A commercial breakthrough in the in-situ conversion process (ICP) of lacustrine low-to-medium maturity shale oil requires that the energy output-to-input ratio (E_out/E_in) exceeds the economic threshold. The selection of the in-situ heating window represents one of the critical factors controlling E_out/E_in. Based on hydrocarbon generation simulations, hydrocarbon yield, and the evolution of porosity and permeability with increasing temperature for shale samples from the third submember of the seventh member of the Triassic Yanchang Formation (Chang 7₃ submember) in the Ordos Basin and the second member of the Cretaceous Nenjiang Formation (Nen 2 Member) in the Songliao Basin, China, this study demonstrates that the optimal control of the in-situ heating window can effectively reduce development costs. The results indicate that kerogen type and structure govern the main hydrocarbon generation window and the composition of “synthetic hydrocarbons”. The kerogen in the Chang 7₃ submember is abundant in short side-chain hydrocarbon groups, resulting in earlier hydrocarbon generation and lighter products. In contrast, kerogen in the Nen 2 Member is rich in long-chain hydrocarbon groups, leading to later hydrocarbon generation and heavier products, and it exhibits a conversion rate exceeding 80% and peak light hydrocarbon yield at both the in-situ heating temperatures of 346 °C and 363 °C. Clay minerals in shale function as natural catalysts, lowering the activation energy for kerogen cracking and enhancing the yields of gaseous and light hydrocarbons. During heating, light hydrocarbon mass yields from organic matter in montmorillonite-rich shale are 20%-50% higher than in montmorillonite-poor shale. A high clay mineral content effectively reduces the required heating window temperature. Shale porosity and permeability exhibit an inflection point corresponding to the main stage of “synthetic hydrocarbon” formation. During heating, porosity and permeability increase markedly, with much greater enhancement parallel to bedding than perpendicular to bedding. At 350 °C, porosity is increased to 7.3 times and permeability to 265 times. The optimal in-situ heating windows are less than or equal to 350 °C for the Chang 7₃ submember and 330-360 °C for the Nen 2 Member. Within these windows, synthetic hydrocarbon production, E_out/E_in, and product value are maximized, thereby significantly improving the economic viability of the resource.

This study combines sedimentological analogs with surface and subsurface geological analyses to investigate the Kalazha Formation along the southern margin of the Junggar Basin, China. Outcrops, modern sediments, cores, well logs, seismic data, production data, and laboratory test data constrain its formation, distribution, geologic age, and relationship with the Cretaceous Qingshuihe Formation. Four main findings emerge. First, the Kalazha Formation comprises sandy and gravelly distributive fluvial system (DFS) deposits of different scales along the piedmont. These systems include coarse-grained alluvial fans, fluvial fans, sandy fluvial mega-fans, and their associated deltas and fan deltas. Second, the Kalazha Formation is widely distributed as apron-like bodies along the southern basin margin, thinning northward and westward. Third, the Kalazha conglomerates and Qingshuihe sandstones and mudstones form a vertically superimposed, continuous depositional succession with a conformable contact near the piedmont outcrops. Basinward, the Kalazha Formation thins and pinches out, passing laterally into the fine-grained deposits of the Cretaceous Qingshuihe Formation and Tugulu Group. Fourth, the stratigraphic affiliation of the Kalazha Formation depends on the position of the Jurassic-Cretaceous boundary, which remains uncertain in this area. Further biostratigraphic and chronostratigraphic data are therefore needed. Drilling confirms that the Kalazha Formation has the basic conditions for high-quality reservoir development. Quantitative characterization of its depositional systems and reconstruction of lithofacies paleogeographic changes across the Jurassic-Cretaceous transition should provide a stronger basis for hydrocarbon exploration.

