Portable gas detector sizing collapses to a four-axis decision: target gas set, sensor technology, sampling method, and certification envelope, and most spec errors trace back to skipping the first or last axis [S1][S3].
The unit count is split across single-gas, dual-gas, and 4-in-1 (sometimes 5-gas) form factors; sensor packages combine electrochemical (EC), catalytic bead (pellistor), non-dispersive infrared (NDIR), and photoionization detector (PID) cells, each with a distinct response curve and interference profile [S3][S4]. IP rating, ATEX/IECEx zone, and run-time on a single charge are the second-tier filters that decide whether a shortlisted model survives plant acceptance testing.
Single-Gas vs Dual-Gas vs 4-in-1 Multi-Gas: Selection by Hazard Set
Single-gas diffusion monitors (H2S, CO, O2, SO2, ClO2) dominate personal-protection roles because EC cells deliver sub-30-second T90 response, weight stays under 150 g, and unit cost runs a fraction of a multi-gas rig [S4][S5]. A 2-year disposable format (no calibration, no sensor swap, no battery charge) is now standard for H2S in oil and gas, where the worker only needs the one target gas [S5].
Dual-gas units cover the CO+H2S pair that wastewater and steel-plant crews face daily; they share the same diffusion EC architecture but add a second cell on the same PCB, holding weight near 200 g [S4].
4-in-1 (and 5-gas) portables combine LEL (catalytic bead or NDIR), O2, CO, and H2S in a single pumped or diffusion body; the GC310-class 4-in-1 ships IP67-rated, weighs 200-350 g, and targets confined-space entry where the atmosphere is unknown and could swing on any of the four axes [S3][S4]. Sizing rule: if any two of LEL, O2, CO, H2S are unknowns in the worker's breathing zone, the 4-in-1 is the only defensible pick.
Sensor Technology Match: EC, Catalytic Bead, NDIR, PID
Electrochemical cells are the default for toxic gases (H2S, CO, SO2, ClO2, NO2) and O2; they draw microamps, run 1-3 years, and are poisoned by prolonged high-gas exposure, which is why bump-testing before each shift is the universal discipline [S3][S4].
Catalytic bead (pellistor) sensors detect combustible gases as a percentage of LEL, but they require at least 10% O2 to oxidize the target gas, fail in inert or oxygen-depleted atmospheres, and are poisoned by silicone, lead, and sulfur compounds [S3].
NDIR sensors (CH4, CO2, hydrocarbons) sidestep the O2 requirement and survive silicone/lead poisoning, making them the right call for landfill, biogas, and CO2 storage work where a pellistor would zero-shift within weeks [S3].
PID (photoionization) cells detect VOCs at ppm levels and are the only practical option for benzene, toluene, isocyanates, and solvent cleanup; lamp energy (9.8 eV, 10.6 eV, 11.7 eV) is the spec axis that decides which VOC subset the unit can see [S3].
Diffusion vs Pumped Sampling: When the Pump Is Mandatory
Diffusion sampling relies on natural airflow across the sensor face; it is lighter, cheaper, and the format of choice for clip-on personal monitors worn in the breathing zone throughout a shift [S4][S5].
Pumped sampling draws air through a hose and water/dust filter, lets the worker sample a tank head, manhole, or wall cavity before entry, and is required by most confined-space-entry procedures, where the OSHA-style "test the atmosphere before entry" rule physically cannot be met by a diffusion unit [S3][S7].
Response lag is the trade-off: a 1 m sample hose adds 3-10 seconds of transport delay on top of sensor T90, so the instrument reading lags the true concentration during a rapid release, a failure mode that diffusion-on-a-clip avoids but diffusion-on-a-rope cannot [S3].
Detection Range, Accuracy, and Calibration Interval
Portable units carry a narrower detection range and lower sensitivity than fixed units because the sensor cell and optics are physically smaller, so a portable rated 0-100 %LEL should not be substituted for a fixed 0-100% v/v combustible sensor [S2].
