Title: Development of lean, efficient, and fast physics-framed deep-learning-based proxy models for subsurface carbon storage

Journal Article · · International Journal of Greenhouse Gas Control
 [1];  [1];  [1]
  1. Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
+ Show Author Affiliations

In this work, we present deep-learning-based surrogate models for CCUS developed with four different algorithms and a physics-framed two-phase flow problem involving displacement of water by CO2. The deep-learning models were trained using 3D datasets describing the pressure plume, CO2 saturation plume, and water extraction rate generated by numerical simulation. The hyperparameters defining the architecture of the neural networks were optimized to determine the slimmest network size and training parameters that give the most efficient performance at the least training cost. To develop a robust model that closely mimics the governing physical laws, the discretized form of the two-phase fluid transport equation was used to formulate the supervised deep-learning task. The algorithms investigated in this study predicted the data to above 95% accuracy, with the multi-layer perceptron model demonstrating the best performance by balancing training speed, prediction time, and prediction accuracy with lean network capacity. Furthermore, the surrogate models simultaneously predict reservoir pressure and CO2 saturation in every grid block, including the surface well extraction rate and bottomhole pressure, at all simulation times for a given static model realization in just a few seconds on a standard desktop computer. A key outcome of this study is that limits can be placed on network design parameters to avoid over designing neural networks, with associated efficiencies in training and prediction times. This is very useful because large volumes of data may be generated in CCUS projects and over-design of neural network architectures imposes penalties that are antithetical to the goal of near-real time forecasting.

Research Organization:
Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
Sponsoring Organization:
USDOE Office of Fossil Energy (FE)
Grant/Contract Number:
AC02-05CH11231
OSTI ID:
1884778
Journal Information:
International Journal of Greenhouse Gas Control, Journal Name: International Journal of Greenhouse Gas Control Vol. 114; ISSN 1750-5836
Publisher:
ElsevierCopyright Statement
Country of Publication:
United States
Language:
English

References (24)

A data-driven reservoir simulation for natural gas reservoirs journal March 2021
A benchmark study on problems related to CO2 storage in geologic formations: Summary and discussion of the results journal July 2009
Meeting CCS communication challenges head-on: Integrating communications, planning, risk assessment, and project management journal January 2011
Building Statistical Proxy Models for CO2 Geologic Sequestration journal January 2014
A Global Big Data Assessment of Public Attitudes Towards CCS Through the Media journal January 2014
Code intercomparison builds confidence in numerical simulation models for geologic disposal of CO2 journal July 2004
A data-driven CO2 leakage detection using seismic data and spatial–temporal densely connected convolutional neural networks journal November 2019
Application of an artificial neural network in predicting the effectiveness of trapping mechanisms on CO2 sequestration in saline aquifers journal July 2020
Towards a predictor for CO2 plume migration using deep neural networks journal February 2021
Bayesian deep convolutional encoder–decoder networks for surrogate modeling and uncertainty quantification journal August 2018
A deep-learning-based surrogate model for data assimilation in dynamic subsurface flow problems journal July 2020
A deep learning approach to anomaly detection in geological carbon sequestration sites using pressure measurements journal June 2019
Data-driven proxy at hydraulic fracture cluster level: A technique for efficient CO2- enhanced gas recovery and storage assessment in shale reservoir journal November 2015
Machine learning for surveillance of fluid leakage from reservoir using only injection rates and bottomhole pressures journal September 2019
Development of an adaptive surrogate model for production optimization journal September 2015
Deep-learning-based surrogate model for reservoir simulation with time-varying well controls journal September 2020
Deep Convolutional Encoder‐Decoder Networks for Uncertainty Quantification of Dynamic Multiphase Flow in Heterogeneous Media journal January 2019
Predicting CO 2 Plume Migration in Heterogeneous Formations Using Conditional Deep Convolutional Generative Adversarial Network journal July 2019
A New Approach for the Development of Fast-analysis Proxies for Petroleum Reservoir Simulation journal July 2012
Efficient Processing of Deep Neural Networks: A Tutorial and Survey journal December 2017
Mean squared error: Love it or leave it? A new look at Signal Fidelity Measures journal January 2009
Image Quality Assessment: From Error Visibility to Structural Similarity journal April 2004
Carbon Capture, Utilization, and Storage: An Update journal December 2018
Simulating the Behavior of Reservoirs with Convolutional and Recurrent Neural Networks journal May 2020

Similar Records

Related Subjects

58 GEOSCIENCES
Carbon sequestration
Carbon storage
Deep-learning
Fast proxy model
Machine-learning
Physics-guided