To systematically examine the paradigm shift of sedimentary paleogeography research driven by artificial intelligence, and clarify the technical integration pathways and core challenges, this study combines theoretical analysis and typical case analysis to summarize the multi-dimensional characteristics of the paradigm shift and explore a new framework for intelligent paleogeographic reconstruction. The results show that the paradigm of sedimentary paleogeography research is undergoing a systematic transformation across four dimensions: the data system shifts from experience-driven to data-driven, the cognitive mode from qualitative description to quantitative characterization, the prediction target from static representation to dynamic prediction, and the decision-making mode from single-discipline analysis to systematic decision-making. Accordingly, a comparative framework of traditional empirical, digital, and intelligent paradigms is constructed. In three key aspects, i.e. paleogeographic reconstruction, sedimentary process simulation, and reservoir parameter prediction, the methods such as knowledge graphs, physics-informed neural networks, and ensemble learning have demonstrated significant integration effectiveness. Furthermore, a technical workflow framework for intelligent paleogeographic reconstruction is proposed, encompassing core modules including multi-source data fusion, knowledge graph constraints, machine learning prediction, expert validation, and feedback iteration. This work provides a systematic theoretical reference and technical pathway for the digital-intelligence transformation of sedimentary paleogeography.

Based on seismic, drilling and outcrop data, fault geometry-kinematics analysis and balanced-section restoration were used to define the geological architecture and boundaries of the Awati Sag in Northern Depression, Tarim Basin, China, reconstruct its structural evolution sequence, and identify its main hydrocarbon accumulation controls and exploration direction. The Awati Sag presents a framework of wedge-shaped structure with larger thickness in the west than in the east, as a result of multi-stage differential reworking of the three boundary faults of South Kalayuergun, Shajingzi, and Aqia-Tumuxiuke. Internally, conjugate and strike-slip fault systems form a multi-stage, vertically differentiated, heterogeneous 3D deformation network with a composite structural style. Its structural evolution history includes a pre-sag stage and a sag stage. The former stage can be divided into five sub-stages: the Jinningian rifting center, the Early-Middle Caledonian intracratonic depression, the Middle-Late Caledonian post-uplift of the peripheral foreland basin, the Early-Middle Hercynian northwestern Tarim paleo-uplift, and the Late Hercynian rifting center, while the later stage into 3 sub-stages: the Indosinian embryonic sag, the Yanshanian low-amplitude uplift, and the Himalayan final shaping. Two sets of principal source rocks are derived from marine Cambrian Yuertusi and Ordovician Saergan formations, supplemented by the Ordovician Yingan Formation marine shale and Permian-Triassic coal measures and dark mudstones. Thermal evolution shows the characteristics of long-term low temperature and shallow burial followed by rapid deep burial temperature rise. Within the whole petroleum system, near-source fractured-vuggy/tight sandstone and far-source structural-lithologic oil and gas are expected to form a composite accumulation area, with in-source shale oil/gas as a potential exploration target, and the eastern slope showing the greatest exploration potential.

The concept of integrated source-to-sink system (IS2S) is proposed to integrate fine-grained (smaller than 62.5 μm) clastic sediments into the source-to-sink system, break the traditional theoretical framework of coarse-grain dominated sedimentary reservoir research, and unravel the complete sediment-routing process and multi-factor coupling from source area to sedimentary basin. The concept is based on multi-proxy provenance tracing and a comparison of marine and continental sedimentary records, and its key elements and the coupling relationship between coarse- and fine-grained clastic systems are discussed. The IS2S framework extends source-to-sink analysis from coarse-grained clasts to the full grain-size spectrum of terrigenous clastic sediments, places chemical weathering on an equal footing with physical denudation as a primary control, and unifies characterization of the dynamically coupled clastic and associated dissolved/colloidal fluxes within the same framework. Chemically and biogenically induced precipitates are explicitly treated as independent, parallel pathways outside the clastic grain-size spectrum research. The framework further recognizes that coarse-grained deposition is governed mainly by topographic gradients and tractive currents, whereas fine-grained deposition is dominated by suspended transport and flocculated settling. Although terrigenous coarse- and fine-grained clasts share a common provenance, they differ markedly in transport and depositional mechanisms. Compared with coarse-grained source-to-sink systems, fine-grained systems are still constrained by non-unique provenance proxies, complex bottom-current reworking, and marked differences between marine and continental settings. The IS2S is intended primarily to characterize depositional material composition and depositional-body architecture, and can provide sedimentological constraints on quantification of fine-grained sediment provenance contributions and modeling of the coexistence of conventional and unconventional hydrocarbon.

