Geologic context: the measurement problem in H2 exploration

Summary

A single field reading of “140 ppm H2” is not by itself evidence of anything. Hydrogen is already present everywhere at ~0.53 ppm background, it diffuses faster than any other common gas, soil bacteria consume it as fast as it arrives, and weather (rainfall, barometric pressure, time of day) changes how much reaches a sensor. On top of that, the act of drilling a sampling hole can manufacture hydrogen from the rock itself, in amounts large enough to swamp a genuine deep signal. That 140 ppm could be a real seep, a rain-suppressed non-event, a drilling artifact, nitrogen-fixing clover, or an instrument quirk.

This is the field-practice companion to the technology-specific sensor pages. It does not re-derive any instrument’s internal physics (see the electrical, electrochemical, and optical sections) but lays out the transport theory, environmental confounds, and emerging protocol that decide whether any instrument’s reading can be trusted as geologic signal.

Context numbers used throughout

Atmospheric H₂ background ≈ 0.53 ppm (531 ± 6 ppb, 1991–1996 NOAA/CMDL global mean), drifting at +1.4 ± 0.7 ppb/yr with a ~50 ppb interhemispheric gradient [1][2]. Exploration anomalies span roughly 1 ppm to percent level; wellhead gas has reached ~98 vol% H₂ at Bourakebougou, Mali [3][4][5]. LEL = 4 vol% (40,000 ppm) [6].

How it works

Why H2 resists detection. H2 is a homonuclear diatomic with no permanent dipole, so its fundamental vibration is infrared-forbidden and only weak quadrupole transitions exist. NDIR analyzers, the workhorse of CH₄/CO₂ sensing, cannot detect H2 at all (Why H2 is hard), which forces geologic-H2 work onto electrochemical, catalytic, thermal-conductivity, metal-oxide, Pd-alloy, GC, and MS methods. H2’s small kinetic diameter (~2.9 Å) and low mass give it the highest diffusivity of any common gas, roughly 3–4× CH₄ and 4–5× CO₂ near STP [7]. Fast diffusion underlies nearly every interpretation problem here: H2 escapes a disturbed soil column faster than co-evolved gases, equilibrates with the atmosphere faster, and is more sensitive to barometric pumping.

Governing transport equation. Soil-gas H₂ concentration $C(z,t)$ follows a 1-D advection–diffusion–reaction equation:

$\frac{\partial C}{\partial t} = D_s\frac{\partial^2 C}{\partial z^2} - v\frac{\partial C}{\partial z} - k_s\,\Theta_a\,C + Q(z,t)$

Symbol	Meaning
$D_s$	effective soil-gas diffusivity (function of air-filled porosity $\Theta_a$ and tortuosity)
$v$	advective (Darcy) velocity, $v = -(k/\mu)\nabla P$
$k_s$	microbial first-order removal-rate constant
$Q(z,t)$	deep source term (upward flux from reservoir or fault)

Every term is independently modulated by weather: $D_s$ and $\Theta_a$ by moisture, $k_s$ by temperature and moisture, $v$ by barometric transients. This is why a single spot reading is under-determined: the same surface concentration can come from a strong source under wet, diffusion-limited soil or a weak source under dry, advection-prone soil during a barometric low.

Soil is a net sink. Microbial uptake dominates local production almost everywhere except active seeps, so soil–atmosphere exchange is parameterized by a deposition velocity, $F = v_d\,(C_{atm} - C_{soil,eq})$. Ehhalt & Rohrer’s two-layer expression,

$v_d = \frac{D_{S,I}}{\delta + \dfrac{D_{S,I}}{D_{S,II}}\cdot\dfrac{1}{\sqrt{k_s\Theta_a\,D_{S,II}}}\,\tanh\!\left(\sqrt{\dfrac{k_s\Theta_a}{D_{S,II}}}\,\ell\right)}$

reproduces measured $v_d \approx 0.01$–$0.1\ \mathrm{cm\,s^{-1}}$ across biomes [8], and the resulting global soil sink accounts for ~71% of the atmospheric H2 sink, roughly 60 Tg H₂/yr [2]. Newer work finds high-affinity H₂-oxidizing bacteria active down to soil water potentials of −70 to −100 MPa, far drier than the −3 to −10 MPa in older models, raising the calibrated global deposition velocity to ~0.09 mm/s for a 2010 baseline [9]. The exploration consequence: soil is an active biological sink whose baseline consumption must be understood before a surface enrichment can be attributed to upward flux.

