Multi-objective optimization algorithms are crucial for addressing real-world problems, particularly with regard to optimizing well control parameters, which are often computationally expensive due to their reliance on numerical simulations. Surrogate-assisted models help to reduce this computational burden, but their effectiveness depends on the quality of the surrogates, which can be affected by candidate dimension and noise. This study proposes a novel surrogate-assisted multi-objective optimization framework (MOO-SESA) that combines selective ensemble support-vector regression with NSGA-II. The framework’s uniqueness lies in its adaptive selection of a diverse subset of surrogates, established prior to iteration, to enhance accuracy, robustness, and computational efficiency. To our knowledge, this is the first instance in which selective ensemble techniques with multi-objective optimization have been applied to reservoir well control problems. Through employing an ensemble strategy for improving the quality of the surrogate model, MOO-SESA demonstrated superior well control scenarios and faster convergence compared to traditional surrogate-assisted models when applied to the SPE10 and Egg reservoir models.

Surrogate-assisted management of water-flooding, as a reservoir production optimization strategy, dynamically adjusts development schemes at each production step, leading to improved economic benefits and enhanced recovery. However, many computationally expensive numerical simulation runs are needed to build the surrogate model in most existing surrogate-assisted reservoir production optimization methods. With that in mind, this study proposes an efficient intelligent optimization method based on active learning strategy and surrogate ensemble for multi-objective reservoir production optimization named ALSA-MOPO. In this proposed ALSA-MOPO method, three frequently-used surrogate models, the radial basis function network, Gaussian process regression, and support vector regression are adopted to construct the surrogate ensemble. In addition, an active learning strategy is adopted to reduce the sample of establishing surrogate model and query for the worse and best samples based on the surrogate ensemble using particle swarm optimization which are infilled to the dataset for improving the accuracy and quality of the surrogate ensemble. The ALSA-MOPO method stands out due to its unique use of an active learning strategy to enhance the accuracy of the surrogate model, combined with a surrogate ensemble to improve robustness. Furthermore, two synthetic reservoirs with different scales and one complex fault block reservoir were utilized to test the effectiveness and practicability of the ALSA-MOPO method. The optimization results indicated that the ALSA-MOPO framework outperformed numerical simulation-based methods by approximately 50, 20, and 35 times in the three respective cases.

Relative permeability (RP) curves which provide fundamental insights into porous media flow behavior serve as critical parameters in reservoir engineering and numerical simulation studies. However, obtaining accurate RP curves remains a challenge due to expensive experimental costs, core contamination, measurement errors, and other factors. To address this issue, an innovative approach using deep learning strategy is proposed for the prediction of rock sample RP curves directly from mercury injection capillary pressure (MICP) measurements which include the mercury injection curve, mercury withdrawal curve, and pore size distribution. To capture the distinct characteristics of different rock samples' MICP curves effectively, the Gramian Angular Field (GAF) based graph transformation method is introduced for mapping the curves into richly informative image forms. Subsequently, these 2D images are combined into three-channel red, green, blue (RGB) images and fed into a Convolutional Long Short-Term Memory (ConvLSTM) model within our established self-supervised learning framework. Simultaneously the dependencies and evolutionary sequences among image samples are captured through the limited MICP-RP samples and self-supervised learning framework. After that, a highly generalized RP curve calculation proxy framework based on deep learning called RPCDL is constructed by the autonomously generated nearly infinite training samples. The remarkable performance of the proposed method is verified with the experimental data from rock samples in the X oilfield. When applied to 37 small-sample data spaces for the prediction of 10 test samples, the average relative error is 3.6%, which demonstrates the effectiveness of our approach in mapping MICP experimental results to corresponding RP curves. Moreover, the comparison study against traditional CNN and LSTM illustrated the great performance of the RPCDL method in the prediction of both S_o and S_w lines in oil–water RP curves. To this end, this method offers an intelligent and robust means for efficiently estimating RP curves in various reservoir engineering scenarios without costly experiments.

In the realm of reservoir development, the optimization of well placement constitutes a cornerstone challenge with significant implications that directly determine the recovery rate and economic benefits of oil and gas production. This research proposes a novel approach to optimizing well placement in reservoirs by integrating reservoir numerical simulations with intelligent optimization algorithms. The quintessence of this inquiry revolves around the strategic placing of wells amidst the complex geological fabric of reservoirs, where the objective function landscape often manifests with non-smooth, multimodal characteristics. To address those issues, the Weighted Mapping of Productivity Potential (WMPP) technique, fortified by the Simulated Annealing algorithm to judiciously ascertain specific weighting coefficients for the computation of WMPP across reservoirs is introduced in this study. Furthermore, an emblematic carbonate reservoir model serves to corroborate the adaptability and viability of WMPP for well placement optimization, underscoring its efficacy as a swift, economically viable instrument for the delineation of prospective reservoir zones and the guidance of drilling initiatives. The optimization results show that the well placement scheme guided by WMPP, which required 7 fewer wells than the oil initially in place (OOIP)-based scheme, improved 21.74% oil production over the twenty years production period. This comprehensive workflow proffers invaluable insights and benchmarks for the formulation of well placement strategies, with the proposed methodology, in its apparent simplicity, showcasing remarkable efficiency.