Probabilistic Analysis of Radionuclide Transport for Radioactive Waste Disposal Systems in Soil and Fractured Rocks
The safe management of radioactive waste disposal facilities is increasingly becoming a necessary condition for the future development of nuclear industry. To ensure long-term safety of these disposal facilities, performance assessment models need to be developed that can quantitatively estimate the potential impact of disposal on biosphere. These models facilitate in envisioning the extent of safety achieved due to the isolation of waste and estimate the amount of risk caused by the failure of disposal systems by considering various scenarios of release and pathways of intrusion to the geosphere. They also treat the uncertainties associated with input characteristics and quantify them using probabilistic techniques. In this thesis, efficient performance assessment models are developed focusing mainly on understanding the migration process of low and intermediate level radioactive wastes through the geological medium into the biosphere by predictive radionuclide transport models. These models are developed for different geological environments which include soil and fractured rocks. As an intrinsic part of performance assessment, the aleatory (spatial variability due to inherent randomness in soil properties) and epistemic uncertainties (due to parameter and model uncertainties) in the geological and transport properties of the medium and radionuclides are quantified using various probabilistic methods. These methods have been implemented to estimate the limits above which the eventual release of radioactive wastes from disposal facility will pose unacceptably high risks. Also, the critical parameters affecting the radionuclide migration are evaluated using various sensitivity methods. Several algorithms are developed and implemented in PYTHON and MATLAB to add new features that introduce complexity in the numerical models and also, interface the deterministic and probabilistic analyses.
- Civil Engineering (CiE)