P. Sangvai, A. Biswal, M. Mathur - Reliance Industries Ltd.
J.F. Fruehn, P. Smith, I.F. Jones, D.G. King, M.C. Goodwin, V. Valler - ION GX Technology Corp.
Imaging in complex environments requires a wide range of tools to suppress various classes of noise and multiples. This must be accomplished in the pre-stack domain so that automated dense picking can be performed on migrated gathers to permit reliable model update.
Use of such an approach for data offshore eastern India improves image quality compared to a recent pre-stack time migration, avoiding the structural distortion introduced by severe localized velocity variation in the near surface sediments, delivering gathers suitable for attribute work.
Imaging in deep water environments poses a specific set of challenges, both in data pre-conditioning and imaging. These challenges include scattered complex 3D multiples, aliased noise; and low velocity shallow anomalies associated with channel fills and gas hydrates.
Transition from the shallower coastal waters to the deep shelf off the east coast of India often encounters significant topographical variation in the seabed. In addition to deep channels and steep slopes, there are buried channels with low velocity fills and gas hydrates. These give rise to numerous effects which the processing geophysicist has to handle. Diffracted and “out-of-plane” multiples are the norm in these environments, and must be dealt with in order to derive a reliable velocity model for an acceptable structural image.
To address multiples, differential velocity based methods such as Parabolic Radon often are used in deepwater. To some extent, the problem of aliasing of the multiples on far offsets can be addressed either by interpolation and/or use of a de-aliased (“beam”) Radon transform. However, Radon-based techniques fail for complex multiples, as the apex of the events in the CMP domain does not fall on zero offset for ray paths not in the plane of the shot-receiver axis. In these cases, an alternative is necessary.
In recent years, surface-related multiple elimination (SRME) has become popular in deepwater. Near offset multiples in particular are better attenuated than with Parabolic Radon. Cascading 2D SRME and Radon has become an industry standard. However, the complexity of the multiple generator and “out-of-plane” effects can limit even this combination. With the advent of 3D SRME, a theoretically more correct approach is available. The following demonstrates its effectiveness compared with the “conventional” approach.
Results from 2D SRME compared to those from 3D SRME show complex ray-paths for the first seabed multiple and associated sedimentary layers give rise to a shifted apex aspect to the moveout behavior as seen in the CMP domain. Following either 2D or 3D SRME, additional de-noise techniques can be applied to deal with the aliased noise and other classes of noise.
A comparison as a QC stack (prior to 3D pre-stack depth migration [PSDM]) of the 2D SRME shows a swath of noise over an area of interest (a major unconformity). This remnant multiple energy will be spread around during migration and will be difficult to remove at that stage. Conversely, following 3D SRME, the QC stack is mainly free of this multiple contamination.
Velocity model building and PSDM
In an environment with punctual discontinuous velocity anomalies, such as narrow channel fills or gas hydrate accumulations, a purely layer based velocity model is inadequate. Furthermore, a purely gridded approach also may encounter problems.
This project used a hybrid-gridded approach that combines conventional gridded tomography, high resolution gridded tomography, auto-picked layers, and detailed manually interpreted layers. The initial depth interval velocity was derived from the time-stacking velocity (smoothed and converted to depth interval velocity), and the water bottom was picked from a water-velocity depth migration and inserted in the initial model as an explicit layer.
Following this, several iterations of gridded tomographic model updates were performed. This involves running an autopicker (in this instance based on plane-wave destructors) on densely sampled CRP depth gathers, and inputting the autopicked velocity errors and dip information to the gridded tomographic 3D solver.
For gas hydrate accumulations, this project relied primarily on high resolution gridded tomography, and could resolve small-scale velocity features with Vi ~ 1250 m/s, compared with the background sediment velocities of ~ 1600 m/s. For detailed narrow channels, the project relied on manual interpretation of the top and base of the channel features, and a scan over potential channel-fill velocities.
In the first kilometer of data near the seabed are deeply incised seabed canyons and also some small localized channels just below the seabed. These channels result in a severe pull-down distortion of the underlying sediments due to their low-velocity fill. A smooth velocity model would not resolve these small-scale features (typically 200 m [656 ft] in width). Hence, detailed manual picking is required. The 3D PSDM image shown was created using a smooth background velocity field. Hence, the pull-down is visible. The detailed channel-fill velocity model is superimposed. A migration velocity scan determines the best channel-fill velocity: This case used 1,200 m/s.
Comparing 3D PSDM result after migrating with a smooth background velocity field (no punctual channels included) versus the result incorporating the low-velocity fill channels, improves the deeper section. The shallow channel problems are not solved, but incorporating them in this way enables better resolution in the deeper section. Ignoring them is not a viable option.
For these data there are no classical multi-pathing problems. Hence for the final migration, an amplitude preserving Kirchhoff migration was used.
Acknowledgements
Our thanks to our colleagues at Reliance Industries and ION GX Technology for help and advice during this project, and to our respective employers for permission to present this work.
This article is based on a paper presented at 2008 EAGE.
References
Claerbout, J.F., 1992, Earth Soundings Analysis: PVI, Blackwell Scientific Publications.
Hardy, P.B., 2003, High resolution tomographic MVA with automation, SEG/EAGE summer research workshop, Trieste.
Jones, I.F., 2003, A review of 3D preSDM velocity model building techniques First Break, 21, No.3, pp45-58.
Jones, I.F., Sugrue, M.J., Hardy, P.B., 2007, Hybrid Gridded Tomography. First Break, v25, No.4, p15-21.
Stewart, P., 2004: Multiple attenuation techniques suitable for varying water depths, Proceedings of the CSEG annual meeting.
Stewart, P.G., Jones, I.F., Hardy, P.B., 2007, Solutions for deep water imaging, Geohorizons, January, p8-22.
