A modeling framework for CO2-storage in depleted gas reservoirs

This work performs a complete framework of numerical simulation of CO2-Injection into depleted gas reservoirs against the background of enhanced gas recovery and CO2-Storage. This framework ranges from model development to site-specific scenario simulations and result interpretation. Numerical simulations of gas related applications such as CO2 sequestration, geothermal energy production, or natural gas storage have to consider non-isothermal effects caused by gas compression or expansion. This mathematical approach results in a system of coupled non-linear PDEs, which have been implemented into the open-source software platform OpenGeoSys. For model verification purposes, a number of well-known benchmark tests and analytical solutions of simplified or adapted conditions has been utilized to prove the validity of the developed simulation tool. Fluid material parameters are obtained by applying highly accurate and state-of-the-art property correlations. However, the accuracy of these correlations is strongly depending on the precision of the chosen equation of state, which provides a relation between the system state variables pressure, temperature, and composition. To guarantee a high level of accuracy, four commonly used equations of state (EOS) have been chosen from literature and have been evaluated by comparison using a large number of measurement datasets. Complex EOS reach a much better precision than simple ones, but lead to expansive computing times. Therefore, comparative simulations have been performed to investigate the effects of EOS differences on numerical simulation results. The comparison shows, that little differences in the density determination may lead to significant discrepancies in simulation results. Applying a compromise among precision and computational effort, a cubic EOS has been chosen to simulate the continuous injection of carbon dioxide into a depleted natural gas reservoir. This simulation allows to investigate physical phenomena which appear during injection and to predict the evolution of reservoir pressures and temperatures. Investigating multiple scenarios, this model helps to find the best injection strategy for enhanced gas recovery applications.