ABSTRACT In one-dimensional (1D) geomechanical modeling, precise predictions are essential for ensuring wellbore stability, defining safe mud weight windows, and determining hydraulic fracturing pressure ranges. However, accurately estimating geomechanical parameters, especially in data-limited environments, remains a significant challenge. This study addresses these challenges by integrating advanced machine learning algorithms with well log data from two wells in one of Iran’s hydrocarbon fields. The dataset includes shear wave velocity (Vs) logs, static Young’s modulus (Est) samples, uniaxial compressive strength (UCS) samples, internal friction angle ( φ ) measurements, and pore pressure (PP) samples. To tackle the challenge of predicting Vs with high precision, three sophisticated algorithms were employed: Bidirectional Long Short-Term Memory (BiLSTM), BiLSTM+Transformer, and a hybrid model (CBiT) combining 1D Convolutional Neural Network (CNN), BiLSTM, and Transformer mechanisms. Feature selection using Pearson correlation and Mutual Information and SHapley Additive exPlanations (SHAP) analysis identified the most influential parameters for each algorithm. The models were evaluated against traditional empirical methods using statistical metrics, such as mean square error (MSE), root mean square error (RMSE), and coefficient of determination (R2). The results revealed that the machine learning models significantly outperformed the empirical methods, with the hybrid CBiT model showing superior accuracy, robustness, and generalization capability, establishing a new benchmark in predictive performance for shear wave velocity due to its ability to capture complex nonlinear relationships and long-term dependencies. Particle Swarm Optimization (PSO) was applied to further optimize model hyperparameters, achieving R2 values of 0.9989 (training) and 0.9985 (testing), with RMSE for unseen data reduced to 0.0269 and R2 reaching 0.9973. Dynamic elastic parameters, including Young’s modulus, bulk modulus, and shear modulus, were calculated using the optimized model. Strong correlations between dynamic and static properties, such as dynamic Young’s modulus and UCS, were extrapolated across the entire depth range of both wells. The study also introduced a novel approach to estimate internal friction angle using Plumb’s correlation and laboratory data, alongside improved pore pressure predictions based on Eaton’s empirical method. Finally, vertical and horizontal stresses were calculated using poroelastic equations for comprehensive stress modeling. This study demonstrates how the integration of hybrid machine learning algorithms, advanced optimization techniques, and rigorous data analysis can significantly improve the accuracy, reliability, and practical applicability of 1D geomechanical models, providing a robust and efficient framework for subsurface analysis and offering substantial advantages over conventional approaches.

ABSTRACT Proper prediction of the length of perforation penetration plays an essential role in maximization of well productivity and improved reservoir connectivity, but the standard empirical models are still limited in their scope, often necessitating recalibration and inadequate coverage of gun and well parameters. This study mitigates these shortcomings by building an advanced machine learning model that is trained on 1,648 API-19B real-world standard perforation tests, which constitute 16,480 data points under a variety of operating conditions. Routinely measured input features were used to evaluate ten machine learning algorithms. These features include shot density, casing grade, casing nominal weight, casing outer diameter, explosive type, temperature rating, explosive weight, cement compressive strength, and gun diameter. Ensemble and neural network models recorded the best accuracy with AdaBoost, Random Forest, Gradient Boosting, and Neural Network recording root mean square errors of 0.059, 0.072, 0.078, and 0.084, respectively. These findings confirm that machine learning has a profoundly better performance compared to traditional empirical models; not only does it provide effective, low-cost, and real-time estimations of the length of perforation penetration, but also it is more adaptable to a wide range of perforation guns and well completion parameters.

