ML20141H751
Text
r
~
e FEB 141985 Mr. Richard A. Duval, Manager San Francisco Operations Office U.S. Department of Energy 1333 Broadway Wells Fargo Building Oakland, CA 94612
Dear Mr. Duval:
LLNL Technical Assistance to the Division of Engineering, NRR
Subject:
NRC "Probabilistic Assessment of Seismic Hazard for EUS
~
Nuclear Power Plant Sites" (A0448)
The enclosed NRC Form 173, Standard Order for DOE Work, is hereby submitted in accordance with Section III B.2 of the DOE /NRC Memorandum of Understanding, dated February 24, 1978.
Funding authorizations in the amount of $136,000 have been provided for this project to date. The purpose of this letter is to modify the LLNL proposal dated April 11, 1984.
FY 1985 funding will be provided upon the receipt of an acceptable proposal.
If you have any questions concerning the acceptance of this order, please contact Ms. D. Money on FTS 492-8378.
Sincerely, Ortginal signed by:
James P. Knight, Acting Director Division of Engineering Office of Nuclear Reactor Regulation
Enclosures:
1.
NRC Fonn 173
- 2. ~' Modification to Statement of Work cc:
R. Barber, HQ/ DOE D. Bernreuter, LLNL B. Bowman, LLNL D. Chung, LLNL p%21kog4851205 J. Johnson, LLNL BELL 85-535 PM DEGSB [J3 SBrocoum g
f
\\a) 9/A/DC A
_h_IR D
'DE
- PPAS
- DE:1 FC
- DE:GSB 5
~
____:... _ m l..::DE:
_ _ _ _ _ :..i AME:PSobe1[Im# :LReiter ex
- TDi oma
- JKr ht
__________.h._.[,........_:__________..:..___..__.
____:....________:___________:_......____:j/7/85 3 /85 ATE:1/24/85
- 1/L9/85 O/ 5-/85
/
0FFICIAL RECORD COPY hh
k U S NUCLE.AR REGULATORY COMuSSON On0ER NUMGER NRC FOmW 173 20-85-244 (144)
OATE STANDARD ORDER FOR DOE WORK FEB 1419g5 ACCOUNTING CITATION ISSUED BY; (NRC Othce) esSUED TO: (DOE Othce)
^*'" '"' ANON SweOL Office of Nuclear Reactor 31X0200.204 San Francisco Operations Office DivNylatifnEngineering asR NuMeER R
ion o 20-19-50-53-2 PenFoneseNG ORGANIZATION AND LOCATION M NUMGER Lawrence Livermore National Laboratory A0448 4 Livermore, California WORK PERIOD - TH.S ORDER FIXED E ESTIMATED C gnou; 7o; Probabilistic Assessment of Seismic Hazard for EUS Nuciear 01/17/84 09/30/85 FIN TITLE Power Plant Sites OsuGAT10N AVAILAS4UTY PROVIDED BY;
$ A.
TMS OADER TOTAL OF ORDERS PLACED PMCR TO TMS DATE WIT 8 2,939.000 S.
(TOTAL A a si s2,939,000 C TOTAL onoERS TO DATE AMOUNT mCLuCED W "C" APPUCam 8 TO THE "M NUMSER" CUE 0 m THIS ORDER 136,000 O.
3 FUNOS WILL NOT BE REPAOGRam80 SETWEEN FMS LWE O C FmANCiAt rtExeuTY-0 FUNOS MAY SE REP 40GRAMMED NOT TO EXCEED 2 oGUGATONS AUTHOAl210 S$ NOTED STANDARD TERMS ANO CONDITIONS (see NAC Manuar Cnepest 1102. Appenen Port a) A SECUNTY ATTACHMENTS THE FOLLOWWG ATTACMMENTS AAE HERESY MADE A PART OF THts
~ WORM ON TNf3 ORDER INVOLVES CLASStFIED INFORMATION NAC FOAM 187 IS ATTACHED.
