A data-driven physics-informed deep learning approach for estimating sub-core permeability from coreflooding saturation measurements

Keyphrases

• Deep Learning 100%
• Core Permeability 100%
• Core Flooding 100%
• Physics-informed Neural Networks 100%
• Physics-informed Deep Learning 100%
• Data-driven Physics 100%
• Deep Neural Network 83%
• Neural Network System 50%
• Numerical Simulation 33%
• Inverse Problem 33%
• Misfit 33%
• Limited Information 16%
• Core Scale 16%
• Surrogate Model 16%
• 3D Data 16%
• Contaminated Data 16%
• Specific Scenario 16%
• Scaling Properties 16%
• Multiphase Flow 16%
• Output Parameters 16%
• Forward Problem 16%
• Enhanced Oil Recovery 16%
• Capillary Heterogeneity 16%
• Permeability Structure 16%
• Core Level 16%
• Saturation Profile 16%
• Core Saturation 16%
• Coupling Approach 16%
• Fast Prediction 16%
• CO2 Sequestration 16%
• Two-phase Flow Problem 16%
• Stochastic Solutions 16%
• Heterogeneous Permeability 16%
• Gas Recovery 16%
• Contaminant Treatment 16%

Chemical Engineering

• Deep Learning Method 100%
• Physics Informed Neural Network 100%
• Deep Neural Network 83%