Migration Study of the Uranus Prospect in the Barents Sea: Evaluation of Velocity Sensitivity, Wave Equation Migration and Acquisition Geometry

Abstract

Imaging of salt is a known geophysical challenge, since salt has complex geology, with varying lithology and complicated structures. Salt diapir can have steep dips and obtaining a correct velocity models are some of the issues related to imaging salt structures. Three 2D seismic lines and one velocity model from the Uranus prospect drilled in the Nordkapp Basin in the Barents Sea were investigated. Two depth migrations have been applied and compared; the reverse time migration and split-step Fourier migration. Since reverse time migration is known to be costly, but better suited for complex structures, the objective was to assess the value of reverse time migration. Then two single-spread lines have been migrated with pre-stack reverse time migration, compared, and combined into one split-spread line. The objective was to examine the value of acquiring costly split-spread surveys. A third objective was to study the velocity sensitivity of pre-stack reverse time migration by systematically altering the velocity model. Reverse time migration proved to handle steep dips, diffractions and large lateral velocity changes when applied to seismic containing two salt diapirs. The split-step Fourier migration produced good results for smaller dips, and had higher resolution of layers. In the presence of multiples, near vertical dips and large velocity contrast, the split-step Fourier migration was unable to image an adequate result. Since the salt diapirs had dips in opposite directions, the single-spread lines each provided shots in up-dip direction and thereby different imaging of the salt flanks. In most cases, the split-spread derived a depth and angle between the two single-spreads, which made the interpretation of the split-spread ambiguous. As the velocity was increased, the interpretations of the flanks showed an overall improvement and reflections were moved to greater depths. The opposite was true for a velocity decrease. Deviation from the initial velocity model increased with depth. None of the measures improved the imaging of the flanks when the data quality was poor. The reverse time migration proved to handle the complexities associated with great dips and large lateral velocity changes better, and is preferred for imaging salt diapirs. Split-step Fourier migration offers a reasonable alternative if dips are small. Acquiring a split-spread line was evaluated to be desirable when only 2D data is provided. Single-spread data can be viewed separately to avoid ambiguous interpretations. Results indicate that the initial velocity model estimated too low velocities in the sediment layers. Improved imaging of the salt flanks shows that pre-stack reverse time migration is velocity sensitive. Throughout the study no base salt reflection was imaged, and changes in the velocity model or acquiring of 3D data could improve the results. In addition, one should strive to obtain high quality data when dealing with complex structures.