Governing relations by transduction class (each derived on its own page):

Class	Governing relation / basis	Field role	Ref
Amperometric electrochemical	$i_L = nFAD\,C_\infty/l$, current linear in concentration [23]	Dominant soil-gas workhorse	Electrochemical cells
Catalytic bead (pellistor)	Heat of combustion in a Wheatstone bridge; needs ≥~12 vol% O₂	Percent/LEL safety, not ppm	Catalytic pellistors [10][11]
Thermal conductivity	H₂’s ~7× air thermal conductivity vs. a reference gas	Percent-level wellhead/reservoir gas	Thermal conductivity [12]
Metal-oxide (MOX)	Depletion-layer change at heated SnO₂ (300–450 °C)	Cheap, humidity-cross-sensitive screen	MOX sensors
Pd/Pd-alloy solid-state	Reversible PdH$_x$ formation, intrinsically H₂-specific	Process analyzers	Palladium-resistive, Work-function devices [13]
Neutron well-logging	High hydrogen index of free H₂ gas	Indirect downhole proxy (Bourakebougou)	[5]
Fiber-optic / Raman	Surface interrogator, no downhole electronics	R&D for quantitative H₂ logging	Fiber-optic sensors [14]
Isotope-ratio / noble-gas MS	δD-H₂, δ¹³C-CH₄, ³He/⁴He source fingerprinting	Genetic layer no concentration sensor provides	Mass spectrometry, Isotopes, Helium

Deployment geometries

Orthogonal to the physics, and arguably more decision-relevant for a campaign:

• Static/passive soil probes. A buried tube or perforated chamber at ~0.8–1 m, equilibrated before a spot or continuous reading: MONHyTOR (non-pumping, hydrophobic membrane at 950 mm) [19] and the EVA passive comparator [27].
• Pumped/active sampling. A pump draws soil gas to an analyzer at 300–550 mL/min (GA5000 ≈550, WHALI ≈300, PARHyS ≤500 mL/min) [21][27][22]; faster but with its own artifacts.
• Flux (accumulation) chambers. An inverted sealed chamber measuring headspace buildup to derive a flux (mol m⁻² d⁻¹), demonstrated at Chimaera and for Brazilian depression flux of ~0.03–0.04 m³ day⁻¹ m⁻² [26][28].
• Fixed continuous stations. Telemetered, battery- or solar-powered, logging sub-minute to hourly for weeks to a year: PARHyS (~1-year autonomy, hourly, 2 ppm sensitivity) [22], MONHyTOR (1 Hz capable, months) [19], WHALI/EVA (weeks) [27][24].
• Borehole headspace analysis. For water-logged terrain, a sealed headspace is equilibrated with a water or drilling-fluid sample, exploiting H₂’s low aqueous solubility to concentrate it into gas (Dissolved gas in groundwater).
• Downhole wireline logging. Standard petrophysical logs, with H₂ detected indirectly via its neutron-porosity anomaly rather than a dedicated sensor [5].

Performance envelope

As-published, field-relevant specifications for instruments actually named in the geologic-H₂ literature. Deeper per-technology tables live in the companion pages.

Instrument / sensor	Technology	Range	LOD / resolution	Response (T90)	Selectivity notes	Cost class	Ref
Geotech/QED GA5000 (Biogas 5000)	Electrochemical (H₂), IR (CH₄/CO₂), electrochemical O₂/H₂S/CO	H₂: 0–1000 ppmv (vendor); tested to 2000 ppm	±2.5% FS precision (independent field test)	30–60 s	Depressurization causes H₂ overestimation; CO and RH interference documented	~$5,000–8,000	[21][17]
Axiom Sensing WHALI (pumped autonomous logger)	Electrochemical H₂/H₂S + IR CO₂	H₂: 0–2000 ppm	±10 ppm precision	Designed for continuous field logging	Paired T, P sensors logged for covariate correction	Research/instrument-grade	[18][27][24]
iBrid MX6 (Industrial Scientific)	Electrochemical multi-gas handheld	H₂: 0–2000 ppm (as configured)	±10 ppm precision (as reported)	Seconds-scale	Manufacturer publishes a separate cross-sensitivity table for the CO channel	Low thousands USD	[18][27]
PARHyS (Engie)	Electrochemical + T/RH compensation, LoRa telemetry	Sensitivity-optimized, not full-scale stated	Sensitivity ≤2 ppm	T90 <30 s	Peristaltic pump ≤500 mL/min; ~1-year battery autonomy	Research-grade fleet (~100 units, Brazil)	[22]
MONHyTOR (Geo Numeric/ISTerre)	Electrochemical, non-pumping buried probe	0–1000 ppm	0.1 ppm display resolution; ±5% FS	Adjustable to 1 s	Buried at 950 mm to damp thermal transients; cross-sensitive to CO only	Research-grade	[19]
Variotec 460 Tracergas	Dual sensor: TCD (0–100 vol%) + MOX semiconductor (ppm)	TCD: 0–100 vol%; semiconductor: ppm-level, claimed 0.1 ppm resolution	Manufacturer claims low cross-sensitivity to moisture/methane on the semiconductor channel	(unverified)	Dual-range design spans background ppm to wellhead percent-level	~Several thousand USD	[25]
SRI Natural Gas Analyzer GC / 310 Soil-Gas GC	Gas chromatograph, TCD (+optional FID)	Percent-level down to TCD floor	TCD-only detection limit ~200–300 ppm	Minutes per run (batch, not continuous)	High selectivity via chromatographic separation; tolerant of water vapor (H2 detection capability for these specific models not shown on cited pages, unverified)	~$5,000–20,000+	[12]
H2Scan HY-OPTIMA 2710	Solid-state Pd–Ni thin film	0.1%–10% H₂ (1,000–100,000 ppm)	High accuracy claimed at the low end of its (percent-level) range	(unverified)	Marketed as non-combustion, non-electrolyte, not cross-sensitive to other gases	Several thousand USD	[13]
Alphasense H2-AF	Electrochemical amperometric	0–2,000 ppm (warranted)	Noise-equivalent resolution <0.7 ppm RMS	<45 s (0→400 ppm)	Very low CO cross-sensitivity (“most selective H₂ sensor on the market” per vendor)	Sensor element, tens–low hundreds USD	[29]
Figaro TGS 821	MOX (SnO₂) chemiresistor	Detects from ~50 ppm upward	Coarse relative to electrochemical	Fast (seconds) but slow humidity-driven baseline settling	High sensitivity but poor selectivity; humidity cross-sensitive	Sensor element, single-digit–tens USD	[10]
Catalytic bead / pellistor (generic)	Calorimetric combustion, Wheatstone bridge	Percent/LEL range (not ppm)	Not applicable below roughly hundreds–thousands of ppm	Seconds	Poisoned by silicones/lead/sulfur; reversibly inhibited by H₂S; needs ≥~12% O₂	Sensor element tens USD; instrument hundreds–low thousands	[10][11][30]
Neutron porosity wireline log	Nuclear (indirect)	Qualitative, flags anomalous zones, not ppm	N/A (corroborating signature, not primary readout)	Standard logging-run timescale	No gas cross-sensitivity, but requires careful petrophysical interpretation	Wireline logging-run cost (amortized across a full log suite)	[5]

