A new methodology reduces 4-D seismic acquisition and processing differences and delivers better seismic repeatability and 4-D signal enhancement.

A key objective in time-lapse, or 4-D, projects is to acquire and process seismic data in such a way that the difference between base and repeat surveys only reflects the dynamic changes in the state of the reservoir. At the same time an increase in seismic repeatability between base and monitor surveys can be ensured.

However, in quest of the reservoir's 4-D signal one frequently faces the problem that monitor and baseline surveys do not have the same survey configurations and are differently affected by feathering. Consequently, two or more surveys need to be compared with varying source-receiver azimuths and different sampling in inline and crossline directions. Azimuth differences between base and monitor surveys imply that the wavefield propagated through the overburden has illuminated different parts of the subsurface. Changes in azimuth distribution can therefore lead to degradation of seismic repeatability. In addition, large differences in source-receiver azimuths between base and monitor surveys may obscure the 4-D seismic effects.

A strategy for acquisition of optimal 4-D surveys is comprised of:

• Acquisition with overlap configurations; i.e., the seismic vessel is towing more streamers (adding outer streamers in the spread) than nominally required according to the sail line separation. In fact, the vessel is steered as if it had fewer streamers than actual. This produces an oversampling of the surface fold, making it possible to select matching source-receiver azimuths between multiple surveys during the 4-D seismic data processing phase.
• Accurate steering of the vessel for repeating-shot positions and sailing direction for all lines.
• Dense streamer separation.

The repetition of vessel/source positions in future monitor surveys in an overlap shooting acquisition mode can be more efficient if baseline surveys are also well-conditioned; i.e., infill shooting is minimized and the vessel closely follows the survey's pre-plot sail lines.

High-density 3-D

The ability to tow more (>12) and denser (<165 ft or 50 m) streamers is the ultimate goal in a seismic acquisition experiment. Performing this type of acquisition ensures dense and even sampling in both time and space domains so that aliasing can be avoided and signal-to-noise (S/N) ratio can be increased.
Unfortunately, 3-D streamer acquisition has historically been forced to compromise spatial sampling in the name of efficiency and cost. Figure 1 illustrates a modern seismic vessel which can accommodate massive towing capacities (>16 streamers). These highly efficient vessels have changed the way in which modern seismic data is acquired.

Dual-source shooting has historically dominated streamer acquisition in comparison to single-source shooting. This is because the streamer separation with single-source shooting must be halved to preserve crossline processing fidelity. In the old days of limited streamer capacity, short streamer spreads using small streamer separation were prohibitively expensive to deploy. Recently, the ability to tow 12 to 16 long streamers at 123 ft to 165 ft (37.5 m to 50 m) separation with no loss of efficiency (i.e. no increase in downtime) in a single-source acquisition mode are cost-effective (Hegna et al., 2001; Long et al., 2003). These types of surveys increase fold, improve 3-D spatial resolution, and enhance seismic imaging and data quality.

A synthetic example

In order to investigate the optimal strategy for 4-D acquisition, we first analyze synthetic navigation data. We assume that we have 2? and 10? constant feathering in the base and monitor surveys, respectively. Shot positions are repeated exactly. Thus, changes in azimuth depend on the receiver positions only. The denser we sample in the crossline direction, the less exposed we are to azimuth errors.

In our first example, base and monitor datasets are acquired by a dual-source, eight-streamer configuration with 320-ft (100-m) streamer separation, i.e., 80-ft (25-m) spacing between inlines. In the second example, the 4-D surveys are acquired by an equivalent single source configuration, i.e. 16 streamers with 165-ft (50-m) separation. Figure 2 shows a map with azimuth mismatches for traces in a nominal offset bin at circa 3,600-ft (1,100-m) offset after 3-D binning. The single source configuration offers better azimuth preservation due to the increased number of streamers deployed with a denser separation. In both cases the feathering leads to large azimuth mismatches between base and monitor data along adjacent sail lines.

