ML19256E328

Supplemental Testimony of J Kelleher Re NRC Methods Used to Estimate Strong Ground Motion

ML19256E328
Person / Time
Site:	Skagit
Issue date:	10/03/1979
From:	Kelleher J WOODWARD-CLYDE CONSULTANTS, INC.
To:
Shared Package
ML19256E326	List: ML19256E328 ML19256E329 ML19256E332 ML19256E335
References
NUDOCS 7911020105
Download: ML19256E328 (4)

Text

, ...

.? ,

Woodward Clyde Consultants SUPPLEMENTAL TESTIMONY CF JOHN KELLEHER RELATING TO ESTIMATES OF STRONG GROUND MOTION This supplemental testimony is to respond to the request of the Board for additional information on methods of determining ground motion parameters. Initially the staff would note that no technique for estimating strong ground motion is clearly superior to other methods under all conditions. The choice of method is usually determined by the type of information available for the specific seismic hazard or by site specific considerations. However, to help the Board in its understanding of the dif ferent tech-niques to estimate ground motion, wc provide some site speci-fic remarks on the Skagit site and some general remarks describing the position of the staff on techniques of esti-mating ground motion.

1. Large earthquakes occurring within the near surface environment (H <_ 15KM) are typically accompanied by surface rupture or sur f ace f aulting. For the specific evaluation of the SKAGIT site for potential sources of st rong motion an important consideration is the virtual absence of evi-dence inferring surface faulting. Indeed with the possible exception of a location on the Olympic Peninsula there JLs no evidence that surface rupture has ever accompanied a -

large shock during Quaternary time throughout the entire province in which the site is location.1 Thus while surface f aulting cannot be categorically precluded the evidence indicates that it is minor or absent throughout most of the province and that there is no reason to anticipate major release of seismic energy in a near-surface environment. .

The staff considers the Skagit site to be in the tectonic province formed by the overriding co:cinental lithosphere and bounded on the north by an extension through Vancouver Island of the zone of decoupling between the Explorer and Juan de Fuca plates, e the south by the zone of transition associated with the Mendocino triple junction, on the cast by the volcanic chain and on the west by oceanic lithosphere of the Pacific plate. T province. 1258 3 Dbounds are not included in th 7911 n20 /d f

J Woodward Clyde Consultants Page 2 By way of contrast, major carthquakes in central or southern California commonly are accompanied by surface faulting; major energy release has typically occurred at depths of 10 km or less.

The significance of this tectonic feature is that earthquakes of given magnitudes which have caused the most serious destruction have been extremely shallow earthquakes usually accompanied by sur f ace f aulting. For example the Tangshan, China earthquake of 1976 (M =7. 6 ) and the Managua, Nicaragua earthquake of 1972 (M=6.4) caused great devasta-tion. These earthquakes had major seismic energy release at extremely shallow depths as evidenced by the surf ace rupturing in the central part of the respective cities.

By contrast the largest earthquakes of the tectonic prov-inces of the northwest have consistently occurred at the base of the crust or below. Even for the largest earthquakes of the northwest no single observation of intensity has ever exceeded MM VIII. A key consideration, therefore, in the site specific evaluation of the proposed Skagit facil-ity is the absence of any evidence suggesting the likelihood that large earthquakes would be accompanied by surf ace f ault-ing and by inference by major seismic energy release in the near surface environment.

2. Some studies have estimated relationships among earth-quake magnitude, distance and peak horizontal ground acceler-ation. Such studies include Schnabel and Seed (1973), Tri-funac and Brady (1976), and U.S. Geological Survey Circular 795. Other studies have provided correlations between peak ground accelerations and seismic intensity. These include Trifunac and Brady (1975) and Murphy and O'Brien (1977).

Woodward Clyde Consultants Page 3 These latter sources are used if only intensity information is available, while the other studies are useful if esti-mates can be made of source parameters such as magnitude, focal depth and distance to generating structure.

If instrumental de termina tions of magnitude exist, then estimates of strong ground motion relationships should be used such as Schnabel and Seed (1973), Trifunac and Brady (1976) or U.S. Geological Survey Circular 795. If no instru-mental determination of magnitude exists for earlier earth-quakes, then intensity estimat 2s are the most direct kind of information. In such cases intensity-ground motion rela-tionships such as Trifunac and Brady (1975) or Murphy and O'Brien (1977) should be used.

No instrumental determination of magnitude is avail-able for the 1872 earthquake of the Pacific Northwest.

Only intensity estimates are available and these are gen-erally sparse. The staff concluded that the maximum inten-sity associated with the 1872 earthquake was intensity VIII MM. For reasons discussed below, the estimate of Trifunac and Brady (1975) of .25 g for intensity VIII MM would pro-vide a conservative reference for the specific purpose of bounding the ground motion associated with the 1872 earth-quake in accordance with Reg. Guide 1.60. ,

In the original Safety Evaluation Report (SER) the staff compared the 1872 earthquake to a se ies of attenua-tion curves for an earthquake of magnitude 7.5. This was done because at that time the series of attenuation curves (PSAR for WPPSS No. 3 Figure 2-5-57b) was the most complete '

set available. The set of curves was compiled for earth-

~

9

Woodward Clyde Consultants Page 4 quakes of magnitude 7.5. Thus even though the magnitude of 7.5 was larger than any estimated for the 1872 earthquake, the completeness of the set of curves justified some ccmpar-ison. At no time, however, did the staff make a determin-ation that the 1872 earthquake was of magnitude 7.5 and this fact was so stated in the SER.

The study by Murphy and O'Brien (1977) probably pro-vides better estinates of actual peak acceleration at var-ious intensities than does the study by Trif unac and Brady (1975). The reason is that Murphy and O'Brien (1977) had a larger data set and improved statistical techniques.

Nevertheless, the more conservative relationship of Trifunac and Brady (1975) is considered more appropriate for setting reference accelerations because at intensities V, VI and VII the Reg. Guide 1.60 spec tra determined by the mean of the peaks (i.e., the method of Trifunac and Brady) falls between the mean and 84 th percentile soectrum for the fre-quencies of interest (Agbabian Associates, 1977). For inten-sity VIII, Trifunac and Brady (1975) appears to be conser-vative for their cited data set, in that the mean of observed peak accelerations is .167 g whcteas their relationship predicts .256 g.

3. For near field location of earthquakes (within about ,

10 km for M=6 and within about 20 km for M=7), there is no generally accepted method for estimating strong ground motion. Each such situation requires an extensive and speci-fic examination. For earthquake locations between approx-imately 20 or 30 km to 100 km there is reasonable agreement on estimates of strong ground mot ion among many recent studies including Schnabel and Seed (1973), Trifunac and Brady (1976) and the U. S. Geological Survey Circular 795.

~

_