The hunt for natural hydrogen reserves

They lit the gas and watched it burn clear – a good sign. It was 2018 and prospecting company Natural Hydrogen Energy (NH2E) had been drilling in the US state of Nebraska in the hope of detecting an underground accumulation of naturally occurring hydrogen. Several years of analysing survey data both old and new to assess the regional and local geology and surface geochemistry all pointed to this site, and there was potential access to the North American energy market too. The initial test results were positive.

‘Mixed feelings invade you – you want to celebrate but you are afraid,’ says Viacheslav Zgonnik, founder and chief executive of NH2E, as he recalls his fears they might be reading a false positive. Further analysis confirmed their results, however, and NH2E became the first company to successfully find natural hydrogen accumulation at a new site where no previous accidental discoveries had been reported.

Reserves of hydrogen had been stumbled upon previously as far back as 1888, when Dmitry Mendeleev reported his analysis of the gas seeping from a Ukrainian coal mine. Similar accidental hydrogen discoveries have cropped up all over the world since, but these drilling projects were all targeting hydrocarbon fossil fuels, so observations of hydrogen roused little more than marginal curiosity.

For a long time people claimed that natural hydrogen did not exist on Earth. They were wrong

Now, with the race to decarbonise energy sources, combined with an increasing body of evidence that quantities of hydrogen in underground reserves may be significant, this ‘natural hydrogen’ is finally starting to attract some limelight. At an estimated $1 per kg it is comparatively cheap as well as clean. While questions remain – such as where to drill for it, what flux of hydrogen to expect and for how long – with both research and industry involved, the prospects of a natural hydrogen fuel economy seem to be growing ever brighter. The interest attracted is also helping to shed light on what might be going on in the broader hydrogen cycle, potentially resolving some longstanding conundrums around Earth’s geology. An era of hydrogen fuelled science in more ways than one is upon us.

Hydrogen has long been sought after by the chemical industry, most notably to produce ammonia for fertilisers through the Haber–Bosch process. But concerns over the carbon dioxide released from burning fossil fuels have promoted it as a vector of energy, along with a rainbow of grades indicating the environmental impact of its production. The cheapest at an estimated $0.90–3.20 per kg (£0.70–£2.50), as well as the most established production method, is steam reforming. This works by reacting methane with steam at 700–1000°C and pressures of 3–25 bar to produce carbon monoxide and some carbon dioxide. It has been dubbed grey hydrogen on account of the environmental impact of this process, which significantly impinges on its green credentials. Carbon capture add-ons can mitigate this to a degree, upgrading it to blue hydrogen at a cost. Better still is green hydrogen, produced from electrolysis of water powered by renewables but while efforts to reduce costs continue it is often considered priced out, at $3–7.50 per kg. In this context, the idea of tapping off hydrogen from underground reservoirs ready to go seems almost too good to be true.

From fable to fact

‘For a long time people claimed that natural hydrogen did not exist on Earth. It was just wrong,’ says Alain Prinzhofer, scientific director of the consultancy firm GEO4U in Brazil and one of the early advocates for investigating the resource potential of natural hydrogen around 15 years ago. Yet there was good reasoning behind this assumption. Hydrogen is the smallest and lightest molecule on Earth, comprising just two of the lightest possible atoms. It is highly diffusive and can seep through many types of rock and fluids into the atmosphere. In addition, hydrogen is fairly reactive, so it seemed ridiculous to think it would be lurking in vast quantities underground, except of course for all the historic evidence of it. Yet for a long time even discovered hydrogen reserves were dismissed as ‘extremely marginal – of absolutely no economic interest’, says Prinzhofer.

As it happens these early discoveries were made against the odds too, since for a long time the chromatography in the gas analysers most common for samples from oil and gas drilling used hydrogen as a carrier gas. Any hydrogen in the sample was thus indistinguishable from the baseline. ‘It contributed significantly to missing accidental discoveries of hydrogen,’ says Zgonnik, who first developed an interest in the subject during his PhD in chemistry in the late 2000s.

Now there is a positive boom in the field, prompted in no small part by developments at a site of accidental hydrogen discovery in the Bourabougou field in Mali. ‘When you’re looking for natural gas you often ask villagers have you got any wells with something funny about them,’ says Owain Jackson, as he recounts how  a colleague at a commodity company in 2008 was exploring for oil and natural gas when their search led them to the Malian village of Bourakébougou. Local residents pointed out a well abandoned in the 1980s, where legend has it one of the drillers smoking a cigarette was badly injured by an exploding flux of gas from the well. Undaunted by the previous explosion, the villagers opened the tap on the well and lit a jet of high-pressure gas. Although there was some colour in the flame when gas from the well was lit, indicating at least some natural gas in the mix, analysis revealed the gas to be mainly hydrogen. 