ORDER N ON TMS ORDER INVOLVES UNCLASStFIED C STATEMENT OF WORK SAFEGUARDS. PAOPRIET ARY OR OTHER SENSITIVE C A00ffCNAL TERMS AND CONDITCNS MFORMATION Z OTHER WORK ON, THIS ORDER IS UNCLAS$lFIED AND NOT SENSITfvE.
FE RE E
E OE m_1549 REMARES (Re8s once ene proposer ey nurneer and dere. saa anecese d me anecned seevement of m 11, 1984, and requests that a This order modifies the LLNL proposal dated April revised proposal be submitted within 30 days to D. Money, DE/NRR with provided to M. Kaltman, PPAS/NRR.
of an acceptable proposal.
D. Dandois After acceptance, please send to the NRC Office of Resource Management, AT K. McGrath.
and provide a m ov to the Office of Nuclear Reactor Reaulation. ATTN:
ACCEPTIt0G ORGANIZATION
[ ISSUSNG AUTNORITY
$4GNATURE
,SiM(F t
SGNATURE
(
James P. Knight ints i
l'
./
l TiTtE Acting Directo(, Division of Engineering OATE NRC FORM ITS (144)
.,.m---m
--e,,
-9 r
-e_m.,,-r-,e--_m
,we,, -,
i Hodification to Statement of Work s
Title:
Probabilistic Assessment of Seismic Hazard for EUS Nuclear Power Plant Sites FIN No: A0448 B&R No: 20-19-50-52-2
.1 Revise
^
i Project Manager: Phyllis Sobel, FTS 492-8962
~~
Delete Task 2: Calculate the Seismic Hazard Spectra for all Sites East of the Rocky Mountains Subtask 3.1: Sensitivity Studies Using USGS Model 2
i.
1 Task 4: Historic Hazard ~ Analysis
{
Task 7: User Manual Preparation 1
Addition of Subtasks 3.3 through 3.9 i
Subtask 3.3: Sensitivity Study Com)arini SHC-Project Results To j
Electric Power Research Institute 4azarc Results Estimated Level of Effort: 0.10 psy Perform sensitivity studies for the 10 original sites included in the SHC project so as to facilitate' comparison of the results from the Electric Power Research InstitLte (EPRI) hazard program. This sensitivity study will involve minor software changes, so that the lower bound negnitude in -
the hazard code can be different than magnitude 3.75.
The software should be changed so that any ' lower bound magnitude could be.uSed. Specific calculations for the ten sites should be completed using a lower. bound magnitude of 5.0, compatable with the EPRI program.
Additionally these specific calculations should be completed using the same peak acceleration model used by the EPRI researchers. Hazard code runs using the above peak acceleration model should be completed for the full seismicity panel uncertainty analysis.
Subtask 3.4: Assess Difference between LLNL's Catalog, USGS's Catalce and EPRI's Catalog
-Estimated Level of Effort: 0.10 psy
. Obtain EPRI's catalog of historical earthquakes and compare to the catalog used as part of the LLNL project and the USGS catalog published by Stover, Reagor and Algermissen (1984). The comparison is to be made for magnitudes greater than 3.75 or MM Intensity IV (cutoff values for the LLNL catalog).
Identify the differences between catalogs. For the larger earthquakes (MM VII and larger) briefly examine the basis for the Feedback with the experts only to determine.if they relied differences.
primarily on the LLNL data base or primarily used other catalogs.
Assessment of the Impact of using a Uniform Classical Subtask 3.5:
Approach for the Development of the Earthquake Recurrence Models for Seismicity Panel Member's Zonations Estimated Level of Effort: 0.20 psy The LLNL panel members used various catalogs and methods for removing aftershocks, methods for correcting for incompleteness and estimating the Also implicit parameters of earthquake recurrence models for each zone.is the met the events in the catalog. Both EPRI and the USGS used uniform (but different) approaches to estimate the parameters of the earthquake recurrence models for each zone. Apply a uniform method to correct af tershocks and incompleteness and to estimate the paramete'rs of the earthquake recurrence models and their uncertainty. These parameters would then be used in a hazard analysis (for PGA) to assess the impact this approach would have on the seismic hazard at 4 of the test sites.
Only apply this approach to the zones that contribute significantly to seismic hazard at each site. Consider both the LLNL and EPRI' catalogs.
