A Coreflood Simulation Study of a Systematic Co2 Huff-N-Puff Injection for Storage and Enhanced Gas Condensate Recovery

Recently some core flood experimental data were reported following a new CO₂ Huff-n-Puff (H-n-P) injection technique. This technique optimized CO2 injection pressure and volume to improve CO₂/resident fluid interactions for enhanced gas and condensate recovery and CO₂ storage purposes. This simulation study aims to complement and generalize the corresponding core flood experimental results. The simulation data confirm the dominant governing mechanism and the importance of using appropriate CO2/gascondensate kr data while accounting for the effect of compositional changes on gas and condensate mobility during CO2 H-n-P injection. Laboratory PVT tests were performed to generate relevant data sets that describe the complex phase behavior changes when CO2 interacts with gas condensate systems. These data sets were applied for EOS tuning, phase behavior prediction, and quantifying the level of CO2/gas condensate interactions. A CO2 H-n-P injection core-flood simulation model was developed. H-n-P injection cycles with the incremental injection of CO2 volumes were simulated to replicate experimental procedures performed on a high-permeability Berea sandstone core. Experimental data showed that conventional CO2 H-n-P injection treatment significantly improves hydrocarbon gas and condensate recovery efficiency but at the cost of injecting and producing high volumes of CO2. While the proposed method applied at the maximum condensate saturation for the corresponding CO2/gas-condensate mixture can match the recovery efficiency achieved when applying the conventional injection technique, but with much lesser volumes of CO2 injected and produced. The relative permeability data measured for gas and condensate fluids (GC-kr) were significantly affected by the compositional changes resulting from CO2/resident fluid interactions below the saturation pressure. The numerical model predicted a close match for the pressure profile after adjusting the GC-kr data. However, it could only match the production profile for the pre-CO2 and first CO2 injection cycle, where the volume of CO2 injected was small and had a negligible effect on condensate recovered relative to the volume of condensate in place. Sensitivity analyses were performed on GC-kr data attempting to history match the experimental and simulated data. The generated data were analyzed to quantify the effects of CO2/resident fluid interactions on condensate revaporization and the model's predictability. These data will aid in bridging the gap in the level of CO₂/gas-condensate interactions during CO2 flooding, which is vital for designing an efficient CO₂ H-n-P injection process.