Calibrating a catalytic-bead or IR LEL sensor with the wrong reference gas is one of the fastest ways to make a combustible gas detector go silently blind, and the gap between a methane and a propane span cylinder is the most common source of that failure [S1][S2].
Across oil and gas, LPG distribution, HVAC service, and confined-space entry, the field rule is: calibrate to the target gas when you know it, default to methane when you do not, and use a methane blend labelled "propane equivalent" only when a single cylinder has to drive a multi-gas detector on a quad-mix regulator [S7].
Why methane is the default span gas
Methane's lower flammability limit sits at 5.0 vol% in air, and a 50% LEL span is therefore 2.5 vol% CH4 in air, the most common certified cylinder on the market and the reference that almost every combustible gas detector ships with from the factory [S3].
Methane also has the highest auto-ignition temperature of the common calibration gases, roughly 537 degC, against about 260 degC for pentane, so a sensor that can still see pentane may already be unable to see methane [S5]. A sensor whose catalytic activity has degraded by poisoning tends to lose methane response first, which is why methane calibration is the conservative choice in mixed atmospheres and around unknown backgrounds [S1][S4].
Correction-factor guidance from the major LEL sensor makers goes the same way: if you calibrate on methane and then display another gas, methane calibration tends to overestimate the other gas's %LEL as the sensor ages, which is a fail-safe direction, while pentane or propane calibration can quietly under-read methane in a natural-gas leak [S4][S9].
When propane (or propane equivalent) is the right pick
Propane's LEL is 2.1 vol%, so a 50% LEL propane span is 1.05 vol% C3H8, and the cross-sensitivity pattern on a catalytic bead is different from methane: most LEL sensors are less sensitive to propane than to methane, which is why a methane blend labelled 1.62% methane / 50% LEL propane equivalent produces the same bridge signal as 1.05 vol% propane on a healthy sensor [S7][S8].
That maths is exactly why a 1.62% methane / 50 ppm CO / 25 ppm H2S / 18% O2 quad-mix exists, it lets one regulator, one cylinder, and one bump test verify a portable gas detector against the propane reading the instrument is set up to display, instead of carrying a separate propane span bottle into a confined space [S7].
For pure LPG work, propane bulk terminals, cylinder filling bays, forklift refuelling rooms, autogas dispensers, the cleaner answer is still a real propane span gas, because it removes the equivalent correlation from the loop and the bump test reads the same number the instrument will display on leak [S3][S5]. Correction factors for propane against methane calibration run about 1.7 to 2.0 on common catalytic-bead sensors, so a 50% LEL methane bump on a propane-set instrument reads around 25 to 30% LEL [S4][S8].
Decision matrix: methane, propane, or equivalent
Three decision dimensions sort the choice in the field. First, target gas: natural gas, landfill gas, coal-mine CH4, sewer gas all point to methane; LPG, autogas, refrigerant-grade hydrocarbons, and most solvents push to propane or pentane [S3][S5]. Second, sensor health margin: methane is conservative because lost methane sensitivity is the first symptom of a poisoned or aged catalytic bead, so methane calibration over-reads rather than under-reads as the sensor degrades [S1][S4]. Third, cylinder logistics: a methane/propane-equivalent quad-mix on a single regulator wins on bump-test speed for service teams carrying one multi-gas detector across mixed sites [S7].
Concretely, for a fleet of portable detectors on a mixed LPG and natural-gas contract, a 2.5 vol% methane (50% LEL CH4) cylinder is the correct factory-span choice; a 1.62% methane blend labelled 50% LEL propane equivalent is the correct bump-test span; and a 1.05 vol% propane (50% LEL C3H8) cylinder belongs in the bench station if the instrument is locked to a propane reading [S7][S8].
Oxygen, temperature, and humidity constraints that drive the choice
All catalytic-bead LEL sensors need at least 10 vol% O2 to read correctly up to 100% LEL of nearly every common hydrocarbon, and that floor drops to roughly 5 vol% O2 only at about 10% LEL methane before the sensor starts to under-read [S4]. Pushing a sensor from air into pure nitrogen causes a transient that decays over several minutes, so bump testing a methane span immediately after entering a purged vessel will read low and should not be trusted.
Across 0 to 40 degC the bridge signal shifts by under 4%, and from 20% to 90% RH by about 2%, which is small enough to ignore for bump tests but large enough to fix a 50% LEL span on a climate-controlled bench rather than in a hot truck [S4]. These are the same constraints that make infrared LEL sensors attractive for low-oxygen or high-humidity sites, but they still need their own certified span gas, normally methane or propane at 50% LEL, on a quarterly cadence.
Pentane equivalent: the middle ground most teams should avoid as a primary span
A "pentane equivalent" methane cylinder is a methane concentration chosen so that a new sensor's bridge signal matches the signal a 25% or 50% LEL pentane span would produce, which puts a manufacturer-supplied correlation right in the middle of the calibration loop [S2]. On a new sensor the worst-case accuracy is roughly plus or minus 25%, and once the sensor starts losing methane sensitivity, that correlation is no longer the same number it was at the factory [S2][S4].
The practical consequence is that pentane-equivalent span gas hides methane poisoning rather than catching it, which is the opposite of what a safety instrument is supposed to do [S2]. Two better patterns exist: either run a real pentane span on the bench and bump test with methane before each use, or use a methane/propane-equivalent cylinder as a bump-only gas while keeping a real methane or propane span for the formal calibration [S2][S4].
How to read a methane-calibrated instrument on a propane leak
If the instrument was span-calibrated on 50% LEL methane and the atmosphere contains 50% LEL propane, the display will read about 25 to 30% LEL rather than 50%, which still trips a 10% or 20% LEL alarm with a 2-3x safety factor on top [S2][S4]. That is the conservative direction the field expects, and it is why running a methane-calibrated fixed gas detector on an LPG line is acceptable as long as the alarm setpoints and the correction factor are documented.
The dangerous case is the reverse, propane calibration on a methane leak: the sensor reads 50% LEL propane equivalent but the actual methane is well above 100% LEL, the alarm is late, and the building is already past the explosive threshold [S4][S9]. This is the failure mode the correction-factor tables exist to prevent, and the reason the standard operating answer in mixed atmospheres is still methane calibration with documented cross-sensitivity factors in the instrument's setup.
Field checklist for selecting a calibration cylinder
Match the span to the target gas first (methane for natural gas service, propane for LPG service, pentane for solvent or VOC work) and only deviate to an equivalent gas when a multi-gas cylinder is the only practical option [S3][S5]. Confirm the cylinder concentration against the sensor's span setting, 2.5 vol% CH4 for 50% LEL methane, 1.05 vol% C3H8 for 50% LEL propane, 1.62 vol% CH4 for 50% LEL propane equivalent, and roughly 0.7 vol% n-pentane for 50% LEL pentane on most catalytic-bead sensors [S7][S8][S9].
Verify oxygen is above 10 vol% at the bump-test point, certify the cylinder on a 6 to 12 month shelf-life rotation, and after any failed bump test, run a real-gas methane bump before the instrument goes back on a confined-space job [S1][S4]. For deeper guidance on how a methane span interacts with detector mounting height around natural gas and LPG releases, the Gas Detector Mounting Height: Methane, Propane, H2S Spec Guide sits next to this article in the LEL workflow.