Analytical and Inverse Models for Leakage Characterization of Carbon Dioxide Storage

CO2 Storage (CS) in deep saline aquifers is a means of reducing atmospheric emissions of CO2 to mitigate anthropogenic climate change. Injecting large amounts of CO2 is proposed over a relatively short period of time. CO2 injectivity depends on number of parameters including the salt dry-out due to vaporization of the native brine by injected CO2. In the first part of this dissertation, salt dry-out during CO2 injection is studied. Salt dry-out can plug the reservoir and reduce CO2 injectivity. An analytical solution to evaluate the amount of salt precipitation and the region over which it occurs is developed. A time-dependent skin factor is introduced to assess the effect of salt dry-out on the permeability. The analytical model is used to study the sensitivity of precipitated salt saturation and CO2 plume extension to different reservoir parameters including: aquifer pressure, temperature, salinity, relative permeability functions, and volume change upon mixing.;One of the essential concerns in geologic storage of CO2 in deep saline aquifers is the risk of CO2 leakage. The sealing capacity of the cap-rock overlying the storage aquifer must therefore be evaluated, so characterization of the cap-rock is required even before starting the storage operations. Aqueous fluids can be injected in the target storage aquifer and the pressure change due to leakage can be monitored in an upper aquifer separated by a cap-rock. The main focus of this dissertation is on leakage characterization through pressure monitoring. Two new analytical forward solutions are derived which relate leak properties to pressure change at the monitoring aquifer. Using the analytical solutions significantly reduces the computational cost of solving leakage inverse problem, assist with decomposing the effect of different leakage parameters on pressure, and improve understanding the leakage dynamics. Obtaining the location and the transmissibility of the leak based on the pressure measurement is investigated. Uniqueness and stability of the solution is analyzed based on the inversion analysis techniques. Leak parameters are estimated over a wide confidence interval in response to pressure measurement error. A graphical method to obtain the leak parameters through analysis of pressure derivative is presented. The graphical method findings are used in a two-step manner to estimate the leak parameters with higher confidence. Design considerations to maximize the capability of leakage characterization are presented including pressure sampling frequency, pulsing, and multiple injectors/monitors. Placement of the monitoring well is investigated to ensure acquiring sufficient information for closely estimating the leak location and transmissibility.