Nuclear power generation is being regarded as a solution to ever increasing demand for electricity, and concerns over global warming and climate change due to the use of fossil fuels. Although nuclear power generation is considered to be reliable, economical, clean, and safe, the wastes produced from the nuclear fuel cycle are not, and can remain hazardous for hundreds of thousands of years. An international consensus has developed over the past several decades that deep geologic disposal of low, intermediate, and high level radioactive wastes is the best option to protect the biosphere. In this thesis, both regional scale and sub-regional scale models are created to simulate groundwater flow and transport for a representative Canadian Shield setting, honouring site-specific topography and surface water features. Sub-surface characteristics and properties are derived from numerous geoscience studies. In addition, a regional scale model is developed, centred on the Bruce Nuclear Power Development (BNPD) site near Tiverton, Ontario, and located within the Michigan Basin. Ontario Power Generation (OPG) has proposed a Deep Geologic Repository (DGR) for low & intermediate level waste (L&ILW) at the BNPD site. Paleoclimate simulations using various combinations of parameters are performed for both the Canadian Shield Sub-Regional model, and the Michigan Basin Regional model. Fracture zone permeability is a very important parameter when modelling crystalline rock settings. Migration of a unit tracer representing glacial recharge water can occur to depth in fractures of high permeability. Representative rock compressibility values are necessary as compressibilities are used to calculate storage coefficients, and the one-dimensional loading efficiency; these affect the subsurface propagation of elevated pore pressures due to glacial loading at surface. Coupled density-dependent flow and transport in paleoclimate simulations affects deep flow systems and provides a measure of flow stability, as well as increasing the mean life expectancy at depth. Finally, hydromechanical coupling is a very important mechanism for reducing vertical hydraulic gradients during a glaciation event when a hydraulic boundary condition equal to the pressure at the base of an ice-sheet is applied at ground surface. Pore water velocities are reduced, thereby retarding migration of surface waters into the subsurface environment.

Nuclear power generation is being regarded as a solution to ever increasing demand for electricity, and concerns over global warming and climate change due to the use of fossil fuels. Although nuclear power generation is considered to be reliable, economical, clean, and safe, the wastes produced from the nuclear fuel cycle are not, and can remain hazardous for hundreds of thousands of years. An international consensus has developed over the past several decades that deep geologic disposal of low, intermediate, and high level radioactive wastes is the best option to protect the biosphere. In this thesis, both regional scale and sub-regional scale models are created to simulate groundwater flow and transport for a representative Canadian Shield setting, honouring site-specific topography and surface water features. Sub-surface characteristics and properties are derived from numerous geoscience studies. In addition, a regional scale model is developed, centred on the Bruce Nuclear Power Development (BNPD) site near Tiverton, Ontario, and located within the Michigan Basin. Ontario Power Generation (OPG) has proposed a Deep Geologic Repository (DGR) for low & intermediate level waste (L&ILW) at the BNPD site. Paleoclimate simulations using various combinations of parameters are performed for both the Canadian Shield Sub-Regional model, and the Michigan Basin Regional model. Fracture zone permeability is a very important parameter when modelling crystalline rock settings. Migration of a unit tracer representing glacial recharge water can occur to depth in fractures of high permeability. Representative rock compressibility values are necessary as compressibilities are used to calculate storage coefficients, and the one-dimensional loading efficiency; these affect the subsurface propagation of elevated pore pressures due to glacial loading at surface. Coupled density-dependent flow and transport in paleoclimate simulations affects deep flow systems and provides a measure of flow stability, as well as increasing the mean life expectancy at depth. Finally, hydromechanical coupling is a very important mechanism for reducing vertical hydraulic gradients during a glaciation event when a hydraulic boundary condition equal to the pressure at the base of an ice-sheet is applied at ground surface. Pore water velocities are reduced, thereby retarding migration of surface waters into the subsurface environment.

