ML20141H681

Responds to Discussing Comparative Evaluations of Lll & EPRI Seismic Hazard Studies

ML20141H681
Person / Time
Issue date:	05/08/1985
From:	Reiter L Office of Nuclear Reactor Regulation
To:	Stepp J ELECTRIC POWER RESEARCH INSTITUTE
Shared Package
ML20140B566	List: IA-85-535, Forwards Unexecuted Mod to Std Order for DOE Work: Probabilistic Assessment of Seismic Hazard for Eastern Us Nuclear Power Plant Sites. Order Modifies Lll Proposal Dtd 840411 ML20140C928 ML20140E702 ML20140E717 ML20140E723 ML20140E760 ML20140F043 ML20141H660 ML20141H673 ML20141H681 ML20141H686 ML20141H692 ML20141H696 ML20141H703 ML20141H751 NUREG/CR-3465, Comments on Encl May 1985 Draft NUREG/CR-3465, Probabilistic Safety Analysis Review Manual, Section 5.2 Re Seismic Hazard Analysis & Section 5.4 Re Fragility/Hazard Interface.W/O Encl
References
FOIA-85-535 NUDOCS 8601170370
Download: ML20141H681 (8)

Text

_ ,. . . .. - 1 ,, . . _ .

. p Af0

/ o,, UNITED STATES 8 o NUCLEAR REGULATORY COMMISSION s j WASHINGTON, D, C. 20655

\...../ NAY 0 81985 Dr. J. Carl Stepp Electric Power Research Institute

/ 3412 Hillview Avenue k

P.O. Box 10412 Palo Alto, CA 94303

Dear Carl:

In response to your letters of April 11, 1985 which discuss the comparative evaluations cf the LLNL and EPRI seismic hazard studies, it may be appropriate to review our basic perspective on this issue. As part of NRC's response to the USGS clarification of its position with respect to the Charleston 1886 earthquake, the staff included a short-term probabilistic assessment of seismic hazard at all nuclear power plants east of the Rocky Mountains. This assessment, an update of a more limited LLNL study, had been initiated prior to the USGS letter but took on greater importance as part of the so-called " Charleston Program." The earlier version of the LLNL seismic hazard study (begun in 1978) had been used very successfully in determining seismic reevaluation criteria for 9 sites in the Systematic Evaluation Program and in several licensing safety reviews. It underwent intensive staff, utility, ACRS and peer-panel review and was considered favorably in several licensing-board hearings. While there were individual

, utility studies, there was no coordinated response that could allow a meaningful comparison to the LLNL study in the manner in which it was used.

As a result we recommended in the " Charleston Program", that a unified utility study be conducted so as to allow such a comparison. - We were very pleased with the positive response by the industry in initiating the ongoing EPRI study. We are also pleased that the approach taken by EPRI is somewhat different than that taken by LLNL. The-present status of seismic hazard evaluation at low probability levels is sufficiently uncertain such that different perspectives on the same problem are needed. Our aim, as in the

, past, is to make decisions primarily based on relative estimates of i seismic hazard.

As a result of our meeting with the Executive Director for Operations (EDO) on November 7,1987, it was decided to delay calculations at most of the sites until the EPRI results 9 for test sites could be compared to the LLNL

calculations at the same sites. This meeting was an important meeting and I am enclosing a subsequent memorandum from the EDO which includes a short i statement on this delay.

The comparison with the EPRI results is an important part of ongoing 1 comparisons with other seismic hazard studies. In the LLNL report (to be i published in May) final results are presented for the 10 test sites along with comparisons made with the SEP studies, individual utility studies, USGS i

calculations and historic hazard analysis. These, plus additional studies

being carried out by the USGS, and of course, EPRI will allow us to make the I

8601170370 851205 PDR FOIA BELL 85-535 PDR k <

. E MAY 0 81905 best use of seismic hazard calculations. Towards the end of the 1985, after j we have seen the EPRI results, and the analyses of differences (if any) between the two studies, we will decide whether and to what extent repolling of the LLNL panels is necessary before calculating the hazard at the remaining 65 sites in the central and eastern United States. We view the USGS/NRC-sponsored conference on seismic hazard to be a very important -

4 element in the comparative evaluation process. We are looking forward to LLNL, EPRI and the USGS presenting results of their studies at this meeting.

i Additional studies by EPRI and the USGS will be considered concurrently with the LLNL calculations at the 65 sites. All available seismic hazard studies

! will be considered by NRC in determining how to apply the results to the

" Charleston Issue."

