Feldspar dissolution and precipitation of clays and quartz cements are important diagenetic reactions affecting reservoir quality evolution in sandstones with detrital feldspars. We examined two sets of sandstone reservoirs to determine whether the sandstone diagenetic systems were open or closed to the mass transfer of products from feldspar dissolution and its impact on reservoir quality. One of the reservoirs is the Eocene fan delta sandstone buried 2.5–4.0 km (1.5–2.5 mi) below sea level (BSL) in the Gaoliu (GL) area of the Nanpu sag, and the other is the Eocene subaqueous fan sandstone buried 1.5–4.5 km (1–2.8 mi) BSL in the Shengtuo (ST) area of the Dongying sag. Both sandstones consist mainly of lithic arkoses and feldspathic litharenites, and have secondary porosity formed by dissolution of feldspars. In the GL sandstones, the absolute amounts of authigenic clays and quartz cements (generally 125°C [257°F]). The low abundance of authigenic clays and quartz cements, and low pore-water salinity indicate that much of the K+, Al3+, and SiO2(aq) released from leached K-feldspars were exported from the GL sandstone system. And the extensive feldspar dissolution enhanced much porosity and permeability. In contrast, the ST sandstones with secondary pores formed by feldspar dissolution generally contain authigenic clays (kaolinite and illite) and quartz cements with almost identical volume of secondary pores. Kaolinite dominates in the ST sandstones at shallower depth (3.1 km [2 mi] BSL) where temperature exceeds 125°C (257°F). The presence of abundant clays and quartz cements indicates that Al3+ and SiO2(aq) released from leached feldspars were retained in the ST sandstone system. The dominance of authigenic illite at greater depth indicates that sufficient K+ should have been retained within the sandstones for occurrence of illitization of kaolinite and feldspars. Secondary porosity in thin sections can be up to 3%, but little porosity (<0.25%) is enhanced. Primary macropores are lost as clays and quartz precipitate whereas the proportion of microporosity increases, occurring mainly between clay crystals. The overall result is that permeability is degraded. The diagenetic difference between the GL and the ST sandstones can be interpreted by assessing pore-water evolution in these two areas. The current pore waters with low salinity and negative hydrogen isotopic compositions in the GL sandstone system indicate the significant impact of meteoric water, whereas the current pore waters with high salinity and the paleofluids with positive oxygen isotopic compositions in the ST sandstone system indicate little trace of meteoric water. Access of meteoric freshwater to the GL area probably occurred during the late Oligocene to Neogene through widely developed faults in the Paleogene and Neogene strata. The low-salinity water could have been responsible for flushing of solutes derived from feldspar dissolution. As such, diagenesis in the GL sandstones is considered to have occurred in an open geochemical system, whereas with limited faults and high water salinity, the ST sandstones acted as a closed geochemical system where precipitation of kaolinite, illite, and quartz cements occurred following dissolution of feldspars.

The purpose of this work was to study the depositional mechanisms and significance of the Longmaxi shale in the Sichuan Basin in southern China. Seven lithofacies were identified based on the detailed observation of outcrops and cores using petrographic and scanning electron microscope examination of thin sections and other data analyses: (1) laminated calcareous mudstone, (2) laminated carbonaceous mudstone, (3) laminated silty mudstone, (4) laminated claystone, (5) laminated siliceous shale, (6) siltstone, and (7) massive mudstone. The laminated mudstone and laminated claystone originated from suspension deposition, and siliceous shale is associated with ocean upwelling, whereas massive mudstone and siltstone were primarily deposited by turbidity currents. The depositional mechanisms have a great effect on the source rock and reservoir properties. Suspension deposition near oceanic upwelling zones can provide favorable conditions for the production and preservation of organic matter and are thus conducive to the formation of high-quality source rocks (total organic carbon content up to 5.4%). The reservoir storage spaces are primarily interlaminated fractures and organic pores with good physical reservoir properties (high porosity, permeability, and brittle mineral content). Turbidity currents may carry a large quantity of oxygen to the seafloor, resulting in the oxidation of organic matter, which is unfavorable for its preservation. The lithofacies formed by turbidity currents have relatively low total organic carbon contents (average: <1%). Structural fractures and intergranular pores are the primary storage spaces that are present in the reservoir. In summary, organic-rich shale and siliceous shale that was deposited from suspension near upwelling zones are key exploration targets for shale oil and gas. The widely distributed, multilayer, tight sandstone is important in the exploration for tight oil. A better understanding of the deposition mechanism and its effect on oil reservoirs may assist in identification of favorable areas for exploration.

Recent years have witnessed the development of the applications of machine learning technologies to well logging-based lithology identification. Most of the existing work assumes that the well loggings gathered from different wells share the same probability distribution; however, the variations in sedimentary environment and well-logging technique might cause the data drift problem; i.e., data of different wells have different probability distributions. Therefore, the model trained on old wells does not perform well in predicting the lithologies in newly-coming wells, which motivates us to propose a transfer learning method named the data drift joint adaptation extreme learning machine (DDJA-ELM) to increase the accuracy of the old model applying to new wells. In such a method, three key points, i.e., the project mean maximum mean discrepancy, joint distribution domain adaptation, and manifold regularization, are incorporated into extreme learning machine. As found experimentally in multiple wells in Jiyang Depression, Bohai Bay Basin, DDJA-ELM could significantly increase the accuracy of an old model when identifying the lithologies in new wells.