Automation of Fracture Counts Using Volumetric Seismic Attributes

Guo, Hao¹, Kurt Marfurt¹ (1) University of Houston, Houston, TX

Fracture sets are developed with different intensity and preferred orientations in response to changing stress regimes throughout geologic time. Fracture sets with preferred orientations are often interpreted as the cause of seismic P-wave velocity (Vp) anisotropy and linear features in seismic attribute cubes. Field experiments show that stress orientation can often be measured by P-wave anisotropy. On contrast the lineaments seen in the volumetric curvature attributes are related to paleo-stress through classical models of geologic deformation. Our challenge is how to predict open fractures, given a velocity estimate of present day stress and a curvature estimate of paleo-deformation. To exploit the fracture-intensity and fracture-orientation information carried in these attributes, we develop a program that can automatically count the azimuth distribution of fast P-wave velocity (or the strike of the lineaments) weighted by the amount of anisotropy (or the intensity of lineaments) at any point in the volume for a given analysis window. The output is a series of rose diagrams, which we hope can be related to the variation of fracture intensity and azimuth across space and depth (or geologic time). We find there is a visual similarity between rose diagrams generated from anisotropy and those from curvature. Eventually, the key to this proposed workflow is relationship of anisotropy (lineaments) and fractures to image log breakouts, production and tracer data.