ABSTRACT
One significant advance in the science or art of seismic interpretation in the past few years has been an improvement in techniques of depth conversion. This is certainly due in part to an awareness of the limitations and errors involved in the standard and widely accepted principles, but is also due to an improvement in computer facilities and programs.
The different types of velocity field used in making a depth-map from a seismic time-map can be classified as follows:
Constant average velocity. This is essentially a copy of the time-map with depth labels on the contours.
Constant interval velocities. This is a “layer-cake” method which ignores increases of velocity due to compaction.
Average time-depth curve. The method can allow for compaction but may ignore differences in formation thicknesses.
Each of these methods can be improved by allowing the velocities or the curve to change with geographic location, and the accuracy is then limited by the density of data points.
Average functions. Used on a layer-cake basis, these take compaction or other geological phenomena into account, and allow for better interpolation between data points.
Functions with varying parameters. This may be the most accurate method to use if the geology changes significantly over the prospect. A comparison of these methods applied to synthetic data shows that major errors can be introduced by using simplistic methods, and more reliable depth-maps can be obtained by using algorithms which allow, for instance, for the increase of seismic velocity with burial depth. Other geologic changes such as inversion can be taken into account by making the parameters in the equations vary with location.
