Optimizing CO2 Storage Monitoring with Enhanced Rock Physics Modeling
| Title | Optimizing CO2 Storage Monitoring with Enhanced Rock Physics Modeling |
| Publication Type | Presentation |
| Year of Publication | 2024 |
| Authors | Abhinav Prakash Gahlot, Felix J. Herrmann |
| Keywords | Amortized Variational Inference, augmentation, Bayesian inference, conditional normalizing flows, data assimilation, deep learning, FWI, Imaging, Inverse problems, ML4SEISMIC, rock physics, RTM, SLIM, Summary Statistics, Uncertainty quantification |
| Abstract | Based on the latest data-assimilation and machine-learning techniques, Digital Twins (DTs) have shown promise for high-fidelity monitoring and control of underground CO2 storage. While the use of these techniques have important advantages, they do rely on certain assumptions. If these assumptions are not met, the DT’s neural networks may no longer infer the state of the CO2 plume (pressure/saturation) accurately. By augmenting the forecast ensemble, we address this issue. |
| URL | https://slim.gatech.edu/Publications/Public/Conferences/ML4SEISMIC/2024/gahlot2024ML4SEISMICosm |
| Citation Key | gahlot2024ML4SEISMICosm |