Constraining Reservoir Uncertainty With Frequent 4D-Seismic Data at Valhall Field

摘要

Permanent ocean-bottom seismic cables were installed in 2003 at Valhall Field, offshore Norway, as part of a Life of Field Seismic (LoFS) program. Six seismic surveys have been acquired from the permanent system through 2005, in addition to two historical to wed-str earner surveys acquired in 1992 and 2002. The various 3D seismic data sets generally show strong time-lapse or 4D effects resulting from primary production.In a previous paper we described a workflow used to integrate 4D seismic data with conventional well data in a Top-Down Reservoir Modeling (TDRM) computer-assisted history match of a portion of the Valhall field. Two seismic attributes were used to help constrain the reservoir model - a two-way time shift from the surface-tow surveys and a sum of negative amplitude difference between two of the LoFS surveys. These two seismic attributes seemed to provide independent spatial information about the reservoir that was not contained in the conventional well measurements. History matching was also conducted on the conventional well data alone, without the 4D constraints. The results indicated that when run in prediction mode, the history matches that included the additional 4D seismic constraints provided a narrower uncertainty range on the resulting forecasts than did the models that only honored the well data.In this paper, we examine some additional issues associated with the potential use of multiple time-lapse seismic attributes from the LoFS program in our ongoing quest to further constrain the reservoir history match. We focus in particular on the considerations for choosing an appropriate set of "target" seismic attribute maps for the LoFS history matching effort, which deals with time-lapse intervals that are much smaller than those associated with typical 4D surveys.