A deposição de rejeitos radioativos em repositórios geológicos profundos vem sendo estudada nos últimos anos em diversos países. Materiais à base de cimento são utilizados nesses repositórios como material estrutural, matriz de imobilização de rejeitos ou material de preenchimento. Compreender o desempenho desse material é essencial para garantir a segurança da instalação durante o seu tempo de vida útil (de milhares a centenas de milhares de anos, dependendo do tipo de rejeito). Este trabalho objetiva modelar o comportamento em longo prazo do cimento Portland e estudar a influência de diversos fatores na hidratação e na evolução desse material. A modelagem descritiva abordou a hidratação do cimento nas condições ambientais esperadas no repositório e os efeitos desses fatores em propriedades mecânicas, mineralógicas e morfológicas do cimento. Os fatores ambientais considerados relevantes neste trabalho foram: alta temperatura e pressão, penetração de água subterrânea contendo íons quimicamente agressivos ao cimento e a presença do campo de radiação proveniente dos rejeitos. Ensaios acelerados de degradação também foram realizados para corroborar com o modelo descrito. Observou-se uma sinergia entre diversos fatores na degradação do cimento, como a influência da temperatura e da radiação em reações deletérias ao material. O resultado da modelagem apontou três principais possíveis causas de falha nas barreiras artificiais: a) a formação de um caminho preferencial; b) a perda de resistência e coesão do material; e c) o aumento na corrosão das estruturas metálicas. A descrição do modelo apresentada é a base para a modelagem matemática e a análise de segurança dos repositórios estudados no Brasil. / The radioactive waste disposal in deep geological repositories has been studied for many countries in the last years. Cementitious materials are used in these repositories as structural material, immobilization matrix and as backfill material. The understanding of the performance of these materials is essential to ensure the safety of the installation during its life time (from thousand to hundreds of thousands of years, depending on the type of waste). This works aims at modeling the long-term performance of Portland cement and study the influence of many environmental factors in the hydration and evolution of this material. The modeling approached the cement hydration in the conditions expected in the repository and the effects of these factors on cement mechanical, mineralogical and morphological properties. The environmental factors considered relevant was: high temperature and pressure, the penetration of groundwater containing aggressive chemical ions, and a radiation field from the waste. Degradation accelerated tests were done to corroborate with the descriptive model. It was observed a synergism between some factors on the cement degradation, as the influence of temperature and radiation field in some deleterious reactions in the material. The results of modeling pointed to three main causes of engineered barrier failure: a) the formation of a preferential pathway; b) loss of resistance and cohesion in the material; and c) the increase in the metallic structures corrosion process. The descriptive model is the basis for a mathematical modeling and to perform the safety assessment of the repositories studied in Brazil.

A deposição de rejeitos radioativos em repositórios geológicos profundos vem sendo estudada nos últimos anos em diversos países. Materiais à base de cimento são utilizados nesses repositórios como material estrutural, matriz de imobilização de rejeitos ou material de preenchimento. Compreender o desempenho desse material é essencial para garantir a segurança da instalação durante o seu tempo de vida útil (de milhares a centenas de milhares de anos, dependendo do tipo de rejeito). Este trabalho objetiva modelar o comportamento em longo prazo do cimento Portland e estudar a influência de diversos fatores na hidratação e na evolução desse material. A modelagem descritiva abordou a hidratação do cimento nas condições ambientais esperadas no repositório e os efeitos desses fatores em propriedades mecânicas, mineralógicas e morfológicas do cimento. Os fatores ambientais considerados relevantes neste trabalho foram: alta temperatura e pressão, penetração de água subterrânea contendo íons quimicamente agressivos ao cimento e a presença do campo de radiação proveniente dos rejeitos. Ensaios acelerados de degradação também foram realizados para corroborar com o modelo descrito. Observou-se uma sinergia entre diversos fatores na degradação do cimento, como a influência da temperatura e da radiação em reações deletérias ao material. O resultado da modelagem apontou três principais possíveis causas de falha nas barreiras artificiais: a) a formação de um caminho preferencial; b) a perda de resistência e coesão do material; e c) o aumento na corrosão das estruturas metálicas. A descrição do modelo apresentada é a base para a modelagem matemática e a análise de segurança dos repositórios estudados no Brasil. / The radioactive waste disposal in deep geological repositories has been studied for many countries in the last years. Cementitious materials are used in these repositories as structural material, immobilization matrix and as backfill material. The understanding of the performance of these materials is essential to ensure the safety of the installation during its life time (from thousand to hundreds of thousands of years, depending on the type of waste). This works aims at modeling the long-term performance of Portland cement and study the influence of many environmental factors in the hydration and evolution of this material. The modeling approached the cement hydration in the conditions expected in the repository and the effects of these factors on cement mechanical, mineralogical and morphological properties. The environmental factors considered relevant was: high temperature and pressure, the penetration of groundwater containing aggressive chemical ions, and a radiation field from the waste. Degradation accelerated tests were done to corroborate with the descriptive model. It was observed a synergism between some factors on the cement degradation, as the influence of temperature and radiation field in some deleterious reactions in the material. The results of modeling pointed to three main causes of engineered barrier failure: a) the formation of a preferential pathway; b) loss of resistance and cohesion in the material; and c) the increase in the metallic structures corrosion process. The descriptive model is the basis for a mathematical modeling and to perform the safety assessment of the repositories studied in Brazil.