arXiv subjects

Hannah P. Menke

Publications and source records attributed to Hannah P. Menke.

At least 19 recordsLinked to original sources

arXiv2026

Machine learning enhanced data assimilation framework for multiscale carbonate rock characterization

Carbonate reservoirs offer significant capacity for subsurface carbon storage, oil production, and underground hydrogen storage. X-ray computed tomography (X-ray CT) coupled with numerical simulations is commonly used to investigate the multiphase flow behaviors in carbonate rocks. Carbonates exhibit pore size distribution across scales, hindering the comprehensive investigation with conventional X-ray CT images. Imaging samples at both macro and micro-scales (multi-scale imaging) proved to be a viable option in this context. However, multi-scale imaging faces two key limitations: the trade-off between field of view and voxel size necessitates resource-intensive imaging, while multi-scale multi-physics numerical simulations on resulting digital models incur prohibitive computational costs. To address these challenges, we propose a machine learning-enhanced data assimilation framework that leverages experimental drainage relative permeability measurements to achieve efficient characterization of micro-scale structures, delivering a data-driven solution toward a high-fidelity multiscale digital rock modeling. We train a dense neural network (DNN) as a proxy to a multi-scale pore network simulator and couple it with an ensemble smoother with multiple data assimilation (ESMDA) algorithm. DNN-ESMDA framework simultaneously infers the CO2-brine drainage relative permeability of microporosity phases with associated uncertainty estimation, revealing the relative importance of each rock phase and guiding future characterization. Our DNN-ESMDA framework achieves a computational speedup, reducing inference time from thousands of hours to seconds compared with the usage of conventional multiscale numerical simulation. Given this computational efficiency and applicability, the machine learning-enhanced ESMDA framework presents a generalizable approach for characterizing multiscale carbonate rocks.

physics.geo-ph↗

arXiv2026

Surrogate models for Rock-Fluid Interaction: A Grid-Size-Invariant Approach

Modelling rock-fluid interaction requires solving a set of partial differential equations (PDEs) to predict the flow behaviour and the reactions of the fluid with the rock on the interfaces. Conventional high-fidelity numerical models require a high resolution to obtain reliable results, resulting in huge computational expense. This restricts the applicability of these models for multi-query problems, such as uncertainty quantification and optimisation, which require running numerous scenarios. As a cheaper alternative to high-fidelity models, this work develops eight surrogate models for predicting the fluid flow in porous media. Four of these are reduced-order models (ROM) based on one neural network for compression and another for prediction. The other four are single neural networks with the property of grid-size invariance; a term which we use to refer to image-to-image models that are capable of inferring on computational domains that are larger than those used during training. In addition to the novel grid-size-invariant framework for surrogate models, we compare the predictive performance of UNet and UNet++ architectures, and demonstrate that UNet++ outperforms UNet for surrogate models. Furthermore, we show that the grid-size-invariant approach is a reliable way to reduce memory consumption during training, resulting in good correlation between predicted and ground-truth values and outperforming the ROMs analysed. The application analysed is particularly challenging because fluid-induced rock dissolution results in a non-static solid field and, consequently, it cannot be used to help in adjustments of the future prediction.

cs.LG↗

arXiv2026

Beyond Critical Minerals Targets: Digital Rock Physics as Infrastructure for Secure and Circular Supply Chains

The United Kingdom and Europe are moving rapidly from critical minerals target-setting to implementation. The EU Critical Raw Materials Act and the UK's Vision 2035 create ambitious benchmarks for domestic extraction, processing, recycling, circularity, and supply-chain resilience, but many prospective regional resources remain complex, under-explored, historically worked, or economically marginal. This paper argues that implementation will depend not only on permitting reform and project designation, but also on pre-competitive measurement, modelling, and data infrastructure capable of determining which ores, brines, waste streams, and recycling feedstocks can be processed viably and with lower environmental impact. Digital Rock Physics (DRP) should therefore be understood as enabling infrastructure for resource policy rather than as a specialist laboratory method alone. By combining three-dimensional imaging, correlative chemistry, AI-enabled image analysis, and pore-scale modelling, DRP can connect mineral texture and reactive pathways to decisions about ore characterisation, liberation prediction, leaching, Direct Lithium Extraction, mine-waste valorisation, and battery recycling. The paper sets out a UK-European policy agenda built around translational demonstrators, cross-disciplinary training, a Digital Ore Passport standard, a federated Digital Ore Database, and integrated geo-reactive end stations. Treated as shared implementation infrastructure, DRP could help turn critical minerals strategies into practical routes for supply security, resource efficiency, circularity, and more environmentally responsible development.

