:within the body of (1) criteria for
~ "~'~~ ~ ~ "'" ~ ~^~" ^~~ ~ ~~~ ~ ~ ~~ ~^ ~ ~ ~ ~"~ t 6 W nd v i iiG ~% tG(~
{. The Honorable Ivan Selin 2 February 14, 1992 seismic where design they belong. of a plant, now in Part 100, are moved to Part 50, And (2), many requirements in Appendix A, that were state of the art when the appendix was written in 1973, are being brought up to date and are being removed from the regulations andtoplaced up date inin atheregulatory future. guide where they can be more easily kept reorganization. We commend the staff for this proposed It should make the licensing process more rational, and perhaps simpler, and will have no adverse effect on risk. In addition to the proposed reorganization, two of the proposed changes to the content of the regulations deserve comment. The proposed Appendix 8 to Part 50 redefines the Operating Basis Earthquake (OBE) in a way that leads to more rational consideration of the OBE in design and operation. Studies are being made to ensure increases thatinthe risk.proposed changes will not lead to significant right direction. We believe that this change is a step in the Two new Regulatory Guides, DG-1017 and DG-1018, have been proposed to provide guidance on inspections, evaluations, shutdown, greater thanandtherestart, OBE atfollowing a plant. the occurrence of an earthquake The other change is a more significant departure from current requirements. SSE ground motion Proposed Appendix B to Part 100 requires that the be determined "using both probabilistic and deterministic approaches." The statf does not claim that this new requirement will have any sionificant effect on safety. does believe, however, that a probabilistic approach willThe make staff it casier to determine an SSE ground motion in the face of unknowns or uncertainties, resistant and that the resulting value will be more robust and to challenge. In our view, that would argue for the use of a probabilistic approach, not for the use of a dual approach. Although we are not convinced that the proposed dual approach is either necessary or desirable, we have no objection to the staff proposing and publishing it for public comment. In summary, we have no reservations or concerns at this time that would argue againtst publication for comment of the several proposed revisions considered in our review. ! documents considered were not in final form.We note, Somehowever, thatbeen had not yet the edited, others were still being modified by the staff, and none had yet been reviewed by the Committee to Review Generic Requirements. If substantial changes are made in any of these documents before i I
- e r:co a_u w - ~ - - -
..v .-_
. d 1
The Honorable Ivan selin 3 February 14, 1992 I they them. are submitted to you, we expect the staff to inform us of Sincerely, David A. Ward Chairman Referenges: l
- 1. Memorandum dated October 11, 1991, from Themis P. Speis, I Nuclear Regulatory commission, for Raymond F. Fraley, Advisory Committee on Reactor Safeguards,
Subject:
Proposed Revision of 10 CFR Part 100, Reactor Site Criteria, Revisions to 10 CFR Part 50, New Appendix B to 10 CFR Part 100 and Appendix S to ! 10 CFR Part 50, and Associated Regulatory Guides Predeoisional), enclosing: (Draft i
- a. Draft Revision to 10 CFR Part 50 i
- b. Revised Part 100, Reactor Site Criteria i
- c. Proposed Revisions to Regulatory Guide 4.7, !
" General Site Suitability Criteria for Nuclear Power Stations"
- 2. Memorandum dated January 21, 1992, from Lawrence C. 8hao, I
Nuclear Regulatory Commission, for Raymond F. Fraley, Advisory Committee- on Reactor Safeguards,
Subject:
Revision of Appendix A to 10 CFR Part 100 -- Geological and Seismological Sitinga.Criteria for Nuclear Power Plants, enclosing Draft 10 CFR Part 100, Appendix B, Criteria for The Seismic and Geologic Siting of Nuclear Power
- b. Plants After (Effective Date?
Draft 10 CFR Part 50, Appendj.x S, Earthquake c. Engineering criteria for Nuclear Power Plants i Draft Regulatory Guide DG-1015, Identification l and Characterization of Seismic Sources, Expected ! Maximum Earthquakes and Ground Motion d. Draft Regulatory Guide DG-1016, Second Proposed Revision 2 to Regulatory Guide 1.12, Nuclear Power Plant Instrumentation for Earthquakes
- e. Draft Regulatory Guide DG-1017, Pre-Earthquake Planning and Immediate Nuclear Power Plant operator Post-Earthquake Actions
- f. Draft Regulatory Guide DG-1018," Restart of a g.