I I might add that we are already using the LLNL studies outside the context of this issue. The LLNL studies have been very helpful in reviewing several seismic PRAs, operating license applications and in a seismic reevaluation program. In one case (Hope Creek), which was not one of the 10 initial l sites, the applicant chose to use LLNL data and the LLNL methodology to

calculate hazard at the site and presented the results to the Advisory Cemittee on Reactor Safeguards. With respect to schedule we see three important meetings:

1. First Comparative LLNL/EPRI Evaluation Meeting, June. 18, 1985. We hope 1 that by this meeting LLNL will have completed calculations for 9 sites using the LLNL methodology with the EPRI ground motion model and the EPRI lower bound cut off. This will allow a direct comparison of results. We also hope that as part of this meeting EPRI will be able to present a short tutorial on its methodology and results.

2. Second Comparative LLNL/EPRI Evaluation Meeting, November 5,1985. At this meeting the bulk of the LLNL and EPRI comparative evaluations would be presented along with any the USGS has performed.

3. NRC/USGS Seismic Hazard Conference, November 25, 26, and 27,1985 in San Francisco. Presentation and discussion of comparative LLNL/EPRI

, evaluations, in the context of general seismic hazard evaluation, to i

the larger seismic hazard comunity.

We expect LLNL, EPRI and the USGS to come out with separate reports and comparisons. It has beer our experience that in seismic hazard evaluation, as in many other areas of the earth science, the most robust conclusions that NRC can effect come about through the encouragement of separate and independent analysis. Whether and how these analyses will appear in one volume is one of the items that needs clarification, j With respect to the National Academy of Science / National Research Council Panel or seismic hazard evaluation, we view this NRC initiated effort as a

longer range view of what the capabilities are and where we expect to go in the future. We do not expect this panel to recommend particular

,_.m .m. _,

__-.3

MAY 0 81965 methodologies although they have expressed a keen interest in receiving copies of both the LLNL and EPRI studies and participating in the NRC/USGS conference in November.

In conclusion, NRC views the LLNL, and EPRI studies as playing important roles in resolving ongoing seismological problems. We do not expect work in this area to stop with the publication of these and other studies.

Additional updated methodologies and calculations will undoubtedly be needed as our understanding increases.

I would be happy to meet with you again to discuss any additional coordination that you feel is necessary. I look forward to hearing from ,

you.

Sincerely Leon Reiter, Leader Seismology Section Geosciences Branch cc: w/ enclosure P. Sobel G. Giese-Koch J. Knight Sherwood H. Smith, Jr.

Ruble A. Thomas John J. Taylor EPRI Walter B. Loewenstein, EPRI Ian B. Wall, EPRI Jerry King, EPRI Don Bernreuter, LLNL i

. \

i United States Department of the Interior

=# # GEOLOGICAL SURVEY RESTON, VA. 22092 In Reply Refer To:

Mail Stop 905 May 17,1985 Memorandum To: Keiiti Aki, University of Southern California Ted Algermissen, USGS Don Bernreuter, LLNL Ken Campbell, USGS Kevin Coppersmith, Geomatrix Inc.

Allin Cornell, Stanford University A. Der Kiuregian, University of California, Berkeley Robin McGuire, Risk Inc.

Chris Mortgat, TERA Corp.

Otto Nuttli, St. Louis University David Perkins, USGS Maurice Power, Geomatrix Inc.

Leon Reiter, NRC Haresh Shah, Stanford University Burton Stemmons, University of Nevada Carl Stepp, EPRI Nafi Toksoz, MIT Daniel Veneziano, MIT From: Walter Hays, Chairman of Steering Committee

Subject:

Workshop on "Probabilistic Earthquake llazards Assessments,"

San Francisco, California, November 25-27, 1985 On behalf of the steering committee, I am inviting you to participate in the subject workshop cosponsored by the U.S. Nuclear Regulatory Commission and the U.S. Geological Survey (USGS) .

We are requesting that you prepare a "draf t paper" of at least 8 pages on the theme identified by-your name in the enclosed preliminary program. Please follow the format of the enclosed example. You will have 60 days af ter the meeting to finalize your paper before publication as a USGS Open-File Report.

We will reimburse non government participants for their travel costs in accordance with current government regulations. Please let me know within 30 days if you can participate in what promises to be a very interesting and valuable meeting.

Enclosure cc: Joe Berg k

m -

I

\ \

WORKSHOP ON PROBABILISTIC EARTHQUAKE HAZ,ARDS ASSESSMENTS Fisherman's Wharf Matriott Hotel San Francisco, California November 25-27,,1985 Sponsors : U.S. Nuclear Regulatory Commission and U.S. Geological Survey Objectives: 1) To review the methodology and results of recent studies to assess earthquake hazards probabilistically, and 2) to identify practical and innovative ways to improve the overall state-of-knowledge.

Preliminary Program (Note: The program will be refined to incorporate suggestions up to the date of the meeting.)

