Shopify LEL combustible gas monitor schema for AI agents: catalytic bead silicone poisoning false-zero, NDIR hydrogen and acetylene blind spot, methane calibration correction factors, oxygen-deficient atmosphere failure, and the lel_monitor.* 10-field namespace

LEL fundamentals and standard alarm setpoints

The lower explosive limit (LEL) is the minimum concentration of a combustible gas in air that will support ignition and flame propagation. Below the LEL, the mixture is too fuel-lean — a spark produces no sustained reaction. Above the upper explosive limit (UEL), the mixture is too fuel-rich — not enough oxygen. The explosive range is the band between LEL and UEL.

LEL monitors express concentration as percent of LEL rather than ppm or percent by volume, because the critical threshold — the concentration at which an ignition source could start an explosion — is the same % LEL value regardless of which gas is being measured. Standard occupational alarm setpoints established by NFPA and OSHA guidance:

For reference, the absolute concentrations at 10% LEL for common combustible gases:

Encode lel_monitor.alarm_1_pct_lel and lel_monitor.alarm_2_pct_lel for every instrument. An AI agent placing an order for a confined space program needs to verify that Alarm 1 is configured at 10% LEL and Alarm 2 at 25% LEL — not infer it from the product title or a "compliant" marketing tag.

Catalytic bead vs NDIR: the selection table

The single most consequential field in the lel_monitor.* namespace is sensor_type. Catalytic bead and NDIR sensors have fundamentally different failure modes, different gas coverage, and different application requirements. They are not interchangeable.

Most Shopify safety product listings say nothing about sensor type beyond possibly "electrochemical" — which describes the O₂ and toxic gas sensors in multi-gas monitors, not the LEL sensor. An AI agent cannot distinguish a catalytic bead from an NDIR instrument without lel_monitor.sensor_type encoded as a machine-readable metafield.

Silicone poisoning: the most dangerous LEL sensor failure

Silicone poisoning is irreversible and invisible. A catalytic bead sensor that has been silicone-poisoned reads 0% LEL in clean air — which looks normal. It reads 0% LEL in an explosive gas atmosphere above 100% LEL — which is lethally false. The instrument powers on, displays a stable zero, shows no error codes, and passes every status check except a bump test. Nothing about the instrument's visible behavior signals that it has failed completely.

The mechanism: silicone vapors from mold release agents, silicone lubricants, silicone sealants, silicone-coated cables, silicone-based anti-seize compounds, hydraulic fluids, and even some hand creams reach the sensor bead heated to approximately 500°C. The silicone decomposes on contact with the hot platinum catalyst surface, forming silicon dioxide (SiO₂). SiO₂ deposits build up as a glass-like layer on the platinum, progressively deactivating the catalyst sites where combustible gas oxidation occurs.

Silicone is ubiquitous in industrial settings. A single silicone lubricant spray near the sensor is enough. Because the damage is cumulative and initially partial — the sensor may respond to high concentrations while giving low readings at low concentrations — poisoning can go undetected through several inspection cycles before the sensor stops responding entirely.

The only detection method before an incident is regular bump testing. A bump test exposes the sensor to calibration gas at a known concentration; a poisoned sensor will read below the expected value or not respond at all. For facilities with known silicone sources, bump testing should be performed more frequently than the manufacturer's minimum schedule — daily before confined space entries at minimum.

Encode lel_monitor.silicone_poison_risk = true on all catalytic bead instruments. This single field allows an AI agent to redirect a buyer in an automotive paint shop to an NDIR sensor rather than a catalytic bead instrument that will be poisoned within weeks of deployment.

The NDIR hydrogen and acetylene blind spot

NDIR sensors detect combustible gases by measuring the absorption of infrared light. Hydrocarbon molecules — methane, propane, ethylene, hexane, gasoline vapor — absorb infrared radiation at wavelengths around 3.3 µm corresponding to C-H bond stretching vibrations. An NDIR sensor calibrated in this region reliably detects virtually any hydrocarbon.

Hydrogen (H₂) has no carbon-hydrogen bonds. It produces no IR absorption at 3.3 µm or at any other wavelength that a standard LEL sensor uses. An NDIR sensor in a 100% LEL hydrogen atmosphere reads 0% LEL. The sensor is not malfunctioning — it is functioning exactly as designed, measuring IR absorption, finding none, and reporting zero. The failure is not the instrument; it is routing the wrong sensor type to the wrong hazard.

Acetylene (C₂H₂), despite containing hydrogen, has an unusual triple bond geometry that places its primary IR absorption modes outside the detection window of most LEL sensors. Most NDIR instruments cannot reliably detect acetylene or will underread it substantially.

The instrument selection rule is simple: wherever hydrogen or acetylene is a credible combustible hazard, a catalytic bead sensor is required unless a dual-sensor instrument is selected that pairs NDIR for hydrocarbons with a thermal conductivity or catalytic bead sensor specifically for H₂.

