Optimizing seismic survey designs and configurations for full waveform inversion: a study using physical modeling data and an industry software package

Christina Schumacher, Kristopher A. Innanen, Daniel O. Trad, Joe Wong

This study examines the development of modeled datasets from a physical modeling facility designed to simulate real-world seismic acquisition surveys and optimize them for Full Waveform Inversion (FWI) applications. The data modeling process was meticulously structured to ensure compatibility with industry-standard seismic processing software, facilitating the implementation of realistic processing workflows.

Two seismic datasets were acquired over a model subsurface structure for comparative analysis in the context of time-lapse monitoring in a simulated marine environment. The Pre-Injection dataset (baseline survey) consists of a four-layer model with a sand mound inside a water tank, representing a typical marine 3D survey environment. The Post-Injection model (monitor survey) retains the same four-layer sand mound configuration, however incorporates gas injection to simulate conditions relevant to carbon capture and storage (CCS).

Both datasets underwent identical processing workflows, enabling a direct comparison of their potential for time-lapse applications and their suitability for FWI applications, including the feasibility of time-lapse FWI comparisons. The results highlight that the controlled conditions of the physical modeling facility offer valuable insights into the applicability of pseudo three-dimensional FWI methodologies on laboratory-acquired seismic datasets. This research contributes to advancing subsurface imaging and monitoring techniques, demonstrating the efficacy of physical models in optimizing acquisition strategies for future FWI applications.