--> Abstract: Modeling Flow Barriers in Distributary Systems Using a Numerical Analog from a Process-Based Model, by Hongmei Li and Tao Sun; #90078 (2008)

Datapages, Inc.Print this page

Modeling Flow Barriers in Distributary Systems Using a Numerical Analog from a Process-Based Model

Hongmei Li1 and Tao Sun2
1Energy Resources Engineering, Stanford University, Califonia, CA
2Reservoir Characterization, ExxonMobil Upstream Research Company, Houston, TX

Scour lobes are the fundamental depositional elements in deep-water distributary systems. One common feature of a scour lobe is the fining of grain sizes in both vertical and horizontal directions. The finest grained deposits are potential flow barriers which could compartmentalize the reservoir. On the other hand, erosion (due to scour) of the fine-gained layer during deposition of the overlaying lobes may lead to vertical amalgamation of lobes that enhance the reservoir’s connectivity. Therefore, modeling flow barrier and scour distribution is key to predicting reservoir connectivity.

Since barrier distribution is associated with individual lobe geometry, we developed a workflow for extracting lobe stacking patterns from process-based models. These process-based models simulate the fundamental geologic process and produce realistic 3D numerical representations of the reservoir architecture. Therefore, their results provide ideal reservoir analogs for collecting lobe stacking pattern statistics.

Using the process-based model as a reservoir analog, a hierarchic approach is used for modeling flow barrier distribution. First, the lobe distribution is simulated using a surface-based modeling technique. Individual lobe placement is determined based on the statistics of stacking pattern and topographic constraint. Second, barriers are simulated within each lobe. This is done either by explicitly adding barrier to lobe top surface, or by simulating permeability using vertical and horizontal trends. The resulting simulated models successfully reproduce the pattern features observed in process-based models.

 

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