The goal of natural fracture prediction technologies is to discern the likely location, orientation, and relative aperture of natural fractures from the structural configuration of an area and the mechanical qualities of the reservoir rock. This "geomechanical approach" can augment or be augmented by direct detection methods, or it can provide a cost-effective alternative where advanced seismic is not feasible.

The geomechanical approach is based on the mapping of fault systems within the reservoir from existing seismic data. Paleo-stress concentrations around the faults are calculated based on fault geometry or paleo-stress tensors. Minimum tensile failure and modified Coulomb shear failure criteria are then determined to estimate the potential for natural fracturing.

The fracture detection optimization project  completed in the Piceance basin of Colorado showed the enormous potential of the geomechanical approach. A 3-D, P-wave, multi-azimuth seismic survey was collected over a 4.5 square mile area within the Rulison field. This data was analyzed for direct fracture detection, but the results were ambiguous and highlighted the basic technical problems that may hinder this approach in many areas. However, advanced processing of the data allowed the recognition of many subtle faults within the field. Geomechanical modeling of the stress and failure implications of the fault deformation identified areas of likely fracture clustering within the field that correlated well with higher well productivities. This approach was subsequently applied in the nearby Mamm Creek field with similar favorable results.

A follow-on test of the geomechanical approach was conducted in the Frontier Sandstone (15,000 feet depth) in the Table Rock field, eastern Greater Green River basin, Wyoming. Existing 3-D seismic surveys were re-processed and analyzed for detailed fault mapping. The predictions of the model were consistent with the observed production from four wells in the area, and a general location for a deep horizontal well was determined. The geomechanical model was then rigorously applied to a smaller, 80-section area, and enhanced fracture concentration was predicted for the proposed well location. The well's initial production of 14 million cubic feet/day exceeded all other wells in the area.

Further testing and extension of the geomechanical approach is planned. Testing in true exploration settings at three separate field sites (Greater Green River, Wind River, and Anadarko basins) is planned. Work to fully integrate the geomechanical model into an integrated system that includes state-of-the-art fracture simulators is also underway.

More on Exploration Technologies