The commercial development of unconventional oil and gas reservoirs using multiply fractured horizontal wells is vital to meeting the world's energy demand. However, the modeling of hydraulic fracture propagation and the production from such fractured wells still poses a challenge today. For instance, the need to re-mesh the computational grid as the fracture propagates (in standard numerical discretization schemes such as the finite element or finite volume schemes) limits the practical application of such methods.
A few authors have suggested the solution of the coupled partial differential equations that govern the flow, mechanical deformation, and hydraulic fracturing using a combination of an embedded discrete fracture model (EDFM) and the extended finite element model (XFEM). This is because EDFM allows flow modeling in the reservoir matrix with structured grids that do not have to conform to the orientation and geometry of the fracture grids. Similarly, XFEM allows the modeling of hydraulic fracture propagation without re-meshing the matrix. This work proposes the use of an unconditionally stable sequentially implicit scheme that couples the projection-based EDFM (pEDFM) with XFEM to accurately model low-conductivity fractures efficiently. Unlike previous coupled XFEM and EDFM attempts that only model flow in fractures using porous media with high fracture porosity, we provide the option of using either this or the lubrication theory for flow in parallel plates.
The validation studies presented indicate the accuracy of our model at reproducing the analytical solutions to coupled geomechanics and fracture propagation problems. We show that the iterative coupling of pEDFM with XFEM accounts for the presence of low-conductivity fractures in the vicinity of a hydraulic fracture, whereas EDFM does not. This is important when modeling hydraulic fracturing and the subsequent production from multiply fractured hydraulic wells. We also show that the use of lubrication theory yields more accurate results in unpropped fractures. The iterative coupling approach used in this work provides the flexibility and simplicity needed to model complex fluid and rock behaviors in unconventional reservoirs.