Concentration units: ppm, ppb, %LEL, mg/m³, and partial pressure

Summary

Every instrument in this wiki reports “how much hydrogen” in a different unit because each sensor’s physics responds to a different physical quantity. A “ppm” reading counts molecules per million total molecules; a mass-based reading weighs them; a safety alarm’s “%LEL” is a percentage of a danger threshold, not of the gas mixture. Hydrogen makes the differences unusually large because it is the lightest molecule there is: the same sample reads 14.4× smaller by mass than by mole count. A geologist reading ppm off a field meter and a safety officer reading %LEL off a pellistor can be describing the same gas cloud or situations four orders of magnitude apart, and the number on the screen does not say which.

This page is the unit algebra under every other number in this wiki: what each unit counts, exact conversions between them, and where the conversions hide errors of 14×, 9%, or six orders of magnitude.

Context numbers used throughout

• Atmospheric H₂ background ≈ 0.53 ppmv (530 ppb)
• atmospheric He ≈ 5.24 ppmv
• geologic H₂ soil-gas exploration anomalies span roughly 1 ppmv to percent level
• H₂’s Lower Explosive Limit (LEL) = 4 vol% = 40,000 ppmv
• a representative field electrochemical analyzer’s vendor H₂ range is 0–1,000 ppmv

These are the anchor numbers used across every sibling page; see Geologic context and Electrochemical cells.

Mole fraction: what “ppm” counts

“1 ppm” is a counting ratio: one molecule of the target per million total. The formal quantity is the mole fraction (IUPAC: amount fraction) [6]:

$$x_{H_2} = \frac{n_{H_2}}{n_{tot}} = \frac{n_{H_2}}{n_{H_2}+n_{N_2}+n_{O_2}+\dots}$$

It says nothing about molecular weight, only counts. ppm, ppb ($10^{-9}$), and ppt ($10^{-12}$) are the same ratio at different decimal scales:

$$1\ \mathrm{ppm} = 1{,}000\ \mathrm{ppb} = 1{,}000{,}000\ \mathrm{ppt}$$

The global H₂ background in outdoor air, ≈0.53 ppmv = 530 ppbv = 530,000 pptv, is one concentration written three ways. Instruments advertising sub-ppm or ppb detection limits are trying to see below that atmospheric floor; see Reading a sensor datasheet for what such claims actually promise.

Why ppmv and “ppm by mole” are identical

At the same temperature and pressure, equal volumes of any ideal gas contain equal numbers of molecules (Avogadro), so one mole of any gas occupies the same molar volume $V_m = RT/P$. Extracting the H₂ from a mixture and re-compressing it to the mixture’s own $T$ and $P$ gives $V_{H_2} = n_{H_2}V_m$, while $V_{tot} = n_{tot}V_m$, so

$$\frac{V_{H_2}}{V_{tot}} = \frac{n_{H_2}\,V_m}{n_{tot}\,V_m} = \frac{n_{H_2}}{n_{tot}} = x_{H_2}$$

The $V_m$ cancels: volume fraction equals mole fraction with no approximation beyond ideal-gas behavior [6], which holds well at atmospheric and most soil-gas or wellhead pressures and bends only at deep-downhole pressures. This is Dalton’s law restated: partial pressure $p_i = x_i P_{tot}$, mole fraction, and volume fraction carry the same information.

Unless a source says otherwise, a gas-phase “ppm” means ppmv = ppm by mole, which is what essentially every field instrument, GC, and mass spectrometer here reports. The unit that is not interchangeable is ppm by mass.

Mass units: ppmm and mg/m³

ppmm: counting vs. weighing

Counting and weighing disagree whenever the target’s molar mass differs from the mixture’s. H₂ is the extreme case: $M_{H_2}=2.01588$ g/mol [1] against dry air’s $\bar M_{air}=28.9652$ g/mol [3], a factor of 14.4. Mass fraction (ppmm or ppmw) is

$$w_{H_2} = \frac{m_{H_2}}{m_{tot}}, \qquad \boxed{\;\text{ppmm} = \text{ppmv}\times\frac{M_{H_2}}{\bar M_{mix}}\;}$$

