Integrating seismic lithofacies prediction and depositional geometry analysis for reservoir delineation in a North Sea turbidite field

Abstract

Summary In this study we delineate reservoir sands in a North Sea turbidite field by combining lithofacies prediction from pre-stack seismic amplitudes and quantitative analysis of seismic scale depositional geometries. This Late Paleocene field has been problematic because of complex sand distribution and non-reservoir seismic anomalies. Two of the three most recent exploration wells failed to encounter reservoir sands. Our goal is to improve our ability to forecast reservoir sands in this and similar turbidite fields in the North Sea. First we recognize and classify different lithofacies from well-log data, including sandstone, shale, marl, limestone and tuff. Since the Vp/Vs ratio together with impedance better discriminate lithology than does impedance alone, we conduct AVO analysis to predict seismic lithofacies from seismic data. We derive probability density functions (pdfs) for each of the facies in terms of zero offset reflectivity (R(0)) and AVO gradient (G). R(0) and G from inversion of real seismic data is used to predict the most likely facies distribution along selected seismic lines. Subsequently, we analyze and quantify the seismic scale depositional geometries in the area. Where reservoir sands have been identified from seismic interpretation, there is a good correlation between reservoir sand thickness and a thicker, better defined Late Paleocene seismic interval. We take advantage of this correlation and predict thickness of reservoir sands given the Late Paleocene interval thickness. These results are integrated with the sand predictions from the AVO inversion. A blind test is conducted on a well drilled at a location where post-stack seismic amplitudes indicated reservoir sands, but where only shales, tuffs and some carbonates were encountered. Our lithofacies prediction results indicate that the most likely facies at that well location in terms of AVO response is tuff, whereas the geometry analysis indicate only local presence of a thin sand unit around the well. We conclude that the seismic anomaly around the well is caused by an intra-Late Paleocene tuff unit, and this unit likely explains the local thickening of the Late Paleocene interval .