(295d) Study of Injected Water Recovery Based on New Discrete Fracture Network Approach

Due to the significant advances in drilling and fracturing technologies, the United States has seen a significant surge in unconventional oil and gas production. Despite this unqualified success, these technologies have spurred socioeconomic concerns over the amount of fresh water used in hydraulic fracturing treatments that are implemented. Recycling fracturing fluids for future fracture treatments is environmentally desirable and also leads to savings to operators. However, currently, less than one-half of injected hydraulic fracturing fluid is recovered. In order to increase the recovery of injected water after treatments, it is necessary to understand the fate of the injection fluids and improving load recovery. This has been done by tracing fluid chemical composition changes, analyzing field operations data, simulating reservoir systems and so on. In this paper, the latter tactic is adopted in order to investigate injected water recovery mechanisms by performing numerical simulations. The computation platform is a new Discrete Fracture Network (DFN) methodology.

Existing literature cites using commercial reservoir simulation software to simulate post-treatment water recovery. Some of those studies did not capture non-orthogonal, complex fracture geometries of fractures; this is a disadvantage of finite difference simulators. The protocol adopted was to implement a new DFN to a Control Volume Finite Element Method (CVFEM) reservoir simulator. This new DFN methodology has several advantages over many traditional DFN protocols: (1) fracture segments have their own control volumes, rock-fluid properties, pressures, and saturations, and (2) fracture geometries are independent of matrix discretization. The simulator has been verified by comparing simulation results with both Miller’s analytical solution. The new simulator has been applied to analyze the recovery of injected water in a multiphase, mixed wettability environment. This has included:

• Demonstrating the role of different rock-fluid properties such as relative permeability on the water recovery factor, and,
• Comparing the impact of various fracture geometries - such as orthogonal fracture systems (characteristic of dual porosity simulations) and non-orthogonal fracture systems where self-propping may be generated (at least temporarily) due to shearing.