Applications of Automatic Fault Extraction (AFE) in a Variety of Geologic Environments: Results, Limitations, and Suggested Improvements

Dorn, Geoffrey A.¹, Huw E. James², Laura Evins², Jonathan Marbach¹, Francis A. Coady¹ (1) TerraSpark Geosciences, L.P, Boulder, CO (2) Paradigm Geophysical, Houston, TX

Automatic Fault Extraction (AFE) is a process designed to automatically extract fault surfaces from 3-D continuity (e.g., Coherency) volumes. AFE combines signal processing technology with intelligent geologic rule based steps to enable automated and semi-automated fault interpretation. The output is a relative fault probability volume called the Fault Enhanced (FE) attribute volume, fault polylines and fault surfaces.

The workflow for AFE is comprised of a combination of 2-D image processing algorithms, modified to accommodate the characteristics of discontinuities in 3-D seismic data. A Lineament Enhanced (LE) volume is generated by applying 2-D image processing algorithm to time-slices of continuity. The FE volume is generated from the LE volume by applying a 3-D generalization of a 2-D image processing algorithm that searches for small dipping planar features in the data. 3-D faults are extracted from the FE volume through a combination of extraction of fault cuts on inlines, crosslines and time-slices through the FE volume, and linking the fault cuts into discrete fault surfaces.

Experience with the application of AFE in a variety of geologic settings demonstrates the interpretive value of the FE attribute volume created by AFE for structural interpretation, and the limitations of the AFE process. It also suggests several potentially more reliable approaches to the extraction of the 3-D fault pattern from the FE attribute volume.