Pre-stack estimation of time-lapse seismic velocity changes : an example from the Sleipner CO2-sequestration project

Carbon dioxide (CO2) is being injected into a shallow sand formation at around a 1,000 metre depth at the Sleipner Field located in the North Sea. It is expected that the CO2 injected in the bottom of the formation, will form a plume consisting of CO2 accumulating in thin lenses during migration up through the reservoir. Several studies have been published using stacked seismic data from 1994, 1999, 2001 and 2002. A thorough analysis of post-stack seismic data from the Sleipner CO2-Sequestration Pilot Project was conducted. Interpretation of seismic data is usually done on post-stack data. For a given subsurface reflection point, seismic data are acquired for various incidence angles, typically 40 angles. These 40 seismic signals are stacked together in order to reduce noise. The term pre-stack refers to seismic data prior to this step. For hydrocarbon-related 4-dimensional seismic studies, travel time shift estimations have been used. This paper compared pre-stack and post-stack estimation of average velocity changes based on measured 4-dimensional travel time shifts. It is more practical to compare estimated velocity changes than the actual travel time changes, since the time shifts vary with offset for pre-stack time-lapse seismic analysis. It was concluded that the pre-stack method gives smaller velocity changes when estimated between two key horizons. Therefore, pre-stack travel time analysis in addition to conventional post-stack analysis is recommended. 6 refs., 12 figs