Open Access iconOpen Access

ARTICLE

crossmark

Prediction of Porous Media Fluid Flow with Spatial Heterogeneity Using Criss-Cross Physics-Informed Convolutional Neural Networks

Jiangxia Han1,2, Liang Xue1,2,*, Ying Jia3, Mpoki Sam Mwasamwasa1,2, Felix Nanguka4, Charles Sangweni5, Hailong Liu3, Qian Li3

1 State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum, Beijing, 102249, China
2 Department of Oil-Gas Field Development Engineering, College of Petroleum Engineering, China University of Petroleum, Beijing, 102249, China
3 Exploration Production Research Institute, Sinopec, Beijing, 102249, China
4 Ministry of Energy, Tanzania Petroleum Development Corporation, Tower, Dar es Salaam, 2774, Tanzania
5 Ministry of Energy Building, Petroleum Upstream Regulatory Authority in Tanzania (PURA), Dar es Salaam, 11439, Tanzania

* Corresponding Author: Liang Xue. Email: email

(This article belongs to the Special Issue: Modeling of Fluids Flow in Unconventional Reservoirs)

Computer Modeling in Engineering & Sciences 2024, 138(2), 1323-1340. https://doi.org/10.32604/cmes.2023.031093

Abstract

Recent advances in deep neural networks have shed new light on physics, engineering, and scientific computing. Reconciling the data-centered viewpoint with physical simulation is one of the research hotspots. The physics-informed neural network (PINN) is currently the most general framework, which is more popular due to the convenience of constructing NNs and excellent generalization ability. The automatic differentiation (AD)-based PINN model is suitable for the homogeneous scientific problem; however, it is unclear how AD can enforce flux continuity across boundaries between cells of different properties where spatial heterogeneity is represented by grid cells with different physical properties. In this work, we propose a criss-cross physics-informed convolutional neural network (CC-PINN) learning architecture, aiming to learn the solution of parametric PDEs with spatial heterogeneity of physical properties. To achieve the seamless enforcement of flux continuity and integration of physical meaning into CNN, a predefined 2D convolutional layer is proposed to accurately express transmissibility between adjacent cells. The efficacy of the proposed method was evaluated through predictions of several petroleum reservoir problems with spatial heterogeneity and compared against state-of-the-art (PINN) through numerical analysis as a benchmark, which demonstrated the superiority of the proposed method over the PINN.

Keywords


Cite This Article

APA Style
Han, J., Xue, L., Jia, Y., Mwasamwasa, M.S., Nanguka, F. et al. (2024). Prediction of porous media fluid flow with spatial heterogeneity using criss-cross physics-informed convolutional neural networks. Computer Modeling in Engineering & Sciences, 138(2), 1323-1340. https://doi.org/10.32604/cmes.2023.031093
Vancouver Style
Han J, Xue L, Jia Y, Mwasamwasa MS, Nanguka F, Sangweni C, et al. Prediction of porous media fluid flow with spatial heterogeneity using criss-cross physics-informed convolutional neural networks. Comput Model Eng Sci. 2024;138(2):1323-1340 https://doi.org/10.32604/cmes.2023.031093
IEEE Style
J. Han et al., "Prediction of Porous Media Fluid Flow with Spatial Heterogeneity Using Criss-Cross Physics-Informed Convolutional Neural Networks," Comput. Model. Eng. Sci., vol. 138, no. 2, pp. 1323-1340. 2024. https://doi.org/10.32604/cmes.2023.031093



cc This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
  • 411

    View

  • 489

    Download

  • 1

    Like

Share Link