Applied Geochemistry 145 (2022) 105396

Contents lists available at ScienceDirect

Applied Geochemistry
journal homepage: www.elsevier.com/locate/apgeochem

Natural hydrogen migration along thrust faults in foothill basins: The North
Pyrenean Frontal Thrust case study
N. Lefeuvre a, *, L. Truche a, **, F.-V. Donzé a, F. Gal b, J. Tremosa b, R.-A. Fakoury c, S. Calassou c,
E.C. Gaucher c, 1
a
Université Grenoble Alpes, CNRS, ISTerre, F, 38058, Grenoble, Cedex 9, France
b
BRGM, 3 Avenue Claude Guillemin, 45060, Orléans, France
c
Total SA, CSTJF, F, 64018, Pau Cedex, France

A R T I C L E I N F O A B S T R A C T

Editorial handling: Marcello Liotta The existence of geological fluids rich in natural hydrogen (H2) raises the question about the energy potential of
this carbon-free resource. However, to date there is no exploration strategy based on robust methodologies and
Keywords: pathfinders. Therefore, it is important to develop an exploration guide that is not only focused on surface gas
* The western Pyrenees as a fertile geological monitoring, but that also considers the local deep geological setting integrating the entire hydrogen system from
setting for H2 production and migration
source to trap or leakage into the atmosphere.
* Soil gas and geophysical surveys carried out
The northwestern Pyrenees, and particularly the Mauléon Basin, represent a promising geological environ­
to detect gas migration
* A deep-fluid migration along the North ment for natural H2 exploration for at least four reasons. First, an ultramafic mantle body is emplaced at shallow
Pyrenean Frontal thrust (NPFT) depth below the basin under pressure-temperature conditions favorable to serpentinization. Second, major faults
such as the North Pyrenean Frontal Thrust constitute large-scale fluid flow convergence and drainage. Third,
hydraulic gradients imposed by sharp reliefs and combined with temperature and pressure gradients trigger fluid
migration. Fourth, impermeable sedimentary formations or caprocks such as evaporites or claystones overly
porous reservoir rocks that could constitute traps for accumulating H2.
To investigate H2 migration at the fault scale, we present new geochemical and geophysical data recorded
along the North Pyrenean Frontal Thrust. Based on both soil gas and electromagnetic transects, we reveal the
presence of a gas-draining fault. Soil gas concentration (H2, CO2, CH4 and 222Rn) recorded at 1 m depth increases
when approaching the North Pyrenean Frontal Thrust. The maximum H2, CO2 and 222Rn concentrations recorded
in the fault zone are 822 ppmv, 10.3 vol% and 57 kBq.m− 3, respectively - whereas their local background
concentrations are by 1–2 orders of magnitude lower: 10 ppmv, 0.2 vol% and 0.3 kBq.m− 3 respectively. Our
geochemical and geophysical data support the concept of a deep-fluid migration along the detected fault plane.
In addition, the study of historical well data combined with the most recent geological and geophysical surveys
carried out in the region, highlights zones where H2 could accumulate at depth. The Triassic salt formations,
located at 2800 to 4000 m deep beneath the Mauléon Basin, represent the most promising trap for H2 in the
northwestern Pyrenees.

1. Introduction Faults may act as drains or conduits, thereby promoting H2 migration

toward the surface and the atmosphere. Since hydrogen is scarce in the
Chemical and mechanical processes in crustal rocks can produce atmosphere (531 ppb) and shallow subsurface environments, its pres­
dihydrogen – hereafter hydrogen or H2 (Charlou et al., 2002; Truche ence at elevated levels in soil gases, water springs and gas seepages may
et al., 2020; Klein et al., 2020; Boreham et al., 2021). Once released, H2 be a good proxy to reveal deep fluid migration.
is highly mobile, poorly soluble in water, and may migrate freely in On some aspects, H2 mobility may be compared to helium (He), as
fracture networks (Nivin, 2019; Zgonnik et al., 2019; Xiang et al., 2020). both compounds display similar migration properties and may be deeply

* Corresponding author.
** Corresponding author.
E-mail addresses: nicolas.lefeuvre@univ-grenoble-alpes.fr (N. Lefeuvre), laurent.truche@univ-grenoble-alpes.fr (L. Truche).
1
Present address: Institute of Geological Sciences, University of Bern, CH-3012 Bern, Switzerland

https://doi.org/10.1016/j.apgeochem.2022.105396
Received 28 April 2022; Received in revised form 13 July 2022; Accepted 14 July 2022
Available online 22 July 2022
0883-2927/© 2022 Elsevier Ltd. All rights reserved.
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

sourced in the Earth’s crust. However, unlike He, H2 can be particularly Sauveterre-de-Béarn.
reactive. Hydrogen is a reducing agent that can be involved in plethora
of redox reactions either at high temperature (T > 200 ◦ C) or at low 2. Geological setting
temperature (T < 100 ◦ C) in the presence of bio-mediators. There is only
a small temperature window (100 < T < 200 ◦ C) where H2 can be The Pyrenean orogenic belt is a N100◦ -trending mountain belt
considered as relatively inert because the kinetics of both biologically- formed during late Cretaceous to Early Miocene convergence between
mediated and abiotic reactions, like thermochemical sulfate reduction the Eurasian and Iberian plates (Choukroune, 1989; Muñoz, 1992). The
or Fischer-Tropsch type reactions, are inhibited or sluggish (Truche North Pyrenean Zone (NPZ), including the Mauléon Basin, is a stripe of
et al., 2009; McCollom, 2013). In addition, H2 may also be produced by about 10–40 km wide belonging to a fold-and-thrust belt tightly con­
microbial activity, like fermentation (Schwartz and Friedrich, 2006; nected with the presence of evaporite formations and salts diapirs
Kubas, 2007; Gregory et al., 2019). Hydrogen abiotic reactivity, com­ (Fig. 1). The NPZ is also characterized by numerous ultramafic outcrops
bined with microbial production and consumption may thus complexify that reveal the occurrence of several serpentinization stages (Ducoux
the interpretation of both the origin and the concentration of H2 in soil et al., 2021; Fabriès et al., 1991, 1998; Tichadou et al., 2021). The NPZ is
and surface seeps (Truche et al., 2009; Gregory et al., 2019). bounded by major east-west tectonic accidents especially prominent to
Up to now, H2 monitoring in soil was mostly used to identify active the north of this zone as highlighted by the North Pyrenean Frontal
faults. Wakita et al. (1980) reported strikingly high level of H2 (up to 3 Thrust (NPFT) separating the NPZ from the Aquitaine Basin (Ford and
vol%) in soil gases near the Yamakasi fault (Japan). This discovery gave Vergés, 2020). Moreover, Barré et al. (2021) have shown that the NPFT
rise to numerous studies that confirmed anomalously high H2 concen­ is a major structure responsible for the leakage of deep hydrocarbons
tration in soils overlying active faults, particularly before earthquake reservoirs and, thus, would correspond to a favorable pathway for deep
occurrences (Sato and Mcgee, 1982, Sato et al., 1984, Sugisaki, 1984; fluids migration.
Ware et al., 1984; Dogan et al., 2007; Zhou and Coauthors, 2010; Li In this study, we focus on the Mauléon Basin located in the western
et al., 2013; Gong et al., 2015; Fang et al., 2018; Xiang et al., 2020). The part of the NPZ (Fig. 1). This basin was formed during the Cretaceous
release of occluded gases during rock fracturing (Giardini et al., 1976) hyper-extension. At this period, mantle rocks were exhumed toward the
has been first proposed to explain gas burst during stress change, but southern part of the basin. The Mauléon Basin was inverted as a tectonic
Wakita et al. (1980) revealed that surface reactions between water and pop-up delimited by two major detachment faults: the NPFT to the north
freshly crushed rocks may also produce H2. This observation was further and the North Pyrenean Fault (NPF) to the south (Masini et al., 2014).
confirmed by laboratory experiments (Freund et al., 2002; Kameda The basin in itself is mainly composed of Mesozoic formations and is
et al., 2003; Saruwatari et al., 2004; Hirose et al., 2011; Telling and decoupled from the basement by evaporite formations (Biteau et al.,
Coauthors, 2015; Torre and Coauthors, 2020). However, many other 2006; Ducoux, 2017).
H2-producing reactions may be at play in the geological environment Recent geophysical studies have revealed an atypical structure under
(Truche et al., 2020). H2 may also be massively produced during ul­ the Mauléon Basin: i) a strong positive magnetic anomaly (>60 nT),
tramafic rocks serpentinization and by water radiolysis (Klein et al., resulting potentially from magnetite production during serpentinization
2020). The occurrence of H2 seepages or H2-rich geological fluids is of mantle rocks (Malvoisin et al., 2012; Oufi et al., 2002; Toft et al.,
particularly well documented in ultramafic environments like in sea­ 1990), ii) a strong Bouguer gravity anomaly (>20 mGal) (Grandjean,
floor hydrothermal vents or in ophiolites (Abrajano et al., 1988; Charlou 1994; Pedreira et al., 2007; Pedrera et al., 2017; Vacher and Souriau,
et al., 2002, 2010; Chavagnac et al., 2013a, 2013b; Crespo-Medina et al., 2001), and iii) high seismic velocities (Vp ≈ 7.3 km s− 1, Vs ≈ 4.2 km
2017; Etiope et al., 2011a,b; 2013, 2017; Lyon and Hulston, 1984; s− 1). The two latter anomalies are interpreted as reflecting the presence
Monnin and Coauthors, 2014; 2021; Neal and Stanger, 1983). High H2 of dense mantle materials at shallow depth (Wang et al., 2016; Gar­
concentrations in soils in connection with putative deep H2 sources have cía-Senz et al., 2020). All these geophysical features are very well
also been reported in intra-cratonic basins (Donzé et al., 2020; Larin correlated in space (Lefeuvre et al., 2021). Moreover, the Maupasacq2
et al., 2015; Moretti et al., 2020; Prinzhofer et al., 2019; Zgonnik et al., passive seismic survey carried out in 2017 above the Mauléon Basin
2015). Thus, the relevance for H2 monitoring in soil is not restricted to revealed a very unusual seismic sequence of 150 very closely spaced
active seismic faults, but can also be extended to any other location events with similar waveforms at four-km depth near the village of
where H2-fertile geological settings are envisioned. Natural H2 sources Sauveterre-de-Béarn (Sylvander et al., 2019; Lehujeur et al., 2021;
could potentially represent a primary energy resource, and surface Chevrot et al., 2022). The behavior of these seismic events corresponds
detection of H2 seeps may probably represent the most self-evident to a seismic swarm without mainshock-aftershock relationship. A slow
pathfinder provided more advance exploration guides could be set up. upward migration of these seismic events is also recorded. These ob­
A fault zone is a favorable pathway for fluids migration in the crust, servations are interpreted as the result of fluid migration and diffusion
due to the presence of numerous connected and opened fractures along (Sylvander et al., 2019).
the fault plane (Baubron et al., 2002; Géraud et al., 2006). Faults may All together, these data corroborate the presence of a dense mantle
thus connect a deep H2 source or reservoir to the surface. Here, we test body, located at 8–10 km below the Mauléon Basin that may trigger H2
this hypothesis by monitoring soil gases (H2, CO2, CH4, and 222Rn) along generation through ultramafic rocks serpentinization. From this source
the North Pyrenean Frontal Thrust located to the north of the Mauléon rock, H2 can migrate along the major faults of the region, such as the
basin in the French Pyrenean foothills. This site is chosen because a NPFT, and reach the surface or accumulates within a potential trap.
previous regional soil gas survey coupled with recent geophysical data Lefeuvre et al. (2021) revealed several areas of high H2, 222Rn and CO2
suggested the presence of a favorable geological setting for H2 migration concentrations to the north of Mauléon Basin and more precisely near
and drainage in the northwestern Pyrenean foothills (Lefeuvre et al., Sauveterre-de-Béarn, which is located close to the NPFT (Fig. 2).
2021). Thus, to better investigate H2 migration at the fault scale, we
present new geochemical and geophysical data sets near

