Flow Simulation and Characterization of Fracture Systems Using Fast Marching Method and Novel Diagnostic Plots
MetadataShow full item record
Recently, the industrial trend of hydraulic fracturing is reducing the cluster spacing while increasing the fluid and proppant usage, which often generates complex fracture networks. The challenge from this trend is to understand and characterize the complex fracture networks. Recently, a novel approach has been proposed based on the high-frequency asymptotic solution of the diffusivity equation leading to the Eikonal equation. The Eikonal equation governs the pressure front propagation and can be solved by a front-tracking algorithm called Fast Marching Method (FMM). In this dissertation, we extend this method to complex fracture networks characterization and simulation, using novel diagnostic plots and FMM-based simulation. First, we develop novel diagnostic plots for complex fracture networks characterization. We directly use the field data to calculate the well drainage volume, instantaneous recovery ratio (IRR) and w(r) function. The w(r) function serves as a diagnostic plot to detect fracture geometry and flow regimes and the IRR plot is used to detect fracture conductivity. Second, we extend the FMM-based simulation workflow to local grid refinements (LGRs). The detailed workflow is proposed to generate the computational grid for the diffusive time of flight (DTOF) calculation. We use various models to validate the accuracy and computational efficiency of this workflow. In addition, we investigate various discretization schemes for the transition between local and global domain. Third, we extend the FMM-based simulation workflow to embedded discrete fracture model (EDFM). We utilize a novel gridding to link the embedded discrete fractures and the matrix based on Delaunay triangulation. Using the DTOF as a spatial coordinate, the FMM-based flow simulation reduces the 3D complex fracture networks simulation to an equivalent 1D simulation. Multiple examples are shown to validate the accuracy and computational efficiency of this workflow. Lastly, we investigate the impact of tighter cluster spacing of the hydraulic fractures using the Eagle Ford field data. The hydraulic fracture propagation simulator Mangrove® is used to generate the fracture patterns based on the completion data. A manual history matching is conducted to match the field injection treatment pressure. The impact of cluster spacing is examined through the calibrated fracture models.
Xue, Xu (2019). Flow Simulation and Characterization of Fracture Systems Using Fast Marching Method and Novel Diagnostic Plots. Doctoral dissertation, Texas A&M University. Available electronically from