Geophysics brings clarity to the potential of natural hydrogen

Natural hydrogen is considered a potential low-carbon energy source. But how much is generated underground—and how quickly? An international research team led by the LIAG Institute for Applied Geophysics in Hannover has published the first process-based estimates for the Western Pyrenees and Northern California in Nature Communications. The calculated production rates are lower than earlier theoretical estimates. The study's new evaluation method, through the open-source tool PoNHy, allows more realistic assessments and can be applied to other regions to identify promising geological systems more precisely and evaluate their potential.

Subscribe to our newsletter for the latest sci-tech news updates.

A promising low-carbon energy source

Deep underground, natural hydrogen can form when water reacts with iron-rich rocks from Earth's mantle. During this natural process, known as serpentinization, molecular hydrogen (H₂) is continuously released—unlike green hydrogen, which requires electricity to produce, or gray hydrogen, which relies on fossil fuels. Seeps of natural hydrogen have already been measured at the surface at numerous locations worldwide. In some places, such as Mali and the Philippines, communities have even used it as a local energy source.

Study areas: Western Pyrenees and Northern California

The challenge for science has been to understand how much is truly available and over what timescales it can be tapped. To address these questions, researchers in geophysics, geology, geochemistry and physics from the LIAG Institute for Applied Geophysics, the Federal Institute for Geosciences and Natural Resources, the Czech Academy of Sciences, the Universities of Montpellier and Toulouse, the United States Geological Survey and MARUM – Center for Marine Environmental Sciences at the University of Bremen worked together.

The team investigated two geologically well-documented regions: the Western Pyrenees and Northern California. In both areas, two different ultramafic rocks from the upper mantle react with water at great depth. In the Western Pyrenees, these are fertile lherzolites that were brought to comparatively shallow depths during earlier rifting processes. In Northern California, the team examined depleted harzburgites of the Coast Range Ophiolite. Hydrogen is detectable at the surface at both sites.

Geophysics makes production-rate estimates more robust

Because serpentinization takes place deep underground, it cannot be observed directly. This is where geophysics comes in. Using gravity, magnetic and seismic data, the team reconstructed the distribution, volume, temperature and degree of alteration of the mantle rocks in three dimensions. The geophysical models were then coupled with thermodynamic and kinetic calculations of fluid-rock reactions.

Rather than estimating theoretical maximum potential, the study models the actual rates at which hydrogen is generated under real geological conditions, while accounting for physical and chemical constraints that previous studies largely overlooked. The result is a more realistic and actionable picture of natural hydrogen as a resource.

Several hundred tonnes per year—less than expected

For the Western Pyrenees, the study estimates production of around 308 tonnes of hydrogen per year, compared with around 515 tonnes per year for Northern California. These values are substantially lower than earlier theoretical estimates, which projected hundreds of thousands to millions of tonnes per year. At the same time, they are consistent with measurements of natural hydrogen fluxes from comparable geological systems worldwide. This increases confidence that the model captures the true order of magnitude of the process.

"Our results provide a realistic baseline that tells us where to look, which conditions favor accumulation and what time scales need to be considered. That is exactly what responsible exploration of any new energy resource requires," says Dr. Rodolfo Christiansen, lead author of the study, which was carried out as part of his scientific work at the LIAG Institute for Applied Geophysics.

The lower production rates arise mainly because earlier approaches focused predominantly on the available rock volumes. The new modeling also accounts for processes that limit natural hydrogen release. These include the saturation of dissolved hydrogen in pore fluids, which slows formation once the fluid reaches its solubility limit, and the rate at which rock surfaces come into contact with water. Once these factors are incorporated into the model, production rates fall significantly.

More realistic estimates create new opportunities for exploration

At the same time, the study regards the more realistic determination of production rates as an opportunity: It enables exploration programs to be targeted more effectively.

"Large accumulations of hydrogen remain possible, but they require time and the right geology. With this knowledge and the open-source tool PoNHy, scientists and the exploration industry can investigate more systematically where favorable generation and storage conditions coincide," Christiansen emphasizes.

Newly developed PoNHy tool enables global exploration

As part of the study, the team developed the open-source tool PoNHy, short for Potential for Natural Hydrogen. The modeling tool integrates three-dimensional geophysical inversion with the thermodynamics and kinetics of fluid-rock reactions.

PoNHy is freely available on GitHub and Zenodo and can be applied to any serpentinizing geological system worldwide. It provides the scientific community and the exploration industry with a shared basis for further development, enabling geological settings to be compared, regional potential to be assessed more realistically and particularly promising study areas to be selected more systematically.

More information: Rodolfo Christiansen et al, Controls on natural hydrogen generation during serpentinization of mantle rocks, Nature Communications (2026). DOI: 10.1038/s41467-026-73920-5

Rodolfo Christiansen, PoNHy – Potential for Natural Hydrogen: Reproducible code and results for natural H₂ generation modelling (v3.0), Zenodo (2026). DOI: 10.5281/zenodo.18733249

Provided by LIAG-Institut für Angewandte Geophysik (LIAG)