No abstract

The Seismic Safety Margins Research Program (SSMRP) is a multiyear, multiphase program whose overall objective is to develop improved methods for seismic safety assessments of nuclear power plants, using a probabilistic computational procedure. The % program is being carried out at the Lawrence Livermore National Laboratory and is sponsored by the U.S. Nuclear Regulatory Commission, Office of Nuclear Regulatory Research. Phase I of the SSMRP was successfully completed in January 1981: A * probabilistic computational procedure for the seismic risk assessment of nuclear power plants has been developed and demonstrated. The methodology is implemented by three computer programs: HAZARD, which assesses the seismic hazard at a given site, SMACS, which computes in-structure and subsystem seismic responses, and SEISIM, which calculates system failure probabilities and radioactive release probabilities, given (1) the response results of SMACS, (2) a set of event trees, (3) a family of fault trees, (4) a set of structural and component fragility descriptions, and ( 5) a curve describing the local seismic hazard. The practicality of this methodology was demonstrated by computing preliminary release probabilities for Unit 1 of the Zion Nuclear Power Plant north of Chicago, Illinois.Studies have begun aimed at quantifying the sources of uncertainty in these computations. Numerous side studies were undertaken to examine modeling alternatives, sources of error, and available analysis techniques. Extensive sets of data were amassed and evaluated as part of projects to establish seismic input parameters and to produce the fragility curves.

Various magnitude scales commonly used and the interrelationships among them are reviewed. It is shown that problems exist with all of the magnitude scales being used in the United States. When using regional catalogs, for example, it is often necessary to determine how the reported magnitudes were determined. Often such information is not available, although the potential errors are quite large. Both the MS and the mb scales were designed to be universal scales. However, both MS and mb magnitudes are often determined beyond the applicable range of the equations used to define the two scales. Furthermore, the MS magnitudes are not generally available for moderate to small earthquakes in most earthquake catalogs. The mb magnitudes are more generally available than MS values; however, there is also much greater variation in the way mb is determined. In particular, a significant change in the mb scale occurred in the early 1960’s when the World‐Wide Standard Seismograph Network (WWSSN) was established. This change in instrumentation used to determine mb values had a significant effect on estimated magnitudes (post‐1960 values are lower) and the saturation level of the mb scale. The older, longer‐period instruments recorded larger mb magnitudes than can be recorded with the WWSSN instruments. In addition, great care must be taken when selecting the mb magnitudes of western U.S. earthquakes, because the values are often in considerable error owing to the fact that they were determined at distances less than 25° and were not properly corrected for attenuation in the upper mantle, or asthenosphere. The seismic body wave magnitude mb of an earthquake is strongly affected by regional variations in the Q structure, composition, and physical state within the earth. Therefore because of differences in attenuation of P waves between the western and eastern United States, a problem arises when comparing mb values for the two regions. A regional mb magnitude bias exists which, depending on where the earthquake occurs and where the P waves are recorded, can lead to magnitude errors as large as ⅓ unit. There is also a significant difference between mb and ML values for earthquakes in the western United States. An empirical link between the mb of an eastern U.S. earthquake and the ML of an equivalent western earthquake is given by ML = 0.57 + 0.92(mb)east. This result is important when comparing ground motion between the two regions and for choosing a set of real western U.S. earthquake records to represent eastern earthquakes.

A thorough review of the geology and past seismic history of the area surround ing the Lawrence Livermore Laboratory site is given. This information is used to identify those active faults on which destructive earthquakes could occur. Earthquake magnitudes are assigned to these postulated earthquakes, and the approaches used to assign magnitudes are discussed. Estimates of maximum possible ground This report will define the earthquake hazard to which the facilities at the Lawrence Livermore Laboratory in Livermore, California might be subjected. The location and magnitude of earthquakes that could cause damage to Laboratory facilities will be postulated and ground motions at the Laboratory site resulting from these postulated earthquakes de veloped. These site ground motions are described in terms of maximum ground acceleration (peak g level), acceleration response spectra, duration, and accelerograms. Unfortunately, defining earthquake hazard for structural design calculations is not subject to the same degree of cer tainty or the same exercise of judgment as are methods of design analysis. For this acceleration, response spectra, and duration of the strong shaking at the Laboratory site resulting from the postu lated earthquakes are given. A discussion of the methods used to predict these quantities is also presented. Using the foregoing information, design-basis earthquakes, which define the ground motion for Laboratory structural design purposes, are devel oped. reason, such factors as risk evaluation, probability, economics, and good engi neering judgment are essential elements in earthquake engineering design pro cedures. Definitive results cannot be given and considerable judgment must be used; however, this study will develop reasonable estimates of the maximum ground shaking to which the Laboratory facility could be subjected, an estimate of the possibility of suca occurrences, and a design-basis earthquake (DBE) with general comments on its-ilication to structural design. Considerable detail as to the methods used in the development of the estimates and the DBE will be given. Although many investigators have studied the probability of occurrence of earthquakes, their studies are not yet

No abstract