Massive liquid injection into tight sandstone enhances flow channels and energy recovery. However, the coupling among fluid injection, microscopic pore-structure response, and flow evolution remains insufficiently understood. Furthermore, the effects of key controlling factors are still unclear. Here we employ online nuclear magnetic resonance (NMR) to quantitatively track microfracture response and flow behavior during this process. Stress analysis and grey relational analysis (GRA) are used to interpret the mechanical mechanism and summarize the relative controls of key influencing factors on flow capacity. The results indicate that microfractures open in tension because pore pressure reduces effective normal stress on fracture surfaces and concentrates stress near weak zones, with the opening direction perpendicular to effective minimum principal stress. Relative fracture toughness decreases with pore pressure, and the difference between matrix and fracture rocks narrows at high pore pressure. Matrix rocks respond more strongly because pre-existing fractures guide fluid preferentially and delay pressure sweep into the matrix, resulting in porosity and permeability increase of 9.69–17.22

A flow-regime-based analytical approach to rate transient analysis (RTA) based on single fracture in a single producer is extended to a multi-fractured horizontal well (MFHW) producer. This method can be used for an existing multi-fractured horizontal producer for production performance diagnosis and forecast. The procedures in developing the RTA analysis approach include: identifying bilinear and formation linear flow regimes, determining the slope values and the switch time between those two flow regimes based on historical production data; summarizing the slope relationship between those two flow regimes from multiple MFHWs in the same reservoir. There are mainly two scenarios to forecast production performance: (1) For MFHWs with both bilinear and formation linear flow regimes appeared; their production performance can be forecast by extrapolating the formation linear flow regime for a rationally longer period before entering other radial flow regime or boundary dominated flow (BDF). In the meantime, a slope value relationship between those two flow regimes can be summarized from field production data; (2) For MFHWs with only bilinear flow regime during the available production history: the RTA analysis approach can be applied to predict the formation linear flow regime for the slope value and switch time based on the relationship summarized; then to forecast the remaining bilinear flow followed by formation linear flow for a rationally longer period. A model study clarifies the numerical sensitivity of timestep in RTA analysis approach and the procedure in constructing the RTA model forecast. A field case study demonstrates the successful production performance forecast of single MFHW or an entire reservoir block with multiple MFHWs, as well as to apply the forecast as signals for diagnosing and optimizing production performance to single MFHW in field surveillance study.

CO2 injection into shale integrates resource exploration with carbon sequestration. Prior investigations have relied predominantly on crushed samples or CO2-brine mixtures to study long-term geochemical interactions during soaking. Consequently, the essential mechanisms governing the evolution of flow capacity and microstructure across different injection methods remain poorly understood under realistic formation conditions. This study employs nuclear magnetic resonance (NMR) and multiple characterization techniques to investigate the microscopic mechanisms of flow evolution and sequestration efficacy across different CO2 injection methods under actual formation conditions. Quantitative criteria are established to evaluate the contributions of distinct processes, and the influences of key factors and their interactions are systematically analyzed. Furthermore, a prediction model with multifactor coupling is constructed based on correlation analysis. The results show that CO2 soaking under thermodynamic equilibrium transitions from physical to chemical control as pressure increases. At low pressure, limited particle migration rises blockage risks in micropores in zones above 0.06 mD. Elevated pressure enhances elastic energy, improving flow capacity by 100-200 % in zones below 0.06 mD despite increased macropore blockage risks. Conversely, CO2 flooding remains physically controlled by pressure gradients. Low pressure causes sharp flow decline due to particle blockage in narrow pore-throats, while increased pressure promotes high-speed CO2 flow, reducing flow capacity by 7.79 %-50.48 % but increasing CO2 gas column height by 1.67 %-18.32 %. For engineering practice, CO2 flooding should be avoided in highpermeability, low-porosity formations under low pressure. Instead, during the middle-to-late stages, massive flooding in ankerite-rich zones is recommended to couple capillary and mineralization trapping.

