Tight conglomerate reservoirs exhibit strong pore-scale heterogeneity and extremely low permeability, in which spontaneous imbibition is primarily governed by capillary and viscoelastic effects. In this study, the imbibition dynamics of four representative fracturing fluid systems, including slickwater, 3% potassium chloride (KCl) brine, hydrolyzed polyacrylamide (HPAM) viscoelastic fluid, and a nanoemulsion (NE), were investigated using a temperature-controlled nuclear magnetic resonance (NMR) monitoring system. This approach enables real-time quantification of fluid uptake and pore-scale redistribution through time-resolved T-2 spectral analysis. The experimental results reveal a three-stage imbibition process consisting of rapid capillary-driven uptake, viscoelastic-retarded transition, and final equilibrium. Among the four fracturing fluid systems, the nanoemulsion exhibits the lowest interfacial tension (1.72 mN/m), the strongest wettability alteration, and the highest equilibrium recovery (0.76), which is nearly 80% greater than that of slickwater. Based on these observations, a multiscale capillary-viscoelastic coupling model was developed by extending the Lucas-Washburn framework to incorporate pore-size distribution, time-dependent wettability evolution, and viscoelastic damping. The model fits the experimental data well (R-2 > 0.90) and identifies viscosity as the most influential parameter controlling the imbibition rate (sensitivity = 0.78). Energy analysis further indicates that capillary energy dominates the early stage, whereas viscoelastic energy storage sustains fluid transport during the later stage. SEM observations were further used to qualitatively corroborate pore heterogeneity and pore-mineral associations, supporting the NMR-based pore-scale interpretation. This study provides a quantitative framework for describing non-Newtonian capillary flow in tight conglomerate rocks and enhances the understanding of capillary-viscoelastic interactions relevant to multiphase fluid migration.

[Objectives and Methods]Deep coalbed methane(CBM)reservoirs commonly exhibit well-developed bed-dings,strong mechanical heterogeneity,and high in-situ stress gradients.These characteristics result in pronounced non-linear fracture propagation and strong multi-field coupling effects during hydraulic fracturing.Consequently,it is chal-lenging to accurately describe the mechanisms governing fracture complexity in deep coal reservoirs using conventional mechanical models for fractures.Using a super-large true triaxial system with dimensions of 2.0 m × 2.0 m × 1.0 m,this study conducted physical simulation experiments on hydraulic fracturing under varying injection rates and viscosities of fracturing fluids.In combination with fracture mechanics and energy conservation theory,this study established an en-ergy balance equation for fracture propagation,a convection-diffusion equation for proppant transport and settling,and a model for the coupling relationships among fracture complexity and the injection rate and viscosity of fracturing fluids.Accordingly,both the dynamic mechanisms behind fracture evolution and the pattern governing the fracture network complexity were systematically elucidated.[Results]The results indicate that fracture propagation is jointly controlled by the in-situ stress field,fluid pressure field,and bedding structures,representing a unsteady energy conversion process.The fracture propagation rate exhibits a power-law relationship with the energy release rate.The injection rate of fractur-ing fluids primarily determines the energy input rate and fracture propagation velocity.A high injection rate results in energy concentration in the front of the primary fracture,promoting fracture interconnectivity while suppressing branch development.Accordingly,fracture complexity is reduced.In contrast,a low injection rate corresponds to a more uni-form energy distribution,enhancing the accumulation and lateral diffusion of energy.This facilitates multi-point initial cracking and fracture branching,increasing fracture complexity by approximately 25％-35％.Fracturing fluid viscosity significantly influences the energy transfer between fluids and solids,as well as proppant settling behavior.A high vis-cosity(45 mPa·s)is associated with a significant decrease in the proppant settling velocity.Compared to a low viscosity of 15 mPa·s,the high viscosity increases the proppant transport capacity by approximately 40％,promoting more uni-form proppant placement in far-wellbore zones and creating favorable conditions for the formation of continuous hy-draulically conductive pathways.[Conclusions]Empirical relationships derived from experiments and fitting indicate that the fracture complexity exhibits power-law coupling relationships with the injection rate and viscosity of fracturing fluids.Notably,the low-injection-rate and high-viscosity combination is more favorable for the development of 3D frac-ture networks,with a fractal dimension reaching up to 1.46.The proposed theoretical-experimental coupling framework reveals the energy transfer mechanisms governing fracture propagation and proppant transport in deep coal reservoirs,providing a quantitative theoretical basis for optimizing hydraulic fracturing parameters and predicting fracture complex-ity in deep unconventional reservoirs.