Focusing on Shunbeilong 2, a well with commercial oil/gas flow from the Ordovician Penglaiba Formation in the dome structure of Shunbei area in the Tarim Basin of China, the drilling, logging, testing and thin-section data were integrated for petrologic analysis, microscopic temperature measurement of fluid inclusions, C/O isotopic composition, trace element analysis, and U-Pb dating. The reservoir space, fluid properties/origin, and controlling factors of the Penglaiba Formation reservoirs were delineated to explore their genetic mechanisms. Finally, a reservoir development model was established. The results indicate that the dolomite reservoirs are distributed in the form of thin interbeds and quasi-layers, with the reservoir space consisting of fabric-selective vugs mainly, and superimposed by reticular fractures. Image logging data reveal a weak correlation between vugs and fracture intervals, demonstrating that vugs were formed primarily under the control of high-frequency sequence boundaries. The dolomite matrix is predominantly composed of medium-to-coarse crystalline dolomite, with minor fine crystalline dolomite and late calcite cements. The U-Pb age of medium-to-coarse crystalline dolomite is dated to (464±14) Ma; this dolomite shares inherited seawater-derived dolomitizing fluid signatures with fine crystalline dolomite and vug-filling dolomite cements. The homogenization temperatures of fluid inclusions hosted in secondary quartz and late calcite are concentrated at 150-160 °C and 130-140 °C, respectively, and the in-situ U-Pb age of late calcite yields (450±3.4) Ma. The hydrothermal fluid activity is related to the Late Ordovician tectonic-thermal event. It is proposed that the development model of dolomite reservoirs in Penglaiba Formation of dome structure is: selective dissolution controlled by high-frequency sequence boundaries during the penecontemporaneous period, dolomitization under shallow burial conditions, hydrothermal fluid dissolution of carbonate rocks along fault-fracture systems, and reconstruction of pre-existing vuggy layers by superimposed reticular fractures, ultimately forming fracture-vug composite dolomite reservoirs.

Focusing on the Neoproterozoic-Lower Paleozoic dolomite reservoirs in the Tarim Craton, this study constructed a comprehensive geological database encompassing 16 conventional oil and gas fields in dolomite reservoirs and 40 regional exploration wells drilled through dolomite reservoirs with no hydrocarbon discovery in the Tarim Basin, and global representative carbonate reservoirs. Based on this database, reservoir characterization and comparison were performed to reveal the coupling relationship among types, spatial distribution, and geometric configurations of large-scale dolomite reservoirs, and the seismic response characteristics were summarized. Results indicate that sedimentary facies, karstification, and tectonic fracturing are the primary controlling factors governing both dolomite reservoir spatial distribution and internal architecture. Accordingly, large-scale dolomite reservoirs are classified into three types and ten subtypes. Sedimentary-type reservoirs, developed in six types of sedimentary facies belts within three kinds of carbonate platforms, exhibit (quasi-)layered architectures, characterized by chaotic mound- or cone-shaped or reverse S-shaped seismic reflections. Karst-type reservoirs are subdivided into two subtypes: fault structural zones and regional uplifts, predominantly distributed within karst highlands and slopes, displaying a massive architecture with truncation seismic features. Tectonic-type reservoirs are subdivided into subsalt structures and strike-slip fault zones, also presenting a massive architecture, with flexural or bead-like seismic patterns. This paper summarizes the geological models for the ten subtypes of reservoirs above, providing both geological evidence and practical guidance for dolomite reservoir exploration.

The Sinian Dengying Formation in central Sichuan Basin is characterized by complex hydrocarbon accumulation processes, significant divergences in timing of paleo-oil reservoir formation, and a limited availability of absolute age data. In this study, petrographic observations, fluid inclusion analysis, carbonate U-Pb dating, and paleo-oil-water interface reconstruction were integrated to decipher the diagenetic-hydrocarbon accumulation evolution sequence of the Dengying Formation reservoirs in central Sichuan Basin and to precisely clarify the formation timing of paleo-oil reservoirs. The results reveal that the first-stage paleo-oil reservoirs with a limited scale were formed in the Late Cambrian-Early Ordovician, and the second-stage large-scale paleo-oil reservoirs were formed during the end of Middle Triassic, approximately 30 Ma earlier than the previously proposed end of Late Triassic. The paleo-structural high on the top of the Dengying Formation was more extensive during the Middle Triassic than in the end of Late Triassic, and it controlled a broader distribution of paleo-oil reservoirs. These paleo-oil reservoirs exhibited the features of widespread contiguous oil-bearing, quasi-continuous distribution, and relative enrichment in structural highs, providing liquid hydrocarbon for the subsequent development of large-scale oil-cracking gas reservoirs. The findings improve our understanding of natural gas exploration potential for the Sinian strata in the Sichuan Basin and suggest favorable gas exploration prospects for structural-lithologic traps within the Central Sichuan paleo-uplift and its peripheral areas.

