This study developed a data-driven surrogate model based on a deep neural network (DNN) to evaluate gas–liquid multiphase flow occurring in horizontal pipes. It estimated the liquid holdup and pressure gradient under a slip condition and different flow patterns, i.e., slug, annular, stratified flow, etc. The inputs of the surrogate modelling were related to the fluid properties and the dynamic data, e.g., superficial velocities at the inlet, while the outputs were the liquid holdup and pressure gradient observed at the outlet. The case study determined the optimal number of hidden neurons by considering the processing time and the validation error. A total of 350 experimental data were used: 279 for supervised training, 31 for validating the training performance, and 40 unknown data, not used in training and validation, were examined to forecast the liquid holdup and pressure gradient. The liquid holdups were estimated within less than 8.08% of the mean absolute percentage error, while the error of the pressure gradient was 23.76%. The R² values confirmed the reliability of the developed model, showing 0.89 for liquid holdups and 0.98 for pressure gradients. The DNN-based surrogate model can be applicable to estimate liquid holdup and pressure gradients in a more realistic manner with a small amount of computating resources.

The integrated injection of low-salinity water (LSW) and carbon dioxide (CO₂) into the water-alternating-gas (WAG) process offers advantages, primarily increasing oil recovery and reducing operating costs. However, CO₂ has challenges in sweep efficiency due to significant differences in density and viscosity compared with oil. While LSW and dimethyl ether (DME) have shown promise in improving recovery through wettability alteration and reducing minimum miscible pressure, interfacial tension (IFT), and CO₂ mobility, their synergistic integration with CO₂-WAG remains poorly understood. Existing DME-based enhanced oil recovery (EOR) studies have not explored low-salinity water injection as a cost-effective alternative to mitigate high DME operating costs. This study introduces the CO₂/DME-LSWAG method, systematically evaluating the effect of DME concentrations (0%, 10%, 25%) and LSWs (seawater, twice-diluted seawater, ten-times-diluted seawater) on sweep and displacement efficiencies, oil recovery, and CO₂ storage in a 2D cross-sectional carbonate reservoir model. Results showed that DME dramatically reduces IFT (67% and 95% at 10% and 25% DME solvent, respectively) while salinity effects are relatively small. Compared with CO₂-LSWAG, the oil recovery factor improved by 5.2–13.1% depending on DME concentration and water salinity, with DME performance maximized at higher salinity water. CO₂ storage efficiency showed opposing trends. Structural trapping decreased, while solubility trapping increased with lower salinity. The sensitivity analysis identified DME concentration as the dominant factor for CO₂ storage. The composition modeling and simulation of the CO₂/DME-LSWAG process provide critical engineering guidance for the design of future EOR and CO₂ storage projects that utilize DME in carbonate reservoirs.