Revisitation of Migration Experiments: Final report

The transport parameters of solutes, diffusion coefficient and retardation factor, in Boom Clay can be determined using various types of laboratory experiments. To gain confidence in the determined values, it is important to assess the consistency of the different used set-ups. The first aim of this work was to investigate the consistency in the transport parameters of the non-reactive tracers, HTO and 125I, estimated via various laboratory migration experiments, through-diffusion (TD), pulse injection and back-to-back experiments. The consistency was checked for different BC samples, i.e. clay-rich vs silt-rich, with orientation parallel or perpendicular to the bedding. The second objective was to assess the sensitivity of the set-ups to some selected factors: the design of the cell in pulse injection experiments, the size of the BC cores in pulse injection and TD experiments and the experimental set-up in TD experiments, with the use of the two non-reactive tracers HTO and 125I on clay-rich BC sample with a perpendicular orientation to the bedding. The consistency study showed a good agreement in the apparent and effective diffusion coefficients and accessible porosity for HTO which didn’t vary for more than a factor 2 and 1.6, respectively between the three set-ups investigated, i.e through-diffusion, pulse injection and back-to-back experiments (for which the water content is measured instead of the accessible porosity). On the other hand, the transport parameters obtained for 125I were shown to be more sensitive to the used experimental set-up. In pulse injection experiments and at the scale of few cm, the dispersion length cannot be neglected for 125I and needs to be known/assessed to extrapolate the diffusion coefficients. Pulse injection experiments also provided significantly higher accessible porosities than through-diffusion experiments which confirms that is set-up is not appropriate for the determination of the anionic accessible porosity. The cutting of the cores and extrapolation of the activity/concentration profiles in back-to-back experiments is practically more challenging and the experimental datasets displayed high scatter. Yet overall the three experimental set-ups, the determined diffusion coefficients for 125I didn’t vary more than a factor 3 for the different samples and orientation investigated. Unfortunately, the two through-diffusion set-ups that were investigated could eventually be considered identical. The original objective was to maintain, in one set-up, a zero boundary condition in the outlet while in the other set-up the concentration of the outlet compartment was allowed to increase with time. However, the second set-up didn’t run for long enough to depart from the zero boundary condition in the outlet preventing any conclusion to be drawn regarding the consistency of the transport parameters determined with these two set-ups. The two investigated cell designs in pulse injection experiments provided significantly different results highlighting the need to appropriately choose the set-up by reducing as much as possible dead-volumes. In through-diffusion experiments, the size of the BC cores, ranging from 1.5 to 6 cm, didn’t show any significant effect on the determination of the diffusion and dispersion coefficients and accessible porosity for both 125I and HTO. However, with pulse injection experiments performed in through-diffusion cells, the smaller the samples, the more deviation was observed between the determined and the expected transport parameters. This trend was attributed to the increased impact of the dead volumes of the cell with the decrease of the sample length.