简介：目前在深水进行的勘探成功率非常令人失望，需要新技术对深处的复杂结构进行更精确的成像，从而可以更加准确的布井。

Patterns of progress 

In 1986 Diamond Shamrock drilled through hundreds of feet of Gulf of Mexico salt and found a thousand feet of reservoir-quality rock beneath it ? and the race was on. 

WAZ designs used on three Recent Gulf of Mexico surveys show acquisition geometry as well as relative coverage. (Image courtesy of WesternGeco)

The ability to image complex structures has evolved fairly recently, largely driven by the discovery of reservoir-quality structures lying beneath salts and basalts or in complex carbonates. Facilitated by prestack depth migration (PSDM) techniques, geophysicists mapped subsalt sediments in the deep waters of the Gulf of Mexico in the early 1990s. When the first discovery well was drilled beneath the salt, the conventional wisdom of more than 70 years went out the window.

Almost instantly great pressure was brought to improve our ability to resolve exploration difficulties, not only beneath the salt, but also with regard to complex structures everywhere. PSDM was just the tip of the iceberg that hinted at techniques that could deliver better quality and higher resolution images. And the wide open spaces of the deepwater Gulf of Mexico provided the ideal place to test them.

Early deepwater exploration techniques paid off. Large numbers of deepwater structures have been identified in the Gulf of Mexico, Mexico, Brazil, Nigeria, Angola, Egypt, and Australia as well as smaller structures offshore from a dozen more countries around the world. The drilling industry answered the call with high?performance deep- and ultra deepwater drilling units, and a few notable operators provided the investment capital to launch extensive drilling campaigns. But exploration success rates were disappointing. In the five years spanning 2002 to 2006, the Gulf of Mexico recorded only 25 discoveries for 170 attempts, a ratio of 6.8%. Accordingly, it became clear that improving drilling success rates was a critical driver of continued exploration. Better techniques were required to image the deep, complex structures so well placement could be more precise.

But new techniques alone would not suffice. Technology breakthroughs in both acquisition and processing were required to leverage the improvements promised by innovative shooting designs. Some of the innovations that have proven to be of significant value include calibrated single-sensor data acquisition, shot-to-shot source output calibration, enhanced source and streamer positioning accuracy, and full dynamic streamer steering on both horizontal and vertical planes.

Processing innovations included velocity model-building based on tomography, accurate and efficient migration algorithms, 3-D multiple attenuation techniques, data regularization and interpolation, deepwater velocity correction to compensate for subsea currents and temperature inversions, space-adaptive wavelet processing, and micro-modeling using iterative discrete sparse spike inversion algorithms. And simultaneous inversion to rock properties provided valuable information used by drillers and reservoir engineers alike.

With capable acquisition equipment and processing technology keeping pace, explorationists turned to acquisition designs to improve subsalt image clarity. Several highly effective designs evolved and have been described in the literature. The first two tried were wide-azimuth (WAZ) and multi-azimuth (MAZ).

Wide-azimuth acquisition

Particularly suitable for large-area 3-D acquisition, WAZ uses different combinations of source boats and receiver boats (which can also tow sources). The surveys illuminate deep structures using near and far offsets, both inline and crossline. Hydrophone interval, streamer spacing, and streamer length can be varied to achieve the desired shot-receiver density as well as the numbers of source boats and receiver boats used. Multiple source boat positioning provides the desired inline offsets. 

On Shell’s Friesian project, a single Q-Marine receiver boat was combined with two source boats. Successive passes were made with one source boat positioned ahead of and one behind the streamer spread. The multiclient E-Octopus project that has currently surveyed 951 blocks in the Gulf of Mexico was shot in four phases. WesternGeco is in the process of acquiring a further 229 blocks for a total of 1,180 blocks. Plans are already being made for phases 5,6, and 7.

Phase 1 of E-Octopus used a single Q-Marine source/receiver boat and two laterally offset source boats. In Phase 2, an additional source/receiver boat was employed to speed up the acquisition. As can be seen by the polar plots in Figure 1, a much wider azimuth was achieved in Phase 2, quite similar to the one achieved in the Friesian project.

