HR: 14:40h
AN: S13C-04    [Abstracts]
TI: Challenges to Seismic Hazard Analysis of Critical Infrastructures
AU: * Kl�gel, J
EM: jkluegel@kkg.ch
AF: Nuclear Power Plant Goesgen, Kraftwerkstrasse, Daeniken, 4658 Switzerland
AB: Based on the background of the review of a large scale probabilistic seismic hazard analysis (PSHA) performed in Switzerland for the sites of Swiss nuclear power plants- the PEGASOS project (2000-2004) - challenges to seismic hazard analysis of critical infrastructures from the perspective of a professional safety analyst are discussed. The PEGASOS study was performed to provide a meaningful input for the update of the plant specific PRAs (Probabilistic Risk Assessment) of Swiss nuclear power plants. Earlier experience had shown that the results of these studies to a large extend are driven by the results of the seismic hazard analysis. The PEGASOS-study was performed in full compliance with the procedures developed by the Senior Seismic Hazard Analysis Committee (SSHAC) of U.S.A (SSHAC, 1997) developed for the treatment of uncertainties by the use of a structured expert elicitation process. The preliminary results derived from the project did show an unexpected amount of uncertainty and were regarded as not suitable for direct application. A detailed review of the SSHAC-methodology revealed a number of critical issues with respect to the treatment of uncertainties and the mathematical models applied, which will be presented in the paper. The most important issued to be discussed are: � The ambiguous solution of PSHA-logic trees � The inadequate mathematical treatment of the results of expert elicitations based on the assumption of bias free expert estimates � The problems associated with the "think model" of the separation of epistemic and aleatory uncertainties � The consequences of the ergodic assumption used to justify the transfer of attenuation equations of other regions to the region of interest. Based on these observations methodological questions with respect to the development of a risk-consistent design basis for new nuclear power plants as required by the U.S. NRC RG 1.165 will be evaluated. As an principal alternative for the development of a meaningful design basis contemporary methods of deterministic seismic hazard analysis will be discussed. It will be demonstrated that modern deterministic methods can be employed successfully both for the design of critical infrastructures as well as a starting point for the development of modern probabilistic scenario-based methods. As a challenge for future developments the question of an alternative presentation of the results of a seismic hazard analysis will be considered. In addition to traditional seismic hazard maps based on spectral or peak ground accelerations it is suggested to develop similar maps in terms of Arias-intensity and of CAV-values. This will support the selection of appropriate earthquake time histories by civil engineers responsible for the design of critical infrastructures.
DE: 7299 General or miscellaneous
SC: Seismology [S]
MN: Fall Meeting 2005