The volcanic rocks of the Es3 Formation (the third member of the Eocene Shahejie) in the Liaohe Eastern Sag can be divided into four facies and twelve subfacies. The porosity spectrum, porosity bin, variation coefficient (VC), and porosity width derived from electrical imaging log data were applied to study and characterize the heterogeneity of four facies and nine subfacies, both qualitatively and quantitatively. However, the VC and porosity width cannot be used to quantitatively classify heterogeneity when the VC is small and the porosity width is large. In the present study, the authors propose a new parameter, P_vcd = variation coefficient × porosity width. Based on this parameter, using a combination of porosity spectra, porosity bin features, VC, and porosity width, lithofacies heterogeneity is divided into three categories. The first is weak heterogeneity, which has a P_vcd < 1.1, a VC < 0.15, and a porosity width < 6. The second is moderate heterogeneity, which has a P_vcd 1.1–1.6, a VC 0.15–0.25, and a porosity width 6–9. The third is strong heterogeneity, which has a P_vcd > 1.6, a porosity VC > 0.25, and a porosity width > 9. In these three cases, the porosity spectra mainly display unimodal features, the porosity bins are concentrated, the bimodal features with tails and porosity bins are less concentrated, or the multimodal features with tails and porosity bins are scattered, respectively. Favorable reservoirs of volcanic rocks are controlled by lithofacies. In the study area, the favorable reservoirs appear to be the diatreme subfacies with low or medium heterogeneity, the pyroclastic flow subfacies with low heterogeneity, the compound lava flow subfacies with low or medium heterogeneity, and the outer zone subfacies with strong heterogeneity.

To address the challenges of seismic attribute selection and the low accuracy of characterizing single-phase river channels with a single seismic attribute, a 3D seismic attribute fusion technology based on lithology probability curve analysis has been developed. This method uses lithology probability curve analysis to optimally select amplitude attributes that highlight the distribution characteristics of sedimentary sand bodies and coherence attributes that emphasize variations in sand body boundaries. These attributes are then fused to form a new attribute volume, effectively solving the issues of seismic attribute selection and the low precision of single attributes in characterizing river channels. The approach enables precise identification of single-phase river channels and improves the sand body prediction accuracy in the study area by 20.2% points. The results have been applied to guide the optimization of 112 well interventions, resulting in an accumulated oil increase of 10,861 tons. This demonstrates favorable outcomes and provides a geological foundation for the accurate exploitation of residual oil in high-water-cut late-stage reservoirs.

Although water and gas drives are still the most commonly used methods for most reservoirs, the high clay content adversely affects production efficiency. This study investigates three development methods (water, CO ₂ , and N ₂ ) in high-clay reservoirs. It aims to identify factors affecting productivity and explain the underlying mechanisms. The mineral composition of natural cores was analyzed using X-ray fluorescence (XRF) and X-ray diffraction (XRD) techniques. We studied water, CO ₂ , and N ₂ drives in tight oil reservoirs through both core experiments and molecular simulations. The simulations focused on quartz and clay minerals (kaolinite and montmorillonite). Due to the limitation of reservoir fracturing pressure, the recovery rates of the natural cores were found to be as follows: water drive > CO₂ drive > nitrogen drive. Molecular simulations showed that all three replacement media performed best on quartz surfaces, with water drives being most effective on kaolinite surfaces and CO₂ drives being most effective on montmorillonite surfaces. The negatively charged surface of kaolinite, affected by the adsorption capacity of minerals, leads to weak CO₂ adsorption, resulting in poor recovery. Increasing the pressure to 35 MPa increased the CO₂ drive efficiency to 96.3%, confirming that the pressure increase contributes to the formation of a mixed phase. In kaolinite reservoirs, nitrogen drive outperformed CO₂ thanks to its non-polar properties and gas expansion effect, resulting in better penetration into micropores and less interference from clay minerals. Finally, the water drive performed poorly on both clay surfaces due to hydration and swelling of the clay caused by low salinity. The higher ion exchange capacity of montmorillonite, meanwhile, exacerbated pore plugging.