Hydration of the bentonite buffer in a KBS-3 repository: Fid-Bau Portal

Hydration of the bentonite buffer in a KBS-3 repository

Hökmark, Harald

AbstractThe present study addresses the problem of bentonite hydration in a KBS-3 type repository. A finite element program, Code_Bright, that has been specifically developed to analyze coupled Thermo-Hydro-Mechanical (THM) processes in unsaturated porous media, is used to estimate the time required to reach full saturation of the bentonite buffer. The finite element model analyzed here concerns a horizontal slice of the annular bentonite-filled space between canister and rock wall at canister midheight and considers T-H aspects of the bentonite material behavior. Processes included in the model are transport of heat and moisture. Moisture movements caused by suction gradients as well as by thermally induced vapor mass fraction gradients are considered. All these T-H transport processes are coupled and highly nonlinear. In particular, the thermally driven vapor transport is sensitive to temperature gradients and to temperature levels. Therefore, an important issue is to determine relevant thermal boundary conditions for the FEM model. This is done here by use of analytical expressions which take the initial canister power, the fuel power decay characteristics, the repository layout and the rock thermal properties into account. Results are presented that show the buffer saturation as function of distance to the canister surface at different times. Effects of incomplete saturation on the temperature development are analyzed for a number of assumptions regarding bentonite properties. In particular are conditions that may lead to delayed saturation, such as low permeability, high vapor diffusivity, etc., analyzed. For the reference KBS-3 repository, a general result is that the buffer filling out the annular space between canister and rock wall will be fully water saturated after about 3 or 4 years, provided that there is access to water at the buffer/rock interface of the individual deposition holes. This general result seems to apply for a wide range of assumptions regarding values of transport parameters, gas confinement, porosity, canister power etc. There are, however, a number of uncertainties in the material model. Some of these are either unimportant or can be handled conservatively, while others, such as the validity of the saturation/suction relations are concluded to be important and will need further attention in the future. A general conclusion is that the saturation time scale may turn out to depend more on hydrological conditions around individual deposition holes, on the hydrological buffer/rock interaction and on the repository-scale restoration of the ground water pressure after repository closure, than on details of the bentonite material model.