Accuracy is typically quoted as ±5% of reading (or ±10% of LEL for catalytic bead) at calibration, and that figure drifts between bump tests; the industry standard is a bump test before each shift and a full calibration at 30- or 90-day intervals depending on sensor and manufacturer [S3][S5].
Method 21 fugitive-emission monitoring uses a portable FID or PID pumped at 0.5-2.0 L/min against a 500 ppm or 10,000 ppm methane-equivalent reference standard, which is why the screening instrument and the quantitation instrument are specified separately, not as a single unit [S7].
Certification, Ingress, and Run-Time
ATEX/IECEx zone rating, IP rating, and battery endurance are the three pass/fail filters that the hazard analysis cannot negotiate. Group I (mining) and Group II (surface) categories, plus the Ex d (flameproof) or Ex ia (intrinsically safe) protection concept, are written into the procurement spec, not left to the vendor [S3][S4].
IP67 is the common industrial floor for wash-down and rain exposure, and most 4-in-1 portables ship at that rating [S4]. Run-time on a single charge is 8-16 hours for pumped multi-gas units and 2-3 years for 2-year disposable single-gas H2S monitors, so the duty cycle (8 h shift, 12 h shift, multi-day turnaround) decides which architecture survives the cost-per-shift math [S5].
Who Should Not Buy the Mainstream 4-in-1
For benzene-only refinery turnarounds, a 4-in-1 is the wrong pick because catalytic bead, EC, and NDIR cells are blind to BTEX at regulatory thresholds; the right tool is a dedicated PID with a 9.8 eV or 10.6 eV lamp, even at 2-3x the unit cost [S3].
For CO2 storage, beverage dispense, or dry-ice handling, an NDIR-only CO2 portable is correct because the O2 and LEL readings on a 4-in-1 add cost and weight without information; the CO2 hazard is real, the explosion hazard is not [S2]. For pure H2S field work at oil and gas wells, the 2-year disposable single-gas H2S monitor is the right call because it removes calibration labour, sensor-replacement inventory, and charging dock cost from the safety budget, a pattern documented across downstream operators [S5].
If the application is fixed-area perimeter monitoring of a compressor skid or battery room, a fixed gas detector with 4-20 mA or relay output is the correct architecture, and a portable is a complement, not a substitute, for routine worker entry [S2][S3].
Comparison Table: Portable Detector Classes on Decision Criteria
The four common portable classes line up against duty, sampling, sensor mix, and entry-test suitability: single-gas disposable (diffusion, one EC cell, 2-year life, suitable for ambient personal H2S or CO only); dual-gas personal (diffusion, two EC cells, 8-24 month life, suitable for CO+H2S or CO+O2 personal monitoring); 4-in-1 multi-gas (pumped or diffusion, EC + catalytic bead or NDIR, 8-16 h rechargeable, suitable for confined-space entry and refinery turnaround); dedicated PID or NDIR (pumped, single technology, suitable for VOC or CO2-only duty) [S1][S3][S4][S7].
The decision pivot: choose by the worst-case gas in the breathing zone, not the most common one. If the worst-case is an unknown atmosphere inside a vessel, a pumped 4-in-1 with LEL, O2, CO, H2S is the minimum defensible spec; if the worst-case is benzene at 1 ppm, no multi-gas rig clears the requirement, regardless of sensor count [S3][S7]. For plant-wide architectures, a portable gas detector fleet sized against shift count and bump-test throughput sits inside a broader instrument-and-control procurement plan, and engineers specifying adjacent systems (from valves to alarms) benefit from cross-referencing selection logic such as the gas alarm controller sizing guide and the pinch valve selection criteria for the chemical-plant envelope.
Trackable next signals: bump-test gas cylinder consumption per shift (proxy for instrument count and calibration discipline), increased specification of NDIR LEL sensors on new 4-in-1 tenders (proxy for silicone-poisoning awareness in oil and gas), and the rise of 2-year disposable H2S units in shale and downstream turnaround crews as a measurable shift away from rechargeable personal monitors [S1][S3][S5].
Spec-level background on the components involved: linear guide, and crossed roller guide.