On the inclusion of forest exposure pathways into a stylized lake-farm scenario in a geological repository safety analysis

Abstract

<p>Geological disposal of radioactive waste has been recognized as the ‘reference solution’ to ensure the safety required for the present and future society and environment. To study the possible exposure pathways from groundwater to humans, radioactive transport modelling is used. One of the ecosystems that may play a significant role when assessing the dose conversion factor (i.e. the dose resulting from a nominal release of 1 Bq/year of each radionuclide) for humans is forest. In this paper we have developed a model of a lake-farm system with a forest component. The biosphere system used in this study represents a typical agricultural scenario in Finland, amended with a typical forest. A lake is assumed to form due to post-glacial land uplift. The main features of this future lake have been obtained from our probabilistic shoreline displacement model. Both deterministic calculations and sensitivity analysis were carried out to simulate the model. The deterministic simulation demonstrates the behaviour of the studied radionuclides (<sup>36</sup>Cl, <sup>135</sup>Cs, <sup>129</sup>I, <sup>237</sup>Np, <sup>90</sup>Sr, <sup>99</sup>Tc and <sup>238</sup>U) and the proportions of different exposure pathways to humans. Particularly for <sup>135</sup>Cs and <sup>129</sup>I, forest pathways make a notable contribution to the dose conversion factor. The sensitivity analysis was done using two methods: EFAST and Sobol’. With both methods, the parameters related to the farm contribute the most to the variance of the dose conversion factor for humans. The study demonstrates that the exposure pathways related to forest products may make a considerable contribution to the dose conversion factor in a lake–farm–forest system. It is also confirmed that an advanced sensitivity analysis for a radionuclide transport and dose assessment model on such a landscape scale is feasible even with moderate computational efforts.</p>