Taking the shale of Cretaceous Qingshankou Formation in northern Songliao Basin, China, as an example, this study adopts experimental methods including X-ray diffraction analysis, field emission electron microscopy, atomic force microscopy, two-dimensional nuclear magnetic resonance, combined with the test of pressure-preserved sealed core ultra-low temperature frozen samples, to discuss the coupling relationship between the diagenetic evolution of organic-clay complex and the four properties of shale, i.e., reservoir property, oil-bearing property, mobility and fracability. (1) The oil content of organo-clay complex initially increases and then decreases with enhanced diagenesis, with the maximum oil content exceeding twice that of the corresponding bulk shale, indicating a significant contribution to shale oil production. The decrease in oil content of organo-clay complex during the middle diagenetic stage B suggests that a portion of the crude oil has micro-migrated into other pores in shale for storage. (2) The Brunauer-Emmett-Teller (BET) specific surface area and pore volume of organo-clay complex show a trend of first increasing and then decreasing with diagenetic evolution, reaching the maximum values of 65 m²/g and 0.15 cm³/g, respectively, at the end of the middle diagenetic stage A₁. From middle diagenetic stage A₂ to middle diagenetic stage B, the volume shrinkage of the organo-clay complex generates abundant organo-clay complex pores and bedding-parallel fractures, resulting in an increase in effective porosity of the shale from 5% to 10% or so. (3) The transformation of clay minerals within the organo-clay complex and the formation of authigenic minerals such as quartz result in improved shale fracability. In the middle diagenetic stage B, the shale exhibits an elastic modulus greater than 20 GPa, a Poisson's ratio less than 0.2, and a brittleness index greater than 65%. (4) Both the organo-clay complex and the bulk shale show increasing free oil content with progressive diagenesis, and improved shale oil mobility. During the middle diagenetic stages A₂ and B, dehydration of clay minerals results in an increase in pore water content of the shale, reaching a maximum of 11.8 mg/g, which suggests alterations in the wettability of shale oil flow pathways that are favorable for water-oil imbibition and displacement during hydraulic fracturing. Integrating the shale oil enrichment model and the coupling relationships among the parameters of four properties, three types of shale oil sweet spot zones (intervals) are categorized. The findings of this study hold significant theoretical implications for the exploration evaluation and development practice of continental high-clay-content shale oil.

Based on core, well-logging, well-testing and seismic data, the reservoir characteristics and controlling factors of favorable reservoirs in a semi-restricted environment of carbonate ramp were investigated through a case study of the Lower Cretaceous Yamama Formation in Oilfield A, Central Arabian Basin. During the Early Cretaceous, the study area was a carbonate ramp in a semi-restricted environment. The Yamama Formation is dominated by wackestone, packstone, and bindstone, with poor lateral correlatability. Overall, low- to medium-energy shallow-water sedimentary facies predominate, characterized by extensive lagoonal deposits and locally developed shoal-front, grain shoals, patch reefs, back-shoals and tidal flat deposits. The sediments are particularly rich in bioclasts, including algae, benthic foraminifera, bivalves, bacinella and peloids. The Yamama Formation is dominated by porous reservoirs. Reservoir spaces mainly consist of micropores, moldic pores and intrafossil pores, whereas primary intergranular pores are poorly developed. Pore throats are dominated by medium and micro throats. The reservoirs generally exhibit low to moderate porosity and extra-low to ultra-low permeability. The medium- to high-permeability reservoirs are relatively poorly developed. Large-scale high-quality reservoirs with good physical properties developed where dissolution was relatively intense. High-quality reservoirs in the Yamama Formation are primarily controlled by four factors: local high-energy deposition, enrichment of soluble bioclasts, intense dissolution and abnormally high formation pressure. In local high-energy grain shoals, well-preserved primary intergranular pores contribute to good reservoir properties, forming small-scale favorable reservoirs. In contrast, low- to medium-energy facies, such as lagoons and back-shoals, are enriched in soluble bioclasts such as algae and bacinella. Intense penecontemporaneous dissolution of these bioclasts generates abundant moldic and intrafossil pores, effectively improving reservoir properties and developing relatively large-scale high-quality reservoirs. Formation overpressure compartment effectively inhibits the effects of compaction and cementation, allowing the secondary pores in high-quality reservoirs to be consistently preserved. Favorable reservoirs in the Yamama Formation are mainly developed in the YA and YB intervals, and large-scale reservoirs are concentrated in the central and northern parts of the study area, representing key targets for future exploration and development.

