Uncertainties in THM-coupled Integrity Calculations
Country / Region: Deutschland
Begin of project: November 1, 2021
End of project: October 31, 2024
Status of project: November 14, 2023
Fig.1: Illustration of the sample positions according to the sparse grid method
Source: BGR
In the project, BGR is working on tasks to extend the methodology for numerical integrity analyses of repositories for radioactive materials. In order to demonstrate the preservation of those rock properties that are responsible for the confinement of radioactive waste, thermally, hydraulically, mechanically (THM) coupled numerical analyses are carried out to test the integrity of the containment providing rock zone. These analyses include models of the geological situation, the description of the material behaviour of the rock units, and their parameterisation. Much of the input data of the numerical analyses is subject to uncertainties for various reasons.
In this project, methods shall be developed to quantify the influence of the uncertainties in the input parameters of the chosen material description on the results of the THM coupled numerical integrity analyses. Input parameters, or their combinations, with the greatest influence on the results, will be investigated using different sensitivity analysis methods. To demonstrate the applicability, the developed methods will be applied to two test cases, and the findings and experience gained will be transferred into recommendations for action for the consideration of parameter uncertainties in integrity analyses.
In cooperation with the partners from the Institute of Geotechnics at TUBA Freiberg, the Department of Environmental Informatics at the UFZ in Leipzig, and the Professorship of Numerical Mathematics at Chemnitz University of Technology, as well as the BGR departments "Geotechnical Safety Analyses " and "Rock Characterisation for Storage and Final Disposal", the project brings together different areas of expertise to make modern methods from mathematics and information technology suitable for practical use. As part of a BGE research cluster on "Uncertainties and Robustness with regard to the Safety of a repository for high-level radioactive waste" (URS), there is an exchange with other collaborative projects.
Fig. 2: Model of a generic repository system in claystone
Source: BGR
The methods developed in the project are to be applied to a generic repository system in claystone, which was developed in previous projects. Statistical information on the input data will form the basis for the evaluation of the resulting uncertainties in the target variables. This requires a suitable evaluation of the stochastic results to assess the integrity.
Fig. 3: Different contributions to the measurement uncertainty in rock mechanics laboratory investigations
Source: BGR
In accordance with the requirements of the case studies, the available database on relevant parameters for the scenario simulations is collected and processed for statistical treatment. Possible physical origins of parameter uncertainties are to be identified and, as far as possible, described in terms of their relevance and statistical characteristics. For the dominant input parameters identified on the basis of the numerical case studies, possibilities and limits of further experimental specification of these parameters will be investigated. In so far as possibilities are identified, they will be characterised and their practicability evaluated with regard to the type of measures required, e.g. extension of the data base; optimisations with regard to sample availability; improvements in the experimental procedure.
