Florez-Nino, Juan-Mauricio¹, Gary Mavko¹
(1) Stanford University, Stanford, CA

ABSTRACT: Probabilistic Modeling of Fracture Density Constrained by Outcrop Data and Geomechanics

We use the descriptive statistics of fault and fracture systems, collected at an outcrop, to create geostatistical models of joint frequency distribution. Faults and fractures occur at different scales in a hierarchical fashion, controlled by the stratigraphic architecture. Joint frequency has an exponential distribution, where the mean depends not only on the degree of deformation but also on the average bed-thickness. Faults of different hierarchies are characterized by the width of the damage zone and their dimensions, which are constrained by the confining stratigraphic interval. Mean joint frequency within the damage zone is about one order of magnitude higher than the background joint frequency. Correlation length, mean joint frequency and, in some cases, orientation may vary significantly between different fault hierarchies.
We have used sequential gaussian simulation to generate models of joint frequency when all the faults and fractures are parallel. In order to generate models of joint frequency when large faults, small faults and background joints have different orientation, we have used sequential indicator simulation. This technique allowed us to reproduce the impact on anisotropy introduced by the diverse orientation of fault hierarchies and background joints, however it could not reproduce the extensive continuity of large faults. As an alternative we have used a more deterministic approach, where we modeled the distribution of fault zones apart from small faults and background joints.

AAPG Search and Discovery Article #90026©2004 AAPG Annual Meeting, Dallas, Texas, April 18-21, 2004.