The efficient development of deep coalbed methane (CBM) faces challenges including complex geological conditions, sensitivity of fracturing parameters, and data scarcity. This study focuses on a block within the Ordos Basin and proposes a hybrid Artificial Intelligence modeling methodology integrating data augmentation, ensemble learning, and interpretability analysis. The Synthetic Minority Over-sampling Technique (SMOTE) was employed for data enhancement. Based on 17 geological and engineering parameters, a Stacked Generalization ensemble model integrating multiple algorithms including Random Forest, Support Vector Machine, and Gradient Boosting was constructed through randomized search hyperparameter optimization. Furthermore, the Shapley Additive Explanations (SHAP) method was introduced to identify dominant controlling factors, combined with Particle Swarm Optimization (PSO) to achieve collaborative optimization of fracturing parameters. Results demonstrate that geological parameters are the primary controlling factors for post-fracturing productivity. Among geological parameters, gas content, reservoir pressure, and Young's modulus show significant influence, while among engineering parameters, low-viscosity slickwater volume, pad fluid volume, highviscosity slickwater volume, and pumping rate exhibit considerable impact. After SMOTE and Stacking integration modeling, the production prediction model achieved acceptable prediction accuracy. The optimal fracturing parameter intervals were determined as: low-viscosity slickwater volume 100-150 cubic meters (m3), pad fluid volume 200-400 m3, high-viscosity slickwater volume 100-300 m3, and pumping rate 18-20 cubic meters per minute (m3/min). This study provides an interpretable and scalable methodological framework for fracturing optimization under data-scarce conditions in deep CBM development, offering valuable references for intelligent development of unconventional oil and gas resources.

To reveal the energy transfer mechanism of water injection and the dynamic response characteristics of pore pressure in tight sandstone reservoirs, and to clarify the influence of lithology, injection pressure, and injection method on the energy enhancement effect of water injection, a high-pressure energy injection and response testing system and nuclear magnetic resonance (NMR) testing technology were used to conduct systematic water injection energy enhancement experiments on three rock types: mudstone, sandstone, and naturally fractured sandstone. Combined with pressure dynamic monitoring and pore structure evolution analysis, the pressure response characteristics and energy enhancement mechanism of rock samples under different experimental conditions were explored. The experimental results showed that the NMR T2 distribution of the three rock samples exhibited bimodal characteristics, corresponding to small pores (pore size < 1000 nm) and large pores/microcracks (pore size > 1000 nm), respectively. There were significant lithological differences in the evolution of pore structure during water injection, with a cumulative decrease of 7.2% in the proportion of large pores in mudstone and an increase of 9.3% in the proportion of large pores in sandstone with natural fracture development. There is a positive correlation between injection pressure and the energy enhancement effect. Under an injection pressure of 40 MPa, the pressure increment at the outlet end of sandstone with natural fracture development reaches 8.06 MPa, and the energy enhancement effect is 24% higher than that under the 30 MPa working condition, while the mudstone only increases by 15%. The energy enhancement effect of intermittent water injection is significantly better than that of depleted water injection, and the energy enhancement effects of the three rock samples are increased by 18.6%, 12.0%, and 6.9%, respectively. Overall, sandstone with natural fractures has the best energy enhancement effect, followed by sandstone, and mudstone has the worst. The connectivity of pores and the degree of fracture development are the core factors that dominate the water injection energy enhancement effect and pressure transmission efficiency. The research results can provide reliable experimental basis and theoretical support for optimizing water injection development plans, improving energy efficiency, and dynamically regulating stress fields in tight sandstone reservoirs.