Rolled up: across instruments deployed in published geo-H₂ surveys, realistic field LOD sits in the 1–10 ppm range, rarely below ~0.7 ppm even for the best electrochemical cells under lab conditions [29], only marginally below or comparable to the 0.53 ppm background [1]. Most fielded instruments cannot cleanly resolve background from a sub-ppm anomaly, so single-digit-ppm anomalies require independent corroboration. Full-scale ranges span three orders of magnitude (1,000 ppm handhelds to 100 vol% TCD channels), and no single fielded instrument covers that span with ppm resolution at the low end.

Cross-sensitivity & environment

Environmental covariates drive field interpretation, and this is where the 2023–2025 literature has moved fastest.

Covariate	Effect and field magnitude	Ref
Temperature	Modulates diffusivity, microbial kinetics, and sensor sensitivity, but correlates only weakly with measured flux ($r^2 = 0.13$). An observed diurnal midday peak was correlative, not causal: soil temperature at ~1 m stays stable while air temperature oscillates, so a diurnal cycle is not by itself confirmation of a temperature-driven process	[18][27]
Soil moisture / rainfall	Strongest covariate, acting three ways: sets diffusivity $D_s$ and air-filled porosity, modulates the microbial sink, and at high saturation caps the gas physically. A WA seep fell from 493 ppm (dry, Feb 2022) to 52 ppm (post-rain, Feb 2023) at the same site as rising water sealed the pore network; rainfall suppression confirmed more broadly. Campaigns during or shortly after rain systematically under-detect	[27][18]
Barometric pressure	Two confounded mechanisms: geologically, H₂’s low density makes it responsive to pressure-driven advective pumping (daily barometric dips coincide with H₂ peaks); instrumentally, a chamber test found depressurization below atmospheric makes the GA5000 overestimate H₂, while above-ambient pressure has minor effect. Hard to separate from a single spot reading	[27][17]
Relative humidity	For electrochemical cells, shifts electrolyte hydration and the zero point; for MOX, adsorbed water competes for SnO₂ surface sites. A probe pulled from cool humid soil into warm dry air can produce a transient reading unrelated to H₂ (the “humidity-shock” artifact)
Microbial sink	High-affinity H₂-oxidizing bacteria, ~71% of the global sink (~60 Tg/yr), active to −70 to −100 MPa water potential, so it rarely fully switches off. Baseline consumption must be characterized before attributing a surface enrichment to deep flux	[2][9]
Biological production	Nitrogenase reduces some protons to H₂ as an obligate side reaction of N₂ fixation, so legume nodules and free-living diazotrophs vent H₂ into soil. Surveys near legume cover crops or N-fixing mats carry a non-geologic positive

Cross-gas interferents. CO is a documented electrochemical H₂-cell interferent; low-CO-cross-sensitivity cells (Alphasense H2-AF) are marketed on this basis [29]. CH₄ does not directly interfere with amperometric H₂ chemistry but is co-measured (IR/TCD) and is itself a source diagnostic [21][26]. H₂S is measured as a co-analyte and is a reversible catalytic-bead inhibitor, not usually an H₂-channel cross-interferent [27][21][11]. CO₂ (IR-active, unlike H₂) is co-measured and high values flag possible microbial/fermentation sources [21][12]. O₂ is required at ≥~12 vol% for pellistors and is the amperometric counter-electrode reactant, so severe depletion degrades electrochemical response [10][11]. Catalytic beads are permanently poisoned by silicones, lead, and sulfur/phosphate species and reversibly inhibited by H₂S and halogenated compounds [11][30], which matters where a wellhead safety monitor sits near the H₂S the well also produces.