Worked example. $\frac{2.01588}{28.9652} = 0.0696$, so 1 ppmv H₂ = 0.0696 ppmm. Equivalently, the same H₂ reads 14.37× smaller by mass than by mole/volume: a mixture correctly called “1 ppmv H₂” holds 0.0696 mg of H₂ per kilogram of gas. Other co-tracer gases sit much closer to 1:1:

Gas	Molar mass	ppmm/ppmv ratio
CH₄	16.04 g/mol	0.554
CO₂	44.01 g/mol	1.519 (larger by mass, the opposite direction)
N₂	28.01 g/mol	almost exactly 1:1

No atmospherically relevant gas has a larger ppmv-to-ppmm gap than H₂, so intuition built on CH₄/CO₂ work fails on hydrogen. In practice nearly everything quoted for H₂ is ppmv: calibration cylinders are certified gravimetrically but reported as mole fraction (ISO 6142; see Standards & regulations), and H₂ has no OSHA PEL or ACGIH TLV at all, being classified as a simple asphyxiant controlled by a minimum-oxygen rule [10]. Mass-based reporting appears mainly in cross-gas comparisons and European mg/Nm³ industrial-hygiene or emissions documents [11].

mg/m³: mass per physical volume, pinned to T and P

Mass concentration answers “how many milligrams of H₂ per cubic meter of gas, at the temperature and pressure it is actually at”:

$$\boxed{\;C\;[\mathrm{mg/m^3}] = \text{ppmv}\times\frac{M_{H_2}\;[\mathrm{g/mol}]}{V_m(T,P)\;[\mathrm{L/mol}]}\;}$$

$V_m = RT/P$ is not a constant, so an mg/m³ number is only as good as its stated $(T,P)$. With the exact $R=8.314462618\ \mathrm{J\,mol^{-1}K^{-1}}$ [13], the molar volumes in circulation:

Reference condition	$T$	$P$	$V_m$ (L/mol)	Who uses it
”Ambient”/lab, 25 °C, 1 atm	298.15 K	101.325 kPa	24.465	Rigorous ideal-gas value
ACGIH/OSHA industrial-hygiene convention	298.15 K	101.325 kPa	24.45	The standard ppm↔mg/m³ formula [4]
Classic STP, 0 °C, 1 atm	273.15 K	101.325 kPa	22.414	Pre-1982 IUPAC STP; still what most chemists mean by “STP”
Modern IUPAC STP, 0 °C, 1 bar	273.15 K	100.000 kPa	22.711	IUPAC’s post-1982 STP, a different number from classic STP
20 °C, 1 atm	293.15 K	101.325 kPa	24.06	Some European/industrial-hygiene conventions

Worked example: 1 ppmv H₂ at each condition, using $M_{H_2}=2.01588$ g/mol [1]:

Reference condition	$V_m$ (L/mol)	$C$ for 1 ppmv H₂
25 °C, 1 atm (rigorous)	24.465	0.0824 mg/m³
25 °C, 1 atm (ACGIH/OSHA)	24.45	0.0824 mg/m³
Classic STP (0 °C, 1 atm)	22.414	0.0899 mg/m³
Modern IUPAC STP (0 °C, 1 bar)	22.711	0.0888 mg/m³
20 °C, 1 atm	24.06	0.0838 mg/m³

First row arithmetic: $C = 1\times(2.01588/24.465) = 0.0824$ mg/m³. The same 1 ppmv spans 0.0824–0.0899 mg/m³, a ~9% spread, purely from the silently assumed reference condition. “Normal cubic meter” (Nm³) and “standard cubic foot” (scf) are not single conventions either: 0 °C/101.325 kPa, 0 °C/100 kPa, 15 °C/101.325 kPa, and 60 °F/14.73 psia are all in active use [12]. Never accept an mg/m³, mg/Nm³, or scf-based number without its $(T,P)$.

The safety ladder: vol%, LEL, and %LEL

Percent by volume is ppmv with the decimal moved: $\mathrm{vol\%} = \mathrm{ppmv}/10{,}000$.

