Analysis of colloids erosion from the bentonite barrier of a high level radioactive waste repository and implications in safety assessment

To investigate the dominant mechanisms of colloid formation from compacted and confined bentonite innovative experiments were conducted. Chemical or physical processes that can affect the erosion of the bentonite surface were analyzed (ionic strength of the water, Ca in the water and in the exchange complex of the clay, dry density of the clay and presence of a water flow rate at the bentonite surface).Hydration, swelling and extrusion of clay into pores or fractures are primary steps for the formation of free colloidal particles in the aqueous phase, and the chemistry of the clay/water system is the most important parameter controlling the generation and stability of colloids. Ca-bentonite formed colloids quantities below the detection limit of our techniques, even in deionised water, but a percentage of Na approximately 20–30% in the clay exchange complex, as that present in the FEBEX bentonite, is enough to allow the formation of colloidal particles in quantities very similar to those produced by the Na-bentonite.The results for bentonite colloid generation obtained at a laboratory scale allowed the estimation of a range of colloid generation rates under different chemical conditions. Results were compared with in situ experimental investigations carried out at the FEBEX gallery emplaced in a granite massif at the Grimsel Test Site (Switzerland). The quantitative analysis of laboratory and in situ data can be used as input for models and performance assessment (PA) of high level radioactive waste (HLRW) repositories.

► Colloids may affect radionuclide migration in a deep geological repository (DGR).
► Colloid formation from the bentonite barrier of a DGR was investigated by novel tests.
► Salinity of the water and Ca concentration affect colloid formation and stability.
► A range of colloid generation rates under different chemical conditions was estimated.
► Laboratory and in situ data: input for models for DGR safety assessment.