High resolution velocity as a 4D attribute

Introduction 4D (also known as time-lapse or repeat) seismic has in the past few years emerged as a significant technique for monitoring fluid movement within reservoirs. In recent years, an improved understanding of the petrophysics of the reservoir (e.g. Wang 2001) has enabled geoscientists to establish relationships between observed seismic attribute changes and corresponding rock property changes. Once it was established that such additional information could be extracted from surface seismic data (e.g. Lumley 1995, 2001), changes were made to seismic data processing sequences so as to better preserve relative amplitude changes, especially with regard to pre-stack differences between data vintage subsets. Monitoring of changes as diverse as temperature (as in the case of steam injection: Lumley 1995) and movement of fluid contacts have proved attainable and successful (Jack 1998). Previously, one of us (Jones & Baud 2001) demonstrated the potential of using a dense velocity estimate as a high-resolution tool for visualizing subtle changes not clearly discernable in the seismic amplitude response. (The success of such techniques depends to a large part on the data being correctly pre-stack migrated, so that all the diffraction energy is correctly collapsed.) In this work, we assess the potential of high resolution velocity as an indicator of reservoir change in the context of a 4D study (Jones & Folstad 2002).