Taking the Permian clastic buried-hill reservoirs in the Jiyang Depression, Bohai Bay Basin, China, as the research object, this study investigates the reservoir characteristics and diagenetic fluid types and assemblages using core samples, thin sections and laboratory test data. The reservoir controlling factors are analyzed, and a reservoir-controlling model is established. Four types of diagenetic fluids are developed in the clastic buried-hill reservoirs, namely alkaline water associated with tuffaceous alteration, meteoric freshwater, hydrothermal fluid, and organic acid. Diagenetic fluid assemblages vary distinctly among different buried-hill reservoirs. The Gubei buried-hill is characterized by the assemblage of meteoric freshwater + volcanic hydrothermal fluid + organic acid. The Gaoqing buried-hill develops the assemblage of alkaline water associated with tuffaceous alteration + meteoric freshwater (relatively weak) + organic acid, and the Yihezhuang-Dawangzhuang buried-hills develops the assemblage of alkaline water associated with tuffaceous alteration + meteoric freshwater + organic acid. Among these fluids, meteoric freshwater and organic acid exert constructive modification on reservoir quality. Reservoirs are synergistically controlled by multiple factors such as fault-related fractures, weathering crust, soluble components and diagenetic fluids. The established “three-element and six-factor” reservoir-controlling model yields predictions that display a significant positive correlation with measured reservoir physical properties. Sand bodies with coarse grains, large thickness and abundant tuffaceous components formed during the sedimentary period provide the material basis for high-quality reservoir formation. Meteoricfreshwater dissolves soluble components within reservoirs during the uplift stage, whereas organic acid improves reservoir quality via further dissolution during the burial stage. The proposed model supplies new geological theoretical support for the prediction of clastic buried-hill reservoirs.

To reveal the apparent hydrogen diffusion characteristics of typical rocks, their controlling factors, and the sealing performance of different caprock, a two-chamber gas diffusion apparatus was used to determine the apparent diffusion coefficient (D_app) of 25 sedimentary, igneous, and metamorphic rock samples under isobaric conditions at 0.1 MPa. Representative samples were tested for sensitivity to temperature, effective stress, water content, and counter-diffusing gas composition, followed by one-dimensional diffusion simulations using the measured parameters. The experimental results show that: (1) At an effective stress of 1 MPa, D_app ranges from 2.54×10^-9 m²/s to 358.43×10^-9 m²/s, with most values between 10×10^-9 m²/s and 100×10^-9 m²/s, and shows a strong positive correlation with porosity. Hydrogen diffusion is relatively high in sandstone, coals, and some high-porosity shales and carbonate rocks. The tested basalt, granite, and dolerite samples show low to moderate diffusion, whereas the tested serpentinite and salt rock samples exhibit lower diffusion. (2) Increasing temperature from 278.15 K to 318.15 K increases D_app by 11.9%-34.4%. Increasing effective stress causes an exponential decrease in D_app, with stress sensitivity decreasing in the order of salt rock, serpentinite, dolerite, shale, sandstone and limestone. Increasing water content markedly suppresses hydrogen diffusion. (3) Compared with N₂, CH₄ induces only minor changes in D_app, whereas CO₂ reduces D_app by an average of 8.6%. (4) Increasing caprock thickness and water content reduces hydrogen diffusion and loss. Among the selected caprock samples, dry salt rock shows the strongest sealing performance, while shale performs no worse than dolerite at comparable water contents. Overall, the apparent hydrogen diffusion coefficient is jointly controlled by porosity, pore-throat structure, connectivity, temperature, effective stress, and water content. Caprock sealing evaluation should therefore consider lithology, thickness, effective connectivity, and in-situ temperature, stress and water conditions.

