InterPore2026

Multiscale Pore-Network Model of Carbonate Reservoirs: Experimental Validation and Wettability Analysis

22 May 2026, 11:50

15m

Oral Presentation (MS09) Pore-Scale Physics and Modeling MS09

Speaker

Rafael Arenhart (LTrace)

Description

Pore-Network Models (PNM) provide a computationally efficient framework for simulating flow in porous media. However, many economically significant carbonate reservoirs exhibit multiscale porosity: a term identifying a pore space with sizes spanning multiple orders of magnitude. When using PNM approaches, two main challenges arise: imaging resolution and computational complexity. Resolution constraints stem from the micro-CT imaging trade-off between field of view and voxel resolution; achieving a Representative Elementary Volume (REV) for larger pores often results in a resolution too coarse to resolve the finer porosity. Regarding complexity, representing every individual pore across all scales within an REV can lead to a total pore count that renders flow simulations unfeasible. To address these issues, we implemented a multiscale PNM approach based on the micro-link model proposed by Bultreys (2016).
The adopted approach utilizes a hybrid combination of pores: resolved porosity is extracted into pores and throats and solved using the methods of Valvatne (2004), while unresolved porosity is addressed via an implicit model based on structural assumptions. We diverge from the original micro-link model by employing a bundle-of-tubes assumption for the unresolved porosity structure, and also by characterizing the microporosity as Darcy-types pores and throats, instead of the distinct structure of the micro-links. These modifications allows for the integration of experimental Mercury Injection Capillary Pressure (MICP) data and utilizes the OpenPNM (Gostick, 2016) library for fluid flow simulation.
The current work applies this multiscale multiflow relative permeability method to a suite of 20 carbonate rock samples. The primary objective is to verify the validity of two-phase flow simulations as a characterization tool for samples with limited experimental information. A significant challenge in this context is the accurate characterization of complex wettability behavior. To overcome this, a sensitivity analysis was performed by applying multiple wettability scenarios to the same network. The rock's specific characteristics are then inferred by identifying the parameters that yield a relative permeability curve most closely matching experimental results.
Experimental results highlight the necessity of an accurate wettability definition for high-fidelity simulations. Furthermore, the findings demonstrate distinct flow behaviors between pore spaces that percolate through resolved versus unresolved porosity. Finally, the study underscores the importance of High-Performance Computing (HPC) for the practical application of large-scale sensitivity testing in digital rock physics.