Experimental Optimization and Empirical Correlations of CO2 Huff-and-Puff in Tight Sandy Conglomerate Reservoirs

Abstract The efficient development of tight sandy conglomerate reservoirs remains challenging due to their complex pore structures and ultra-low permeability. This study aims to provide operational guidance for CO2 huff-and-puff, a promising enhanced oil recovery (EOR) technique for such reservoirs. A systematic series of core-flooding experiments was conducted to quantify the impact of five key operational parameters: huff-and-puff timing, permeability, CO2 injection volume, soaking time, and cycle number. The results provide quantitative design criteria: initiating the process early at a higher reservoir pressure (37.0 MPa) yielded a recovery factor of 17.86%, which is 3.2 times greater than that achieved at the bubble-point pressure. Increasing permeability from 0.05 to 2.94 mD enhanced recovery by 11.15% by improving sweep efficiency. An injection volume of 0.50 pore volume (PV) was identified as the optimal trade-off between recovery factor (17.86%) and CO2 utilization efficiency (0.19 m3/m3). Soaking beyond 120 min offered diminishing returns (yielding <1% incremental recovery), and three operational cycles were determined as the economic limit, achieving a cumulative recovery of 26.83%. Furthermore, high-fidelity empirical correlations (R2 > 0.96) were developed, revealing distinct hierarchies of parameter influence: for the recovery factor, the order is huff-and-puff timing > injection volume > permeability > soaking time. For the CO2 oil displacement efficiency, the hierarchy is huff-and-puff timing > permeability > injection volume > soaking time. This work translates experimental insights into practical models and clear operational guidelines, providing a direct pathway to optimize CO2 huff-and-puff performance in analogous tight conglomerate reservoirs.