ML19271A944
| ML19271A944 | |
| Person / Time | |
|---|---|
| Site: | Vallecitos File:GEH Hitachi icon.png |
| Issue date: | 06/20/1980 |
| From: | Ang A S ILLINOIS, UNIV. OF, URBANA, IL |
| To: | Igne E Advisory Committee on Reactor Safeguards |
| References | |
| NUDOCS 8008110267 | |
| Download: ML19271A944 (2) | |
Text
- " h :iversity of L inois at Ui aana-C7ampaigri DEPARTMENT OF CIVl1 ENGIN!! RING UEANA, tiltNOIS 61601 (217) 333-3812 2
i ECGt, a June 20, 1980 CM /J g 33 T0:
Mr. E. Igne
. lSW L* ' rig.CC ;M FROM:
A. H-S. Ang Jich CC.[Co[,EE CIl 3
ng3 RE:
Comments on Subcommittee Meeting on GETR of 16-17 June 1980 My coments are principally limited to questions of probabilities and the models used to calculate such probabilities.
(1) From all indications, the probability of significant offset (e.g. one meter) at the foundation of the General Electric Test Reactor (GETR) appear to be small, of the order of 10-5 per year.
On the basis of an " intuitive" approach and simple calculations, this probability appears reasonable.
For instance, on the basis of available information, including the geologic infonnation of strain rates on the shears, the annual probability of a one meter offset on the shears would be of the order of 10-4 per year. With this, and the. conditional probability of 0.06 for an offset underneath the GETR foundation given an offset around the site, the total probability of a significant offset underneath the GETR foundation therefore is of the order of 10-5, No complex model is really necessary to arrive at this number for the probability of a significant (1 meter) offset.
In fact, because of the lack of pertinent information and proper data to evaluate the input parameters necessary in the more complex model of JBA as well as the nodel of TERA /LLL, the results obtained by these more complex models are not any more credible than the results obtained with "the intuitive model".
(2) The probability model used by J. Benjamin and Associates to develop the interaction curve for combining earthquake acceleration and
" cantilever" under the foundation needs to be re-examined and revised.
Obviously, the cantilever is a consequence of an offset occurring under-neath the GETR foundation.
An offset, however, is heavily correlated with the earthquake acceleration, whereas the JBA model for the interaction curve is based on the assumption that the acceleration and offset are statistically independent.
More correctly, the interaction curve ought to be developed on the basis of the assumption that the offset and acceleration are correlated.
The shape of the interaction curve based on this assumption can be expected to change from that obtained thus far by JBA. This may be done by prescribing a reasonable joint probability distribution for the offset size and acceleration level, with various values of correlation coefficient.
8008110$6[
~tr 2
I do not expect that the resuits will change any of t,he present conclusions relative to combined acceleration and offset; however, the probability basis for combining offset effect and acceleration ought to be made more consistent.
A cc:
Dr. W. Kerr Dr. C.P. Siess Dr. D. Okrent t
I
.,-