2
https://www.convergent-margins.com.

2
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

Fig. 1. A) Geological map of the North Pyrenean Zone (NPZ) with the location of the studied site near Sauveterre-de-Béarn (red box). This site is close to the North
Pyrenean Frontal Thrust (NPFT) that also encompasses the Saint Palais Thrust (St-PT) and the Saint Suzanne Thrust (St-ST). The southern part of the NPZ is bounded
by the North Pyrenean Fault (NPF). The locations of existing oil&gas wells and proposed H2 exploration boreholes are also shown on this map. B) Structural map at
the continental scale showing the location the Iberian Peninsula and the Pyrenean belt in-between France and Spain (modified from Ducoux et al., 2021). (For
interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

3. Materials and methods Sensitivity (CS) forward models to provide electrical conductivity
models from which we can deduce the local resistivity fields.
3.1. Geophysical survey

An electromagnetic (EM) survey using a GF® Instrument’s CMD 3.2. Soil gas survey
DUO electromagnetic conductivity meter which works on 925 Hz, has
been performed along a NNE-SSW 200-m long profile located on the A soil gas survey was conducted at about 1 km to the north of Sau­
south side of the NPFT (profile #C shown on Fig. 2). This tool allows us veterre-de-Béarn as shown in Figs. 1 and 2. GPS coordinates of the site
to image resistivity contrasts induced by different lithologies in the are given in Fig. 2. The studied site extends over 16.9 Ha (41.76 acres)
shallow subsurface. It is also possible to detect zones of fluid circulation. and is crossed by the NPFT that is geomorphologically visible through a
The EM data were acquired using three different inter-coil distances of steepening slope ranging from 12.8% to 35.5% at its steepest (Fig. 2).
10, 20 and 40 m, allowing a maximum depth of investigation of 60 m. The fault scarp is about 40 m high and runs through the landscape over
For each measurement, both vertical and horizontal positioning of the several tens of kilometers along a NNW direction. A total of 240 in situ
coils were used. The transmitter generates sine wave of magnetic field gas measurements were carried out during spring and summer seasons
and the induced secondary magnetic field from the ground is recorded. in 2018 and 2021 (Fig. A1). The slope that characterizes the fault scarp
Then, the data are inverted using a Python-based open-source EMI is mostly covered by meadows and groves, whereas cornfields are
inversion software, EMagPy (McLachlan et al., 2021). The inversion located on top of the hills and on its backdrop to the north. The first
algorithms utilize either Maxwell-based FS (Full Solution, refers to the meter of soil is mostly composed of pebble-rich orange colored alluvial
full solution of Maxwell’s Equation) forward models or Cumulative formations. A 125 cm deep soil core sample was collected in June 2021
next to well #1 (Fig. 2). A clear redox transition highlighted by a color

3
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

Fig. 2. 3D block of the studied area at Sauveterre-de-Béarn. Drone aerial photogrammetry is used to map the local topography on top of a geological cross-section.
The studied area is crossed by a major fault highlighted by the topography and the 40 m high fault scarp that separates the Cenomanian formation to the south from
the Senonian formation to the north. Soils are mostly composed of alluvial deposit (clays and gravels). The locations and associated GPS coordinates of the four soil
gas transects (#A, #B, #C and #D) and the three soil-gas accumulation wells are shown on the map and given as Supplementary. The location of the electromagnetic
profile (EM) corresponds to the profile #C.

change from orange to green/blue is visible at about 1 m depth (Fig. 3). Soil gas concentrations of H2, CO2, CH4, and 222Rn were also
The clay fraction is mostly composed of smectite, muscovite, and measured in the same area at 1-m depth using both a multi gas analyzer
kaolinite as revealed by X-ray diffraction on oriented mounting of the GA-5000 (GeoTech®) and an AlphaGuard DF2000 radon detector
<2 μm fraction of the soil samples. (Bertin instrument®) according to the method described by Lefeuvre

Fig. 3. Schematic diagram of a gas accumulation well and its connection to a multi gas analyzer (GA-5000®), a radon monitor (Alphaguard®) or a gas cylinder
(Swagelok®). Picture assemblages of soil samples collected along a 125 cm long core located next to accumulation well #1 are shown in the inset (lower middle).
4
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

et al. (2021). Soil gas probes were inserted at 1 m depth immediately using an IRMS (Isotope Ratio Mass Spectrometer) Isoprime (Ele­
after drilling of a 1 cm orbital diameter (OD) hole using a portable drill mentar®) with a stable C isotopic ratio referenced to Vienna Peedee
in percussion mode only (no rotation of the drill bit). In addition, three Belemnite (VPDB).
gas accumulation wells were installed at different altitudes along the
fault scarp (Figs. 2 and 3). These boreholes were drilled at 1 m depth, 4. Results
fitted with a PVC tube (4 cm OD, bottom screened section of 10 cm
length), clogged by a silicone plug to be airtight and connected to a 4.1. Electromagnetic profile
sampling valve at the surface end. The sampling valve allowed to con­
nect and fill a gas tight Swagelok® stainless steel gas cylinder (40 mL) The electromagnetic profile (Fig. 4) recorded along the transect #C
with soil gas accumulated in the borehole. The gas cylinders were pre­ shows a significant vertical variation of resistivity that decreases from
viously evacuated (primary vacuum). Gas samples were then analyzed 1000 to 10.Ω m at a depth of approximately 5 m. Laterally, the resistivity
in the lab to quantify H2, CO2, and CH4 concentration using a Perkin increases sharply from 0 to 100 Ω.m close to the top of the slope (i.e. to
Elmer® CLARUS 500 Gas Chromatograph (GC) equipped with a Thermal the north of the profile). In addition, two high resistivity spots ranging
Conductivity Detector (TCD), a 2-m long column (Restek® Shin Carbon from 500 to 1000 Ω.m are detected at about 15 m depth and at a distance
ST80/100), and Ar as a carrier gas. The gas concentration ranges and of about 130 and 170 m from the Northern part of the profile.
typical accuracies of all analytical tools are shown in Table 1. Isotopic
data (δ13C–CO2 and δ13C–CH4) of these sampled gases were acquired

Table 1
222
Gas concentration ranges (H2, CO2, CH4, O2, Rn) and typical accuracies of the gas analyzers used in this study: GA5000, Alpha Guard DF2000 and GC.
Gas Range Typical accuracy

GA 5000

CH4 0- 100 vol% 0 - 70 ± 0.5 vol% 70 - 100 ± 1.5 vol%

CO2 0- 100 vol% 0 - 60 ± 0.5 vol% 60 - 100 ± 1.5 vol%
O2 0- 25 vol% 0 - 25 ± 1 vol%
H2 0- 1000 ppmv ±2 ppmv

Alpha Guard DF 2000

222
Rn 2 – 2 × 106 Bq.m− 3
±5 cpm at 100 × 106 Bq.m− 3

Gas chromatography

H2 Quantification limite 5 ppm Precision of 1% at the 95% confidence level

CO2 Quantification limite 500 ppm Precision of 1% at the 95% confidence level
CH4 Quantification limite 10 ppm Precision of 1% at the 95% confidence level

Fig. 4. Cross-section of the electromagnetic survey carried out along profile #B (see Fig. 2 for the exact location).

