ML20148P807

Reviews 780531 Final Rept, Simulation of Earthquake Ground Motions for Facility. Requests Applicant Supply Math Models Used in Swis,Prose & Faltung Programs.Analysis Remains in Linear Elastic Domain & Is Crude Approximation of Reality

ML20148P807
Person / Time
Site:	San Onofre
Issue date:	09/28/1978
From:	Zudans Z FRANKLIN INSTITUTE
To:	Wright E Advisory Committee on Reactor Safeguards
References
FOIA-79-86, FOIA-79-98 ACRS-CT-1054, NUDOCS 7811290045
Download: ML20148P807 (2)

Text

A . -

lJFranklin Research Center A Division of The Franklin Institute g September 28, 1978 Lew.to ADWE Y C ' ".:lT TEZ C?.

Mr..Ed Wright EN Sem u. :c.4 Advisory Committee on Reactor Safeguards U.S. Nuclear Regulatory Commission 00i 3 10/ij Washington, D.C. 20555 .

p$'qn.'.*

'I ~ " '

@? .i ,9% *.l ' *

Subject:

Review of San Onofre Seismic Analysis g,

References:

1. " Simulation of Earthquake Ground Motions for San Onofre Nuclear Generating Station, Unit,"

Final Report, May 1978.

2. " San Onofre Nuclear Generating Station, Unit 1,"

e NRC Docket 50-206, April 29, 1977.

Dear Ed:

As per your request the following has been accomplished as of this date. Report Ref. 1 was reviewed and it is noted that significant analytica1'and numerical effort has been invested to design some capability to predict the best estimate most severe carthquake that could reasonably occur at the San Onofre site.

It is noted that pred etions are given only for the earthquake horizontal component envelope. Presumably the reason for it is that the analytical tools used (SWIS, PROSE, FALTUNG) predicted the earthquake vertical component consistently in excess of that observed in the events used for computer program validation.

This particular circumstance raises a question with regard to the validity of the results for horizontal component as well.

While Ref. l provides very clear discussion of the methodology used in SWIS, P QSE and FALTUNG, actual mathematical formulations are not given. ccordingly, this reviewer can not evaluate the adequacy of the athematical modelling on basis of the available information in Ref. 1 alone. It is requested that the applicant supply the mathematical formulations used in the above computer programs.

Assuming that all formulations represent the best that any mind can ccnceive witl\ today's knowledge, the entire analysis is still in the linear elhstic domain and hence only a very crude approximation to the rebl non-linear process that takes place during an earthquake.

7 8112 9 001f'

\

f ir'.' ,Ed3 Wright - September 28, 1978

\CRS .

Irrespective of the above comments I believe the authors of Ref. 1 have done a magnificant job in advancing the state-of-art of carthquake maximum response prediction at least in a qualitative To accept the quantitative prediction I believe it is sense.

ndcessary to resolve the inability of the programs to predict the vertical component of the earthquake (consistent overprediction).

With respect to Ref. 2, I have reviewed the seismic analysis methods and models and find the same satisfactory, in fact far more detailed than normally found on other power plants.

The review of the analysis results of Ref. 2 is in progress and will be available by 1 November, 1978 (I am out of the country 30 September to 21 October, 1978).

/

Very truly yours,

)

• Senior Vice President, ces Engineering

\

\ *

\

\

\

\, .

4

\

\ \