Radioprotection, vol. 44,n 5 (2009) 333–338 C EDP Sciences, 2009 DOI: 10.1051/radiopro/20095065 Modelling the radionuclide transfer from bedrock to surface systems at Forsmark site (Sweden) F. Grandia 1 , C. Sena 1,2 , D. Arcos 1 , J. Molinero 1 , L. Duro 1 and J. Bruno 1 1 AMPHOS XXI Consulting S.L., Passeig de Rubí 29-31, 08197 Valldoreix, Barcelona, Spain 2 I&DGeoBioTec, Geosciences Dept., University of Aveiro, 3810-193 Aveiro, Portugal Abstract. Quaternary sediments and soils at the Forsmark site (Sweden) would constitute the last barrier for radionuclide migration interposed between a deep repository of high level nuclear wastes (HLNW) and surface ecological systems. The retention capacity of these sediments is evaluated by means of reactive transport simulations. Two different scenarios have been modelled: (1) Inflow of deep fluids carrying radionuclides into shallow aquifers hosted in the carbonate glacial till, and (2) Inflow of deep fluids carrying radionuclides into organic matter-bearing clays. The model results predict that caesium is very efficiently retained in both scenarios for periods longer than 3000 years due to the strong affinity of this ion with the frayed edge sites of the illite layers. Strontium is also retained via cation exchange in illite and, in a lesser extent, via co-precipitation in calcite. The retention efficiency for strontium, however, quickly decreases to 0 after 1000 years. Uranium is retained in both scenarios with retention efficiencies higher than 40% at long term (>3000 years) due to precipitation of amorphous uraninite and adsorption on Fe(III) oxyhydroxides. Radium is only retained in the till aquifer since co-precipitation with barium sulphate does not occur in the clay porewater. 1. INTRODUCTION The concepts of deep geological disposal of high level nuclear waste (HLNW) include a number of natural and engineered barriers designed to prevent radionuclide migration to surface systems in case of isolation failure of the waste canisters [1]. The Swedish concept of HLNW storage is based on the deep disposal in a granite environment, and two candidate sites, Forsmark and Laxemar, are currently under investigation by the Swedish Nuclear Fuel and Waste Management Company (SKB). In both areas, Quaternary regoliths overlying the granite bedrock would constitute the last natural barrier for the retention of radionuclides before reaching the surface systems. The retention capacity of these deposits depends on the chemical behaviour of the radionuclides and the geochemical features of the sediments (mineralogy, porewaters, organic matter content). Deterministic calculations of radionuclide releases from the geosphere (granite) [1] show that Ra- 226, Cs-135 and I-129 would be the main dose contributors to the surface systems at long term in case of canister failure. The determination of the hydrogeochemical behaviour of these radionuclides when in contact with Quaternary soils is, therefore, very relevant. The behaviour of radium and caesium, along with other radionuclides of interest such as uranium and strontium, is predicted in this work by using 2D reactive transport calculations in which retention is simulated under two different hydrodynamic environments in the Forsmark area. Iodine has not been modelled at this stage since the thermodynamic database for this element is still not well developed. The important outcome of the project is the categorization of the retention mechanisms considered. 2. CONCEPTUAL MODEL AND NUMERICAL IMPLEMENTATION 2.1 Surface geology and groundwaters at Forsmark The transfer of radionuclides from geosphere to surface systems may take place in different geological environments at Forsmark (Fig. 1): (1) direct contact between crystalline bedrocks and Article published by EDP Sciences and available at http://www.radioprotection.org or http://dx.doi.org/10.1051/radiopro/20095065