During the energy storage fracturing process of tight sandstone reservoirs, the pre-injection of fracturing fluid is used to supplement the formation energy, and the physical properties of rocks change under hydration. To reveal the damage mechanism of hydration on tight sandstone, the tight sandstone surrounding the Daqing Changyuan in the northern part of the Songliao Basin was taken as the research object. Through indoor static hydration experiments, combined with scanning electron microscopy (SEM), nuclear magnetic resonance (NMR), Nano-indentation experiments, and other methods, the evolution laws of rock micro-pore morphology, microfracture parameters, Young's modulus, hardness, and other mechanical indicators under different hydration durations and soaking pressures were systematically explored. The research results show that the water-rock interaction of acidic slick water fracturing fluid significantly changes the mineral composition and microstructure of mudstone and sandstone, controls the development of induced fractures, and degrades the micro-mechanical properties of rocks, with significant lithological differences. In terms of mineral evolution, the soaking time causes the clay minerals in mudstone to increase by up to 12.0%, while pressure causes the carbonate minerals in sandstone to decrease by up to 23.3%. In terms of induced fracture development, the induced fracture widths of sandstone and mudstone under 30 MPa of pressure increase by 122.4% and 85.7%, respectively. The fracture width of mudstone shows a trend of "increasing first and then decreasing" with time, while that of sandstone decreases monotonically. In terms of micro-mechanical properties, after soaking for 168 h, the Young's modulus of mudstone decreases by up to 66.9%, much higher than that of sandstone (29.5%), while the decrease in hardness of both is similar (58.3% and 59.8%); the mechanical parameters at the induced fractures are only 53.0% to 73.6% of those in the matrix area, confirming the influence of microstructural heterogeneity. This research provides a theoretical basis and data support for optimizing hydraulic fracturing parameters, evaluating wellbore stability, and predicting the long-term development performance in tight sandstone reservoirs.

Coalbed methane (CBM) is an important alternative energy source, while its efficient development relies on multistage hydraulic fracturing technology. A large amount of fracturing fluid is injected underground; however, an accurate evaluation of fracture network properties is a challenge. Meanwhile, the massive consumption of water resources and flowback efficiency has also drawn concern from the industry. In this study, we first identify the flow regimes of successive depletion of primary and secondary fractures based on flowback data from field cases, using the traditional flowing-material-balance (FMB) method. Based on this finding, a subdivided FMB model is proposed to analyze two pseudosteady-state (PSS) flows during flowback. Then, a workflow is established to extract the information on primary and secondary fractures, and its accuracy has been validated through numerical simulations conducted by a commercial simulator. Furthermore, the new approach is applied to flowback data from five multifractured wells, and the correlation analysis between the inversion results and the fracturing completion parameters is conducted. The results show that the porosity of primary fractures has no significant correlation with the amounts of proppants injected but increases with the sand-to-liquid ratio. The volume of secondary fractures is positively correlated with the total injected fluid volume. However, as the vertical depth and closure pressure increase, an obvious reduction in fracture volume is demonstrated, attributable to fracture closure. For shallow CBM, the forecasted flowback efficiency is generally higher than 67%, which is recommended to determine the appropriate treatment processes for recycling and reuse. For deep CBM, to ensure the effectiveness of proppant filling within fractures, it is advisable to increase the proportion and amounts of small-particle proppants.

