Quantitative Fluvial Facies Models as Guides to Subsurface Interpretations
Interpretation of the planform style of a river is not typically achievable from 1D samples of its deposits, such as those offered by cores and image logs: facies, trends thereof, and even accretion styles when inferred, are not truly diagnostic. Notwithstanding, as the architectures of channel deposits of braided and meandering rivers differ significantly, reservoir intervals are still commonly classified on the supposed channel patterns of their formative rivers for purposes of reservoir characterization, often on the sole basis of the application of idealized facies models or selected analogs. In spite of the awareness, at least in the scientific community, of the inadequacy of 1D fluvial facies models as templates for interpretations of river planform, traditional fluvial facies models are still widely applied, particularly by non-specialists, in the interpretation of cored intervals and for the erection of conceptual models of subsurface architecture. With the aim to provide measures of uncertainty in subsurface interpretations, this study presents a quantitative assessment of relationships between attributes of facies in channel deposits and classes of fluvial systems.
A database-driven analysis is undertaken for over 70 reaches of modern rivers in which vertical facies arrangements can be constrained with confidence through the study of exposures and cores. The studied rivers are classified in terms of channel pattern, but also on additional attributes that further describe their depositional context (e.g., hydrology, gradient, dominant mode of sediment transport, climate). Channel deposits that form barforms and channel fills of different types are characterized quantitatively with respect to their proportions, thicknesses, and vertical transitions. By synthesizing data on these features for classes of fluvial systems, quantitative facies models are compiled, which highlight what facies characteristics tend to be associated with particular types of depositional systems, and to what degree. These models demonstrate that modal facies architectures of sand-rich braided and meandering systems are not distinguishable statistically, but also that certain characteristics, particularly extreme values in facies proportions, might have some value for guiding core interpretations.
The results serve to facilitate the communication of uncertainty in interpretations of cores and image logs, and in resulting conceptual models of reservoir architecture.
AAPG Datapages/Search and Discovery Article #90350 © 2019 AAPG Annual Convention and Exhibition, San Antonio, Texas, May 19-22, 2019