Peter Whiting

Variable Depth Streamer - The New Broadband Acquisition System

The importance of recording the full range of frequencies (low as well as high) is widely accepted. High-fidelity, low-frequency data provides better penetration for the clear imaging of deep targets, as well as providing greater stability in seismic inversion. Broader bandwidths produce sharper wavelets and both low and high frequencies are required for high-resolution imaging of important features such as thin beds and stratigraphic traps. The industry has been facing many issues that have limited the performance of marine seismic surveys with respect to bandwidth. Among them, we find mechanical and acoustic noise, source and receiver ghosts and attenuation with depth. Until recently, conventional de-ghosting was found to be sub-optimal. Thanks to recent advances in technology and also in operational capabilities, we have seen several improvements, in particular with the use of solid streamers, deep towing and notch diversity. We describe a different technique to achieve broadband marine streamer data. The proposed solution is a new combination of streamer equipment, novel streamer towing techniques, and a new de-ghosting and imaging technology. The technique takes full advantage of the low noise and low-frequency response of the new generation of solid streamers (see Figure 1). It then uses receiver notch diversity to yield a broadband spectrum. Conventional acquisition (see Figure 2) with its receiver ghost represents a tuned receiving array which enhances some frequencies and completely cancels others (at the ghost notches). In variable depth streamer acquisition (see Figure 3) a variable depth streamer is used so that the receiving array is detuned and receives all frequencies. As a result, the method creates an exceptionally sharp and clean wavelet for interpretation. It can be optimized for different water depths, target depths and desired output spectra. Figure 4 shows how the variable depth configuration improves low frequency response (it is the average streamer depth that is key parameter here), while at the same time using notch diversity to avoid higher frequency notch problems. This approach to towed streamer broadband seismic is particularly efficient, flexible and customizable for a range of environments and applications. The acquisition parameters such as variable depth streamer profile, maximum streamer depth and source depth can be tuned to provide the maximum possible bandwidth for a given geological setting and water depth. In particular this technique can take full advantage of towing solid streamers at what are currently considered as extreme depths to benefit from the improved low-frequency response of the hydrophones and reduced sea-state noise. To date a variety of test lines have been acquired in different settings with streamer maximum depths as large as 60m. The novel approach to deghosting is fully 3D. It makes no 2D assumptions and has no limitations in the cross-line direction making it suitable for wide-azimuth as well as 3D surveys. This flexibility means that the technique can be used for a range of applications. The increase in penetration from the extension of the bandwidth at the low end will benefit the imaging of deep targets and those below complex overburdens. Shallow targets (such as shallow hazards) will benefit from the fully from the total bandwidth available and recordable. Recent trials have achieved usable bandwidths between 2.5 and 150 Hz.

Reflection Tomography Without Picking

Reflection tomography can be implemented as an algorithm that does not required manual traveltime picking and leads to reliable interval velocity models. Traveltime information can be automatically picked for any locally coherent reflection event, without the need for explicitly defined reflectors. Entropy constraints and staged smoothing help to reduce spurious velocity anomalies and to avoid local minima. A field data example shows that this algorithm can significantly improve an initial interval velocity model without any manual picking or interpretation. The velocity model obtained from this reflection tomography algorithm resulted in a much improved prestack depth migration.

Practical reflection tomography and maximum entropy imaging

Ever since tomographic procedures were introduced to reflection seismology, a lot of research effort has been expended trying to find simple and efficient algorithms for extracting the subsurface velocity information from the pre-stack traveltimes. These efforts have either simplified the conventional algorithm for traveltime tomography or completely reformulated it into an easier to use structure. Improved refectiontomography requires simpler and easier traveltime data picking, as well as methods for avoiding solution image non-uniqueness and local minima. These requirements are fulfilled by a method of raytracing that does not require reflectors, the incorporation of the maximum entropy principle for the selection of the solution image, and by stages of decreasing smoothing to help avoid local minima. The maximum entropy principle has been extensively used in other disciplines and ensures that no correlations exist between pixels that are not required by the traveltime data, as well as providing a unique and consistent image. These changes or upgrades do not simplify the basic inverse problem in any way, but do make it easier to obtain the traveltime data and make the resulting solution images more reliable.

Harriet : a 3D imaging case history.

Summary The Harriet field is late in its life and requires improved imaging to take advantage of infill/attic opportunities. The struc turing in the area is apparently simple, but the seismic image is affected by a noise cone, even after prestack time migration, created b y faulting and shallow velocity variations. Post stack depth migration resulted in mild improvement. Prestack depth migration wa s used effectively to improve the imaging. Automated reflection tomography was successful in further improving the prestack depth migration results.