Natural hydrogen the fuel of the 21st century

Much has been learned about natural hydrogen (H2) seepages and accumulation, but present knowledge of hydrogen behavior in the crust is so limited that it is not yet possible to consider exploitation of this resources. Hydrogen targeting requires a shift in the long-standing paradigms that drive oil and gas exploration. This paper describes the foundation of an integrated source-to-sink view of the hydrogen cycle, and propose preliminary practical guidelines for hydrogen exploration.


Let's go targeting natural hydrogen
The quest for sustainable energy supply at low environmental and economical cost is a major driver in the current context of the energy transition toward a low carbon society. Hydrogen (H2) is a carbon-free fuel par excellence because its oxidation (i.e. combustion, fuel cells) emits only water. However, to this point, H2 has only been regarded as an energy vector and not as a credible large-scale alternative to hydrocarbons since most production methods (e.g. methane steam reforming) currently in use only postpone CO2 emissions. Its massive production through water hydrolysis using renewable energy is also very challenging, as it requires both pure water resources and local storage infrastructures.
The discoveries of hundreds of natural H2 seepages, generally connected with circulation of hydrothermal fluids through ultramafic rocks [1][2][3], both under the seafloors and on the continents, remove these obstacles but raise important questions regarding the energy potential that these sources can represent [4] (Fig. 1). In addition, recent observations of intra-cratonic seepages and accumulations with no obvious genetic link to ultramafic formations challenges current understanding of H2 production and fate in the crust [5][6][7]. The natural production of hydrogen is a recent area of scientific research, but little is known about H2 generation, migration, consumption and potential accumulation. To date, there is neither exploration strategy nor any resource assessment, as practical guidelines for hydrogen targeting are lacking. It now appears necessary to study how H2 migrates and becomes trapped in the Earth's crust in order to reasonably assess the recovery potential of this primary natural production.

Hydrogen generation: challenging the "olivine monopoly"
Hydrogen is naturally produced by different reaction processes occurring in the Earth's crust (Fig. 2). The bests known are (i) the hydrothermal alteration of ultramafic rocks [8], (ii) the radiolysis of water due to U, Th, and K radioactive decay [9], (iii) the activity of certain fermentative anaerobic bacteria and cyanobacteria [10], (iv) the formation of FeS2 from FeS [11], (v) the decomposition of methane into C-graphite and H2 under upper mantle conditions [12], or even (vi) mechano-radical processes triggered by friction of fresh silicate mineral surfaces exposed in active faults [13]. Among these potential H2 sources, serpentinization of ultramafic rocks remains by far the most studied. Oxidation of Fe(II)bearing minerals (e.g. olivine, pyroxene) mainly into magnetite (Fe(II)/Fe(III) mixed oxide) leads to the concomitant reduction of water into H2 [1,8], according to: The first observations of natural gas emissions containing H2 date back almost a century [1,14], but these occurrences were considered as anecdotal until about twenty years ago. It was only with the discovery of H2-rich submarine hydrothermal vents in the late 1970s that systematic studies on the origin of natural H2 were undertaken [2,3]. New field studies were then conducted onshore, using the occurrence of obducted ultramafic rocks (e.g. Oman, Philippines, Turkey, Liguria-Italia, New Caledonia) as a prospecting guide [15], by analogy with the geological context of the mid-oceanic ridges (Fig. 2).
This brief history of both the first discoveries of natural H2 sources and their associated generation mechanisms led to the paradigm that "only hydrothermal alteration of olivinerich ultramafic rocks can produce significant amount of hydrogen". However, several geological observations suggest alternative processes to the olivine serpentinization for H2 generation.
(1) Hydrogen is a major gas component (up to 30%) trapped in evaporite formations, and in particular in salt deposit containing significant amount of carnallite and other hydrated potassic salts [16]. This hydrogen may have several origins: (i) production during early biodegradation of organic matter, (ii) water radiolysis due to elevated concentration of 40 K and 87 Rb, and (iii) exogenic sources and subsequent migration into evaporite formations.
(2) Hydrogen may be particularly enriched in coals. Levshounova [17] reports hydrogen content ranging from 2.9 to 40% in former Soviet Union coal derived gases. Hydrogen, produced at the early stage of organic matter maturation may remains trapped or adsorbed within the micropores of coals, as it is the case for methane.
Clearly, olivine serpentinization is not the only process able to trigger H2 production. Bacterial fermentation, water radiolysis, and hydrothermal alteration of other Fe(II)-bearing minerals such as amphibole, mica, chlorite or siderite are likely to play this role.

