ASSESSING FRACTURE CHARACTERISTICS FROM BOREHOLE IMAGE LOGS USING VARIATIONS OF THE AZIMUTH VERSUS TRAVERSE DISTANCE (AVTD) METHOD AND STATISTICAL CURVATURE ANALYSIS TECHNIQUE (SCAT)

The azimuth versus traverse distance (AVTD) method and the statistical curvature analysis technique (SCAT) are used to analyze fracture data derived from borehole image logs. The AVTD method was originally designed to assess sub-vertical fractures on continuous outcrop whereas SCAT was originally developed to analyze structural features (e.g., folded strata) derived from borehole logs. Variations of these two methods are conjunctively used here to display fracture data derived from optical and acoustic televiewer logs.

Fracture data derived from borehole image logs are first plotted in AVTD plots before the data are visually enhanced by using an algorithm that smoothes two-dimensional histograms of the scatterplots. The resulting smoothed density plots are identical to the azimuth vs. depth plot of SCAT except that the original SCAT does not involve contouring fracture data. The smoothed AVTD plot differs from the original AVTD method in that the fracture data are plotted against depth rather than distance. Other plots from SCAT are also used to analyze the fracture data; fracture dip is first plotted against depth and strike followed visual enhancement and classification.

Other fracture characteristics derived from borehole logs can also be analyzed. For example, apparent fracture aperture can be plotted against depth, azimuth or dip to investigate the distributions of fracture characteristics. The scatterplots created in this manner can also be visually enhanced to highlight major trends.

The method proposed here is not intended to replace traditional methods used to assess fracture data from borehole image logs. Rather, the method is intended to provide additional means for assessing fracture pattern distributions that may not be apparent in traditional techniques. For example, equal area projections are useful for assessing general trends in data because they lump data from different depths. However, local variations in fracture distributions in the subsurface may not appear on these projections, but may show up on azimuth vs. depth, and dip vs. depth plots. The major advantage of the method outlined in this paper is that it enhances the value of the results derived from borehole surveys that ultimately lead to improved understanding of fracture distributions in the subsurface.