Perpustakaan Digital ITB

This study evaluates the performance of CO2 storage in a deep saline aquifer using a line-drive well configuration with brine production. The main objective is to analyze how brine production and selected reservoir-operational parameters affect pressure buildup, CO2 plume migration, trapped CO2, dissolved CO2, and overall storage performance. A three-dimensional compositional reservoir model was developed using a commercial simulator to represent CO2 injection into a conceptual homogeneous deep saline aquifer. The model consists of one CO2 injection well and one brine production well arranged in a line-drive configuration. CO2 solubility in brine was modeled using the HENRY-MOD1-CO2 option, while residual trapping was represented through relative permeability hysteresis. Operational constraints were defined based on injection rate, maximum injector bottom-hole pressure, minimum producer bottom-hole pressure, liquid production rate, and allowable CO2 breakthrough. Sensitivity analysis and proxy modeling were then performed using a commercial simulator by varying permeability, salinity, reservoir depth, injection rate, and injector position. The base case result shows that CO2 can be injected continuously until the end of the simulation period, although the injector operates close to the maximum bottom-hole pressure constraint. The final reservoir pressure remains below the defined maximum injection pressure, indicating that the base case is still within the operational pressure limit. The CO2 plume develops around the injection well and does not reach the producer, as shown by zero gas rate at the production well. At the end of the base case simulation, trapped CO2 reaches approximately 7.4 × 109 mol, while dissolved CO2 reaches approximately 6.0 × 108 mol. The commercial simulator results show that CO2 trapped varies from 2.814 × 109 mol to 3.797 × 1010 mol, while dissolved CO2 ranges from 6.856 × 105 mol to 2.321 × 109 mol. The proxy model for CO2 trapped gives a good prediction quality, with an R-square value of 0.952 and prediction R-square of 0.884. In contrast, the dissolved CO2 proxy model has lower prediction quality, so it is more suitable for evaluating parameter trends rather than absolute prediction. Sensitivity analysis shows that salinity is the most dominant parameter for both trapped CO2 and dissolved CO2. This study adds value by combining brine production, line-drive configuration, sensitivity analysis, and proxy modeling in one integrated workflow for CO2 storage evaluation. The results provide a practical screening approach to identify dominant parameters and support early-stage optimization of CO2 storage performance in deep saline aquifers.