← Back to Sessions

How to Present and Interpret Probabilistic Fault Displacement Hazard Analysis (PFDHA) Results to Improve Engineering Decision-Making

Session: Seismic Hazard Analysis for Critical Infrastructure

Type: Oral

Date: 4/19/2021

Presentation Time: 10:45 AM Pacific

Description: 

Our recent experience indicates that there are important concepts associated with application of PFDHA hazard curves for engineering evaluation of critical facilities that need more attention. For example, it is common in project design criteria to adopt probabilities of exceedance for fault displacement hazard that are equal to those selected for ground motion hazard (e.g., the 2% in 50 year). In tectonically active areas, these probability levels commonly yield ground motions that represent larger than average shaking for controlling earthquake magnitudes and distances (positive epsilon values in deaggregation plots). However, it is common in PFDHA for these probability levels to correspond to negligible hazard (the mean hazard curve lies below the selected probability level), or the displacement is less than the median expected outcome (i.e., a negative epsilon value). This result is due to the low slip rates (a few millimeters or less) of many active faults and the fact that in PSHA many seismic sources contribute to hazard whereas in PFDHA hazard is usually dominated by a single fault source. Evaluating facility performance against a mean negative epsilon displacement may not give a complete view of risk, especially given the large epistemic and aleatory uncertainties in current PFHDA models. We propose in PFDHA practice to clearly document the epsilon values associated with mean displacements at different probability levels and to plot fractile hazard curves that highlight the immature state of fault displacement model development. We encourage practitioners to consider whether to adopt mean hazard results in cases of negative epsilons, or whether to evaluate critical facilities against higher hazard fractiles and/or a “floor” of a median deterministic displacement for projects that are known to cross an active fault. We believe such additional evaluations will lead to a better understanding of risk mitigation and evaluating performance versus cost.

Presenting Author: Stephen C. Thompson

Student Presenter: No

Authors