Hydrogen migration and reactivity: the known unknown
Once produced, H2 can react with oxidized elements -mineralized or dissolved in geological fluids -or diffuse toward the surface and escape into the oceans or the atmosphere. This hydrogen can then be either a source of energy for bacterial developments [9,20], or as a reagent for abiotic hydrocarbons synthesis [8], but rarely, if ever, as a carbon-free energy resource [4]. Its high mobility and reactivity at high T, as well as at low T in the presence of bacteria, are considered to prevent its accumulation in the geological media. This view was recently challenged by two major discoveries of geological environments where H2 is trapped in deep sedimentary formations overlying intra-cratonic crystalline basements.
i. Widespread H2 enrichment in water-saturated clay-rich rocks at 20°C surrounding the Cigar-Lake uranium deposits, Athabasca, Canada [21]. Thermal desorption measurements reveal that H2 is enriched up to 500 ppm (i.e. 0.25 mol.kg -1 of rock) in these water-saturated rocks having a very low total organic content (<0.5 wt%). Such H2 uptake is comparable and even exceeds adsorbed methane capacities reported elsewhere for pure clay minerals or shales. Up to 17% of H2 produced by water radiolysis over the 1.4 Galifetime of the Cigar Lake uranium ore deposit, accounting for about 500 tones, have been trapped in the surrounding clay alteration haloes. As a result, sorption processes on layered silicates must not be overlooked as they may exert an important control on the fate and mobility of H2 in the crust.
ii. Large accumulation of H2 in the Taoudeni Basin, Bourakebougou field, Mali [7]. Recent exploratory wells in Mali, close to the village of Bourakebougou, confirm the presence of an extensive H2 field featuring at least five stacked reservoir intervals containing significant H2 content (up to 98 vol.% of gas) and covering an estimated area higher than 8 km in diameter. The relatively pure H2 reservoirs are associated with traces of methane, nitrogen and helium. The geological stratigraphic accumulation of H2 is linked to the presence of multi overlaid doleritic sills and aquifers that seem to prevent upward gas migration and leakage.
New experimental data and reactive transport models are needed to investigate H2 transport and storage in deep geological environments. Fundamental data such as H2 solubility, vapor-liquid partitioning, or mineral adsorption under geologically relevant conditions are still lacking. The mechanism for H2 trapping may rely on a combination of both H2 adsorption properties and low solubility acting in concert as an impermeable barrier. Any attempt at modeling H2 behavior in deep geological media will remain misleading in the absence of quantitative data on these two key processes.

Hydrogen targeting: providing practical exploration guidelines
Finally, a practical application of the understanding of H2 behavior in the continental crust would be to develop a specifically design exploration guide. Currently, apart from targeting ophiolitic massifs or circular structures in flat sedimentary terrain overlaying cratonic basements, there is no extended exploration guide, nor any dedicated exploration wellapart the very recently drilled exploratory wells, that have been completed close to a fortuitous discovery made in the late 1980's near Bourakebougou, Mali [7]. Knowledge of hydrogen migration and accumulation in the crust is too limited to seriously consider an extensive exploration campaign. This is not surprising since 99% of the deep drilling programs are dedicated to oil and gas exploration/production in geological contexts that are not particularly relevant for hydrogen targeting. Here, we describe what could be a preliminary exploration guide based on a global 'source-transport-accumulation' understanding of H2-concentrating process and combining methods used for ore targeting (so-called strategic, tactic and punctual targeting).
First, one of the key points related to potential source areas is the presence of gravity and magnetic anomalies that are typically characteristic of iron-rich rocks. These rock formations with strong anomalies, such as ultramafic rocks, or peralkaline granites, can be found at the outcrops but also are observed a few thousand meters below the continental surface. These formations often correspond to ophiolitic sutures or to peridotite massifs sandwiched during orogenic phases. The presence of Archean greenstone belts (e.g. Canadian and Fennoscandian Shields) containing ultramafic rocks may also represent excellent H2-producing zones either via serpentinization, or water radiolysis [21]. There are also intra-cratonic zones (e.g. Russia, North Carolina in the USA, or Mali) that may display significant H2 fluxes [5][6][7], with no clear link with ultramafic rocks. Hydrothermal alteration of other Fe(II)-bearing minerals such as those listed above (amphibole, mica, chlorite) may be envisaged in these latter cases.
Second, the structural/tectonic context and the presence of faults deeply rooted in the basements capable of draining a potential deep and scattered source will certainly play a very important role. A structural, gravimetric and seismic survey will be relevant to target potentially active deep faults able to drain hydrogen produced at depth where the P-T conditions are optimal.
Third, if there are storage areas (i.e. a reservoir) of H2 at depth, it will also be necessary to consider the nature of the different lithologies (e.g. mineralogy, porosity, total organic content) and develop prospection methods similar to those implemented for ore targeting. Surface seepages may be either in connection with the source rock or with an intermediate leaking reservoir. Thus, mapping H2 concentration anomaly in soils and outcropping rock formations at different scales will be extremely instructive when superimposed to geological, seismic, and gravimetric data. Many unknown seepages are awaiting for explorers to reveal their potential.