Optimization of Site-Exploration Programs for Slope-Reliability Assessment

Site investigation through boreholes, such as vane shear tests or cone penetration tests, provides an effective means to assess the performance of geotechnical sites. Boreholes should be located such that they provide a maximum amount of information. This study develops a systematic approach for optimizing site-exploration programs based on maximizing the value of information. To compute the value of information, test data are randomly generated based on the prior model of soil parameters. Bayesian updating of the slope reliability is carried out by application of an adaptive version of the Bayesian updating with structural reliability methods (BUS) in combination with subset simulation, which is efficient in tackling problems with discrete random field representations. The approach is used for the reliability assessment of a slope in undrained clay, wherein the site-exploration program is optimized with vane shear tests to identify the undrained shear strength. Results show that the approach is able to identify the optimal location and number of boreholes.