In the Bonan area, the lower fourth member of the Shahejie Formation (Es4x) is buried beneath a sedimentary pile ranging from 2500 to 5000 m. Understanding the impact of diagenetic alterations on these deeply buried reservoirs is crucial for effective hydrocarbon exploration and production. This study employs a terminal fan sedimentation model, encompassing depositional environments such as feeder channels, distributary channels, floodplains, and basinal zones, to provide insights into the spatial distribution of reservoir properties and their influence on the localization of optimal reservoirs within the sag. The analysis integrates diagenetic facies with well log responses, subsurface porosity trends, and permeability variations across the formation. The petrographic analysis indicates that the sandstone is composed primarily of litharenite, feldspathic litharenite, lithic arkose, and minor amounts of arkose. The dominant clay cement is illite, accompanied by mixed-layer smectite/illite, chlorite, and kaolinite. Thin section observations reveal secondary porosity formed through the dissolution of quartz grains, volcanic rock fragments, and feldspar, along with their associated cements. These sandstones exhibit relatively good sorting, with average porosity and air permeability values of 14.01% and 12.73 mD, respectively. Diagenetic alterations are categorized into three processes: porosity destruction, preservation, and generation. Key diagenetic mechanisms include compaction, cementation, replacement, and dissolution, with compaction exerting the most significant control on reservoir porosity reduction. Statistical analysis indicates that the average porosity loss due to compaction is approximately 13.3%, accounting for about 38% of the original porosity. The detrital rock cement predominantly comprises quartz (42%), feldspar (32%), clay minerals (14%), and carbonate (12%). Under the prevailing depositional conditions, porosity is enhanced by dissolution and fracturing, while late-stage diagenetic cementation by clay and carbonate minerals—excluding chlorite—adversely affects reservoir quality. Consequently, the distributary zone is identified as the primary target for exploration.

Lacustrine shale in continental rift basins is complex and features a variety of mineralogical compositions and microstructures. The lithofacies type of shale, mainly determined by mineralogical composition and microstructure, is the most critical factor controlling the quality of shale oil reservoirs. Conventional geophysical methods cannot accurately forecast lacustrine shale lithofacies types, thus restricting the progress of shale oil exploration and development. Considering the lacustrine shale in the upper Es₄ member of the Dongying Sag in the Jiyang Depression, Bohai Bay Basin, China, as the research object, the lithofacies type was forecast based on two machine learning methods: support vector machine (SVM) and extreme gradient boosting (XGBoost). To improve the forecast accuracy, we applied the following approaches: first, using core and thin section analyses of consecutively cored wells, the lithofacies were finely reclassified into 22 types according to mineralogical composition and microstructure, and the vertical change of lithofacies types was obtained. Second, in addition to commonly used well logging data, paleoenvironment parameter data (Rb/Sr ratio, paleoclimate parameter; Sr %, paleosalinity parameter; Ti %, paleoprovenance parameter; Fe/Mn ratio, paleo-water depth parameter; P/Ti ratio, paleoproductivity parameter) were applied to the forecast. Third, two sample extraction modes, namely, curve shape-to-points and point-to-point, were used in the machine learning process. Finally, the lithofacies type forecast was carried out under six different conditions. In the condition of selecting the curved shape-to-point sample extraction mode and inputting both well logging and paleoenvironment parameter data, the SVM method achieved the highest average forecast accuracy for all lithofacies types, reaching 68%, as well as the highest average forecast accuracy for favorable lithofacies types at 98%. The forecast accuracy for all lithofacies types improved by 7%–28% by using both well logging and paleoenvironment parameter data rather than using one or the other, and was 7%–8% higher by using the curve shape-to-point sample extraction mode compared to the point-to-point sample extraction mode. In addition, the learning sample quantity and data value overlap of different lithofacies types affected the forecast accuracy. The results of our study confirm that machine learning is an effective solution to forecast lacustrine shale lithofacies. When adopting machine learning methods, increasing the learning sample quantity (>45 groups), selecting the curve shape-to-point sample extraction mode, and using both well logging and paleoenvironment parameter data are effective ways to improve the forecast accuracy of lacustrine shale lithofacies types. The method and results of this study provide guidance to accurately forecast the lacustrine shale lithofacies types in new shale oil wells and will promote the harvest of lacustrine shale oil globally.

