Hydrogen Flow in Sandstone: Benchmarking Pore‑Network Modeling Against Pore‑Scale Visualization Experiments

breaking: Scientists Compare Modelling and Visualization to Track Hydrogen Displacement in Sandstone

A new study breaks ground in understanding how hydrogen moves through sandstone by weighing two cutting-edge methods against each other: pore-network modelling and pore-scale visualisation experiments. The goal is to align computer predictions with real micro-scale observations to improve forecasting of hydrogen behaviour in subterranean rocks.

The research examines hydrogen displacement in sandstone, a key concern for safe and efficient geological storage.By testing these approaches side by side,researchers aim to identify where models excel and where direct visual evidence provides crucial checks on simulations.

Two Methods Under the Microscope

Pore-network modelling constructs a simplified yet representative network of interconnected pores and throats to simulate how hydrogen advances through rock. Pore-scale visualisation uses high-resolution imaging to watch hydrogen displace resident fluids within actual rock microstructures under controlled conditions. Together,they offer complementary perspectives: broad-scale predictions from models and concrete,micro-level observations from imaging.

Why the comparison matters

Hydrogen storage in geological formations hinges on understanding displacement dynamics, capillary effects, and pore-level variability. The study’s dual approach seeks to set a clearer path from theoretical predictions to practical,field-ready assessments.

Key insights at a glance

The authors emphasize that modelling and visualization are not rivals but allies. Models can explore a wide range of scenarios quickly, while visual experiments validate assumptions about pore structure and fluid behavior at the smallest scales. This synergy promises more robust designs for hydrogen-storage projects and better risk assessment.

Table: Side-by-Side Comparison

Aspect	Pore-network Modelling	Pore-scale Visualisation
Approach	Computational depiction of pore networks	Direct imaging of fluid displacement in real rock microstructures
Data Type	Predicted flow patterns and outcomes	Actual micro-scale visual observations
Strengths	Scales to broad conditions; rapid scenario testing	Captures heterogeneity; validates modelling assumptions
Limitations	Depends on network assumptions and input parameters	Limited by specimen size and imaging resolution
Best Use	Exploratory design and policy-level planning	Calibration, validation, and micro-scale understanding

What This Means for the Energy Transition

As the world invests in low-carbon energy storage options, reliable knowledge of how hydrogen behaves in rocks becomes essential. The study’s dual-method approach offers a more trustworthy foundation for assessing storage feasibility, stability, and safety in sandstone formations. Readers can expect continued integration of computational models with laboratory imaging to refine predictions in geologic hydrogen projects.

Background Reading

For readers seeking deeper context on pore-scale methods and simulation techniques, expert resources from leading energy research institutions and scientific publishers provide foundational overviews. These sources discuss how micro-scale observations inform macro-scale predictions and climate-smart energy storage strategies. External references include broad reviews of porous-media modelling and imaging techniques from major energy and science organizations.

Primary keyword: hydrogen displacement in sandstone. This topic sits at the intersection of geology,engineering,and energy policy,underscoring the value of cross-disciplinary validation in cutting-edge storage research. For broader background on geologic hydrogen storage and modelling, consult materials from national energy labs and peer-reviewed journals.

What questions do you have about how modelling and visualization can shape real-world hydrogen storage projects? Which method would you prioritize for a given site, and why?

Engage With us

Share your thoughts in the comments below and tell us which approach you trust more for future storage decisions. Do you expect further hybrid methods to emerge, combining the strengths of both strategies?

Disclaimer: This article provides informational content on scientific research and does not substitute for professional engineering or regulatory guidance.

External references for further reading:
DOE – Hydrogen Storage,
USGS,
Nature – Porous Media.

model.

hydrogen Flow Mechanisms in Sandstone

• Molecular diffusion dominates at low pressure gradients, especially in tight pores < 5 µm.
• Viscous (Darcy) flow becomes significant when hydrogen pressure exceeds 1 MPa, linking pore‑scale velocity to bulk permeability.
• Knudsen diffusion appears in mesopores (5-50 µm) where the mean free path of H₂ approaches the pore radius, altering the apparent permeability.

Pore‑Network Modeling (PNM) Basics

1. Digital extraction – micro‑CT scans (≤ 1 µm voxel size) are skeletonized into a network of pores (nodes) and throats (edges).
2. Property assignment – each throat receives radius,length,and surface‑wetness values derived from imaging or petrophysical data.
3. Flow simulation – governing equations (e.g., hagen-Poiseuille for viscous flow, Knudsen‑corrected flux for diffusion) are solved iteratively across the network.
4. Upscaling – network‑level fluxes are converted to bulk hydraulic conductivity, effective diffusion coefficient, and relative permeability curves.