The Lower Explosive Limit is the minimum fuel concentration in air, in vol%, that sustains a flame. For hydrogen, LEL = 4 vol% = 40,000 ppmv [8][9]. %LEL is an instrument reading expressed as a percentage of the LEL value itself, not of the mixture:

$$\mathrm{ppmv} = \frac{\%\mathrm{LEL}}{100}\times 40{,}000\ \mathrm{ppmv}$$

%LEL	vol% H₂	ppmv H₂
1%	0.04%	400
10%	0.4%	4,000
25%	1.0%	10,000
50%	2.0%	20,000
100% (= LEL)	4.0%	40,000

Worked check, 10% row: $10/100\times4\,\mathrm{vol\%} = 0.4\,\mathrm{vol\%} = 4{,}000$ ppmv [8].

10% LEL

10% LEL is the near-universal industry alarm threshold for combustible-gas detectors, one order of magnitude below flammability [9]. For hydrogen it is exactly 4,000 ppmv: the boundary this wiki treats as the line between exploration signal and safety event.

The two instrument families read in different units for a reason. A pellistor is calibrated and certified (IEC 60079-29-1) on the %LEL scale because its job is the binary safety question, and its ~1 %LEL ≈ 400 ppmv floor is two to three orders coarser than exploration anomalies (Catalytic pellistors). Electrochemical, MOX, and optical instruments calibrate against ppmv standards because their value is the range below what a pellistor sees (Electrochemical cells, MOX sensors).

Partial pressure: what many sensors actually respond to

Every unit above is a ratio, and ratios don’t change when the whole mixture is compressed. Several transducers respond instead to the absolute partial pressure (Dalton’s law):

$$\boxed{\;p_{H_2} = x_{H_2}\cdot P_{tot}\;}$$

Families that respond to $p_{H_2}$, not to ppmv:

• Palladium-resistive (Sieverts’-law) sensors: $[\mathrm H]/[\mathrm{Pd}]=K_S\sqrt{p_{H_2}}$ (Palladium-resistive sensors)
• Work-function / MIS-FET / Kelvin-probe devices: coverage and voltage shift follow $\ln p_{H_2}$ (Work-function devices)
• Absorption spectroscopy (NDIR/TDLAS/QCL): Beer–Lambert absorbance scales with molecules in the light path, i.e. $p_{H_2}$ (IR absorption and proxies)
• Mass spectrometers/RGAs: the ion source literally measures partial pressure; “ppm” appears only after dividing by total pressure (Mass spectrometry)

Worked example: 100 ppmv at the surface vs. downhole. The mole fraction is $x_{H_2}=10^{-4}$ in both cases:

Location	$P_{tot}$	$p_{H_2}=x_{H_2}\cdot P_{tot}$
Surface	1 bar	$10^{-4}$ bar = 100 µbar = 10 Pa
Downhole	200 bar (~2,900 psi)	$2\times10^{-2}$ bar = 20 mbar = 2,000 Pa

The partial pressure is 200× higher downhole for the numerically identical reading. Consequences: a Sieverts’-law device’s raw signal scales as $\sqrt{p_{H_2}}$, so the downhole signal is $\sqrt{200}\approx14.1\times$ larger before any thermal or poisoning effect; and 20 mbar sits at the Pd α→β hydride transition threshold (reported near 20–30 mbar at room temperature), so the same ppmv that is hysteresis-free at the surface pushes the sensor into a nonlinear two-phase regime at depth. A log-linear work-function device shifts by $\ln(200)\approx5.30$ calibration units, correctable only with an independent total-pressure measurement. This is why Downhole deployment treats downhole H₂ sensing as an open transduction problem, not a housing-rating problem.

Dissolved phase: molarity and Henry’s law

For H₂ in groundwater, brine, or produced water, the unit becomes moles of solute per liter of solution: molarity [5], almost always in nanomolar (nM) or micromolar (µM) for natural waters. Henry’s law bridges gas and dissolved phases:

$$C_{aq} = k_{H,cp}\cdot p_{H_2}, \qquad k_{H,cp}\ \text{in mol}\,\mathrm{L^{-1}atm^{-1}}$$

Some sources give the reciprocal volatility form $p_{H_2}=k_{H,pc}\cdot C_{aq}$ with $k_{H,pc}=1/k_{H,cp}$ [7]; using the wrong one un-inverted is a ~six-order-of-magnitude error for H₂. NIST’s value for H₂ in water at 25 °C is $k_{H,cp}\approx7.8\times10^{-4}\ \mathrm{mol\,L^{-1}atm^{-1}}$ ($k_{H,pc}\approx1{,}300\ \mathrm{L\,atm\,mol^{-1}}$) [2]. H₂ is poorly soluble: the dimensionless constant $H_{cc}=C_{gas}/C_{aq}\approx52$ means ~52× more H₂ in the gas phase than in an equal volume of water at equilibrium.

