Lars O. Ericsson, Chalmers University of Technology (Sweden)
Johan Holmén, Golder Associates AB (Sweden)
Ingvar Rhén, SWECO VIAK AB (Sweden)
Niklas Blomquist, Aqualog AB (Sweden)
One of many geoscientific questions in connection with the siting of a final repository for spent nuclear fuel has to do with understanding the large-scale flow patterns of the naturally circulating groundwater. A project was conducted in order to: 1.) evaluate conceptual simplifications and model uncertainties in super-regional groundwater modelling, 2.) carry out an in-depth of regional flow conditions in south-eastern Sweden. Achieving these goals has required an approach based on the use of available geoscientific data combined with an analysis of different conceptual assumptions.
The area studied consists of five regional catchment areas. The groundwater model represents a very large area, about 80×130 km2. The bedrock in the model area is dominated by rock types belonging to the "Transscandinavian igneous belt". Due to the large potential groundwater recharge in relation to the limited permeability of the rocks in the region the calculation cases were run with topography-determined top boundary conditions.
Flow paths from theoretical repository areas were studied with respect to: Flow pattern for groundwater passing repository depth, Outflow time (breakthrough time), Flow lengths, Specific flows at repository depth, Evolution of salinity in connection with density-driven flow during the land uplift phase.
The following system properties were studied: Local topographic undulation, Quaternary deposits, Anisotropy in the conductivity field, Depth-decreasing conductivity, Varying conductivity depending on different lithological units, Regional vertical deformation zones, Regional horizontal deformation zones, Dolerite dykes, Local heterogeneity, Density-dependent flow and time-dependent aspects of the studied system.
The following general conclusions were drawn from the study: The factor of greatest importance for the regional flow pattern is the topography. Different lithological units, regional deformation zones, local heterogeneity, Quaternary deposits are of less importance than the undulation of the topography. For areas described and analyzed with the most realistic assumptions, the groundwater flow pattern can be described as a primarily local flow process. The median flow path length in the study is on the order of 2 km, and the fraction of super-regional flow paths (longer than 10 km) is very small. If hydraulic conductivity decreases with depth, flow cells of a smaller size are obtained than if this was not the case. Anisotropy in the conductivity field with greater horizontal than vertical hydraulic conductivity results in relatively larger flow cells and long flow paths are extended. Anisotropy in the conductivity field with greater vertical than horizontal hydraulic conductivity results in relatively smaller flow cells. Hydraulic conductivity that decreases with depth and horizontal anisotropy has a greater influence on the flow pattern than lithological units, deformation zones, dolerite dykes and Quaternary deposits.