Gas detector selection starts with the gas list, not the brand. Across industrial guidance from 2024-2026, the consistent first step is to inventory every gas the workforce can realistically contact, then match a sensor technology and a form factor to that list [S1][S3].
Three sensor families cover roughly 90% of industrial duty: catalytic bead / pellistor for combustible gases, electrochemical cells for toxic gases and oxygen, and NDIR or PID for VOCs and refrigerants. Detector bodies are then split into portable single-gas, portable multi-gas (typically 4 or 5 sensors), handheld leak-checkers, and fixed-point wall or duct transmitters, each with different coverage and response-time profiles [S1][S2][S3].
Sensor technology versus target gas
Catalytic bead sensors remain the default for combustible gases from methane through hexane vapours, with alarm setpoints commonly anchored at 10% LEL of the target gas to leave a 90% safety margin before the lower explosive limit is reached [S4].
Electrochemical cells dominate toxic-gas detection for CO, H2S, NH3, and O2; their low power draw makes them the workhorse inside multi-gas portables such as the 4-gas and 5-gas formats widely specified for confined-space entry [S1][S3]. For fixed ammonia detection in cold storage and refrigeration plants, dedicated NH3 electrochemical heads are recommended rather than general combustible sensors, because ammonia toxicity thresholds sit well below its flammability range [S1].
For VOCs the choice is between PID (photoionisation) and NDIR. PID lamps at 10.6 eV detect a broad VOC spectrum down to low ppm levels; lamp energy selection (9.8, 10.6, 11.7 eV) trades sensitivity against compound coverage, and detection range must be compared against expected exposure rather than chosen by default [S5]. NDIR cells are more selective and poison-resistant, useful for CO2 and refrigerants like HFCs and PFCs where the gas is known and cross-interference must be minimised [S1].
Form factor: portable, handheld, or fixed
Portable single-gas wearables are sized for one dominant hazard, such as H2S in oil and gas or CO in steelwork; they trade sensor count for lighter weight, longer battery life, and a lower unit cost [S3]. Multi-gas portables accept up to five sensor slots, are the practical default for mixed-hazard sites, and increasingly ship with pre-entry check modes for confined-space work [S3].
Handheld leak-checkers are leak-location tools, not personal safety alarms: they sniff around flanges, valves, and fittings to find a release source, and they are not a substitute for a worn monitor [S3]. Fixed-point detectors cover continuously occupied or unmanned areas where a portable workforce cannot be relied on, and key installation considerations include sensor height relative to gas density, airflow patterns, and accessibility for calibration [S1][S8].
Selection criteria that drive the shortlist
Five criteria consistently appear in current selection guidance: hazard list, sensor reliability, alarm clarity, battery and maintenance burden, and intrinsically safe certification for the zone [S1][S2][S3]. On reliability, dual-redundant sensor architectures such as Industrial Scientific's DualSense Technology use two independent sensors for the same gas, which removes the single-point sensor-failure mode that drives most instrument-related incidents [S2].
Alarm design should pair a high-decibel sounder with bright LEDs and a programmable action message such as "ventilate" or "evacuate," so the user's next step is unambiguous under stress [S2]. Maintenance is the hidden cost: docking stations automate bump tests, calibration, and firmware updates, and any shortlist should be filtered on whether the vendor's accessories cover the fleet size, not just the device [S2][S9].
When a 2-in-1 or 3-in-1 detector is the wrong answer
Combined natural-gas and propane alarms are a common failure mode in the residential and light-commercial market, and the failure mechanism is physical: methane is lighter than air and pools near the ceiling, while propane is heavier and pools at floor level, so a single sensor location cannot cover both [S4]. The same logic applies to CO versus natural-gas combination units, which generally do not detect both accurately and give occupants false confidence [S4].