Overlap shooting

Provided that shot positions and sailing directions are repeated, the overlap shooting mode (additional outer streamers) can be employed in order to minimize the strong 4-D azimuth mismatches.
In order to achieve an optimal azimuth match, it is more important to repeat shot positions than receiver positions. This is because marine streamer acquisition is much more sparsely sampled on the source side than on the receiver side. However, steering for repeat shot positions in the presence of feathering may cause holes in the surface coverage, especially at large offsets. Therefore, it is important to use overlap shooting when repeating shotlines. The effect of not repeating shot positions results in a match in source-receiver azimuths that is significantly degraded.

A case study

In the following we analyze and compare the azimuth match for 4-D surveys at two different fields in the North Sea.

At Field A, a six-streamer dual-source vessel operated in overlap mode by steering as if four streamers were deployed. The base survey was acquired with four streamers. The vessel steering during the monitor survey was based on surface coverage without any attempt to repeat the shot lines of the base survey.

At Field B, a vessel towing 10 streamers acquired monitor data with overlap. In addition, the shot lines of the monitor survey were repeated as closely as possible according to the base survey. The base survey was a normal eight-streamer survey.

A subset of the navigation data covering approximately 20,000 common midpoint (CMP) locations at Field A and Field B has been arbitrarily chosen for a quantitative 4-D azimuth comparison. Note that the 4-D survey at Field B, due to its wider towing spread, is much more exposed to azimuth errors than the Field A survey. However, the following analysis shows that azimuths at Field B have been better preserved because of the improved acquisition strategy.

The navigation data have been sorted into offset bins with a nominal fold of one. Both base and monitor surveys show oversampled areas, areas with a fold higher than one.

The randomly distributed oversampled areas for both base surveys result from a combination of feathering, vessel steering for surface coverage and infill shooting. The fold maps for the monitor surveys, however, show a regular increase of fold at the sail-line boundaries due to the overlap geometry. It can be demonstrated for Field B that the additional data between the sail lines allows searching for trace pairs from base and monitor surveys with almost perfect azimuth match.

4-D-friendly binning

A 4-D-friendly binning procedure is necessary to take full advantage of the acquisition strategy outlined above. The surplus of data in the overlap zones makes it possible to achieve a better match by using selective binning schemes. Three binning approaches are discussed here:

• Independent binning using distance to bin center as the trace-selection criterion for base and monitor data. Azimuth-dependent binning using azimuth closest to the sail-line azimuth as the trace-selection criterion for the monitor survey. The corresponding trace in the base survey is selected based on minimum difference in source-receiver azimuth.
• Simultaneous 4-D binning using minimum difference in source-receiver azimuth between the base and monitor trace as the selection criterion.

Figure 3 shows azimuth-difference maps resulting from these three methods. The left figure is based on independent binning of base and monitor. The observed azimuth differences are generally large. Azimuth-dependent binning of the same 4-D data results in an improved match, but azimuth mismatches at the sail-line boundaries are still visible (middle). Simultaneous 4-D binning (right) reveals the real repeatability potential of this 4-D dataset. It proves that the acquisition applied on Field B (shotline repetition combined with overlap shooting) resulted in proper oversampling, which minimizes the azimuth-difference footprint. The remaining azimuth differences in cross-line direction on the right are related to the large streamer separation as illustrated in Figure 2 (left).

The advantage of this acquisition strategy becomes even clearer when comparing the azimuth-difference map from Field A (overlap only) with the one from Field B (shotline repetition combined with overlap). The acquisition performance with respect to the 4-D source-receiver azimuth differences at Field B is superior.

Conclusion

With a dense streamer separation, source-receiver azimuths can be repeated very accurately. As a result, this methodology reduces 4-D seismic acquisition and processing artefacts and delivers better seismic repeatability and 4-D signal enhancement.

Comments