ML20069A056

Comments on ACRS 780615 & 0707 Meetings Re Use of Effective Peak Acceleration & Underestimation Response to Offshore Fault,Resulting in Large Reduction in tau-effect & Small Torsion Excitation

ML20069A056
Person / Time
Site:	Diablo Canyon
Issue date:	09/22/1978
From:	Luco J CALIFORNIA, UNIV. OF, SAN DIEGO, CA
To:	Mckinley J Advisory Committee on Reactor Safeguards
Shared Package
ML20027A248	List: ML20069A056
References
NUDOCS 7811010388
Download: ML20069A056 (5)

Text

.

. '7~ C 5 ,

UNIVERSITY OF CALIFORNIA. SAN DIECO . -

\

I mutstry . Davis a snv NE = tos Ascar.ss

• nrvertsio:

• SAN Det.CO

• SAtt FHANCISCO .< $ SANTA DAlulAAA

• SANTA CHL"3

.,e DEPARTMENT CF APPLIED MECllANiC3 AND

• ENGINEERING SCIENCES. MAIL CODE 3 010 LA JCLLA. CALIFCRNIA 9*093 .s ..,

si .

September 22, 1978 This ri;2rt m:icidsindirW.ir)' .

or c'te::ta::::a rf:'i :t-.d ti 4 , .w, , , I.'" '. ' . . . ,

ptsM:t:0P a;:Y.': fit:4:ar?. It err.: o: 5 .T- c.::: '- ;c.:

shoc!d c:t ha r:!: :-h;t!.3 r::!n 9.!.t).g d::um:stlias tsc3 c; plet:d., g , r.; p,,r IP -

Mr. John McKinley %C ;C '"."-; .' -

ACRS r' U.S. Nuclear Regulatory Commis sion Washington, D.C. 20555 ,.

DISTK;CU720 70 f.q 3 , -,:,,,;-

Dear John:

r Enclosed please find my conunents to the ACRS meetings of June 14-15 and July 6-7, 1978, re ga rding Diablo Canyon Nuclear Power Plant.

The. travel e:cpense and compensation forms have been sent to Ms. Dundr.

Sincerely, 4 e */

ysw C. u*J< m J. E. Luco JEL:bs Enc.

0 7817.910 W3 pt. U. t, + >

.t T.9.b cc

Comments on the ACRS Subcommittee Meeting of 6/14/78 regarding Diablo Canyon Nuclear Power Plant - J. E. Luco The concept of an ' effective' peak ground acceleration was discussed again during the meeting. In previous reports I have expressed the opinion that not enough justification has been presented for the use of ' effective' peak acceleration lower than the value recommended by the USGS. A recent study by Bertero, Mahim and Herrera ('A Seismic Design Implications of Near-Fault San Fernando Earthquake Records,' Earthquake Engineering and Structural Dynamics, Vol. 6, pp. 31-42,1978) provides additional information of the need to treat the use of ' effective' accelerations with utmost care.

After a detailed study of the Olive View Hospital buildings which suHered major damage during the San Fernando earthquake, Bertero et al.

conclude: r "Usually large velocities may be developed at near-fault sites. Methods for constructing elastic and inelastic design response spectra should reflect the larger vale.2s recorded at such sites. In particular, methods which use standard ground spectrum shapes and effective acceleration values smaller than the expected peak values may not be reliable at near-fault sites."

In addition, in the text of the paper the authors write:

"... some procedures reduce the peak ground acceleration of standard type earthquakes (sometimes by 50 percent or more) to an ' effective' accelera-ation level that occurs a suHicient number of times to effect the structural response. The use of these reduced acceleration values in conjunction with

. . . standard ground spectrum shapes . . . may lead to further underes'timation of the peak ground velocity and displacement at near-fault sites. Thus , to avoid unconservative designs, derivation of smoothed linear-elastic response spectra for nes:-fault sites should be based on realistic estimates of peak ground acceleration, velocity and displacement. This would result in higher values of ground velocity and displacement than are usually considered at p re s ent. " .

In seversi occasions it has been argued that the peak acceleration of 1.25 g recorded at Pacolma Dam does not refice: the damage potentis1, but, rather the local response associated with the irregular topography of the site.

The aforementioned study of Bertero et al. sheds come light into this matter.

Bertero et al. used modified Pacolma Dam records with peak accelerations of -

0.4 and 0.8 g and predicted permanent drifts for a nonlinear model of the Olive View Hospital of 5 and 20 in, respectively. Since the observed permanent drifts were in excess of 30 in, it seems that the peak acceleration at the site of the Olive View Hospital was not significantly lower than that recorded at Pacolma Dam. In my opinion these results indicate that the Pacoima records are repre-sentative of the motion in the near-fault region and that the USGS recommenda-tion of a peak acceleration of 1.15 g is not exe,ssive for a 7.5 magnitude carth-quake. .

e e

e O

e 4 e 5 e

e *

• O e

e 9

e 4

4 0

9

• Comn$ents on the ACRS Committee Meeting of 7/7/78 regarding Ciablo Canyon Nuclear Power Plant - J.E. Luco During the meeting Dr. Frasier consultant for the applicant indicated that he had conducted studies to establish the type of waves that would arrive at the Diablo Canyon Site from an offshore fault. He said that his studies .

showed that most of the excitation would be arriving in the form of almost vertically incident waves. If his calculations are correct, this indicates that the reduction of the input motion for the so-called tau-effect may not be justi-fled. This additional information, which was not made available to the Com-mittee in time, suggests that the use of an exaggerated reduction for tau-effect coupled with the use of a very small torsional excitation may underestimate the response of the plant.

If one considers the alternative of assuming that a significant portion of the energy a/ rives in the form of surface. waves, again one finds that the analysis presented by the applicant underestimates the response. I have indicated in a previous report that if one assumes horizontally incid:nt waves then the reduction for tau-effect used by the applicant is too large and the amount of torsional input conside :ed is too low. Dr. Seed consultant for the applicant mentioned at the meeting that consideration of Rayleigh surface waves could lead to an increase in the calculated response. Again, these studies of the response of the plant to Rayleigh surface waves that lead to a higher response were not made available to the Committee.

On the basis of the statements by Drs. Frazier and Seed, it becomes apparent that the reduction of the input motion by tau-effect is not justifled ,

for Diablo Canyon.

A considerable amount of time was devoted to the discussion of the amount of damping used in the analysis. Most of the information presented by NRC to justify the 7 percent of critical damping was taken from a report from Werne'r and Reddy (Equivalent Viscous Damping for Seismic Analyscs of Nuclear 11 ants, Report SAN /1011,103 prepared by Agbabian Associates for U.S. E. R. D; A. , 1975). This report gives considerable empha sis to the damp-ing values reported by Hart et al. (1973), Hart and Vesudevan (1975), and Iemura and Jennings (1974). The damping values reported in these studies .

i. .

• L~ ;, .

correspond to the attenuation characteristics of the complete structure-foun-dation-soil system and are heavily invluenced by, the effects of soII-structure interaction. Since the Diable Canyon Plant is suppo'rted on hard material the effects of soil-structure interaction will be small. In these conditions the damping calues used should reflect only that portion of the energy dissipated in the structure.

J e

9 e

G ems O

I .

i 9

4 t

t