' Assessment of EPRI's A)proach for the Development of Subtask 3.6:
Earthquake Recurrence 4odels Estimated Level of Effort: 0.20 psy Veneziano has developed for EPRI an approach to remove aftershocks and using the culled set to simultaneously correct for incompleteness and Veneziano's approach requires the estinate the "a" and "b" parameters.
use of a number of smoothing (parameters and other analyst decis
- 1) to assess the implications of Veneziano's subtask has two objectives:
approach and (2) to assess the impact that Veneziano's approach might have on the results of our study. The assessment would be made for the zones identified in subtask 3.5.
NRC will request that EPRI make Veneziano's (ERRI's) software available to Determine if it is reasonable to get the software up on the LLNL LLNL.
computer system or if it would be simpler, after reviewing the computer program, for LLNL to use it on EPRI's computer system. Then apply
[
t 1
3-
.Veneziano's approach to develop the historic hazard at the same 4. test sites used in Subtask.3.5.
i.
Subtask 3.7: Sensitivity Studies / New Sites Estimated Level of Effort: 0.20 psy j
To better understand the differences between the.LLNL results and EPRI's results perform the following sensitivity studies:
i i
(1)
Impact of site correction. LLNL has included a complex site
' correction methodology in their analysis 'in contrast to EPRI which has not considered this question. As LLNL has up to eight catagories of sites, show the impact of.the various categories on the hazard.
Also include one-or-two site specific results to address the difference between the LLNL category approach tid site specific.
(2)
Importance of confidence level assigned to the zones.
LLNL panel members assign estimated levels of confidence in its existence to different. zones. Explore the significance of these confidence i
estimates by a sensitivity study in which LLNL alters the values i
assigned by.the experts and explores the sensitivity that the levels of confidence have on the hazard curve.
t (3) Cortribution of the complimentary zone.
LLNL analysis. terms the region complimentary to all of the specifically defined zones - the i
complimentary zone (in the SEP it is called the background zone).
All of the panel members assigned seismicity to the complimentary
,1 zone-and in some cases assured that large earthquakes could occur in this zone. This is significantly different from the EPRI approach j
which does not have a large complimentary zone. -To study the
. significance of the complimentary; zone, develop the hazard assuming j_
zero seismicity in the complimentary zone.
~
(4) Contribution of different distance / magnitudes to the hazard. For the j
best estimates case examine the contribution.to the hazard for different-magnitude bins and different distance zones. For several sites, examine how the contribution to the uncertainty changes as a function of magnitude and distance bin included in the analysis.
This study will allow us to assess the importance of any systematic differences in the upper magnitude cutoff between the~LLNL study and l
EPRI's study.
Also, as requested by NRC, develop estimates of the seismic hazard at selected sites using LLNL data files and current methodology.
i
,7~
.[
.....,y.[
,_.--...-((
,1,
-,,.,%,~,_r y
w.,,
,.[
e.-
,. ~,.,
..r
...v.
t
~
.. Subtask 3.8: Assessment of Differences Between EPRI's Methodology to Assess The Uncertainty and LLNL's Approach Estimated Level of Effort: 0.10 psy Given the hazard models', e.g. zonations, weights, earthquakes recurrence models for each zone, etc., there are different approaches to propagate the uncertainty through the analysis.
LLNL used continuous distributions except for zonation where discrete maps-up to 30 per expert-were used and a full Monte Carlo approach to propagate the uncertainty. The EPRI approach is to use discrete approximations for c6ntinuous distributions and a branching logic to propagate the~ uncertainty.
EPRI has only briefly.
discussed their approach in general terms; they may follow the approach used by Yankee Atomic Electric Co. for the Yankee Rowe site.
In the Rowe study, except for zonation, typically only three (except for the "b" value-when only two values were used) discrete values were used.
In addition, perfect spatial dependence was assumed (e.g. if the upper magnitude cutoff is high for one zone it is high for all zones). The difference between such a discrete approach to propagating the uncertainty and a complete Monte Carlo simulation could be significant.
