Application of reliability methods to the design of underground structures

Abstract

Uncertainty in rockmass and in situ stress parameters poses a critical design challenge in geotechnical engineering. This uncertainty stems from natural variability (aleatory) due to the complex history of formation and continual reworking of geological materials as well as knowledge-based uncertainty (epistemic) due to a lack of site specific information and the introduction of errors during the testing and design phases. While such uncertainty can be dealt with subjectively through the use of conservative design parameters, this leads to a lack of understanding of the variable ground response and the selection of an over-conservative design that can have a negative impact on both the project cost and schedule. Reliability methods offer an alternative approach that focuses on quantifying the uncertainty in ground conditions and utilizing it directly in the design process. By doing so, a probability of failure can be calculated with respect to a prescribed limit state, providing a measure of design performance. When multiple design options are considered, reliability methods can be paired with a quantitative risk analysis to determine the optimum design on the basis of safety and minimum cost rather than subjective conservatism. Despite the inherent benefits of such an approach, the adoption of reliability methods has been slow in geotechnical engineering due to a number of technical and conceptual challenges. The research conducted pertaining to this thesis aims to address these issues and remove the perceived “cloak of mystery” that surrounds the use of reliability methods. The scientific and engineering research in this thesis was divided into four sections: (1) the assessment of uncertainty in geotechnical input parameters, (2) a review of reliability methods in the context of geotechnical problems, (3) the development of a reliability-based, quantitative risk approach for underground support design and (4) the application of such a method to existing case studies. The completion of these areas is critical to the design of underground structures and may bring about a shift in design philosophy in the geotechnical industry.