Gas Sensors
Electrochemical Sensors
The workhorse toxic-gas and oxygen sensors in every multi-gas monitor. Cheap, specific, low-power — and full of subtle ways to fool you.
OPS core survey monitoring TECH exotic-sensor selection & cross-sensitivity interpretation
How it works
An electrochemical (EC) sensor is a tiny fuel cell tuned to one gas. The target gas diffuses through a membrane into a liquid-electrolyte cell and reacts at an electrode. That reaction either releases or consumes electrons, producing a tiny electric current proportional to the gas concentration. More gas → more current → higher reading. Different electrode chemistries and membranes make a sensor selective for CO, or H₂S, or Cl₂, and so on.
Oxygen sensors are a special case. Most traditional O₂ sensors are consumable galvanic cells with a lead anode: oxygen entering the cell is reduced while the lead is oxidized, generating current. Because the lead is literally used up by the reaction, the sensor has a built-in lifespan — it wears out whether you use it or not.
What it's good for
- The core life-safety gases: oxygen (O₂), carbon monoxide (CO), and hydrogen sulfide (H₂S) — the standard "4-gas" set alongside a combustible/LEL channel.
- Specific toxic industrial chemicals (TICs) at low ppm with good selectivity: chlorine (Cl₂), ammonia (NH₃), hydrogen cyanide (HCN), phosphine (PH₃), sulfur dioxide (SO₂), nitrogen dioxide (NO₂), and more.
- Continuous personal and area monitoring — low power, small, alarms in real time, good for confined-space entry and clandestine-lab/agricultural scenarios.
- Quantifying a known gas in ppm, which colorimetric spot checks and broadband tools can't do continuously.
What it CANNOT do / limitations
- It only detects the gas it was built for. A CO sensor is blind to chlorine; you need a channel (a sensor) for each target. If you don't have a phosphine sensor, your meter says nothing about phosphine.
- It cannot identify an unknown. EC sensors quantify a pre-selected gas; they don't tell you what a mystery vapor is.
- Limited range. Designed for low-ppm toxic levels; high concentrations can saturate or damage the cell.
- Cross-sensitivity means a reading on one channel may actually be a different gas (see below). The number can be real current from the wrong chemical.
- A dead sensor can read zero — indistinguishable from clean air without a bump test.
Cross-sensitivity & interference
EC sensors are selective, not perfectly specific. Other gases can react at the electrode and produce a reading on the "wrong" channel — sometimes positive, sometimes negative (suppressing a real reading). This is one of the most important things to understand about your multi-gas meter, because a cross-sensitivity can either raise a false alarm or, worse, hide a real hazard.
Hydrogen (H₂) reads strongly on most CO sensors. In smoldering fire, battery, and overhaul atmospheres, a big "CO" number can be substantially hydrogen. That's not harmless — it means you may be misjudging the actual hazard mix. Cross-check with other tools and context rather than treating the CO channel as gospel.
| Sensor (channel) | Interfering gas | Typical effect |
|---|---|---|
| CO | Hydrogen (H₂) | Large positive — reads much of the H₂ as "CO." |
| Hydrogen sulfide (H₂S) | Positive — H₂S bleeds onto the CO channel. | |
| Unsaturated hydrocarbons (ethylene, acetylene) | Positive interference. | |
| H₂S | Sulfur dioxide (SO₂), mercaptans | Positive — related sulfur species read across. |
| Methanol, some alcohols/solvents | Can produce a positive response. | |
| HCN | H₂S, SO₂, NO₂, HCl | Positive interference; HCN sensors are notoriously cross-sensitive and drifty. |
| Age / humidity | Baseline drift; needs frequent verification. | |
| Cl₂ / NO₂ (oxidizing) | Each other, ozone, ClO₂ | Oxidizers cross-read on oxidizing-gas sensors (often positive). |
| Reducing gases (H₂S, SO₂) | Can read negative — may suppress or null a real Cl₂/NO₂ reading. | |
| SO₂ / NO₂ | Cross-interference between the two | Positive/negative depending on pairing; verify with the chart. |
| Oxygen | CO₂ (high), strong oxidizers | High CO₂ can slightly depress O₂ reading; oxidizers may perturb it. |
Reading a single toxic channel as if it's specific. If your "HCN" or "CO" alarms in an atmosphere full of other combustion gases, don't announce a confirmed HCN/CO level — announce that a channel alarmed and that it may be cross-sensitivity. Confirm with a colorimetric tube or a second technology before you commit tactics to it.
Degradation, poisoning & failure modes
Finite lifespan (they age out)
- Toxic EC sensors typically last ~2–3 years from manufacture — and the clock runs even in storage. Sensitivity slowly fades; eventually a bump won't pass.
