We introduce two algorithms that invert simulated gravity data to 3D subsurface rock/flow properties. The first algorithm is a data-driven, deep learning-based approach, and the second is a post-processing scheme built on top of our first algorithm that combines traditional inversion with DL-based inversion. The target application of these proposed algorithms is the prediction of subsurface CO2 plumes as a complementary tool for monitoring CO2 sequestration deployments. Each proposed algorithm outperforms traditional inversion methods and produces high-resolution, 3D subsurface reconstructions in near real-time. These results indicate that combining 4D surface gravity monitoring with deep learning techniques represents a cost-effective and more environmentally friendly method for monitoring CO2 storage sites.
Authors: Adrian Celaya (Rice University), Yen Sun (TotalEnergies EP Research & Technology US), Bertrand Denel (TotalEnergies), Antony Price (TotalEnergies) and Mauricio Araya-Polo (TotalEnergies EP Research & Technology US)