Monday, November 25 1-1/2 hrs. Review of Methodology--Studies conducted by Lawrence Livermore National Laboratory KLLNL) and Electric Power Research Institute 1 (EPRI) will be featured. Differences will be identified and discussed.

l --Carl Stepp, EPRI (will organize presentation)

--Don Bernreuter, LLNL (will organize presentation)

BREAK 1-1/2 hrs. Discussion of Methodology--Two participants will lead the initial discussion of the methodologies used by LLNL and EPRI. Everyone will take part.

--Haresh Shah, Stanford University

--Otto Nuttli, St. Louis University LUNCH 1-1/2 hrs. Review of Results and Comparisons

--Don Bernreuter, LLNL (will organize ~ presentation)

--Carl Stepp, EPRI (will organize presentation) 1-1/2 hrs. Discussion of Results--Two participants will lead the initial

-discussion of the results obtained by EPRI and LLNL and their compa risons. Everyone will take part.

--Maurice Power, Geomatrix, Inc.

--Ken Campbell, USGS 9

rwn - ~,--

"" y -, *y --y

~

\ l

~

Tuesday, November 26 1 hr. Historical Seismicity and Tectonic Information--What is the role of historical methods of analysis versus zonation methodologies now being applied in the Eastern and Western United States?

--Daniel Veneziano, MIT

--David Perkins, USGS 1/2 hr. Discussion of Presentations--One participant will lead the initial discussion of this issue. Eieryone will take part.

--Chris Mortgat, TERRA Corp.

BREAK 4 hrs. Problems Associated with Describing Seismic Source Zones Probabilistically--A series of presentations focusing on current problems of interest and concern. Discussion .to clarify issues will be permitted during presentation.

(Note: Lunch will be taken midway in the session.)

Type of Faulting, Subduction Zones, etc.

--Kevin Coppersmith, Geomatrix Inc.

Earthquake Frequency Distribution, Characteristic Earthquakes, Non-Poissonian Models, Attenuation, Seismic Gaps, etc.

--Robin McGuire, Risk Inc.

Directivity, Finite Faults, Complex Descriptions of Ground Motion, Uncertainty, etc.

--A. Der Kiuregian, University of California, Berkeley Maximum and Minimum Magnitude Estimation, etc.

--Burton Slemmons, University of Nevada 1/2 hr. Additional Discussion of Presentations--One participant will lead the initial discussion of these problems. Everyone will take part.

--Nafi Toksoz , MIT Wednesday, November 27 1 hr. Where do we go from here?--Two presentations suggesting innovative ways to increase the state-of-knowledge needed to make significant advances in probabilistic hazards assessments.

--Allin Cornell, Stanford University

--Ted Algermissen, USGS s

e i-

/

i

,.. \ -

1/2 hr.- Discussion of Presentations ,

BREAK 2 hrs. What are the present capabilites, limitations, and the appropriate next steps?--A panel discussion describing where we are now and i suggesting practical plans or proposed actions to increase the current state-of-knowledge.

--Keiiti Aki, University of Southern California

--Robin McGuire, Risk Inc.

--Carl Stepp. EPRI

--Don Bernreuter, LLNL

--Leon Reiter, NRC

--Ted Al germissen, USGS 1/2 hr. Closure -

2 3

i l

i h

f i

i Y

i

, , . , .---- - - - . _ . , , ,,y,, y-- ., ,,-3 . , ,_ . , - --, , ,,- , v.

-- - = -. .. _ _

,L i

. s,

\.

a INCREASING BAZARD AWARENESS AND PERSONAL PREP WESS by Os  ?

- $p +

of #

Risa I. Pale t Osiversity of Colorado 4N#f g j Boulder, Colorado 80309 Og#4 /e 0, *^ b,,

to #e ,Or IETRODUCTION

• d*

bopolo' 00 E 0 oq*#^7 0 The title of this session suggests that there is an association between O 7 l # f 0o,*

i increased hazard awareness and increased personal preparedness. In this paper, th'e extent to which such a linkage exists will be discussed, as well as the implications of this relationship for public policy particularly public education campaigns.

At the outset, it is important to determine both the current and the optimum levels of-awareness of the earthquake hazard in the Virgin Islands. Although the Virgin Islands have experienced numerous earthquake.s, the level of 1

i awareness of such earthquake susceptibility by either the resident or visiting population is not known. However, one suspects that resident awareness in the Virgin Islands is less than that observed in a state such as California with a history of major damaging earthquakes and an extremely active program of

public and private efforts to increase awareness and mitigation behavior. The adoption of mitigation acasures directly related to the Virgin Islands earthquake hazard could also be assumed to be less than the level observed for' California - which means that it is highly likely that only a minority of the papulation have earthquake insurance policies on their horses,. and few people have taken preparedness measures related to the earthquake hazard.