Fracture interference strongly controls cross-scale CO2 mass transfer and flow behavior between matrix and fractures during huff-n-puff, while residual water further complicates these processes. However, the governing mechanisms remain unclear. Here we construct artificial fractures with finite lengths in core samples to simulate distinct flow regions near fractures. Then, we investigate how fracture interference and residual water control the pore-scale mass transfer through online nuclear magnetic resonance (NMR). A recognition method is established to identify the distinct flow behaviors and quantify the effects of fracture length, fracture number, and their coupling. The results indicate that fracture interference with short length intensifies distal liquid redistribution and micropore backflow, but this response cannot be efficiently converted into recovery because of the rapid decay of pressure gradient and the long return distance. Recovery therefore mainly depends on the expansion and carrying capacity associated with CO2 phase transition, with a final recovery of 46.72%. Long fractures advance fluid redistribution and improve its conversion during single-phase flow due to the shorter return distance, giving a final recovery of 65.16%. In later cycles, however, bubble-assisted carrying remains confined near the fractures because gas propagation lags behind pressure propagation in the distal matrix. Residual water weakens fracture effectiveness and redirects CO2 through locally less water-blocked pathways. As cumulative water film migration strengthens the barrier near fractures in later cycles, the additional recovery is only 6.49%, 16.63% lower than water-free case. These findings provide theoretical support for fracturing design and CO2 injection optimization.

Horizontal heterogeneity is a key geological factor that governs fluid migration and interphase mass transfer during CO2 huff-n-puff in shale. Existing studies mainly focus on macroscopic vertical heterogeneity in layered geological systems, whereas microscopic horizontal heterogeneity in shale developed by volume-fractured horizontal wells remains poorly understood, despite its potential to amplify pore-scale disturbances into largescale multiphase flow behavior. Here we propose a novel method for quantitatively characterizing microscopic horizontal heterogeneity based on spatial T2 spectra and geometric mean through online nuclear magnetic resonance (NMR), emphasizing coupled pore-throat size and connectivity. Dynamic evaluation indicators are established to analyze how microscopic horizontal heterogeneity and spatial location control two-phase mass transfer during CO2 miscible and immiscible huff-n-puff. The results indicate that in miscible process, CO2-oil mass transfer in low-permeable layers shows strong distance dependence. Full miscibility develops near the injection end through multiple contacts during early phase transition, reducing oil viscosity by 45.87%. In midpermeable layers, distal miscibility is controlled by enriched gas during soaking, while proximal miscibility occurs mainly during production. High-permeable layers further amplify this near-distal disparity. Weak internal heterogeneity in immiscible process may contribute to short-term advantages of selective extraction during initial soaking. In later cycles, light component depletion and trapped CO2 bubbles aggravate interlayer conflicts caused by ultra-high-permeable layers. Compared with immiscible processes, miscible huff-n-puff more effectively reduces this later interaction heterogeneity and reactivates ineffective flow pores in high-permeable layers. These findings provide theoretical support for CO2 injection optimization and CO2 retention assessment.

Hydraulic fracturing creates artificial fracture networks that provide primary flow pathways for oil production from shale oil reservoirs. However, crude oil trapped in the ultra-tight matrix pores remains difficult to mobilize, resulting in limited recovery enhancement. To improve oil displacement from the matrix, a CO2 micro–nano bubble (MNB) structure was prepared in fracturing fluid. Under the high pumping pressure of hydraulic fracturing, these MNBs can be forced into matrix pores and promote oil displacement into fracture channels through imbibition, thereby enhancing oil recovery.① The micro–nano bubbles prepared with CO2 and N2 in the fracturing fluid exhibited median bubble sizes (D50) of 215.6–311.9 nm and concentrations of 25.7–33.3 × 10ˆ8 bubbles/mL. Compared with those prepared directly in clean water, the bubbles generated in the fracturing fluid had smaller particle sizes and higher concentrations.② A PVT apparatus was used to simulate six temperature–pressure conditions, with temperatures of 30 and 70 °C and pressures of 5, 15, and 35 MPa. After static aging of the CO2 micro–nano bubble fracturing fluid for 6, 12, and 24 h, the D50 values remained below 400 nm and the bubble concentrations remained above 24 × 10ˆ8 particles/mL, indicating good temperature and pressure resistance. As the standing time increased, the average bubble size increased and the concentration decreased. The bubble size expanded with increasing temperature and was compressed with increasing pressure, while the bubble concentration continuously decreased with increasing temperature and pressure.③ Drag reduction tests conducted at flow rates of 10–30 L/min showed that, relative to clean water, the drag reduction rates of the CO2 and N2 micro–nano bubble fracturing fluids were 14.99%–69.18% and 15.75%–67.62%, respectively. Compared with conventional fracturing fluid, the drag reduction rate increased by 0.54%–6.60% for the CO2 system and by 1.12%–4.96% for the N2 system, demonstrating that micro–nano bubbles have a drag reduction effect. The calculated contributions of bubble concentration and particle size to the drag reduction rate were approximately 70% and 30%, respectively, with bubble concentration playing the dominant role.④ Using nuclear magnetic resonance (NMR) and imbibition displacement simulation, the imbibition displacement performance of CO2 nano-bubble and CO2 micro-bubble fracturing fluids in natural shale cores was compared. During the first 120 h of continuous imbibition, the CO2 micro-bubble fracturing fluid showed a higher imbibition displacement rate. After 336 h of continuous imbibition, the CO2 nano-bubble fracturing fluid gradually gained an advantage, which was maintained during further imbibition. The results demonstrate the preferential imbibition behavior of CO2 nano-bubble fracturing fluid for mobilizing crude oil in small-scale pores of shale reservoirs.The results demonstrate that CO2 micro-nano bubbles can be feasibly generated in fracturing fluid, that the bubble structure can withstand the temperature and pressure environment of shale reservoirs, and that the system provides drag-reduction benefits. After entering the reservoir, CO2 micro-nano bubbles are expected to improve the efficiency of imbibition-driven oil displacement in small matrix pores and thereby assist enhanced oil production.

