Migration of Caesium in Bentonite: Influence of the Pore Water Composition

The transferability of empirical distribution coefficients (KD) derived from batch sorption experiments to compacted systems is evaluated in this study. Sorption pa-rameters of Cs+ on bentonite (Wyoming type MX-80) were studied at different dry bulk densities of the clay (1.3, 1.6, 1.9 g/cm3) in both batch tests and diffusion ex-periments. Here, an innovative approach developed at Paul Scherrer Institute (CH) was used for modelling the pore water composition of the compacted clay [1]. The obtained parameters shall be used for long-term prognoses of diffusion processes in the near-field of final repositories of nuclear waste.
The KD values derived from batch and diffusion tests, respectively, differ by ap-proximately 20 to 40 %. This result is acceptable, when taking into consideration the experimental uncertainties of both methods. The use of batch derived KD’s to de-scribe the migration behaviour of Cs+ can be recommended, provided that an ade-quate pore water composition is used.
In contrast, the effective diffusion coefficients (De) of Cs+ in compacted ben-tonite exceed the De’s of tritiated water (HTO) significantly. This effect is attributed to a phenomenon described in the literature as “surface diffusion” indicating an addi-tional diffusive flux of adsorbed species.
Hence, the combination of diffusion coefficients of HTO and batch derived KD’s systematically underestimates the diffusive flux of Cs+. In order to implement this ap-proach in the future, the surface diffusion phenomenon has to be elucidated by ad-vanced spectroscopic methods and to be quantified.
Deviations in the KD are only related to the transient state of the diffusion proc-ess. Thus, they influence the initial state of the transport process. The batch derived KD’s are transferable to the compacted system. However, the use of an adequate diffusion coefficient De is of extreme importance for any prediction of long-term pre-dictions in risk assessment.