Jackson and his colleague had been looking for natural gas and oil at the time and so moved on, but in 2012 Malian millionaire and entrepreneur Aliou Boubacar Diallo reopened the well and plugged it into a generator that provided the village with electricity for artificial lighting and freezers for the first time. The discovery also piqued Jackson’s interest and he got together with a group of around 12 scientists to work out a geological model for the hydrogen reservoir in Mali, identifying a possible source of the hydrogen, capping rocks to trap it and a fault system along which the gas could migrate. ‘We realised this wasn’t going to just happen in Mali, these geological ingredients happen all over the world,’ said Jackson, now chief executive of hydrogen exploration company H2Au.

Identifying the source

There are three main processes for generating underground hydrogen. The jury is still out as to the proportional contributions of each one but the oxidation of ferrous minerals (rocks containing iron in the +2 oxidation state) is a common favourite for generating the most. Wherever water is present alongside peridotite – an igneous rock made up largely of the silicate minerals olivine ((Mg,Fe)₂ SiO₄) and pyroxene XY(Si,Al)₂ O₆ (where X can be iron(ii) and Y can be magnesium, among other smaller metal ions) – a chemical reaction converts the silicates into serpentinite (X₃ Si₂ O₅ (OH)₄ (where X is for example Mg or Fe), brucite (Mg(OH)₂), and magnetite (Fe₃ O₄), releasing hydrogen. The low activity of silica plays a key role in enabling water to oxidise the iron in these compounds to form magnetite, since the extraction of silica from the olivine stabilises magnetite and produces the highly reducing conditions needed. This ‘serpentinisation’ is particularly favourable at 200–350°C.

‘What is new for the last couple of years is that in fact the sedimentary rock – iron-rich rock ­– onshore are considered to be a major generator of onshore hydrogen,’ says Isabelle Moretti, a hydrogen researcher at Université de Pau et des Pays de l’Adour in France. Her interest in natural hydrogen was piqued during several years in industry with the company ENGIE, where she was chief scientific officer. She points out that only 10 years ago people would have considered hydrogen generation as limited to serpentinisation. Although it has long been known that this process is active in mid ocean ridge ‘smokers’, extracting the hydrogen here would be economically unviable on account of the depth of water above the rocks releasing hydrogen and the distance to land where it might find use. Although serpentinisation can occur through the reactions of ophiolitic nappes, similar geological structures to the mid ocean ridges but onshore, we now know that hydrogen can also be produced by reactions with onshore banded ironstone formations as well.

When you sit down and do the science you see that hydrogen behaves just like natural gas

Another source of hydrogen is radiolysis of water. Here energy imparted by the decay of naturally occurring radioactive minerals can split water molecules into hydrogen and oxygen – nature’s free version of the more costly artificial ‘green hydrogen’ processes that mimic it. The process also produces helium when the radioactive alpha particles emitted gain electrons. This natural process is slow, however, so that it would take millions of years to generate a significant reserve. That the hydrogen-powered turbine in Mali could run for years without any discernible drop in pressure indicates there is a process replenishing the hydrogen stock at much greater rates, such that Prinzhofer among others describes it as a renewable resource. However, as Moretti points out, the coexistence of hydrogen and helium in some reserves shows that some hydrogen may be produced this way. Similarly, the detection and monitoring of hydrogen in coal mines suggests it may also be generated from buried organic matter in sedimentary basins at around 200°C.

How much for how long?

The discovery of hydrogen reserves – accidental or otherwise – rebuts arguments that the molecule is too diffusive and reactive to accumulate underground. ‘When you sit down and do the science you see that once it leaves the source it’s just like natural gas,’ says Jackson. ‘It can trap under very typical trapping cap rocks, it can reservoir in very typical reservoir rocks, and you can explore for it much the same way you explore for methane.’ The difference, he adds, is where it is found.

Once hydrogen emerges into the atmosphere it soon reacts with oxygen to make water, which is then available to produce more hydrogen again either from hydrolysis or from serpentinisation, and so the cycle continues. Radiolysis requires nothing other than the radiation from naturally occurring radioactive minerals, which have lifetimes of several billion years, hence the assumption that hydrogen generation through this process is very slow. As for the olivine oxidised to serpentinite in the serpentinisation process, this may be replenished eventually.

‘It goes through a big convective cell, through subduction zones, and about 200 million years later the same rocks are turned around,’ says Prinzhofer, describing the recovery of the source rocks for serpentinisation, olivine and pyroxene. He suggests that sulfur likely plays a role in cycling the rocks in the mantle, turning the iron(iii) serpentine minerals back into the iron(ii) olivine form but further research is underway. In any case the stores of these rocks already available are considered so vast there can be no concern over natural serpentinisation processes depleting them.

‘Estimates for the flux of geologic hydrogen have been increasing by one order of magnitude every decade or so, which is very surprising,’ says Zgonnik, whose own estimate of natural hydrogen quantities from an arithmetic sum of the values in the literature also notched the figure up yet another order of magnitude. ‘I believe that the next will be higher because we still know very little about natural hydrogen.’

For some the lack of information need not be an impediment to extraction, since there was little or no information on natural gas and oil when those industries started up, and they lacked what knowledge we may be able to transfer from the past 50 years of research into fossil fuels. Nonetheless, it would be helpful to understand them better. What is more this hydrogen may play a role in explaining other key geological processes.

