The Impact of Automated Fault Detection and Extraction Technology on Seismic Interpretation

Abstract

3D wide azimuth seismic data plays a vital role in fault interpretation, which has significant importance during exploration and development stages. Interpreting faults in 3D seismic data is one of the most time consuming and challenging process especially when dealing with poor quality seismic data. This paper provides a complete workflow and example of its application from seismic pre-conditioning to fault detection and extraction automatically based on published concepts by Dave Hale. With recent advancement in computer technology, multi-threaded algorithms and data driven methodologies, geoscientists can automatically detect and interpret virtually all discontinuities in seismic data in an efficient manner. This workflow involves random and coherent noise suppression, seismic likelihood attributes generation to enhance the discontinuities, detect faults and extract them from thinned fault likelihood volume. Unlike other fault tracking methods that use local seismic continuity attributes, such as coherency, this automated method incorporates aspects of Hale's fault-oriented semblance algorithm, which highlights fault planes with unprecedented clarity. This methodology has been successfully applied on complex faulted reservoirs. It contributes to the extraction of detailed discontinuity information (minor and major) from 3D seismic data. The traditional manual interpretation step that follows the detection of faults was time consuming and error prone. Automated fault interpretation improves the fault tracking accuracy, consistency and significantly reduces fault interpretation time in prospect generation. This workflow will optimize and reduce uncertainty associated with the seismic fault interpretation process.