Operational efficiency in highly heterogeneous (channelized or fractured) formations requires unique characterization methods. Reservoir characterization often is performed in sequence: 1) build the static 3D geologic model, 2) generate the saturation governing inputs (capillary pressure, relative permeability curves and PVT data), 3) upscale, and 4) modify the upscaled model to "history match" the dynamic performance data. There has been much recent emphasis on upscaling the static model to preserve appropriate heterogeneity within the simulator; however, during the history match the upscaled model may be substantially altered. This paper describes an alternative approach of incorporating both static and dynamic information at the beginning of the reservoir characterization process. This approach decreases both the modifications to the geologic and saturation inputs and the time required for history matching.

This paper provides an example from the Marathon operated Yates Field showing how flow data compiled by well, area, and total reservoir was used to assess the large scale "plumbing" of the reservoir. Tools developed for 3D geologic modelling have been deployed to distribute the connected secondary-porosity flow features (caves, solution-enhanced fractures, connected vugs, etc.) and barriers. Static geological data from the well scale is honored along with deterministic large-scale flow characteristics. These identified large-scale flow enhancements and barriers were incorporated into field operational planning and simulation construction early in the characterization process. The characterization efforts have lead to improved workover success in the Yates Field. Simulation and Yates Field examples demonstrating the value of simultaneous static and dynamic characterization are provided.

You can access this article if you purchase or spend a download.