--> Empirical Characterization of the Geometry of Fluvial Channel Deposits: Implications for Object-Based Reservoir Modeling

AAPG ACE 2018

Datapages, Inc.Print this page

Empirical Characterization of the Geometry of Fluvial Channel Deposits: Implications for Object-Based Reservoir Modeling

Abstract

Channelized fluvial deposits representing the infill of channels, valleys or the aggradation of channel belts form the main components of most fluvial reservoirs. A characterization of reservoirs of this type requires inference of the possible distribution and geometry of channelized bodies in the inter-well volume, and this is commonly attempted by means of object-based reservoir-modeling techniques, which require families of channelized units to be parameterized on quantities describing their planform and cross-sectional geometries.

The geometric realism of object-based reservoir models ensures plausibility in volume and production forecasts: quantitative constraints of geologic sensibility are therefore advocated, and for this reason it is common to refer to geologic analogs. Given their ability to reproduce complex geometries and to draw upon the analog experience, object-based models are considered inherently realistic. Yet, this is merely a qualitative perception of realism, as a systematic quantified assessment of the outputs of these techniques against the geologic record has never been undertaken.

This work presents a synthesis of data on channel bodies derived from a large database that describes the sedimentary architecture of fluvial systems. The aim is to provide a set of tools for constraining object-based fluvial reservoir models in data-poor situations. Analog data are used to test the intrinsic realism of four object-based modeling algorithms by comparing characteristics of the modelled architecture against fundamental patterns identified across the analogs.

Based on data from >7,000 large-scale channel bodies identified by 26 analog studies, an empirical characterization of the geometry of fluvial channel bodies is undertaken that describes distributions in, and relationships between, sandbody thickness, cross-stream width, and mean planform wavelength and amplitude. 780 object-based models are built running simulations conditioned on six alternative analog-informed parameter sets, using the four algorithms according to seven different modelling strategies. A closeness of match between the analogs and the models is then determined on a statistical basis.

Results indicate what modeling approaches return architectures that more closely resemble the organization of known fluvial systems. However, no algorithm fully reproduces the expected results, suggesting that future modeling tools should try to incorporate additional geologic knowledge.