Online Journal for E&P Geoscientists

Wharton, Stanley R.^*1; Kumpas, Michael ²; Bakhiet, AbdelFattah ¹; Tang, David ¹; Gregory, Arthur ¹; Lawrence, Paul ¹
(1) Saudi Aramco, Dhahran, Saudi Arabia. (2) Consultant, Marlow, United Kingdom.

Seismic sequence stratigraphy has traditionally been applied to 2D transects to validate regional interpretation models. Using 3D surveys for sequence stratigraphy, spatial horizon extractions and 3D visualization depends on fast computing and graphics hardware. Seismic sequence stratigraphy’s evolution has thus been dependent on advances in computer technology and software development.

Initially, 3D methodology for sequence stratigraphy was applied in small areas of field or prospect size. Recently, this methodology has started being used in regional exploration settings, but the large size of these 3D surveys has prohibited the application of seismic stratigraphy, because of the unmanageable size of the data volumes and the processing time required to extract a meaningful interpretation density.

Recent advances in computer processing power and data volume management, paired with advances in software capability to map entire seismic data volumes, have paved the way for the study of large 3D seismic volumes.

The spatial seismic horizon tracking described in this paper relies on analysis and alignment of the dip attribute of each seismic sample in the seismic volume. Data is pre-conditioned with a dip filter, which also removes outliers based on specified parameters. This conditioning is termed “dip steering” and it forms the basis for subsequent calculation of chronostratigraphy and horizon extractions from the 3D volume.

At Saudi Aramco, we have expanded the application of automated 3D chronostratigraphic horizon mapping from detailed studies of fields, to exploration targets covering vast areas. For these large areas, we apply the chronostratigraphic calculations to survey subsets. Minor mis-ties occur at the subset edges when merged, but these are considered insignificant when exploring in a regional context.

The automatic volume horizon tracking accuracy was blind tested on horizons, with moderate relief, against wells spread across several hundreds of square kilometers of 3D seismic coverage. The seismic volume was decimated for expedience, but with a subsequent loss of lateral resolution. Seismic-to-well accuracy was nevertheless better than +/-8 ms at 70% of the blind test intersections. This level of accuracy is acceptable for regional scale assessment purposes, but in areas of steep dip where higher lateral resolution is required, the bin size after decimation must be kept small to maintain optimum performance of the dip steering algorithm.