This project is part of the research theme "Model Legitimacy and Uncertainty Estimation for Large-scale Models" within the Research Training Group (RTG) "Integrated Hydrological Modeling." The RTG consists of a group of doctoral students from the University of Tübingen, the University of Stuttgart, and the University of Hohenheim. Its goal is to collaborate with Canadian partner institutions to further develop flow and reactive transport models for coupled hydrological systems in a unique environment. The Canadian partner for the current project is Prof. Kevin Mumford from Queen's University, Kingston.
Gas migration in porous media is a complex, nonlinear process controlled by multiphase mass transfer, multiphase flow, and diffusion within detailed pore geometries. However, we aim to predict gas migration paths in larger systems without needing to know the geometry of each individual pore. For such predictions, we want to find suitable mathematical models for computer-based simulations.
The overarching question is: What strategy should be used to construct such a model? A simple model with few parameters that is easy to calibrate and runs quickly? A complex, detailed model that could be highly accurate but requires a lot of experimental data to be created and calibrated, and then runs slowly on a computer? While many possible model concepts could be proposed, we are interested in how to choose optimally between these models.
The main goal of the project is, therefore, the development of a systematic, objective, and quantitative framework for comparing different versions of a specific model class (Invasion Percolation models) with laboratory experiments of gas injections into saturated sand under various flow regimes. The framework aims to assist in selecting a suitable model for prediction under the respective flow conditions.
