Effect of a nitrate plume on migration in clay rock. Experimental study at a pluridecimetric scale

Document available in extended abstract form only. An important issue for the development of nuclear industry arises from the most dangerous waste from the back-end of the fuel cycle. The Callovo-Oxfordian formation (COx, Bure, Meuse/Haute-Marne, France) is investigated as a potential host rock for deep geological disposal by the National Radioactive Waste Management Agency (ANDRA). There is already a good understanding of COx clay chemistry, ion migration in COx, and a large dataset on their transfer parameters. However, there is still a lack of knowledge concerning behaviour and properties of co-contaminants and their potential effect on COx confinement properties. In particular, salinity gradients will take place at dm-m scale around salt rich medium activity long lived radioactive waste. Thus, one needs to evaluate the propagation of a saline plume eventually released by waste packages and quantify the potential effects of such macroscopic perturbation. This study was developed within the laboratory group 'TRANSFERT' led by ANDRA. Two experiments at pluri-deci-metric scale were dedicated to quantify the effect of a saline plume on the migration of neutral, anionic and cationic tracers in COx clay. Two clay samples were cored from borehole DIR 1004 (Bure, depth: -496 m, no. 1004-27320), namely 'S1' and 'S2', with dimensions Diam. x H = 29 x 25 cm2. These samples were first of all drilled in order to set an internal diffusion chamber and sealed in steel diffusion cells. Samples were re-hydrated over a year with a synthetic pore water solution of small volume compared to poral volume. This step ensured the hydration of the rock, in order to reach an equilibrium state close to in situ conditions. It also ensured that the transport of tracers is mainly diffusive. Then a solution with extra sodium nitrate ([NaNO3] ∼ 0,5 mol.L-1) and several radiotracers was injected in the hydraulic system in contact with the rock. Diffusion of HTO and 36Cl- was studied in 'S1' and diffusion of 22Na+ and 137Cs+ in 'S2' for several months, simultaneously with the nitrate plume, and the decrease of tracer concentrations was monitored in the hydraulic system. When the diffusion period was finished, experiments were dismantled and concentration profiles were measured in the rock. Meanwhile, HTO, 36Cl-, 22Na+, 137Cs+, Na+ and NO3- concentration profiles were obtained along X, Y and Z directions. We used both experimental results (monitoring of the tracer concentration in the injected solution and post-mortem concentration profiles in the rock) to determine transfer parameters values (De, εa, and Kd) for each species in COx clay in the presence of the nitrate plume. A direct modelling of the experimental data was performed using the simulation platform ALLIANCES. Simulations were performed with a 3D mesh including the geometry of the experiments and the exact shape of the solid solution interface in the injection chamber. A classical transport code was used to calculate the evolution of concentration of tracers, including poral water and sorption on clay, in order to fit adjustable parameters for each tracer: effective diffusion (De) and accessible porosity (ea) or solid solution distribution ratio (Kd) for sorbing elements. Sorption was also modelled with an experimental non linear sorption isotherm (Kd = f[C]) when necessary. These fitted transport parameters are compared to classical data obtained in cm-thick pristine rock samples. This work gives valuable results to quantify effect of ionic strength on diffusion of ions in clay rock. It also provides an interesting experimental setup with deci-metric samples, intermediate between centimetric laboratory cells and in situ experiments at metric scale, thus allowing experimental conditions similar to the expected near-field perturbations, among which the nitrate plume. Results confirmed the efficiency of sorption-diffusion model to reproduce 3 dimensional effects, like anisotropy or solid/solution interface geometry on the diffusion in the rock, as well as the spatial heterogeneities induced by the nitrate plume. (authors)