Statistical Curvature Analysis Techniques - SCAT Diagrams
Traditional dipmeter analysis techniques used for structural analysis involve pattern recognition on the dip arrow, or tadpole, plot. While this approach can be learned with study and practice, there are other approaches that can be applied. Alternatives to the conventional arrow plots have been proposed, mainly because of the effects of statistical variations and ambiguous patterns which sometimes make arrow plots hard to use.

The most successful technique is called statistical curvature analysis, better known as SCAT. The method lends itself to interactive computer programming, and was described by C.A. Bengtson in "Statistical Curvature Analysis Techniques for Structural Interpretation of Dipmeter Data", published in AAPG Bulletin in 1981. The paper was also printed in Oil and Gas Journal, June 1980 and in Geobyte, May 1988.

Microcomputer programs for analyzing dipmeter data in this way were presented by Robert Elphick in the May 1988 and March 1989 issues of Geobyte. These programs do not seem to be available from the major service companies.

SCAT is based on four unfamiliar, but empirically well verified, geometric concepts:
   1. structural curvature
   2. transverse and longitudinal structural directions
   3. special points on dip profiles
   4. dip isogons or trend lines

Flat or dipping planes have zero or planar curvature. Horizontal or plunging folds have one degree of curvature. Doubly plunging folds have two. Drag and rollover on faults have structural curvature and can be analyzed in the same way as folds. Illustrations of typical surfaces and their dip angle vs dip azimuth plots are shown in below.

Dip angle vs azimuth plots - basic shapes

The obvious difference between SCAT and the conventional approach is that SCAT uses the dip angle vs dip azimuth plot plus four other machine plotted dip vs depth displays, whereas the conventional method relies on an all purpose display, the arrow plot, augmented by azimuth frequency plots over selected intervals.

The five plots used in SCAT are:
   1. dip angle vs dip azimuth
   2. dip azimuth vs depth
   3. dip angle vs depth
   4. transverse section dip angle vs depth
   5. longitudinal section dip angle vs depth

The patterns on dip angle vs dip azimuth plots may be simple or complex. However, they are usually simpler and never more complex than patterns on arrow plots.

Arrow plots show complex patterns when a well crosses a crestal plane, but transverse dip component plots show smooth trend lines that cross the zero dip axis. Because angle of dip on an arrow plot is neither positive nor negative, there is no chance for a negative scatter to cancel positive scatter in a flat dip situation. Therefore, a zone of zero dip is falsely perceived as a zone of a few degrees average dip with varying dip azimuth. On a dip component vs depth plot, however, half of the points will fall to the right of the zero dip axis and half to the left, correctly indicating zero average dip.

SCAT resolves the data into mutually perpendicular transverse and longitudinal (or T- and L-direction) components, using the dip rotation arithmetic described elsewhere in this Handbook The T-direction is defined as the direction of cross section through the well that shows the greatest structural change, and the L-direction as the direction that shows the least structural change. These directions are chosen from the locations of the maximum and minimum dip angle scatter on the dip angle vs azimuth plot, marked T and L. They are usually orthogonal directions and can be picked by eye or by statistical analysis.

Average L-direction component of dip is zero for planar and nonplunging fold settings and equal to the angle of plunge for plunging fold settings. On plunge reversal settings the average L-direction component of dip shows a reversal of dip (and hence plunge) with depth. The only exceptions occur in wells cut by cross faults. However, longitudinal dip component plots may show considerable scatter in zones of steep dip.

The shape of the statistical trend line on a transverse dip vs depth plot defines the bedding curvature on a transverse cross section. A trend line conforming to constant dip indicates planar curvature. A smoothly curved trend line with no bends or reversals indicates uniform or smoothly varying curvature, whereas a trend line with bends or reversals will show one or more of eight mathematically definable patterns or special points.

Six of these points serve to locate and identify structural surfaces (axial planes, kink planes, inflection planes, secondary inflection planes, minimum curvature planes, and zero strain boundaries) that intersect the well, and two serve to locate dip-slip faults, distinguishing faults that dip to the right from faults that dip to the left.

Finally, it should be stressed that SCAT has the capacity to find the bearing and plunge of crestal and trough lines of folds, the strike and dip of crestal, axial, and inflection planes of folds, and the strike and direction of dip of dip-slip faults. Dip arrow plots do not handle this function very well.

The concept that there are only a few types of structural curvature greatly simplifies interpretation. Beds are either planar or curved; if planar, the beds are either horizontal or dipping. A zero dip homocline shows no structural change in any direction and hence has no T- or L-directions. In the low and higher homoclinal dip settings (Figures 32.02 bottom, 32.03 top) the T-direction parallels the dip and the L-direction parallels the strike. Patterns on T, L, and azimuth vs depth plots are vertical and a maximum density of points will occur at the average regional dip on the dip vs azimuth plot.

SCAT plots for zero dip setting

SCAT plots for homocline and fold settings

If the beds are curved, they are either singly or doubly curved. If single curved, their crestal or trough lines are either horizontal or plunging at a constant angle. In either situation, the T-direction is perpendicular to the crestal or trough lines and the L-direction is parallel. The L component graph will be vertical. The others will be curved. The depth of crestal, axial, and inflection planes are found by observation of the bends in the trends.

SCAT plots for plunging fold settings

If the beds are doubly curved, their structure contours are either elliptical or circular in plan. If elliptical, their geometry can be approximated by two singly curved plunges joined by a non-plunging central sector, in which case the T-direction is perpendicular to the crestal or trough lines and the L-direction parallels the long dimension. If the structure contours are circular, the transverse directions will converge radially toward the center, and the longitudinal directions will be disposed circumferentially around the center. L-component patterns also have bends.

SCAT plots through faults show the pattern of the structural setting around the fault and the drag is superimposed on it. The fault usually creates a cusp on the transverse dip section, pointing in the direction of the dip of the fault for normal faults and opposite to the dip for reverse faults. (Figures 32.05 and 32.06). The drag patterns are quite distinctive on SCAT plots and help to differentiate faults from folds. Rollover creates a half cusp pattern. These patterns are similar to red and blue patterns seen on dip arrow plots of faults.

SCAT plots for fault settings

More SCAT plots for fault settings