A Fluid Flow‐Based Deep Learning (FFDL) Architecture for Subsurface Flow Systems With Application to Geologic CO2 Storage
Abstract Prediction of the spatial‐temporal dynamics of the fluid flow in complex subsurface systems, such as geologic CO2 storage, is typically performed using advanced numerical simulation methods that solve the underlying governing physical equations. However, numerical simulation is computationa...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2025-01-01
|
| Series: | Water Resources Research |
| Subjects: | |
| Online Access: | https://doi.org/10.1029/2024WR037953 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849687851843715072 |
|---|---|
| author | Zhen Qin Yingxiang Liu Fangning Zheng Behnam Jafarpour |
| author_facet | Zhen Qin Yingxiang Liu Fangning Zheng Behnam Jafarpour |
| author_sort | Zhen Qin |
| collection | DOAJ |
| description | Abstract Prediction of the spatial‐temporal dynamics of the fluid flow in complex subsurface systems, such as geologic CO2 storage, is typically performed using advanced numerical simulation methods that solve the underlying governing physical equations. However, numerical simulation is computationally demanding and can limit the implementation of standard field management workflows, such as model calibration and optimization. Standard deep learning models, such as RUNET, have recently been proposed to alleviate the computational burden of physics‐based simulation models. Despite their powerful learning capabilities and computational appeal, deep learning models have important limitations, including lack of interpretability, extensive data needs, weak extrapolation capacity, and physical inconsistency that can affect their adoption in practical applications. We develop a Fluid Flow‐based Deep Learning (FFDL) architecture for spatial‐temporal prediction of important state variables in subsurface flow systems. The new architecture consists of a physics‐based encoder to construct physically meaningful latent variables, and a residual‐based processor to predict the evolution of the state variables. It uses physical operators that serve as nonlinear activation functions and imposes the general structure of the fluid flow equations to facilitate its training with data pertaining to the specific subsurface flow application of interest. A comprehensive investigation of FFDL, based on a field‐scale geologic CO2 storage model, is used to demonstrate the superior performance of FFDL compared to RUNET as a standard deep learning model. The results show that FFDL outperforms RUNET in terms of prediction accuracy, extrapolation power, and training data needs. |
| format | Article |
| id | doaj-art-50463552e30c47eaadd616a826182e55 |
| institution | DOAJ |
| issn | 0043-1397 1944-7973 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Water Resources Research |
| spelling | doaj-art-50463552e30c47eaadd616a826182e552025-08-20T03:22:12ZengWileyWater Resources Research0043-13971944-79732025-01-01611n/an/a10.1029/2024WR037953A Fluid Flow‐Based Deep Learning (FFDL) Architecture for Subsurface Flow Systems With Application to Geologic CO2 StorageZhen Qin0Yingxiang Liu1Fangning Zheng2Behnam Jafarpour3Mork Family Department of Chemical Engineering and Materials Science University of Southern California Los Angeles CA USAMing Hsieh Department of Electrical and Computer Engineering University of Southern California Los Angeles CA USAMork Family Department of Chemical Engineering and Materials Science University of Southern California Los Angeles CA USAMork Family Department of Chemical Engineering and Materials Science University of Southern California Los Angeles CA USAAbstract Prediction of the spatial‐temporal dynamics of the fluid flow in complex subsurface systems, such as geologic CO2 storage, is typically performed using advanced numerical simulation methods that solve the underlying governing physical equations. However, numerical simulation is computationally demanding and can limit the implementation of standard field management workflows, such as model calibration and optimization. Standard deep learning models, such as RUNET, have recently been proposed to alleviate the computational burden of physics‐based simulation models. Despite their powerful learning capabilities and computational appeal, deep learning models have important limitations, including lack of interpretability, extensive data needs, weak extrapolation capacity, and physical inconsistency that can affect their adoption in practical applications. We develop a Fluid Flow‐based Deep Learning (FFDL) architecture for spatial‐temporal prediction of important state variables in subsurface flow systems. The new architecture consists of a physics‐based encoder to construct physically meaningful latent variables, and a residual‐based processor to predict the evolution of the state variables. It uses physical operators that serve as nonlinear activation functions and imposes the general structure of the fluid flow equations to facilitate its training with data pertaining to the specific subsurface flow application of interest. A comprehensive investigation of FFDL, based on a field‐scale geologic CO2 storage model, is used to demonstrate the superior performance of FFDL compared to RUNET as a standard deep learning model. The results show that FFDL outperforms RUNET in terms of prediction accuracy, extrapolation power, and training data needs.https://doi.org/10.1029/2024WR037953physics‐encoded approachgeologic carbon sequestrationdeep learningmultiphase flowextrapolationpredictive modeling |
| spellingShingle | Zhen Qin Yingxiang Liu Fangning Zheng Behnam Jafarpour A Fluid Flow‐Based Deep Learning (FFDL) Architecture for Subsurface Flow Systems With Application to Geologic CO2 Storage Water Resources Research physics‐encoded approach geologic carbon sequestration deep learning multiphase flow extrapolation predictive modeling |
| title | A Fluid Flow‐Based Deep Learning (FFDL) Architecture for Subsurface Flow Systems With Application to Geologic CO2 Storage |
| title_full | A Fluid Flow‐Based Deep Learning (FFDL) Architecture for Subsurface Flow Systems With Application to Geologic CO2 Storage |
| title_fullStr | A Fluid Flow‐Based Deep Learning (FFDL) Architecture for Subsurface Flow Systems With Application to Geologic CO2 Storage |
| title_full_unstemmed | A Fluid Flow‐Based Deep Learning (FFDL) Architecture for Subsurface Flow Systems With Application to Geologic CO2 Storage |
| title_short | A Fluid Flow‐Based Deep Learning (FFDL) Architecture for Subsurface Flow Systems With Application to Geologic CO2 Storage |
| title_sort | fluid flow based deep learning ffdl architecture for subsurface flow systems with application to geologic co2 storage |
| topic | physics‐encoded approach geologic carbon sequestration deep learning multiphase flow extrapolation predictive modeling |
| url | https://doi.org/10.1029/2024WR037953 |
| work_keys_str_mv | AT zhenqin afluidflowbaseddeeplearningffdlarchitectureforsubsurfaceflowsystemswithapplicationtogeologicco2storage AT yingxiangliu afluidflowbaseddeeplearningffdlarchitectureforsubsurfaceflowsystemswithapplicationtogeologicco2storage AT fangningzheng afluidflowbaseddeeplearningffdlarchitectureforsubsurfaceflowsystemswithapplicationtogeologicco2storage AT behnamjafarpour afluidflowbaseddeeplearningffdlarchitectureforsubsurfaceflowsystemswithapplicationtogeologicco2storage AT zhenqin fluidflowbaseddeeplearningffdlarchitectureforsubsurfaceflowsystemswithapplicationtogeologicco2storage AT yingxiangliu fluidflowbaseddeeplearningffdlarchitectureforsubsurfaceflowsystemswithapplicationtogeologicco2storage AT fangningzheng fluidflowbaseddeeplearningffdlarchitectureforsubsurfaceflowsystemswithapplicationtogeologicco2storage AT behnamjafarpour fluidflowbaseddeeplearningffdlarchitectureforsubsurfaceflowsystemswithapplicationtogeologicco2storage |