Fracture morphology in unconventional reservoirs after hydraulic fracturing directly affects production. Evaluating this morphology is crucial for analysing the effects of reservoir stimulation. This study proposes a method for evaluating complex fracture networks in microseismic data from fracturing using topological reconstruction. Traditional microseismic interpretation methods struggle to distinguish effective fracture responses from outlier events and unconnected regions. At the same time, static geometric connections overlook the dynamic growth of fractures and the constraints imposed by stress. During the research process, microseismic events were preprocessed using spatial density, stress direction, and manual review. This approach helped construct a highly confident, effective event set and reduce interference from abnormal events. Based on the basic process characteristics of fracture propagation—namely, initiation, expansion, and branching—a time-series-driven fracture topology growth model is proposed. This model simulates the dynamic expansion of fractures along the path of minimum resistance in the formation. Building on fracture propagation paths, an SRV/ESRV evaluation method is introduced. This method uses response unit slicing, the convex hull, and magnitude heterogeneity correction to distinguish between the spatial sweep volume and the effective stimulated volume. These steps aim to improve the physical rationality and engineering interpretability of complex fracture network evaluation. The field application of the Y1 well in the Ordos Basin shows that, compared with the traditional geometric envelope method, the SRV proposed in this paper avoids the virtual high or irregular distribution underestimation of the outer envelope. The overall ESRV improves by about 9.22% compared with the geometric envelope method and shows a more stable, positive linear correlation with injection volume. The inversion of fracture propagation effectively identifies multi-cluster competition, asymmetric propagation, and differences in the development of natural fractures. It is in good agreement with geological modelling and construction pressure-drop characteristics. This method provides a new approach to quantitatively evaluating fracture networks after hydraulic fracturing, combining physical consistency with engineering practicality.

Objective Rock bridges exert a controlling influence on the deformation, fracture field evolution, and structural stability of mining-induced fractured rock masses. Therefore, it is necessary to investigate their deformation and failure characteristics, as well as the underlying controlling mechanisms, under different confinement conditions. Such investigations will provide theoretical support for the safe mining of underground coal seams and the prevention and control of mining-induced disasters. Methods Mechanical models of fractured rock blocks containing rock bridges were established. Specimens with different fracture dip angles, rock bridge widths, and material strengths were prepared using rock-like materials. Furthermore, an independently developed compressive shear testing system capable of applying lateral confinement was employed to conduct compressive shear failure tests on the specimens under two conditions: in the absence of lateral confinement and under unilateral confinement. Results and Conclusions The load-displacement curves reveal that in the absence of lateral confinement, the specimens underwent three evolutionary stages: compaction accompanied by fracture initiation and propagation, post-peak failure induced by fracture penetration, and residual frictional equilibrium. Without lateral confinement, the rock bridge failure load increased linearly with material strength and rock bridge width, while showing a negative linear correlation with the sine of the fracture dip angle. Furthermore, the cohesion and internal friction angles obtained by analyzing the stresses acting on the rock bridge interfaces in the mechanical models exhibited relative errors of consistently less than 5.8% compared to the results of double-sided shear tests. Under unilateral confinement, the rock bridge failure load increased by an average of 41.64%, with the failure mechanisms governed by the fracture dip angle. Specifically, specimens with fracture dip angles of 57° or 75° primarily exhibited lateral contraction prior to rock bridge failure, while the normal load remained relatively low. Following rock bridge failure, the specimens experienced outward expansion, resulting in a sharp increase in normal load. In contrast, specimens with a fracture dip angle of 90° underwent lateral expansion during the initial loading stage due to the tension-shear effect. Concurrently, the normal load applied to the specimens increased with continued loading and gradually stabilized after the penetration failure of rock bridges. The vertical load transfer coefficient increased as the fracture dip angle decreased and increased slightly with specimen strength due to the limited influence of the latter. These findings indicate that rock blocks with high strength and low fracture dip angles exhibit greater load transfer efficiency. The results of this study provide a theoretical basis and novel experimental approaches for evaluating the stability and determining the strength parameters of mining-induced fractured rock masses.

Traditional desorption methods for determining coalbed methane gas content suffer from high costs, lengthy cycles, and difficulties in regional coverage, making them unsuitable for rapid economic viability assessments and cost-saving efficiency improvements. To address this, a coalbed methane gas content prediction method based on an improved Stacking fusion algorithm is proposed. This method optimizes the hyperparameters of five heterogeneous base models—Support Vector Machine (SVM), Random Forest, Extreme Gradient Boosting, Lightweight Gradient Boosting, and Multi-Layer Perceptron (MLP)—through grid search. It innovatively constructs a new meta-training set by weighting the average of raw training set features and predictions from each base model, then inputs this set into a Linear Regression meta-model for secondary learning. This achieves dual integration of raw features and base model prediction information. When applied to coalbed methane blocks in the southern Qinshui Basin, results demonstrate that the improved Stacking fusion model achieves a test set coefficient of determination of 0.94, representing improvements of 10.64% and 6.38% over the optimal single base model and standard Stacking model, respectively. The root mean square error, mean absolute error, and mean squared error were reduced by 14.3%, 25.2%, and 43.7%, respectively, compared to the standard Stacking model. SHAP interpretability analysis indicates that coal seam dip angle, natural potential, and natural gamma are the primary features influencing gas content prediction. Field application demonstrated good agreement between predicted and measured values for eight wells, with a mean absolute error below 0.84 m³/t and relative error controlled within 6%. This model significantly enhances the accuracy and reliability of coalbed methane gas content prediction, providing a new technical approach for coalbed methane resource evaluation and efficient development.

