First of a Series: 3-D Design Philosophy

The geometry of onshore 3-D seismic recording grids is based on five parameters: source-station spacing, receiver-station spacing, source-line spacing, receiver-line spacing and recording swath size.

At each prospect, these design parameters can be specified by answering four simple questions:

  • What is the narrowest lateral dimension of the geologic target that is to be imaged?
  • What is the depth of the shallowest target?
  • What is the depth of the deepest target?
  • What stacking fold is required to create an acceptable image at the depth of the principal target?

The first thee questions can be answered if the geology of the prospect is known. The fourth question can be answered by examining local seismic data or by guessing.

Once these four questions are answered, the methodology described in this four-article series allows many onshore 3-D seismic surveys to be designed without having to use commercial design software.

About this series: Some of the more technical aspects of 3-D seismic design, such as ensuring that there are proper azimuth and offset distributions between source-receiver pairs, are not considered in this simple treatise on first principles of 3-D survey design. The intent is to describe a method that allows non-geophysicists to plan simple 3-D surveys – and to recognize flaws in proposed 3-D programs.

The design procedure is diagrammed on figure 1.

Narrow Targets and Stacking Bins

The horizontal resolution provided by a 3-D seismic image is a function of the trace spacing within the 3-D data volume. As the separation between adjacent traces decreases, horizontal resolution increases.

At the conclusion of 3-D data processing, each trace in a migrated 3-D seismic data volume is positioned so that it passes vertically through the midpoint of a small area called a stacking bin or image bin (figure 2). These stacking bins can be square or rectangular, depending on an interpreter’s preferences.

The trace spacing in a given direction across a 3-D image is the same as the horizontal dimension of the stacking bin in that direction. As a result, horizontal resolution is controlled by the areal size of stacking bins.

The imaging objective dictates how small a stacking bin should be. Smaller bins are required if the resolution of small stratigraphic features is the primary imaging objective.

As a general rule, there should be a minimum of three stacking bins – and preferably four bins – across the narrowest feature that needs to be resolved in a 3-D data volume so that the targeted anomaly is expressed on three or four adjacent seismic traces.

A variation in seismic reflection character on three to four adjacent traces is usually noticed by interpreters; anomalous reflection behavior on fewer traces tends to be ignored, or may not even be seen when viewing a 3-D data volume.

For example, if the interpretation objective is to image meandering channels that are as narrow as 100 meters, then stacking bins should have lateral dimensions of approximately 25 meters or less (figure 3), which would cause a 100-meter channel to affect at least four adjacent traces.

One of the first 3-D design parameters to define is the physical size of the stacking bin that should be created. That bin size, in turn, can be determined by first developing a stratigraphic model of the target that is to be imaged and using that model to define the narrowest feature that needs to be seen.

Once this minimum target dimension is defined, stacking bins having lengths and widths that are about one-fourth of this minimum target width should be created if the target is to be recognized in a 3-D data volume.

Source and Receiver Station Spacings

Geophysical texts show that the dimension of a 3-D stacking bin in the direction that receiver lines are deployed is one-half of the receiver-station spacing along the receiver lines, and the dimension of a stacking bin in the direction that source lines are oriented is one-half of the source-station spacing along the source lines.

Thus, when a choice is made about the narrowest target that must be imaged, not only have the dimensions of stacking bins been determined, but source-station and receiver-station spacings also are defined.

Because of the 2-to-1 ratio between station spacings and stacking-bin dimensions, source-station and receiver-station spacings should be one-half (or less) of the narrowest horizontal dimension that needs to be interpreted from 3-D data.

Referring to figure 1, we have now defined parameters – source-station and receiver-station spacings – required for the first calculation step of the 3-D design process.


One appeal of the 3-D seismic design philosophy described here is that the design procedure uses geological concepts to guide the selection of geometrical parameters – the first concept being the physical size of the smallest stratigraphic feature that the seismic data must image.

This non-mathematical approach to 3-D seismic design is intended for non-geophysicists – and it produces good 3-D images for most geologic problems that do not involve severe structural dip.

Where rock units have steep dips or where large changes in rock velocity across fixed horizontal planes, a more rigorous calculation of station spacing should be made using commercial 3-D design software.

Comments (0)

 

Regions and Sections Column - Adrienne Pereira

Adrienne Pereira began working with the AAPG in the Asia Pacific Region in December, 2009.

Geophysical Corner

Geophysical Corner - Bob Hardage
Bob A. Hardage, senior research scientist at the Bureau of Economic Geology, the University of Texas at Austin served as editor of the Geophysical Corner from 2006 to 2011.

The Geophysical Corner is a regular column in the EXPLORER that features geophysical case studies, techniques and application to the petroleum industry.

VIEW COLUMN ARCHIVES