Flammable Gas Detection
Technologies and what each can and cannot see.
13 min
The measurement
Flammable gas is measured as a percentage of the lower explosive limit — the concentration below which the mixture is too lean to burn. Typical alarm settings are around twenty per cent LEL for a first alarm and forty to sixty per cent for executive action, giving warning well before a flammable atmosphere exists.
Catalytic bead (pellistor)
A heated catalytic bead oxidises the gas, raising its temperature and its resistance, measured against a compensating reference bead. Long established, responds to almost any flammable gas, and inexpensive.
Its limitations are important and often forgotten:
- It requires oxygen to work, so it cannot detect gas in an inert or oxygen-deficient atmosphere or inside a purged enclosure.
- It can be poisoned by silicones, sulphur compounds, lead and halogens, losing sensitivity permanently and — critically — without any indication. A poisoned pellistor reads zero in gas.
- High gas concentrations can burn out the bead.
- It requires regular gas testing because there is no way to confirm sensitivity otherwise.
Infrared point detectors
Hydrocarbons absorb infrared at characteristic wavelengths. The detector compares an active wavelength with a reference wavelength, so dirt and drift affect both equally and largely cancel.
Advantages are substantial: no oxygen required, immune to poisoning, unaffected by high concentrations, fail-safe behaviour because loss of signal is detectable, and a longer life with less maintenance. They are now the default for hydrocarbon gas detection.
The limitation is that infrared detects only gases that absorb at the chosen wavelength — hydrocarbons. It does not detect hydrogen, which has no infrared absorption in the usable band. A facility with hydrogen needs catalytic or dedicated hydrogen detection, and this is a recurring gap on plants that standardised on infrared.
Open path detectors
An infrared beam across an open distance between a transmitter and receiver, measuring the total gas in the path. It covers a line rather than a point, which suits perimeters, along pipe racks, and around the boundary of a process module where a point detector would have to be lucky.
The reading is in LEL-metres, which is a path-integrated quantity rather than a concentration — a small dense cloud and a large diffuse one can give the same reading. Alignment, fogging, heavy rain, blocked line of sight and structural movement all affect it, and it must be part of a design that also uses point detectors.
Other technologies
- Ultrasonic gas leak detectors — detect the ultrasound of a pressurised gas release rather than the gas itself. They respond within milliseconds, independent of wind direction and ventilation, which is exactly the weakness of every concentration-based detector in an open, windy offshore or desert environment. They do not confirm that gas has reached anywhere, so they complement rather than replace gas detection.
- Laser and tunable diode detectors — very high selectivity for specific gases.
- Acoustic and camera-based leak detection — increasingly used for early detection and for locating leaks.