Reservoir Monitoring Over a Producing Norwegian Oil Field

Abstract

Abstract A pilot study over Norsk Hydro's Oseberg field was conducted to confirm that it is possible to monitor hydrocarbon fluid movement in this producing reservoir using repeat 3D seismic data and reservoir simulation. Two 3D surveys were collected about 16 months apart using the same vessel and equipment and processed using identical parameters. It was important to maneuver the monitor survey in 3D space to the base survey to allow for an optimum comparison. A detailed interpretation of both surveys were made and even intra-Brent events showed a high similarity between interpretations. The stratigraphic parameters of the geologic model and the related petrophysical parameters were used to build an acoustic model of the reservoir. Detail was added by seismic inversion such that the refined geologic model best fatted both the initial geologic model and the seismic data. The refined model was history matched with a reservoir simulator to estimate fluid saturations at dates corresponding to the two seismic surveys, and these saturations were used for comparison with the observed seismic differences. The comparison showed a seismic anomaly between the two 3D surveys at a position where the Gas Oil Contact was expected to move. This had spatial continuity and extent, of a similar shape to the simulated saturation changes. The seismic data provided evidence of the configurations of fluid fronts between wells, and not directly observable from well data. Summary A pilot reservoir monitoring study was conducted in the North Sea over Norsk Hydro's Oseberg field. This study established that it is possible to witness hydrocarbon fluid movement in this producing reservoir using repeated 3D seismic surveys. In 1989 a 3D survey was collected over part of the Oseberg Field. Sixteen months later the survey was repeated using the same field equipment. The two data sets were processed in an identical manner. To allow a comparison of the two surveys by differencing, the second data set was repositioned using the first data set to account for variations in such factors as source position, cable feather and tides. Detailed interpretations of the two 3D data volumes, which included intra-Brent events, showed a high degree of similarity. An initial acoustic impedance model of the reservoir was built incorporating geologic data, related petrophysical parameters and reservoir fluids. The model was updated using seismic inversion methods to match the seismic data collected during the 1989 survey. Reservoir simulations of the refined model provided the expected fluid saturation changes between the two survey dates. The 3D seismic difference volume showed a seismic anomaly where movement of the Gas Oil Contact was expected. This anomaly had a spatial continuity, and extent similar to that predicted from ft simulated saturation changes. The seismic data provided evidence of the shape of the fluid fronts between wells, which cannot be directly observed on non-seismic data. THE SEISMIC DATA Two 3D seismic surveys were recorded using essentially the same equipment, separated by sufficient calendar time for saturation changes in the reservoir to be observable using seismic methods. The Base Survey was acquired using the m.v. Polar Princess during 13 December 1989 – 27 January 1990. The Monitor Survey was acquired using the same vessel during 29 April 1991 – 9 May 1991. Thus the two surveys were separated by 16 months production. Each measured approximately 12.5 km long by 2.5 km wide as shown by the pilot area of Figure 1. P. 119^