Anisotropy is often observed in rock masses that contain fractures and affects several properties including flow behavior, which is controlled by how fractures are clustered in space. While Rose diagrams are often used for delineating "fracture sets" in different directions, the role of anisotropy in terms of fracture clustering remains unaddressed. This research attempts to capture clustering anisotropy in network geometries and investigate how it affects the anisotropy of flow patterns in such networks. We employ lacunarity for evaluating directional clustering of fracture networks by modifying the standard “gliding-box” algorithm. Instead of a square-box of increasing size, a rectangular-box of increasing width is used and glided in a direction along which the clustering is captured. Next, flow anisotropy is assessed by considering the fracture continuum model and using a Darcy based streamline simulator to determine “recovery” in a given direction. A set of known fractal-fracture networks and natural maps are analyzed in terms of their clustering and flow anisotropy. The results show that overall fluid recovery tends to be higher in the direction of highly clustered fractures and that clustering anisotropy, even at smaller scales, can possibly influence this recovery curve. These findings imply that DFN models can be improved in terms of predicting flow behavior, if they can take into account directional variations in fracture clustering observed in well-data and reservoir analogs.
Considerable research over the last 25 years has focused on quantifying the heterogeneity of fracture networks. Various geostatistical techniques, ranging from semivariograms (LaPointe and Hudson, 1985; Chiles, 1988) to fractal descriptors (La Pointe, 1988; Berkowitz and Hadad, 1997; Roy et al., 2007), have been applied to this issue. Since fractures result from deformation processes that are inherently directional (Kruhl, 2013), it is to be expected that, in addition, to being heterogeneous, fracture networks will also display some form of anisotropy.