Probabilistic failure analysis of the Shek Kip Mei landslide in Hong Kong

This paper presents a probabilistic failure analysis approach that uses failure samples generated in Monte Carlo Simulation (MCS) and analyzes these failure samples to assess effects of various uncertainties on slope failure probability. The probabilistic failure analysis approach contains two major components: hypothesis tests for prioritizing effects of various uncertainties and Bayesian analysis for further quantifying their effects. The approach is implemented in a spreadsheet environment and applied to analyze the 1999 Shek Kip Mei landslide in Hong Kong. Hypothesis tests show that the uncertainty of pore water pressure ratio has the most significant effect on slope performance and its effect is further quantified by Bayesian analysis. Bayesian analysis results are validated against independent sensitivity studies. It is shown that probabilistic failure analysis provides results that are equivalent to those from additional sensitivity studies, but it avoids additional computational times and efforts for repeated runs of MCS in sensitivity studies.