Investigation of the thermo-hygro-chemico-mechanical performance of the Bentonite barrier at the high-level radioactive waste canister interface

Abstract

This research focuses specifically on the geotechnical and physicochemical evolution of the highly-compacted Bentonite at the radioactive waste canister interface, with respect to the multi-barrier system for deep geological disposal of radioactive waste. The research investigated the geomechanical (i.e. Swelling pressure, self-healing, permeability and rheology) and mineralogical (i.e. Fe enrichment/ integration into the aluminosilicate structure and alteration to the accessory minerals) evolution of the Bentonite barrier under key life phases of the near-field regarding thermo-hygro-mechanical-chemico factors. Such life phases were derived from leading concepts and were simplified into three fundamental stages replicating key stages of the near-field. The first phase was “early-life” replication where high temperatures/ drying conditions were exposed to the interface. The second phase replicated the “mid-life” conditions, subjecting the interface to resaturation using high saline ground water with high temperatures but lower than the initial thermal exposure. The final phase replicated the “late-life” conditions which subject the interface to background thermal loading, representative of background temperatures of a crystalline host rock, with a completely saturated/ hydrostatic saline pore-water environment. Post analysis examined the impact on the geomechanical and physicochemical properties with respect to how these relate to the desired functionality of the passive barrier system. Previous work has seldom addressed this interface system from a coupled engineering and physicochemical perspective. A comprehensive investigation looking at how the near-field environment inhibits the desired engineering function and the related physicochemical state will enable insight into building a UK specific safety case. This research employed an experimental approach exploring the boundary conditions that the interface would experience as inferred from leading international radioactive waste repository concepts. Experimental methods included batch testing, interface replication and geotechnical constant volume tests. Batch tests looked at a mix of variable boundary conditions relevant to the near-field interface environment and isolated specific conditions to measure the intrinsic alterations to the desired engineering function. Interface replication tests subject MX-80 Bentonite, compacted to in-situ emplacement conditions, and carbon steel to an interface environment under three separate “life-phases” (as stated earlier) to assess what phase had the most significant impact on the geomechanical and physicochemical state of the barrier. Finally, the constant volume experiments subject oedometer samples to “mid-life” conditions as well as isolated “mid-life” variables respectively. Post exposure measurements looked at the impact of the various “mid-life” boundary conditions and their associated impact on the swelling pressure, swelling kinetics, permeability and rheology. “Mid-life” conditions were chosen to carry forward into the oedometric tests as this phase was considered to subject the Bentonite to the highest thermal loading under the most complex chemical environment. This work has highlighted key near-field factors that alter the interface barrier function after relatively short-term exposure. Such as thermal loading and exposure to high salinity, both induce alterations to the swelling pressure, swelling kinetics and rheological properties. It was observed that high temperature-saline exposure caused loss in the dielectric constant and reduction in the Diffuse Double Layer (DDL), as well as the dissolution/ precipitation of silica which inhibited the swelling and plasticity of the MX-80. Exposure to corrosion products derived from S275 carbon steel appear not to significantly affect the physical properties of the Bentonite compared to thermal and saline factors after short-term experimental contact. However, corrosion integrated into the Bentonite matrix after thermal loading in the presence of Deionised (DI) water displayed a slight increase in the kinetic behaviour. This was concluded to be due to the high volumetric capacity of the hydrated corrosion products within the clay matrix. Furthermore, interface tests indicate some alteration to Ferro-saponite type smectites, indicating some Fe integration into the octahedral layer resulting in a trioctahedral smectite. The outer zone furthest away from the canister was characteristic to Mg-type saponite. Nonetheless, XRD displayed that the clay component remained high charge smectite with no alteration to the d-spacing after swelling. Green rust and Goethite was observed at the interface and mid zones respectively, the ferrous mineral appeared to be meta-stable when exposed to ambient conditions. Physicochemical analysis indicated the presence of ferrous hydroxides complexed to the edge sites of the clay. High diffusion rates and dissolution rates of iron corrosion and secondary minerals respectively were greater in “mid-life” and “late-life” conditions. This indicates that thermal loading, as well as a high saline hydraulic gradient, aided the corrosion process and displacement of interlayer cations away from the corrosion rich zone. This resulted in the development of a high pH environment which ultimately increased dissolution rates of the secondary minerals. The dissolution of silica minerals induced precipitation of siliceous by-products upon cooling which essentially “fused” clay particles together, ultimately inhibiting the macro swelling properties and ductility. Results show that the macro-scale geomechanical properties of the Bentonite do not deviate below the design limits set out by leading concepts, but do come close to the minimum limits for swelling pressure (1 – 2MPa). Concern is however highlighted for combined thermo-saline-chemical exposure, which appears to cause significant loss in the ductility of the MX-80. This may be disadvantageous to the stress distribution around the canister which is necessary to ensure that the canister is not breeched due to potential rock displacements etc. Neither comparative data, nor design limits were set out for barrier stiffness, therefore the study suggests that high thermal exposure and duration is kept to an absolute minimum. The research has further highlighted the need to investigate the impact of a completely reduced iron-rich environment on the engineering function of the Bentonite barrier, as completely reduced conditions were difficult to obtain within the time-scale of this study. Other research suggests that a rich environment would accelerate clay alteration as well as alterations to layer charge and lattice stresses potentially inducing 2:1 to 1:1 alteration, all of which potentially lead to complete loss in the swelling behaviour of the clay.