A geostatistical solution to estimating ground-water inflows in deep rock tunnels with validation through case studies

Estimating ground-water inflows in rock deep tunnels is critical to the safety of personnel constructing the tunnel and the cost of construction. This paper presents two case studies, including the Chattahoochee and Nancy Creek tunnels, for estimating steady-state ground-water inflows into deep rock tunnels. Statistical methods were applied to packer test data obtained from exploratory borings during geotechnical site investigations, and a semiempirical procedure was used to estimate ground-water inflows. The Chattahoochee and Nancy Creek tunnels are deep rock tunnels that exhibit radial flow conditions. For the radial flow condition, a statistical analysis presented in this paper indicates that the semiempirical procedure works well; however, variations in the procedure and/or an insufficient amount of testing could lead to underestimations of the inflow quantities. Modeling using Monte Carlo simulations was observed to be an effective tool for removing data distribution irregularities and incorporating high-end permeability data to derive reasonable ground-water inflow estimates.