The 2023–2025 measurement-artifact controversy

The central fight is whether ppm-to-hundreds-of-ppm soil-gas spikes recorded during standard 1 m spot surveys are real upward flux or generated by the measurement itself. The strongest case for artifacts: a systematic study across 24 sub-circular depressions in the Yilgarn Craton and Perth Basin drilled 83 holes and made repeated measurements without redrilling, finding drilling-induced hydrogen up to 464 ppm at 85% of locations, concentrated in high-strength, low-moisture soils, with only ~15% attributable to natural hydrogen [16][20]. A companion paper attributes this to mechanochemical reactions at the drill bit (“drill-bit metamorphism,” DBM), strongly lithology-dependent: shallow drilling with water in ultramafic and carbonate rocks generates far more H₂ because ultramafics supply Fe²⁺ and reactive silicates at freshly fractured surfaces [31]. A separate study found co-located H₂–CO₂–CH₄–H₂S near wet soil more consistent with near-surface microbial fermentation than deep degassing [12], and the GA5000 chamber test documented genuine depressurization, temperature, and CO sensitivity [17].

On the genuine-signal side: MONHyTOR observed installation peaks decaying exponentially with decay parameters that track soil type (clay: higher initial, slower decay than sand), consistent with diffusion physics rather than a pure instrument artifact [19]. Telemetered monitoring (PARHyS, WHALI/EVA) recorded diurnal and multi-day pulses persisting over months, beyond any single drilling or pumping event [5][18][27][19][28][22].

Synthesis: where both sides agree

1. Single spot measurements taken immediately after drilling, without a settling period or repeat protocol, are unreliable and can be dominated by DBM artifacts, especially in ultramafic/iron-rich or high-strength, low-moisture soils [15][16][26].
2. Pumped extraction and pressure/humidity transients bias electrochemical cells at ppm levels, now quantified for the GA5000 [17].
3. Genuine natural seepage exists, corroborated by recurring diurnal/seasonal cycles over months, lithology-dependent decay kinetics, lab-GC-confirmed wellhead compositions, and downhole neutron-log anomalies no soil-probe artifact can explain [5][18][27][19][28][22].
4. Emerging best practice is convergent: percussion (not rotary) drilling, repeated measurements without redrilling, multi-day-to-multi-week continuous monitoring rather than spot readings, and multi-gas/soil-property context alongside every H₂ reading [15][16][26].

Use for geologic H₂

Field summary

No single soil-gas reading should be treated as confirmed geologic signal on its own. Soil-gas surveys are the workhorse and are fit for purpose only when paired with percussion drilling, a settling/repeat protocol, and covariate logging; used as bare spot readings, they have reported drilling artifacts as “natural seepage” at the majority of sampled locations in at least one systematic study [16]. Continuous monitoring stations are the strongest tool for separating signal from artifact, but the field runs almost entirely on bespoke research instrumentation, a genuine procurement gap. Electronics-free downhole logging is achievable today only via the indirect, qualitative neutron-porosity proxy; a direct, quantitative, electronics-free downhole H₂ measurement does not yet exist as a field-proven product.

Soil-gas surveys (the workhorse). Electrochemical handhelds (GA5000, iBrid MX6) and research probes (MONHyTOR, PARHyS) are fit for purpose only with a repeat-measurement/settling protocol and covariate logging; percussion drilling is preferred over rotary to minimize mechanochemical H₂ [15][26]. Minimum-viable protocol: percussion-drill to ~1 m, let the hole settle, take a first reading, repeat after a defined interval to distinguish exponential decay (artifact) from a stable or rising reading (seepage); log soil moisture and temperature at depth; and where possible schedule after a dry period rather than post-rain [27].

Continuous monitoring stations. The strongest tool for distinguishing signal from artifact, because a multi-week-to-multi-month series lets covariates be regressed out and genuine multi-day pulses (which a drilling artifact cannot reproduce weeks after installation) be identified [18][27][28][22]. The main constraint is power and telemetry in remote terrain, solved with battery + LoRa (PARHyS, ~1-year autonomy) [22] or solar + local logging (MONHyTOR, months) [19]; both are mature enough to field but are bespoke research builds, not shelf products, with procurement and support-lifecycle implications.

Downhole/borehole logging. The “no downhole electronics” preference is most binding here and the literature is thinnest. The one field-proven electronics-free approach is retrospective use of standard petrophysical wireline logs, especially neutron porosity, as an indirect corroborating signature, as at Bourakebougou [5]; it works with existing infrastructure and high-T/P ratings but is qualitative and needs careful petrophysical interpretation. Direct downhole composition today comes from wireline gas samplers bringing a physical sample to surface for GC, not any downhole electronic sensor. Optical/Raman fiber approaches that keep electronics at surface are the most promising long-term path, demonstrated to ~450 °C in fiber-coating studies but not in an actual well, and not yet for quantitative natural-H₂ concentration [14]; treat as a 3–5 year R&D bet (Fiber-optic sensors, Raman spectroscopy).

