Long-term safety for the final repository for spent nuclear fuel at Forsmark. Main report of the SR-Site project

The central conclusion of the safety assessment SR-Site is that a KBS-3 repository that fulfils long-term safety requirements can be built at the Forsmark site. This conclusion is reached because the favourable properties of the Forsmark site ensure the required long-term durability of the barriers of the KBS-3 repository. In particular, the copper canisters with their cast iron inserts have been demonstrated to provide a sufficient resistance to the mechanical and chemical loads to which they may be subjected in the repository environment. The conclusion is underpinned by: - The reliance of the KBS-3 repository on i) a geological environment that exhibits long-term stability with respect to properties of importance for long-term safety, i.e. mechanical stability, low groundwater flow rates at repository depth and the absence of high concentrations of detrimental components in the groundwater, and ii) the choice of naturally occurring materials (copper and bentonite clay) for the engineered barriers that are sufficiently durable in the repository environment to provide the barrier longevity required for safety. - The understanding, through decades of research at SKB and in international collaboration, of the phenomena that affect long-term safety, resulting in a mature knowledge base for the safety assessment. - The understanding of the characteristics of the site through several years of surface-based investigations of the conditions at depth and of scientific interpretation of the data emerging from the investigations, resulting in a mature model of the site, adequate for use in the safety assessment. - The detailed specifications of the engineered parts of the repository and the demonstration of how components fulfilling the specifications are to be produced in a quality assured manner, thereby providing a quality assured initial state for the safety assessment. The detailed analyses demonstrate that canister failures in a one million year perspective are rare. Even with a number of pessimistic assumptions regarding detrimental phenomena affecting the buffer and the canister, they would be sufficiently rare that their cautiously modelled radiological consequences are well below one percent of the natural background radiation