Natural fractures, fissures, and microfractures are well-known contributors to production performance of shale reservoirs. Complex fracture geometries can be generated by either small-scale fracturing, DFIT, or large volume fracture stimulation, because the activation of pre-existing natural fractures, fissures, and microfractures plays a significant role on generation of induced hydraulic fractures. Therefore, much more attention should be paid to the model development of DFIT complexity.
In previous work, Chen et al. (2017a) built a complex fracture-network model for after-closure analysis of DIFT by considering natural fractures. Based on that, this paper introduces a generalized model with random fracture geometry caused by natural fractures, fissures, and microfractures. First, the model flexibility is demonstrated by different complex fracture cases, namely opening-fissure fracture network, tree-like fracture network, radial multiple fracture network, and mutually orthogonal fracture network. It is found that the pressure derivative reaches constant level, no matter what the fracture geometry is. Furthermore, the reservoir permeability of field examples from actual DFIT tests in Marcellus shale reservoir is quantified using the log-log diagnostic plots based on the model solutions. Finally, the Nolte G-function is applied to verify the estimated results. We find that the results from the two methods are consistent. This work primarily focuses on quantifying the reservoir permeability, while in the future more efforts will be made to identify the fracture properties using the proposed model.