On the basis of the transient flow fields, transport calculations with the code r3t and the newly implemented approach are performed. In Fig. 1 the conditions for the environmental parameters are shown for a model time of 160,000 a. At this stage, the distribution of the environmental parameters has changed according to the flow field, the boundary conditions and the geochemical interactions of the environmental parameters.
Fig. 1: Distribution of environmental parameters after 160,000 a model time for [Ca] (mol m-3, top), [DIC] (mol m-3, 2nd line), pH (without unit, 3rd line) and corresponding smart Kd-values for Cs (m3 kg-1, bottom).
The distribution of environmental parameters at 160,000 a model time can be construed for different parts of the model area. Exemplarily, it is explaned for the areas of groundwater recharge in the northern and southern part of the upper aquifer here. The rainwater has a low pH-value and low contents of Ca and DIC compared to their initial concentrations in the upper aquifer. Due to the influence of the modestly mineralized rainwater, the Ca and DIC concentrations in the recharge areas decrease. As a consequence of the low concentrations of Ca and DIC, calcite dissolves causing an increase of the pH-value. In the center of the upper aquifer, the groundwater is still affected by the groundwater recharge in the south and in the north. Calcite is dissolved, the pH-value increases, and the Ca and DIC concentrations are decreased.
After 160,000 a model time the smart Kd‑value very well reflects the evolution of the environmental parameters, in the upper aquifer as well as in the aquitard. Since the pH-value is the most important parameter here, the smart Kd‑value is higher in areas with higher pH and vice versa. Additionally the higher content of clay minerals causes the higher smart Kd‑values in the aquitard. The smart Kd‑value in the aquitard decreases, especially above the northwestern rim syncline, where the pH‑value is lowest. Where groundwater from the lower aquifer rises to the upper aquifer and is transported to the south, lowest smart Kd‑values in the upper aquifer are found. Elevated smart Kd‑values can be observed in the groundwater recharge areas. High smart Kd‑values are found in the vicinity of the southern hydraulic window according to the highest pH‑values found there.
On that basis the transport of Cs-135 is calculated, assuming a stylized inflow as point source with a constant flux in a time frame between 12,500 and 111,500 a [Noseck et al. 2012]. The distribution of Cs-135 after 160,000 years model time is shown in Fig. 2. After upward and northward transport through the lower aquifer, the aquitard is reached. Due to high smart Kd-values and mainly diffusive transport in this unit, Cs-135 has only entered about one third of the aquitard thickness. This observation is plausible. However, it is necessary to apply longer model times considering climatically driven changes to enable a deeper insight into the radionuclide transport through the aquitard and the upper aquifer. Such simulations will be performed in the near future.
Fig. 2: Distribution of Cs-135 after 160,000 a, logarithmic scale [mol m‑3].
Summarizing the calculations shown here for a model time of 160,000 a with constant boundary conditions, the transport of the radionuclides is dominated by advection. The environmental parameters reflect well the interaction of the initial and boundary conditions as well as the mixing processes and thus the changes of the geochemical environment. The impact of the environmental parameters on the distribution coefficients of Cs can be well retraced. A comparison of CPU times needed for the transport simulations employing the conventional Kd‑value with those for the new approach using the smart Kd‑concept shows only an increase by a factor of 3 – 3.5, which is fully feasible for long-term safety assessment calculations. A detailed description of the transport simulations can be found in Noseck et al. (2012).
Noseck, U., Brendler, V., Flügge, J., Stockmann, M., Britz, S., Lampe, M., Schikora, J., Schneider, A.: Realistic integration of sorption processes in transport codes for long-term safety assessments, Final Report GRS-297, (Gesellschaft für Anlagen- und Reaktorsicherheit (mbH)), Braunschweig 293 p (2012).