To address production decline and safety concerns associated with wellbore liquid loading during natural gas production, this study proposes a method integrating physical mechanisms and intelligent algorithms for liquid loading prediction and early warning of flow cessation caused by water accumulation in the wellbore. Constrained by liquid-carrying mechanisms and driven by time-series production data, the method uses mechanistic modeling and numerical simulation to augment the training data. A dual-channel liquid loading prediction model and a flow cessation early-warning model integrating multiple strategies are developed. Physical constraints are incorporated to correct the model predictions, enabling progressive assessment from liquid loading detection to early warning of flow cessation. Applications on 56 test wells in a tight gas field in China yield a liquid loading detection accuracy of 91.6%. In the long-term and intermittent liquid loading case studies, liquid loading is identified 5 d and 3 d before the annotated onset dates, respectively. In a representative case, a warning is issued 7 d before flow cessation. Early warnings are successfully issued 49 times for flow cessation events (87.5%), with a mean lead time of 6.5 d for the successfully warned events. These results are conducive to optimizing maintenance decisions, reducing the risk of unplanned production cessation, and extending the production cycle of gas wells. The proposed models are promising for engineering application.

To address the limitation of conventional macroscopic wettability measurements in characterizing pore-scale wettability distributions in shale, a quantitative method for microscopic wettability characterization was developed. The method improves the conventional Amott method by integrating nuclear magnetic resonance (NMR) and high-pressure mercury intrusion porosimetry. Fluid distributions during spontaneous and pressure-assisted imbibition in oil-water and oil-fracturing-fluid systems were measured, and a model for calculating the pore-size-dependent microscopic wettability index was established. The average microscopic wettability index derived from the proposed method agrees well with those obtained from gravimetric and macroscopic NMR measurements, validating the reliability of the method and revealing mixed-wettability characteristics that may be obscured by macroscopic averaging. The method was applied to interbedded, laminated, and mud-laminated shale cores from Well H181-1 in the Chang 7 Member of the Triassic Yanchang Formation, Ordos Basin. The results show that the original macroscopic wettability of the shale is generally oil-wet, whereas the overall water-wetness of all three core types increases markedly after fracturing-fluid interaction. The microscopic wettability responses vary significantly among the three shale types. In the interbedded shale, local oil-wetness increases in small pores, whereas the medium and large pores generally become more water-wet. In the laminated shale, water-wetness increases markedly in small pores, the medium pores exhibit heterogeneous responses, and water-wetness decreases in large pores. In the mud-laminated shale, the small and medium pores shift toward more water-wet conditions, whereas the large pores exhibit heterogeneous wettability changes. These results clarify the pore-scale wettability distribution characteristics of shale reservoirs and provide a theoretical basis for fracturing-fluid selection, stimulation-parameter optimization and enhanced oil recovery.

Application of CO₂ pre-fracturing to displace the shale oil of the Permian Fengcheng Formation in the Mahu Sag of Junggar Basin, China, suffers from gas channeling, asphaltene precipitation, emulsification and viscosity increase of crude oil under fracturing impact, and low utilization rate of CO₂. To address these problems, a high-temperature and high-pressure visual silicon tube experimental system was used, together with Raman-infrared microscope and stereoscope, to study the interaction between CO₂ and crude oil in fractures and microfissures, and to reveal the influence mechanism of solubilizer on the dissolution characteristics, foam oil behavior and flow stability of CO₂-crude oil system. During displacement at higher pressure, the solubilizer can greatly reduce the start-up pressure and oil-water interfacial tension of CO₂-crude oil system, enhance the dissolution and retention capacity of CO₂ in crude oil, form a stable and fine foam oil structure, inhibit bubble aggregation and delay gas release, thereby delaying the process of gas channeling, and improving the sweep and oil displacement efficiency of CO₂. The effect is better when the mass fraction of solubilizer is 1.0%. During production at lower pressure, the solubilizer can inhibit the precipitation of resin and asphaltene through peptization, thus reducing the risk of plugging near wellbore, and improving the mobility of crude oil in hydraulic fractures. The field data confirm that the solubilizer performs well in increasing production, and it provides a technical option for the efficient development of shale oil in the Mahu Sag.

