Technical Focus

Figure 1. Comparison of the Strike vs. dip amplitude density at a target surface froma 3D ray tracing study. Botha acquisition directions yield fundamental illumination degradationon the flanks of very steep structures, although strike shooting is somewhat better. A cross-line 'footprint' between adjacent sail lines compounds the illumination problems assaciated with dip shooting. From Long et al. (2004)

 

Figure 2. Modelled subsurface illumination maps showing the subsurface 'hit density' for the Varg survey area in the North Sea. Each result is different none is ideal. From Reksnes et al. (2002)

 

Figure 3. Modelled subsurface illumination maps showing the offset distribution (i.e. number of unique offsets) for the Varg survey area in the North Sea. The density and uniformity of offset coverage improves as more shooting directions are combined (single-azimuth vs. dual-asimuth vs. triple-azimuth). From Reksnes et al,(2002).

 

Figure 4. Amplitude maps for the Varg taget horizoni n the North Sea, derived from a single-azimuth dip 3D survey (left) and multi-azimuth acquisition (right). Improved illumination and subsurface sampling has resulted in significantly greater data quality, clarity and resolution. From Reksnes et al. (2002).

 

Figure 5.Interpreted time structure map at the main Varg producing horizon in the North Sea. Multi-azimuth acquisition (left) has provided a far more detailed field understanding than single-azimuth acquisition (right). Reservoir compartmentalisation has been completely revised, enabling the succesful drilling of two step-out wells.

Multi-Azimuth Seismic
By Andrew Long, PGS Marine Geophysical, andrew.long@pgs.com

Introduction
In the ideal world, 3D streamer seismic acquisition satisfies three main criteria:

1. Complete illumination of the target geology. In other words, a high density of energy is uniformly reflected from each subsurface point on the target

2. The reflected seismic wavefield is spatially sampled in all directions without aliasing as it encounters the surface

3. Efficiency must be as high as possible in the field

Unfortunately, target illumination is never uniform, and therein lies a weakness in the seismic method, as conventionally applied. In the least case, irregular target illumination will result in processing artefacts, acquisition footprint, amplitude anomalies, increased noise, and degraded resolution. In the worst case, data quality and resolution will be so poor that confidence cannot exist in interpretation, exploratory drilling will be cancelled, field development will be incomplete or stalled, and field production is mismanaged. The risk of acquiring seismic data with bad target illumination is the wasted cost of seismic acquisition, and the cascading effects upon delayed/cancelled/incorrect exploration and production drilling programmes.
Multi-azimuth 3D seismic acquisition is a robust solution to target illumination problems, that involves the following:

  • A 3D streamer survey is acquired in more than one direction over the same survey location
  • Different shooting directions illuminate different parts of the target
  • Collectively, the overall target illumination will be far more uniform and complete
  • The datasets are collectively processed to output a single 3D seismic cube
  • The multi-azimuth processing can be 'targeted' to combine azimuths or isolate specific azimuths as appropriate to optimise data quality and resolution at specific locations in the subsurface

Multi-azimuth 3D is a transparent and robust effort to overcome a fundamental weakness in the seismic method. The acquisition and processing technologies required are conventional, although efficient acquisition technologies will significantly reduce the overall survey cost. PGS has acquired several multi-azimuth surveys with 14 streamers at close separation with Ramform vessels to efficiently maximize data resolution.

Overall, multi-azimuth acquisition and processing has quickly established itself as providing a significant demonstrable return on investment (ROI) when applied to established reservoirs. The Varg field, operated by Pertra, is a case in point. It lies on the Norwegian side of the North Sea. Multi-azimuth 3D data have conquered an historical inability to plan production drilling because of poor resolution seismic data. New data delivered a step change in quality, and two step-out wells were subsequently drilled, prolonging the life of the late-stage field, and contributing more than two billion Norwegian kroner in additional revenues by mid-2004. Several multi-azimuth projects have also been completed by PGS elsewhere in the North Sea and West Africa. In the most innovative effort yet, BP recently completed the acquisition of their Raven multi-azimuth 3D survey in offshore Egypt. Five new azimuths were acquired over an existing (single-azimuth) 3D dataset, yielding six available azimuths to optimise target quality and resolution. A new standard for target illumination will be established.

Target illumination
vs.
acquisition direction

What exactly is 'illumination'? In the strictest sense, every subsurface point at the target should have reflected seismic energy with a uniform distribution of source-receiver offsets and azimuths. If this criteria is satisfied for all target points, then a very high quality, high resolution seismic image of the target should result from processing (Vermeer, 1999). In reality, subsurface illumination is always irregular, and may even be characterised by large 'holes' when complex geology exists between the surface and the target.

The illumination of the target and the sampling of the reflected wavefield are both a function of how the sources and receivers are deployed on the surface during the acquisition experiment. As always, the success of pre-survey planning is constrained by the properties of the Earth at each specific 3D survey location, even when rigorous acquisition standards are applied. In both the land and the seafloor acquisition cases the source and the receivers are completely decoupled. Therefore, they can be located independently of each other. Provided that there are no obstructions to the deployment of the sources and receivers, it is possible in principle to deploy the sources and receivers over a large area, thereby providing a complete illumination of the subsurface target, whilst simultaneously spatially sampling the reflected wavefield very densely.