cond-mat.other↗

arXiv2026

Machine learning for sustainable geoenergy: uncertainty, physics and decision-ready inference

Geoenergy projects (CO2 storage, geothermal, subsurface H2 generation/storage, critical minerals from subsurface fluids, or nuclear waste disposal) increasingly follow a petroleum-style funnel from screening and appraisal to operations, monitoring, and stewardship. Across this funnel, limited and heterogeneous observations must be turned into risk-bounded operational choices under strong physical and geological constraints - choices that control deployment rate, cost of capital, and the credibility of climate-mitigation claims. These choices are inherently multi-objective, balancing performance against containment, pressure footprint, induced seismicity, energy/water intensity, and long-term stewardship. We argue that progress is limited by four recurring bottlenecks: (i) scarce, biased labels and few field performance outcomes; (ii) uncertainty treated as an afterthought rather than the deliverable; (iii) weak scale-bridging from pore to basin (including coupled chemical-flow-geomechanics); and (iv) insufficient quality assurance (QA), auditability, and governance for regulator-facing deployment. We outline machine learning (ML) approaches that match these realities (hybrid physics-ML, probabilistic uncertainty quantification (UQ), structure-aware representations, and multi-fidelity/continual learning) and connect them to four anchor applications: imaging-to-process digital twins, multiphase flow and near-well conformance, monitoring and inverse problems (monitoring, measurement, and verification (MMV), including deformation and microseismicity), and basin-scale portfolio management. We close with a pragmatic agenda for benchmarks, validation, reporting standards, and policy support needed for reproducible and defensible ML in sustainable geoenergy.

cond-mat.dis-nn↗

arXiv2026

A Benchmark Dataset for Machine Learning Surrogates of Pore-Scale CO2-Water Interaction

Accurately capturing the complex interaction between CO2 and water in porous media at the pore scale is essential for various geoscience applications, including carbon capture and storage (CCS). We introduce a comprehensive dataset generated from high-fidelity numerical simulations to capture the intricate interaction between CO2 and water at the pore scale. The dataset consists of 624 2D samples, each of size 512x512 with a resolution of 35 μm, covering 100 time steps under a constant CO2 injection rate. It includes various levels of heterogeneity, represented by different grain sizes with random variation in spacing, offering a robust testbed for developing predictive models. This dataset provides high-resolution temporal and spatial information crucial for benchmarking machine learning models.

physics.chem-ph↗

arXiv2026

Feature-Modulated UFNO for Improved Prediction of Multiphase Flow in Porous Media

The UNet-enhanced Fourier Neural Operator (UFNO) extends the Fourier Neural Operator (FNO) by incorporating a parallel UNet pathway, enabling the retention of both high- and low-frequency components. While UFNO improves predictive accuracy over FNO, it inefficiently treats scalar inputs (e.g., temperature, injection rate) as spatially distributed fields by duplicating their values across the domain. This forces the model to process redundant constant signals within the frequency domain. Additionally, its standard loss function does not account for spatial variations in error sensitivity, limiting performance in regions of high physical importance. We introduce UFNO-FiLM, an enhanced architecture that incorporates two key innovations. First, we decouple scalar inputs from spatial features using a Feature-wise Linear Modulation (FiLM) layer, allowing the model to modulate spatial feature maps without introducing constant signals into the Fourier transform. Second, we employ a spatially weighted loss function that prioritizes learning in critical regions. Our experiments on subsurface multiphase flow demonstrate a 21\% reduction in gas saturation Mean Absolute Error (MAE) compared to UFNO, highlighting the effectiveness of our approach in improving predictive accuracy.

