Designing a system for the safe disposal of spent fuel and high-level defense waste will require being able to predict repository performance over time spans exceeding the total known time of human existence on this planet.
Historically, primary reliance has been placed on natural (geological) barriers to prevent or retard release of radionuclides to the accessible environment. These have included crystalline rock (Sweden, Finland, Switzerland and Canada), salt domes (Germany), clay (Belgium), and tuff (U.S.A.). In all countries except the United States, the potential disposal environments are saturated with moisture. By contrast, the site at Yucca Mountain, Nevada, currently being characterized by the United States, is located in an arid climate, and the zone proposed for the repository is an unsaturated zone well above the water table. As a result, one can say that the site under evaluation in the U.S.A. is unique. It provides special opportunities and challenges with respect to radioactive waste disposal.
Recently, there has been some “erosion,” of confidence in international circles about the ability of geological barriers to provide the degree of containment originally assumed. This has led to an increased interest in a number of countries in the development of engineered barriers of various kinds that work in concert with the natural barriers. Designs under consideration cover a wide range of concepts, for example, special materials for more robust waste containers; new concepts for the emplacement of waste containers--such as room or drift emplacement; chemical control of the container environment; use of thermal energy as a design parameter, and growing interest in a more “systems oriented approach” to the storage, transportation and emplacement of spent fuel and high-level waste through the use of multi-purpose containers.
Because of increased interest in multi-purpose containers and because they may allow us to improve our ability to predict radionuclide containment over long periods of time, corrosion behavior of the various materials remains a major consideration. As experiments to predict and measure corrosion behavior are designed, we must also develop statistical methods to evaluate the data collected during these experiments.