Multi-rate mass transfer modeling of two-phase flow in highly heterogeneous fractured and porous media

TL;DR

This paper develops a multirate rate mass transfer (MRMT) model to accurately simulate two-phase flow in highly heterogeneous fractured and porous media, emphasizing the role of mass transfer between mobile fractures and immobile matrix blocks.

Contribution

It introduces two methods to estimate the global memory function for MRMT, enabling better modeling of multiphase flow in complex heterogeneous media.

Findings

Both estimation methods yield predictive MRMT models.

The methods are validated on fracture networks of varying complexity.

The approach effectively captures flow behavior in heterogeneous media.

Abstract

We study modeling of two-phase flow in highly heterogeneous fractured and porous media. The flow behaviour is strongly influenced by mass transfer between a highly permeable (mobile) fracture domain and less permeable (immobile) matrix blocks. We quantify the effective two-phase flow behaviour using a multirate rate mass transfer (MRMT) approach. We discuss the range of applicability of the MRMT approach in terms of the pertinent viscous and capillary diffusion time scales. We scrutinize the linearization of capillary diffusion in the immobile regions, which allows for the formulation of MRMT in the form of a non-local single equation model. The global memory function, which encodes mass transfer between the mobile and the immobile regions, is at the center of this method. We propose two methods to estimate the global memory function for a fracture network with given fracture and matrix geometry. Both employ a scaling approach based on the known local memory function for a given immobile region. With the first method, the local memory function is calculated numerically, while the second one employs a parametric memory function in form of truncated power-law. The developed concepts are applied and tested for fracture networks of different complexity. We find that both physically based parameter estimation methods for the global memory function provide predictive MRMT approaches for the description of multiphase flow in highly heterogeneous porous media.