A novel CO2-responsive hydrogel for intelligent control of gas channeling in CO2- enhanced oil recovery (CO2EOR) and geological CO2 storage has been developed. A monomeric long-chain tertiary amine surfactant (HXB-2) that has specific amide and carboxyl groups was synthesized. The surfactant can interact with CO2 in aqueous solution to increase the viscosity and induce gelation. The hydrogel is irreversible and does not revert to solution phase after N2 bubbling. It shows excellent structural stability and thermal resistance and the viscosity remains four times higher than that of the initial solution upon heating. For the mechanism, HXB-2 protonates in CO2 environment and self-assembles into worm-like micelles (WLMs) under synergistic forces of hydrophobic interaction, hydrogen bonding, and electrostatic interaction, which further crosslink to form a three-dimensional (3D) network to induce gelation. The hydrogel can be formed in-situ to control gas channeling intelligently and redirect the gas to unswept low-permeability channels. It can enhance the recovery rate by 23.53 % and the maximum seepage resistance reaches 29.45 MPa & sdot;min & sdot;cm- 3 for water-alternatinggas flooding. Moreover, by having spontaneous association and shear-dissociation properties, the hydrogel in the rock pores causes minimal damage to the reservoir. This study provides valuable insights and empirical support for the development of irreversible CO2-responsive hydrogels for CO2 chemical sequestration and gas channeling control to help EOR and geological CO2 storage.

Addressing gas channeling, limited sweep efficiency, and restricted CO2-oil miscibility in heterogeneous reservoirs during CO2-enhanced oil recovery (CO2-EOR) remains a critical challenge. Herein, a novel pressure-regulated, hydrophilic CO₂-responsive hydrogel-to-foam transition system is used to enhance oil recovery and synergistically facilitate geological CO₂ sequestration. By molecular engineering of a quaternary ammonium surfactant (HXA-1) with unsaturated alkyl chains and hydrophilic moieties, the system achieves ultralow oil-water interfacial tension and adaptive micellar structures, enabling in-situ gel-to-foam phase transition under CO2 pressure. Core flooding experiments demonstrate that the Chemical-Assisted Water-Alternating-Gas (CWAG) technique forms a stable piston-like displacement front, boosting oil recovery by 16 % compared to conventional WAG while achieving 91.2 % plugging efficiency. In-situ visualization and molecular dynamics simulations reveal that the hydrophilic surfactant mitigates water-phase shielding and hydrophobic interactions, promoting CO2/oil miscibility and emulsification. Furthermore, the system exhibits dual functionality: under reservoir conditions, a 0.35 wt% HXA-1 solution absorbs twice as much CO2 as WAG, significantly enhancing sequestration, while the dynamic phase transition enables deep conformance control in heterogeneous reservoirs. This work pioneers a multifunctional CO2-responsive material platform, offering transformative solutions for sustainable fossil fuel extraction and carbon capture, utilization, and storage (CCUS) technologies.

