Effects on surface hydrology and near-surface hydrogeology of an open repository in Laxemar Results of modelling with MIKE SHE

This report presents the methodology and the results from the modelling of an open repository for spent nuclear fuel in Laxemar. Specifically, the present work analyses the hydrological effects of the planned repository during the construction and operational phases when it is open, i.e. air-filled, and hence may cause a disturbance of the hydrological conditions in the surroundings. The numerical modelling is based on the SDM-Site Laxemar MIKE SHE model. The modelling was divided into three steps. The first step was to update the SDM-Site Laxemar model with a new hydrogeological bedrock model. The other main updates were an increase of the depth of the MIKE SHE model domain, enhanced vertical computational resolution and that the drainage of the Aespoe Hard Rock Laboratory was included in the model. The resulting model was used to simulate undisturbed natural conditions. The next step was to describe the open repository conditions, using Laxemar layout D2, by implementing the access tunnel, the repository tunnels and shafts in the model, and to simulate the consequences for the surface hydrology caused by an open repository under different conditions. The final step was a sensitivity analysis that aimed to investigate the sensitivity of the modelled effects of the open repository to the hydrogeological properties of the bedrock and the Quaternary deposits, the sediments under the sea, and changes in boundary conditions. The model covers an area of 34 km2. The groundwater divides were assumed to coincide with the surface water divides; thus, a no-flow boundary condition was used at the horizontal boundaries, except in the Quaternary deposit layers towards the sea where a time-varying boundary condition describing the sea-level in the area was used. In the bedrock layers, however, a no-flow boundary condition was applied. Also the bottom boundary was described as a no-flow boundary. The transient top boundary condition was based on meteorological data gathered at the local SKB stations during the period 2004-2006. The groundwater modelling was performed with the MIKE SHE code, a process-based modelling tool that calculates the groundwater flow in three dimensions. It takes the whole hydrological cycle into consideration and describes the water flow from rainfall to river flow. The coupling to the pipe flow model MOUSE was used to implement the repository. A development of the coupling code, compared to the code used in earlier open repository modelling, was made. The repository was described as a number of pipe links in MOUSE and the inflow of water from MIKE SHE to MOUSE, i.e. the flow of water from the aquifer to the tunnels, was calculated. The shafts were described as cells with atmospheric pressure. The results from the updated MIKE SHE model for undisturbed conditions agrees with the results obtained from the SDM Laxemar model presented in the final (SDM-Site) version of the site description. The average specific runoff in the simulation for the year 2006 was calculated to 139 mm and the total evapotranspiration to 398 mm. The groundwater table in the area is rather deep; the mean depth to the groundwater table for the year 2006 was calculated to 3.7 m below ground surface (the sea area excluded). The discharge in the water courses is transient during the year and is dependent on the meteorological conditions. The impact of the open repository on the groundwater table position is extensive, which is also the case with the head change in the bedrock, and reaches the model boundary in the northern and southern parts of the model domain. The largest drawdown of the groundwater table is developed above the central parts of the repository. The calculated groundwater table drawdown and the size of the associated influence area (here defined as the area where the drawdown is larger than 0.3 m) are somewhat dependent on the level of grouting in the access tunnel and the deposition tunnels. Three levels of grouting were tested corresponding to hydraulic conductivities (K) of 1x10-8 m/s in all tunnels, or 1x10-9 m/s or 1x10-10 m/s in deposition tunnels and 1x10-8 m/s elsewhere. The lowest level of grouting leads to an influence area of 16.7 km2, as an average for 2006 (the last year in the simulation period). When the highest level of grouting is applied in the repository, the average influence area is calculated to 13.9 km2 for the year 2006. The temporal variation of the influence area during a year is rather small, with small differences in influence areas between most months in 2006. However, periods with heavy rain or snowmelt decrease the influence areas and cause a large difference between the maximum and minimum influence areas. The water level in Lake Frisksjoen, as well as the discharges in the water courses, are affected by the open repository; for example, the average water level of Lake Frisksjoen decreases with 0.1 m and the mean discharge of Laxemaraan decreases with 8% compared to undisturbed conditions, with a grouting level of K=1x10-8 m/s. The simulated total inflows to the open repository vary between 88 L/s and 55 L/s depending on the applied level of grouting. More than two thirds of the repository inflow comes from an increased vertical inflow from the land area. The influence areas and inflows specified above refer to a case with the whole repository open, i.e. with all the transport and deposition tunnels open at the same time. However, this is a hypothetical worst case scenario, because the repository will be constructed and taken into operation in five development phases. In the present modelling study, the second of these development phases and the initial construction phase were investigated in separate open repository simulations and the results compared to those obtained with the whole repository open at the same time.