Tome Architecture and management of a geological repository - Andra

3 – Design study of a repository in a granite formation• A physical and chemical environment suitable for waste packagesDisposal cell design seeks to provide a suitable physico-chemical environment for waste and packagesin order to control changes in state over time and help limit release of radionuclides. Such anenvironment is ensured i) by the materials used for waste over-packs, which are selected according totype, volume, radiological inventory and chemical nature of the waste, and ii) as well as by engineeredbarriers.For B waste containing metal elements (B1, B3, B4 and B5 reference packages), the aim is to limitcorrosion by providing a favourable chemical environment (reducing potential, pH 10 to 12.5), inparticular by using concrete for waste over-packs. For bituminised B waste, the aim is to maintain, onthe long-term, bitumen confinement properties (B2 reference package) by controlling chemicalconditions and temperature (between 20 and 30°C).For C waste and spent fuel, emplacing clay buffers between package and granite rock attenuateschemical interaction between packages and granite groundwater.• Leak-tight or very low permeability disposal packages over a sufficiently longperiod of timeIn order to ensure complementarity with the geological barrier, primary packages are inserted inadditional containers, to constitute disposal packages. A study has been carried out in order to ensureleak-tightness or very low permeability over sufficiently long periods of time which depends on kindsof waste and their radiological inventories.A concrete disposal package has been chosen for B waste. For some types (B1 and B5 packages,which have major radioactive content and do not release gas), disposal packages have long-termconfinement properties (around ten thousand years). This performance is achieved by using a speciallyadapted concrete formulation (with very low permeability and porosity) and a specific design (closingmethod). This type of container limits the amount of water reaching primary packages as well asradionuclides release for this period of time.For C waste packages, the aim is to prevent water from coming into contact with the glass for severalthousand years. This period concerns the thermal phase (i.e. the period when the temperature at theheart of the glass is over 50°C) during which glass alteration phenomena are accelerated. Theproposed design is based on a very thick steel container.For spent fuel, a copper container is proposed, with long term leak-tightness property (up to severalhundred thousand years). In contrast to C waste, radionuclides are not trapped in a confinement matrix(a fraction of the radionuclides is released upon contact with water, and the remainder is releasedgradually as the uranium oxide matrix dissolves). This option is based on the ‘KBS-3’ coppercontainer a concept adopted in Sweden (SKB) and Finland (Posiva). It was adopted by Andra at thisgeneric design phase. Site data could justify revision of this option if architecture adaptation to granitemassif fracturing and engineered structures (backfills and seals) provide a sufficiently long transfertime in the geological medium to ensure radioactive decay of radionuclides.3.3.3 Limiting granite disturbances by the repositoryWhile the repository design aims to take the most of the favourable properties of granite, it should beensured that repository construction and its long-term evolution do not aversely affect the properties ofthe granite medium. The various arrangements studied involve structure dimensioning, choice ofmaterials for engineered components and the disposal process.Dossier 2005 Granite - ARCHITECTURE AND MANAGEMENT OF A GEOLOGICAL REPOSITORY45/228