An expert system for Seismic data interpretation using visual and analytical tools

-Seismic images are extensively used by petroleum geologists to delineate subsurface geological structures. This interpretation is dependent on interpreter's individual skills, and expertise causing a complex seismic image to be interpreted differently by different seismologists. To overcome this uncertainty a prototype expert system SeisExpert, has been developed in which the experience and heuristic knowledge of the experts have been incorporated in the form of a set of rules. Over 120 rules have been buil t into a commercially available expert system shell, ‘Flex’ (from Logic Programming Associates, UK). Information received from the user is chained through these rules to interpret the seismic images. Additionally, analytical procedures have been provided to refine the interpretation in case of complex subsurface structures. These include statistical cross-correlation across the seismic traces and secondary attributes such as instantaneous phase and reflection strength. Using this system, several horizons were successfully tracked across seismic traces in four different seismic images containing field data from existing oil fields. Discontinuities such as faults, wherever present, were also correctly identified.

[1]  K. Marfurt,et al.  Seismic attributes — A historical perspective , 2005 .

[2]  Klaus D. Tönnies,et al.  Exploiting 3D Spatial Continuity for Robust Automatic Horizon Matching across Faults , 2006, Third International Symposium on 3D Data Processing, Visualization, and Transmission (3DPVT'06).

[3]  Arthur E. Barnes,et al.  Redundant and useless seismic attributes , 2007 .

[4]  Alistair R. Brown Understanding seismic attributes , 2001 .

[5]  Monique Thonnat,et al.  A Knowledge-based Approach of Seismic Interpretation : Horizon And Dip-fault Detection By Means of Cognitive Vision. , 2008 .

[6]  M. Taner,et al.  Complex seismic trace analysis , 1979 .

[7]  T.S. Durrani,et al.  Seismic horizon picking using an artificial neural network , 1992, [Proceedings] ICASSP-92: 1992 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[8]  Klaus D. Tönnies,et al.  Horizon correlation across faults guided by geological constraints , 2002, IS&T/SPIE Electronic Imaging.