Worked examples ($p_{H_2}=C_{aq}/k_{H,cp}$):

• Atmospheric anchor: 0.53 ppmv at 1 atm gives $p_{H_2}=5.3\times10^{-7}$ atm, so $C_{aq}=(7.8\times10^{-4})(5.3\times10^{-7})=4.13\times10^{-10}$ mol/L ≈ 0.41 nM.
• A 0.1 nM detection floor: $p_{H_2}=1.28\times10^{-7}$ atm ≈ 130 nbar ≈ 128 ppb equivalent at a 1-atm headspace, below atmospheric background.
• 10 nM (typical of microbial redox boundaries): $p_{H_2}=1.28\times10^{-5}$ atm ≈ 13 µbar ≈ 12.8 ppmv equivalent, 100× the 0.1 nM case. The diagnostically meaningful dissolved range is compressed.

Related conventions: mass concentration (mg/L) is $C\times M_{H_2}$, and H₂’s tiny molar mass makes 1 nM only 2.016 ng/L; molality (per kg of solvent) is preferred in high-precision work because a liter expands with temperature and a kilogram does not, a distinction that matters in hot formation brine; the Bunsen and Ostwald coefficients are pre-SI forms of the same solubility ratio [7]. Field methodology and the microbial redox ladder that makes 1–10 nM diagnostic live on Dissolved gas in groundwater.

Common mistakes

Six ways this bites people in the field

1. Treating ppmv and ppmm as interchangeable. For CH₄ or N₂ the ratio is 0.55–1.0; for H₂ it is a 14.4× error, and in the unintuitive direction (the mass number is smaller). Apply $\mathrm{ppmm}=\mathrm{ppmv}\times0.0696$ explicitly.
2. Converting mg/m³ without the reference temperature and pressure. Four “standard” molar volumes circulate (24.45–24.465, 22.414, 22.711, 24.06 L/mol), a ~9% spread for identical ppmv. Treat an mg/m³, mg/Nm³, or scf number without its $(T,P)$ as unusable until the source is asked.
3. Confusing %LEL with vol% or ppm. 100% LEL = 4 vol% = 40,000 ppmv for H₂; misreading the scales produces 25×, 400×, or 10,000× errors. Convert through the gas-specific LEL first.
4. Forgetting partial-pressure scaling downhole. The same 100 ppmv carries 200× the partial pressure at 200 bar, a 14× raw-signal change for a Sieverts’-law device, and can cross a phase-transition threshold never approached at the surface.
5. Getting the Henry’s-law reciprocal backwards: $k_{H,cp}$ (mol·L⁻¹·atm⁻¹) and $k_{H,pc}$ (L·atm·mol⁻¹) are reciprocals ~six orders apart for H₂. Check the printed units.
6. Conflating molarity, molality, and mass concentration in dissolved-phase reporting, especially in deep, hot brine where the liter-vs-kilogram distinction stops being negligible.

How this shapes sensor choice and field practice

Field verdict

Before comparing any two concentration numbers, pin down four things: mass or mole/volume basis; the reference $(T,P)$ if mg/m³ is involved; the total pressure at the sensor if a partial-pressure-driven transducer is involved; and gas phase or aqueous phase. Skipping any one is how the mistakes above become a bad procurement, a missed downhole interpretation, or a false safety call.

• Log raw instrument units alongside converted readings. “342 mV, GA5000 H₂ channel, ambient 970 mbar” can be re-derived against a revised calibration later; “12 ppm” alone cannot. See Calibration, drift & QA/QC.
• %LEL-native instruments are safety triage, not exploration quantification. A %LEL reading of “0” means only that H₂ is below ~400 ppmv. Conversely, a ppmv-native cell may saturate below hazardous %LEL levels, so don’t make it the sole safety instrument either. See Choosing a sensor.
• Any downhole deployment of a partial-pressure-driven sensor needs total-pressure correction designed in from the start: a co-located pressure measurement or a validated high-pressure Sieverts’/Temkin correction. Without one, a downhole “ppmv” is not a well-defined quantity.
• Establish which unit family an instrument natively reports before comparing spec sheets. Two “0.5 ppm resolution” claims on different transduction physics are not comparable unless pinned to the same total pressure. See Reading a sensor datasheet and Figures of merit.
• Bridge gas-phase (ppmv) and dissolved-phase (nM) data through an explicit Henry’s-law step, and remember measured dissolved values are usually a microbially imposed steady state, not passive equilibrium with a deep source. See Dissolved gas in groundwater and Gas vs. aqueous sensing.

