Composite liners are used in landfills to isolate solid waste from the local environment. The combination of a high-density polyethylene (HDPE) geomembrane and compacted clay liner (CCL) is commonly used worldwide. In the Ontario region, bentonite sand mixtures (BSMs) and the local clay i.e. Leda clay, can be considered as appropriate CCL materials. However, the interface failure between smooth HDPE and CCL is a critical issue for landfill safety. The shear stress behavior and strength parameters at the interface between the HDPE and CCL can be affected by many factors, such as temperature and chemicals. The temperature difference between winter and summer in the Ontario region is approximately 50°C, which causes a freeze-thaw (F-T) phenomenon in local landfills. Leachate and heat are generated during the solid waste stabilization process. Landfill leachate usually contains a high concentration of cations, which can carry heat, thus affecting the landfill liner properties. As a result, the interface shear stress behavior and strength parameters are affected by the aforementioned conditions. In this thesis, a series of experiments were conducted on the shear stress behavior at the interface of Leda clay / HDPE and bentonite sand mixture (BSM) / HDPE. In order to understand the influence of the F-T phenomenon, the samples were tested by varying the number of F-T cycles. Meanwhile, in order to understand the combined influence of cations and heat, the samples were saturated with different solutions, i.e. distilled water, potassium chloride and calcium chloride solutions. Then they were cured in an oven with different temperatures and room temperature, respectively. All of the laboratorial shear tests have been performed by using a direct shear machine. Results show that the BSM /HDPE and Leda clay/ HDPE interfaces are both influenced by the F-T cycles. The BSM/HDPE interface shear of the samples between 0 and 5 F-T cycles has more obvious differences, while the friction angle of compacted Leda clay/HDPE exhibits distinct reduction in the first 3 cycles, after which, the difference becomes hard to differentiate. The results also indicate that both high temperature and high concentration of cations from leachate can slight reduce the interface shear stress of BSM/HDPE. However, the combined influence of thermal-chemical conditions is not much more obvious compared to the effects of a single thermal or chemical condition. The BSM materials, which were saturated with different solutions, are also tested by using X-ray diffraction to examine the mineral changes in the BSM. The calcium and potassium cations convert sodium-bentonite into calcium-rich bentonite and illite/semectie mixtures, respectively. Nevertheless, the changess of clay part caused by the combined effect of heat and leachate have limited influence on the BSM/HDPE interface shear behavior.

Dans le but de vérifier l'efficacité des dispositifs de scellement ou des barrières ouvragées dans le stockage géologique des déchets radioactifs, l'Institut de Radioprotection et de Sûreté Nucléaire (IRSN) a mis en œuvre le projet expérimental SEALEX (SEALing performance EXperiments) auquel ce travail est étroitement lié. Dans le cadre de ce projet, des essais in-situ sont effectués à l'échelle représentative et dans des conditions naturelles sur un mélange compacté de bentonite et de sable. Ce matériau de mélange a été choisi pour sa faible perméabilité et surtout pour sa capacité de gonflement qui permet de colmater les vides existant dans le système, notamment le vide technologique correspondant au vide radial entre le noyau de scellement et la roche hôte et qui est inévitable au cours de l'installation du noyau dans le forage. Une fois les vides scellés, le gonflement à volume constant engendre une pression de gonflement aussi bien sur la roche hôte (radiale) que sur les structures de confinement en béton (axiale). Le comportement de ce matériau dans ces conditions de couplages hydromécaniques est alors étudié dans ce travail. La microstructure du matériau à son état initial a été premièrement examinée par micro-tomographie rayons-X. Ceci a permis de voir la distribution des grains de bentonite et de sable ainsi que le réseau de pores dans l'échantillon. Des macro-pores se sont retrouvés concentrés à la périphérie de l'échantillon ainsi qu'entre les grains de sable, ce qui pourra affecter à court terme la perméabilité. L'hydratation du même matériau en condition de gonflement limité a été ensuite observée par une photographie 2D et par la micro-tomographie aux rayons-X. Le mécanisme de gonflement par production de gel de bentonite, la cinétique de gonflement, la diminution de densité et l'homogénéisation du matériau final on été analysés. L'hydratation en conditions de gonflement empêché a été aussi étudiée par des essais où la pression de gonflement a été mesurée dans deux directions : radialement et axialement. La différence retrouvée entre les pressions de gonflement axiales et radiales a évoqué la présence d'une anisotropie de microstructure qui a été analysée en fonction de la masse volumique sèche de bentonite dans le mélange. Des essais en modèle réduit reproduisant à une échelle 1/10ème les essais in situ (SEALEX) ont été également effectués afin d'étudier le comportement du noyau compacté après la reprise des vides au cas d'un accident détruisant les éléments de confinement. Des mesures locales de pression de gonflement le long des échantillons ont permis de mettre en évidence l'évolution du gradient de densité durant le gonflement axial. Finalement une comparaison entre les résultats obtenus dans ce travail et ceux d'un essai in situ (SEALEX) a été faite. Une bonne correspondance entre les valeurs d'humidités relatives a été retrouvée pour les mêmes longueurs d'hydratation tout en prenant en compte la saturation par le vide technologique radial. Par contre, la comparaison des évolutions et des valeurs de pressions de gonflement était plus compliquée vu les différences de configurations des essais / In order to verify the effectiveness of the geological high-level radioactive waste disposal, the French Institution of Radiation protection and Nuclear Safety (IRSN) has implemented the SEALEX project to control the long-term performance of swelling clay-based sealing systems, and to which this work is closely related. Within this project, In-situ tests are carried out on compacted bentonite-sand mixture in natural conditions and in a representative scale. This material is one of the most appropriate sealing materials because of its low permeability and good swelling capacity. Once installed, this material will be hydrated by water from the host-rock and start swelling to close all gaps in the system, in particular the internal pores, rock fractures