Nuclear Power Plant Shut Down By A Seismic Event Proposed Revision Vibratory Ground 3, Standard Review Plan 2.5.2, Motion i
, i
, Q9 3 -1 l F' P 9,.
FEB 3 8
- MEMORANDUM FOR: x Lawrence C. Shao, Director Division of Engineering, RES l
FROM: Andrew J. Murphy, Chief Structural & Scismic Engineering Branch Division of Engineering, RES
SUBJECT:
SUMMARY
OF PUBLIC MEETING 1 THE REVISION OF APPENDIX A,
" SEISMIC AND GE0 LOGIC SITIN RITERIA FOR NUCLEAR POWER PLANTS," T0 10 CFR PART 100 On February 4,1992, the NRC staff met with the st ff of the Nuclear Manage-ment and Resources Council (NUMARC) and other indus ry representatives .o discuss the proposed revision of Appendix A, Seismi and Geologic Siting Criteria for Nuclear Power Plants, to Part 100. Enclosure 1 is a list of attendees.
Public meetings on a proposed revision to the regulations are unusual. However, during a December 10, 1991 meeting with the NRC staff the Advisory Committee on Reactor Safeguards (ACRS) Subcommittee on Extreme External Phenomena stated that discussion of the Subcommittee members with representa-tives of the nuclear industry was impaired by the fact that the related documents being examined were not publicly available. In order to make future meetings more effective with respect to the exchange of views and information, the Subcommittee requested that all of the documents that have been or will be made available for Committee review be made publicly available. The Executive Director for Operations (ED0) agreed in this case that public release would make ACRS review more effective. Accordingly, the documents were placed in the public document room. NUMARC staff and industry representatives were not able to make detailed l comments on the proposed revision of Appendix A to Part 100. They did not ! have access to the documents until a week before the meeting. As a general l comment they did support the reduction in the level of detail that appears in the proposed siting regulation, Proposed Appendix B, Criteria for the Seismic and Geologic Siting of Nuclear Power Plants After [ Effective Date), to 10 CFR Part 100. 74estract7- 'pg
- l. -4.. t L
j . Lawrence C. Shao 2 cr.g ; g j3.g i Raymond Ng of NUMARC provided background information, in the form of two - nugraphs (Enclosure 2), on role of the seismic siting in the industry advanced i- reactor program.
- l. Future meetings may be scheduled after their review has been completed.
l Origind 90ned by Andr(G d Mijrphy l Andrew J. Murphy, Chief l Structural & Seismic Engineering Branch ! Division of-Engineering, RES i L
Enclosures:
As stated - DISTRIBUTION: RESReading PDR RKenneally NChokshi AMurphy RBosnak l LShao
- t. ,
e
'SSEB/DE/RES AMurphy:fkm 2//y/92 l L I l l l
i
)
l l [ i 1 l I i 5 I E
A'/i F4 5 0 < p a. s ,' n f e <,,- f)L. ,,s e j A M rp4f IJ A c/ Mar , 9 o/- 9'ez - 194 s D. t. Qmom f 6;/ Ed/u g/1-yog 9 70 9
. fl - P M etdl Ngc_ f]2E5 30: 4 339,3 Ndu A Clek&- A>Rc/ges goi 4p-s r/L
/26 ffc /'fu//m pgc-)?6S Jo/ 492 Ja 08 L, ou%n Rola/ Ro%
Q' twze-wize NAC- NRR Sol- 5 4-2733 Sof-SD4-3Sof 3 o /- sa q- a 739'
$1yN is Sd LK' C - 22 L Ram %rasaa @D E/'D P- 62 :50 J A 95 - sw? i
Ton o ~}$ u rwa P u n a rz.c_ zoz. - fsi2.- 71tO I
R P.KAsa wuA EPRI 4 II Pff- 23 7 5' C R/ hkwanc @edeyz.-i2.eo 1
. .c ,
Figure I-1: BUILDING BLOCK
SUMMARY
