Even though the first 3-D seismic survey
was acquired almost 40 years ago, it's been in only the last 15
years that 3-D has evolved from an R&D project for major oil
companies to a "commodity" tool that is almost ubiquitous. Accompanying
that evolution has been an improvement in the hardware and software
necessary to design, acquire, process and interpret the resulting
3-D data as efficiently as possible.
Despite the broad acceptance of 3-D seismic, no clear
standard for survey design has emerged. Nor should one be expected.
The best survey is always a function of the geology that needs to
be imaged. As long as the subsurface of the earth is not "standardized,"
there can be no "standard design."
Furthermore, most users aren't just interested in
the best data quality possible; they want the best overall survey.
The difference between the two is that the best survey must also
consider economic and surface issues.
Ultimately, a successful 3-D survey is one that gathers
"good enough" data — good enough, that is, to meet the economic
demands of our industry.
However, in any endeavor that lacks standardization,
there are bound to be a few eight tracks and BetaMaxes.
Therefore, it is worthwhile to take a look at some
of the more common misconceptions that can impact the success of
a 3-D survey.
Wide Azimuth 3-D Equals 'True' 3-D?
There is no short and simple answer to the question
of optimum source — to-detector azimuth. Intuitively, a wide-azimuth
survey that collects long offset data from all directions might
seem to be better — but this isn't always the case.
In fact, most early 3-D seismic surveys were narrow
azimuth, although it was probably a matter of necessity as much
as intentional design. In basins with moderate — to-deep objectives,
the number of channels in the recording system restricted the contractors'
ability to economically acquire wide-azimuth seismic data.
However, most of these early surveys were "good enough"
to be considered successful, or if they weren't, it probably wasn't
the lack of azimuth that caused them to fail.
For deep geologic objectives, equipment limitations
can still exist. Achieving long offsets in the cross-line direction
requires either very widely spaced receiver lines or a lot of lines
in the active recording patch.
Before choosing a wide-azimuth design, a question
that must be asked is how will these different azimuths be used?
If pre-stack, azimuthally dependent analysis of the
data is planned (see, for example, October-November 2002 GeophysicalCorner), then wide-azimuth data is absolutely necessary.
If not, designing a survey to record long offsets
in all directions can easily create more problems than it solves.
Three Different Wide-Azimuth Designs
To help understand the implications of wide-azimuth
shooting, we will compare offset distribution plots from a standard
narrow-azimuth geometry (figure 1, design
A) to three different wide-azimuth designs (B, C and D). But before
we do, let's take a careful look at each of the four different acquisition
strategies.
For all four surveys we will assume a maximum usable
offset of 10-11,000 feet. Other key design parameters are listed
in tables 1 and 2. In particular, notice the "Maximum Cross-Line
Offset" values listed in table 2.
Table 1: Design Parameters
|
|
Design
|
Recording Patch
|
Receiver Line Spacing
|
Source Line Spacing
|
Nominal Fold
|
|
A
|
10 lines of
96 receivers
|
880 feet
|
1,760 feet
|
30
|
|
B
|
10 lines of
96 receivers
|
2,200 feet
|
1,760 feet
|
30
|
|
C
|
24 lines of
96 receivers
|
880 feet
|
3,960 feet
|
32
|
|
D
|
24 lines of
96 receivers
|
880 feet
|
1,760 feet
|
72
|
Table 2: Design Attributes
|
|
Design
|
Maximum
In-Line Offset
|
Maximum
Cross-line Offset
|
Relative
Cost
|
|
A
|
10,450 feet
|
4,290 feet
|
$$
|
|
B
|
10,450 feet
|
10,890 feet
|
$$
|
|
C
|
10,450 feet
|
10,450 feet
|
$$
|
|
D
|
10,450 feet
|
10,450 feet
|
$$
|
As you can see in figure 2, wide-azimuth
design B has greater cross-line offset than narrow-azimuth design
A (figure 1), despite having the same
number of receiver lines, channels and fold. It does this by using
a receiver line spacing that is more than twice the spacing used
for design A.
Design C (figure 3) on
the other hand, has the same receiver line spacing as A (the narrow
design), but uses 24 lines in its patch geometry to achieve the
added width. However, to keep the fold (and cost) about the same
as that of the narrow design, source line spacing for C has more
than doubled.
Finally, we get to design D — the "best" of the
wide designs. It uses the same source and receiver line spacing
as the narrow plan. The major design difference is in its recording
patch — 24 lines of 96 channels versus only 10 lines for A.
As a result, the fold produced by design D will be
more than twice that of the other surveys.
There is one other difference between these two designs:
relative cost. Design D will cost more to acquire, because significantly
more recording equipment will be needed.
The Importance of Offset Distribution
For any particular 3-D survey design, a wide range
of attribute plots can be easily produced and examined. However,
for any given fold, the attribute that will have the most impact
on data quality is offset distribution.
The potential problems created by poor (irregular)
offset distribution are numerous, and in some cases the damage is
irreparable by even the cleverest data processor.
These problems might include (but aren't limited
to) the following processing related issues:
- DMO (Dip Move Out) artifacts.
- Poorly resolved surface-consistent statics solutions.
- Poorly resolved refraction statics solutions.
- Inferior, or highly variable stack attenuation of coherent
noise.
- Degraded AVO analyses.
- Increased appearance of an acquisition footprint.
- Increased difficulty estimating correct processing velocities.
Certainly, not all surveys with poor offset distribution
will be ruined by problems such as these, but it is better to address
them during the design phase than after the data are acquired.
Next month, therefore, we will look at offset distribution
plots for each design mentioned above.