In the June EXPLORER we discussed the idea that the best overall 3-D seismic survey is not necessarily the one with the best quality data. Nor does it have to be the one with long offset data from all azimuths.
The best survey really depends on balancing a combination of factors -- in particular, subsurface geology and economic objectives.
For some projects, wide-azimuth data is a necessity; for others, it can be more of a liability than an asset.
The critical issue is to record seismic data that is "good enough" to image the geology and still meet the economic requirements of the user. This is accomplished by recognizing the important role of survey design in the planning process.
This month we look at offset distribution plots and offset-limited fold plots from several different wide-azimuth designs. Then, we will compare these plots to similar plots from a typical narrow-azimuth design.
This comparison will reveal some of the adverse effects that can result from wide-azimuth shooting.
Design Comparison: Offset Distribution
In June we discussed the importance of source-to-detector offset distribution for each individual cell. We concluded that for any given fold and bin size, offset distribution is the single most important design attribute, especially when it comes to processing and interpreting the final data volume.
One of the best ways to display this offset information is with a trace offset scatter plot -- also known as a "necklace plot," which displays source-to-detector offset distances (along the vertical axis) for every pre-stack trace that belongs within a particular cell.
Adjacent cells are indicated along the horizontal axis, so that entire cell-lines can be examined at one time. Gaps in offset-domain coverage appear as voids in a pattern of overlapping "necklaces."
The larger the void is, the greater the likelihood of noticeable artifacts in the processed data.
In June we also discussed four different 3-D survey designs named A, B, C and D. Figures 1A and 1B are necklace plots that correspond to designs A and B, respectively.
Recall that design A is the narrow-azimuth survey, where the cross-line maximum offset is only about 40 percent of the in-line maximum. Design B, on the other hand, has in-line and cross-line maximum offsets that are approximately equal to each other.
Note that even though designs A and B produce the same fold, the offset distribution for the wide design (figure 1B) is markedly poorer. The same observation also holds true for wide azimuth design C (figure 1C).
In both cases, near and mid-range offsets have been sacrificed in order to achieve large cross-line offsets. As a result, the data volume produced by either design B or design C is likely to be inferior to the volume produced from A -- the narrow design.
Of the three wide-azimuth designs modeled, only design D has better offset distribution (figure 1D) than design A. However, the D design also has more than two and a half times the fold of A, and that extra fold doesn't come free.
The cost of acquiring design D will be substantially higher than any of the other three designs.
Design Comparison: Shallow Fold
In addition to having poor offset distribution, the ability of designs B and C to image shallow events is degraded. We can see this degradation by examining fold plots that have been offset-limited to source-to-detector distances of 5,000 feet or less (figures 2A-D). Limiting the offsets to 5,000 feet or less is representative of the offset mute that is applied to shallow data by the data processors.
For this example, we will consider geologic depths of about four to six thousand feet to be "shallow."
Although the nominal fold for wide-azimuth designs B and C is about the same as narrow-azimuth design A, the offset-restricted fold is quite different. Figure 2A shows offset-restricted fold for design A ranges from 10 to 14, whereas the wide designs B and C (figures 2B and 2C) only have four to eight traces per cell. This means the ability to accurately map a shallow, secondary objective, or to use a shallow marker horizon for isochron mapping, probably will be compromised by using either design B or C.
Only design D achieves wide-azimuth data and effective imaging of shallow events (see figure 2D).
Unfortunately, as we mentioned before, design D will cost more to acquire than any of the other three design options.
Conclusions
The point of this article is not to suggest that designs B, C or D are necessarily better -- or worse -- than design A.
Rather, the point is to call attention to the fact that those extra azimuths are going to cost you in one way or another. Either the price of your seismic survey will go up, or the offset distribution and shallow imaging will deteriorate, or both.
Therefore, you must carefully weigh the pluses against the minuses in the final seismic subsurface image.
What are you getting?
What are you losing?
What will it cost?