cs.LG↗

arXiv2025

A First Look at Hydrogen Generation in an Ultramafic Rock with Micro-CT and SEM-BEX

Natural hydrogen generated by water-rock interaction in ultramafic rocks is increasingly recognised as a potentially important primary energy resource, but the pore-scale processes that control the initiation and early transport of a free gas phase remain poorly constrained. Here we present an in situ X-ray micro-tomography experiment in which an ultramafic granular pack of dunnite from West Papua, Indonesia, saturated with KI-doped brine, is heated to 100C with a pore pressure of 4bar under 10bar confining pressure inside a micro-CT scanner. Time-resolved 4D imaging captures the transition from a fully liquid-saturated pore space to the appearance and growth of a distinct gas phase after an 8h induction period. Bubbles first nucleate near the top of the sample before becoming distributed throughout the imaged volume as a connected ganglia. The nucleating gas phase is most plausibly dominated by molecular hydrogen generated by low-temperature fluid-rock reaction, as indicated by independent hydrogen-presence detectors, although we cannot yet fully exclude minor contributions from other gases. SEM-BEX imaging reveals textural alteration and local changes in elemental signals between reacted and unreacted material. Taken together, these observations provide spatially and temporally resolved evidence for gas generation during low-temperature alteration of ultramafic grains and demonstrate that pore-scale imaging can directly link water-rock reaction kinetics, gas generation and multiphase flow behaviour in natural hydrogen systems.

cond-mat.other↗

arXiv2025

Scalable CFD Simulations in Multi-Billion Voxel Micro-CT Images of Porous Materials Using OpenFOAM on ARCHER2

This study investigates the use of High-Performance Computing (HPC) to simulate flow and transport in ultra-large micro-CT images of porous materials using Computational Fluid Dynamics (CFD). Two distinct rock samples, representative of two different rock formations - Bentheimer sandstone and Estaillades carbonate - are investigated. The Bentheimer sandstone image, with dimensions 1,950x1,950x10,800 voxels at 6 micron resolution, comprising 41 billion voxels, represents a largely homogeneous structure, while the Estaillades carbonate image, at 1,144x1,144x6,000 voxels and 3.9676 micron resolution, amounting to 8 billion voxels, features greater heterogeneity, including micro-porous regions. These images are used for direct CFD simulations with GeoChemFoam, our OpenFOAM-based numerical solver, leveraging the computational resources of the UK supercomputer ARCHER2. One of the key aspects of the study is the use of the Darcy-Brinkman-Stokes approach, for which the solid surface is represented using a volumetric indicator function, rather than a complex mesh. This enables the use of simple Cartesian meshes that can be generated in parallel in an efficient and scalable manner. The study explores both weak and strong scaling through subvolume decomposition, demonstrating that, due to the strong scalability and the computational power of ARCHER2, full-resolution CFD simulations can be carried out without the need for image size reduction. This work illustrates the potential of HPC to perform detailed, full-scale simulations on large, high-resolution micro-CT data. The approach relies on a meshing strategy that leverages simple, parallelisable Cartesian grids derived from volumetric indicator functions, eliminating the need for complex surface-conforming meshes and allowing scalable simulation of flow and transport in geological and engineering applications.

physics.flu-dyn↗

arXiv2025

Inference of microporosity phase properties in heterogeneous carbonate rock with data assimilation techniques