ABSTRACT At present, the research on the adsorption of Alkali-Surfactant-Polymer (ASP) is mostly limited to a single scale, making it difficult to establish a direct correlation between the microscopic mechanism and the macroscopic effect. In order to study the ability of a single chemical oil displacement agent and a compound system of multiple oil displacement agents to change the surface properties of rocks, their influence on the effect of enhanced oil recovery was analyzed, and the adsorption behavior of the ternary composite oil displacement system within the formation rocks was investigated in this study. The multi-scale model of ‘core scale experiment-pore scale analysis-molecular scale simulation’ was established, and the correlation between the adsorption capacity of different oil displacement agents and the effect of enhanced oil recovery was studied. The results show that under the synergistic effect of the ternary system, alkali is preferentially adsorbed on the rock surface by virtue of its strong migration ability, which changes the surface properties of the rock, and then the subsequent adsorption of macromolecular polymers and surfactants is significantly inhibited. The multilayer adsorption structure between polymer molecules is formed through hydrogen bonds and van der Waals forces, which has high adsorption stability and thus provides a guarantee for the long-term effect of the oil displacement agent. However, excessive polymer adsorption on the rock surface increase the complexity of its internal space, resulting in increased retention of adsorbed surfactants and affecting oil recovery. Low-permeability cores have a large specific surface area, poor pore-throat connectivity, and a higher adsorption capacity of chemical agents than high-permeability cores. Excessive adsorption exacerbates pore throat blockage, reduces fluid fluidity, and leads to a reduction in recovery rate by approximately 10%. The pre-polymer slug can saturate the adsorption sites on the core surface, reducing the adsorption and retention amounts of subsequent surfactants and alkali. When the concentration reaches 1500 mg/L, the adsorption regulation effect tends to be saturated. This study clarified the competitive adsorption and rock modification mechanisms among the oil displacement agents in the ASP flooding system from a microscopic perspective. The relationship between the adsorption behavior of chemical agents and the recovery rate was characterized in the macroscopic oil displacement effect, providing important theoretical support for optimizing the oil displacement scheme and enhancing the crude oil recovery rate.

ABSTRACT This study focuses on observing the dynamic behavior of gas in a tight sandstone pore through Molecular Dynamics (MD) simulations, which are based on equilibrium configurations obtained from a Grand Canonical Monte Carlo (GCMC) simulation. The adsorption capacity and density profile of CH4 in tight sandstone are calculated under different influencing factors, including pore size, temperature, pressure, and C2H6 content. The results of the research indicate that in micropores (d < 2 nm), CH4 mainly exists in adsorbed state and has almost no free state, while in mesopores (d = 2－50 nm) and macropores (d > 50 nm), the adsorbed and free state of CH4 coexists. The adsorption capacity of CH4 in tight sandstone declines as temperature rises, while it increases with higher pressure, but the increasing trend slows down. Additionally, the isosteric heat of CH4 ranges from 7.03 to 10.66 kJ/mol and rises with increasing pressure within the pressure range used in the simulation. Therefore, gas adsorption behavior is classified as physical adsorption. For the exploration and development of tight sandstone gas, this work may provide an important theoretical basis and guidance.

ABSTRACT This study explains a highly accurate predictive model for the interfacial tension (IFT) between crude oil and nitrogen, an important parameter for nitrogen-based gas injection in petroleum reservoirs based on a dataset of 148 experimental points, of which 90% (133 points) were used for training the models and the remaining 10% (15 points) for testing their performance. We used a gradient boosting decision tree (GBDT) algorithm, which was optimized by applying four different methods: Gaussian Processes Optimization (GPO), Batch Bayesian Optimization (BBO), Evolutionary Strategies (ES), and Bayesian Probability Improvement (BPI). These models were trained in experimental data and evaluated their performance using statistical and graphical analyses. The data proved suitable for model construction. A sensitivity examination revealed that pressure, crude oil API gravity and temperature, all negatively impact IFT, while pressure is the most significant feature. Among the evaluated models, the GBDTBBO demonstrated superior accuracy. It demonstrated the highest R-squared values and lowest error rates, and it correctly predicted the trends of IFT in relation to changes in pressure, temperature, and crude oil API.

ABSTRACT Aiming at the problems of airlock, reduced pumping efficiency and equipment wear caused by the high gas-to-liquid ratio in the process of carbon dioxide oil driving, this paper designs an active suction-assisted lifting short connection. Through indoor experiments and simulations, the flow velocity distribution law inside the short-joint was studied, and it was found that increasing the throat length and the distance between the inlet and the nozzle would reduce the fluid velocity, and the optimal length-to-diameter ratio of the diffusion tube was 1:7. The experiments showed that, when the liquid volume was greater than 0.194 m3/h, the inlet gas was continuous, and the gas-liquid volume was approximately linearly correlated. After applying the gas inlet pressure, the gas-liquid volume increases with the increase of pressure, but the gas-liquid ratio increases and the liquid-carrying rate decreases, which is especially obvious when the pressure exceeds 0.4 MPa. When the gas volume is 0–6.86 L/min, the liquid-carrying rate decreases rapidly and stabilizes after exceeding 6.86 L/min, and the flow pattern in the tube is always segmental plug flow. This design is of great significance for utilizing formation gas energy, coping with wells with high gas-liquid ratio and improving well production.

