SITE 94. Modelling of groundwater chemistry at Aespoe Hard Rock Laboratory

In this report a model is described, which has been able to give agreement between observed and modelled values for more than ten element concentrations (including pH and pE values). The model makes use of a number of steady state waters which are mixed naturally after which the mixtures react with minerals in the fractures. The end member waters are supposed to have been present in the fracture system during a time interval which is long enough for the rock groundwater system to have reached a steady state. Some elements, e.g. chlorine, is modelled as conservative (inert with respect to the rock). Most element concentrations cannot be explained from mixing alone. Rather reactions with the fracture walls have to be taken into account. The situation is complicated by the fact that a system comprised of groundwater and a number of fracture minerals may violate Gibb's phase rule. In such a system, no global equilibrium state exists, and thus the water can never reach equilibrium with respect to all the fracture minerals. The end member waters eventually formed can be expected to be in a steady state condition rather than equilibrium with respect to the fracture minerals. It should be noted that such a steady state is not an equilibrium state. Rather, the water chemistry has to fluctuate as a result of spatial variability in the local mineral set. In most cases when an end member water is sampled, a large number of local waters are mixed causing the fluctuations to cancel out. The CRACKER is a program which has been developed to handle this complicated chemical situation. It couples chemistry and transport, using elaborate chemical modelling in combination with a simplified transport model. The program simulates chemical reactions of groundwater flowing through a plane fracture. The simulation results show that although the end member waters are far from equilibrium with respect to most of the minerals, they are in a steady state with respect to the rock. The chemistry of the groundwater appears to be mostly a consequence of chemical reactions between water and fracture filling minerals. This has been studied in simulations where rain has been allowed to enter a fracture. Although there are some differences, the results found in this modelling generally are similar to the observed values. The deviations suggest that the fracture filling minerals alone are not enough to explain the observed groundwater chemistry. This should not be expected, since the old waters found hundreds of meters below the surface have been able to react with other mineral sets as well. In the CRACKER simulations it has been found that the simulated waters are similar to observations with respect to redox properties. The Aespoe groundwaters have iron concentrations which are low compared to those found in other deep Swedish groundwaters. By inhibiting reactions in the simulations, it has been found unlikely that the Fe(II) - Fe(III) redox couple is the main controller of the redox properties in Aespoe groundwaters. Rather, it appears that the redox properties can be maintained even if reactions with this couple are inhibited. According to the simulation results, reactions in which dissolved hydrogen, carbonates and methane are participating are important. This does not mean that iron species and minerals are not involved in redox processes. The simulation methods have been applied to possible future events and processes. As examples of such, the effects of elevated temperature on the groundwater properties have been investigated as well as a prediction on possible effects of Baltic Sea water intrusion