In low-permeability reservoirs, the development of a barrier significantly influences fluid migration and remaining oil distribution during CO2 miscible flooding. Thus, a comprehensive study of the migration characteristics and dynamic sweep law of the CO2 miscible flooding front in low-permeability reservoirs with a barrier is crucial for optimizing gas flooding reservoir development design and enhancing oil recovery. Based on the geological features of a low-permeability reservoir block in the Jilin Oilfield, a visualized two-dimensional model with a barrier was designed and fabricated. Then, two sets of CO2 miscible flooding experiments were carried out using different injection-production modes. By analyzing the dynamic images and the key injection-production parameters during the experiments, the impact of the barrier on oil-gas migration and remaining oil distribution during CO2 miscible flooding in low-permeability reservoirs was investigated, and the front migration characteristics and dynamic sweep law of the CO2 miscible flooding in low-permeability reservoirs with barrier were revealed. The results show that: (1) The combined effects of gravity differentiation and barrier significantly influence the migration of oil and gas and the distribution of remaining oil. This impact directly leads to notable differences in the migration characteristics and dynamic sweep law of the flooding front under different injection-production modes. (2) Barrier affects CO2 displacement through dual mechanisms of interference and obstruction. (3) In this experiment, injecting CO2 from one side of the barrier achieved better oil displacement results than injecting from the side without a barrier. Therefore, the position of the barrier should be fully considered before actual field production, and a reasonable injection-production scheme should be formulated accordingly to mitigate the negative impact of barriers on the oil recovery. (4) To achieve efficient reservoir development, injection-production parameters should be monitored in real time during CO2 flooding, and the injection-production rates and positions should be dynamically adjusted based on the flow characteristics of gas flooding.

To meet the requirements of tertiary oil recovery in low-permeability, tight reservoirs, this study synthesized silica-based amphiphilic Janus nanoparticles. The organic hydrophobic groups were grafted onto the surface of silica nanoparticles via in situ generation and chemical fracture methods. The chemical structure of the synthesized Janus nanoparticles was characterized by transmission electron microscopy (TEM), atomic force microscopy (AFM), and Fourier transform infrared spectroscopy (FT-IR). Their interfacial properties were systematically investigated through measurements of interfacial tension, contact angle, and emulsification performance. The results indicate that when the concentration of Janus nanoparticles reaches or exceeds 0.02 wt%, the interfacial tension between the amphiphilic nanoparticle dispersion and crude oil remains on the order of 10–2 mN/m, demonstrating excellent interfacial activity. Additionally, the Janus nanofluid effectively alters rock wettability, converting it from oil-wet to water-wet. Janus-SiO₂ nanoparticles constructed with trimethylsilane as the hydrophobic carbon chain and hydroxyl groups as the hydrophilic end groups exhibit uniform particle size distribution and excellent dispersion stability in mineralized water. Core-scale displacement experiments with low-permeability sandstone indicate that the amphiphilic nanoparticle dispersion at a concentration of 0.02 wt% can further enhance oil recovery by 16.9 % compared to water flooding alone. The mechanisms by which amphiphilic nanoparticles improve oil recovery in low-permeability tight reservoirs mainly include reducing oil–water interfacial tension, altering rock wettability, promoting in situ emulsification, and strengthening the interfacial film, thus demonstrating promising application prospects for the development of low-permeability tight reservoirs.

Physicochemical interactions between CO2 and crude oil induce the deposition or blockage of heavy components. The integration of nuclear magnetic resonance (NMR) and theoretical calculations was employed to elucidate the pore-scale mass transfer mechanisms of CO2-heavy component interactions and quantify their impacts on flow. The results indicate that the interaction between CO2 and heavy components exhibits a pressure threshold that exceeds the miscible pressure of CO2 and heavy components. Thermal effect makes the impact of heavy components on flow approximately 1.8-2.5 times lower than low temperatures. When injection pressure is below the miscibility, low temperature and nano-confinement effect cause heavy components in micropores to gasify after CO2 injection, leading their migration towards macropores for liquefaction and then adsorption or blockage. Conversely, macropores' heavy components migrate towards micropores with thermal effect, resulting in endothermic adsorption. When injection pressure exceeds the miscible pressure, heavy components extracted by CO2 adsorb and form a boundary layer away from the pore wall. As injection pressure increases to the threshold, CO2 repeatedly contacts and extracts this fluid phase, eventually migrating out with the gas flow. This process can increase the maximum flow capacity by 70.09 % and pore volume by 8.12 %.