Cheat sheet

All rows computed from: $M_{H_2}=2.01588$ g/mol [1]; $\bar M_{air}=28.9652$ g/mol [3]; $V_m(25\,^{\circ}\mathrm C,1\ \mathrm{atm})=24.45$ L/mol [4]; LEL = 40,000 ppmv [8][9]; $k_{H,cp}=7.8\times10^{-4}\ \mathrm{mol\,L^{-1}atm^{-1}}$ [2] (dissolved column assumes gas–water equilibrium at 1 atm, 25 °C).

ppmv	ppmm	mg/m³ (25 °C, 1 atm)	vol%	%LEL	$p_{H_2}$ @ 1 atm total	Dissolved equiv. (nM)
0.53 (atm. background)	0.0369	0.0437	0.000053%	0.0013%	53.7 mPa	0.41 nM
1	0.0696	0.0824	0.0001%	0.0025%	0.101 Pa	0.78 nM
100	6.96	8.24	0.01%	0.25%	10.1 Pa	78 nM
1,000 (field-analyzer full scale)	69.6	82.4	0.1%	2.5%	101.3 Pa	780 nM (0.78 µM)
4,000 (10% LEL alarm point)	278	329	0.4%	10%	405 Pa	3.12 µM
40,000 (100% LEL = 4 vol%)	2,783	3,298	4%	100%	4,053 Pa	31.2 µM

Two things to carry forward. At the LEL row, the mass fraction (2,783 ppmm ≈ 0.28%) is over 14× smaller than the volume fraction (4%): a mixture at the flammability threshold is only a quarter of one percent hydrogen by weight. And the dissolved column is a unit conversion, not a field prediction; real groundwater dissolved-H₂ rarely sits at passive equilibrium because of microbial consumption (Dissolved gas in groundwater).

Unit family	What it counts	Pressure-dependent?	Typical native users
ppmv (mole fraction)	molecules of H₂ per million total molecules	No	Electrochemical, MOX, GC, mass spec
ppmm/ppmw (mass fraction)	mass of H₂ per million mass units of mixture	No	Emissions/hygiene contexts, mass-balance engineering
mg/m³, µg/m³	mass of H₂ per physical volume, at stated $(T,P)$	Yes	Occupational exposure limits, emissions standards
vol%	ppmv ÷ 10,000	No	Wellhead/mud-gas percent-level readings, TCD instruments
%LEL	fraction of the flammability threshold	No	Catalytic pellistors, safety monitors
Partial pressure $p_{H_2}=x_{H_2}P_{tot}$	absolute pressure contributed by H₂	Yes	Pd-resistive, work-function, mass spec, absorption spectroscopy
Molarity, nM/µM	moles of H₂ per liter of water	Indirectly, via Henry’s law	Dissolved-gas sensors, bubble-strip/GC, MIMS

Interactive converter

ppmv vol% %LEL mg/m³ partial pressure (Pa) dissolved (nM) total P atm

ppmv		ppmm
vol%		%LEL
mg/m³		p(H₂)
dissolved equiv.

Constants as on this page: M(H₂) = 2.01588 g/mol [1], M̄(air) = 28.9652 g/mol [3], LEL = 40,000 ppmv [8][9], k_H,cp = 7.8×10⁻⁴ mol L⁻¹ atm⁻¹ at 25 °C [2]. The dissolved value is the passive gas–water equilibrium at 25 °C (a unit conversion, not a field prediction; see above). mg/m³ depends on the V_m convention; partial pressure and the dissolved value depend on total P.

