Quantitative Integration of Seismic and Electromagnetics for Enhancing Subsalt Imaging

Abstract

Abstract Thick salt sequences characterized by halite concentrations embedded in evaporitic and clastic layers can form a seal for hydrocarbon accumulations. Prospect generation in this geology involves accurate imaging of the evaporitic package, underlying deep basin structures and the basement. Robust seismic velocity model building in salt geology is a challenge and, at the same time, a mandatory requirement for obtaining accurate seismic depth images of the subsurface. Electromagnetic (EM) methods provide the additional independent structural and rock-physics constraints needed to generate robust velocity models through a numerical integration method called joint inversion. Multiphysics datasets were acquired over three blocks in the study area including wide azimuth seismic, magnetotellurics (MT) and controlled source electromagnetics (CSEM). Wide offset, multi-frequency and multi-component CSEM and marine MT complement each other to map the 3D resistivity distribution from the seabed to several kilometers depth. The resistivity information from EM inversion was utilized to enhance seismic velocity reconstruction though a hierarchical and multiscale joint inversion approach iterating over model dimension, frequencies and offsets. Seismic traveltime residuals and a cross-gradient function, mapping the structural similarity between velocities and a previously derived CSEM/MT resistivity model, are simultaneously minimized through the joint inversion approach to enrich the velocity reconstruction with structural details contained in the resistivity model. As a result, a better velocity description is obtained for the key geological features such as: the top of the layered evaporitic sequence (LES), the flanks of the salt diapers, and of clastic packages within the LES. The new depth-domain seismic images provide enhanced imaging of key geological targets such as the base of salt and the underlying basement structures that are superior to seismic images derived from a seismic-only velocity model building workflow.