A Unified 3-D Seismic Workflow

Abstract A geophysicist who practices seismic data analysis for earth modeling and imaging in depth is overwhelmed by the prolific number of inversion methods to estimate layer velocities and delineate reflector geometries—the two constituents of a seismically defined earth model in depth. Given a specific type of structural play, the key to estimating an accurate earth model in depth, however, is a workflow that is based on a judicious combination of inversion methods appropriately selected for their robustness. We present a unified workflow for processing, inversion, and interpretation of 3-D seismic data that is applicable to low-relief structures and complex structures associated with extensional and compressional tectonics. With some modifications, the workflow also is applicable to complex overburden structures associated with salt and overthrust tectonics. Although doing it right the first time is most desirable, there is never a situation where this is possible when estimating an earth model in depth. A fundamental problem with inversion for earth modeling is velocity-depth ambiguity. This means that an error in layer velocity can be indistinguishable from an error in reflector geometry. The velocity-depth ambiguity that is inherent to seismic inversion makes it very difficult to obtain the right answer (an adequate representation of the true geological model), let alone do it the first time. Limitations in the resolving power of the methods to estimate layer velocities that arise from the band-limited nature of the recorded data and finite cable length used in recording further compound the problem. Additionally, traveltime picking that is needed for most velocity estimation techniques and time-to-depth conversion as well as picking depth horizons from depth-migrated data to delineate reflector geometries are all adversely affected by noise present in the data. All things considered, we can only expect to do our best in estimating what may be called an initial model, and update this model to get an acceptable final model. The objective behind the design of the seismic workflow described in this paper is to attain the best estimate of a structurally consistent initial model based on rms velocities associated with migrated data, so as to minimize the work required to update the model. The unified workflow involves analysis of seismic data both in time and depth, and follows a pathway that starts with the application of 3-D dip-moveout correction and 3-D prestack time migration to derive an rms velocity field. This is followed by estimation of an accurate, structurally consistent initial model by Dix conversion of rms velocities and interpretation of a set of depth horizons from 3-D poststack depth migration. To update the initial model, the image gathers derived from 3-D prestack depth migration are analyzed for residual moveout. The resulting final model is then used to perform 3-D prestack depth migration to obtain an image volume in depth. The final phase of the workflow includes structural and stratigraphic interpretation of the image volume with the ultimate objective of obtaining a seismically derived reservoir model.