Coda-wave interferometry and time-lapse seismic reservoir monitoring

Abstract

Coda-wave interferometry employs the sensitivity of multiply scattered waves to detect minute changes of the propagation velocity. In most applications the underlying assumption is that the velocity changes take place approximately uniformly in a large region. Time-lapse seismic reservoir monitoring, on the other hand, aims to infer local time-lapse changes from seismic reflection measurements at the surface. Hence, coda-wave interferometry, applied directly to time-lapse seismic data for reservoir monitoring, cannot infer such local changes in a reservoir. Recently, we developed a Marchenko-based methodology to isolate the response of a target zone from the full reflection response. We apply this methodology to Troll-field data and successfully predict the internal multiples of the target zone. This confirms that in principle the Marchenko-based isolation method can be used to prepare time-lapse seismic data for coda-wave interferometry, which in turn can be used to infer time-lapse changes in a reservoir caused by production or storage of resources in the subsurface.