I : PREREOUISITES FROM ONGOING PROGRAMS Current Nuclear Imw-Level ! Plant Performance Radioactive Waste i (Utilities) (eel-ACORD) , liigh-level Adequate, Economic Radioactive Waste Fuel Supply (eel-ACORD) (EEI) GENERIC SAFETY / ENVIRONMENTAL REGULATION & INDUSTRY STANDARDS Predictable Licensing & ALWR Utility Requirements Stable Regulation (NUMARC) wr+n ISOJECT-SPECIFIC ACTIVITIES NRC Design Cenification Siting 1 (Plant Designers) (EPRI-USC/NUMARC) Enhanccd St " diration l First-of-a-Kind Engineering gey9 d ig (ARC-EPRI) (INPO) INSTITUTIONAL STEPS Enhanced Public Clarification of Ownership & Acceptance Financing (USCEA) (EEI) ! pWWa@,c l pp^ ~ State Economic Enhanced Govemmental Regulatory Issues Support (EEI) (ANEC)
- 14. Executive Summary i
/ '
/
INITIAL Ai!WR :g LICENSING / CONSTRUCTION
, E j 1991 1992 1993 1994 ? 1995 1996 1997 1996 1999 2000 2001 M
S able be u n h f ??;5flNf$ .j$$Y.ll$ff 'l l 1 y 1 1 ALWR Requirements p- . y' ., { u.< a soasan = ra.s-am a s Document
~
MWW~ ~- 4 g I a no u d ME *r=~ """." E [EmT Y i Design (77 yh y . ,7 Y [ } . sat = *.= c= sa.= = me== we y} ,,,,m,,,,,,,,,,,,,,, i Certification -
,5 gqfQyyjgj , , , , _ _ _
i .Sitingt PT1Q Q "f"Q.tf..Y e.N * = ** * **"
- W a"*c" *'
j Early Site Permit '
*YA 1 [0" MM *'*****""*"*"
4 ; . == ,ve == ==.as ====
"7,,,
- . ..,=
i Project Decision g,,, : m"* L**J* '" i ea **
- ,*J" ',*,",,a*",",;l"
. c.. . A . 7N Y Y - * - = " ' "
\ L Process s"' _ i ., a N Og mem%%T&N Plant Construction Q ;? h' fN g %" :lL, c"*v / g NRC Verflication ActMtles a 5 A n s nsa w - rua. i eme to.s [ .a1 m x
\
F. : November 19
% ___ _ .. - 1. - _ . . - .-~ >
i
+
--_..-,,-._,,_,__..-.._,....,..-...-..,-,..-.,,._m.., , , . _ . , . . _ . , _ , - , _ - , , . _ .
.-.,,-..,,.,v. . . , . . - . - a-
,I I
! j lw **:% UNITED STATES l$ %g NUCLEAR REGULATORY COMMISSION i
( ADVISORY WASHINGTON, COMMITTEE ON REACTOR SAFEGUARDS '; j D. C. 20555 l
%*** December 24, 1991 l
f James M. Taylor
,![l Executive Director for Operations MEMORANDUM FOR: I i
e A raley, / Executive Director Ra and . J FROM: Advisory Committee on Reactor Safeguards l
SUBJECT:
PROPOSED REVISIONS OF APPENDIX A TO 10 CFR PART 100, SEISMIC AND GEOLOGIC SITING CRITERIA ; FOR NUCLEAR POWER PLANTS The ACRS is actively reviewing the proposed revisions of Appendix A l 100 consistent with the staff's request for ACRS to 10 CFR Part 1991. l comments dated October 11, During the Extreme External Phenomena (EEP) Subcommittee meeting onthe d December 10, 1991, to review this matter, J Subcommittee members with representatives of the nuclear industry (e.g., NUMARC, EPRI, etc.) was who have a specific interest fact that the related documents being ceriously impaired by the these individuals. The EEP examined were not available ..to 1992, to Subcommittee is planning a second meeting on February 5,Following this meeting, , i continue its discussion of this 1992) matter.the and provide full its Committee is expected toI advice to the l 382nd meeting (February 6-8, l t Commisrion. l In order to make these planned meetings more effective ! l of the documents Committee that have review regarding beenbeormade this matter will bepublicly mAeavailabic available for ; as soon as possibic. This will include the proph 'd rule changes, in draft l proposed regulatory guides, and other related dococents, form if that is all that is available. l CC: S. J. Chilk, SECY T. P. Speis, RES L. C. Shao, RES , C. M. Abbate, NRR I EDO --- 007329
$~ ' 'A
$$&&Q 1
. ._ . _ . . . . - _ . . - ~ ~- - - . - - - - .-
- t- - ,
,, ,y UNITED sfATES f NUCLEAR REGULATORY COMMISSION i WASHINGTON,0. C. 309ts j l January 24, 1992 l
,r
*\**..*
MEMORANDUM FOR: Raymond F. Fraley, Executive Director Advisory Committee on Reactor Safeguards i FROM: James M. Taylor Executive Director for Operations I
SUBJECT:
