Numerical Investigation of Fracture Network Formation under Multiple Wells

Abstract

A two-step fracturing method is proposed to investigate the hydraulic fracture evolution behavior and the process of complex fracture network formation under multiple wells. Simulations are conducted with Rock Failure Process Analysis code. Heterogeneity and permeability of the rocks are considered in this study. In Step 1, the influence of an asymmetric pressure gradient on the fracture evolution is simulated, and an artificial structural plane is formed. The simulation results reflect the macroscopic fracture evolution induced by mesoscopic failure; these results agree well with the characteristics of the experiments. Step 2, which is based on the first step, investigates the influence of preexisting fractures (i.e., artificial structural planes) on the subsequent fracturing behavior. The simulation results are supported by mechanics analysis. Results indicated that the fracture evolution is influenced by pressure magnitude on a local scale around the fracture tip and by the orientation and distribution of pore pressure on a global scale. The constant pressure in wellbore H2 can affect fracture propagation by changing the water flow direction, and the hydraulic fractures will propagate to the direction of higher local pore pressure. Furthermore, the artificial structural planes influence the stress distribution surrounding the wellbores and the hydraulic fracture evolution by altering the induced stresses around the preexisting fractures. Finally, fracture network is formed among the artificial structural planes and hydraulic fractures when multiple wells are fractured successively. This study provides valuable guidance to unconventional reservoir reconstruction designs.

Numerical Investigation on the Hydraulic Fracture Evolution of Jointed Shale

“…Developed for heterogeneous materials (e.g., rock or concrete), the RFPA code is based on finite element and statistical damage theory, and it can simulate the failure process of heterogeneous and permeable geomaterials. In the simulation of heterogeneity and random distribution of material defects, the mesoscopic elements are assumed to be isotropic and homogeneous, and their mechanical properties (e.g., Young's modulus, Poisson's ratio, and strength properties, among others) are assumed to be linear in the RFPA model (Figure 1) [30]. ese linear mesoscopic elements are also statistically distributed (e.g., normal, Poisson's, and Weibull distributions).…”