Accurate digital rock modeling of carbonate rocks is limited by the difficulty in acquiring morphological information on small-scale pore structures. Defined as microporosity phases in computed tomography (micro-CT) images, these small-scale pore structures may provide crucial connectivity between resolved pores (macroporosity). However, some carbonate rocks are heterogeneous, and high-resolution scans are resource-intensive, impeding comprehensive sampling of microporosity phases. In this context, we propose the usage of the ensemble smoother multiple data assimilation (ESMDA) algorithm to infer the multiphase flow properties of microporosity phases from experimental observations for digital rock modeling. The algorithm's effectiveness and compatibility are validated through a case study on a set of mm-scale Estaillades drainage image data. The case study applies ESMDA to two capillary pressure models to infer the multiphase flow properties of microporosity phases. The capillary pressure curve and saturation map were used as observations to predict wetting phase saturation at six capillary pressure steps during iterative data assimilation. The ESMDA algorithm demonstrates improved performance with increasingly comprehensive observation data inputs, achieving better prediction than recently published alternative techniques. Additionally, ESMDA can assess the consistency between various forward physical models and experimental observations, serving as a diagnostic tool for future characterization. Given the diverse application conditions, we propose that ESMDA can be a general method in the characterization workflow of carbonate rocks.

physics.geo-ph↗

arXiv2025

Enhanced Permeability Estimation in Microporous Rocks Using a Hybrid Macropore-Darcy Approach

This study presents a novel workflow for constructing hybrid macropore-Darcy models from micro-CT images of microporous rocks. In our approach, macropore networks are extracted using established methods, while the microporosity is characterised through segmented phase classification and incorporated into the model as Darcy cells. Effectively, Darcy cells capture the micro scale connectivity variations that are missing in the macroscopic networks. This dual entity model thus incorporates both the conventional macroscopic pore structure and the critical flow pathways present in the under-resolved microporous regions. The proposed workflow is rigorously validated by comparing the permeability estimates with direct numerical simulation (DNS) results and experimental measurements. Our findings demonstrate that this hybrid approach reliably reproduces fluid flow behaviour in complex porous media while significantly reducing computational demands, offering a promising tool for advanced groundwater modelling and water resource management.

physics.flu-dyn↗

arXiv2024

Local Padding in Patch-Based GANs for Seamless Infinite-Sized Texture Synthesis

Texture models based on Generative Adversarial Networks (GANs) use zero-padding to implicitly encode positional information of the image features. However, when extending the spatial input to generate images at large sizes, zero-padding can often lead to degradation in image quality due to the incorrect positional information at the center of the image. Moreover, zero-padding can limit the diversity within the generated large images. In this paper, we propose a novel approach for generating stochastic texture images at large arbitrary sizes using GANs based on patch-by-patch generation. Instead of zero-padding, the model uses \textit{local padding} in the generator that shares border features between the generated patches; providing positional context and ensuring consistency at the boundaries. The proposed models are trainable on a single texture image and have a constant GPU scalability with respect to the output image size, and hence can generate images of infinite sizes. We show in the experiments that our method has a significant advancement beyond existing GANs-based texture models in terms of the quality and diversity of the generated textures. Furthermore, the implementation of local padding in the state-of-the-art super-resolution models effectively eliminates tiling artifacts enabling large-scale super-resolution. Our code is available at \url{https://github.com/ai4netzero/Infinite_Texture_GANs}.

cs.CV↗

arXiv2024

Multi-scale flow, permeability, and heat transport in low-carbon and traditional building materials

Permeability and heat transport through building materials ultimately dictates their insulatory performance over a buildings service lifetime. Experiments combining XCT with numerical modelling are an accepted method of studying pore scale processes and have been used extensively in the oil and gas industry to study highly complex reservoir rocks. However, despite the obvious similarities in structure and application, these techniques have not yet been widely adopted by the building and construction industry. An experimental investigation was performed on the pore structure of several building materials, including conventional, historic, and innovative, using XCT and direct numerical simulation. Six samples were imaged at between a 4 and 15 micron resolution inside a micro-CT scanner. The porosity and connectivity were extracted with the grain, throat, and pore size distributions using image analysis. The permeability, velocity, and thermal conductivity were then investigated using GeoChemFoam, our highly-versatile and open source numerical solver. It was found that each material had a unique, heterogeneous and sometimes multi-scale structure that had a large impact on the permeability and thermal conductivity. Furthermore, it was found that the method of including sub-resolution porosity directly effected these bulk property calculations for both parameters, especially in the materials with high structural heterogeneity. This is the first multi-scale study of structure, flow and heat transport on building materials and this workflow could easily be adapted to understand and improve designs in other industries that use porous materials such as fuel cells and batteries technology, lightweight materials and insulation, and semiconductors.