Limitations

1. No NDIR option exists for H₂ itself: any multi-gas instrument’s H₂ channel is electrochemical, MOX, TCD, or catalytic, while its CH₄/CO₂ channels can be true NDIR.
2. The best fielded electrochemical noise floor (~0.7–several ppm) sits close to the 0.53 ppm background [1][29], so “detecting an anomaly relative to background” is far harder than “detecting hundreds of ppm at a seep.” Calibrate expectations for percent-level or high-tens-to-hundreds-of-ppm anomalies against a noisy few-ppm floor, not background-resolution work.
3. Percent-level wellhead gas (up to 98 vol% at Bourakebougou) is a different, much easier regime (TCD/GC, catalytic LEL monitors) than ppm soil-gas prospecting [4][5]; do not conflate the two.

Vendors & products

Vendor	Product line	Technology	Key specs	Price class	Ref
Geotech / QED Environmental Systems	GA5000 (Biogas 5000) portable analyzer	Electrochemical H₂/H₂S/O₂/CO + IR CH₄/CO₂	H₂ 0–1000 ppmv, IP65, ATEX Zone 1/IECEx/CSA/MCERTS, −10…+50 °C, ~8 hr battery, 1.6 kg	~$5,000–8,000	[21]
Industrial Scientific	iBrid MX6	Electrochemical multi-gas handheld	H₂ 0–2000 ppm, 10 ppm precision (as configured in geo-H₂ studies)	Low thousands USD	[18][27]
Axiom Sensing (Australia)	WHALI (pumped autonomous logger), EVA (passive logger)	Electrochemical H₂/H₂S + IR CO₂, T/P sensors	H₂ 0–2000 ppm, ±10 ppm; ~300 mL/min pump; weeks-long autonomous deployment	Research/instrument-grade custom build; commercial soil-gas survey service offered	[27][24]
Engie (R&D/GRTgaz)	PARHyS	Electrochemical + T/RH compensation, LoRa telemetry	Sensitivity ≤2 ppm, T90 <30 s, ≤500 mL/min pump, ~1-year battery autonomy, ~100 units deployed in Brazil	Research fleet deployment, not commercial SKU	[22]
Geo Numeric / ISTerre (France)	MONHyTOR	Electrochemical, non-pumping buried probe	0–1000 ppm, 0.1 ppm resolution, ±5% FS, 1 Hz capable, months of autonomy, buried at 950 mm	Research-grade, deployed as a “fleet” in France since 2024	[19]
Hermann Sewerin GmbH (Germany)	VARIOTEC 460 Tracergas	TCD (0–100 vol%) + MOX semiconductor (ppm)	Claimed 0.1 ppm resolution on semiconductor channel; 4×AA batteries, ≥8 hr operation	Several thousand USD	[25]
Alphasense (UK)	H2-AF, H2-BF	Electrochemical amperometric	H2-AF: 0–2000 ppm, <0.7 ppm RMS noise, <45 s T90, −30…+50 °C; H2-BF: up to 5000 ppm	Sensor element, tens–low hundreds USD	[29]
Figaro Engineering (Japan)	TGS 821	MOX (SnO₂) chemiresistor	Detects from ~50 ppm; fast but humidity-sensitive baseline	Sensor element, single-digit–tens USD	[10]
H2Scan (USA)	HY-OPTIMA 2710 (and 2700/5000/700B series)	Solid-state Pd–Ni thin film (ex-Sandia National Labs technology)	0.1%–10% H₂ range, non-consumable sensor, low cross-sensitivity by mechanism	Several thousand USD (process analyzer class)	[13]
SRI Instruments (USA)	Natural Gas Analyzer GC; 310 Soil-Gas GC System; 8610C	Portable GC, TCD (+FID option)	TCD-only LOD ~200–300 ppm; tolerant of water vapor; field-portable, battery-capable; H2 detection capability for these named models not confirmed by cited source (unverified)	~$5,000–20,000+	[12]
HyTerra Ltd. (Australia/USA)	Nemaha Project (Kansas), Geneva Project (Nebraska)	Exploration operator (wellhead GC/wireline, not a sensor vendor)	Well results up to 96.1% H₂ (Sue Duroche-3), 83% H₂ (McCoy 1), Nebraska up to 44% H₂/13% He; Schlumberger wireline logging post-drilling; Kansas’s hydrogen-rich wells were first flagged in a 1986 USGS soil-gas study	Operator, not instrument vendor	[15][32]
Gold Hydrogen Ltd. / CSIRO (Australia)	Ramsay Project soil-gas survey	CSIRO-operated soil-gas survey (instrument model not stated in available sources)	Certified unrisked prospective resource of 1.3 billion kg H₂ at Ramsay; multi-location soil-H₂ detections reported as “moderate”	Operator/survey program	[14][33]
La Française de l’Énergie / CNRS-GeoRessources (France)	Folschviller well, Lorraine basin (Regalor project)	Wellhead GC and downhole measurement program (University of Lorraine/CNRS)	15% H₂ measured at 1,093 m depth; ~98% H₂ estimated (not directly measured) at ~3,000 m; potential ~46 Mt H₂ resource	Operator/academic partnership	[34]
Engie Research & Innovation (Brazil)	Long-term Brazil monitoring (São Francisco Basin, Campinas/Baru structures)	~100 PARHyS units, hourly sampling at 80 cm	Diurnal H₂ cycles resolved over 6–8 months of continuous monitoring	Research program	[28][22]