PROPOSED REVISIONS OF APPENDIX A TO 10 CFR PART 100, SEISMIC AND GEOLOGIC SITING CRITERIA FOR N9 CLEAR POWER PLANTS . i requesting that the This is in response to your letter of December 14,1991, l pre-decisional document containing proposed revisions of Appendix A to 10 CF i Part 100 be made publicly available as soon as possible. l While we do not normally make pre-decisional material publicly available until l after the Comission has had an opportunity to review it, we agree Accordingly, in thiswe case public release will make the ACRS review more effective. will place in the public document room as soon as possible a draft copy of the 1 This draft copy will be the same as that to be provided ! requested material.to the Comittee in preparatton for the February 1992 Sub j Committee meetings. However, it should be recognized that this draft copy is subject to change as a result of reviews by the Sommittee and completion of i f the staff review and approval process. l 1
- - 1 l
J s xecutive Director i J for Operations i ! cc: The Chaiman Commissioner Rogers i Commissioner Curtiss .
- Comissioner Remick
! Comissioner de Planque . , SECY } i OGC I i 1 l ca2 m -2i-92 oi:ieru T/'-74g ,,
j
'o UNITED STATES g
8 o NUCLEAR REGULATORY COMMISSION pyA g E WASHINGTON, D C. 20555
%.....) AFR E8 HZ
. MEMORANDUM FOR: Lawrence C. Shao, Director Division of Engineering, RES FRON: Andrew J. Murphy, Chief Structural & Seismic Engineering Branch Division of Engineering, RES
SUBJECT:
SUMMARY
OF PUBLIC MEETING ON THE REVISION OF APPENDIX A, I
" SEISMIC AND GE0 LOGIC SITING CRITERIA FOR NUCLEAR POWER PLANTS," TO 10 CFR PART 100 i
On April 23, 1992, the NRC staff met with the staff of the Nuclear Management I and Resources Council (NUMARC) and other industry representatives to discuss i the proposed revision of Appendix A, " Seismic and Geologic Siting Criteria for Nuclear Power Plants," to 10 CFR Part 100. A public meeting notice appeared , in the Federal Reaister on April 7,1992, Vol 57, page 11691. Enclosure 1 is I a list of attendees. The meeting was scheduled at the request of the NUMARC staff. The meeting provided NUMARC staff, other industry representatives, and the public with an opportunity to discuss the draft proposed revision of Appendix A to 10 CFR Part 100 that was placed in the NRC Public Document Room (PDR) at 2120 L Street NW (Lower Level), Washington, DC (Hemorandum from Lawrence C. Shao to Raymond F. Fraley, dated January 21, 1992,
Subject:
Revision of Appendix A to 10 CFR Part 100 - Geological and Seismological Siting Criteria for Nuclear Power Plants). i I started the meeting with a brief summary of the Appendix A revision status. The sumary consisted of the following major points (1) the staff has received approval from the Advisory Committee on Reactor Safeguards to publish the proposed regulations for public comment, (2) minor technical changes have been made to the document that is in the PDR before it was submitted to the Committee to Review Generic Requirements (CRGR), (3) the initial CRGR meeting was in March, the package should be resubmitted to them having incorporated their initial comments in May, (4) a Commission briefing is scheduled for June, (5) the regulations should be published for a 90 day public comment period in late July or early August, (5) It is anticipated that the final regulations will be published about this time next year. Raymond Ug, NUMARC, followed with a brief explanation of why NUMARC requested the meeting. He stated that NUMARC has established an Ad Hoc Advisory Committee on Appendix A Revision. The Committee is scheduled to meet with NUMARC's Seismic Issues Working Group on May 5, 1992. They wanted to meet with the staff to make sure they understood what the proposed regulations would require and the staff rational for the requirements.