Multi-stage multi-cluster hydraulic fracturing in conglomerate reservoirs is often characterized by strong cluster-to-cluster variability in fluid distribution, which can reduce stimulation efficiency. However, field-scale observations that constrain how injected fluid is partitioned among clusters remain limited, especially in strongly heterogeneous formations. In this study, wide-field electromagnetic (WFEM) monitoring was applied to a horizontal well completed in the Baikouquan Formation sandstone–conglomerate reservoir of the Mahu Sag, Junggar Basin. The monitored treatment consisted of 13 fracturing stages, each containing six perforation clusters. Time-lapse electromagnetic data acquired during pumping were inverted to reconstruct the spatiotemporal evolution of the effective conductive fluid-swept region. Based on the inversion results, we introduce a set of quantitative proxy indicators (swept area, swept length, cluster-specific sweep, and an asymmetric index) to support relative comparison of fluid distribution patterns at both stage and cluster scales. Results show pronounced non-uniformity within and between stages, even under similar pumping conditions. A limited number of clusters exhibit stronger and farther-reaching WFEM-inferred conductive-fluid responses, whereas other clusters show weaker or more localized responses. Asymmetric sweep patterns on opposite sides of the wellbore are also commonly observed. These patterns are consistent with the combined influences of reservoir heterogeneity, local structural/stress disturbances, and operational factors, although WFEM alone does not uniquely validate causal mechanisms of fracture growth. Overall, this study demonstrates that WFEM monitoring provides a field-scale proxy tool for delineating effective conductive fluid-swept regions and for evaluating cluster-to-cluster variability under consistent acquisition and inversion settings. The findings offer practical guidance for interpreting fluid distribution and optimizing multi-cluster fracturing in strongly heterogeneous unconventional reservoirs.

The railway hump is a critical facility in train classification, and its geometric design directly affects operational efficiency and investment. Existing studies mainly focus on vertical profile optimization while keeping the horizontal alignment fixed, which restricts the search space and may overlook superior design alternatives. To resolve this issue, a 3D track layout optimization model for railway humps is developed, jointly optimizing horizontal and vertical hump alignments under their coupling constraints. Rolling efficiency and construction cost are formulated as bi-objective functions. To efficiently solve the constrained optimization problem, a particle swarm optimization (PSO) algorithm with a feasibility-bound preprocessing strategy is proposed to avoid infeasible search regions and enhance the search efficiency. In a real-world case study, the recommended solution reduces rolling time and intervals between successive railcars by 15.21% and construction cost by 4.02% compared with the best manual design. Comparative experiments with genetic algorithm (GA), original PSO, and a customized PSO variant demonstrate that the proposed method achieves improved solution quality and computational efficiency.