5
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

4.2. Soil gas profile and 222Rn concentrations. However, subsequent measurements carried
out on June 25th, 2020 and June 11th, 2021 revealed very high H2
All the measurements were acquired under similar climatic condi­ concentration values (up to 407 ppmv) close to the fault zone. Note that
tions to reduce the effects of soil moisture, temperature and atmospheric the radon concentration in soil gas remains low in comparison to profiles
pressure, but this study is not intended to document the effect of such #B and #D.
parameters. All the gas measurements were done under dry and sunny Profile #D (Fig. 6c), carried out in June 7th, 2021, is comparable to
weather conditions during spring and summer seasons (Fig. A1; the profile #B as it also shows a steady concentration increase for all gases
weather forecast recorded during the sampling period is shown in while approaching the fault zone. Hydrogen, CO2 and 222Rn concen­
Supplementary Information). For purpose of comparison, the atmo­ trations remain at low values far from the fault, at the bottom of the fault
spheric concentration of CO2, H2 and 222Rn are 415 ppmv, (https://gml. scarp: 44 ppmv, 0.1 vol% and à 0.15 kBq.m− 3, respectively. At the vi­
noaa.gov/ccgg/trends/), 531 ppbv (Novelli et al., 1999) and 0.01 kBq. cinity of well #1, close to the fault zone, the H2, CO2 and 222Rn con­
m− 3 (Baubron et al., 2002), respectively. centration values are remarkably higher: 98 ppmv, 0.6 vol% and 11 kBq.
The soil gas profile #A was collected on March 29th, 2018 and was m− 3, respectively. Subsequent measurements carried out on June 11th,
designed to cross the NFPT fault over several hundreds of meters on both 2021 near the fault zone on profile #D have revealed even higher H2
sides in order to monitor the gas emissions possibly related to the fault concentration (up to 407 ppmv).
but also to evaluate the local background values (Fig. 5). At both ends of
the profile (0 and 525 m), H2 and CO2 concentration values remain
below 10 ppmv and 0.5 vol%, respectively. However, in the vicinity of
the fault (located at about 220 m in this profile) H2 and CO2 concen­
tration values increase up to 177 ppmv and 1.8 vol%, respectively. One
can also notice that this profile displays other spots of high H2 (up to 140
ppmv) and CO2 (up to 2.2 vol%) concentrations downside the hill along
the fault scarp (i.e. toward the south). Subsequent measurements carried
out on June 25th, 2020 revealed even higher H2 and CO2 concentration
values of 177 ppmv and 10.3 vol%, respectively, at the level of the fault
zone.
Three others 200-m long transects were monitored in the study area
(Fig. 2). Profile #B and #C are perpendicular to the fault, whereas
profile #D is slightly oblique to the fault and intersects the three accu­
mulation wells shown in Fig. 2.
The profile #B (Fig. 2) is located in the western part of the study area
and was carried out in June 25th, 2020. All gas concentration values (H2,
CO2 and 222Rn) increase steadily along the slope and reach their
maximum at the top of the fault scarp (Fig. 6a). At the bottom of the
slope, 190 m away from the fault, the concentration values of H2, CO2
and 222Rn are 15 ppmv, 0.2 vol% and 0.3 kBq.m− 3, respectively, while
they reach 244 ppmv, 2.3 vol% and 3.8 kBq.m− 3 in the vicinity of the
fault plane, at the top of the slope. Along this profile, a preliminary
measurement carried out on March 29th, 2018 revealed even higher H2
and CO2 concentration within the fault zone at the top of the slope: 632
ppm and 5.7 vol%, respectively.
Along profile #C carried out in June 8th, 2021 (Fig. 6b), CO2 and
222
Rn concentrations also increase while approaching the fault zone
with maximum values of 2.9 vol% and 0.3 kBq.m− 3, respectively.
Hydrogen concentration remains at a constant and high value (around
100 ppmv) all along the profile, with no particular correlation with CO2

Fig. 6. H2, CO2 and 222Rn concentrations measured at 1 m depth along the
three profiles #B, #C and #D (see Fig. 2 for location). (a) and (b) Profiles #B
and #C located perpendicularly to the fault and starting at the top of the fault
Fig. 5. H2 (yellow dots and stars) and CO2 (red triangles and stars) concen­ scarp (0 m) and ending at the bottom of the slope (200 m away from the top).
tration at 1 m depth in soil along the profile #A (see Fig. 2 for exact location). (c) Profile #D intersects the three accumulation wells, with well #1 being the
The circles and triangles stand for measurements recorded on March 29th, highest and closest to the fault. The yellow stars correspond to measurement
2018. The stars stand for measurements performed on June 25th, 2020. (For carried out along the profile at different periods (given as supplementary). (For
interpretation of the references to color in this figure legend, the reader is interpretation of the references to color in this figure legend, the reader is
referred to the Web version of this article.) referred to the Web version of this article.)

6
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

4.3. Soil gas accumulation ppmv, 52 ppmv and 57 kBq.m− 3, respectively. Thus, the trend in gas
accumulation indicates again an increase in gas concentration when
Three wells (1-m deep) were drilled at three different distances from approaching the fault zone and confirms a H2-rich gas flow near the
the fault trace, such that well #3 is the furthest and the well #1 is the fault.
closest (Fig. 7). Directly after drilling, the wells were flushed with air, Two isotopic analyses of δ13C–CO2 and δ13C–CH4 were performed on
fitted with PVC tubes and then sealed. Gas concentration measurements gas samples from the well 1# at one-week intervals. The isotopic values
were initially carried out with the GA-5000 instrument on August 26th, are − 12.6‰ and − 19.0‰ for δ13C–CO2 and -49.9‰ to − 50.1‰ for
2020 immediately after flushing the wells. These measurements δ13C–CH4. such values could indicate a thermogenic origin for CH4. For
revealed the absence of H2 and CO2 in the boreholes (concentrations CO2, the enriched ratio (− 12.6‰) could also point to a possible ther­
below quantification limit of 5 ppmv and 0.1 vol% respectively). The mogenic origin but the depleted ratio (− 19.0‰), measured on a sample
boreholes accumulated gas over 189.48 h (7.89 days) and were then coming from the same location at a different time, does not allow to
sampled using the evacuated gas cylinders. The concentrations of H2, confirm this hypothesis, as − 19‰ (and 0.8% of CO2) can typically be
CO2 and CH4 and the δ13C–CO2 and δ13C–CH4 were measured in the lab, produced by respiration processes in the soil under temperate climate.
but the 222Rn activity was measured directly in the well with the radon Complementary investigations are needed to clearly assess the origin of
detector. the CO2 (Fig. 8). In addition, a O2 vs CO2 concentration plot (Fig. A2)
After one week of accumulation, the well #3 has no measurable shows values that are clearly above the respiration trend line and
concentrations of both H2 and CH4 (<5 ppmv), suggesting the absence of consistent with exogenous CO2 addition. This observation plays also in
measurable H2 and CH4 flow towards the surface at the level of well #3. favor of a deep source of CO2 migrating along the NPFT.
However, there is a noticeable increase of both CO2 (996 ppmv) and
222
Rn (1 kBq.m− 3) concentrations. The increase of CO2 concentration at 5. Discussion
a level equivalent to two times the atmospheric one proves that the
installation is airtight. In comparison, the well #2 is much more 5.1. Gas concentration anomalies related to the fault
enriched in CO2 and 222Rn with concentrations reaching 4383 ppmv and
20 kBq.m− 3, respectively. Methane is also present at 18 ppmv concen­ The electromagnetic data allow the identification of subsurface
tration (ten times higher than the atmospheric concentration of 1888 structures down to 60 m depth (Fig. 4). The very high resistivity spots
ppbv in June 2021, https://gml.noaa.gov/ccgg/trends_ch4/), whereas (>500 Ω.m) recorded at about 140 and 170 m along the profile might be
H2 concentration is still below the quantification limit of the GC tech­ due to the presence of more resistive geological formations such as
nique (5 ppm). The highest CO2, H2, CH4 and 222Rn concentrations are alternating layers of marl and limestone. These alternating layers,
recorded in well #1 that is located close to the fault zone: 2 vol%, 822 known as flysch, are well visible below the Sauveterre bridge (road

Fig. 7. Drone aerial photogrammetry of the study area with location of the fault and accumulation wells. A geological cross section is superimposed to this image.
The concentrations of H2, CO2, CH4 and 222Rn recorded in September 3rd, 2020 after 8 days of accumulation in the three wells drilled to 1 m depth are plotted on top
of this figure.

