Improving Reservoir Modeling through Integration of Seismic Data in Eocene Turbidites for West Brae Field, Central North Sea, United Kingdom

Abstract

The understanding of the reservoir in the West Brae field in the North Sea has improved because of the incorporation of reprocessed seismic data into reservoir characterization and modeling. The field was discovered in 1975 with initial production in 1997 from two early Eocene turbidite sands in the Balder and Sele formations (Flugga sand member). Both turbidite sands are of good quality, with an average porosity of 30%, an average net-to-gross ratio of 85%, and permeability up to 7500 md. The field produces mainly black oil (22 API) with a dry gas cap and has two distinct oil-water contacts. A high-quality four-dimensional seismic data set was acquired in 2007, which was parallel processed with the 1993 baseline seismic data. These new data prompted a rebuild of the reservoir model to assess the potential for bypassed hydrocarbons. The West Brae model is the result of a multidisciplinary reservoir characterization study that has incorporated attributes from the 1993 reprocessed seismic survey into the static geologic model. The key to incorporating the three-dimensional seismic data into the reservoir model was an elastic simultaneous inversion attribute that clearly identified the good-quality reservoir sands. The integration of the new seismic data into the West Brae reservoir model has improved reservoir understanding by (1) providing a stratigraphic framework for the geomodel, (2) refining the depositional model, and (3) creating more consistency in the geostatistical distribution of reservoir properties in the model. Colocated cokriging of the well data and a “soft” seismic attribute volume (Poisson impedance) has helped reduce the uncertainty of sand distribution and the prediction of flow potential in the West Brae field. This case study has shown that using a multidisciplinary team (geophysics, geology, petrophysics, and reservoir engineering) and an integrated data set significantly reduces the uncertainty for a reservoir characterization study.