and technological voids. Afterwards, swelling pressure will develop. In the present work, laboratory experiments were performed to investigate the sealing properties under this complex hydro-mechanical conditions taking into consideration the effect of technological voids. The microstructure of the material in its initial state was first examined by microfocus X-ray computed tomography (µCT). This allowed identification of the distribution of grains of sand and bentonite as well as the pores in the sample. Macro-pores are found concentrated at the periphery of the sample and between the grains of sand, which could affect in the short term the permeability. The hydration of the same material in limited swelling conditions was then observed by 2D photography and 3D µCT. The swelling mechanism with bentonite gel production, the swelling kinetics, the density decrease and the homogenisation of the material were analyzed. The hydration in the conditions of prevented swelling was also studied by swelling pressure tests with radial and axial measurements of swelling pressure. The difference found between the axial and radial swelling pressures suggested the presence of an anisotropic microstructure. Mock-up tests at a 1/10 scale of the in situ SEALEX tests were carried out for the study of the recovery capacity of the mixture in case of an accident causing the failure of the confining structures. Local measurements of swelling pressures along the sample allowed analysis of the density gradient evolution during axial swelling. Finally, a comparison between the laboratory results and those from an in-situ test was done, showing a good fitting in the relative humidity curves for the same infiltration length while considering the saturation effect from the technological void. The swelling pressure comparison was more complex because of the different configurations of the tests (existence of technological void in-situ that could affect the kinetics)

Electricity generation at nuclear power plants produces a large amount of high-level radioactive waste (HLW) every year, which has long-term detrimental effects on humans and the environment. Other applications of nuclear technology (e.g., medicine, research, nuclear weapons, industry) also produce radioactive waste (e.g., low-level radioactive waste, LLW, Intermediate-level waste, ILW). The potential of deep geological repositories (DGRs) as an option for disposal of radioactive waste (HLW, ILW, LLW) has been examined in several countries, including Bulgaria, Canada, China, Finland, France, Germany, India, Japan, Russia, Spain, Sweden, Switzerland, Ukraine and the United Kingdom and are still under discussion. In Ontario, Canada, DGRs with a multi-barrier system comprised of a sedimentary rock formation (i.e., a natural barrier) and an engineered barrier system (EBS) are currently under consideration. An EBS consists of various components, such as waste containers, buffer, backfill, and tunnel sealing materials, intended to prevent the release of radionuclides. Several engineered barrier materials, including a mixture of bentonite and sand, are currently being considered for use in DGRs for nuclear waste in Ontario. Bentonite has some advantageous physical and chemical properties, such as low permeability, high plasticity, and high swelling potential, which provide it with a good sealing ability and thus make it an effective barrier. However, interaction between the compacted bentonite–sand mixture and underground water chemistry fluids (chemical factor) in the DGR could significantly alter the favourable properties of bentonite (e.g., swelling potential), thus influencing its performance when used in an EBS and eventually jeopardizing the overall safety of DGRs. In addition, other parameters, such as the clay content, initial dry density and moisture content of the compacted barrier (physical factors), as well as the presence of salts in groundwater may affect the physical and physiochemical properties of barrier materials. Moreover, during the lifetime of a DGR for used spent fuel, the bentonite–based barrier material will not only be exposed to a broad range of groundwaters with different chemical compositions, but also to high temperatures (heat generated by the nuclear wastes) (thermal factor). Thus, the interaction between the compacted bentonite–sand mixture, the surrounding groundwater and the heat from the nuclear waste material could jeopardize the favourable properties of the bentonite-based (bentonite-sand) barrier material. Properties of a bentonite-sand barrier is an important characteristic to study while designing and constructing an EBS for a DGR. Thus, to understand and assess the operations of DGRs in Ontario, comprehensive studies must be performed on engineering properties like swelling behaviour, permeability, and thermal conductivity. The goal of this research study is to experimentally investigate the physical, chemical and thermal factors that influencing the engineering properties of a barrier material made up of bentonite-sand composite used in DGRs for nuclear waste in Ontario. Compacted samples are subjected to one-dimensional free swell test to understand the swelling behaviour of the material. Hydraulic conductivity was investigated using a flexible wall permeability test. Thermal conductivity and diffusivity were tested using Decangon KD2 Pro with TR-1 and and KS-1 sensors. The specimens contain different bentonite–sand mixture ratios (20:80, 30:70, 50:50, and 70:30 dry mass), and they are tested under conditions with differing bentonite content, dry density, groundwater chemistry, and temperature. Additional tests were conducted to investigate the microstructure of the specimens. These tests include X-ray diffraction (XRD) analysis, mercury intrusion porosimetry (MIP), and thermogravimetric analyses (TG/DTG). The results reveal that the time and strain required to achieve maximum swelling of compacted bentonite–sand specimens increase with the increase of initial dry density. The simulated saline solutions of Guelph and Trenton groundwater are found to suppress the swelling of the bentonite–sand specimens. This in turn leads to the increase of hydraulic conductivity and decrease of thermal properties of the barrier material. However, the impact of the salinity is significantly reduced by increasing the dry densities and sand content of the compacted material. Moreover, the coupled effect of salinity and temperature decreases the swelling potential of the bentonite-sand mixture. Also, some transformation of Na-montmorillonite into Ca-Montmorillonite was observed. The results also indicate that some montmorillonites might have been transformed into illites, thereby further decreasing the swelling potential of the bentonite-based barrier.