Open questions

The most active groups (the Edith Cowan/Western Australia cluster on drilling artifacts and seep characterization [15][16][18][27], the ISTerre/Grenoble instrumentation group [17][19], the Moretti/Prinzhofer network on flux modeling and Bourakebougou [4][5][20][28][35], and the USGS Energy Resources Program [36][37]) converge on closing the artifact-vs-signal debate with longer, better-instrumented time series and independent geophysical corroboration. Specific items for Hestia:

1. Get the exact GA5000 pressure-correction curve. The chamber test shows depressurization overestimation but the ppm-magnitude correction polynomial is not confirmed here [17]; pull the full paper before it enters an SOP.
2. Confirm the full author list and quantitative tables for the DBM mechanism paper [31] before external use; only the mechanism and lithology dependence are confirmed here.
3. Build or procure a repeat-measurement soil-gas SOP: percussion drilling, settle time, repeat reading, moisture/temperature/pressure logging, and explicit flagging of high-strength/low-moisture/ultramafic sites as high artifact risk.
4. Evaluate build-vs-license for a MONHyTOR- or PARHyS-class continuous probe, since both are research instruments today [19][22]; needs an academic collaboration or in-house build (Electrochemical cells).
5. Treat fiber-optic/Raman downhole H₂ sensing as a 3–5 year R&D bet; current work shows fiber survivability but not validated quantitative retrieval [14].
6. Quantify the nitrogen-fixation confound for candidate sites: any area with legumes, pasture, or N-fixing mats needs a pre-survey land-use screen; this has not been systematically subtracted in any published survey.
7. Track the USGS prospectivity release [37] and cross-reference HyTerra’s Nemaha/Geneva results [15] as a validation case.
8. Clarify the widely quoted ~23 ± 8 Tg H₂/yr flux figure: recent commentary cautions it reflects estimated subsurface production, not measured atmospheric emission [3].
9. Add a dedicated page on wellhead/downhole physical gas sampling tools (bomb samplers, U-tube samplers, mud-gas canisters) that bring a percent-level sample to surface for GC/IRMS.

Related pages

• Why H2 is hard: the selection-rule and transduction framing behind every technology here
• Electrochemical cells: the default soil-gas workhorse
• MOX sensors: the cheap, dense screening layer with a large covariate burden
• Palladium-resistive sensors and Work-function devices: intrinsically H₂-specific solid-state alternatives, mostly unproven at ppm background levels
• Thermal conductivity & acoustic sensors: the percent-level wellhead/reservoir regime
• Catalytic pellistors: %LEL safety monitoring, not ppm exploration signal
• Fiber-optic sensors and Raman spectroscopy: leading candidates for electronics-free downhole H₂ profiling, still R&D
• Enhanced Raman: approaches that could push Raman toward ppm field sensitivity
• IR absorption and proxies: why direct IR fails for H₂ and what proxy-gas strategies substitute
• UV/EUV and fluorescence: an alternative optical route for source-gas discrimination
• Plasmonic metamaterials: an emerging H₂-specific optical route
• Mass spectrometry: the isotope-ratio and noble-gas instrumentation behind source attribution
• Isotopes & fingerprinting: the full δD/δ¹³C/R-Ra toolkit
• Helium & co-tracers: the companion noble-gas tracer for pressure and source diagnosis
• Dissolved gas in groundwater: the aqueous-phase analog for water-logged terrain

References

Full reference list (37 entries)

[1] Novelli, P.C., Lang, P.M., Masarie, K.A., Hurst, D.F., Myers, R., Elkins, J.W. (1999). “Molecular hydrogen in the troposphere: Global distribution and budget.” Journal of Geophysical Research 104(D23), 30427–30444. DOI: 10.1029/1999JD900788.

[2] Paulot, F., Pétron, G., Crotwell, A.M., Bertagni, M.B. (2024). “Reanalysis of NOAA H2 observations: implications for the H2 budget.” Atmospheric Chemistry and Physics 24, 4217–4229. DOI: 10.5194/acp-24-4217-2024. https://acp.copernicus.org/articles/24/4217/2024/

[3] Zgonnik, V. (2020). “The occurrence and geoscience of natural hydrogen: A comprehensive review.” Earth-Science Reviews 203, 103140. https://www.sciencedirect.com/science/article/abs/pii/S0012825219304787

[4] Prinzhofer, A. et al. (2018). “Discovery of a large accumulation of natural hydrogen in Bourakebougou (Mali).” International Journal of Hydrogen Energy. https://www.sciencedirect.com/science/article/abs/pii/S0360319918327861

[5] Maiga, O., Deville, E., Laval, J., Prinzhofer, A., Diallo, A.B. (2023). “Characterization of the spontaneously recharging natural hydrogen reservoirs of Bourakebougou in Mali.” Scientific Reports. DOI: 10.1038/s41598-023-38977-y. https://pmc.ncbi.nlm.nih.gov/articles/PMC10363119/

[6] Industrial hydrogen safety references on flammability: hseblog, “What’s LEL? Lower Explosive Limit Guide for Safety Pros.” https://www.hseblog.com/lel/ ; Industrial Monitor Direct, “Hydrogen Gas LEL Limits & Exposure Safety.” https://industrialmonitordirect.com/blogs/knowledgebase/hydrogen-gas-lel-limits-and-exposure-safety-for-industrial-facilities