In the streamer case, the source and the receivers are coupled, and there is consequently far more restriction upon the flexibility of 3D data acquisition. Figure 1 presents a real 3D modelling study from deep water offshore Brunei. The surface plotted is the target surface beneath a structurally complex overburden. The colour scale represents the density of reflections from each point on the subsurface, derived by full-offset 3D ray tracing. Illumination is degraded for both the strike and dip shooting cases, wherever steeply-dipping geology occurs.

Figure 2 presents a 3D modelling study based on the Varg field. The colours represent the illumination density at the target for three different acquisition directions. Each of these results are different. These fundamental observations confirm a weakness in the seismic method: Illumination is never perfect. The colour scale in Figure 3 represents the range of source-receiver offsets being reflected from each subsurface point at the target. It is clear when moving from left to right that the uniformity of illumination quickly builds up from any one shooting direction, to a combination of two acquisition directions, and then finally a combination of three acquisition directions. A clear demonstration is made that a combination of surveys that are acquired in different directions provides the best quality target illumination.

Two new 3D surveys were acquired over Varg, separated by 120° (roughly NW-SE and SW-NE), providing two new azimuths to complement an existing E-W 'dip' survey. Figure 4 demonstrates a profound improvement in data clarity and resolution at the Base Cretaceous target horizon was achieved when all three azimuths were combined in processing to output the final multi-azimuth dataset. The step change in data resolution has significantly revised the reservoir model, relocating several major faults and introducing many local structures affecting reservoir compartmentalisation (Figure 5). As noted earlier, the successful drilling programmes would have been impossible without the new multi-azimuth seismic data.

So how many azimuths are really required to yield a desired improvement in data quality and resolution? 3D seismic modelling via ray tracing will usually yield robust and accurate results when a reasonable 3D model can be built of the reservoir and the overburden. In the Varg case, the coarse original model extracted from the right side of Figure 5 was used to produce the modelling results in Figures 2 and 3, these proving to have provided a reliable guide for the subsequent acquisition surveys. Figure 3 demonstrates that the most significant illumination improvement at Varg arose from the addition of two (orthogonal) surveys, and the addition of a third survey was less significant. Nevertheless, the availability of three azimuths proved useful when attempting to build high quality seismic images around certain faults and important compartmentalising structures within the reservoir. PGS experience shows that the decision on the number of azimuths required is location-specific. The advantage of pre-survey modelling is that discrete locations in the reservoir can be isolated or targeted, such that all possible multi-azimuth acquisition scenarios can be tested, quantified, and calibrated.

Some multi-azimuth operational considerations

Figures 1 - 5 have demonstrated that appropriate shooting strategies can be taken to improve the quality of target illumination in a 3D seismic survey. Improved target illumination will translate to less artefacts in processing, better data quality, higher signal-to-noise ratio, and better resolution (Vermeer, 1999; Figure 4). Processing must be customised to optimise the multi-azimuth data benefits (Hegna & Gaus, 2003). Theory says that pre-stack time migration (PSTM) should be adequate for both isotropic and anisotropic geology, provided that lateral velocity gradients are smooth. More complex geology will necessitate pre-stack depth migration (PSDM) to be used, as with the Varg case. Specific locations can be targeted, such that the optimum combination of different azimuths is used during seismic imaging. This selection will be location-specific, and demonstrates a great power and flexibility inherent in the multi-azimuth methodology.

Multi-azimuth acquisition configurations are typically tailored to optimise survey efficiency. Infill requirements are reduced, as traces can be 'borrowed' from overlapping azimuths. The cost of multi-azimuth will therefore not scale linearly with the number of azimuths. The Varg multi-azimuth survey was acquired with eight streamers at 100 m separation. Several other multi-azimuth surveys have been acquired with 14 streamers at 37.5 m separation and single-source shooting to maximize trace density and optimize 3D spatial sampling.

Conclusions

Multi-azimuth 3D seismic is for exploration and production applications plagued by the effects of poor target illumination during 'conventional' 3D seismic surveys. Target illumination is never ideal, and is an inherent weakness in the standard seismic method. Multi-azimuth 3D uses more than one acquisition direction to improve uniform target illumination, thereby providing a robust geophysical solution to the problem. Several case experiences demonstrate a significant return on investment (ROI), with accelerated drilling successes, prolonged field lives, and significantly improved data clarity, interpretability, and resolution.

Acknowledgements

I thank PGS Marine Geophysical for permission to publish this paper, and Pertra for permission to publish the Varg examples.

References

HEGNA, S., & GAUS, D., 2003. Improved imaging by prestack depth migration of multi-azimuth towed streamer seismic data. Annual Meeting Abstracts, EAGE, Paper C-02.
LONG, A.S., RAMSDEN, C.R.T., & HOFFMANN, J., 2004, On the issue of strike vs. dip streamer shooting for 3D multi-streamer acquisition. Exploration Geophysics 35, 2, 105-110.
REKSNES, P.A., HAUGANE, E., & HEGNA, S., 2002, How PGS created a new image for the Varg field. First Break, 20, 773-777.
VERMEER, G.J.O., 1999, Factors affecting spatial resolution. Geophysics, 64, 942-953.