Application of simultaneous inversion (FWI and nonlinear LS-RTM) for improved imaging.

Imaging of complex faulted structures requires the most advanced imaging algorithm that handles multi-pathing, an accurate background velocity model for correct structural positioning and an element of least-squares migration to mitigate limitations of the imaging system (migration and acquisition process). In this paper, we address all these challenges with a single data-driven inversion process that updates both the background velocity model and the reflectivity model simultaneously.

With a seismic data example from the Outer Vøring basin in the Norwegian sea, we demonstrate how the inversion heals the fault shadow zone, improves the structural imaging compared to the underlying Kirchhoff pre-stack depth and reverse time migration results. The inversion provides a better well to seismic tie compared to the alternatives and the de-coupled parameters (velocity and reflectivity) can be used to directly derive reservoir property attributes such as relative density. The seismic data example shows low-density sand layers at the target that correlate with the measured properties at the well. The applied methodology has the potential for both turnaround reduction, quality improvements and prospect de-risking by combining velocity estimation, imaging, and reservoir property estimations into a pure data-driven inversion process.