Evaluation of heat propagation from a KBS-3 type deep repository for Aberg, Beberg and Ceberg in SKB's SR 97

The following conclusions and remarks may be drawn from this review work: The phenomenology of heat propagation in the complex system of a repository is generally well understood, and permits a reliable prediction of temperature evolution. Some simplifications such as the representation of incomplete saturation of the buffer are justified at the present stage of knowledge, and do not seem to impair the reliability of the results. These simplification may become unnecessary in the future as experimental work presently going on leads to a better predicting capacity of these mechanisms. Important parameters (thermal conductivity and heat capacity) are known through a large number of measurements, and furthermore do not span a wide range. The main uncertainty seems to lie in the near field parameters (description of the contact between container and buffer, thermal conductivity of unsaturated bentonite). Generally speaking, the weak point of thermal modelling is generally not in the representation of a particular component, but in the heat transfer between different components. Although no systematic sensitivity analysis was performed in this study, the various situations modelled do give an idea on the effect of varying parameters such as thermal conductivity, heat capacity, initial heat output. The modelling work presented is consistent with the phenomenology described. A powerful computing capacity allows a detailed description of mechanisms at two different scales. However, the transition from one scale (unit cell) to the other (global scale) would deserve a more rigorous analysis. From a reviewers' viewpoint, the description of calculation cases is generally not sufficiently detailed to judge precisely the results, or to try and duplicate them, should this be needed. While this situation is acceptable in a generic exercise such as SR 97, a more detailed description would be required in a real site assessment. Apart from thermal parameters and repository location and geometry, the results would be highly sensitive to hypotheses regarding the waste production (burn-up, intermediate storage). These aspects are not addressed in this work, but the modelling done shows well, inversely, how these parameters may be optimised from a temperature criterion. Consequences of the thermal evolution on hydrological, mechanical and chemical phenomena are considered, but have not yet, as far as we can tell, been updated considering the latest heat transfer simulations. Chemical aspects are by far the most complex, and will require an important work in the future. As a general conclusion, the study shows the feasibility of calculations aimed at dimensioning the repository geometry as well as its exploitation