Geotechnical modelling of a deep tunnel excavation in the Boom Clay formation

Abstract

In this study a deep tunnel excavation in the Boom Clay (BC) formation, in the context of constructing a repository for the deep geological disposal of radioactive waste in the Netherlands, is investigated by means of numerical modelling. First the selection of a constitutive soil model, that captures adequately the known aspects of the mechanical behaviour of natural BC, is addressed. To this purpose, three conventional, drained, strain-controlled triaxial compression tests and one high pressure oedometer test, performed on intact BC samples originating from Essen (Belgium) which were extracted from the literature, were numerically simulated utilising the lab test facility of the PLAXIS 2D 2011 software. The Linear Elastic Perfectly Plastic-Mohr Coulomb model, the Modi?ed Cam-Clay model, the Soft Soil-Creep model and the Hardening Soil (HS) model were calibrated on the basis of the test results. Based on the comparison of the numerical results with the test data, it was concluded that the HS model performed best, especially in terms of reproducing the evolution of sti?ness and mobilised shear strength under triaxial compression, which are deemed to be the most predominant processes a?ecting the behaviour of the BC host rock due to a tunnel excavation. Subsequently, the level of in?uence of the geotechnical properties of BC, used as input parameters for the HS model, on the radial extent of the fully Plastic Zone (PZ) and of the Hydro-Mechanical (HM) disturbance, caused by an undrained tunnel excavation, is investigated. The level of in?uence of the soil properties on the magnitude of the hoop forces (N) in the tunnel liner and on the magnitude of the generated pore water pressures (u) around the excavation is examined as well. This was achieved by numerically simulating an undrained tunnel excavation in the BC host rock, in two-dimensional plane strain conditions, with the PLAXIS 2D 2011 ?nite element program. A mechanical sensitivity analysis was performed by varying, individually, selected HS model input parameters to upper and lower bound values. It was found that the thickness of the PZ was mainly in?uenced by the magnitude of the elastoplastic shear sti?ness of the material. The radial extent of the HM disturbance was mostly a?ected by the elastic shear sti?ness of the soil. Finally, the magnitude of N in the liner and the magnitude of u in the vicinity of the excavation were found to be primarily in?uenced by the value of the e?ective friction angle (?') of the BC material.