Detection of time-lapse changes using 4D full-waveform inversion at King Field, Gulf of Mexico, USA

Production-related saturation and/or pressure changes in hydrocarbon reservoirs are commonly investigated using 4D seismic acquisition and processing. A critical requirement for conventional 4D processing is that both base and monitor surveys have high repeatability. However, as acquisition techniques evolve over time, different types of surveys are often acquired over the same fields. In this study, we test a novel 4D FWI approach to improve the 4D signal as well as overall imaging. We compare a conventionally 4D-processed Narrow Azimuth (NAZ) base and monitor pair with the same surveys after running 4D FWI. This method utilizes the full wavefield to build kinematically accurate velocity models from different vintages with minimal signal processing. Joint inversion of both baseline and monitor models minimizes acquisition-related differences between the surveys. We focus on King field in the Gulf of Mexico, USA, where business decisions regarding potential infill well locations as well as future field development scenarios benefit from knowing bypassed zones, even years after seismic survey acquisition. High-frequency joint 4D FWI produces very similar results compared to legacy conventional 4D processing. However, 4D FWI renders better stratigraphic resolution, FWI-derived reflectivity (FDR) volumes improve imaging of events near steep salt flanks, and significantly reduces processing times. Given this successful proof of concept, future work will investigate whether 4D FWI is also a viable option for OBN to streamer 4D processing.