Broader geological ramifications

Hydrogen gas is often reported at volcanoes, despite the exposure to oxygen in the air and rocks nearby, which would react with some of it. ‘Some researchers have even suggested that hydrogen can be the driving force for volcanic activity,’ says Zgonnik, explaining how the oxidation of a flux of deep-seated hydrogen could provide the heat needed to melt rocks and vaporise water causing the build-up of pressure that leads to a volcanic explosion. ‘It’s not the mainly accepted theory of volcanism but myself I find it incredibly appealing, and as a chemist it makes perfect sense to me.’

As to how hydrogen might come to be in the mantle in the first place, this may have been addressed in a recent study of a possible planetary formation pathway for Earth by Edward Young and Hilke Schlichting at the University of California in Los Angeles, alongside Anat Shahar at the Carnegie Institution of Washington in the US. Noting the prevalence of hydrogen-rich atmospheres in exoplanets large enough for their gravitational force to hold onto an atmosphere of such light molecules, the researchers hypothesised that all planets start out with a hydrogen-rich atmosphere, but the smaller planets soon lose theirs. The assumption makes sense not least because hydrogen is the most abundant element in the universe.

Young, Shahar and Schlichting modelled the Earth forming from large clumps of matter – ‘planetary embryos’ – that have hydrogen-rich atmospheres, and aggregate due to gravity. Researchers in Japan had already hypothesised the Earth starting out with a hydrogen-rich atmosphere that became the source of water on the planet as it oxidised, but by modelling the evolution of this formation in terms of the equilibrium of the planet as a whole, Young, Shahar and Schlichting hit on a number of points that resolve other longstanding riddles in geoscience, and may fill in some of the details of how hydrogen might find its way into the mantle.

The flux of primordial hydrogen can be the driving force for many geological processes

The aggregation of these planetary embryos to form Earth would convert a vast amount of gravitational potential energy into heat, giving rise to temperatures of thousands of degrees. In the absence of a crust – it would take hundreds of millions of years for the Earth to cool enough for a crust to form – not only could hydrogen from the atmosphere dissolve into the magma of the mantle but the frenzied convection currents from such high temperatures would whisk the mantle magma, carrying some of the dissolved hydrogen to the core. Traditionally it has been assumed the core is made of iron and nickel as this matches the composition of meteorites. As long ago as the 1950s, however, people began to notice that the speed of seismic waves travelling through the Earth did not add up for a core with such a high density as iron and nickel. With hydrogen present the environment would be sufficiently reducing for silicon to displace some of the iron in the core. As a lighter element silicon would contribute to lowering the core density, as would hydrogen itself. Compositional analysis of rocks in the crust suggests there is not as much silicon as expected from comparisons with meteorites, and a previous study by Shahar of the isotopic ratios of silicon in the context of which isotopes might bond more readily with the metals in the core had already suggested an isotopic argument for the presence of silicon in the core too.

It is conceivable within such a model that hydrogen might be rejected into the outer core as the inner core crystallised. The mismatch of concentrations with the mantle would then encourage hydrogen to diffuse into the magma where solubility changes for different temperatures and pressure could cause it to degas from the solution. However, this all hinges on guesswork around the solubility of hydrogen in metals at the relevant temperatures and pressures, which experiments struggle to verify, since the equipment needed to achieve those conditions is incompatible with taking direct measurements. ‘Hydrogen solubility in metal as a function of temperature and pressure is one of the new holy grails, I would say, of planetary research,’ says Young.

Zgonnik is already convinced that the Earth’s core and mantle are enriched with primordial hydrogen. ‘The flux of primordial hydrogen can be the driving force for many geological processes over the history of the planet,’ he says. He flags the work of Russian geologist Vladimir Larin, who spent a lifetime developing the theory of a primordially hydrogen-rich Earth. Zgonnik’s own recent work appraising the chemical composition of the Earth and other planets alongside researcher Hervé Toulhoat of the Sorbonne University in France also suggests the Earth formed from material extremely rich in hydrogen. While most of this hydrogen would have dissipated into space, Zgonnik says up to 4 wt% could have been trapped, bonded in the form of various hydrides. These, he suggests, would continue to decompose, sustaining a flow of hydrogen that contributes to a number of geologic processes, including volcanism and the formation of oceans, among others.

While these remain points for further scholarly interrogation, the presence of accessible hydrogen reserves in the Earth’s crust is far from being of purely academic interest. Both France and Australia have already adapted their resource law to accommodate licenses to look for natural hydrogen, and there are numerous plans to drill wells for hydrogen in Australia and the US in the coming months, and the list of companies with vested interests is growing longer. If the site of NH2E’s success in drilling for hydrogen in Nebraska can be developed into a working centre for hydrogen retrieval, Zgonnik hopes it will demonstrate a gear change for the sector, showing once and for all ‘that the full cycle from exploration to production is being mastered’.

Anna Demming is a science writer based in Bristol, UK