3D muti-scale tomography: A case study from Oklahoma mountain front

Elive Menyoli1

¹Marathon Oil Corporation

In this study, we applied a multi-scale tomographic strategy on a wide-azimuth onshore dataset along the Oklahoma mountain front. The Oklahoma region has complex geology where several major sedimentary basins are set amongst mountain ranges and uplifts. In some areas eroded mountains created prolific granite wash formation which are difficult for seismic imaging. Vertical and overturned beds due to thrusting, plus associated wrench faulting make seismic imaging difficult. Strong lateral velocity variations and raypath bending would require a detail velocity model to identify edge of granite thrust and image sub-granite reservoir traps. Therefore, a prestack depth imaging workflow is a necessity for any exploration and production project in this area.

To estimate an accurate depth imaging velocity model, a single-scale ray-based tomography is routinely applied industry-wide. For example once the number of azimuth to be used for tomography is chosen, it becomes an invariant. In the case of wide azimuth survey as in this onshore study, a single-scale approach might fail to resolve features which change with azimuth, thus leading to azimuthal velocity smearing.

Variables for the tomographic scale decomposition include; azimuth range, velocity grid parameterization and ray shooting density. Horizons as well as locally coherent seismic patches were used as input into the tomography. This case study shows that adopting azimuth in the scaling strategy resolve azimuthal velocity details, and ensures a reasonable degree of convergence. The final depth migration images with the velocity model showed high quality sub-granite events as well as the definition of the granite-wash geometry.

AAPG Search and Discover Article #90097 © 2009 Tulsa Geological Society, Tulsa, Oklahoma

 

 

��������������� �������������������������������������������������������������������������������������������������������������������������������