To address the problem of high-precision microstructure identification in the deep No. 8 coal seam of Carboniferous Benxi Formation in the Ordos Basin, “elemental fingerprints” of the deep coal seam and its roof and floor strata were constructed, and measurement points along the horizontal well were mapped to their corresponding reference positions in the vertical profile of the pilot well, thereby enabling the identification of microstructural features along the horizontal section of deep coalbed methane horizontal wells. First, based on elemental logging data and constrained by the lithological framework, candidate elements were determined for distinguishing the coal seam from its roof and floor strata and for identifying different coal lithotypes. Then, the candidate elements were screened using the coefficient of variation to obtain the characteristic elements for microstructure identification. The concentrations of the characteristic elements were adjusted using Z-score normalization, and the combination of characteristic-element concentrations at the same measurement point was expressed as a vector, termed an “elemental fingerprint”. Subsequently, the elemental fingerprints of the pilot well were matched with those of the horizontal section to determine the vertical-profile positions and distributional variations of the horizontal section. Finally, horizontal-well drilling information was integrated to determine the locations and types of microstructures developed along the horizontal section, thereby guiding adjustments to the drilling direction of the bit. Field drilling and hydraulic- fracturing practices demonstrate that the elemental fingerprint-based microstructure identification method can accurately identify microstructural features (e.g. small-scale folds, superimposed faults, and tectonically disturbed zones) along the horizontal section, and provide technical support for the optimized design of drilling and hydraulic fracturing in deep coalbed methane horizontal wells.

A carbon dioxide (CO2)-hybrid fracturing approach combines the benefits of CO2 and water-based fracturing techniques, where liquid CO2 is injected to initiates hydraulic fractures (HFs), and is then followed by the injection of proppant-laden water-based fluid to further extend and support the HFs. After being heated by the wellbore and reservoir, liquid CO2 generally transforms into its supercritical state. There is an interface between supercritical CO2 (Sc-CO2) and the water-based fluid, which advances inside the HF. Due to the strong contrast between the two fluid viscosities, the interface dynamics can affect fluid pressure and HF width distributions in two regions occupied by dissimilar fluids. For this study, we developed a fully coupled HF propagation model to deal with two-phase fluid flow in fractures by describing the time-dependent movement of interface with a volume of fluid (VOF) equation. First, we verified the model capability to track interface movement using an analytical solution to the fracture problem with constant inlet pressure. Then, we performed a parametric study to determine the mechanisms affecting HF propagation in layered reservoirs. Numerical results revealed that HF height is significantly smaller than its length at the end of Sc-CO2 fracturing, as the higher-stress layers strongly delay HF vertical growth due to higher-rate leakoff of low-viscosity Sc-CO2 into rock mass and bedding planes (BPs). At this stage, the Sc-CO2 front coincides with the horizontal HF tip, but as fracturing proceeds, the HF propagates vertically into the adjacent higher-stress layers, and horizontal growth is temporarily suppressed until the water-based fluid catches up with the HF tip. The duration of coexistence of two fluids in the HF depends primarily on the pumping rate of water-based fluid. The subsequent fracturing of water-based fluid promotes an increase in HF length and height. For the case studied, the leakoff distance of Sc-CO2 along BPs can reach approximately 56 m, which is about 13 times greater than that of high-viscosity linear gel. The large-scale infiltration of Sc-CO2 elevates the pressure within the BPs, limiting the leakage of linear gel in facilitating HF extension. Obviously, there is an optimization in Sc-CO2 leakoff and HF growth rate by adjusting the pumping rate and amount of liquid CO2, and the optimal operational parameters of CO2-hybrid fracturing play a role. Finally, we performed an optimization study for a vertical shale oil well, and the proposed model and findings offer practical guidance for designing CO2-hybrid fracturing treatments in multilayered reservoirs.

The complex interaction behavior between hydraulic fractures (HFs) and gravels in heterogeneous glutenite reservoirs makes the near-wellbore HF propagation geometries still unclear. In this study, fracturing experiments combined with CT scanning and rate step-down tests were conducted to analyze the effects of horizontal stress difference (Δσ), fluid viscosity (μ), gravel size (dg), and stiffness ratio of gravel to matrix (Eg/Em) on the pressure response and near-wellbore HF geometries. Importantly, a novel method was developed to quantitatively characterize the near-wellbore fracture tortuosity by introducing fitting coefficients Knw and m. Results indicate that there are four types of injection pressure responses, with type-3 and type-4 responses accounting for the highest proportion of 37.5% and 42.9% in slick-water and gel fracturing, respectively. With the increase of Δσ and μ or with the decrease of Eg/Em, the HF geometry shifts from the complex gravel-bypassing fracture network to a bi-wing main fracture penetrating gravels, and Eg/Em has the most significant impact on fracture complexity in glutenite. It is found that the interpreted values of Knw and m are 0.016∼2.772 and 0.384∼1.1, respectively, with them following a logarithmic correlation; as dg rises and μ decreases, Knw increases while m and equivalent fracture width reduce. Besides, a Knw-m plate is established to predict fracture tortuosity by incorporating the effects of gravel cementation, gravel size, and fluid viscosity. This investigation aims to give theoretical insights into HF-gravel interaction behavior and near-wellbore fracture geometry, providing guidance for fracturing optimization in glutenite reservoirs.

