Abstract Technological advances in tools and processing techniques are playing a crucial role in evaluating discoveries of the past. This paper offers a unique workflow to assess the potential of complex turbidite reservoirs which would have otherwise been overlooked using conventional wireline logging tools and traditional petrophysical techniques. In highly heterogenous reservoirs below wireline logging tool resolution, conventional interpretation techniques do not provide an accurate representation of reservoir quality. The primary focus of this paper is to demonstrate how a robust understanding can be established on fine-scale variations in reservoir quality of complex turbidite sandstones by integrating geological and advanced petrophysical techniques. The proposed workflow integrates a neural net approach by utilizing data from cores, petrography, borehole images, and NMR log. Petrography and MICP data were utilized to understand diagenetic control on pore-throat radii and reservoir quality change due to grain properties (size, sorting, maturity), dissolution, and calcite cementation. Core calibrated image logs were used to classify rock types and train the neural net. The outcome is a high-resolution core calibrated rock-type model which has played a critical role in distinguishing sweet spots from poor quality reservoir intervals in these complex turbidites. The model has proven over 80% accurate when blind tested against cores and dynamic data in offset wells. This has subsequently helped in generating high confidence net to gross estimations across the field to substantiate the potential in these turbidite units. This novel methodology has shown extremely promising results in understanding the degree of heterogeneity in complex turbidites. The new technique has allowed identification, location, and scaling of by-passed zones which were overlooked previously using low resolution logging tools and traditional petrophysical processes.
Zaher et al. (Tue,) studied this question.