7
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

very well correlated with the location of the fault zone as highlighted by
the terrain geomorphology and our EM field investigation (Figs. 4 and
6). Moreover, the carbon isotopic fractionations of both CO2 and CH4
display a thermogenic signature that supports the existence of a deep gas
source. Dissolved CO2 and CH4 from the Salies-de-Béarn salty spring, a
neighboring location located along the same fault, display similar car­
bon isotopes signature, thus suggesting a similar origin for these mole­
cules (Fig. 8). The NPFT may thus act as a preferential drain for deep
fluid migration. This fault crosses several geological domains, alter­
nating claystones and limestones. Considering its important vertical
extent of several kilometers with a low pressurization above 6000 m
depth (well data log, Bellevue-1), we can indeed expect an important
clay smeared gouge inducing a low lateral permeability (Wibberley
et al., 2017). This means that channeling will indeed occur along the
fault. We also point out that the channeling zones can be very narrow,
less than 10% of the connected fracture sets. In the present case, H2
Fig. 8. Genetic diagram of δ13C–CO2 vs δ13C–CH4 (from Milkov and Etiope, 2018) produced by mechano-radical reaction along the fault (Kita et al., 1982)
superimposed on the isotopic data from gas samples (CR: CO2 reduction; F: could be excluded as no earthquakes with magnitude >0.7 Mw at less
Fermentation). Salies-de-Béarn isotopic data were acquired by BRGM (Tremosa than 10 km depth were recorded 30 km around the study area between
et al., 2022).
2018 and 2022. Currently, the Pyrenean geological setting is in a
post-orogenic situation which induces low-magnitude earthquakes.
D933) along the Gave d’Oloron river stream (1 km to the south east of Third, the high 222Rn concentrations (>1 kBq.m− 3 up to 57 kBq.m− 3)
the studied area). The high horizontal resistivity contrast located at 5 m testify to a gas flux that cannot be sourced from the soil and upper layer
depth to the north and decreasing towards the south may well corre­ of the ground. Indeed, radon is a gas produced during the radioactive
spond to a geological boundary between a water-depleted formation and decay chains of Th and U. Due to its short half-life (3.82 days), 222Rn
the water-saturated clays. This subsurface structure agrees with litho­ must be rapidly transported upwards using a carrier gas such as CO2 to
logical description from drilling in the vicinity of Sauveterre (Tronel, maintain a high concentration. High concentrations of both 222Rn and
1996). On the other hand, the thickening of the fluid-rich formation CO2 in soil gases thus highlight active fluid circulations (Fig. A2; Bau­
recorded near the top of the profile (NNE) is a possible evidence of a bron et al., 2002). Faults are well known to act as channels for rapid fluid
high-water saturation level in relation to the fault (Figs. 1 and 4). This migrations and their presence is often correlated with soil gas anomalies
latter interpretation of the EM data is well supported by soil gas moni­ (Byrne et al., 2020; Ciotoli et al., 2007). All together, these geochemical
toring in the three accumulation wells. Indeed, soil gas concentrations data support deep-fluid migration along the detected fault plane.
along the profiles, and in particular along profiles #B and #D, appear to
be very well correlated with both the terrain geomorphology and the 5.3. An H2 production system at Sauveterre-de-Béarn?
shallow geological structures as imaged by EM.
Sauveterre-de-Béarn is located next to Salies-de-Béarn, a village
5.2. A deep gas leakage along the fault? known for its salty springs. These two neighboring locations are crossed
by the NPFT fault. The temperature of the salty springs is relatively cold
The H2 soil gas concentration anomalies may reflect deep geological ranging from 17 to 18 ◦ C, but the >10,000 years long residence time of
fluid migration. Hydrogen in particular may be produced at depth by water (apparent age), as deduced from δ14C and tritium dating, implies
various abiotic processes such as mechano-radical reactions, water deep fluid circulations. The high salinity of these springs (up to 300 g.
radiolysis, and serpentinization. However, one cannot ignore that soil L− 1 eq. NaCl) is due to the presence of both a salt diapir and Triassic
gas concentration and composition, including H2, can be strongly evaporites at depth (Fig. 9; Berard and Mazurier, 2000; Tremosa et al.,
affected by biological, meteorological (e.g. temperature, hygrometry, 2022). The Salies-de-Béarn springs show an enrichment in He (in both
atmospheric pressure) and pedological factors (e.g. soil composition, dissolved and free gas phases; 0.03 vol% and ~4 × 10− 6 mol.L− 1
vegetation). Such surface processes may overprint the signature of deep respectively) with a 3He/4He ratio of 0.115 Ra. This isotope signature
fluid migration. Long term monitoring, coupled to stable isotopes frac­ indicates an origin of helium predominantly in the continental crust
tionation and noble gases analysis may help to resolve the contribution with c.a. 2% input from the subcontinental mantle (Tremosa et al.,
of different sources to the surface signal, but this is not always the case 2022). These He isotopes data together with the carbon isotope signa­
(Gal et al., 2011; Romanak et al., 2012). In the following section, we tures of CO2 and CH4 (Fig. 8) support a deep origin of these gases and
gather three strong arguments that support at least the contribution of probable migration along the NPFT faults.
one deep source of H2. The Arbailles massif, located to the south of the Mauléon basin
First, the hydrogen-producing bacteria are composed of a metallo- (Fig. 1), is interpreted as an early Cretaceous half-graben with deeply
enzyme called hydrogenase, which can be used both to produce and to rooted faults connected to the Triassic evaporites and affecting the
consume H2 (Gregory et al., 2019; Peters et al., 2015). The most com­ basement (Ducasse et al., 1986; Lagabrielle et al., 2010; Saspiturry et al.,
mon process for producing H2 is the fermentation of organic matter in 2019). In addition, this massif is composed of numerous caves located
soil. However, the H2 concentration in soil gas usually remains at low close to these faults. The karst formation is supposed to be linked to a
level, typically below 5 ppm, due to symbiosis relationships between H2 mixing of surficial water with H2S-rich fluids originating from the
producers and consumers (Chen et al., 2015; Kessler et al., 2019; Thauer thermochemical sulfate reduction of the Triassic evaporites (Laurent
et al., 1996). The high H2 concentrations recorded in the present study et al., 2019). Fluid mixing leads to the formation of sulfuric acid and
exceed by two orders of magnitudes those reported so far in soils. In thus the dissolution of the carbonate host rock. The presence of this very
addition, the vegetation (meadows and groves) and the nature of the soil peculiar sulfidic karst (Jones and Northup, 2021) demonstrates the
is very homogeneous over the studied area. Thus, there is no particular occurrence of deep fluid migration along faults connected with the
reason for a specific near-surface ecosystem to develop on top of the Triassic evaporites and probably the basement. Therefore, the meteoric
fault scarp and not along the slope or at the bottom of the hill. water infiltrating the Arbailles massif may percolate at depth thanks to
Second, the high H2, CO2 and 222Rn concentration anomalies are the topography-related hydraulic gradient and the connected fault

8
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

Fig. 9. Geological cross section of Mauléon Basin representing the H2 production, migration pathways and trapping zones as proposed in this study. The Sauveterre
area is indicated by the black box above the NPFT (cross section modified from Zolnai, 1973, and Saspiturry et al., 2019).

