Greenhouse Gas Capture and Storage in Depleted Reservoirs to Limit Greenhouse Gas Emissions and to be Used for Enhanced Oil Recovery in Cuu Long Basin, Vietnam

ABSTRACT

This study evaluates the simultaneous co‐optimization of carbon dioxide storage and enhanced oil recovery (CO ₂ ‐EOR) in the X Oilfield, Cuu Long Basin, Vietnam, to assess long‐term CO ₂ storage integrity over 300 years. A dynamic reservoir model was developed from a geological model constructed in Petrel and validated using ECLIPSE simulation, achieving approximately 90% accuracy against historical production data. On the basis of the validated model, sensitivity analysis was conducted to investigate the impact of key operational parameters on CO ₂ ‐EOR performance and CO ₂ sequestration. The results indicate that a CO ₂ injection rate of 20,000 Mscf/day maximizes oil recovery. In contrast, a higher injection rate of 40,000 Mscf/day yields the most significant economic benefit, driven by carbon credit incentives and government subsidies. To address the computational cost and time limitations of conventional numerical simulation, a surrogate reservoir model (SRM) based on long short‐term memory (LSTM) neural networks was developed. The SRM model rapidly produces forecasting and facilitates uncertainty quantification through 100 Monte Carlo realizations. Probabilistic results confirm the project's technical and economic feasibility, with a total CO ₂ storage capacity of approximately 16.1 million tons. Long‐term simulations extending to the year 2300 demonstrate stable CO ₂ plume migration and effective geological containment, confirming the viability of the proposed co‐optimization strategy for oilfields approaching late‐stage production.