For industrial buyers the equivalent trap is overloading a 4-gas monitor with sensors that fight each other, such as placing a low-range H2S cell next to a high-concentration CO cell in a ventilation shaft, or specifying one PID lamp energy for a process that streams both aromatics and chlorinated solvents [S3][S5]. If the hazard list is genuinely mixed, separate fixed detectors with different ranges are usually a safer and more maintainable answer than a single overloaded portable [S8].
Comparison: main detector types against four buyer criteria
Single-gas portable versus multi-gas portable versus fixed-point transmitter line up against coverage, response, cost per gas channel, and maintenance burden. Single-gas wins on cost per channel and battery life but loses on coverage in mixed-hazard zones; multi-gas portables are the best general-purpose trade for maintenance and turnaround work but require docking infrastructure to stay compliant; fixed-point systems win on continuous unmanned coverage and integration with building management, but lose on flexibility and have higher installed cost per point [S2][S3][S8].
Within PIDs, the trade is lamp energy against detection range: a 10.6 eV lamp covers most common VOCs to single-ppm resolution, while an 11.7 eV lamp extends to harder-to-ionise compounds but shortens lamp life and increases false positives from humidity [S5]. Buyers should size PID range to the highest expected exposure, not the lowest detectable concentration, to avoid saturating the sensor on a release [S5].
Real-world applications and the duty each one demands
Cold storage and ammonia refrigeration need NH3-specific electrochemical detectors with alarm setpoints tied to occupational exposure limits well below flammability, plus fixed-point coverage in the compressor room because leaks often occur when the plant is unmanned [S1]. Wastewater plants need H2S single-gas wearables for routine inspection and 4-gas multi-gas monitors for confined-space entry, with explicit pre-entry testing protocols [S3].
Oil and gas upstream operations carry the widest mix: combustible gases from wellheads, H2S from sour crude, and oxygen-deficiency risk in confined vessels, so a 5-gas monitor with LEL, O2, CO, H2S, and a fifth toxic slot is the typical minimum specification, with PID added whenever benzene or other VOCs are credible [S1][S3]. Pharmaceutical and chemical batch reactors where solvent leaks are credible add PID plus a fixed array of NDIR or electrochemical sensors around the reactor bay [S5].
Calibration, bump testing, and lifecycle cost
Every detector in the field requires regular calibration and a bump test before each shift, and skipping either is the single most common root cause of "the alarm did not go off" investigations [S9]. The buying decision should price the docking station, test gas, and consumables across the fleet life, not just the unit hardware, because over a 5-year lifecycle the accessories routinely exceed the hardware cost on dense fleets [S2][S9].
Sensors themselves have a finite service life, typically 2-3 years for most electrochemical cells, after which accuracy drifts even with calibration; the spec sheet should publish the expected sensor life and the replacement cost, not just the warranty [S9]. Specifiers who want a full cross-reference for fixed-instrument lifecycle and HVAC integration can read the fixed gas detector installation considerations reference page and the broader industrial equipment selection context for siting logic. For comparator decisions on adjacent equipment such as lighting upgrades in the same facility, the lamps and light fittings sizing reference and the lighting equipment and electric lamps spec page cover rated life and lumen-maintenance trade-offs that pair with detector siting in the same control rooms.
Trackable signals to watch after specification
Two signals confirm the shortlist is working: first, bump-test pass rate stays above 99% across the fleet after the first 90 days, which indicates the chosen sensor mix is not drifting or poisoning in the actual process atmosphere [S2][S9]. Second, the ratio of nuisance alarms to real actionable alarms trends down over the first quarter, which is the cleanest indicator that alarm setpoints and sensor ranges were sized to the real exposure profile rather than copied from a generic default [S2][S5].
For buyers comparing fixed-system architectures, the hydraulic and mechanical system spec reference and the linear motion spec reference sit alongside detector selection when planning the same plant upgrade, and a cross-discipline maintenance study like rough terrain forklift failure modes during shutdowns applies the same failure-mode mapping logic that detector shortlists should be built on. If the new detector fleet will be deployed alongside an optical or vision system, the vision light source sizing and selection guide follows the same hazard-first methodology used here.