networks. Fluid-rock interactions with the mantle body, and subsequent The Triassic formations, and in particular the flank of the Salies
flow back to the surface in contact with the Triassic evaporites located diapir, is one of the most promising formation for H2 trapping in the
under the Mauléon Basin, may be envisaged when considering both the studied area because of its excellent sealing capability and its relatively
structural context and the geochemistry of the subsurface water. The inert nature with respect to H2 (Fig. 10; Sainz-Garcia et al., 2017; Tar­
unconformity between the Mauléon Basin and the fractured crystalline kowski, 2019; Zivar et al., 2020). Indeed, it is composed of gypsu­
basement is located at about 7 km depth. The mantle body in itself is m/anhydrite interbedded with limestone/dolomite whose porosity
well imaged at 8–10 km depth and is connected to the Mauléon Basin by ranges from 1 to 10%. This formation has already shown its potential to
the NPFT (Fig. 9b; Chevrot et al., 2015, 2018; García-Senz et al., 2020; form gas reservoirs (Pochitaloff, 1954).
Lacan, 2008; Lehujeur et al., 2021; Wang et al., 2016). The Saint Palais well, drilled in 1954, has revealed a very peculiar
Based on the regional geothermal gradient of the Aquitaine Basin of gas accumulation mainly composed of CO2 (89.4 vol%); N2 (9.8 vol%);
approximately 27 ◦ C.km− 1 (Bonté et al., 2010), the exhumed mantle and minor CH4 (0.2 vol%) at ~2800-m deep. Unfortunately, H2 con­
body is expected to be within the 200–270 ◦ C temperature range centration was not measured at that time, simply because H2 was the
(Lefeuvre et al., 2021). This temperature range is optimum for efficient carrier gas of the GC technique used for gas analysis. Two pumping tests
serpentinization processes triggered by the infiltrated meteoric water, were also carried out in that well and revealed a flow rate of 150 L min− 1
thereby producing magnetite and H2 (Malvoisin et al., 2012; Klein et al., and a pressure of 400 bar within the Triassic formations (Pochitaloff,
2020). The significant positive magnetic anomaly (>60 nT) located 1954).
above the mantle body might be well correlated with an active or past The temperature of the Triassic formation is 85 ◦ C at its top and 150
serpentinization zone of mantle rocks (García-Senz et al., 2020; Lefeuvre ◦
C at its base. Such a temperature is ideal for long term H2 trapping.
et al., 2021). According to this scenario, H2 produced by serpentiniza­ Indeed, H2 is relatively inert within the 100–200 ◦ C temperature win­
tion of mantle rocks could migrate along the NPFT fault system until it dow for both kinetic and biological reasons. At temperatures lower than
reaches the Triassic formations or leaks to the surface (Fig. 9). 100 ◦ C, H2 could be consumed by micro-organisms and above 200 ◦ C, it

Fig. 10. Schematic representation of H2 trapping conditions near Sauveterre-de-Béarn as a function of the local geological structures (Trap geometry modified from
Hudec, 2012). The thermal gradient, H2 solubility (Bazarkina et al., 2020) and reactivity are also shown as a function of depth. The location of the putative H2-filled
reservoirs and the local geometry of the traps are shown in light green within the 100–200 ◦ C temperature window. (For interpretation of the references to color in
this figure legend, the reader is referred to the Web version of this article.)

9
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

can react with oxidized species such as sulfate and carbonate (Fig. 10; Data availability
Truche et al., 2009, 2010; McCollom and Shock, 1997; Schwartz and
Friedrich, 2006). H2-consuming reactions may not totally deplete an https://doi.org/10.26022/IEDA/112346. Natural hydrogen migra­
incoming H2 flux as a balance may exist in-between the input flux, the tion along thrust faults in foothill basins: The North Pyrenean Frontal
reaction zone and H2 residence time as well as the leaks. Under certain Thrust case study (Original data) (Earth/Chem).
circumstances microbial reactions may not be as efficient as expected.
This may be due to competition between different consortia or to mi­ Acknowledgments, Samples, and Data
crobial inhibition due to pH values or water salinity for examples. A
recent study of Karolytė et al. (2022) showed that an introduction of This work was conducted in the framework of the Convergence
O2-rich meteoric water in an H2 accumulation zone may catalyze sudden project (https://convergent-margins.com/), funded by TOTAL S.E..
microbial blooms and by way of consequence consume H2. On the Laurent Truche acknowledges support from the Institut Universitaire de
contrary, Sherwood Lollar et al. (2006) suggest that natural H2 may be France. Dominique Duclerc, Isabelle Mitteau, Magali Pujol, Marc Buis­
preserved in a hypersaline environment where hydrogenotrophic son and Maxime Ducoux are warmly thanks for support during the field
metabolism are inhibited. investigation and analytical measurements. Supplementary data re­
Parts of H2 migrating along the fault planes may either reach the ported in this study are given in supporting information and all soil gas
surface or react with different lithologies during its ascent, thus modi­ data are stored in EarthChem repository: https://doi.org/10.26022/IE
fying the composition of the gas mixture. In this study, the soil gas DA/112346. The authors thank Associate Editor Pr Marcello Liotta,
accumulation in well #1 allows us to estimate the surface gas flow Isabelle Moretti and an anonymous reviewer for instructive comments
assuming only a vertical flux and no lateral drainage (no pumping, only during the review of this manuscript.
passive accumulation). Based on the above assumptions, the calculated
daily flow rate for H2 ranges from 0.07 to 0.15 m3 m− 2.d− 1 in the fault Appendix A. Supplementary data
zone. Such a flow rate is comparable to other recently estimated H2 flow
rates in soils from the Sao Francisco basin in Brazil (0.04–0.9 m3 m− 2 Supplementary data to this article can be found online at https://doi.
d− 1; Prinzhofer et al., 2019; Moretti et al., 2020) and the Semail org/10.1016/j.apgeochem.2022.105396.
ophiolite in Oman (0.073–0.147 m3 m− 2.d− 1; Zgonnik et al., 2019).
Thus, the western Pyrenean foothills stand for a good case study for References
natural H2 exploration. All the elements that make up an active H2
system are gathered: water infiltration at depth, thermal gradient suit­ Abrajano, T., Sturchio, N., Bohlke, J., Lyon, G., Poreda, R., Stevens, C., 1988.
able for H2 production, serpentinization of mantle rocks, upward fluid Methanehydrogen Gas seeps, zambales ophiolite, Philippines: deep or shallow
origin. Chem. Geol. 71 (1–3), 211–222.
migration and potential reservoirs. Following the above-mentioned Barré, G., Fillon, C., Ducoux, M., Mouthereau, F., Gaucher, E.C., Calassou, S., 2021. The
considerations, it might be wise to consider first a preliminary drilling North Pyrenean Frontal Thrust: structure, timing and late fluid circulation inferred
through the NPFT at few hundreds meter depth to ensure a steady gas from seismic and thermal-geochemical analyses of well data. BSGF - Earth Sciences
Bulletin 192, 52. https://doi.org/10.1051/bsgf/2021046.
flux along the fault and then to target the upper Triassic anticline at Baubron, J.-C., Rigo, A., Toutain, J.-P., 2002. Soil gas profiles as a tool to characterise
3000–4000 m deep below Sauveterre-de-Béarn or to the north near active tectonic areas: the Jaut Pass example (Pyrenees, France). Earth Planet Sci.
Berenx. This latter formation composed of interbedded rocks of dolo­ Lett. 13 https://doi.org/10.1016/S0012-821X(01)00596-9.
Bazarkina, E.F., Chou, I.-M., Goncharov, A.F., Akinfiev, N.N., 2020. The behavior of H2 in
mitic limestone and evaporite with a total thickness of about 200 m may aqueous fluids under high temperature and pressure. Elements 16 (1), 33–38.
constitute a promising trap for H2. https://doi.org/10.2138/gselements.16.1.33.
Berard, P., Mazurier, C., 2000. Ressources en eaux thermales et minérales des stations du
département des Pyrénées-Atlantiques. Station thermale de Salies-de-Béarn. Rapport
6. Conclusion
BRGM/RP 50176-FR 28, 3 annexes de 26 pages. https://infoterre.brgm.fr/rappo
rts/RP-50174-FR.pdf.
Based on soil gas analysis and electromagnetic surveys we have Biteau, J.J., Le Marrec, A., Le Vot, M., Masset, J.M., 2006. The aquitaine basin. Petrol.
Geosci. 12 (3), 247–273. https://doi.org/10.1144/1354-079305-674.
confirmed the existence of a fluid-draining fault system. Soil gas con­
Bonté, D., Guillou-Frottier, L., Garibaldi, C., Bourgine, B., Lopez, S., Bouchot, V.,
centration (H2, CO2, CH4 and 222Rn) increases when approaching the Lucazeau, F., 2010. Subsurface temperature maps in French sedimentary basins: new
North Pyrenean Frontal Thrust near Sauveterre-de-Bearn. The maximum data compilation and interpolation. Bull. Soc. Geol. Fr. 181 (4), 377–390. https://
H2, CO2 and 222Rn concentration values recorded in the fault zone at 1 m doi.org/10.2113/gssgfbull.181.4.377.
Boreham, C.J., Sohn, J.H., Cox, N., Williams, J., Hong, Z., Kendrick, M.A., 2021.
depth are 822 ppmv, 10.3 vol% and 57 kBq.m− 3, respectively; whereas Hydrogen and hydrocarbons associated with the neoarchean Frog’s Leg Gold camp,
their local background values are one to 2 orders of magnitude lower: 10 Yilgarn craton, western Australia. Chem. Geol. 575, 120098.
ppmv, 0.2 vol%, 0.3 kBq.m− 3, respectively. Gas accumulation in 1 m Byrne, D.J., Barry, P.H., Lawson, M., Ballentine, C.J., 2020. The use of noble gas isotopes
to constrain subsurface fluid flow and hydrocarbon migration in the East Texas
deep wells confirmed the existence of an active fluid circulation. In Basin. Geochem. Cosmochim. Acta 268, 186–208. https://doi.org/10.1016/j.
addition, the geochemical analysis carried out at Salies-de-Béarn ther­ gca.2019.10.001.
mal springs have revealed a deep fluid circulation along the NPFT. These Charlou, J., Donval, J., Fouquet, Y., Jean-Baptiste, P., Holm, N., 2002. Geochemistry of
high h2 and ch4 vent fluids issuing from ultramafic rocks at the rainbow
data support the idea of deep fluid migration in the studied area in hydrothermal field (36 14 n, mar). Chem. Geol. 191 (4), 345–359.
agreement with previous studies (Lefeuvre et al., 2021; Tremosa et al., Charlou, J.L., Donval, J.P., Konn, C., OndreAs, H., Fouquet, Y., Jean-Baptiste, P.,
2022). Finally, the compressive structure located at 2800–4000 m depth Fourre, E., 2010. High production and fluxes of H2 and CH4 and evidence of abiotic
hydrocarbon synthesis by serpentinization in ultramafic-hosted hydrothermal
composed of the interbedded Triassic formation represents a promising
systems on the mid-atlantic ridge. Diversity of hydrothermal systems on slow
trap for H2 and has already demonstrated the capacity to accumulate spreading ocean ridges 188, 265–296.
gas. Chavagnac, V., Ceuleneer, G., Monnin, C., Lansac, B., Hoareau, G., Boulart, C., 2013a.
Mineralogical assemblages forming at hyperalkaline warm springs hosted on
ultramafic rocks: a case study of Oman and ligurian ophiolites. G-cubed 14 (7),
Declaration of competing interest 2474–2495.
Chavagnac, V., Monnin, C., Ceuleneer, G., Boulart, C., Hoareau, G., 2013b.
The authors declare that they have no known competing financial Characterization of hyperalkaline fluids produced by low-temperature
serpentinization of mantle peridotites in the Oman and ligurian ophiolites. G-cubed
interests or personal relationships that could have appeared to influence 14 (7), 2496–2522.
the work reported in this paper. Chen, Q., Popa, M.E., Batenburg, A.M., Röckmann, T., 2015. Isotopic signatures of
production and uptake of H2 by soil. Atmos. Chem. Phys. 15 (22), 13003–13021.
https://doi.org/10.5194/acp-15-13003-2015.
Chevrot, S., Sylvander, M., Diaz, J., Ruiz, M., Paul, A., the PYROPE Working Group,
2015. The Pyrenean architecture as revealed by teleseismic P-to-S converted waves