The current status of volume fracturing technology is reviewed for continental shale oil reservoirs in China. Considering the limited hydraulic fracture propagation and proppant transport under the prevailing “long-fracture” volume fracturing model in horizontal wells, a new “wide-short fracture” volume fracturing model is proposed to fully stimulate the reservoir in the near-wellbore zone, in accordance with four principles: multi-cluster fracture placement, balanced fracture propagation, appropriate treatment scale, and economics-based parameter selection. Beyond the energy allocation pattern in current fracturing practices, the new model is designed logically to concentrate fracturing energy in the near-wellbore zone. This redistribution enables balanced propagation of multiple fractures and full-domain propping, thereby enhancing the connectivity between the fracture system and the oil-bearing matrix, expanding the fracture-controlled reservoir volume, and achieving coordinated reservoir control by wells and fractures. Five core elements underpin the proposed fracturing design: refined modeling technique of the in-situ stress field, techniques for controlling balanced fracture propagation, high-density perforating with uniform entry-hole diameters, techniques for full-domain propping of complex fracture systems, and a framework for jointly optimizing well patterns and well spacing. Comparisons of numerical model predictions of different fracturing models indicate that the proposed model can substantially increase both total ultimate recoverable reserves and the ultimate oil recovery factor at the well-pad scale in continental shale oil reservoirs. These findings can provide a theoretical basis and engineering support for the large-scale, cost-effective development of continental shale oil reservoirs in China.

Targeting the continental tight sandstone reservoirs of the Chang 8 Member of the Triassic Yanchang Formation in the Ordos Basin, China, a hydraulic fracturing test site was constructed in the Xifeng Oilfield. By deploying two differentially fractured wells and one coring well that served both experimental testing and production, multi-source technologies including core observation, microseismic monitoring, imaging logging, and ceramic proppant tracing were integrated to investigate directional fracture propagation patterns, fracture parameters, and proppant placement characteristics. The differentiated characteristics and identification methods are clarified for different fracture types. A total of 17 hydraulic fractures are identified in the 223.6 m-long core, revealing that the hydraulic fractures propagation is generally dominated by in-situ stress, with minor diversion locally, and the main fractures propagate in the direction of NE 73°. The hydraulic fractures predominantly extend as single fractures, with local inter-well coupling, and exhibit a non-uniform clustered distribution. The supported effective of the main fractures is approximately 180 m. A low-cost proppant design combining particle size grading and color coding is adopted, showing a non-uniform unimodal concentrated distribution along the fractures, with mixed particle sizes dominating the near-wellbore zone and small-sized particles prevailing at the distal end. The tight oil hydraulic fracturing test site in the Xifeng Oilfield has established a distinctive testing mode for precise fracturing of continental tight oil reservoirs, providing a practical basis for the cost-effective development of tight oil and a technical support for the stimulation of tight sandstone reservoirs similar to the Chang 8 Member reservoirs in the Xifeng Oilfield.

Visualization experiments of axial-torsional coupled percussion (ATCP) drilling in hot dry rock (HDR) were conducted using digital image correlation technique and multiscale characterization method. The evolution mechanism of high-temperature granite under ATCP mode was revealed, spanning from dynamic strain response and mesoscopic crack propagation to macroscopic rate of penetration (ROP) improvement. The results indicate that ATCP induces a coupled tensile-compressive strain response in granite, which develops a fracture network dominated by a transgranular-intergranular composite mode, thereby facilitating a transition in the rock-breaking pattern from high-energy crushing and grinding to low-energy volume spalling. The ATCP mode promotes alternation of rock principal strain between the positive and negative values by enhancing the coupled tensile-compressive strain response to increase the ROP. This reduces the weight on bit and its fluctuation during drilling, ensuring a stable rock-breaking while improving the ROP and lowering the mechanical specific energy (MSE). Field test results from the geothermal well Fushen-1 showed that the ATCP mode significantly increases ROP and reduces MSE in HDR formations, demonstrating its great applicability for improving HDR drilling performance.

