Efficient system reliability analysis of soil slope using generalized subset simulation

Various uncertainties exist in slope stability analysis. These uncertainties can be rationally taken into account in a probabilistic framework, where the plausibility of slope failure is quantified as the occurrence probability of soil masses sliding along a slip surface. Slope failure mechanisms, e.g., locations of potential slip surfaces, are unknown prior to slope stability analysis. Several possibilities are usually assumed in slope stability analysis, leading to a number of possible failure modes. Different failure modes may have different contributions to slope failure. How to account rationally for various failure modes in slope stability reliability analysis and how to identify key failure modes that have significant contributions to slope failure are critical questions in slope engineering. These questions can be rationally addressed under a system reliability analysis framework using direct Monte Carlo simulation (MCS). However, it is well-known that direct MCS suffers from a lack of efficiency and resolution at small probability levels though it has advantages of simplicity and robustness. This paper develops an efficient computer-based simulation method for slope system reliability analysis using an advanced MCS method, called generalized Subset Simulation (GSS). The proposed approach decomposes a slope system failure event into a series of sub-events representing possible failure modes and calculates their failure probabilities by a single run of GSS. Based on the reliability analysis results from GSS, representative failure modes that are considered relatively independent are identified from possible failure modes using probabilistic network evaluation technique. Their relative contributions are then assessed quantitatively, based on which key failure modes are determined. The proposed approach is illustrated using a practical engineering slope, Congress Street cut slope in Chicago. Results show that the proposed approach provides proper estimates of failure probabilities of slope system and different failure modes by a single GSS run, which avoids repeatedly performing simulations for each failure event. Compared with direct MCS, the proposed approach significantly improves computational efficiency, particularly for failure modes with small failure probabilities. Key failure modes of slope stability are determined among possible failure modes in a cost-effective manner. Such information is valuable in making slope design decisions and remedial measures.