Multiphase Flow

The complex interaction of liquids, gases and solids is of interest in many areas of geoscience including enhanced oil recovery, hydraulic fracturing and carbon sequestration. These interactions are heavily influenced by properties and phenomena which occur at the pore scale, such as the relative wettability of phases, interfacial physics and capillary actions. A significant amount of insight can be gained by accurately simulating multiphase fluid flow at the pore scale, however this has proven to be a challenging area for existing numerical methods to address.

We have developed a simulation framework capable of characterizing a variety of physical reservoir phenomena via numerical experiments on pore-scale core samples. This has included development in three key areas, namely, a plug-in solver framework for a range of numerical methods, a distribution technique to facilitate asynchronous, parallel computation on multicore hardware, and a library of methods including smooth particle hydrodynamics (SPH), the lattice Boltzmann method (LBM), the discrete element method (DEM) and finite difference (FD).

The numerical framework supports the simulation of multiple phases, including surface tension effects such as contact angle and surface wettability, and has been validated for porous media flows in both synthetic and actual geometries. The library of numerical methods provides flexibility when deciding the best approach for solving an engineering problem. From a mathematical and algorithmic standpoint these methods are all reasonably mature, making it easier to choose the most appropriate one(s) based on an assessment of its comparative strengths and weaknesses. For example, SPH excels at naturally reproducing the interfaces of multiple fluid phases whilst the LBM is computationally more efficient and can readily be coupled with DEM and or FEM to solve fluid structure interaction problems. Using both SPH and the LBM the absolute permeability of reservoir core samples can now be predicted robustly, whereas these data are traditionally determined experimentally from cored samples of rock. This procedure takes as input a segmented μCT image of the sample and produces a permeability-porosity relationship with data points from a number of analyzed subsamples. This has been extended to incorporate non-Newtonian fluids (e.g. shear-thickening/thinning), and the behavior of two-phase flows in synthetic porous media has also been assessed.

As the range of implemented and verified modeling capabilities increases, so too will the scope of applications. For example, it is intended that the permeability characterization process is tested for applicability in lower porosity rock samples such as those found in shale gas formations, and that the combination of non-Newtonian fluids and suspended particles is applied to simulate the behavior of proppants in hydraulic fractures.