Did You Forget the Measurement Uncertainty Doing Your Connectivity Analysis?

Many engineering applications in crystalline rocks use fracture intercepts mapped in boreholes as foundation. From this mapping the distributions of intensity, spatiality and orientation can be inferred. These three distributions combined with the size distribution steer the connectivity of the fracture network and hence the nature of groundwater flow and transport of solutes. This study, however, only focuses on the impact that the orientation uncertainty has on the connectivity. The orientation of a fracture intersecting a borehole can be calculated using four angles, each afflicted with uncertainty. These uncertainties are used to distinguish between natural variability and uncertainty using a ?² test of a contingency table of fracture poles. Two DFN models are developed, the rock mass fracture and the “measured” model, and the differences in connectivity between the models are analyzed. The rock mass fracture model have 30% more connected fractures, 60% more connected fracture area and is more elongated than the “measured” model at the connectivity level when the first fracture of the two clusters hits any boundary of the modelling cube.