Geologic Modelling of External and Internal Reservoir Architecture of Fluvial Depositional Systems

Abstract

An integrated study of a petroleum reservoir was conducted to delineate the facies and sequence-stratigraphic architectures, and structural framework for generation of 3D geologic models for use in improved estimates of reserves, flow-simulation studies, and field-development planning. The study incorporated information from 3D seismic interpretations, well-log correlations, facies and petrophysical analyses of cored intervals, and interpretations derived from outcrop exposures of the reservoir interval.

The reservoir intervals are interpreted to represent fluvial depositional systems that vary systematically along an up-dip to down-dip transect. Proximal (up-dip) fluvial-facies belts are interpreted as lowstand fluvial sandstones characterised by amalgamated channel complexes that form widespread sheets. Medial fluvial-facies belts are interpreted as lowstand sandstones of amalgamated to semi-amalgamated fluvial-bar complexes that are thinner and less laterally persistent. Distal (down-dip) fluvial-facies belts are interpreted as lowstand sandstones characterised by thin, yet laterally extensive, fluvial-bar complexes. Within a sequence-stratigraphic context, the composite sequences delineate the external architecture of the reservoir interval. The sequence sets and sequences define the small-scale or internal architectural variations.

The internal architecture of the reservoir intervals was modelled by means of two techniques. Proximal fluvial facies were generated using typical object-based modelling software to populate the zones with channel elements that are clustered to form channel complexes. Medial and distal fluvial facies, however, were modelled using recently developed software that populates the zones with discrete, user-defined, fluvial-bar elements that are distributed along thalwegs. The thalwegs are clustered to form amalgamated to semi-amalgamated bar complexes. This capability, referred to as bar-train modelling, gives a better description of the bar geometry and spatial distribution. The resulting geologic models improve the reservoir description of the external and internal fluvial architecture for the subsurface reservoir. In particular, the geologic models provide a more accurate description of the complex architecture of the lowstand fluvial sandstones as well as the internal architecture of the mudstones Finally, the geologic models more closely matched pressure-transient well tests and short-term production tests than did previous modelling methods.