Shale reservoirs, characterized by abundant reserves and predominantly nano-scale pores and fractures, represent critical sources of unconventional oil and gas production. However, accurately capturing deformation behaviors of nano-to-micro-scale pore-fracture systems and describing microscale fluid-solid coupling phenomena under in situ stress remains challenging. In this study, we propose a homogenization-based approach to upscale micro-scale mechanical deformation and fluid flow results to the mesoscopic (core) scale, establishing effective fluid-solid coupling equations consistent with Biot's framework. Equivalent parameters are explicitly derived through three analytical expressions. Integrating centrifugal-nuclear magnetic resonance) and microfluidic experiments to characterize nano-confinement effects, simulation results indicate that, within the elastic deformation stage, the lower bound of producible oil in shale decreases from 20 to 16.19 nm. For organic pores, the minimum apparent permeability increases from 1.29x10(-5) to 1.86x10(-5) mD and the maximum from 2.97x10(-3) to 4.34x10(-3) mD, with the nonlinear flow regime boundary shifting from 90 to 73 nm. Inorganic pores exhibit negligible stress-induced deformation. Moreover, storage and Biot coefficients ( alpha(eff), gamma(eff)) increase with pore-fracture porosity. The effective elastic modulus ( aeff) varies significantly across reservoir types, with felsic shale exhibiting higher stiffness than clay-rich shale. Accordingly, the degree of fluid-solid coupling follows the order: quartz < feldspar < clay minerals < organic matter. For apparent permeability, small pores influence under a serial configuration, while large pores govern under parallel. Crude oil flow in complex nanopore networks depends on organic matter distribution, whereas multi-mineral interactions are dictated by bulk elastic modulus and Poisson's ratio.

Massive liquid injection is an effective energy replenishment method based on optimizing flow channels. However, the microscopic flow mechanism underlying massive flooding and massive soaking is not fully understood, and the influence characteristics of key injection-production factors on oil recovery remains unclear. Therefore, nuclear magnetic resonance (NMR) was employed to investigate the mass transfer process and microscopic flow mechanism of massive flooding and soaking. The impact of injection methods, injection liquids, fracture development, and displacement pressure differences was analyzed using the quantitative method. The performance of combination between massive flooding and soaking was also discussed. The findings indicate that the mass transfer rate and volume of oil and water between opened microfractures and matrix increase significantly during massive flooding, resulting in a minimum 9.65% higher recovery degree to conventional injection conditions. Macropores dominate oil drainage, and micropores experience a dynamic competitive phenomenon between oil backflow and recovery. The addition of surfactants slightly increases the oil recovery in micropores. Massive soaking transforms from dynamic imbibition dominated by macropores to static imbibition dominated by micropores along with soaking time, improving overall recovery by 4.63% compared to spontaneous conditions. Recovery speed is 4.56 times that of spontaneous conditions. The primary interaction target varies with surfactant properties. The combination of massive flooding and soaking takes full advantage of the two modes to enhance the oil recovery of full-scale pores.

Addressing gas channeling challenges in CO2-enhanced oil recovery (EOR) and geological sequestration, this study builds upon previous research on irreversible hydrogels to develop a novel CO2-responsive hydrogel-tofoam transition system based on a monomeric quaternary ammonium surfactant (HXA-1) with an unsaturated long-chain alkyl structure. The increased degree of unsaturation in the long-chain alkyl structure confers superior water solubility relative to the conventional HXB-2 surfactant, thereby endowing the system with unprecedented pressure-driven phase transition reversibility and enhanced CO2 absorption capabilities. CO2 bubbling at atmospheric pressure triggers hydrogel formation (viscosity (eta 0) surge from 166.75 to 2823 MPa.s), while highpressure CO2 induces a foam phase, with seamless reversion to hydrogel upon depressurization. Mechanistic studies reveal that protonation of HXA-1 under CO2 fosters self-assembly into entangled wormlike micelles (WLMs) via synergistic electrostatic, hydrogen-bonding, and hydrophobic interactions, forming a 3D hydrogel network. Under high pressure, WLMs dissociate into monomers that migrate to gas-liquid interfaces, generating stable foam. Microscopic characterization and molecular dynamics simulation validate this pressure-dependent micellar reorganization. Remarkably, the system demonstrates exceptional CO2 absorption (133.61 mg.g-1 at 10 MPa and 0.5 wt%, tripling pure water's capacity) and maintains robust thickening performance across pressures. This single-component, additive-free platform overcomes the limitations of multi-phase separation, chromatographic effects, and poor water solubility in traditional systems, offering a transformative solution for intelligent gas channeling mitigation and scalable CO2 storage. The work aligns with global carbon neutrality goals, paving the way for next-generation responsive materials in sustainable energy and environmental applications.

