This paper presents the challenges of constructing adequate static and flow models of a field from disparate data sets, and the workflow methodologies adopted for integrating, reconciling, and extrapolating available data from multiple disciplines. This study uses a data set of about 300 wells penetrating a 2500-meter thick section of fractured and vuggy carbonates. A 30-year production history was used to calibrate the model; the main recovery mechanism of the complex is rock and fluid expansion, combined with a weak aquifer.

A reservoir simulation study was constructed to investigate the field behavior under natural depletion, water injection and gas injection, and to provide technical foundation supporting investments in future development projects. The static model comprises of a three-dimensional structural model, a matrix property model, and a fracture property model. There was no 3-D seismic. The complex system of normal and reverse faults dividing the structure into 70 fault blocks was created from well picks. The matrix property model was derived from limited routine and special cores, and vug analysis. The fracture property model was obtained by integrating fracture indicator logs, including DSI, image logs, conventional logs of different vintages, with fault orientation from structural conditioning, core analysis, PTA, well production and static pressure histories.

The complex structure, long production history and the various recovery mechanisms posed additional challenges in constructing and calibrating the flow models to historical performances. The fluid properties for the thick reservoir sections were derived using limited available fluid samples, and verified by history matching reservoir pressure and gas-oil ratio, which declined over time until secondary gas cap encroachment. The different vintage and accuracies of the pressure and rate measurement devices were also considered in evaluating the history match quality.

You can access this article if you purchase or spend a download.