In order to make sure the sedimentary characteristic and evaluation of Dongying Formation, we quantitatively divide the sequence framework and distribution of sedimentary facies, comprehensive application the wave coefficient curve and maximum entropy spectrum analysis, on the base of sequence boundaries recognize. The result is that, there are 2 long-term and 3 mid-term cycle boundaries in Dongying Formation. We divide the Dongying Formation into 4 mid-term cycles, like MSC1, MSC2, MSC3 and MSC4. The fan delta plains and submarine fans are concentrated on the Xinanzhuang Fault, Baigezhuang Fault and Gaoliu Fault, they change to the fan delta front and shallow lake gradually, from the boundary fault to the central of Nanpu Sag. The evaluation of sedimentary in Dongying Formation is controlled by the union of fault, provenance and base level cycle. The base level is rising quickly in MSC1, the sediment are mainly conglomerate, medium sandstone, fine sandstone in fan delta plains and fan delta front. The base level rising slowly, and attain the maximum in MSC2, the sediment are mainly mudstone in shallow lake and semi-deep lake, it is the main source rocks in study area. The subsidence rate decrease in MSC3 and MSC4, and the base level began to decline, the sediment are mainly siltstone, fine sandstone in fan delta front, they are the main reservoirs and productive series in the study area.

Carbon capture, enhanced oil recovery (EOR)-utilization and storage (CCUS-EOR) is recognized as an effective approach to mitigate greenhouse gas emissions while delivering economic benefits. However, its practical deployment is limited by the absence of advanced deep learning models for petroleum tabular data, the limited adaptability of existing optimization methods, and the lack of comprehensive evaluation for full-cycle CCUS-EOR. Here, we introduce a generalizable framework that integrates mechanism experiments, numerical simulations, and deep learning methods to address these challenges. Three-stage experiments are conducted to clarify microscopic displacement mechanisms and provide key parameters for numerical simulation. Based on field-scale simulations of 20 years of CO₂ water-alternating-gas (WAG) injection followed by 19 years of pure CO₂ storage until 2060, we develop a TabPFN-based meta-learning surrogate model for joint prediction of oil recovery, CO₂ storage, and net present value (NPV), achieving high accuracy (prediction error <2%, R² > 0.97) compared to baseline models. We further apply an improved multi-objective optimization using the Adaptive Crossover and Adaptive Mutation Non-dominated Sorting Genetic Algorithm Ⅱ (ACAM-NSGA-Ⅱ) to obtain optimal Pareto solutions. Compared to baseline cases, the proposed framework significantly enhances CCUS-EOR performance, enhancing oil recovery by 27.05% (from 5.95 × 10⁵ t, 35.17% to 1.05 × 10⁶ t, 62.22%), tripling CO₂ storage capacity (from 1.33 × 10⁶ to 4.45 × 10⁶ t), and improving NPV by 68.0% (from ＄344 million to ＄578 million). The Pareto front is further divided into three different solution regions, thereby elucidating the underlying physical mechanisms associated with each cluster and providing clear operational insights for target-oriented CO₂-WAG design. This study offers a scalable blueprint framework for large-scale engineering design in petroleum engineering, particularly in tabular prediction and multi-objective optimization contexts.

The utilization and storage of CO₂ emissions from oil production and consumption in the upstream oil industry will contribute to sustainable development. CO₂ flooding is the key technology for the upstream oil industry to transition to sustainable development. However, there is a significant challenge in achieving high recovery and storage efficiency in unconventional reservoirs, particularly in underdeveloped countries. Numerous studies have indicated that the limited sweep range caused by premature gas channeling of CO₂ is a crucial bottleneck that hinders the enhancement of recovery, storage efficiency and safety. This review provides a comprehensive summary of the research and technical advancements regarding the front sweep characteristics of CO₂ during migration. It particularly focuses on the characteristics, applicable stages, and research progress of different technologies used for regulating CO₂ flooding sweep. Finally, based on the current application status and development trends, the review offers insights into the future research direction for these technologies. It is concluded that the front migration characteristics of CO₂ play a crucial role in determining the macroscopic sweep range. The focus of future research lies in achieving cross-scale correlation and information coupling of CO₂ migration processes. Currently, the influence weight of permeability, injection speed, pressure and other parameters on the characteristics of ‘fingering-gas channeling’ is still not well clear. There is an urgent need to establish prediction model and early warning mechanism that considers multi-parameters and cross-scale gas channeling degrees, in order to create effective strategies for prevention and control. There are currently three technologies available for sweep regulation: flow field intervention, mobility reduction, and gas channeling plugging. To expand the sweep effectively, it is important to systematically integrate these technologies based on their regulation characteristics and applicable stages. This can be achieved by constructing an intelligent synergistic hierarchical segmented regulation technology known as ‘flow field intervention + mobility regulation + channel plugging chemically’. This work is expected to provide valuable insights for achieving conformance control of CO₂-EOR and safe storage of CO₂.

Research on reservoir rock stress sensitivity has traditionally focused on unary granular structures, neglecting the binary nature of real reservoirs, especially tight reservoirs. Understanding the stress-sensitive behavior and mathematical characterization of binary granular media remains a challenging task. In this study, we conducted online-NMR experiments to investigate the permeability and porosity evolution as well as stress-sensitive control mechanisms in tight sandy conglomerate samples. The results revealed stress sensitivity coefficients between 0.042 and 0.098 and permeability damage rates ranging from 65.6% to 90.9%, with an average pore compression coefficient of 0.0168–0.0208 MPa⁻¹. Pore-scale compression occurred in three stages: filling, compression, and compaction, with matrix pores playing a dominant role in pore compression. The stress sensitivity of binary granular media was found to be influenced by the support structure and particle properties. High stress sensitivity was associated with small fine particle size, high fines content, high uniformity coefficient of particle size, high plastic deformation, and low Young's modulus. Matrix-supported samples exhibited a high irreversible permeability damage rate (average = 74.2%) and stress sensitivity coefficients (average = 0.089), with pore spaces more slit-like. In contrast, grain-supported samples showed low stress sensitivity coefficients (average = 0.021) at high stress stages. Based on the experiments, we developed a mathematical model for stress sensitivity in binary granular media, considering binary granular properties and nested interactions using Hertz contact deformation and Poiseuille theory. By describing the change in activity content of fines under stress, we characterized the non-stationary state of compressive deformation in the binary granular structure and classified the reservoir into three categories. The model was applied for production prediction using actual data from the Mahu reservoir in China, showing that the energy retention rates of support-dominated, fill-dominated, and matrix-controlled reservoirs should be higher than 70.1%, 88%, and 90.2%, respectively.