Fractures are present in almost all hydrocarbon reservoirs, but it is only when fractures form an interconnected network that their effect on fluid flow becomes important. Fractures not only enhance the overall permeability of many reservoirs, they also create significant permeability anisotropy. Knowledge of the orientation and magnitude of the horizontal permeability anisotropy has significant economic importance in developing and managing a reservoir. Such knowledge allows optimization of (1) location of production wells for maximum recovery and drainage of the reservoir with the fewest number of wells, and (2) placement of waterflood injection wells to prevent early breakthrough in producing wells, thereby achieving optimum sweep efficiency and maximum oil recovery.

In order to assess the role of natural fractures on hydrocarbon production and reservoir permeability anisotropy, characterization of naturally fractured reservoirs has focused primarily on the distribution and orientation of fractures in a reservoir. For reservoirs with only one fracture set (e.g., regional vertical extension fractures) the horizontal direction of preferred fluid flow is parallel to the trend of the fractures. For reservoirs with more than one set of fractures in different orientations it is often assumed that the intensity of fracturing controls reservoir permeability anisotropy and the maximum permeability direction is closely aligned with the dominant fracture trend. Although this can be demonstrated to apply in many simple geologic settings, this predictive concept must commonly be modified for stress changes caused by the post-fracture geologic history of the reservoir, including local variations in stress magnitude and orientation caused by geologic structures. In cases where the local stresses and fractures are superimposed on regional fractures, fractures that are parallel to the in-situ maximum horizontal stress may provide the dominant control on reservoir permeability anisotropy, because fractures can occur even if the stress-parallel fractures are significantly fewer in number than fractures that trend oblique to the maximum horizontal stress, especially where the stress anisotropy is high. The set of fractures that is open and conductive may change with position around a structure as a function of local stress variations.

Prediction of reservoir permeability anisotropy must include knowledge of the subsurface fracture trends in conjunction with knowledge of the principal in-situ stress direction and magnitudes. Reservoir permeability anisotropy may also change over the life of a reservoir because perturbations in stress state, caused by drilling, production, and waterflood activities, create changes in the three-dimensional effective stress field, and thus in fracture conductivity. Steeply-dipping fractures aligned with the local maximum horizontal stress will have the smallest decline in conductivity as the reservoir is produced.

These conclusions are supported by cores analyses, in-situ stress measurements, well tests, and production histories of tight-gas, naturally-fractured sandstone reservoirs in Colorado and New Mexico and naturally-fractured clastic and carbonate oil reservoirs in West Texas.