10
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

recorded along two dense transects. Geophys. J. Int. 200 (2), 1094–1105. https:// Hirose, T., Kawagucci, S., Suzuki, K., 2011. Mechanoradical H2 generation during
doi.org/10.1093/gji/ggu400. simulated faulting: implications for an earthquake-driven subsurface biosphere.
Chevrot, Sébastien, Sylvander, M., Diaz, J., Martin, R., Mouthereau, F., Manatschal, G., Geophys. Res. Lett. 38 (17).
et al., 2018. The non-cylindrical crustal architecture of the Pyrenees. Sci. Rep. 8 (1), Hudec, M.R., 2012. Influence of Salt on Petroleum Systems Iii: Migration. AGL.
9591. https://doi.org/10.1038/s41598-018-27889-x. Jones, D.S., Northup, D.E., 2021. Cave decorating with microbes: Geomicrobiology of
Chevrot, S., Sylvander, M., Villaseñor, A., Díaz, J., Stehly, L., Boué, P., et al., 2022. caves. Elements 17 (2), 107–112. https://doi.org/10.2138/gselements.17.2.107.
Passive imaging of collisional orogens: a review of a decade of geophysical studies in Kameda, J., Saruwatari, K., Tanaka, H., 2003. H2 generation in wet grinding of granite
the Pyrénées Imagerie passive des orogènes collisionnels: une revue d’une décennie and single-crystal powders and implications for H2 concentration on active faults.
d’études géophysiques dans les Pyrénées. Bull. Soc. Geol. Fr. 193 (1). Geophys. Res. Lett. 30 (20).
Choukroune, P., 1989. The Ecors Pyrenean deep seismic profile reflection data and the Karolytė, R., Warr, O., van Heerden, E., Flude, S., de Lange, F., Webb, S., et al., 2022. The
overall structure of an orogenic belt. Tectonics 8 (1), 23–39. https://doi.org/ role of porosity in H2/He production ratios in fracture fluids from the Witwatersrand
10.1029/TC008i001p00023. Basin, South Africa. Chem. Geol. 595, 120788.
Ciotoli, G., Lombardi, S., Annunziatellis, A., 2007. Geostatistical analysis of soil gas data Kessler, A.J., Chen, Y.-J., Waite, D.W., Hutchinson, T., Koh, S., Popa, M.E., et al., 2019.
in a high seismic intermontane basin: Fucino Plain, central Italy. J. Geophys. Res. Bacterial fermentation and respiration processes are uncoupled in anoxic permeable
112 (B5), B05407 https://doi.org/10.1029/2005JB004044. sediments. Nature Microbiology 4 (6), 1014–1023. https://doi.org/10.1038/
Crespo-Medina, M., Twing, K.I., Sanchez-Murillo, R., Brazelton, W.J., McCollom, T.M., s41564-019-0391-z.
Schrenk, M.O., 2017. Methane dynamics in a tropical serpentinizing environment: Kita, I., Matsuo, S., Wakita, H., 1982. H 2 generation by reaction between H 2 O and
the santa Elena ophiolite, Costa Rica. Front. Microbiol. 8, 916. crushed rock: an experimental study on H 2 degassing from the active fault zone.
Dogan, T., Mori, T., Tsunomori, F., Notsu, K., 2007. Soil H2 and CO2 surveys at several J. Geophys. Res. Solid Earth 87 (B13), 10789–10795. https://doi.org/10.1029/
active faults in Japan. In: Terrestrial Fluids, Earthquakes and Volcanoes: the Hiroshi JB087iB13p10789.
Wakita. Springer, pp. 2449–2463. II. Klein, F., Tarnas, J.D., Bach, W., 2020. Abiotic sources of molecular hydrogen on earth.
Donzé, F.-V., Truche, L., Shekari Namin, P., Lefeuvre, N., Bazarkina, E.F., 2020. Elements 16 (1), 19–24. https://doi.org/10.2138/gselements.16.1.19.
Migration of natural hydrogen from deep-seated sources in the sao francisco basin, Kubas, G.J., 2007. Fundamentals of H2 binding and reactivity on transition metals
Brazil. Geosciences 10 (9), 346. underlying hydrogenase function and H2 production and storage. Chem. Rev. 107
Ducasse, L., Velasque, P.C., Muller, J., 1986. Glissement de couverture et panneaux (10), 4152–4205.
basculés dans la région des Arbailles (Pyrénées occidentales): Un modèle évolutif Lacan, P., 2008. Activité sismotectonique plio-quaternaire de l’ouest des Pyrénées. https:
crétacé de la marge nord-ibérique à l’Est de la transformante de Pamplona. Comptes //tel.archives-ouvertes.fr/tel-01783939.
rendus de l’Académie des sciences. Série 2, Mécanique, Physique, Chimie, Sciences Lagabrielle, Y., Labaume, P., de Saint Blanquat, M., 2010. Mantle exhumation, crustal
de l’univers, Sciences de la Terre 303 (16), 1477–1482. denudation, and gravity tectonics during Cretaceous rifting in the Pyrenean realm
Ducoux, M., 2017. Structure, thermicité et évolution géodynamique de la Zone Interne (SW Europe): insights from the geological setting of the lherzolite bodies: pyrenean
Métamorphique des Pyrénées. http://www.theses.fr/2017ORLE2026. lherzolites, gravity tectonics. Tectonics 29 (4). https://doi.org/10.1029/
Ducoux, M., Masini, E., Tugend, J., Gómez-Romeu, J., Calassou, S., 2021. Basement- 2009TC002588.
decoupled hyperextension rifting: the tectono-stratigraphic record of the salt-rich Larin, N., Zgonnik, V., Rodina, S., Deville, E., Prinzhofer, A., Larin, V.N., 2015. Natural
Pyrenean necking zone (Arzacq Basin, SW France). GSA Bulletin. https://doi.org/ molecular hydrogen seepage associated with surficial, rounded depressions on the
10.1130/B35974.1. European craton in Russia. Nat. Resour. Res. 24 (3), 369–383. https://doi.org/
Etiope, G., Baciu, C.L., Schoell, M., 2011a. Extreme methane deuterium, nitrogen and 10.1007/s11053-014-9257-5.
helium enrichment in natural gas from the homorod seep (Romania). Chem. Geol. Laurent, D., Gaucher, E.C., Durlet, C., Carpentier, C., Barré, G., Collon, P., et al., 2019,
280 (1–2), 89–96. January. A sulfuric acid Speleogenesis in the northern pyrenees? Example of the
Etiope, G., Schoell, M., Hosgormez, H., 2011b. Abiotic methane flux from the chimaera Arbailles karstic region (west pyrenees, France). In: Geophysical Research Abstracts,
seep and tekirova ophiolites (Turkey): understanding gas exhalation from low 21.
temperature serpentinization and implications for mars. Earth Planet Sci. Lett. 310 Lefeuvre, N., Truche, L., Donzé, F., Ducoux, M., Barré, G., Fakoury, R., et al., 2021.
(1–2), 96–104. Native H 2 exploration in the western pyrenean foothills. G-cubed 22 (8). https://
