Image-Based Pore-Scale Modeling of Inertial Flow in Porous Media and Propped Fractures

Abstract

Non-Darcy flow is often observed near wellbores and in hydraulic fractures where relatively high velocities occur. Quantifying additional pressure drop caused by non-Darcy flow and fundamentally understanding the pore-scale inertial flow is important to oil and gas production in hydraulic fractures. Image-based pore-scale modeling is a powerful approach to obtain macroscopic transport properties of porous media, which are traditionally obtained from experiments and understand the relationship between fluid dynamics with complex pore geometries. In image-based modeling, flow simulations are conducted based on pore structures of real porous media from X-ray computed tomographic images. Rigorous pore-scale finite element modeling using unstructured mesh is developed and implemented in proppant fractures. The macroscopic parameters permeability and non-Darcy coefficient are obtained from simulations. The inertial effects on microscopic velocity fields are also discussed. The pore-scale network modeling of non-Darcy flow is also developed based on simulation results from rigorous model (FEM). Network modeling is an appealing approach to study porous media. Because of the approximation introduced in both pore structures and fluid dynamics, network modeling requires much smaller computational cost than rigorous model and can increase the computational domain size by orders of magnitude. The network is validated by comparing pore-scale flowrate distribution calculated from network and FEM. Throat flowrates and hydraulic conductance values in pore structures with a range of geometries are compared to assess whether network modeling can capture the shifts in flow pattern due to inertial effects. This provides insights about predicting hydraulic conductance using the tortuosity of flow paths,which is a significant factor for inertial flow as well as other network pore and throat geometric parameters.