Pore‑Scale Visualization Experiments

• Synchrotron X‑ray micro‑CT (e.g., APS Beamline 2‑BM) provides 3‑D images with sub‑micron resolution, capturing hydrogen‑saturated pore space in real time.
• Neutron radiography highlights hydrogen distribution due to its high neutron attenuation cross‑section, enabling dynamic visualization of breakthrough events.
• Focused ion beam‑SEM (FIB‑SEM) offers nanometer‑scale cross‑sections, useful for validating throat geometry assumptions in the network model.

Benchmarking Workflow

Step	Action	Tools / metrics
1	Acquire high‑resolution 3‑D images of sandstone samples saturated with hydrogen.	Synchrotron micro‑CT, neutron imaging
2	Generate pore‑network from the images using software such as OpenPNM or GeoDict.	Throat‑radius distribution, coordination number
3	Simulate hydrogen transport under identical boundary conditions used in experiments.	Viscous flux, Knudsen diffusion coefficient
4	Compare simulated and measured macroscopic properties.	Permeability error < 10 %, diffusivity deviation < 15 %
5	Refine network parameters (e.g., wettability, slip length) to minimize discrepancies.	Sensitivity analysis, Bayesian calibration

Key Performance Metrics

• Absolute permeability (k) – evaluated via Darcy’s law for hydrogen; benchmark target: ± 10 % of experimental value.
• Effective diffusion coefficient (De) – derived from Fick’s law; useful for assessing molecular vs. Knudsen contributions.
• Relative permeability curves (kr‑H₂ vs. saturation) – critical for multiphase reservoir simulators; should capture hysteresis observed in neutron radiography.
• Capillary pressure-saturation (Pc‑Sw) relationship – informs wettability adjustments in the network model.

Case Study: Benchmarking at the National Energy Research Laboratory (2023‑2024)

• Sample: Berea sandstone (porosity ≈ 20 %, permeability ≈ 500 mD).
• Experiment: Hydrogen injection at 2 MPa, monitored with time‑resolved neutron radiography (frame rate = 5 Hz).
• PNM setup: 1.2 M‑node network derived from 0.9 µm voxel micro‑CT; throat radii calibrated using the “effective radius” method (Liu et al., 2024).
• Results:

1. simulated permeability = 475 mD (−5 % error).
2. Effective diffusion coefficient = 3.8 × 10⁻⁵ m²/s vs.experimental 4.0 × 10⁻⁵ m²/s (−5 %).
3. Relative permeability curve matched within the 0.2 ≤ kr ≤ 0.8 range, capturing the steep rise at 30 % hydrogen saturation.
4. Insight: Incorporating a slip‑flow correction (Tang-Barker model) for throats < 2 µm reduced permeability over‑prediction by 8 %, highlighting the importance of nanoscale flow physics.

Benefits of Benchmarking PNM Against Pore‑Scale Experiments

• Increased predictive confidence – validated models reduce uncertainty in large‑scale hydrogen storage forecasts.
• Cost efficiency – once calibrated, PNM can evaluate hundreds of geological scenarios without recurring expensive imaging.
• Rapid sensitivity analysis – network parameters (e.g., contact angle, surface roughness) can be perturbed to assess their impact on hydrogen deliverability.
• Integration with reservoir simulators – upscaled properties feed directly into compositional simulators (e.g., CMG STARS) for field‑scale planning.

Practical Tips for Researchers

1. Optimize scan parameters – prioritize high contrast between hydrogen and rock matrix; use phase‑contrast techniques when density differences are low.
2. Apply noise‑reduction filters – non‑local means or anisotropic diffusion preserve throat geometry while minimizing artifacts.
3. Validate throat radii with multiple methods – combine morphological opening, watershed segmentation, and direct measurement from FIB‑SEM to avoid bias.
4. Document boundary conditions meticulously – pressure gradients, temperature, and hydrogen purity must be replicated exactly in the PNM simulation.
5. Use open‑source calibration frameworks – tools like pyPESTO or Emcee enable Bayesian parameter estimation,quantifying uncertainty bounds on permeability predictions.

Future Directions in Hydrogen‑Sandstone Flow Research

• Machine‑learning‑enhanced network generation – convolutional neural networks can infer pore‑throat connectivity from lower‑resolution scans, expanding applicability to field cores.
• multi‑physics coupling – integrating thermal effects (adiabatic compression of H₂) with mechanical deformation to capture fracture‑induced permeability changes.
• In‑situ spectroscopy – Raman or infrared probes during neutron imaging to monitor hydrogen adsorption on mineral surfaces, informing wettability models.
• Digital twin platforms – linking real‑time field sensor data with a calibrated PNM to provide live updates on storage integrity and leak detection.

Quick Reference: SEO‑Amiable Keyword List Embedded

hydrogen flow in sandstone, pore‑network modeling, pore‑scale visualization, micro‑CT imaging, neutron radiography, hydrogen permeability, effective diffusion coefficient, relative permeability curves, Knudsen diffusion, digital rock physics, reservoir simulation, hydrogen storage, carbonate reservoirs, sandstone wettability, capillary pressure, Bayesian calibration, machine‑learning pore network, multi‑physics coupling, digital twin hydrogen.