1 September 2024 in Physics of Fluids
This study combines convolutional neural networks, spatial pyramid pooling, and long short-term memory networks (LSTM) with self-attention (SA) mechanisms (abbreviated as CSAL) to address the problem of production dynamics prediction in tight reservoirs during the CO2 water-alternating-gas (CO2-WAG) injection process. By integrating DenseNet and SPP modules, this method effectively captures and processes complex spatial features in tight reservoirs. Concurrently, the LSTM enhanced with SA mechanisms improves the prediction capability of temporal data during the CO2-WAG process. Experimental results demonstrate that the CSAL model performs excellently in both the training and testing phases, achieving a coefficient of determination (R2) exceeding 0.98, significantly enhancing the model's prediction accuracy. Compared to models without attention mechanisms, the CSAL model increases the R2 value in time series prediction by 10%. Furthermore, employing the Ensemble Smoother with Multiple Data Assimilation algorithm, the CSAL model achieves high-precision history matching, significantly reducing the error between predicted values and actual observations. This study validates the application potential and superiority of the CSAL model in the CO2-WAG process in tight reservoirs.
1 October 2024 in Physics of Fluids
This paper proposes a novel meshless numerical simulation method for fractured-vuggy reservoirs. Initially, model discretization involves extracting fracture and vug feature nodes along the contours of the reservoir bodies and in alignment with fracture orientations, tailored to the characteristics of the fractured-vuggy oil deposit. Subsequently, a connection element model is established within the influence domain. Based on the seepage equations of the connection system, a computational method for the parameters of the connection elements (connection transmissibility and connection volume) is defined. In addition, a sequential method is employed to solve for the pressure and saturation, achieving rapid prediction of the reservoir's production dynamics. On this basis, an automatic history matching algorithm is utilized for real-time fitting of oil–water dynamic indicators, inverting the characteristic parameters of the connection element model, and quantitatively characterizing the injection–production connection elements. The research findings indicate that the method is capable of rapidly fitting and predicting the dynamics of oil reservoir production. Furthermore, conceptual model examples have substantiated that it achieves comparable computational efficiency and superior computational accuracy when compared to the traditional finite difference (volume) method under the same nodal conditions. Additionally, the node arrangement for fractures and vugs is comparatively flexible, and it can ensure the integrity of the flow paths with a limited number of nodes to enhance predictive accuracy. Therefore, this method can effectively meet the rapid prediction requirements for fractured-vuggy reservoirs and also provides a novel meshless computational approach for the numerical simulation of such reservoirs.
1 February 2025 in Physics of Fluids
Hydraulic fracturing is a critical technology for the efficient development of unconventional reservoirs. It involves multi-physics coupling, making its numerical simulation an extremely challenging task. In this paper, an efficient hydraulic fracturing simulator is developed by integrating a simplified proppant transport model into the multi-fracture fracturing framework. In the fracture propagation model, the finite volume method is used to solve fluid flow, while the displacement discontinuity method is employed rock deformation. Proppant transport is realized by the simplified Eulerian–Eulerian method, which only needs to consider the one-dimensional fluid flow. The high-precision weighted essentially non-oscillatory scheme is used to solve the nonlinear proppant transport equation. Compared with the Eulerian–Lagrangian method, the proposed method enhances simulation efficiency by 2–3 orders of magnitude. The numerical model is verified through physical experiments and fracture analytical solutions. The fracture propagation and proppant transport models are implemented via a unidirectional coupling strategy. First, the fully coupled model calculates the fracture morphology, followed by the computation of proppant distribution at each time step. Four cases were established to analyze the effects of fracturing fluid injection time, injection rate, fracture cluster spacing, and the number of fracture clusters on the industrial-scale proppant distribution in fractures. This method can be applied to large-scale fracturing simulation, optimization of fracturing parameters and other fields requiring high computational efficiency.
1 July 2025 in Physics of Fluids
In order to improve the field gas recovery (FGR) and CO2 sequestration ratio (CSR) for the edge/bottom-water gas reservoirs, a hybrid framework is designed to combine Long Short-Term Memory (LSTM) with Non-dominated Sorting Genetic Algorithm II (NSGA-II) to obtain the optimal CO2 injection and production scheme at different injection and production rates. In this framework, the Latin hypercube sampling method is used to generate the training and testing samples within a given range of injection and production rates, and the corresponding results are obtained by a calling numerical simulator. The proxy model is obtained by using the samples generated from LSTM training to replace the numerical simulator and improve the simulation efficiency. NSGA-II is utilized to determine the Pareto front of the optimal outcomes to maximize FGR and CSR. The robustness of the framework is verified with a typical edge/bottom water-drive gas reservoir. CO2 sequestration process in this gas reservoir after 100 years is simulated based on the optimal solution. The results show that the R2 of the LSTM proxy model for FGR and CSR prediction is greater than 0.99, which indicates that it can replace the numerical simulator accurately. Compared with the basic scheme without optimization, FGR is increased by 6.09% and CSR is increased by 3.91%. Under the optimal scheme, the CO2 distribution for long-period storage is predicted, and the results show CO2 will eventually migrate to the bottom of the reservoir under the influence of gravity, which is contrary to CO2 sequestration in oil reservoirs. It can provide theoretical guidance for CO2 injection for enhanced gas recovery and sequestration in edge/bottom water-drive gas reservoirs.
1 July 2025 in Physics of Fluids
This paper has developed a novel meshless numerical simulation method for fractured reservoirs with stress-sensitive effects. This method breaks through the limitations of traditional grid methods by employing meshless node technology to discretize the reservoir computational domain, thereby more accurately simulating the complex fracture network geometry. In the simulation of two-phase oil–water flow, the method uses the generalized finite difference method to discretize the pressure equation with stress-sensitive and calculate the transmissibility between matrix nodes, fracture nodes, and the control volume of nodes. The pressure and saturation equations are solved using a fully implicit method. This approach can directly apply nonlinear solvers in reservoir simulation, simplifying the process of developing a universal simulator. Through case studies, we found that an increase in the stress-sensitive coefficient leads to a decrease in the micropores of the rock framework as pressure changes, resulting in a decrease in formation permeability, and consequently reducing daily oil production and cumulative oil production. This method demonstrates the flexibility of meshless nodes in simulating irregular reservoir boundaries and fracture geometries, as well as their high-precision computational capabilities.