Volume 13, Issue 1
Fast Prediction of Immiscible Two-Phase Displacements in Heterogeneous Porous Media with Convolutional Neural Network

Wei Feng & Haibo Huang

Adv. Appl. Math. Mech., 13 (2021), pp. 140-162.

Published online: 2020-10

Preview Full PDF 300 5996
Export citation
  • Abstract

A convolutional neural network is developed for rapidly predicting multiphase flow in heterogeneous porous media. Some direct numerical methods can acquire accurate results of multiphase flow in porous media. However, once the geometry of the porous media changes, it takes much computational time to perform a new simulation. Here, a deep neural network model in the field of semantic segmentation is developed. It takes the two-dimensional microstructure of heterogeneous porous media as inputs and is able to predict corresponding multiphase flow fields (pressure and saturation fields). Compared to the direct lattice Boltzmann simulations, the inference time on new geometry of porous media can be reduced by several orders of magnitude. Our results show that the machine learning method is a good prediction tool in a wide range of porosity and heterogeneity. Besides, to better understand the inherent process, a visible explanation is presented on what our neural networks have learned.

  • Keywords

Porous media, multiphase flow, convolutional neural network, porosity, sorting.

  • AMS Subject Headings

76S05, 76T99, 68T99

  • Copyright

COPYRIGHT: © Global Science Press

  • Email address
  • BibTex
  • RIS
  • TXT
@Article{AAMM-13-140, author = {Wei Feng , and Haibo Huang , }, title = {Fast Prediction of Immiscible Two-Phase Displacements in Heterogeneous Porous Media with Convolutional Neural Network}, journal = {Advances in Applied Mathematics and Mechanics}, year = {2020}, volume = {13}, number = {1}, pages = {140--162}, abstract = {

A convolutional neural network is developed for rapidly predicting multiphase flow in heterogeneous porous media. Some direct numerical methods can acquire accurate results of multiphase flow in porous media. However, once the geometry of the porous media changes, it takes much computational time to perform a new simulation. Here, a deep neural network model in the field of semantic segmentation is developed. It takes the two-dimensional microstructure of heterogeneous porous media as inputs and is able to predict corresponding multiphase flow fields (pressure and saturation fields). Compared to the direct lattice Boltzmann simulations, the inference time on new geometry of porous media can be reduced by several orders of magnitude. Our results show that the machine learning method is a good prediction tool in a wide range of porosity and heterogeneity. Besides, to better understand the inherent process, a visible explanation is presented on what our neural networks have learned.

}, issn = {2075-1354}, doi = {https://doi.org/10.4208/aamm.OA-2019-0377}, url = {http://global-sci.org/intro/article_detail/aamm/18344.html} }
TY - JOUR T1 - Fast Prediction of Immiscible Two-Phase Displacements in Heterogeneous Porous Media with Convolutional Neural Network AU - Wei Feng , AU - Haibo Huang , JO - Advances in Applied Mathematics and Mechanics VL - 1 SP - 140 EP - 162 PY - 2020 DA - 2020/10 SN - 13 DO - http://doi.org/10.4208/aamm.OA-2019-0377 UR - https://global-sci.org/intro/article_detail/aamm/18344.html KW - Porous media, multiphase flow, convolutional neural network, porosity, sorting. AB -

A convolutional neural network is developed for rapidly predicting multiphase flow in heterogeneous porous media. Some direct numerical methods can acquire accurate results of multiphase flow in porous media. However, once the geometry of the porous media changes, it takes much computational time to perform a new simulation. Here, a deep neural network model in the field of semantic segmentation is developed. It takes the two-dimensional microstructure of heterogeneous porous media as inputs and is able to predict corresponding multiphase flow fields (pressure and saturation fields). Compared to the direct lattice Boltzmann simulations, the inference time on new geometry of porous media can be reduced by several orders of magnitude. Our results show that the machine learning method is a good prediction tool in a wide range of porosity and heterogeneity. Besides, to better understand the inherent process, a visible explanation is presented on what our neural networks have learned.

Wei Feng & Haibo Huang. (2020). Fast Prediction of Immiscible Two-Phase Displacements in Heterogeneous Porous Media with Convolutional Neural Network. Advances in Applied Mathematics and Mechanics. 13 (1). 140-162. doi:10.4208/aamm.OA-2019-0377
Copy to clipboard
The citation has been copied to your clipboard