The Friesian project was a classic approach that focused on the desired result, then worked backward to come up with the acquisition design. The objective was to fully illuminate the subsurface, optimally suppress inherent noise, and build a database that was easy to process. Company geoscientists decided they could best achieve those goals if they acquired a range of source/receiver azimuths while maintaining a uniform sampling regime to facilitate processing. They also wanted to ensure full area receiver coverage for each shot to allow for efficient wave equation migration with the highest possible trace density.

Using two source boats and a single receiver boat, they conceived an innovative way to position the source boats for maximum efficiency and overlap. Called the FLIP-450 technique, it called for the lead source boat to be positioned just ahead of the starboard streamer array while the tail source boat was offset 1,476 ft (450 m) to the left toward the center of the streamer spread. Subsequent streamer passes were rolled 2,953 ft (900 m), and on consecutive source positions the inline offset pattern would flip from front to back. The design achieved results close to the desired full-area receiver coverage but at greatly reduced cost and complexity.

The biggest challenge in subsalt imaging is building the velocity model. Shell geoscientists believe that well-sampled prestack volumes representing a wide range of azimuths are required to illuminate the salt structures. According to Shell, the biggest benefit of WAZ is noise suppression, which greatly improves processing effectiveness and subsequent interpretation.
Multi-azimuth and rich-azimuth acquisition

Believed to be an excellent way to image a point structure such as a salt diapir, the MAZ technique can provide high-fidelity images from all angles. Like an art connoisseur surveying a fine sculpture, the survey vessel approaches the structure from several angles, usually in 60° intervals. In the simplest sense, a single source/receiver boat can be used. MAZ surveys can illuminate the flanks of the structure and areas under the overhang, which are known to form closure for multiple fault traps.

On BHP Billiton’s Shenzi project, multiple source boats were used to provide an even more comprehensive technique called a rich-azimuth (RAZ) survey. RAZ acquisition is a combination of multi-azimuth and wide-azimuth acquisition. The results were clearly superior to conventional 3-D illumination of the key field structure. With source boats on either flank of the Q-Marine source/ receiver boat, 100% coverage with greater offset was achieved. According to the operator, wells on the structure are estimated to cost about US $100 million each, so an investment in high-quality seismic was a fundamental step.

The Shenzi project offered an opportunity for WesternGeco to try a new technique. In marine surveying, considerable time is required to make a line change ? the term that describes turning the survey vessel and streamer array through 180° in a wide arc to resume surveying in the opposite direction. With conventional surveys, acquisition is suspended during line changes because the long streamers converge as they make the turn, sometimes even becoming tangled with one another. But with dynamic steerable streamer technology it was believed that the streamers could be maintained parallel and at the appropriate intervals. Ten 23,000-ft (7,000-m) cables were deployed in a spread 3,937 ft (1,200 m) wide yielding a 7,874-ft (2,400-m) crossline offset in three different azimuths. The combination of enhanced source and streamer positioning accuracy and full dynamic streamer steering enabled high-quality surveys to be acquired with turning. The efficiency implications of this technique are significant.

Focus on efficiency

With quality enabled through WAZ and RAZ techniques, WesternGeco addressed efficiency. The Shenzi line change experiment proved that high-quality surveys could be obtained while turning. Why not use turning to create an infinite number of shot/receiver positions resulting in the most comprehensive illumination ever achieved? By sailing in overlapping circles, a single source/receiver boat can produce unprecedented quality images with superior noise suppression at double the shot density. Perfect for high illumination of specific targets, the technique is called coil-shooting.

Tested in the Gulf of Mexico and the Black Sea, coil-shooting has proven to be highly efficient. The main reason is that the vessel is continually surveying ? there are no non-productive intervals for line changes. On a simulated survey comprising 350 sq miles (900 sq km), a four-vessel WAZ survey acquired 160,000 shots in 62 full-production days, whereas a single vessel using coil-shooting acquires 321,706 shots in 61 full-production days. Coherent noise attenuation, a benefit of WAZ, is expected to be comparable due to the larger number of shots, higher fold, and wider range of azimuths.

Coil-shooting results compare quite favorably to WAZ and are significantly better than conventional acquisition technology in subsalt applications. For very large-area surveys, WAZ would likely be the choice, but for superior illumination of a specific target area in a deepwater subsalt environment or one with complex structures such as salt diapirs or reef plays, coil-shooting is the pattern of progress.