Datapages, Inc.Print this page

Bridge, John Stuart1
(1) Binghamton University, Binghamton, NY

ABSTRACT: Fluvial Facies Models for Petroleum Geologists

Petroleum geologists want to predict the geometry, orientation, spatial distribution, proportion and connectedness of fluvial reservoirs using typical subsurface data (mainly well logs, not many cores, variable-quality seismic data). Facies models for the main river types (meandering, braided) are now 3-D, detailed and quantitative, as a result of studies (using GPR and coring) of modern river deposits, plus laboratory and theoretical modeling. These facies models allow prediction of width of individual channel belts from their thickness and lithofacies. It is not necessary to know paleochannel pattern to do this, which is fortunate as paleochannel patterns cannot be determined from most types of subsurface data. Also, reliance on subjective correlation of channel belts between wells is unnecessary.
The spatial distribution, proportion and connectedness of channel-belt deposits within floodplain alluvium (alluvial architecture) depends on intrinsic controls such as: channel-belt geometry (width, thickness); floodplain width; deposition rate; location, frequency and duration of channel diversions (avulsions); and intra-basinal tectonism. Process-based (forward) models can predict quantitatively how 3-D alluvial architecture is determined by the intrinsic controls. Intrinsic controls are determined by tectonic setting, climate, and base-level change. However, most models that attempt to relate alluvial architecture to tectonism, and to change in base level and climate, are still qualitative, 2-D and controversial. Alluvial architecture of reservoirs is normally simulated using stochastic models based on subsurface data and ancient and modern analogs. However, this procedure can result in unrealistic reservoir characterization. The fitting of process-based (forward) models to subsurface data should be considered as an alternative.


AAPG Search and Discovery Article #90026©2004 AAPG Annual Meeting, Dallas, Texas, April 18-21, 2004.