COMMUNICATION CHANNELS What sorts of campaigns should be taken then, both to increase general awareness of the earthquake hazard, and more importantly, to induce personal preparedness?

I

1. " 'o

~,, UNITED STATES

,E g NUCLEAR REGULATORY COMMISSION

• y WASHINGTON, D. C. 20555

' ' .. . . .

• DEC 2 41984 MEMORANDUM FOR: Chairman Palladino Comissioner Roberts Comissioner Asselstine Comissioner Bernthal .

Comissioner Zech FROM: William J. Dircks Executive Director for Operations

SUBJECT:

PROGRESS REPORT AND SCHEDULE CHANGE IN NRC PROGRAM PLAN RELATING TO CLARIFICATION OF U. S. GE0 LOGICAL SURVEY POSITION REGARDING SEISMIC DESIGN EARTHQUAKES IN THE EASTERN SEAB0ARD OF THE UNITED STATES For the purpose of licensing nuclear facilities in the Southeastern U. S.,

the NRC staff has taken a position, based primarily on the advice of the U. S. Geological Survey (USGS), that any reoccurrence of the 1886 Charleston, S.C. earthquake would be confined to the Charleston area; that is, the Charleston earthquake is assumed to be associated with a geologic structure in the Charleston area. The effect of this position is that nuclear power plants in the region east of the Appalachian Mountains are usually controlled in their seismic design, according to Appendix A to 10 CFR Part 100, by the maximum historical earthquake not associated with a geologic structure. This controlling earthquake is typically a Modified Mercalli Intensity (MMI) VII or VIII whereas the Charleston earthquake was a MMI X. Since 1974, the NRC has funded an extensive research project to gain further information on the causative mechanism of the Charleston earthquake.

In a letter dated November 18, 1982 from James F. Devine, USGS to Robert E. Jackson, NRC, the USGS~ clarified its position indicating that:

"Because the geologic and tectonic features of the Charleston region are similar to those in other regions of the eastern seaboard, we conclude that although there is no recent or historical evidence that other regions have experienced strong earthquakes, the historical record is not, of itself, sufficient grounds for ruling out the occurrence in these other regions of strong seismic ground motions similar to those experienced near Charleston in 1886. Although the probability of strong ground motion due to an earthquake in any given year at a particular location in the eastern seaboard may be very 1cw, deterministic and probabilistic evaluations of the seismic hazard should be made for individual sites in the eastern seaboard to establish the seismic engineering parameters for critical ,

facilities." l

Contact:

R. H. Vollmer, NRR 492-7207 f$ M 7 L i l l

..' l

)

l

.- . 1 l

In a Comission Paper dated February 5,1982 (SECY-82-53), we informed the Comission of the possibility of modification in the USGS position and in a memorandum dated November 19, 1982, the USGS clarification was forwarded to the Comission along with an assessment of significance and a pre-liminary plan to address the clarified USGS position. This plan was discussed with the Comission in its November 19, 1983 meeting and a joint NRR/RES program was initiated. The joint program consisted of a short tem probabilistic program that has as its core a Lawrence Livermore Laboratory (LLNL) estimation of seismic hazard at all nuclear power plant sites east of the Rocky Mountains, and a long term deterministic program through RES to determine the causes of large earthquakes, such as the Charleston earthquake, in the eastern seaboard.

With regard to the short term probabilistic program, final calculations by -

LLNL are almost complete for 10 test sites. As a means of comparison we recomended in our original program plan that a utility sponsored study also be carried out. The utility study, being conducted through the Electric Power Research Institute.(EPRI), is currently scheduled for completion in April 1985. We have decided to defer LLNL's estimation of the seismic hazard at the 65 remaining eastern sites for approximately 1 year so as to await completion of the EPRI program and allow a thorough comparison of EPRI and other seismic hazard estimates to be.made. The original plan called for this comparison to take place concurrent with LLNL's calculations, however, we believe this change to be useful in light of the rapidly evolving technology of seismic hazard estimation. As a result the probabilistic portion of the plan to address the clarification of the U. S. Geological Survey's Position on the 1886 Charleston Earth-quake will be available at the end of 1: 96, instead of in 1985 as originally scheduled.

The long term deterministic program is also progressing. Most signif-icantly, several teams of investigators, mainly funded by NRC, have found evidence. of paleoliquefaction in the Charleston area. This is the first indication of large to moderate earthquakes in the eastern seaboard in It is possible that a seismic recurrence interval for

~

prehistoric times.

large earthquakes in the Charleston area may be developed and that a detemination can be made whether or not this area is seismically unique. '

These conclusions await further investigations' and assessment.

~

(SignsiD William J.Dirck William J. Dircks Executive Director for Operations cc: SECY OPE OGC-

.