Fluid flow in microporous and nanoporous media exhibits unique behaviors that deviate from classical continuum predictions due to dominant surface forces at small scales. Understanding these microscale flow mechanisms is critical for optimizing unconventional reservoir recovery and other energy applications. This review provides a comparative analysis of the existing literature, highlighting key advances in experimental techniques, theoretical models, and numerical simulations. We discuss how innovative micro/nanofluidic devices and high-resolution imaging methods now enable direct observation of confined flow phenomena, such as slip flow, phase transitions, and non-Darcy behavior. Recent theoretical models have clarified scale-dependent flow regimes by distinguishing microscale effects from macroscopic Darcy flow. Likewise, advanced numerical simulations—including molecular dynamics (MD), lattice Boltzmann methods (LBM), and hybrid multiscale frameworks—capture complex fluid–solid interactions and multiphase dynamics under realistic pressure and wettability conditions. Moreover, the integration of artificial intelligence (e.g., data-driven modeling and physics-informed neural networks) is accelerating data interpretation and multiscale modeling, offering improved predictive capabilities. Through this critical review, key phenomena, such as adsorption layers, fluid–solid interactions, and pore surface heterogeneity, are examined across studies, and persistent challenges are identified. Despite notable progress, challenges remain in replicating true reservoir conditions, bridging microscale and continuum models, and fully characterizing multiphase interface dynamics. By consolidating recent progress and perspectives, this review not only summarizes the state-of-the-art but underscores remaining knowledge gaps and future directions in micro/nanopore flow research.

The distribution and mutual conversion of remaining oil during the process of oilfield development constitute an important basis for guiding the exploitation of remaining oil potential. Based on the visual core displacement method of CT scanning technology, CT scanning images of the oil–water phase in core models at different displacement stages were obtained, and the remaining oil types were classified. On this basis, image segmentation technology was employed to establish the transformation analysis method of remaining oil types, and the mutual transformation of microscopic remaining oil types at different displacement stages was clarified. The ability of displacement media to utilize various remaining oils was further clarified. The results demonstrate that there are significant differences in the distribution of remaining oil after the injection of different displacement media. The displacement media can not only spread the continuous-phase oil in large pores to varying degrees but also transform the discontinuous oil into continuous-phase oil in some small pore tubes, showing a “converging” transformation law, thereby enhancing the utilization degree of various remaining oils. Additionally, the surfactant’s unique capabilities of “micellar solubilization, emulsification, and oil carrying” have good adaptability to the discontinuous oil phase and can transform the discontinuous-phase remaining oil into continuous-phase remaining oil, namely columnar–film–cluster–recovery.

Fractured low-permeability reservoirs suffer from severe gas channeling and inefficient hydrocarbon mobilization during CO2-enhanced oil recovery (CO2-EOR). Here, we present a combined experimental and multiscale numerical strategy to unravel CO2 migration mechanisms and optimize recovery. A novel 2D visualized fracturematrix physical model is designed to dynamically track CO2-oil front propagation under varying injectionproduction rates (IPR), revealing that an optimal IPR of 0.3 mL/min maximizes sweep efficiency (77 %) and oil recovery (54.1 %). Deviations from this rate exacerbate gas channeling (high IPR) or limit matrix sweep (low IPR). Furthermore, we develop a breakthrough lattice Boltzmann method (LBM) framework integrating a confinement-adapted Peng-Robinson equation of state and dynamically updated fluid-solid interactions to simulate multicomponent transport in dual-porosity-microfracture matrix (DIMM) models. Simulations reveal that wider microfractures accelerate CO2 saturation but intensify bypassing, whereas narrower fractures promote uniform displacement while delaying diffusion. Crucially, microfracture width dictates hydrocarbon mobilization: medium hydrocarbons (Cm) exhibit width-dependent retention (narrow) vs. efficient extraction (wide), whereas light hydrocarbons (CL) remain unaffected. This work provides a predictive toolkit for CO2-EOR optimization in complex fractured reservoirs, bridging pore-scale physics to field-scale performance while advancing carbon sequestration strategies.