@$D$OCNTOY
l i , APR 2 81992 ! Lawrence C. Shao 2 l Enclosure 2 is a copy of the presentation I made on the scope and content of I the Appendix A revision. A considerable amount of the discussion focused on I the deterministic and probabilistic evaluations associated with plant siting. l l Mr. Ng noted that NUMARC would be able to give the staff detailed comments on l the proposed regulations (Appendix B to Part 100 and Appendix S to Part 50 and supporting regulatory guides) after the May 5 Seismic Issues Working Group meeting. He was advised that the ACRS requested notification if significant changes were made to the package after their review. Major changes could result in additional ACRS and CRGR meetings. l Mr. Ng also noted that NUMARC may request an additional meeting with the staff before the regulations are published for public comment. W '$ Andrew J. Murphy, Chief Structural & Seismic Engineering Branch Division of Engineering, RES
Enclosures:
As Stated cc R. Ng, NUMARC R. Bosnak, RES T. King, RES C. Ader, RES G. Bagchi, NRR N. Chokshi, RES R. Rothman, NRR R. McMullen, RES ora Kenneally? RES P. Sobel, NRR A. Ibrahim, NMSS PDR I
I ENCLOSURE 1 ATTENDEES PUBLIC MEETING REVISION OF 10 CFR PART 100, APPENDIX A, SEISMIC AND GE0 LOGIC SITING CRITERIA FOR NUCLEAR POWER PLANTS April 23, 1992 NRC Headquarters, OWFN, IF7/5 9:00 AM FMiE AFFILIATION
$t er ennu. 5 NRC [rcEs f ssG73
%& hosan3< Mc kES /M Dedu 7' C m w SC/44e l &L. - Siarr@ EPPI l
~0 n i > R Moove san L.
- 6 Tu c. l
~ t
\/'f M'ru Arn DnE e ~s hums $5 S A a % 9' A k s E a hclkl ;
I~ytIuvClue& N Rc /nss /Osa l 1
%,Hs sobel nec/valscai hob OhoRTod MR CAE0LtdA Etsc t- CAR N h v~ $u K- Jbua iv/u; $c1,/ac E'cr/dl/ Mar <-
6AN\ 6 rew e "rV A / M u e%Re_ D 'ur.? A / d j Abfc/AC.412-/ "
/W (%eh IV5W- , ACA,S 1
0ek /?k/Yb//en A//24/ ads MYes A Chh4/ ille cJAw s lQ Gw+ W t).t L % v w
=r;w e ."L ecs Nvmrte
@u A/p Naatc 8b< Whr N', w dlLel >J Mss How M _ C. u e rcw i rit u es w ci s
~9 add Fei, der ocwM I4 @ o /ulvahad -Clule
4 ATTENDEES PUBLIC MEETING REVISION OF 10 CFR PART 100, APPENDIX A, SEISMIC AND GE0 LOGIC SITING CRITERIA FOR NUCLEAR POWER PLANTS April 23, 1992 NRC Headquarters, 0WFN, IF7/9 9:00 AM NAjiE AEFILIATION
& haK tu e m el WlLLJ A AA AACfAdCwl k \Al6STOd _
f W 5EATE $4S hfu e M~ n% urd us'
/
Et ff k k ba.O Datz l 1 l l l I l
ENCLOSURE 2 APPENDIX A REVISION
~~'
April 23,1992 Andrew J. Murphy, Chief Structural and Seismic Engineering Branch Division of Engineering Office of Nuclear Regulatory Research e
