Reliability-based assessment of stability of slopes

Multiple sources of uncertainties often exist in the evaluation of slope stability. When assessing stability of slopes in the face of uncertainties, it is desirable, and sometimes necessary, to adopt reliability-based approaches that consider these uncertainties explicitly. This paper focuses on the practical procedures developed recently for the reliability-based assessment of slope stability. The statistical characterization of model uncertainty and parameter uncertainty are first described, followed by an evaluation of the failure probability of a slope corresponding to a single slip surface, and the system failure probability. The availability of site-specific information then makes it possible to update the reliability of the slope through the Bayes' theorem. Furthermore, how to perform reliability-based design when the statistics of random variables cannot be determined accurately is also discussed. Finally, case studies are presented to illustrate the benefit of performing reliability-based design and the procedure for conducting reliability-based robust design when the statistics of the random variables are incomplete.