Set up a simple comparison between the discrete approach and the full simulation approach to determine if the two approaches appear to lead to significantly different results. Since it would be difficult for LLNL to run the branching methodology, NRC would prevail upon EPRI to work with LLNL to fonnulate one or two reasonable test cases.
EPRI would provide LLNL with the results for the branching case and LLNL would provide EPRI with the results for the full simulation.
Subtask 3.9 : Task 3 Report Estimated Level of Effort: 0.10 psy Provide an integrated report documenting the efforts in the various subtasks including a section giving an overall assessment of the significance of the differnaces between LLNL results and EPRI's results.
Reporting Schedule A draft report (five copies) sunrnarizing the results of Subtasks 3.2 through 3.9 should be submitted to the NRC Project Manager and Geosciences Branch Chief by August 15, 1985. Staff conments will be returned'to the contractor within one to two weeks of receipt of the draft report. The final report (five copies) is to be submitted to the NRC Project Manager and Geosciences Branch Chief within two to three weeks of reciept of the staff's consnents.
I
~
1 5-Level of Effort and Period of Performance This modification will decrease the level of effort by approximately.8
- staff years. The period of perfonnance re:nains unchanged.
r i
i 1
i
)
4 5
0 l
4 1
!~
1.
i i
i 1
4.
l ll I
l
..m,
.--..re,.
.e.
..,,,.m..,
~. _,,,,., _..
Law.ence Livai more Natir.-i Labe:atory C '
c -- -
g 4
,L
'.. /
NUCLEAR SYSTEMS SAFETY PROGRAM i
~ ~ " ' ~
March 8,1985 EG-85-020 Carl Stepp Electric Power Research Institute 3412 Hillview Avenue Palo Alto, CA 94304
Dear Carl:
1985 between NRC, EPRI We thank you for organizing the meeting of February 24,The f the
) in our project additional analyses we have to perform (sensitivity and otherE.U.S." for the pu and LLNL.
" Seismic Hazard Characterization of the h Program".
the results with the EPRI analysis " Seismic Hazard Researcith the necessary As agreed at the meeting, we request that EPHI provides us wFollowing is a list input data and softwares to perform our analyses.
items requested:
including The catalogue of earthquakes used by the EPRI s talogue e
)
with all aftershocks removed (as used in the study.
i The actual ground motion models used by your contractors in the r iated with analysis, together with the weights, or probabilities assoc e
them.
f The software developed by Professor Veneziano for the calculation e
the seismicity parameters.
The software developed by EPHI to calculate the Hazard by the e
" Historical Method".
.?
f
.,., g., q
-i~.*1
..e,...,,9.,.,.
r EG-85-020 March 8, 1985
?
We would appreciate your sending these items to us as soon as they become available.
Thank you in advance for taking actions on this matter and we look forward to seeing you at our next " comparison" meeting.
Sincerely, J
Je
. Savy Engineer, Geosciences Group JBS:se:124s cc:
D. Bernreuter/LLNL G. King /EPRI L. Reiter/NRC I
l
__..._...____.______,......______________._.,.._m._
+
,i. _. s_ ww. A -..wa Am
.. er%.~
,.m.
m
- w. o-e y-
,,,, ~.,
i
)
i l;
l
=
i Bectne Power i
Research institute March 18, 1985 TO:
. Leon Reiter i.
Don Bernreuter Jerry King Jean Savy FROM:
Carl Stepp
SUBJECT:
Minutes o the EPRI-NRC/LLNL Planning Meeting'on Comparative Evaluations, February 27, 1985 The attached are my minutes of the meeting.
Please let me know if you have comments.
JCS:br Attachment cc:
Sherwood H.
Smith, Jr.
R.A. Thomas J.J._ Taylor W.B.-Loewenstein Pr?sube1T ~ t I.' B. Wall 3412 Hillvlew Avenue, Post Office Box 10412 Palo Alto CA 94303 Telephone (415) 855-2000 Washington Once: 1800 Massact usetts Ave., NW, Suste 700, Washington. DC 20036 (202) 872 9222 fy
t.<,.
~ ~ ~s
,- - ~~ ~~ _
s-......