Natural hydrogen does not need to be prepared artificially and features zero carbon emissions over its entire lifecycle, making it a potential resource for filling the gap in hydrogen supply. However, the dynamic nature of natural hydrogen accumulation systems and the coupling relationships among accumulation elements have not been systematically investigated. Based on the systems approach, this paper proposes a dynamic natural hydrogen accumulation system with the coupling of five elements (generation, migration, accumulation, sealing and dissipation), and reviews the generation pathways, occurrence states, migration mechanisms, dissipation processes, and preservation conditions of natural hydrogen. Natural hydrogen is mainly generated by water-rock interactions, water radiolysis and deep fluid activity, with the characteristics of continuity, multi-source superimposition and flux control. Natural hydrogen occurs in free, dissolved, and adsorbed states in the subsurface. Its preservation is constrained by physical diffusion, chemical reaction and microbial action. Its accumulation depends on whether hydrogen input can compensate for losses caused by migration and dissipation over a geological timescale. According to the coupling relationships among five elements under different geological backgrounds, three representative natural hydrogen accumulation models are identified: ophiolite-hosted self-generation and self-storage, cratonic-basement tight sealing, and fault-controlled sedimentary-basin accumulation. Natural hydrogen assessment should shift from static reserve evaluation to integrated constraints on hydrogen generation flux, migration efficiency, reservoir capacity, caprock sealing, and dissipation intensity. The proposed framework of dynamic accumulation provides additional systematic insights for research on geological theory and exploration of natural hydrogen.

Under the dual constraints of carbon neutrality and energy security, this study systematically examines the background, advantages, pathway and strategic significance of integrated development of oil and gas industry with new energy, based on resource endowment in China characterized by abundant coal, limited oil and gas, and vast renewable energy potential, together with its relatively high dependence on imported oil and gas. The principle of “inexhaustible energy” is revealed according to three global energy development laws: succession, decarbonization and inexhaustibility. The oil and gas industry possesses inherent advantages such as integrated source-grid-load-storage-intelligence systems, on-site consumption of renewable resources in oil and gas fields, flexible peak-shaving through natural gas-fired power generation, and the evolution of oil and gas fields toward multi-energy systems encompassing power generation, geothermal energy, hydrogen production and carbon management. These advantages provide a solid foundation for deep integrated development between fossil fuels and renewable energy. In terms of implementation, a comprehensive integration framework covering upstream and downstream sectors is established and advanced through three successive stages: the clean substitution and demonstration breakthrough stage, the strategic replacement and large-scale integration stage, and the green transition and carbon neutrality realization stage. Strategically, the integrated development can enhance national energy self-sufficiency and system resilience, accelerate deep decarbonization of the oil and gas industry and its transition toward integrated energy services, elucidate a Chinese solution for the smart integration of oil, gas and new energy, and strengthen the influence of China in global green energy governance. The integrated development of oil, gas, and new energy represents an indispensable pathway toward safeguarding national energy security, fulfilling the dual-carbon goals, and building an energy powerhouse of China.

To address the requirement for comprehensive analysis of multi-source geological information and intelligent decision-making in shale oil exploration and development, we propose a method for constructing a domain-specific large language model (LLM) for the shale oil, along with a framework for building and evaluating an LLM-based intelligent agent for shale oil seismic processing and interpretation. Using Qwen2.5-72B and Qwen3-32B as foundation models, we construct a specialized corpus of 21 billion tokens by integrating public shale oil literature and domain knowledge graphs. Leveraging expert annotation and model generation, we also construct 5.8×10⁴ question-answer pairs in specific domain with high quality to serve as supervised fine-tuning dataset. We build a multi-stage efficient domain adaptation training framework for the shale oil large language model. First, domain knowledge is injected via task-aware continuous pre-training method, and supervised fine-tuning is executed using Quantized Low-Rank Adaptation (QLoRA) combined with context-parallel algorithms. Then, direct preference optimization (DPO) and group relative strategy optimization (GRPO) are employed to align domain-specific preferences and enhance reasoning capabilities. In addition, a reflective retrieval-augmented generation (RAG) framework is built to optimize the accuracy and interpretability of model output. To enhance the tool-use capabilities of LLM and effectively mitigate hallucinations during tool-use process, we construct a specialized toolset by selecting open-source geoscience tools and employ targeted fine-tuning focused on tool utilization. Building on the foundation, we develop an intelligent scheduling service leveraging the Model Context Protocol (MCP) to deeply integrate core functions, such as seismic data processing, seismic interpretation, visualization, and report generation, thereby creating an LLM-driven agent for seismic processing and interpretation of shale oil. Experimental results demonstrate that the shale-oil domain-specific LLM achieves significantly higher accuracy in specialized Q&A compared to general-purpose foundation models, exhibiting excellent domain adaptability. Furthermore, following supervised fine-tuning using geoscience tools, the tool-use capability of shale-oil domain-specific LLM has been improved significantly compared with open-source foundation models.