Shuyang Liu

Summary: Low-permeability reservoirs are critical frontiers for oil and gas displacement, where CO2 water-alternating-gas (CO2-WAG) injection plays a pivotal role in both enhanced oil recovery (EOR) and carbon capture, utilization, and storage (CCUS). However, development performance is highly sensitive to injection parameters, and conventional optimization methods face significant challenges in efficiently addressing global optimization problems involving multiple parameters—especially the time-varying characteristics of the WAG ratio. To address this challenge, this study proposes an intelligent optimization methodology integrating machine learning (ML) and multiobjective optimization. First, a representative low-permeability reservoir model was established using the Computer Modelling Group (CMG) simulation software, and parameter sensitivity analysis of WAG cycle, injection rate, injection sequence, and WAG ratio identified the WAG ratio as the only parameter with pronounced stage-dependent reversals. Second, the 20-year production period was divided into five distinct stages, with the WAG ratio of each stage treated as a decision variable. A data set of 3,125 samples was generated via full factorial design. Various ML algorithms were trained and compared, and extreme gradient boosting (XGBoost) was ultimately selected to construct high-precision surrogate models for cumulative oil production and CO2 storage, both achieving R2 values exceeding 0.99. Finally, the Nondominated Sorting Genetic Algorithm II (NSGA-II) algorithm was used to perform dual-objective synergistic optimization of continuous WAG ratios across the five stages. The results show that the obtained Pareto optimal frontier clearly illustrates the trade-off between oil production and carbon storage. The optimal WAG ratio sequence exhibits a dynamic evolution pattern: A high WAG ratio is preferred in the early stage to boost reservoir pressure, transitioning to a low ratio in the intermediate stage to facilitate miscible displacement, and adjusting to a moderate ratio in the late stage to maintain efficiency. Compared with fixed WAG ratio schemes, this time-varying strategy significantly enhances synergistic benefits. This research provides an efficient and reliable new approach for the dynamic optimization design of CCUS projects in low-permeability reservoirs. Furthermore, the optimized time-varying WAG strategy derived from the conceptual model is validated against a field-derived heterogeneous model of the Daqing Oil Field. The consistent performance superiority confirms the rationality and engineering applicability of the proposed framework, offering a direct reference for field-scale CCUS project design.

The in situ conversion process (ICP) of oil shale is governed by tightly coupled thermal–hydraulic–mechanical–chemical (THMC) interactions, yet the evolution of rock-failure proximity within a fully coupled framework remains poorly quantified. This study develops a THMC coupled framework for oil shale ICP that integrates multiphase, multicomponent flow, heat transfer, linear thermo-poroelastic deformation, pyrolysis kinetics, and stress- and reaction-driven porosity–permeability evolution. Tensile and shear failure factors, derived from the tensile-strength and Mohr–Coulomb criteria, are incorporated to quantify the proximity to mechanically induced failure. The coupled spatiotemporal evolutions of temperature, pressure, kerogen decomposition, and both failure factors are systematically analyzed through a sensitivity study of geological and operational parameters. The results demonstrate that appreciable reductions in the failure factors precede the rapid-production stage, indicating that the formation approaches the mechanical failure condition before peak hydrocarbon output. The low-failure-factor zones are localized in the high-temperature, high-pressure region on the outer flank of the heater wells─jointly controlled by the temperature gradient, pore pressure, and effective stress─rather than coinciding with the locations of maximum temperature or kerogen consumption. Horizontal permeability strongly affects cumulative oil production and failure-factor evolution: under low permeability, impeded fluid expulsion promotes pore-pressure accumulation and extensive failure-activation zones. In contrast, cumulative gas production is mainly influenced by kerogen concentration and porosity. Heating temperature and thermal conductivity regulate the gas–oil ratio through secondary cracking, whereas kerogen concentration and porosity govern the overall product scale. This study provides quantitative bases and practical guidance for field-scale ICP development.

Shale oil reservoirs possess abundant resources yet face low primary recovery (<10%). CO2 huff-n-puff has attracted significant attention as a promising enhanced oil recovery (EOR) method for these reservoirs. However, based solely on experimental investigations, the underlying EOR mechanisms remain poorly understood. To address these challenges, we propose an integrated method combining experiment and numerical modeling to distinguish and quantify the mechanisms during CO2 huff-n-puff. Experimentally, we employed nuclear magnetic resonance to monitor real-time fluid migration within shale cores. Analysis of dynamic T2 spectra and one-dimensional frequency-encoded signals enabled calculation of real-time recovery rates and effective distances. Numerically, a core-scale compositional simulation model was built and calibrated to match experimental data. Experimental results show that CO2 effective distance is primarily governed by diffusion and limited to the centimeter scale, resulting in low oil recovery. The effective distance increases significantly with soaking time but is unaffected by injection pressure, indicating that pressure alone is insufficient to mobilize deeper oil. The calibrated model quantified the contributions of four mechanisms: oil swelling, diffusion, dissolved gas drive, and elastic energy release. Dissolved gas drive was identified as dominant, accounting for over 40% of total oil recovery. Parameter variations showed that higher injection pressure enhances diffusion of oil molecules into the CO2 phase, improving recovery via oil-molecule diffusion, while prolonged soaking time facilitates CO2 diffusion into the matrix, boosting recovery through oil swelling and dissolved gas drive. These findings provide mechanistic insights for optimizing CO2 huff-n-puff strategies and improving shale oil recovery efficiency.

Summary: Numerous numerical modeling studies have been conducted to investigate carbon dioxide (CO2) leakage risk through various mechanisms, such as deep saline aquifer sequestration. As one of the major leakage pathways, the presence of natural fractures and their potential reactivation or opening induced by injection pose a considerable challenge to long-term storage safety. Explicitly representing fracture networks in simulation grids can substantially increase computational cost, particularly when hydrodynamic, mechanical, and geochemical processes are fully coupled. In this work, we propose a novel geomechanics-coupled embedded discrete fracture model (GeoEDFM) that couples mechanical and chemical dynamics and analyzes the CO2 migration in brine formation containing natural fractures. The fracture opening processes are captured using tensile and shear failure criteria. Benchmark tests demonstrate that the proposed approach reduces computational cost by more than 30 times compared with COMSOL® without compromising solution accuracy. Further numerical experiments are carried out to investigate the fracture opening processes and CO2 leakage through both the brine formation and the caprock. Results indicate that fracture permeability evolves nonmonotonically due to coupled flow-stress effects, characterized by pressure-induced fracture opening during injection and progressive closure after CO2 enters the fracture system and pressure dissipates, eventually stabilizing at a lower residual level. Large-scale CO2 leakage mainly occurs when fractures penetrate both the storage aquifer and the barrier layer. Leakage risk increases with fracture dip angle, fracture conductivity (i.e., zero-stress permeability), and fracture density but decreases with higher initial in-situ stress. Furthermore, our simulations demonstrate that beyond certain site-specific critical limits of fracture aperture and density, the coupled hydromechanical opening of fractures significantly accelerates CO2 migration, causing cumulative leakage to rapidly exceed commonly accepted safety margins.