As one of the key indicators for evaluating acid fracturing effectiveness， the conductivity of acid-etched fractures is crucial for optimizing acid fracturing designs and assessing productivity. Currently， experimental methods are commonly used to fit and calculate the conductivity of acid-etched fractures， but there is a lack of corresponding numerical models. Therefore， this study proposed a conductivity calculation model for acid-etched fractures based on the mechanical properties of the real fracture surface. Firstly， a series of experiments on the conductivity of acid-etched fractures and tests on the mechanical parameters of the acid-etched fracture surfaces were carried out to obtain the conductivity of acid-etched fractures and the mechanical properties of the acid-etched fracture surfaces under different acid-etching characteristics. Additionally， 3D laser scanning was employed to acquire the etched fracture morphology. Based on linear elastic theory， the model introduced a contact ratio to calculate the deformation of acid-etched fractures， establishing a conductivity model for acid-etched fractures under varying closure stresses and solving it via programming. The model was calibrated and validated using experimental data， including fracture morphology， mechanical parameters of acid-etched fracture surfaces， and fracture conductivity. By using this model， the effects of fracture corrosion morphology and the fracture surface’s Young’s modulus on the conductivity of acid-etched fractures were studied. The results show that fracture width varies significantly with closure stress for acid-etched fracture surfaces with different corrosion morphologies. Observable apertures and continuous flow paths remain at 80 MPa closure stress for grooves formed by corrosion in rock plates， leading to relatively stable conductivity. In contrast， the conductivity decreases rapidly as closure stress increases for rock plates with vertical-bar flow barriers or uniform corrosion. The conductivity of acid-etched fractures declines as the Young’s modulus of the fracture surface decreases. Rock plates with vertical-bar flow barriers exhibit near-zero conductivity at 50 MPa closure stress when the Young’s modulus is 5 GPa.

Studies at hydraulic fracturing test sites (HFTSs) in North America and the Changqing oilfield, China, show that hydraulic fractures (HFs) commonly occur as closely spaced groups, or fracture swarms, with fracture numbers greatly exceeding perforation clusters. To investigate the formation mechanism and evolution of fracture swarms, continuous scratch testing and nanoindentation were used to characterize the mechanical heterogeneity associated with different lithologies, lithologic interfaces (LIs), and bedding planes (BPs). Hydraulic fracturing experiments were conducted on representative specimens from the Changqing HFTS using an improved small-scale true-triaxial system integrated with high-resolution CT scanning, high-frequency pressure monitoring, and acoustic emission (AE) monitoring. The results show that mechanical heterogeneity associated with LIs and BPs destabilizes cross-layer propagation of the main fracture and promotes fracture deflection, arrest, stepwise propagation, and secondary fracture re-initiation near weak planes, thereby favoring swarm development, whereas the homogeneous sandstone control specimen failed to generate such fracture patterns. When the interface contrast index Hi of a BP exceeds 0.1, the HF is more likely to be captured by the weak plane, which suppresses swarm development; when the Hi of an LI exceeds 0.1, local fracture deflection and secondary fracture re-initiation are more likely, whereas lower Hi favors direct crossing. Mechanistically, when weak planes satisfy the activation condition (Iact,i ≥ 1) but not the re-initiation condition (Ire,i < 1), fracture propagation is limited to local slip or minor deflection; when both conditions are satisfied (Iact,i ≥ 1 and Ire,i ≥ 1), secondary fractures are more likely to form and evolve into swarms. Main fractures exhibit larger apertures than secondary fractures, accompanied by alternating pressure peaks and troughs and AE signatures marked by increased event numbers, shear-event enrichment near weak planes, and sustained multipeak energy release. These findings provide new experimental insights into fracture swarm formation in shale reservoirs.