.-~.,--.~...,n~.
. -.. ~. _.. _
r i
Minutes s
Meeting with NRC/LLNL February 27, 1985 Westin St. Francis San Francisco, CA The meeting was begun at 9:00 AM.
Present Leon Reiter (NRC)
Carl Stepp (EPRI)
Jean Savy (LLNL)
Jerry King (EPRI)
Donald Bernreuter (LLNL)
Purpose i
To form a schedule and preliminary plans for comparative evalua-l tions of the NRC/LLNL and EPRI seismic hazard methodologies at nine test sites.
Schedule The comparative evaluations will be initiated with an information exchange workshop on May 23, 1985.
The following key milestones were established.
e May 23, 1985 Workshop to accomplish technical re-view of the NRC/LLNL and EPRI seismic hazard methodologies and finalize scope of comparative evaluations.
e May 28, 1985 Final comparative evaluation plan.
e June 3, 1985 to Perform comparative evaluations.
September 3, 1985 e
September 3-27, 1985 Prepare draft comparative evaluation report.
e - October 3, 1985 W rkshop on comparative evaluations.
o e
November 15, 1985 Complete comparative evaluation report.
General Structure of Workshop 1 (1)
EPRI presentation of methodology including proposed parametric analyses.
m.
.2 -
s (2)
NRC/LLNL presentation of methodology including proposed parametric analyses.
(3)
Finalize scope of comparative evaluation report.
-- General Structure of Workshop 2 (1)
EPRI presentation of draft comparative evaluTtions.
(2)
NRC/LLNL presentation of draft comparative evaluation report.
(3)
Resolution of comparative evaluation issues.
An issue to be resolved prior to Workshop 2 is the appropriate form of the comparative evaluation report and the procedure for making it a part of the permanent record.
Role of USGS USGS staff is expected to participate in the two workshops'and e
to provide comments on both the EPRI and LLNL reports and pro-
~
posed parametric analyses.
The USGS has done some parametric analyses based on alterna-e tive tectonic models (perhaps other considerations (?)).
Leon Reiter will determine the scope of this effort and assess whether it can be included as part of the comparative evalua-tion effort.
Preliminary Parametric Analysis Proposed by LLNL 1.
Test lower bound magnitude cut-off.
-- EPRI ground motion model will be used for this comparison.
2.
Test impact of different' earthquake catalogs used by EPRI and
~
LLNL.
-- Determine differences in catalogs.
-- Perform computations to test impact on final hazard estimates.
3.
Perform independent calculations of seismicity model para-meters a and b using a statistical model to be~ developed by LLNL.
Both EPRI catalog and LLNL catalog will be used.
4.
Use EPRI's seismicity parameter evaluation method to estimate a and b values for LLNL source zones.
~ 5 ~.
Perform sensitivity on site corrections.
a 6.
Perform sensitivity on scurce zone uncertainties.
7.
Perfcrm sensitivity on distance and magnitude parameters.
3.
Test key sensitivity of the simulation approach (LLNL) versus the logic tree structure (EPRI) for aggregating source zones.
Appropriate test cases need to be identified.
9.
Test sensitivity on assuming no, partial, full correlation be-tween a and b parameters of the seismicity model.
Preliminary Parametric Analysis Proposed by EPRI 1.
Comparisons to test essential differences between the logic structure approach (EPRI) and the expert opinion simulation approach (LLNL).
-- Specific isolation of sources.
-- Specific isolation of seismicity parameters.
2.
Comparisons to test details of the methodologies.
-- Method for truncating upper bound magnitude.
-- Methods for determining seismicity parameters, a and b.
-- Methods for determining act ivity rates and establishing appropriate constraints.
Impact of site-specific versus general regional interpretations.
-- Upper bound magnitude.
-- Different seismicity models.
3.
Specific evaluations to assess' key elements of the.EPRI methodology.
-- Scientific uncertainty matrix.
-- Data uncertainty matrix.
The final methodology should be site independent.
The NRC feels that this should include interpretation of source zones.
The scope of comparative evaluations will be finalized at the May 23 workshop.
The meeting was adjourned at 11:30 AM.