Related pages

• Foundations overview & learning path
• How the sensor families differ
• Gas vs. aqueous sensing
• Figures of merit
• Why hydrogen is hard to sense
• Choosing a sensor
• Comparison charts
• Reading a sensor datasheet
• Electrochemical cells
• MOX (metal-oxide) chemiresistors
• Catalytic pellistors
• Palladium-resistive sensors
• Work-function devices
• Thermal conductivity & acoustic sensors
• IR absorption and proxies
• Mass spectrometry
• Gas chromatography
• Mud-gas logging
• Dissolved gas in groundwater
• Geologic context
• Downhole deployment
• Helium & co-tracers
• Environmental covariance
• Calibration, drift & QA/QC
• Soil-gas surveys
• Standards & regulations
• Glossary

References

Full reference list (13 entries)

1. NIST Chemistry WebBook, “Hydrogen” — molecular weight 2.01588 g/mol. https://webbook.nist.gov/cgi/cbook.cgi?ID=C1333740&Units=SI
2. NIST Chemistry WebBook, “Hydrogen” — Henry’s law constant data ($k^{\circ}_H\approx7.8\times10^{-4}$ mol kg⁻¹ bar⁻¹ ≈ mol L⁻¹ atm⁻¹; reciprocal ≈1,300 L·atm/mol). https://webbook.nist.gov/cgi/cbook.cgi?ID=C1333740&Mask=10
3. Wikipedia, “Density of air” — molar mass of dry air ≈0.0289652 kg/mol (28.9652 g/mol), citing the CIPM/Picard et al. moist-air density formulation. https://en.wikipedia.org/wiki/Density_of_air
4. Canadian Centre for Occupational Health and Safety (CCOHS), “Converting Occupational Exposure Limits from mg/m³ to ppm” — ppm = mg/m³ × 24.45 / MW, where 24.45 L/mol is the ACGIH-convention molar volume at 25 °C, 760 torr. https://www.ccohs.ca/oshanswers/chemicals/convert.html
5. IUPAC Gold Book, “amount concentration” (A00295). https://goldbook.iupac.org/terms/view/A00295
6. Wikipedia, “Mole fraction” — IUPAC “amount fraction” terminology; ideal-gas volume-fraction/mole-fraction equivalence via Avogadro’s law. https://en.wikipedia.org/wiki/Mole_fraction
7. Wikipedia, “Henry’s law” — dimensionless/volatility/solubility Henry’s-constant forms; Bunsen and Ostwald coefficient definitions. https://en.wikipedia.org/wiki/Henry%27s_law
8. Evikon MCI, “%LEL, %LFL, % vol, ppm – different units in gas detection,” including the worked hydrogen example (10 %LEL = 4,000 ppm; LEL = 4 vol% = 40,000 ppm). https://www.evikon.eu/news/lel-lfl-vol-ppm-different-units-in-gas-detection-a-15/
9. Industrial Monitor Direct, “Hydrogen Gas LEL Limits & Exposure Safety: Industrial Guide” — H₂ LEL 4 vol%/40,000 ppm; 10% LEL as the standard alarm/IDLH-derivation convention. https://industrialmonitordirect.com/blogs/knowledgebase/hydrogen-gas-lel-limits-and-exposure-safety-for-industrial-facilities
10. Search-verified occupational-health sources (NJ Department of Health Right-to-Know Hazardous Substance Fact Sheet: Hydrogen, and corroborating OSHA/NIOSH guidance) — hydrogen classified as a simple asphyxiant with no substance-specific OSHA PEL or ACGIH TLV; exposure control via minimum-oxygen-percentage requirement. https://www.nj.gov/health/eoh/rtkweb/documents/fs/1010.pdf
11. Wikipedia, “Air pollutant concentrations” — ppmv vs. mg/Nm³ reporting conventions in emissions regulation. https://en.wikipedia.org/wiki/Air_pollutant_concentrations
12. Airpressa, “What is Nm3? How is It Different from m3?” — Nm³/scf reference-condition variability (0 °C, 15 °C, and 60 °F conventions all in active use). https://www.airpressa.com/what-is-nm3/ (does not support the claim; this source states only the single 0 °C/1 atm convention. The multi-convention claim is correct but should be re-cited to an authority that lists the alternatives.)
13. NIST, CODATA recommended value of the molar gas constant, $R=8.31446261815324\ \mathrm{J\,mol^{-1}\,K^{-1}}$ (exact, SI-2019). https://physics.nist.gov/cgi-bin/cuu/Value?r