Etiope, G., Sherwood Lollar, B., 2013. Abiotic methane on earth. Rev. Geophys. 51 (2), doi.org/10.1029/2021GC009917.
276–299. Lehujeur, M., Chevrot, S., Villaseñor, A., Masini, E., Saspiturry, N., Lescoutre, R.,
Etiope, G., Samardžić, N., Grassa, F., Hrvatović, H., Miǒsić, N., Skopljak, F., 2017. Sylvander, M., 2021. Three-dimensional shear velocity structure of the Mauléon and
Methane and hydrogen in hyperalkaline groundwaters of the serpentinized dinaride Arzacq Basins (Western Pyrenees) Structure tridimensionnelle des vitesses de
ophiolite belt, Bosnia and Herzegovina. Appl. Geochem. 84, 286–296. cisaillement dans les bassins de Mauléon et Arzacq (Pyrénées occidentales). Bull.
Fabriès, J., Lorand, J.P., Bodinier, J.L., 1998. Petrogenetic evolution of orogenic Soc. Geol. Fr. 192 (1).
lherzolite massifs in the central and western Pyrenees. Tectonophysics 292 (1–2), Li, Y., Du, J., Wang, X., Zhou, X., Xie, C., Cui, Y., 2013. Spatial variations of soil gas
145–167. geochemistry in the tangshan area of northern China. Terr. Atmos. Ocean Sci. 24 (3).
Fabriès, J., Lorand, J.P., Bodinier, J.L., Dupuy, C., 1991. Evolution of the upper mantle Lyon, G.L., Hulston, J.R., 1984. Carbon and hydrogen isotopic compositions of New
beneath the Pyrenees: evidence from orogenic spinel lherzolite massifs. J. Petrol. (2), Zealand geothermal gases. Geochem. Cosmochim. Acta 48 (6), 1161–1171.
55–76. McCollom, T.M., 2013. Laboratory simulations of abiotic hydrocarbon formation in
Fang, Z., Liu, Y., Yang, D., Guo, L., Zhang, L., 2018. Real-time hydrogen mud logging Earth’s deep subsurface. Rev. Mineral. Geochem. 75 (1), 467–494.
during the wenchuan earthquake fault scientific drilling project (wfsd), holes 2 and 3 McCollom, T.M., Shock, E.L., 1997. Geochemical constraints on chemolithoautotrophic
in sw China. Geosci. J. 22 (3). metabolism by microorganisms in seafloor hydrothermal systems. Geochem.
Ford, M., Vergés, J., 2020. Evolution of a salt-rich transtensional rifted margin, eastern Cosmochim. Acta 61 (20), 4375–4391.
north Pyrenees, France. J. Geol. Soc. 178 (1), jgs2019–j2157. https://doi.org/ McLachlan, P., Blanchy, G., Binley, A., 2021. EMagPy: open-source standalone software
10.1144/jgs2019-157. for processing, forward modeling and inversion of electromagnetic induction data.
Freund, F., Dickinson, J.T., Cash, M., 2002. Hydrogen in rocks: an energy source for deep Comput. Geosci. 146, 104561 https://doi.org/10.1016/j.cageo.2020.104561.
microbial communities. Astrobiology 2 (1), 83–92. Malvoisin, B., Carlut, J., Brunet, F., 2012. Serpentinization of oceanic peridotites: 1. A
Gal, F., Joublin, F., Haas, H., Jean-prost, V., Ruffier, V., 2011. Soil gas (222Rn, CO2, 4He) high-sensitivity method to monitor magnetite production in hydrothermal
behaviour over a natural CO2 accumulation, Montmiral area (Drôme, France): experiments. J. Geophys. Res. Solid Earth 117 (B1). https://doi.org/10.1029/
geographical, geological and temporal relationships. J. Environ. Radioact. 102 (2), 2011JB008612.
107–118. https://doi.org/10.1016/j.jenvrad.2010.10.010. Masini, E., Manatschal, G., Tugend, J., Mohn, G., Flament, J.-M., 2014. The tectono-
García-Senz, J., Pedrera, A., Ayala, C., Ruiz-Constán, A., Robador, A., Rodríguez- sedimentary evolution of a hyper-extended rift basin: the example of the
Fernández, L.R., 2020. Inversion of the north Iberian hyperextended margin: the role Arzacq–Mauléon rift system (Western Pyrenees, SW France). Int. J. Earth Sci. 103
of exhumed mantle indentation during continental collision. Geological Society, (6), 1569–1596. https://doi.org/10.1007/s00531-014-1023-8.
London, Special Publications 490 (1), 177–198. https://doi.org/10.1144/SP490- Milkov, A.V., Etiope, G., 2018. Revised genetic diagrams for natural gases based on a
2019-112. global dataset of> 20,000 samples. Org. Geochem. 125, 109–120.
Géraud, Y., Diraison, M., Orellana, N., 2006. Fault zone geometry of a mature active Monnin, C., Coauthors, 2014. Fluid chemistry of the low temperature hyperalkaline
normal fault: a potential high permeability channel (pirgaki fault, corinth rift, hydrothermal system of prony bay (New Caledonia). Biogeosciences 11 (20),
Greece). Tectonophysics 426 (1–2), 61–76. 5687–5706.
Giardini, A., Subbarayudu, G.V., Melton, C.E., 1976. The emission of occluded gas from Monnin, C., Quéméneur, M., Price, R., Jeanpert, J., Maurizot, P., Boulart, C., Donval, J.-
rocks as a function of stress: its possible use as a tool for predicting earthquakes. p., Pelletier, B., 2021. The chemistry of hyperalkaline springs in serpentinizing
Geophys. Res. Lett. 3 (6), 355–358. environments: 1. The composition of free gases in New Caledonia compared to other
Gong, Z., Li, H., Lao, C., Tang, L., Luo, L., 2015. Real-time drilling mud gas monitoring springs worldwide. J. Geophys. Res.: Biogeosciences 126 (9).
records seismic damage zone from the 2008 mw 7.9 wenchuan earthquake. Moretti, I., Prinzhofer, A., Francolin, J., Pacheco, C., Rosanne, M., Rupin, F., Mertens, J.,
Tectonophysics 639, 109–117. 2020. Long-term monitoring of natural hydrogen superficial emissions in a brazilian
Grandjean, G., 1994. Etude des structures crustales dans une portion de chaîne et de leur cratonic environment. sporadic large pulses versus daily periodic emissions. Int. J.
relation avec les bassins sédimentaires. Application aux Pyrénées occidentales. Bull. Hydrogen Energy 46 (5), 3615–3628.
Cent. Rech. Explor.-Prod. Elf-Aquitaine 18 (2), 391–420. Muñoz, J.A., 1992. Evolution of a continental collision belt: ECORS-Pyrenees crustal
Gregory, S., Barnett, M., Field, L., Milodowski, A., 2019. Subsurface microbial hydrogen balanced cross-section. In: McClay, K.R. (Ed.), Thrust Tectonics. Springer
cycling: natural occurrence and implications for industry. Microorganisms 7 (2), 53. Netherlands, Dordrecht, pp. 235–246. https://doi.org/10.1007/978-94-011-3066-0_
https://doi.org/10.3390/microorganisms7020053. 21.

