Time-lapse analysis requires a different emphasis on seismic data processing. Single vintage data processing simply aims at producing the best quality stack volumes or prestack gathers. However, a time-lapse survey aims at obtaining the most meaningful differences between different vintages in the survey (Calvert, 2005). This implies that reflection signal from zones without time-lapse effect should be repeatable and difference between vintages at these zones should be minimized. We propose a time lapse processing flow that requires not only the amplitude preserving simultaneous processing of all vintages as one dataset, we also require a strict QC procedure to ensure each main processing step improves (or at least maintains) the repeatability of its previous step. Generally speaking, amplitude balancing, static time correction, and noise attenuation are the main steps where significant repeatability improvements are expected.
View full article as PDF (1.44 Mb)