The complex propagation behavior of hydraulic fractures (HFs) in strongly heterogeneous conglomerate reservoirs poses significant challenges for effective reservoir stimulation. In particular, the interaction between fractures and gravel-induced heterogeneity often leads to highly tortuous fracture networks and uneven stimulation efficiency. To address this issue, a series of laboratory true triaxial hydraulic fracturing experiments were conducted on artificially prepared conglomerate specimens with controlled gravel size and distribution. A quantitative evaluation index, termed the Fracture Complexity Index (FCI), was proposed to characterize the tortuosity and complexity of fracture networks by integrating multiple geological and engineering factors. The effects of cluster spacing and fracturing fluid viscosity on multi-fracture propagation behavior were systematically investigated. The results show that increasing cluster spacing enhances inter-fracture interaction and promotes fracture tortuosity, while lower fluid viscosity facilitates fracture branching but may limit effective propagation distance due to energy dissipation. To further quantify the trade-off between fracture complexity and propagation extent, a dimensionless fracture length was introduced and combined with FCI to establish a fracture morphology evaluation framework. This framework enables the classification of fracture patterns and reveals the coupling relationship between engineering parameters and fracture geometry. The findings provide new insights into the mechanisms of fracture propagation in conglomerate reservoirs and offer a quantitative basis for optimizing fracturing design, particularly in balancing fracture complexity and effective stimulation range in strongly heterogeneous formations.

A critical strategy for shale reservoir development is the comprehensive reservoir evaluation and the fine division of fracturing grades. However, the diversity of evaluation parameters limits hydraulic fracturing optimization. Therefore, we propose an adaptive-category-gaussian-mixture-model (AC-GMM) based on a geology engineering framework, combining reservoir quality (RQ) and completion quality (CQ) to classify the composite quality index (CQI). The classification serves as the basis for an intelligent algorithm developed for fracturing design. Taking three typical wells from the Lucaogou Formation in the Junggar Basin in China as examples the following research results are summarized. First, the AC-GMM model can finely identify the fracturing grades, achieving a conformity rate of over 90 % with the field production data. Second, the paper obtains three types of fracturing grades (I, II, III) and further refines them into four grades (I, II1, II2, III), the grade I considers both high RQ and CQ, while grade II only regards the better of the double quality, and prioritizes the better CQ. Third, the intelligent algorithm groups similar qualities into the same stage, achieving up to 96 % intra-stage homogeneity, significantly enhancing hydraulic fracturing efficiency for long horizontal wells. Our work provides a data-driven framework for optimizing multi-stage fracturing designs in shale reservoirs.

During heavy oil thermal recovery processes, hydraulic fracturing is an effective method for optimizing steam injection efficiency. However, the fractures created by the fracturing operation may connect with high-water saturation zones, forming high-temperature fracture-induced water channeling pathways. This not only increases water production in the wells but also significantly reduces oil and gas production, negatively affecting development efficiency. Gel-based plugging agents are commonly employed to effectively block these water channeling pathways. However, traditional polymer gels exhibit instability under high-temperature conditions, resulting in leakage and diminished plugging performance. In this study, a high-temperature-resistant copolymer gel suitable for 150°C environments was developed, and its gelation performance, thermal stability, and plugging effectiveness were systematically evaluated. The experimental results indicate that the gelation time of the copolymer gel at 150°C is 10 hours, with the post-gelation viscosity reaching 3483 mPa·s. After 30 days of aging at 150°C, the gel maintained a high viscosity, and its micro-network structure remained stable at approximately 100 μm. In core fracture plugging experiments, the plugging efficiency reached 96.2%. Furthermore, numerical simulations of heavy oil thermal recovery fracture plugging, based on laboratory experimental data, further validated the excellent plugging performance of the copolymer gel under high-temperature conditions. The gel effectively plugs fracture-induced water channeling pathways, reduces the water cut at the production well, and increases oil and gas production, providing strong technical support for the efficient development of heavy oil thermal recovery.