Validation of PSHA results by deterministically derived design earthquakes

摘要

German guidelines for Nuclear Power Plants (KTA 2201.1 [1]) are under revision. Fundamental recommendations for the redrafting were elaborated by a specialist's panel, combining seismologist, engineers and members of the different control and regulatory institutions [2]. The still valid guideline (revised in 1990) is mainly focused on a deterministic approach, being particular for German N.P.P design practice. Therefore, the essential changes are related to the explicit incorporation of Probabilistic Seismic Hazard Assessment procedures (PSHA) and its results in terms of hazard deaggregation, fractiles of spectral and ground motion parameters as well as (macroseismic) site intensities. There is a consensus among the involved experts that probabilistic and deterministic approaches (DSHA) are of similar importance. The definition of the design earthquake(s) shall be based on both approaches. An annual occurrence rate of 10~(-5)/year is recommended as the appropriate design level combined with the median values of hazard describing parameters. Concerning the relation between the results of PSHA and DSHA two major questions are of interest: (1) How to act if both approaches lead to (slightly) different results? (2) Is there a need to introduce further safety margins in case of DSHA? Deterministic Seismic Hazard Assessment (DSHA) becomes - despite of its inherent limits - of importance in cases when the PSHA suggests (as a consequence of methodological aspects) unusually high design parameters and when these results have to be validated. Within the comprehensive study, the relation between PSHA and DSHA is investigated for six model sites, covering the seismic and site conditions for German N.P.P. Because of the diverseness of constellations the results of the model points can be used in general for a number of different site conditions.