Comparison of the PSHA results of the 1993-EUS- update and the 1998-TIP studies for waste bar

Abstract

From 1981 to 1989, Lawrence Livermore National Laboratory (LLNL) developed for the Nuclear Regulatory Commission a method for performing Probabilistic Seismic Hazard Analysis (PSHA) in the eastern US with results documented in NUREG/CR-5250. Improvements in the handling of the uncertainties lead to updated results documented in the 1993-EUS-Update study (NUREG-1488.) These results presented substantial differences with the utilities sponsored study performed by the Electric Power Research Institute (EPRI, 1989.) In order to understand the differences between the two studies, the Nuclear Regulatory Commission (NRC), the Department Of Energy (DOE) and EPRI jointly sponsored a study led by the Senior Seismic Hazard Analysis Committee (SSHAC) the task of which was to explain the differences between the two studies and provide guidance on how to perform a state-of-the-art PSHA. The work and conclusions of the SSHAC are documented in NUREG/CR-6372, 1997. As a follow-up to the 1997 SSHAC study, the Trial Implementation Project (TIP), (UCRL-ID-133494, 1998, NUREG/CR-6607) made use of the SSHAC recommendations and developed a set of more detailed guidance for performing PSHA. The TIP project tested the more complicated issue of development of the seismic zonation and seismicity models on two sites: Watts Bar and Vogtle. It was found that the uncertainty generated by artificial disagreements among experts can be considerably reduced through interaction and discussion of the available data and by identifying the elements common to all experts' interpretation. By concentration on those elements, it was possible to develop a consensus of the group on the way to characterize them and eliminate large unnecessary differences. The present study compares the results of the 1993-EUS-Update and the 1998-TIP studies and identifies the reasons for the differences, which were found to be: (1) Differences in the ground motion (GM) attenuation models; (2) The introduction of the Eastern Tennessee Seismic Zone (ETSZ) in the TIP study. We found that these two factors accounted for factor of 6 difference in mean estimates PGA hazard at high GM levels. The agreement between the two studies improved at lower PGA values. The results were in better agreement and only differed by about a factor of two at high ground motion levels, when the same GM model was used with each seismicity model. Finally, it was found that the composite rate of earthquakes around the Watts Bar site was about a factor of two higher for the TIP composite seismicity model than for the composite 1993-EUS-Update seismicity model.