[7] Massman, W.J. (1998). “A review of the molecular diffusivities of H2O, CO2, CH4, CO, O3, SO2, NH3, N2O, NO, and NO2 in air, O2 and N2 near STP.” Atmospheric Environment. https://www.sciencedirect.com/science/article/abs/pii/S1352231097003919

[8] Ehhalt, D.H., Rohrer, F. (2013). “Deposition velocity of H2: a new algorithm for its dependence on soil moisture and temperature.” Tellus B: Chemical and Physical Meteorology 65(1). DOI: 10.3402/tellusb.v65i0.19904. https://b.tellusjournals.se/articles/10.3402/tellusb.v65i0.19904

[9] Reji, L., Bertagni, M.B., Paulot, F., Qin, Q., Zhang, X. (2025). “Global implications of a low soil moisture threshold for microbial hydrogen uptake.” Nature Communications. DOI: 10.1038/s41467-025-67208-3. https://pmc.ncbi.nlm.nih.gov/articles/PMC12804893/

[10] Figaro Engineering. “TGS 821 — Special Sensor for Hydrogen Gas” datasheet. https://www.figarosensor.com/product/docs/TGS821.pdf

[11] Winsen Sensor. “Catalytic Bead Hydrogen Sensors: Principles, Features, and Applications.” https://www.winsen-sensor.com/knowledge/catalytic-bead-hydrogen-sensors.html; Blackline Safety, “Pellistor (Catalytic Bead) Sensor Technical Notes.” https://support.blacklinesafety.com/articles/pellistor-catalytic-bead-sensors (link dead — redirects to generic help.blacklinesafety.com homepage)

[12] SRI Instruments. “Natural Gas Analyzer GC” and “310 Soil Gas GC System” product pages (the vendor pages do not explicitly list H₂ as a target analyte; H₂-detection capability for these specific models is unverified). https://srigc.com/product/gas-chromatographs/natural-gas-analyzer-gc; https://srigc.com/product/gas-chromatographs/310-soil-gas-gc-system; and “Surprising concentrations of hydrogen and non-geological methane and carbon dioxide in the soil,” Fuel/ScienceDirect. https://www.sciencedirect.com/science/article/pii/S004896972405040X

[13] H2Scan Corp. “H2scan is proud to introduce the HY-OPTIMA 2710, a new hydrogen analyzer specifically designed to provide high accuracy measurements at low hydrogen concentrations.” https://h2scan.com/h2scan-is-proud-to-introduce-the-hy-optima-2710-a-new-hydrogen-analyzer-specifically-designed-to-provide-high-accuracy-measurements-at-low-hydrogen-concentrations/

[14] “Comparative Experimental Study of a High-Temperature Raman-Based Distributed Optical Fiber Sensor with Different Special Fibers.” PMC. https://pmc.ncbi.nlm.nih.gov/articles/PMC6387110/ ; pv magazine Australia (2023), “‘Active natural hydrogen field’ confirmed in SA’s Yorke Peninsula.” https://www.pv-magazine-australia.com/2023/11/01/active-natural-hydrogen-field-confirmed-in-sas-yorke-peninsula/

[15] HyTerra Ltd. “Projects” (Nemaha Project, Kansas; Geneva Project, Nebraska — well composition results; this landing page confirms Nemaha “up to 96% H2, 5% He” and Geneva “44% H2, 13% He,” but does not itself display the more granular Sue Duroche-3 96.1% H2 or McCoy 1 83% H2 figures cited alongside this reference in the Vendors table — those are independently verifiable via HyTerra’s ASX announcements). https://hyterra.com/projects/

[16] Davies, K., Josse, R., Frery, E., Esteban, L., Keshavarz, A., Iglauer, S. (2025). “Distinguishing drilling-induced artifacts from naturally occurring hydrogen in soil gas surveys: Insights from sub-circular depressions.” International Journal of Hydrogen Energy 109, 1230–1240. https://doi.org/10.1016/j.ijhydene.2025.02.094

[17] Auclair, S., Lefeuvre, N., Dutoit, H., Donzé, F-V., Truche, L. (2024). “Assessing the Portable Gas Analyzer GA5000 for Accurate H₂ Monitoring in Geological Exploration.” Geonum Ed. https://hal.science/hal-04861335; https://www.researchgate.net/publication/385938959

[18] Davies, K., Haines, P.W., Thomas, C., Normore, L. (2025). “Natural hydrogen soil gas emissions near Harvey, Perth Basin: a comparative study of survey methods.” Science and Technology for Energy Transition 80, 46. DOI: 10.2516/stet/2025026. https://www.stet-review.org/articles/stet/full_html/2025/01/stet20250031/stet20250031.html

[19] Adjie, N., Bordes, C., Brito, D., Nasri, D., Normandin, É. (2025). “In-soil hydrogen concentration monitoring using MONHyTOR: application for natural hydrogen exploration.” Advances in Geosciences 67, 57–67. DOI: 10.5194/adgeo-67-57-2025. https://adgeo.copernicus.org/articles/67/57/2025/

