Use of V0-K depth conversion from shelf to deep-water: how deep is that brightspot?

As worldwide hydrocarbon exploration moves into progressively deeper waters, there is a growing requirement for the depth conversion of seismic times in areas with little velocity control. The V0-K method is a familiar way of depth converting seismic reflection travel times. While this method can give excellent results, large errors may be produced when it is used for regional surfaces with large relief, for instance when using velocity data from wells on the continental shelf to make prognoses for adjacent deep water areas. This paper uses exploration wells in the UK West of Shetlands area, which span water depths from 100 to 1000 m, to constrain and test the accuracy of different V0-K based depth conversion approaches for the Cenozoic post-rift section of deep water areas. The average velocity method, which implies vertical iso-velocity contours, and the traditional form of the V0-K method, which implies horizontal iso-velocity contours, are used as benchmarks to display the advantages of using the seabed as a datum and of allowing lateral variation in the V0 and K parameters. The seabed datum reduces nonuniqueness in the V0 parameter, particularly in deep water. The best depth conversion results in this study are obtained by determining relationships between both V0 and K and the water depth, although good results can also be achieved with optimized single V0 and K values and a seabed datum.