Numerical investigation of a novel hypothesis for fracture termination and crossing, with applications to lost circulation mitigation and hydraulic fracturing

•Hydraulic fracture crossing requires initiation of incipient cracks ahead of the tip.•Initiation may be slowed by poroelastic responses, especially in low permeability.•We perform analytical and numerical calculations that support this hypothesis.•We perform simulations to investigate application for lost circulation mitigation.

We investigate a novel hypothesis regarding the process of hydraulic fracture termination against a preexisting frictional interface. According to current understanding, crossing occurs when small tensile fractures form ahead of the crack tip, on the other side of the frictional interface, before the concentration of stress at the crack tip causes slip along the interface. Slip blunts the concentration of stress at the crack tip and causes termination. Existing crossing criteria assume that the incipient fractures ahead of the crack tip form instantaneously once the effective stress is sufficiently tensile. However, there is a poroelastic response that causes a reduction in pressure in response to opening. This is counteracted by flow into the crack from the surrounding matrix. In very low matrix permeability formations (shale, coalbed methane, etc.), flow of fluid inward from the matrix is slow, and the opening of these incipient fractures may require a non-negligible amount of time. Using the hydro-mechanical discrete fracture network simulator CFRAC, we performed a series of numerical simulations to qualitatively investigate this process. The simulations confirm that poroelastic response could affect incipient fracture initiation and hydraulic fracture crossing. Based on this mechanism, we developed a heuristic modification to an existing crossing criterion. We applied the new criterion to investigate an injection sequence for prevention of lost circulation in fractured, low matrix permeability formations. Lost circulation occurs if wellbore fluid pressure exceeds the minimum principal stress, causing fluid loss due to propagation of a hydraulic fracture. In our proposed injection sequence: (1) injection is performed at high rate to create near wellbore fracture network complexity and then (2) viscous fluid is injected into the newly formed fractures to create resistance to flow. The simulations show that this sequence may be able to mitigate lost circulation and create a stress cage around the wellbore.