An Integrated Ensemble Based Prediction Approach to Quantify and Propagate Uncertainties of Naturally Fractured Carbonate Reservoirs

Abstract

Naturally fractured reservoirs are holding most of the hydrocarbon proven reserves worldwide. Field development planning and optimization of fractured reservoirs face significant challenges due to fracture system complexity, high reservoir heterogeneity and multiple recovery mechanisms. Fractured systems generally have multiple sets of fractures in different orientations, multiple apertures ranging from tiny fissures to large fracture conduits resulting in complicated fluid flow. Reservoir modeling plays a key role in field optimization by examining reservoir recovery under various scenarios. Traditionally, modeling fractured reservoirs is a deterministic process using a single or few reservoirs model (high, mid, and low cases) without proper integration of seismic-to-simulation reservoir model uncertainties. The full span of possible reservoir models and prediction uncertainty isn't captured nor propagated to economics. This paper presents an integrated forward modeling workflow using a Synthetic 3D model data mimicking a real North Sea field, where static and dynamic models are tightly connected to integrate the impact of uncertainties at different modeling stages (horizon uncertainties, fault uncertainties, petrophysical uncertainties, discrete fracture network (DFN) uncertainties to dynamic simulation).