||In France, deep underground storage of long life radioactive wastes is planned by ANDRA within Bure argillite, a compacted clay rock found at 420-550m depth in the East of France. The storage structure is composed of a network of tunnels drilled in argillite and strengthened by a concrete and steel sheath. Due to excavation operations, macro-cracking of the host rock occurs at the tunnel surface (in a so-called EDZ, Excavation Damaged Zone), and partial de-saturation of argillite is also bound to occur, due to tunnel ventilation during storage, prior to tunnel closure. Although it is aimed at strengthening the argillite tunnel, concrete material also de-saturates partially, and it is potentially damaged and micro-cracked. For safety assessment, it is vital to identify to what extent damaged structural materials (i.e. argillite and concrete) are still able to ensure proper radionuclides retention. In this context, the present contribution aims at determining the advective gas transfer properties of a damaged industrial concrete in varied partially-saturated states.
Firstly, we develop a methodology in order to create damage, and most probably micro-cracks, in an ANDRA industrial concrete, which simulate the cracks occurring in the EDZ, yet at the laboratory scale. To this purpose, several experimental protocols have been designed and tested. Gas permeability in the dry state is used as a damage indicator. Three cycles of freeze/thawing (-18°C for 24h to 100°C for 30min) generate damage only when the material is initially water-saturated. In such case, gas permeability values are multiplied by 10, while sample porosity is not increased significantly. Thermal shock using liquid nitrogen (with one cycle from room temperature, down to -196°C and then up to 100°C) has no significant effect on material permeability, even if it is initially water-saturated.
Secondly, gas relative permeability is measured on concrete samples, in both intact and damaged states, at varying confinement, up to values close to the in situ major stress (12MPa). Two different concrete batches were compared in the intact state. Samples are tested after mass stabilization in hermetic chambers at different fixed relative humidity: RH = 100, 98, 92, 85, 75, 70, 59, 43, 11%, and also in the dry state. Suction curves, relating water saturation level Sw to RH, are identified for both intact and damaged samples, and for both batches in the intact state. The permeability variation of each material has been studied as a function of RH but also with confinement. The influence of damage on gas relative permeability is shown to be significant when plotted as a function of RH, so that greater gas passages exist at higher relative humidity for damaged materials than for intact ones. Nevertheless, gas relative permeability is independent of concrete damage state and of confinement, when plotted as a function of water saturation level Sw. It only depends on concrete batch, i.e. on its microstructure, as developed during maturation.