- Oxygen sensors are consumed by design. The lead-anode chemistry means a traditional O₂ cell has a ~1–2 year life whether or not it ever sees a hazardous atmosphere. (Newer "lead-free"/pumped-optical O₂ options last longer but aren't universal.)
Electrolyte drying out or flooding
The cell depends on its liquid electrolyte. In very low humidity / high heat, the electrolyte can dry out, slowing or killing the sensor. In sustained high humidity, some cells can take on water and behave erratically. Both temperature and humidity shift readings, which is why bumping in field conditions matters.
Temperature & pressure effects
- Temperature changes reaction rates; extreme cold slows response and extreme heat accelerates aging and drying.
- Altitude / barometric pressure changes the O₂ reading. The O₂ sensor measures partial pressure of oxygen, then displays a percentage assuming normal pressure. Go up in altitude (or into a pressure-reduced space) and the same 20.9% air reads lower — which can false-alarm as oxygen deficiency, or mask a real drop. Zero/span at working altitude.
Poisoning & over-exposure
- Some sensors are poisoned by certain gases (e.g., silicones, some solvents) that degrade the electrode.
- Channel saturation: a big hit of gas can drive a sensor off-scale and it may need a long recovery/clear-down time in fresh air before it reads correctly again. Until it recovers, it can under-report.
The most dangerous EC failure: a sensor that has aged out, dried out, or been poisoned frequently reads a flat 0 ppm — which looks exactly like clean air. There is no obvious "I'm broken" signal in the number itself. The only routine defense is the daily bump test: if the sensor doesn't respond to a known gas, it's dead, and you'd never have known from the display.
Calibration & bump test schedule
- Bump test before each day's use (ISEA position) — proves the sensor still responds and the alarm works. This is your primary protection against a silent dead sensor. In RAE fleets this is usually automated with an AutoRAE 2 docking station and a RAE 4-gas calibration mix (CH₄/O₂/CO/H₂S in one cylinder); exotic toxics (Cl₂, NH₃, HCN…) typically need their own dedicated cylinders. A docked pass still doesn't check the inlet, filter, or pump path — inspect those by hand.
- Full calibration per the manufacturer/AHJ — commonly monthly, extendable with a documented daily-bump program if the fleet stays in tolerance.
- Calibrate immediately after a failed bump, a drop/shock, exposure to high concentrations or poisons, sensor replacement, or erratic readings.
- Allow recovery time after a saturating exposure before you trust readings again, and re-bump.
- Track sensor age; replace O₂ (~1–2 yr) and toxics (~2–3 yr) on schedule even if still passing, because they're near the edge.
Field care & storage
- Store in moderate temperature and humidity — not a baking apparatus bay or a freezing compartment. Extreme dry heat dries the electrolyte; extreme cold slows it.
- Keep sensors away from solvent vapors and silicone products in storage (some off-gas and can affect sensors).
- Charge and log the instrument on a rotation; a monitor that's been dead in a drawer for months may have aged-out sensors even if "new."
- Protect the sensor membrane from water, dust, and physical damage; use inlet filters and dust guards.
- Let cold instruments acclimate before zeroing/bumping.
Common rookie mistakes
- Trusting a 0 ppm from a sensor you never bumped — it may be dead, not clean.
- Calling a cross-sensitive reading a confirmed chemical (H₂ read as "CO"; H₂S read as "CO").
- Ignoring an oxygen change at altitude and chasing a false O₂-deficiency alarm — or missing a real one.
- Not waiting for recovery time after a big gas hit, then believing the temporarily-suppressed reading.
- Assuming your meter covers a toxic it has no sensor for (no PH₃ channel = zero phosphine information).
- Storing the instrument somewhere hot/dry and wondering why sensors keep failing bumps.
Representative instruments
Electrochemical sensors live inside essentially every portable multi-gas monitor. In a RAE fleet: the MultiRAE family (pumped multi-gas with swappable EC sensor slots plus PID/LEL), the QRAE 3 (4-gas diffusion monitor for O₂/CO/H₂S + LEL), the ToxiRAE Pro (single-gas personal monitor available with O₂ or a range of toxic EC sensors), and the AreaRAE (wireless area monitor carrying the same sensor types for perimeter work). The MultiRAE's swappable slots let one platform be tailored with "exotic" toxics (Cl₂, NH₃, HCN, PH₃, SO₂, NO₂) for the anticipated hazard. Comparable non-RAE platforms include the Dräger X-am series and Industrial Scientific Ventis/MX6. Brand names are illustrative — your department's specific model and its manual govern.
EC sensors are precise about the one gas they're built for — but they age out, they read the wrong gas via cross-sensitivity, and when they die they often just read zero. Bump daily, know your cross-sensitivity chart, and never let a single toxic channel make a tactical decision alone.
Next: the flammability sensors — Catalytic Bead LEL →