ABSTRACT The primary objective of this study was to develop an interpretable machine learning framework for accurate prediction of weight on bit (WOB) in horizontal wells. With the advancement of oil and gas technologies, horizontal drilling has become a key method for resource recovery, but drag in ultra-long sections reduces efficiency. Traditional WOB calculation methods rely on mathematical models and physical experiments, which are costly and complex. To address this, eight parameters, including well depth and hook load, were selected as model inputs. An ensemble tree model optimized by Bayesian algorithms was built to capture parameter – WOB relationships. SHapley Additive exPlanations (SHAP) were applied to interpret model outputs and assess parameter importance. Performance was evaluated using R2, MSE, MAE, and RMSE. The best-performing ensemble model was further integrated into a CNN-LSTM to enhance temporal feature learning. This hybrid approach achieved an R2 of 0.96, representing about a 10% improvement over standalone models. The proposed framework thus offers a reliable and interpretable tool for WOB prediction, providing valuable reference for improving drilling efficiency in horizontal well sections.

ABSTRACT To address the production forecasting challenges in intelligent wells caused by the wellbore storage effect triggered by changes in the ICV opening,which lead to sudden production fluctuations,this study,based on fluid mechanics principles and the Bernoulli equation,analyzes the coupling effect between annular pressure and production following changes in the ICV opening.A multi-stage dynamic production prediction model is established, integrating ICV opening,pressure differential, and flow rate.The model demonstrated high predictive accuracy in case validations involving six intelligent wells (A1,A2,A3,Y1,Y2,Y3),with predicted annular pressure and production rates showing excellent agreement with measured data (average annular pressure error < 1.0%,standard deviation < 0.35%;average production error < 2.0%, standard deviation < 0.2%).The study shows that annular pressure is negatively correlated with ICV opening and positively correlated with the ICV value outflow pressure, while production is positively correlated with ICV opening and negatively correlated with the ICV value outflow pressure. The wellbore-reservoir coefficient primarily affects the rate of change in annular pressure and production but has no significant impact on annular pressure and production during stable production periods. The findings reveal the nonlinear evolution of production during the well shut-in (pressure buildup) and well opening (pressure relief) phases, providing theoretical support for the capacity control and fine management of intelligent wells.

ABSTRACT Wellbore instability and leakage are common problems in drilling igneous rock gas reservoirs of the Songliao Basin. It is important to define the safe drilling fluid density window with accuracy. This study takes the igneous reservoir in Block A, Northeast China, as the research target and builds a new prediction framework that links experimental data with well‑logging inversion. Rock mechanics experiments and logging data were used to build a model for parameter inversion and dynamic – static conversion. There is a strong correlation (correlation coefficients above 94%) between acoustic travel time and mechanical properties such as compressive strength, tensile strength, and internal friction coefficient. With this relation, the method allows continuous inversion of UCS, Young’s modulus, and Poisson’s ratio in coreless intervals. A multi‑parameter in‑situ logging model was applied to calculate four pressure profiles: pore, collapse, leakage, and breakdown pressures. In this way, the safe density window was defined more precisely. The results show that the method improves prediction accuracy in fracture‑developed intervals (safe density range 1.19–1.22 g/cm3) and matches field observations. The framework combines experiment and logging inversion and provides a practical and quantitative way to reduce drilling risks in igneous reservoirs.

ABSTRACT In response to the challenges of high-speed fluid erosion faced by the central tube of liner hangers in oil and gas extraction, this paper conducts an in-depth investigation into the effects of different displacement and sand ratios on the erosion behavior of the central tube through erosion simulation analysis and structural safety evaluation. The study reveals a positive correlation between the increase in flow rate and sand ratio and the maximum erosion rate of the central tube. Further structural safety analysis uncovers the stress distribution of the central tube under extreme loads, providing critical insights for enhancing the design safety of liner hangers. This paper proposes targeted design and optimization recommendations, aiming to offer scientific support for the long-term stable operation and safety improvement of ultra-deep well liner hangers.