cond-mat.mtrl-sci↗

arXiv2023

Experimental investigation of solubility trapping in 3D printed micromodels

Understanding interfacial mass transfer during dissolution of gas in a liquid is vital for optimising large-scale carbon capture and storage operations. While the dissolution of CO2 bubbles in reservoir brine is a crucial mechanism towards safe CO2 storage, it is a process that occurs at the pore-scale and is not yet fully understood. Direct numerical simulation (DNS) models describing this type of dissolution exist and have been validated with semi-analytical models on simple cases like a rising bubble in a liquid column. However, DNS models have not been experimentally validated for more complicated scenarios such as dissolution of trapped CO2 bubbles in pore geometries where there are few experimental datasets. In this work we present an experimental and numerical study of trapping and dissolution of CO2 bubbles in 3D printed micromodel geometries. We use 3D printing technology to generate three different geometries, a single cavity geometry, a triple cavity geometry and a multiple channel geometry. In order to investigate the repeatability of the trapping and dissolution experimental results, each geometry is printed three times and three identical experiments are performed for each geometry. The experiments are performed at low capillary number representative of flow during CO2 storage applications. DNS simulations are then performed and compared with the experimental results. Our results show experimental reproducibility and consistency in terms of CO2 trapping and the CO2 dissolution process. At such low capillary number, our numerical simulator cannot model the process accurately due to parasitic currents and the strong time step constraints associated with capillary waves. However, we show that, for the single and triple cavity geometry.

physics.flu-dyn↗

arXiv2023

Channeling: a new class of dissolution in complex porous media

The current conceptual model of mineral dissolution in porous media is comprised of three dissolution patterns (wormhole, compact, and uniform) - or regimes - that develop depending on the relative dominance of flow, diffusion, and reaction rate. Here, we examine the evolution of pore structure during acid injection using numerical simulations on two porous media structures of increasing complexity. We examine the boundaries between regimes and characterise the existence of a fourth regime called channeling, where already existing fast flow pathways are preferentially widened by dissolution. Channeling occurs in cases where the distribution in pore throat size results in orders of magnitude differences in flow rate for different flow pathways. This focusing of dissolution along only dominant flow paths induces an immediate, large change in permeability with a comparatively small change in porosity, resulting in a porosity-permeability relationship unlike any that has been previously seen. This work demonstrates that our current conceptual model of dissolution regimes must be modified to include channeling for accurate predictions of dissolution in applications such as geologic carbon storage and geothermal energy production.

physics.flu-dyn↗

arXiv2022

Improved Volume-of-Solid formulations for micro-continuum simulation of mineral dissolution at the pore-scale

We present two novel Volume-of-Solid (VoS) formulations for micro-continuum simulation of mineral dissolution at the pore-scale. The traditional VoS formulation (VoS-psi) uses a diffuse interface localization function psi to ensure stability and limit diffusion of the reactive surface. The main limitation of this formulation is that accuracy is strongly dependent on the choice of the localization function. Our first novel improved formulation (iVoS) uses the divergence of a reactive flux to localize the reaction at the fluid-solid interface, so no localization function is required. Our second novel formulation (VoS-psi') uses a localization function with a parameter that is fitted to ensure that the reactive surface area is conserved globally. Both novel methods are validated by comparison with experiments, numerical simulations using an interface tracking method based on the Arbitrary Eulerian Lagrangian (ALE) framework, and numerical simulations using the VoS-psi. All numerical methods are implemented in GeoChemFoam, our reactive transport toolbox and three benchmark test cases in both synthetic and real pore geometries are considered: (1) dissolution of a calcite post by acid injection in a microchannel and experimental comparison, (2) dissolution in a 2D polydisperse disc micromodel at different dissolution regimes and (3) dissolution in a Ketton carbonate rock sample and comparison to \textit{in-situ} micro-CT experiments. We find that the iVoS results match accurately experimental results and simulation results obtained with the ALE method, while the VoS-psi method leads to inaccuracies that are mostly corrected by the VoS-psi' formulation. In addition, the VoS methods are significantly faster than the ALE method, with a speed-up factor of between 2 and 12.

