Atomized inversion of 4D gravity data for risk management of CO2 leakage

Abstract We propose a new approach for monitoring CO2 storage sites and to assess leakage patterns through inversion of time-lapse gravity data. Our focus is on leakage scenarios involving permeable faults and the formation of secondary plumes at shallow depths. The methodology is based on the concept of Extremely Compact Sources, which provides a subsurface representation in terms of atomized sources. This framework enables the successful determination of plume locations, net mass changes, and CO2 leakage rates. A case study using realistic multiphase simulations demonstrates the effectiveness of the method for reservoir management, including forecasting, operational planning, and safe management of the injection and post-injection phases.