ANALYSIS OF BENTONITE COLLOID GENERATION IN THE FEBEX GALLERY AT THE GRIMSEL TEST SITE (SWITZERLAND): GLOBAL EVALUATION OF THE EXPERIMENTAL DATA OBTAINED FROM 2006 TO 2014

Abstract

The role of colloids on radionuclide transport in a crystalline medium, still presents large uncertainties despite it has been largely analysed in the past. The scarceness of experimental data obtained under realistic conditions and the difficulties related to the interpretation and modelling of experimental data are amongst the main reasons of the existence of these uncertainties. The objective of this study is to analyse the possible formation and migration of bentonite colloids from the engineered barrier of a deep geological repository of radioactive waste under in‐situ conditions. The study was carried out at the FEBEX gallery at the NAGRA’s Grimsel Test Site (GTS, Switzerland). In this gallery the FEBEX experiment, which reproduced at a real scale a high‐level waste repository in granite was installed in 1996. This offered a unique opportunity of studying the chemical effects produced by the bentonite on the surrounding groundwater, including the formation and release of colloidal particles. The experimental work described here lasted from the year 2006 (in the frame of the EC‐FUNMIG Project) to 2014 (in the frame of the EC‐BELBAR Project) and consisted in the analyses of waters, from different intervals of different boreholes at different distances from the bentonite surfaces, to verify the existence of colloidal particles and to identify their nature. Many boreholes already existed, since the construction of the FEBEX experiments and others were specially drilled to the purposes of this study (FUN‐1 and FUN‐2 boreholes) and all were used for water sampling. The study showed that colloidal particles exist at the site, in concentration higher than that expected for Grimsel groundwater (around 0.2 ppm). However, it was shown that different artefacts existed, possibly caused by the excavation of new boreholes, which biased the identification of bentonite colloids. High amounts of metals (Zn, Ni, W, Cu, Pb..) and organic carbon (TOC) were identified. A large effort was done to evaluate the possible contribution of bentonite colloids, combining different experimental techniques. This task was not easy due to the low particle concentration and heavy contribution of contamination in a colloidal form. An important constraint for establishing the existence of bentonite colloids is related to the chemical signals that evidence the affections produced by the bentonite, for example electrical conductivity (or Na and Cl) increase. Elements like Al, Fe, Mg and Ti, were used as chemical indicators of the presence of clay. The aluminium concentration, which is a necessary condition to prove the existence of bentonite colloids in the water, is still at present quite low in all the studied intervals of FUN‐1 and FUN‐2. Only in a few cases, Al concentrations higher than the expected for Grimsel water were observed. The maximum concentration of bentonite colloids (considering that the entire aluminium signal comes from the bentonite) was around 0.8 ppm, but indeed it does not seem enough that these data can be considered representative of the presence of bentonite colloids. Not even in the intervals which water was clearly affected by the presence of the bentonite (as an example FUN1‐4), bentonite colloids were identified with certainty. Many possible agents limiting the stability and migration of bentonite colloids (Ca in fracture surface, presence of iron oxides, increase of electrical conductivity, low water flow rates) have been clearly observed, which support the general conclusion that bentonite colloids, possibly formed at the bentonite/granite interface, have not been mobilised in significant quantity, during the time‐frame of the experiment.