Technischer Bericht NTB 87-18
Modellierung der Grundwasserströmungsverhältnisse am Wellenberg
In August 1986 the Group for Hydrodynamic Modelling (GHM) undertook a hydrogeological modelling study of the potential waste repository site at Wellenberg. The main goal of the study was to identify and quantify those hydrogeological conditions within the project area which could be considered relevant to the long term safety of a waste repository. Additional objectives of the study were the identification of factors which could have some bearing on the planned site investigation programme and the comparison of the effects that various alternatives for the layout of the repository would have on the groundwater flow.
A site model domain, enclosing a volume of approximately 25 km3, was defined. The excavations considered include an upper level repository at 550 m NN, and a deep repository at 250 m NN, together with a vertical shaft between the two levels and that part of an access tunnel lying within the host rock. A near field model enclosing a volume of almost 1.2 km3, and which allows various layouts of the upper repository together with access and location tunnels to be considered, was created separately.
In a parametric sensitivity study, 57 cases were calculated using the site model and, on the basis of the results obtained, a further 14 cases were simulated with the near field model. The calculations were performed by the 3-dimensional steady-state finite element program "FEM301".
The structure of the model is based on the current general understanding of the site geology. The NE-SW striking area of the Valanginien marl, which forms the host rock, lies between the Axen nappe in the south and the limestone of the Drusberg nappe in the north. From this information a conceptual model based on seven hydrogeological units was compiled. The following six geometrical options are included in the model:
- deepened valley filling in the Engelberg valley
- slicing of the nappe under the host rock
- limited extent of the host rock
- a block of neighbouring rock isolated in the host rock
- a rockslide mass over the more permeable valley filling
- an inverse limb of the Drusberg nappe
In addition the water table and the boundary conditions at the edges of the modelled domain may be varied. The geometrical options may be introduced into the model in a cumulative fashion and combined with any desired permeability options. The permeability range for each of the defined hydrogeological units is based mainly on investigations carried out in the geologically comparable Oberbauenstock location before the Phase I investigations of 1987.
The topography of the model domain implies higher hydraulic potentials in the Axen nappe to the south of the host rock, and lower potentials in the limestone of the Drusberg nappe to the north of the Wellenberg. The main consequence is a draining of the host rock layer in the north-west direction towards the Engelberg valley, and to a lesser extent towards the Secklis valley. Over a large area the groundwater flow in the host rock is directed downwards towards the inverse limb in the Drusberg nappe.
The predicted potential distributions along the planned exploratory drillings are influenced by the geometrical options. The strength of this influence varies according to the actual location of the drillings.
The shortest groundwater flow path through the host rock (measured from the repository to the point at which the water leaves the host rock) is about 200 m. For a flow porosity of 0.001 the shortest calculated flow times in realistic cases are between 150 and 1'500 years. In very conservative cases, when a number of unfavourable assumptions are combined, flow times below 10 years may result. The velocity of the water in the marl is of the order 1 m/a in the realistic cases.
In the realistic cases, water flows from the upper repository to the lower, through the vertical shaft. This flow direction can be reversed in the very conservative cases. The shortest flow path lengths and flow times are observed in trajectories starting from the northern part of the upper repository. The longest flow paths emanate from the southern part of the upper repository. Flow paths and flow times for trajectories starting in the deep repository exhibit less variability and lie between the extremes calculated for the upper level.
Operational considerations dictate that access to the upper repository is from the north-west. A direct tunnel from the surface would therefore lie parallel to the direction of groundwater flow from the repository to the biosphere. If the tunnels were to make a loop to the west around the repository, entry could be made from the south, the groundwater inflow side. This layout significantly reduces the chance of a direct flow from the repository through the tunnels into the biosphere.
Trajectory data (flow time, distance and direction) are sensitive to the groundwater table level in the adjacent northern and southern limestone formations and to the geometrical options involving a sliced nappe or the inverse limb. Flow times are also strongly dependent on the porosity, the permeability and the extent of the marl and on the orientation of the repository. In the case of a direct access from the north-west, the permeability of the decompressed zone around the tunnels is also an important parameter.
At the present time the data necessary for the calibration of the model are not available. Only the total flow of groundwater out of the model domain can be used as a calibration value for the permeable hydrogeological units and, to a lesser extent, for the water table. The permeability of the host rock cannot, however, be verified in this way.
The planned investigation program will yield data for all the hydrogeological units and thereby allow the calibration and validation of the model. Reliable data will be obtained from observations of the depressurisation induced in the repository domain as a result of the exploratory excavations.