[20] Patino, C., Piedrahita, D., Colorado, E., Aristizabal, K., Moretti, I. (2024). “Natural H2 Transfer in Soil: Insights from Soil Gas Measurements at Varying Depths.” Geosciences 14(11), 296. https://www.mdpi.com/2076-3263/14/11/296

[21] Geotech / QED Environmental Systems. “GA5000 Portable Gas Analyzer” product specification page. https://www.qedenv.com/products/portable-gas-analyzer-ga5000/ ; Van Walt / ERE Inc. distributor specification pages. https://www.vanwalt.com/equipment/ga5000-portable-landfill-gas-analyser/ ; https://www.ereinc.com/products/ga5000

[22] ENGIE Innovation. “PARHyS” (Permanent Analyses of Renewable Hydrogen with Sensors). https://innovation.engie.com/en/innovation-trophies-2018/parhys/9721; https://www.hnatsummit.com/parhys-system-a-new-approach-to-h2-concentration-measurements-in-the-subsurface/ (link dead — domain does not resolve)

[23] Winsen Sensor. “Electrochemical Hydrogen Sensors: Principles, Applications, and Advantages.” https://www.winsen-sensor.com/knowledge/electrochemical-hydrogen-sensors.html

[24] Axiom Sensing. Home / Soil Gas Survey Services (WHALI system description). https://www.axiomsensing.com/ ; https://www.axiomsensing.com/general-8

[25] Hermann Sewerin GmbH. “VARIOTEC 460 Tracergas” product page and technical documentation. https://www.sewerin.com/en-us/products/gas/gas-combination-devices/variotec-460-tracergas

[26] Lévy, D. et al. (2023). “Natural H2 exploration: tools and workflows to characterize a play.” Science and Technology for Energy Transition 78, 27. DOI: 10.2516/stet/2023021. https://www.stet-review.org/articles/stet/full_html/2023/01/stet20230116/stet20230116.html

[27] Davies, K. et al. (2024). “A natural hydrogen seep in Western Australia: Observed characteristics and controls.” Science and Technology for Energy Transition. DOI: 10.2516/stet/2024043. https://www.stet-review.org/articles/stet/full_html/2024/01/stet20240139/stet20240139.html

[28] Moretti, I. et al. (2021). “Long-term monitoring of natural hydrogen superficial emissions in a Brazilian cratonic environment. Sporadic large pulses versus daily periodic emissions.” International Journal of Hydrogen Energy. https://www.sciencedirect.com/science/article/abs/pii/S0360319920341987

[29] Alphasense Ltd. “Hydrogen Sensors (H2)” product line and H2-AF specification. https://www.alphasense.com/products/view-by-target-gas/hydrogen-h2; https://store.alphasense.com/h2-af/ ; https://www.co2meter.com/products/h2-af-alphasense-hydrogen-gas-sensor

[30] Industrial Scientific. “What You Need to Know about Sensor Poisons and Inhibitors.” https://www.indsci.com/en/blog/what-you-need-to-know-about-sensor-poisons-and-inhibitors

[31] [Author list not independently confirmed] (2025). “Drilling-induced hydrogen: A challenge for natural hydrogen exploration.” International Journal of Hydrogen Energy, vol. 149. https://www.sciencedirect.com/science/article/pii/S0360319925030885

[32] U.S. Geological Survey (1986). “Soil gas studies around hydrogen-rich natural gas wells in northern Kansas.” Open-File Report 86-461. https://pubs.usgs.gov/of/1986/0461/report.pdf

[33] Gold Hydrogen Ltd. “Ramsay Project” page. https://www.goldhydrogen.com.au/ramsay-project/

[34] La Française de l’Énergie (2023). “FDE announces the discovery of natural hydrogen in the Lorraine mining basin.” https://www.francaisedelenergie.fr/en/blog/2023/05/15/fde-announces-the-discovery-of-natural-hydrogen-in-the-lorraine-mining-basin/ ; CNRS News, “A gigantic hydrogen deposit in northeast France?” https://news.cnrs.fr/articles/a-gigantic-hydrogen-deposit-in-northeast-france

[35] Myagkiy, A., Moretti, I., Brunet, F. (2020). “Space and time distribution of subsurface H2 concentration in so-called ‘fairy circles’: Insight from a conceptual 2-D transport model.” BSGF — Earth Sciences Bulletin 191, 13. https://www.bsgf.fr/articles/bsgf/full_html/2020/01/bsgf190083/bsgf190083.html; Science.org (2024 news), “How hydrogen-leaking ‘fairy circles’ might form.” https://www.science.org/content/article/how-hydrogen-leaking-fairy-circles-might-form

[36] U.S. Geological Survey (2024). “The Potential for Geologic Hydrogen for Next-Generation Energy.” https://www.usgs.gov/news/featured-story/potential-geologic-hydrogen-next-generation-energy

[37] Gelman, S.E., Hearon, J.S., Ellis, G.S. (2025). “Prospectivity mapping for geologic hydrogen.” USGS Professional Paper 1900. DOI: 10.3133/pp1900. https://pubs.usgs.gov/publication/pp1900