LAViG-FLOW: Latent Autoregressive Video Generation for Fluid Flow Simulations
This addresses the computational bottleneck for subsurface flow modeling in geological applications, though it is an incremental improvement by applying existing diffusion methods to a specific domain.
The paper tackles the challenge of expensive multiphase fluid flow simulations for applications like CO2 sequestration by proposing LAViG-FLOW, a latent autoregressive video generation diffusion framework that generates saturation and pressure fields orders of magnitude faster than traditional solvers while maintaining consistency.
Modeling and forecasting subsurface multiphase fluid flow fields underpin applications ranging from geological CO2 sequestration (GCS) operations to geothermal production. This is essential for ensuring both operational performance and long-term safety. While high fidelity multiphase simulators are widely used for this purpose, they become prohibitively expensive once many forward runs are required for inversion purposes and quantify uncertainty. To tackle this challenge we propose LAViG-FLOW, a latent autoregressive video generation diffusion framework that explicitly learns the coupled evolution of saturation and pressure fields. Each state variable is compressed by a dedicated 2D autoencoder, and a Video Diffusion Transformer (VDiT) models their coupled distribution across time. We first train the model on a given time horizon to learn their coupled relationship and then fine-tune it autoregressively so it can extrapolate beyond the observed time window. Evaluated on an open-source CO2 sequestration dataset, LAViG-FLOW generates saturation and pressure fields that stay consistent across time while running orders of magnitude faster than traditional numerical solvers.