There is international consensus that the environmentally safest place to dispose of toxic radioactive waste is deep underground. Because transport of contaminants occurs in solution, i.e. via the pore water, repositories of radioactive waste will be built within geological formations where water movement is expected to be slow. The chemical characteristics of pore waters in low-permeability rocks contain valuable archives of geochemical conditions that prevailed in the distant past and of processes that occurred since then, often over millions of years. We can learn from Nature by understanding such processes and use this information to predict the expected future evolution of the repository environment. A multi-barrier concept is envisaged, in which metal canisters holding the waste are embedded in swelling clays (which seal against moving groundwater) and cement, and the entire repository is excavated within a low permeability host rock.
Research is needed to understand in detail how these repository components act as barriers and how radioactive contaminants could migrate outwards once the protecting canister has corroded. Based on laboratory and underground experiments, the future behaviour of the repository in its geological setting is predicted by numerical modelling. This is an unusual case for geologists and geochemists, because here the challenge is to predict the future.
Our group has been involved in numerous research projects targeted at radioactive waste over the past 30 years. We focus on the geochemistry of host rocks acting as aquitards and on the geochemistry and hydrogeology of the surrounding aquifers. A further research field includes the geochemistry of bentonite (a swelling clay material) and cementitious materials used in underground structures. We take an integrated approach that combines field work (e.g. drilling campaigns) with laboratory studies and state-of-the-art reactive-transport modelling.
Our main research partners are national agencies responsible for solving the disposal problem: Nagra (Switzerland), Andra (France), Posiva (Finland), SKB (Sweden) and NWMO (Canada). We also participate in large-scale underground experiments in underground research laboratories (URLs), e.g. with swisstopo, Switzerland at the Mont Terri URL, and national waste-management organisations at the Grimsel Test Site GTS (Switzerland), Äspö HRL (Sweden) and Onkalo URL (Finland).
Bentonite is a natural material rich in swelling clays (smectite). In its compacted form it forms an important barrier component in waste repositories, as buffer material surrounding waste packages, as tunnel back-fill, and as element in gallery sealing systems. Bentonite properties, like swelling behaviour, diffusive and advective transport, or its stability during the interaction with the host rock or other materials present (e.g. cement, steel, copper) are of great importance for the repository design.
Hydraulic binders (cements) in form of paste, mortar, concrete and shotcrete are required as waste matrix, structural support, and gallery plugs in waste repositories. The enormous material contrasts between cement and other materials in the repository lead to chemical and mineralogical interaction, mainly driven by chemical gradients and subsequent diffusion in pore water. Relevant host rock or barrier properties like swelling pressure in case of clay rocks, permeability, or retention of radionuclides might be altered.
Host rocks considered for radioactive waste repositories are typically aquitards. Clay-rich sedimentary aquitards have been one of our prominent research topics over the past 25 years. We have developed a number of novel experimental, analytical and numerical methods to study the geochemistry of these rocks and their porewaters. One of the main goals is to understand the transport of solutes, as well as chemical and radioactive contaminants. A significant part of our research involves international studies in Underground Research Laboratories, in particular the Mont Terri Rock Laboratory and the French Laboratory at Bure.