AbstractsEngineering

Abstract

Due to concerns about rising CO2 emissions from fossil fuel power plants, there has been a strong emphasis on the development of a safe and economical method for Carbon Capture Utilization and Storage (CCUS). One area of current interest in CO2 utilization is the Enhanced Oil Recovery (EOR) from depleted reservoirs. In an Enhanced Oil Recovery system, a depleted or depleting oil reservoir is re-energized by injecting high-pressure CO2 to increase the recovery factor of the oil from the reservoir. An additional benefit beyond oil recovery is that the reservoir could also serve as a long-term storage vessel for the injected CO2. Although this technology is old, its application to depleted reservoirs is relatively recent because of its dual benefit of oil recovery and CO2 storage thereby making some contributions to the mitigation of anthropogenic CO2 emissions. Since EOR from depleted reservoirs using CO2 injection has been considered by the industry only recently, there are uncertainties in deployment that are not well understood, e.g. the efficiency of the EOR system over time, the safety of the sequestered CO2 due to possible leakage from the reservoir. Furthermore, it is well known that the efficiency of the oil extraction is highly dependent on the CO2 injection rate and the injection pressure. Before large scale deployment of this technology can occur, it is important to understand the mechanisms that can maximize the oil extraction efficiency as well as the CO2 sequestration capacity by optimizing the CO2 injection parameters, namely, the injection rate and the injection pressure. In this thesis, numerical simulations of subsurface flow in an EOR system is conducted using the DOE funded multiphase flow solver COZView/COZSim developed by Nitec, LLC. A previously developed multi-objective optimization code based on a genetic algorithm developed in the CFD laboratory of the Mechanical Engineering department of Washington University in St. Louis is modified for the use the COZView/COZSim software for optimization applications to EOR. In this study, two reservoirs are modeled. The first is based on a benchmark reservoir described in the COZSim tutorial; the second is a reservoir in the Permian Basin in Texas for which extensive data is available. In addition to pure CO2 injection, a Water Alternating Gas (WAG) injection scheme is also investigated for the same two reservoirs. Optimizations for EOR Constant Gas Injection (CGI) and WAG injection schemes are conducted with a genetic algorithm (GA) based optimizer combined with the simulation software COZSim. Validation of the obtained multi-objective optimizer was achieved by comparing its results with the results obtained from the built-in optimization function within the COZView graphic user interface. Using our GA based optimizer, optimal constant-mass and pressure-limited injection profiles are determined for EOR. In addition, the use of recycled gas is also investigated. Optimization of the EOR problem results in an increased recovery factor with a…