Main content area

Diffusion-driven transport in clayrock formations

Altmann, Scott, Tournassat, Christophe, Goutelard, Florence, Parneix, Jean-Claude, Gimmi, Thomas, Maes, Norbert
Applied geochemistry 2012 v.27 no.2 pp. 463-478
actinides, anions, argillite, cations, cesium, clay, clay minerals, cobalt, copper, geochemistry, ionic strength, mineral content, models, porosity, prediction, radioactive waste, radionuclides, solutes, sorption, strontium
Clay mineral-rich sedimentary formations are currently under investigation to evaluate their potential use as host formations for installation of deep underground disposal facilities for radioactive waste (e.g. Boom Clay (BE), Opalinus Clay (CH), Callovo–Oxfordian argillite (FR)). The ultimate safety of the corresponding repository concepts depends largely on the capacity of the host formation to limit the flux towards the biosphere of radionuclides (RN) contained in the waste to acceptably low levels. Data for diffusion-driven transfer in these formations shows extreme differences in the measured or modelled behaviour for various radionuclides, e.g. between halogen RN (³⁶Cl, ¹²⁹I) and actinides (²³⁸,²³⁵U, ²³⁷Np, ²³²Th, etc.), which result from major differences between RN of the effects on transport of two phenomena: diffusion and sorption. This paper describes recent research aimed at improving understanding of these two phenomena, focusing on the results of studies carried out during the EC Funmig IP on clayrocks from the above three formations and from the Boda formation (HU). Project results regarding phenomena governing water, cation and anion distribution and mobility in the pore volumes influenced by the negatively-charged surfaces of clay minerals show a convergence of the modelling results for behaviour at the molecular scale and descriptions based on electrical double layer models. Transport models exist which couple ion distribution relative to the clay–solution interface and differentiated diffusive characteristics. These codes are able to reproduce the main trends in behaviour observed experimentally, e.g. Dₑ(anion)<Dₑ(HTO)<Dₑ(cation) and Dₑ(anion) variations as a function of ionic strength and material density. These trends are also well-explained by models of transport through ideal porous matrices made up of a charged surface material. Experimental validation of these models is good as regards monovalent alkaline cations, in progress for divalent electrostatically-interacting cations (e.g. Sr²⁺) and still relatively poor for ‘strongly sorbing’, high Kd cations. Funmig results have clarified understanding of how clayrock mineral composition, and the corresponding organisation of mineral grain assemblages and their associated porosity, can affect mobile solute (anions, HTO) diffusion at different scales (mm to geological formation). In particular, advances made in the capacity to map clayrock mineral grain-porosity organisation at high resolution provide additional elements for understanding diffusion anisotropy and for relating diffusion characteristics measured at different scales. On the other hand, the results of studies focusing on evaluating the potential effects of heterogeneity on mobile species diffusion at the formation scale tend to show that there is a minimal effect when compared to a homogeneous property model. Finally, the results of a natural tracer-based study carried out on the Opalinus Clay formation increase confidence in the use of diffusion parameters measured on laboratory scale samples for predicting diffusion over geological time–space scales. Much effort was placed on improving understanding of coupled sorption–diffusion phenomena for sorbing cations in clayrocks. Results regarding sorption equilibrium in dispersed and compacted materials for weakly to moderately sorbing cations (Sr²⁺, Cs⁺, Co²⁺) tend to show that the same sorption model probably holds in both systems. It was not possible to demonstrate this for highly sorbing elements such as Eu(III) because of the extremely long times needed to reach equilibrium conditions, but there does not seem to be any clear reason why such elements should not have similar behaviour. Diffusion experiments carried out with Sr²⁺, Cs⁺ and Eu(III) on all of the clayrocks gave mixed results and tend to show that coupled diffusion–sorption migration is much more complex than expected, leading generally to greater mobility than that predicted by coupling a batch-determined Kd and Fick´s law based on the diffusion behaviour of HTO. If the Kd measured on equivalent dispersed systems holds as was shown to be the case for Sr, Cs (and probably Co) for Opalinus Clay, these results indicate that these cations have a Dₑ value higher than HTO (up to a factor of 10 for Cs⁺). Results are as yet very limited for very moderate to strongly sorbing species (e.g. Co(II), Eu(III), Cu(II)) because of their very slow transfer characteristics.