Development of a methodology for probabilistic alignment selection and design for drill-and-blast hard-rock tunnels

Abstract

Transportation tunnels are heavily studied and debated before they are eventually constructed. The lead time on tunnel projects is typically years and sometimes decades. During the early phases of tunnel planning, it is often necessary to evaluate a wide range of alignment, excavation method, and support system options to find those which are technically feasible and cost-effective. However, comparing a wide or infinite range of alignment, excavation method, and support system options can be time-consuming, tedious, and expensive. Tunnel alignments are frequently selected not based on any quantitative evaluation and comparison of feasibility but on professional judgment or political whim. There are many components to a tunnel's feasibility including highly uncertain ground conditions, many possible excavation methods, and variable support options. The goal of this research project is to bring some quantifiable metrics to the comparison of alternative tunnel options.

To accomplish this goal, the author has developed a 3D implicit geological model of the EJMT case study in Colorado which includes engineering geologic parameters of interest such as rock mass classification parameters Q and RMR. This model includes quantified uncertainty in each parameter. Using this probabilistic geologic model a new alignment optimization algorithm which seeks the best possible tunnel alignment through a given volume is developed. Further, a numerical finite difference model was calibrated to available EJMT case study data. This calibrated model validates the results of the alignment optimization algorithm. The result is a novel and useful tool and methodology which can effectively evaluate and compare many plausible alignments and select the best possible option for further design studies.