Modeling and Simulation of Carbon Dioxide Storage in Geological Layers

This dissertation advances large-scale simulation of carbon dioxide storage by developing a novel mathematical model and numerical schemes, which are validated by benchmark and real-world cases.

Overview

Carbon Capture and Storage (CCS) is a crucial technology for achieving a net-zero energy system. A deep understanding of the carbon underground storage processes is essential to predict the fate of injected CO₂ and other components in the reservoir. Therefore, modeling and simulating the behavior of carbon dioxide in large geological layers is demanding. This thesis develops a kinetic multi-phase multi-component model for large-scale simulations of CO₂ flow and transport in heterogeneous porous media. Not only are flow and transport considered, but the kinetic dissolution and capillary effects are also taken into account. A fully coupled and implicit numerical framework is proposed for solving the governing equations. The Vertex Centered Finite Volume Method is employed for the spatial discretization of the mass conservation equations. The full upwind scheme and the interface conditions for saturation discontinuity in heterogeneous porous media are implemented. The linearly implicit extrapolation scheme (LIMEX) with the error estimator is adapted for adaptive time stepping. The arising linear system of equations is solved by using the Geometric Multigrid (GMG) method. The parallel implementation is based on the open-source software UG4. The validation of the proposed schemes is verified by comparing them with analytical solutions in benchmark cases. The weak and strong scaling tests are performed on the supercomputer Shaheen III with up to 4096 processor cores. The results show that the implemented software is parallelized and highly scalable for problems with billions of unknowns, which makes it a great candidate for high-resolution reservoir simulations. Last but not least, the performance of CO₂ injection and storage over long periods is tested with real-world settings.

Presenters

Brief Biography

Shuai Lu is a Ph.D. candidate in Computer Science at King Abdullah University of Science and Technology (KAUST), supervised by Professor Gabriel Wittum and working in the Modeling and Simulations Lab. He completed his B.Eng. in Mechanical Engineering from China University of Mining and Technology (CUMT) in 2015. He received his master’s degree in Solid Mechanics at Beihang University in 2018.