Modeling thermally driven migration of brine in bedded salt

Abstract

Stress redistribution after the excavation of a drift leads to the generation of an Ed/DZ (excavation disturbed/damaged zone) in the near field of an opening, with significant changes in the hydraulic and mechanical properties. Further changes can occur under thermal load during the heating period in the post-closure phase of a repository for high-level radioactive waste. Initially more or less randomly distributed intragranular and intergranular fluid in a low-permeability sedimentary rock such as bedded or domed salt can then be mobilised and migrated under the altered hydro-mechanical and the coupled thermo-hydro-mechanical-chemical conditions at a potentially significant rate towards the excavation. To investigate the fluid migration behaviour, a test program BATS (Brine Availability Test in Salt) was carried out as a collaboration between Sandia National Laboratories (SNL), Los Alamos National Laboratory (LANL), and Lawrence Berkeley National Laboratory (LBNL) for US Department of Energy Office of Nuclear Energy in the underground facility WIPP (Waste Isolation Pilot Plant), Carlsbad, NM. Within the international cooperative project DECOVALEX-2023, data from the BATS experiment was systematically analysed by international teams using different model concept. Based on the multi-scale modelling strategy developed during DECOVALEX-2019, the BGR/UFZ team is analysing different type of measured data, including the inflow from the Small-Scale Brine Inflow test, the permeability distribution around the excavation from the Small-Scale Mine-by experiment, and the temperature evolution and inflow from the BATS experiment. The zone of enhanced permeability after excavation, which builds the main pathway for the inflow, is approximately 2.5 times the opening radius. The distribution of the permeability in the near-field around an opening can be well predicted by a failure-index based permeability model. Using the thermo-hydro-elastic model taking into account the creep behaviour of the rock salt, a reasonable prediction of inflow can be obtained under heated and unheated conditions. The flow pattern under heated conditions is characterized by an increase in permeability for BATS 1a and a decrease in the pressure gradient for 1b. The observation of a ‘spike’ behaviour after turning-off the power in the experiment can be explained on a microscale by the cooling contraction of the salt crystal, which leads to a 2 OOMs (order of magnitude) increase in permeability due to tensile stress, but only at high pore pressure and for a short time.