Injecting air or CO 2 into shale reservoirs can significantly enhance oil recovery (EOR) following the initial depletion. However, effectively characterizing the complex pore structure of shale reservoirs poses a challenge, leading to an incomplete understanding of the seepage mechanism and microscopic production characteristics of air/CO 2 flooding at different pore scales. In this study, we characterized the microscopic pore structure of shale reservoirs through the reconstruction of visual and quantitative digital cores in multiple dimensions. Subsequently, the online nuclear magnetic resonance (NMR) air/CO 2 flooding experiments were conducted, and the production characteristics and influencing factors of microscopic pore crude oil were quantitatively studied. The results show that the pore structure characteristics and connectivity of shale reservoirs are highly intricate and the deterioration of reservoir physical properties correlates with a decreasing trend in pore‐throat coordination numbers and heterogeneity. Shale oil primarily occurs in three types of pores (< 0.1, 0.1–1, and 1–10 μ m), and improving micronanopore recovery is urgent for EOR. Crude oil production is observed during the air and oil molecule generation low‐temperature oxidation (LTO) reaction. Additionally, CO 2 accelerates mass transfer and oil and gas extraction through molecular diffusion effects, substantially improving shale oil recovery; however, significant differences exist in the microscopic production characteristics of air/CO 2 flooding. High‐oxygen‐concentration air flooding or high‐pressure CO 2 proves beneficial for EOR, especially for small pores and macropores, which contribute 45.75%–53.42% recovery. This study provides scientific and theoretical support for clarifying the microscopic production characteristics and efficient development of shale oil.

Pore-fracture structure characterization and fluid mobility analysis are key to the effective development of tight reservoirs. However, it is difficult to accurately characterize the pore-fracture structure of tight volcanic rocks using a single method, and the mechanisms of fluid mobility are not fully understood. A combination of X-ray diffraction, casting thin section observations, scanning electron microscopy, high-pressure mercury injection, constant-speed mercury injection, nuclear magnetic resonance (NMR), X-ray computed tomography, and the advanced mathematical algorithms in the AVIZO visualization software was used to analyze the Permian tight volcanic rock in the ZhongGuai area of the Junggar Basin to investigate its overall pore-fracture structure characteristics. On this basis, NMR centrifugation experiments were conducted to monitor the fluid migration dynamics in the tight volcanic rocks, and the mobile fluid migration characteristics were studied based on the NMR T2 spectra. The results show that the porosity and permeability of the samples are 0.2-18.7 % (avg. of 8.3 %) and 0.001-6.680 x 10-3 mu m2 (avg. of 1.602 x 10-3 mu m2), respectively, and the various pore types are due to the differences in cementation and compaction. The distribution of the pore throats are mainly contiguous and isolated, and the connected pores are mainly distributed in enriched bands, which is due to the interconnection of gas pores, intergranular pores, and dissolution fractures; while the disconnected pores are mainly isolated, which is related to the development of inter-gravel dissolved pores and matrix dissolution pores. In addition, the contribution of the pore connectivity to seepage is greater than that at the pore scale. The pores in the tight volcanic rock have a wide range of sizes, from 3 nm to 120 mu m, and are dominated by Gaussian and bimodal distribution patterns. The micron-scale pore radius in this area is mainly 4.47-31.56 mu m. Also, the fractures can be divided into three types according to their occurrence and openings. They are mainly high-angle structural fractures and vertical fractures; the pore-fracture structures have strong heterogeneity, and the fractures play a larger role on the seepage of oil and gas. The connectivity of the pore throats in the tight matrix is poor, and the pore throat structure has a great influence on fluid mobility. Subsequently, the movable fluid saturation increases with increasing permeability, and the fractures and micropores have less flow resistance and are more conducive to water flow than small pores. This study provides new insights for the exploitation of similar tight reservoirs.