Time-depth imaging offshore India

Generalized stratigraphy of the west coast Mumbai High consisting mostly of carbonates and shales. This differs greatly compared with the stratigraphy of the east coast Krishna-Godavari basin. Distal reservoirs of the K-G basin are Miocene/Pliocene submarine fan and channel-levee complexes.

Offshore discoveries

Currently three of the hydrocarbon basins produce: the Mumbai offshore basin off the west coast, and the Cauvery and Krishna-Godavari basins off the east coast. In addition, there are deepwater discoveries being evaluated in the K-G, Cauvery, Mahanadi, and Bengali basins in the east coast and attractive prospects in the Andaman Sea.

The stratigraphic columns of the east coast basins are strikingly similar, with successive sequences of open marine clastics laid down during the Upper Cretaceous, Tertiary, and Plio-Pleistocene. The west coast basins differ because carbonates predominate. The stratigraphy of continental margin basins reflects the evolutionary phases of subcontinental migration and marginal basin accretionary processes during India’s continental drift.

In the east coast, abundant sedimentation occurred after the Eocene with thick sediment wedges in the Oligocene essentially partitioning the deeper water realm from the associated sediments onshore. The basins generally are broad drape anticlines. Most structures along the eastern continental margin of India consist of rotated seaward-tilted faulted blocks with anticline and roll-over features formed in association with block-bounding listric faults.

The west coast carbonates were deposited in several pulses of marine transgression associated with vertical uplift of the basins. The typical seismic section off the west coast differs sharply from the east coast types due to the distinct phases of sediment accumulation in the Arabian Sea.

Processing overview

As mentioned, the best input to PSDM is pre-processed seismic data and unsmoothed final velocities from NMO/DMO legacy stacks or from a contemporaneous PSTM workflow. These velocities are converted to interval velocity in depth and subsequently smoothed. The initial velocity model is built using these velocities. The depth imaging workflow consists of multiple iterations of PSDM and reflection tomography based velocity analysis of residual moveout on common image point gathers (CIP gathers) with final calibration to well data.

Once the final velocity model is built and CIP gathers are flat at the correct depth, the full isotropic depth migration is run. If there are indications of anisotropy, the anisotropic parameters are estimated and propagated through the velocity model during model building for a final anisotropic migration. Interpretation mapping of the PSTM/PSDM paired images provides for new enhanced domain with 3D structure and stratigraphic features correctly placed in depth and adding new parameters to visualize deepwater drilling.

Tomography workflow

Grid-based CIP tomography uses all the primaries at all horizons (Schulz and Canales, 1997). This can handle non-hyperbolic events, mainly velocity anisotropy. The tomography workflow begins with a velocity model derived from the stacking velocities obtained from the PSTM processing. First, an accurate water depth is picked on the PSTM stack converted to depth. The final RMS velocity is converted to interval velocity in depth and is smoothed to build the velocity model. If the gathers are flat, the velocity model is inherently correct and gives rise to the initial PSDM section. However, gathers usually are not flat so further iterations are necessary to correct the residual move-out.

This requires selected residual non-hyperbolic events to be input to 3D-ray-traced linear tomography equations; the solution represents the velocity model update for the iteration under consideration. This is followed by another PSDM step using the new, updated velocity model.

The initial iterations of CIP-tomography are parameterized such that velocity updates are done using large-scale lengths. Choice of scale length depends on lateral and vertical (anisotropic) velocity variations. As the velocity model improves through successive iterations, progressively smaller scale lengths resolve the finer details in the velocity field.

Anisotropy estimation

From the initial velocity model, up to four isotropic iterations are run. Anisotropic coefficients, Epsilon and Delta, can be calculated using the time-depth curve from well data (Ball, 1995; Kirtland Grech et al., 2001). Vertical transverse isotropy (VTI), also called polar anisotropy, is caused by sedimentary layering, whereas intrinsic anisotropy is due to clay particle alignments (shale) during deposition. In anisotropic media, compression wave seismic migration velocity changes with the angle of propagation, as given by the Thomsen equation.

Vp (θ) = Vv (1+δ Sin2 θ Cos2 θ + ε Sin4 θ)

The isotropic gathers indicate the peak event flattens but it doesn’t tie to the top-carbonate depth marker in the well; the trough event also flattens but does not tie to the top-reservoir depth marker in the well. After correcting for anisotropy, the events flatten and the peak ties to the depth of the top of the carbonate marker in the well and the trough ties to the top of the reservoir in the well. Anisotropic updates are run to develop a final velocity model, which is then converted to the final anisotropic PSDM section.

PSTM vs. PSDM

Although interpretation is still conducted on time sections, the inherent distortions these bring need to be understood. In time migration (post-stack or pre-stack) the velocity profile is retrieved at the CMP and the program either computes the travel time via the DSR equation (straight ray) or ray traces through the local model (curved ray). No lateral velocity changes are discerned; the ray path is always symmetric for a flat event (1D velocity model). Upon depth migration, a full travel-time table is built externally. The travel-time generator program comprehends velocity changes both vertically and laterally. The migration program retrieves the travel-time from the table to move the sample. The ray path can be nonsymmetric even for a flat reflector (3D velocity model).

Post-migration processing

One historical paradigm is the interpretation of volumes of data exclusively in time section. This view is changing and geophysics is entering a new phase in which interpreters tend towards exclusive use of depth sections.

Post-migration processing consists of the following steps:

1) Sort the migrated offsets to CMP

2) Convert to time

3) Run a residual multiple attenuation filter with Hi-Res Radon

4) Conduct primal, inverse Q, outside mute

5) SRAC, TVF, RAAC processing.

This results in the APSDM stack. This section then gets spectral whitening frequency enhancement and time variant filtering. At this point the newly obtained APSDM is final. The example indicates the improvement in the imaging of the Tertiary strata above basalt flood and of the Mesozoic strata under the basalt off the west coast of India.

References

Ball, G., 1995, Estimation of anisotropy and anisotropic 3-D prestack depth migration, offshore Zaire, Geophysics 60, 1495-1513.

Fainstein, R., Banik, N.C. and Broetz, R. J., 2008, “Examination of evolving hydrocarbon exploration technology in India”, Regional Technology Center – RTC, Mumbai, 57 pp.

Kirtland Grech, M.G., Cheadle, S. and Lawton, D. C., 2001, Integrating borehole information and surface seismic for velocity anisotropy analysis and depth imaging: The Leading Edge, 20, no. 5, 519-523.

Schultz, P. and Canales, L., 1997, Seismic velocity model building: CE in Dallas, 2 November: The Leading Edge 16, no. 7, 1063-1064