Abstract
Abstract A new machine-learning approach based on a Dynamic Time Warping (DTW) algorithm is introduced to detect faults and fractures in 3D seismic data from an unconventional resource play in Eastern Cis- Caucasia (Russia). This novel approach allows for better edge detection in seismic amplitude volumes because it employs a detailed comparison of two neighbouring traces to detect discontinuity via a minimal horizontal distance. For benchmarking purposes the proposed DTW method is compared to a widely used multi-trace attribute (Variance). Subsequently, both calculated attribute cubes serve as an input for ANT-Tracking to delineate fault strike and fracture corridor trends. A comparison of results shows that better resolution and more complete fault images are obtained when the DTW method is applied.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have