physics.flu-dyn↗

arXiv2021

GeoChemFoam: Direct Modelling of flow and heat transfer in micro-CT images of porous media

GeoChemFoam is an open-source OpenFOAM-based numerical modelling toolbox that includes a range of custom packages to solve complex flow processes including multiphase transport with interface transfer, single-phase flow in multiscale porous media, and reactive transport with mineral dissolution. In this paper, we present GeoChemFoam's novel numerical model for simulation of conjugate heat transfer in micro-CT images of porous media. GeoChemFoam uses the micro-continuum approach to describe the fluid-solid interface using the volume fraction of fluid and solid in each computational cell. The velocity field is solved using Brinkman's equation with permeability calculated using the Kozeny-Carman equation which results in a near-zero permeability in the solid phase. Conjugate heat transfer is then solved with heat convection where the velocity is non-zero, and the thermal conductivity is calculated as the harmonic average of phase conductivity weighted by the phase volume fraction. Our model is validated by comparison with the standard two-medium approach for a simple 2D geometry. We then simulate conjugate heat transfer and calculate heat transfer coefficients for different flow regimes and injected fluid analogous to injection into a geothermal reservoir in a micro-CT image of Bentheimer sandstone and perform a sensitivity analysis in a porous heat exchanger with a random sphere packing.

physics.flu-dyn↗

arXiv2021

GeoChemFoam: Operator Splitting based time-stepping for efficient Volume-Of-Fluid simulation of capillary-dominated two-phase flow

We present a novel time-stepping method, called Operator Splitting with Capillary Relaxation (OSCAR), for efficient Volume-Of-Fluid simulations of capillary-dominated two-phase flow. OSCAR uses operator splitting methods to separate the viscous drag and the surface tension forces. Different time-steps are used for the viscous drag steps, controlled by the injection velocity, and for the capillary relaxation steps, controlled by the velocity of capillary waves. Although OSCAR induces an additional numerical error of order 0 in time resulting from the splitting, it is well suited for simulations at low capillary number. First, the splitting error decreases with the capillary number and at low capillary number, the relaxation steps converge before reaching their last iteration, resulting in a large speed-up (here up to 250x) compared to standard time-stepping methods. The method is implemented in GeoChemFoam, our OpenFOAM-based CFD solver. Convergence, accuracy and efficiency are demonstrated on three benchmark cases: (1) the steady motion of an air bubble in a straight 2D microchannel, (2) injection of supercritical CO2 in a 3D constricted channel leading to a snap-off, and (3) water drainage in a 2D oil-wet micromodel representing a porous media.

physics.flu-dyn↗

arXiv2021

GeoChemFoam: Direct modelling of multiphase reactive transport in real pore geometries with equilibrium reactions

We present the novel numerical model GeoChemFoam, a multiphase reactive transport solver for simulations on complex pore geometries, including microfluidic devices and micro-CT images. The geochemical model includes bulk and surface equilibrium reactions. Multiphase flow is solved using the Volume-Of-Fluid method and the transport of species is solved using the Continuous Species Transfer method. The reactive transport equations are solved using a sequential Operator Splitting method, with the transport step solved using our OpenFOAM-based Computational Fluid Dynamics toolbox, and the reaction step solved using Phreeqc, the US geological survey's geochemical solver. The model is validated by comparison with analytical solutions in 1D and 2D geometries. We then applied the model to simulate multiphase reactive transport in two test pore geometries: a 3D pore cavity and a 3D micro-CT image of Bentheimer sandstone. In each case, we show the pore-scale simulation results can be used to develop upscaled models that are significantly more accurate than standard macro-scale equilibrium models.

physics.flu-dyn↗