Complex Laboratory Investigation of Fluid Flow Properties for Anisotropic Porous Media

Abstract Relative permeability and capillary pressure are the governing parameters that characterize multiphase fluid flow in porous media for diverse natural and industrial applications, including surface water infiltration into the ground, CO2 sequestration, and hydrocarbon enhanced recovery. Although the drastic effects of deformation of porous media on single-phase fluid flow have been well established, the stress dependency of flow in multiphase systems is not yet fully explored. Here, stress-dependent relative permeability and capillary pressure are studied in a water-wet carbonate specimen both analytically using fractal and poroelasticity theory and experimentally on the micro-scale and macro-scales by means of X-ray computed micro-tomography and isothermal isotropic triaxial core flooding cell, respectively. Our core flooding program using water/N2 phases shows a systematic decrease in the irreducible water saturation and gas relative permeability in response to an increase in effective stress. Intuitively, a leftward shift of the intersection point of water/gas relative permeability curves is interpreted as an increased affinity of the rock to the gas phase. Using a micro-scale proxy model, we identify a leftward shift in pore size distribution and closure of micro-channels to be responsible for the abovementioned observations. These findings prove the crucial impact of effective stress-induced pore deformation on multiphase flow properties of rock, which are missing from the current characterizations of multiphase flow mechanisms in porous media.

Download Full-text

Simulating fluid flow in complex porous materials by integrating the governing equations with deep-layered machines

npj Computational Materials ◽

10.1038/s41524-021-00598-2 ◽

2021 ◽

Vol 7 (1) ◽

Author(s):

Serveh Kamrava ◽

Muhammad Sahimi ◽

Pejman Tahmasebi

Keyword(s):

Porous Media ◽

Fluid Flow ◽

Stokes Equations ◽

Computation Time ◽

Mass Conservation ◽

Biological Tissues ◽

Heterogeneous Porous Media ◽

Navier Stokes ◽

Time Interval ◽

Flow Properties

AbstractFluid flow in heterogeneous porous media arises in many systems, from biological tissues to composite materials, soil, wood, and paper. With advances in instrumentations, high-resolution images of porous media can be obtained and used directly in the simulation of fluid flow. The computations are, however, highly intensive. Although machine learning (ML) algorithms have been used for predicting flow properties of porous media, they lack a rigorous, physics-based foundation and rely on correlations. We introduce an ML approach that incorporates mass conservation and the Navier–Stokes equations in its learning process. By training the algorithm to relatively limited data obtained from the solutions of the equations over a time interval, we show that the approach provides highly accurate predictions for the flow properties of porous media at all other times and spatial locations, while reducing the computation time. We also show that when the network is used for a different porous medium, it again provides very accurate predictions.

Download Full-text