The Eocene fourth member of the Shahejie formation (Es4x) in the Bonan Depression, Bohai Bay Basin, records syn-rift sedimentation under alternating arid and humid climates. It provides insight into how orbital-scale climatic fluctuations influenced tectonics, facies patterns, and reservoir distribution. This study integrates 406 m of core data, 92 thin sections, 450 km² of 3D seismic data, and multiple geochemical proxies, leading to the recognition of five facies associations (LFA): (1) alluvial fans, (2) braided rivers, (3) floodplain mudstones, (4) fan deltas, and (5) saline lacustrine evaporites. Three major depositional cycles are defined within the Es4x. Seismic reflections, well-log patterns, and thickness trends suggest that these cycles represent fourth-order lake-level fluctuations (0.8–1.1 Myr) rather than short 21-kyr precession rhythms. This implies long-term climate and tectonic modulation, likely linked to eccentricity-scale monsoon variability. Hyperarid phases are marked by Sr/Ba > 4, δ¹⁸O > +4‰, and thick evaporite accumulations. In contrast, Sr/Ba < 1 and δ¹⁸O < −8‰ reflect humid conditions with larger lakes and enhanced fluvial input. During wet periods, rivers produced sand bodies nearly 40 times thicker than in dry intervals. Reservoir quality is highest in braided-river sandstones (LFA 2) with 12%–19% porosity, preserved by chlorite coatings that limit quartz cement. Fan-delta sands (LFA 4) have <8% porosity due to calcite cementation, though fractures (10–50 mm) improve permeability. Floodplain mudstones (LFA 3) and evaporites (LFA 5) act as seals. This work presents a predictive depositional and reservoir model for arid–humid rift systems and highlights braided-river targets as promising exploration zones in climate-sensitive basins worldwide.

Understanding the heterogeneity of reservoirs is crucial for enhancing the efficiency of hydrocarbon exploration and development. The primary porosity of samples from modern braided river sands and outcrops of braided river sandstone was calculated using a model previously proposed by the authors. The characteristic parameters (Vx) for calculating primary porosity are closely related to the architectural–elemental configurations (AEC), and the AEC of braided river sand bodies (BRSD) has apparent effects on the distribution of the primary porosity heterogeneities. Analysis of our results has established a simple primary porosity heterogeneity model of BRSD. The center of braided river channel and mid-channel bars have excellent strong primary petrophysical properties with high primary porosity exceeding 38%. The contact areas between the braided river channel and channel bars exhibit relatively low primary porosities of less than 33%. The area between the center and edge of the braided bars and channels displays medium primary porosities. The nonlinear correlation in the Q–Q plot of the primary porosity and present porosity of samples from BRSD in the Ahe Formation is mainly caused by chemical diagenesis. The present porosity heterogeneity of BRSD in the Ahe Formation is less influenced by compaction and cementation, it predominantly arises from the differential of dissolution. Q–Q plots attempt to correlate the geological information from an individual sample with the heterogeneity of present porosity in BRSD. In addition, by utilizing Q–Q plots of the primary and current petrophysical properties of the sand body, the relative extent of heterogeneity modification caused by different diagenetic processes can be assessed. This assessment is crucial for modeling macroscopic models of physical properties during geological history periods.

Mechanical compaction is an important diagenetic process in sandstone reservoirs. Particle breakage, which commonly occurs during mechanical compaction, plays a significant role in controlling the physical properties of the reservoir. However, existing numerical simulation methods have limitations in simulating mechanical compaction when considering particle breakage. In this study, a discrete element simulation method of mechanical compaction was proposed based on particle cutting, and the experimental results reported in the literature were used to calibrate the simulation parameters. Finally, this method was applied to the simulation of the mechanical compaction of sandstone to analyze the pore evolution process. The results show that the new simulation method has high computational efficiency and can reflect the process of particle breakage. The simulation results coincide well with the experimental results. In the simulated mechanical compacted process of coarse sandstone, particle breakage is strong in the high-stress stage with a vertical stress of 30 MPa–50 MPa. The porosity and mean radii of pores and throats decreased rapidly, and the number of pores and throats increased rapidly in the high-stress stage. When the vertical stress reached 50 MPa, compared to the simulation results without considering particle breakage, the porosity difference rate caused by particle breakage was 4.63%; the radius difference rates of pores and throats were 2.78% and 6.8%, and the number difference rates of pores and throats were 4.95% and 8.74%, respectively. In the process of mechanical compaction, the pore evolution of the reservoir is controlled by the filling of the pre-existing pore space by the fragments generated through particle breakage and the generation of microfractures. Additionally, the simulation method presented in this study can be applied to complex geological conditions and can be combined with other reservoir simulation methods. The simulation results can provide rich training samples for artificial intelligence and other emerging technologies.