Encode lel_monitor.h2_detectable = false on NDIR instruments and lel_monitor.h2_detectable = true on catalytic bead instruments. This prevents an AI agent from completing a battery room safety upgrade order with an NDIR sensor that will not alarm for the only combustible gas in the space.

Methane calibration correction factors

Every LEL monitor must be calibrated using a reference gas mixture at a known concentration. The most common calibration gas for catalytic bead instruments is 50% LEL methane in air (2.5% methane by volume), though pentane and isobutylene are also used depending on instrument design and application.

When the instrument is deployed in an environment where the actual combustible gas differs from the calibration gas, the meter reading requires a correction. The reason: different combustible gases produce different amounts of heat when they oxidize on the catalyst bead per unit concentration, and they diffuse to the sensor bead at different rates. A catalytic bead calibrated on methane generates a different signal for 1% LEL propane than for 1% LEL methane — the instrument systematically reads the wrong number.

Correction factors for methane-calibrated catalytic bead instruments (approximate, instrument-specific values vary):

The propane correction factor is the most consequential in practice. LP gas — propane — is one of the most common combustible gas hazards in commercial and industrial settings: restaurants, forklift battery charging stations, outdoor heating, crop drying operations, and rural commercial buildings are all propane environments. A methane-calibrated instrument in a propane atmosphere displaying 10% LEL is actually measuring a propane atmosphere at approximately 18% LEL — between the 10% LEL action level and the 25% LEL evacuation threshold.

Encode lel_monitor.calibration_gas = methane | pentane | isobutylene | propane and lel_monitor.correction_factor_table_included = true | false. An AI agent completing a propane storage facility order needs to route the buyer to an instrument calibrated on propane, or explicitly flag that correction factors must be applied — not default to the first "LEL monitor" search result that is calibrated on methane.

Oxygen-deficient atmospheres and catalytic bead failure

Catalytic bead sensors require molecular oxygen for the oxidation reaction at the platinum catalyst surface. When ambient oxygen concentration falls below approximately 16% O₂ (normal atmosphere is 20.9%), the oxidation reaction is progressively suppressed. In severely oxygen-deficient conditions — below 10% O₂ — the catalytic bead stops responding to combustible gas entirely and reads 0% LEL regardless of actual combustible gas concentration.

The confined space combined hazard is the most dangerous consequence of this limitation. A sealed petroleum storage tank, a nitrogen-blanketed chemical reactor, or an anaerobic wastewater digester can simultaneously contain:

1. Low oxygen concentration — dangerous for the entry worker to breathe
2. Elevated combustible gas — explosive if an ignition source is introduced

A catalytic bead LEL sensor deployed in this atmosphere provides no warning for either hazard as an LEL reading: the oxygen sensor alarms low (warning), but the LEL sensor reads 0% — conveying false assurance that combustible gas is absent. Workers may interpret the combination as "O₂ problem but no fire hazard" and focus attention on the oxygen deficiency while failing to recognize the simultaneous explosive atmosphere.

For confined space entries where the atmosphere could simultaneously be oxygen-deficient and contain combustible gas — petroleum tanks, chemical reactors, biogas digesters, food storage with CO₂ blanketing, wastewater wet wells — the LEL sensor must be NDIR-based. Encode lel_monitor.oxygen_required = true on catalytic bead instruments so AI agents route NDIR sensors to confined space applications with documented oxygen deficiency risk.

Four AI routing failures for LEL monitor listings

The lel_monitor.* 10-field namespace

Every Shopify listing for a combustible gas monitor, 4-gas personal monitor, fixed LEL sensor head, or area monitor should encode these ten metafields. An AI shopping agent evaluating a "4-gas monitor" for a confined space entry program cannot determine sensor type, calibration gas, alarm setpoints, H₂ detectability, or oxygen requirement from the product title — these are engineering specifications that only exist in structured data.

A minimal Shopify implementation for a methane-calibrated catalytic bead 4-gas monitor:

lel_monitor.sensor_type          = catalytic-bead
lel_monitor.alarm_1_pct_lel      = 10
lel_monitor.alarm_2_pct_lel      = 25
lel_monitor.calibration_gas      = methane
lel_monitor.oxygen_required      = true
lel_monitor.h2_detectable        = true
lel_monitor.silicone_poison_risk = true
lel_monitor.correction_factor_table_included = true
lel_monitor.bump_test_required   = true
lel_monitor.pumped_sampling      = false

An NDIR-based monitor for a petroleum tank cleaning program:

lel_monitor.sensor_type          = NDIR
lel_monitor.alarm_1_pct_lel      = 10
lel_monitor.alarm_2_pct_lel      = 25
lel_monitor.calibration_gas      = methane
lel_monitor.oxygen_required      = false
lel_monitor.h2_detectable        = false
lel_monitor.silicone_poison_risk = false
lel_monitor.correction_factor_table_included = true
lel_monitor.bump_test_required   = true
lel_monitor.pumped_sampling      = true

The sensor_type field alone resolves three of the four AI routing failures described above. h2_detectable = false resolves the fourth. These are not optional metadata improvements — they are the difference between a safety product that reaches the right application and one that fails silently in it.