Pore-scale modeling of ostwald ripening in porous media
File(s)
Author(s)
Adebimpe, Ademola
Type
Thesis
Abstract
A novel pore network is developed to simulate quasi-static capillary-controlled two-phase displacement in porous media combined with transport of dissolved gas in the aqueous phase. The aim of this research is to quantify the impact of Ostwald ripening, the rearrangement of gas mediated by aqueous phase transport, on residual gas saturation, capillary pressure and relative permeability. The application of this work is the increasingly important subsurface gas (natural gas, carbon dioxide, hydrogen and others) storage.
First, a quasi-static pore network model of two-phase flow is presented. Written from scratch in Python with a modular, extensible design, the model employs NumPy vectorisation, Numba JIT compilation, and efficient data structures to retain computational performance comparable to C++, providing a platform with clear algorithms, modular structure, and maintainable code for implementing new physics and coupling with transport processes.
The novel contribution of the thesis is the incorporation of Ostwald ripening under capillary equilibrium conditions. It is shown that, for sufficiently slow flow, capillary pressure is uniform within connected phases and across trapped gas ganglia. Differences in capillary pressure induce solubility gradients, governed by Henry’s law, driving aqueous-phase transport that causes small, high-pressure ganglia to shrink while larger ganglia grow or merge. This process is described by an equilibrium percolation-without-trapping framework. Using the model, laboratory measurements are re-scaled to predict field-scale equilibrium properties, demonstrating that Ostwald ripening reduces residual gas saturation by approximately (20–25%). Normalized saturation enables corresponding re-scaling of capillary pressure and relative permeability, validated across sandstone and carbonate networks.
The thesis further develops a time-dependent Ostwald ripening model, coupling multiphase displacement with advection–diffusion of dissolved gas on an unstructured pore network. Quantitative comparison with pore-scale imaging experiments in Bentheimer sandstone shows good agreement in the evolution of trapped gas morphology. All code and input files are released publicly.
First, a quasi-static pore network model of two-phase flow is presented. Written from scratch in Python with a modular, extensible design, the model employs NumPy vectorisation, Numba JIT compilation, and efficient data structures to retain computational performance comparable to C++, providing a platform with clear algorithms, modular structure, and maintainable code for implementing new physics and coupling with transport processes.
The novel contribution of the thesis is the incorporation of Ostwald ripening under capillary equilibrium conditions. It is shown that, for sufficiently slow flow, capillary pressure is uniform within connected phases and across trapped gas ganglia. Differences in capillary pressure induce solubility gradients, governed by Henry’s law, driving aqueous-phase transport that causes small, high-pressure ganglia to shrink while larger ganglia grow or merge. This process is described by an equilibrium percolation-without-trapping framework. Using the model, laboratory measurements are re-scaled to predict field-scale equilibrium properties, demonstrating that Ostwald ripening reduces residual gas saturation by approximately (20–25%). Normalized saturation enables corresponding re-scaling of capillary pressure and relative permeability, validated across sandstone and carbonate networks.
The thesis further develops a time-dependent Ostwald ripening model, coupling multiphase displacement with advection–diffusion of dissolved gas on an unstructured pore network. Quantitative comparison with pore-scale imaging experiments in Bentheimer sandstone shows good agreement in the evolution of trapped gas morphology. All code and input files are released publicly.
Version
Open Access
Date Issued
2026-02-01
Date Awarded
2026-05-01
Copyright Statement
Attribution-NonCommercial 4.0 International Licence (CC BY-NC)
License URL
Advisor
Blunt, Martin
Bijeljic, Branko
Sponsor
Petroleum Technology Development Fund (Nigeria)
Publisher Department
Department of Earth Science & Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)