This paper presents a semi-analytical model based on the theory of dislocations to predict the reactivation of a natural fault due to the presence of an approaching hydraulic fracture. In particular the model allows the analysis of the stress field along the interface to determine the most probable location of the induced fracture re-initiation. The influence of fault friction angle, fault orientation and pressure in the induced fracture are shown. In absence of fault cohesion, it is shown that the induced fracture cannot, in most of the cases, cross the fault without an offset.
Hydraulic fracture propagation through natural interfaces or discontinuities is an important topic, which one needs to understand to predict or interpret the stimulation of fractured media such as tight gas sandstones, coal bed methane or shales (Anderson 1981, Medlin & Fitch 1988, Daneshy 2003). Several field (mine-backs) and laboratory experiments have revealed the complexity of the fracture pattern (Teufel & Clark 1984, Warpinski & Teufel 1987, Jeffrey et al. 1992, de Pater & Beugelsdijk 2005). One main issue is whether the induced fracture will cross an interface, be arrested at the interface or reinitiate with an offset, as shown in Figure 1. Depending on the application, one wants to avoid fracture arrest or offsetting to maximize fracture width, or ensure fault re-opening to increase the areas of exposed rock surfaces. Previous works using fully couple numerical models have shown the complexity of the mechanisms of propagation, such as the influence of the fault friction coefficient and earth stresses (Keer & Chen 1981, Shaffer et al. 1984, Jeffrey et al. 1987, Zhang & Jeffrey 2006, Zhang et al, in press). But a full numerical analysis does not allow a quick analysis to be used for field application.