The safety of deep geological repositories is important in the disposal of high-level radioactive waste (HLW). In this study, advection–dispersion experiments were designed to build a transport model through a calibration/validation process, and the transport behavior of tritiated water (HTO) and various iodine species (iodide: I− and iodate: IO3−) was studied on a dynamic compacted granite column. Breakthrough curves (BTCs) were plotted under various flow rates (1–5 mL/min). BTCs showed that the non-sorption effect by anion exclusion was observed only in I− transport because the retardation factor (R) of I− was lower than that of HTO (R = 1). Moreover, equilibrium and nonequilibrium transport models were used and compared to identify the mobile/immobile zones in the compacted granite column. The anion exclusion effect was influenced by the immobile zones in the column. The non-sorption effect by anion exclusion (R < 1) was only observed for I− at 5.0 ± 0.2 mL/min flow rate, and a relatively higher Coulomb’s repulsive force may be caused by the smaller hydration radius of I−(3.31 Å) than that of IO3−(3.74 Å).
In this study, a column experiment was employed to evaluate the nuclide migration behavior in the surrounding rock medium of a near-surface disposal site in China and to investigate the advection–dispersion behavior of tritium (H-3) and plutonium-238 (Pu-238) in highly weathered argillaceous shale. A reasonable numerical model was selected to fit the experimental breakthrough curves (BTCs) and to obtain the relevant migration parameters. The results show the following: (1) the internal structure of the highly weathered argillaceous shale exhibited heterogeneity, and the nuclide migration BTC showed characteristics of a “curve peak moving forward” and a “tail curve trailing”; (2) compared with other models, the stream tube mode could better fit the BTCs and obtain the average dispersion coefficient <D>, average distribution coefficient <Kd>, and other parameters; (3) compared to the results of the batch experiment, the distribution coefficient Kd obtained from the column experiment was smaller than that obtained from the batch experiment, which is speculated to be due to the influence of contact time and the contact area between the nuclide and the medium.