Joint Size Estimation Using Joint Traces on Borehole Walls

One approach to characterizing subsurface joint populations is to assume
surface joint patterns are representative of joints at depth. Yet, many times, either the analogous surface joints are unexposed, or absent because surface rocks did not experience the same deformation history. The alternative of direct subsurface characterization has been limited by joints not being resolvable in seismic data and borehole data not yielding fracture size. The present approach uses the subsurface geometry of joint/borehole intersections to estimate mean joint size (mean joint length and width) and aspect ratio (joint length to width ratio), and presents a new method for determining the volumetric joint intensity as estimated with cycloidal scanline samples. This study focused on bed-normal joints in sedimentary rocks that typically terminate at bedding surfaces, have bed-parallel lengths greater than or equal to the bed-perpendicular lengths, and are rectangular. Rectangular joint/borehole
intersections have six geometries: complete, long-edge, short-edge, corner, end, and pierced, which are differentiated by completeness of borehole intersection and joint trace position on the borehole wall as a function of borehole and joint orientation. The counts for the intersection geometries are used to estimate mean joint size and aspect ratio. The approach yielded accurate mean size estimates for synthetic trace populations. Based on this success, the estimators were applied to borehole joint populations from FMI (Formation MicroImager) data logs in the Mesaverde Group of the Piceance Basin, Colorado. Subsurface estimates yielded bed-parallel and perpendicular lengths greater than for exposed joints along the basin perimeter, but the surface and subsurface shared small aspect ratios of ~1:1. The difference in estimated size may reflect differences in deformation history between the basin center and perimeter.