Learning Pore-scale Multi-phase Flow from Experimental Data with Graph Neural Network
This work addresses a domain-specific problem for researchers in climate change mitigation technologies, offering an incremental improvement over existing numerical models.
The study tackled the challenge of accurately modeling pore-scale multiphase fluid flow in porous media, which is critical for climate technologies like CO2 storage, by developing a graph neural network-based method that learns directly from experimental data, achieving computational efficiency while capturing complex physics.
Understanding the process of multiphase fluid flow through porous media is crucial for many climate change mitigation technologies, including CO$_2$ geological storage, hydrogen storage, and fuel cells. However, current numerical models are often incapable of accurately capturing the complex pore-scale physics observed in experiments. In this study, we address this challenge using a graph neural network-based approach and directly learn pore-scale fluid flow using micro-CT experimental data. We propose a Long-Short-Edge MeshGraphNet (LSE-MGN) that predicts the state of each node in the pore space at each time step. During inference, given an initial state, the model can autoregressively predict the evolution of the multiphase flow process over time. This approach successfully captures the physics from the high-resolution experimental data while maintaining computational efficiency, providing a promising direction for accurate and efficient pore-scale modeling of complex multiphase fluid flow dynamics.