t I REVISION OF APPENDIX A (CONT'D) e IDENTIFICATION OF NEED FOR REVISION OF APPENCiX A e " GEOLOGICAL AND SEISMIC SITING POLICY AND PRACTICE FOR NUCLEAR POWEfI PLANTS", SECY-77-288A, AUGUST,1977. e "lDENTIFICATION OF ISSUES PERTAINING TO SEISMIC AND GEOLOGICAL SITING REGULATION, POLICY AND PRACTICE FOR NUCLEAR POWER t PLANTS", SECY-79-300, APRIL,1979. e "
SUMMARY
REPORT OF THE SYMPOSIUM ON SEISMIC AND GEOLOGIC ' SITING CRITERIA FOR NUCLEAR POWER PLANTS", NUREG/CP-0087, OCTOBER,1986. I e USER REQUESTS i e MEMORANDUM FROM NRR (GILLESPIE) TO RES (BECKJORD) DATED AUGUST 2,1990. e MEMORANDUM FROM EDO TO RES DATED SEPTEMBER 6,1990 APPROVING , i RULEMAKING ACT3VITIES. ! - HIGH PRIORITY STATUS. REVISED REGULATION GENERAL IN NATURE, DETAILS IN REG. GUIDES. SITING / DESIGN DECOUPL;NG. t
,-,,,,--,,,--w-,, .v- .---n---,,.,m,,v.,,,,w., - , - , , ,,,.,-m,,, ,-,,-.,--,v,, , , ~ , , y ., ,-,e,--r,- -, , ,,-.,,---,-v,- - . , , -- -- , , -, ,,,,,-rer~ , - , , , - ~ r-- ~ - - - , , - ,
(. i l. 'l l REVISION OF APPENDIX A (CONT'D) 3 i e SCOPE
- i
- REVISE APPENDIX A (NEW APPENDIX B), NEW APPENDIX S TO PART 50 TO i DECOUPLE DESIGN FROM SITING.
- CREATE NEW GUIDE, IDENTIFICATION AND CHARACTERIZATION OF SEISMIC l SOURCES WITH PROBABILISTIC ACCEPTANCE CRITERIA.
e REVISE SRP SECTION 2.5.2, VlBRATORY GROUND MOTION. l e REVISE GUIDE ON NUCLEAR POWER PLANT INSTRUMENTATION FOR EARTHQUAKES. e DEVELOP TWO NEW GUIDES: PRE-EARTHQUAKE PLANNING AND RESTART FOLLOWING AN EARTHQUAKE. e REVISIT EXISTING REG. GUIDES; IDENTIFY THOSE REQUIRING TECHNICAL AND EDITORIAL CHANGES. SUMMARIZING: e EARTH SCIENCES - 1 REG. GUIDE & 1 S.R.P. SECTION e ENGINEERING - 1 REVISED REG. GUIDE & 2 NEW REG. GUIDES
i l l HIGHLIGHTS OF NEW APPENDIX B, PART 100 e RETAINS DETERMINISTIC GEOLOGICAL INVESTIGATIONS e REQUIRES APPLICATION OF EITHER EPRI OR LLNL PROBABILISTIC SEISMIC HAZARD ASSESSMENT TO PROPOSED SITE
- APPLICATION OF EXISTING / APPROVED COMPUTER CODE l - UTILIZATION OF EXISTING COMPUTER COMPATIBLE DATA BASES i
i e DRAFT FOR PUBLIC COMMENT PROBABILISTIC & DETERMINISTIC ANALYSES HAVE EQUAL WEIGHT E g i ; l 3
i
- SITE i
DETERMINISTIC PROBABILISTIC ANALYSIS
- ANALYSIS Geological, Selsmological
!' and GeophysicalInvestigations i I 1 Conduct an EPRI or LLNL Identify Seismic Seismic Hazard Assessment Sources 4 Compare to Operating Determine Expected Maximum I, Plants to Set Probability i Earthquake for Each Source j , of Exceedance Level i 4 Determine Controlling Determine Controlling i Earthquakes Ms & Ds Earthquakes Ms & Ds i i
; i 4
Compare i 4 3 Compute SSE Ground Motion 3 l . ' - -}}