Reliability of slopes in sensitive clays

Abstract

Risk and probabilistic analyses have now had enough applications that make them effective to use in practice. The approach provides more insight than deterministic analyses alone. They help reduce uncertainty and focus on safety and cost-effectiveness. The paper illustrates the use of reliability methods for the analysis of slopes in sensitive clays with examples of the calculation of probability of failure and run-out for the Finneidfjord and Rissa landslides in Norway. The input, model and results of the probabilistic slope analysis are described, including the uncertainties in the parameters, triggers and calculation model, as well as a brief review of the principles of the reliability approach. Reliability approaches do not remove uncertainty nor do they alleviate the need for judgment. They provide a way to quantify the uncertainties and to handle them consistently. Site investigations, laboratory test programs, limit equilibrium and deformation analyses, instrumentation, monitoring and engineering judgment are necessary inputs to the reliability approach. Landslide events, often unwittingly, are triggered or aggravated by human activity, such as change in topography (e.g. excavation or surcharge) and change in drainage conditions. Climate change can increase the frequency of landslide. The paper proposes that a probabilistic model in an event tree format should be included to ensure that all failure modes and the uncertainties have been covered and that slope failure mitigation measures are quickly available.