Saline groundwaters counteract up-flow of contaminants- implications for radionuclide repositories?

The high-level nuclear waste, HLW, from Swedish and Finnish reactors will be deposited in crystalline rock at depths around 500 m. The waste is enclosed in steel canisters protected against corrosion by a 5 cm thick copper shell, which ensures a lifetime far longer than 100 000 years. Should some canister be breached any leaking nuclides will have decayed to so low activity that even if they reached the biosphere, they would cause minimal risk to humans. The cost of the copper is significant. The dismantling of the nuclear reactors, with induced activity must also be disposed of and this waste volume is much larger than that of the HLW, which makes it impossible to protect it in the same way. This paper explores if by locating the waste at larger depth where the ground water is more saline, and where the hydraulic conductivity of the rock is lower up-flow of contaminated water can be ensured to be negligible because the denser water at larger depth counteracts up-flow due to negative buoyancy. Several processes that could cause local up-flow are addressed, such as infiltration of meteoric water, impact of surface topology, heat production of the waste, geothermal gradient, salinity gradient, hydraulic conductivity heterogeneities and salt migration between seeping water and salt in matrix pore water. Flow and transport simulations using data from extensive field investigations over more than ten years with scores of km deep boreholes suggest that a HLW repository at around one km depth may be sufficient to hinder up-flow to the biosphere.

Scenario Analysis for the Safety Assessment of Nuclear Waste Repositories: A Critical Review

A major challenge in scenario analysis for the safety assessment of nuclear waste repositories pertains to the comprehensiveness of the set of scenarios selected for assessing the safety of the repository. Motivated by this challenge, we discuss the aspects of scenario analysis relevant to comprehensiveness. Specifically, we note that (1) it is necessary to make it clear why scenarios usually focus on a restricted set of features, events, and processes; (2) there is not yet consensus on the interpretation of comprehensiveness for guiding the generation of scenarios; and (3) there is a need for sound approaches to the treatment of epistemic uncertainties.