Abstract: Velocity Estimation for Depth Conversion of a 3-D Data Set in Troll Field: A Comparison of Different Techniques

David McCorkindale, Long-Far Hwang

3-D and 2-D seismic data were interpreted and merged to map Troll Field. The quality of the seismic time-to-depth conversion for the topmost reservoir, the Sognefjord, is dependent on the quality of the velocity predictions. Based on 26 well data points and the mapped seismic times, 4 different methods -- a surface fitting approach, linear regression, ordinary kriging, and kriging with an external drift -- were used to estimate the velocities between the wells and build velocity models for depth converting the reservoir. Average velocities were calculated to the top of the Sognefjord formation from the interpreted well depths and seismic time data, and these were used with the different methods to estimate velocities between the wells. To test the prediction accuracy of t e methods, the average velocity at each well was removed, one at a time, and the remaining 25 well values were used within each technique to estimate the value that was removed. The estimated value was compared back to the actual value to illustrate the accuracy. The accuracy of the 4 methods varied from well to well, but one method, kriging with an external drift, provided significantly better results overall. This method also had the advantage of providing estimates of the uncertainty of the predictions.

AAPG Search and Discovery Article #90960©1995 AAPG Southwest Section Meeting, Dallas, Texas