An Outcrop Derived, Deterministic Base-Case Fluvial Model for Reservoir Analog Studies: Modeling Coastal-Plain Deposits from the Blackhawk Formation of Eastern Utah

Christian Carlsson, Håvard Enge, Simon Buckley, John A. Howell, David L. Moreno, and Sigurd Aanonsen
Centre for Integrated Petroleum Research, Department of Earth Science, University of Bergen, Bergen, Norway

Fluvial deposits act as complex reservoirs on a variety of scales. Outcrop analog studies have long been recognized as a means of dealing with the complexity. The aim of the current study is to use the latest field data collection techniques combined with modern reservoir modeling methodologies to build a “close to deterministic” model of a well exposed fluvial system that can be used for reservoir engineering studies.

The chosen outcrops are from the non-marine portion of the Cretaceous Blackhawk Formation, which outcrop in the Book Cliffs of eastern Utah. The study interval includes a 50 m thick succession of marginal to non-marine coastal-plain deposits, which are very well exposed in Woodside Canyon. The outcrops include channels, crevasse splays, coal, overbank and lagoonal deposits. The sequence stratigraphy of the interval is correlated to downdip marine deposits and well understood.

Data were collected from a 2.5 sq. km area using traditional field techniques and LIDAR. The LIDAR data were used to generate a high resolution virtual outcrop. A series of key surfaces were then digitized and exported to a geocellular reservoir modeling package. Within the modeling package the 3D bedforms were recreated, initial as separate zones and then merged into a single, high-confidence facies model. The facies can then be used to guide the population of the model with petrophysical properties for flow simulation. A key aspect of the model is it’s geometrical accuracy and the freedom to assign a variety of different rock and fluid properties and test different well locations and completion strategies. The model is currently being used to test algorithims based on the Ensemble Kalman Filter method for automated history matching as part of the “Rapid Model Updating Project”. Initial results of this will be presented.

AAPG Search and Discover Article #90078©2008 AAPG Annual Convention, San Antonio, Texas