Scaling in time-lapse seismic processing: does it make a difference?

Ying Zou, Laurence R. Bentley, Laurence R. Lines, Doug Kuervers

Two time-lapse seismic lines, the 1991 survey and 2000 survey in the Pikes Peak heavy oil field have been processed with three scaling methods using the KTI processing package. The three scaling methods are conventional scaling, surface consistent scaling and two mean window scaling. The difference between the sections for the 1991 and 2000 survey was calculated using Pro4D. The three difference sections were obtained for each of the three scaling methods. Comparisons with production activities in this field showed that the conventional scaling method with multiple mean windows above the reservoir and one mean window including the reservoir gives the best result. Strong ground-roll and non-surface-consistent amplitude may cause the surface-consistent scaling to fail. The conventional scaling method may work better than the surface-consistent scaling method for land time-lapse data.