11
N. Lefeuvre et al. Applied Geochemistry 145 (2022) 105396

Neal, C., Stanger, G., 1983. Hydrogen generation from mantle source rocks in Oman. Telling, J., Coauthors, 2015. Rock comminution as a source of hydrogen for subglacial
Earth Planet Sci. Lett. 66, 315–320. ecosystems. Nat. Geosci. 8 (11), 851–855.
Nivin, V.A., 2019. Occurrence forms, composition, distribution, origin and potential Thauer, R.K., Klein, A.R., Hartmann, G.C., 1996. Reactions with molecular hydrogen in
hazard of natural hydrogen–hydrocarbon gases in Ore deposits of the Khibiny and microorganisms: evidence for a Purely organic hydrogenation catalyst. Chem. Rev.
Lovozero massifs: a review. Minerals 9 (9), 535. https://doi.org/10.3390/ 96 (7), 3031–3042. https://doi.org/10.1021/cr9500601.
min9090535. Tichadou, C., Godard, M., Munoz, M., Labaume, P., Vauchez, A., Gaucher, E.,
Novelli, P.C., Lang, P.M., Masarie, K.A., Hurst, D.F., Myers, R., Elkins, J.W., 1999. Calassou, S., 2021. Mineralogical and geochemical study of serpentinized peridotites
Molecular hydrogen in the troposphere: global distribution and budget. J. Geophys. from the north-western pyrenees: new insights on serpentinization along magma-
Res. Atmos. 104 (D23), 30427–30444. https://doi.org/10.1029/1999JD900788. poor continental passive margins. Lithos 406.
Oufi, O., Cannat, M., Horen, H., 2002. Magnetic properties of variably serpentinized Toft, P.B., Arkani-Hamed, J., Haggerty, S.E., 1990. The effects of serpentinization on
abyssal peridotites. J. Geophys. Res. Solid Earth 107 (B5). https://doi.org/10.1029/ density and magnetic susceptibility: a petrophysical model. Phys. Earth Planet. In.
2001JB000549. EPM-3. 65 (1–2), 137–157. https://doi.org/10.1016/0031-9201(90)90082-9.
Pedreira, D., Pulgar, J., Gallart, J., Torné, M., 2007. Three-dimensional gravity and Torre, F., Coauthors, 2020. Room temperature hydrocarbon generation in olivine
magnetic modeling of crustal indentation and wedging in the western Pyrenees- powders: effect of mechanical processing under CO2 atmosphere. Powder Technol.
Cantabrian Mountains. J. Geophys. Res. Solid Earth 112 (B12). https://doi.org/ 364, 915–923.
10.1029/2007JB005021. Tremosa J., Gal F., Kloppmann W., André L., Sültenfuß J., Calassou S., Gaucher E.C.,
Pedrera, A., García-Senz, J., Ayala, C., Ruiz-Constán, A., Rodríguez-Fernández, L.R., 2022. Abiotic Gases (CH4, H2) in Thermal Waters of the French Pyrenees Reveals On-
Robador, A., González Menéndez, L., 2017. Reconstruction of the exhumed mantle Going Serpentinization. submitted to EPSL in March 2022.
across The North Iberian margin by crustal-scale 3-D gravity inversion and Tronel, 1996. Forage Piezometrique 64 Sauveterre-de-Béarn. Forage (Frankfort On The
geological cross section: mantle along the Basque-cantabrian basin. Tectonics 36 Main): BSS002HNPZ BS BRGM data.
(12), 3155–3177. https://doi.org/10.1002/2017TC004716. Truche, L., Berger, G., Destrigneville, C., Pages, A., Guillaume, D., Giffaut, E., Jacquot, E.,
Pochitaloff, A., 1954. Log Fondamental: Sondage Saint Palais 1, Forage: BSS002JYFU 2009. Experimental reduction of aqueous sulphate by hydrogen under hydrothermal
BRGM Data. conditions: implication for the nuclear waste storage. Geochem. Cosmochim. Acta 73
Peters, J.W., Schut, G.J., Boyd, E.S., Mulder, D.W., Shepard, E.M., Broderick, J.B., et al., (16), 4824–4835. https://doi.org/10.1016/j.gca.2009.05.043.
2015. [FeFe]- and [NiFe]-hydrogenase diversity, mechanism, and maturation. Truche, L., Berger, G., Destrigneville, C., Guillaume, D., Giffaut, E., 2010. Kinetics of
Biochimica et Biophysica Acta (BBA) -. Molecular Cell Research 1853 (6), pyrite to pyrrhotite reduction by hydrogen in calcite buffered solutions between 90
1350–1369. https://doi.org/10.1016/j.bbamcr.2014.11.021. and 180◦ C: implications for nuclear waste disposal. Geochem. Cosmochim. Acta 74
Prinzhofer, A., Moretti, I., Françolin, J., Pacheco, C., D’Agostino, A., Werly, J., Rupin, F., (10), 2894–2914. https://doi.org/10.1016/j.gca.2010.02.027.
2019. Natural hydrogen continuous emission from sedimentary basins: the example Truche, L., McCollom, T.M., Martinez, I., 2020. Hydrogen and abiotic hydrocarbons:
of a Brazilian H2-emitting structure. Int. J. Hydrogen Energy 44 (12), 5676–5685. molecules that change the world. Elements. An International Magazine of
https://doi.org/10.1016/j.ijhydene.2019.01.119. Mineralogy, Geochemistry,and Petrology 16 (1), 13–18.
Romanak, K.D., Bennett, P.C., Yang, C., Hovorka, S.D., 2012. Process-based approach to Vacher, P., Souriau, A., 2001. A three-dimensional model of the Pyrenean deep structure
CO2 leakage detection by vadose zone gas monitoring at geologic CO2 storage sites: based on gravity modelling, seismic images and petrological constraints. Geophys. J.
process-based leakage detection. Geophys. Res. Lett. 39 (15) https://doi.org/ Int. 145 (2), 460–470. https://doi.org/10.1046/j.0956-540x.2001.01393.x.
10.1029/2012GL052426. Wakita, H., Nakamura, Y., Kita, I., Fujii, N., Notsu, K., 1980. Hydrogen release: new
Sainz-Garcia, A., Abarca, E., Rubi, V., Grandia, F., 2017. Assessment of feasible strategies indicator of fault activity. Science 210 (4466), 188–190.
for seasonal underground hydrogen storage in a saline aquifer. Int. J. Hydrogen Wang, Y., Chevrot, S., Monteiller, V., Komatitsch, D., Mouthereau, F., Manatschal, G.,
Energy 42 (26), 16657–16666. https://doi.org/10.1016/j.ijhydene.2017.05.076. et al., 2016. The deep roots of the western Pyrenees revealed by full waveform
Saspiturry, N., Razin, P., Baudin, T., Serrano, O., Issautier, B., Lasseur, E., et al., 2019. inversion of teleseismic P waves. Geology 44 (6), 475–478. https://doi.org/
Symmetry vs. asymmetry of a hyper-thinned rift: example of the Mauléon basin 10.1130/G37812.1.
(western pyrenees, France). Mar. Petrol. Geol. 104, 86–105. https://doi.org/ Ware, R.H., Roecken, C., Wyss, M., 1984. The detection and interpretation of hydrogen in
10.1016/j.marpetgeo.2019.03.031. fault gases. Pure Appl. Geophys. 122 (2), 392–402.
Saruwatari, K., Kameda, J., Tanaka, H., 2004. Generation of hydrogen ions and hydrogen Wibberley, C.A., Gonzalez-Dunia, J., Billon, O., 2017. Faults as barriers or channels to
gas in quartz–water crushing experiments: an example of chemical processes in production-related flow: insights from case studies. Petrol. Geosci. 23 (1), 134–147.
active faults. Phys. Chem. Miner. 31 (3), 176–182. Xiang, Y., Sun, X., Liu, D., Yan, L., Wang, B., Gao, X., 2020. Spatial distribution of Rn,
Sato, M., Mcgee, K.A., 1982. Continuous Monitoring of Hydrogen on the South. The 1980 CO2, Hg, and H2 concentrations in soil gas across a thrust fault in Xinjiang, China.
Eruptions of Mount St. Helens, p. 209. Washington, 1250. Front. Earth Sci. 8, 554924 https://doi.org/10.3389/feart.2020.554924.
Sato, M., Sutton, A.J., McGee, K.A., 1984. Anomalous hydrogen emissions from the San Zgonnik, V., Beaumont, V., Deville, E., Larin, N., Pillot, D., Farrell, K.M., 2015. Evidence
Andreas fault observed at the Cienega Winery, central California. Pure Appl. for natural molecular hydrogen seepage associated with carolina bays (surficial,
Geophys. 122 (2), 376–391. ovoid depressions on the atlantic coastal plain, province of the USA). Prog. Earth
Schwartz, E., Friedrich, B., 2006. The H2-metabolizing prokaryotes. The Prokaryotes 7, Planet. Sci. 2 (1), 1–15.
496–563. https://doi.org/10.1007/0-387-30742-7_17. Zgonnik, V., Beaumont, V., Larin, N., Pillot, D., Deville, E., 2019. Diffused flow of
Sherwood Lollar, B.S., Lacrampe-Couloume, G., Slater, G.F., Ward, J., Moser, D.P., molecular hydrogen through the Western Hajar mountains, Northern Oman. Arabian
Gihring, T.M., et al., 2006. Unravelling abiogenic and biogenic sources of methane in J. Geosci. 12 (3), 71. https://doi.org/10.1007/s12517-019-4242-2.
the Earth’s deep subsurface. Chem. Geol. 226 (3–4), 328–339. Zhou, X., Coauthors, 2010. Geochemistry of soil gas in the seismic fault zone produced by
Sugisaki, R., 1984. Relation between hydrogen emission and seismic activities. Pure the wenchuan m s 8.0 earthquake, southwestern China. Geochem. Trans. 11 (1),
Appl. Geophys. 122 (2), 175–184. 1–10.
Sylvander, M., Chevrot, S., Polychronopoulou, K., Martakis, N., Díaz, J., Ruiz, M., et al., Zivar, D., Kumar, S., Foroozesh, J., 2020. Underground hydrogen storage: a
2019, January. Spatio-temporal behavior of an extremely focused seismicity swarm comprehensive review. Int. J. Hydrogen Energy. https://doi.org/10.1016/j.
during the Maupasacq experiment (Pyrenean foreland, France). In: Geophysical ijhydene.2020.08.138. S0360319920331426.
Research Abstracts, 21. https://meetingorganizer.copernicus.org/EGU2019/EGU20 Zolnai, G., 1973. Coupe géologique interprétative n-s passant par: les arbaille - bassin de
19-8926.pdf. mauléon - structure (nappe) de ste surzanne - bassin d’arzacq. (compartiment de
Tarkowski, R., 2019. Underground hydrogen storage: characteristics and prospects. peyrehorade). S.N.P.A.
Renew. Sustain. Energy Rev. 105, 86–94. https://doi.org/10.1016/j.
rser.2019.01.051.

12