Probabilistic Assessment with Heterogeneity Considerations in Numerical Modelling of Squeezing Tunnels

Estimation of tunnel behaviour and designing a suitable support system is a major challenge in tunnel engineering. This problem is particularly important for squeezing tunnels as there are high chances of failure of support system. This is due to high degree of uncertainty in estimating rock mass behavior in terms of strength and deformation potential. In this paper, probabilistic approach is applied to a case of squeezing tunnel under high overburden excavated in weak rock mass conditions. Probabilistic analysis is performed using Monte-Carlo Simulations (MCS) and numerical modelling with spatial heterogeneity (NMSH) on the data collected. Convergence-confinement method is used to perform MCS. The PDF is generated for input parameters and the values are randomly sampled to be used in MCS and NMSH. Results shows that output (convergence and support pressure) from probabilistic estimates frequently exceed that calculated from deterministic method. Also, considerable scatter in output values from both probabilistic methods applied here are obtained. Comparison of both probabilistic approaches shows that less scatter of output values in NMSH as compared to MCS method of analysis.