Confined space entry (CSE) on 2026-07-31 is still governed by a four-gas minimum: combustible gas (LEL), oxygen (O2), carbon monoxide (CO), and hydrogen sulphide (H2S) [S2][S3]. That stack, sold as a dedicated confined space portable gas detector, remains the default in sewage, refining, marine, and chemical-vessel work where crew must enter a permit-required space [S2].
Two sensor families dominate the 4-gas product class: a pellistor/catalytic bead for LEL, plus three electrochemical cells for O2, CO, and H2S — a configuration that MSA, Dräger, and Honeywell all build in their mainstream CSE models [S3][S5]. For toxic CSE work, the same form factor is also sold as a multi gas detector with up to 6–7 channels, swapping or adding sensors for SO2, Cl2, NH3, NO2, or IR-measured CO2 [S3][S5].
Minimum CSE Sensor Stack vs. Why Single-Gas Units Fail in CSE
Atmospheric hazards inside a permit-required confined space fall into three categories — oxygen deficiency/enrichment, combustible atmospheres, and toxic exposure — and each requires a dedicated sensing principle, which is why a single gas detector is the wrong tool for the job [S3]. A CO-only or H2S-only unit will not flag a methane LEL excursion or an O2-displaced atmosphere, and that is the most common fatal-cause pattern documented for CSE incidents [S3].
The minimum stack is LEL + O2 + CO + H2S, the exact four channels listed on a typical CSE datasheet [S2]. The LEL channel uses a catalytic-bead or pellistor sensor; O2, CO, and H2S use electrochemical cells; and an optional infrared (IR) channel can be substituted for the combustible channel in environments where silicone or hydrogen may poison a pellistor [S3][S5]. MSA's Altair 5X, for example, accepts IR sensors alongside the electrochemical toxic sensors to cover those edge cases [S5].
Dräger's X-am 8000 measures up to 7 gases including specialised methane in the ppm range, the X-am 5800 covers 6 channels with a SOS / motionless detection, and the X-am 2800 covers 4 channels and is drop-tested past 2 m on its own — three reference points for a CSE-spec performance bar [S3]. On the Honeywell side, BW fixed and portable lines (Sensepoint XCD for fixed, BW Clip / Micro for portable) sit in the same family but the Sensepoint XCD is a fixed gas detector and is not body-worn, so it is a perimeter tool, not a CSE entry tool [S4].
Intrinsic Safety, ATEX, and IECEx: What the Marking Must Show
A CSE body-worn unit is operated inside a classified zone, so the housing marking must include ATEX (EU) or IECEx (international) certification for Ex ia (Zone 0/1) or Ex d (Zone 1), with a corresponding equipment protection level (EPL) [S1][S3]. Chinese CSE-grade units in the 2026 catalogue explicitly advertise "ATEX and SIL2" certification on the LPG leak detector line, with manufacturer SGS and ISO third-party certification, demonstrating that ATEX is a baseline entry ticket, not a premium option [S1].
SIL2 (Safety Integrity Level 2 per IEC 61508) is the typical functional-safety layer below SIL3 and applies to the detector's alarm-and-logic chain; buyers should treat "ATEX + SIL2" as a matched pair, and confirm that the certificate numbers are listed on the manufacturer's ATEX/IECEx notified-body database rather than a marketing PDF [S1][S3]. A useful engineering rule: if the certificate PDF is older than the firmware version on the unit, the SIL claim is stale. Both Dräger's X-am 5800 and Honeywell's BW portables carry Bluetooth connectivity for live monitoring, which only matters if the matching cloud / gateway (e.g. Dräger Gas Detection Connect + ConHub) is rated for the same zone [S3].
Sample-Draw Pump vs. Diffusion: When the Pump Is Mandatory
For pre-entry testing, the device must draw sample from inside the space before a worker enters, which means a motorised sample-draw pump with hose and probe — not a diffusion-only body-worn unit — is the correct tool for the entry permit [S3]. Dräger's X-am 2800 explicitly lists an "optional pump" for the 4-gas compliance-monitor role, and the X-am 8000 is sold as a CSE unit specifically because pump-equipped multi-gas instruments dominate pre-entry measurement [S3].
A diffusion-only 4-gas unit has a place: it is the body-worn monitor that the entrant wears continuously after entry, sounding personal alarms if the atmosphere shifts while they are inside [S2]. A site running a proper CSE programme therefore needs two instruments per entry team: one pumped (pre-entry + standby at the hole), and one diffusion (on the entrant's lapel or harness). Trying to do both jobs with a single diffusion unit is the most common CSE equipment mistake.
For thicker CSE work in heavier confined spaces — petrochemical reactors, marine ballast, foundries — IR sensors for combustible gas and CO2 are preferred because pellistor LEL sensors can be poisoned or under-read in low-oxygen or silicone-laden atmospheres, and the same hardware accepts the IR swap [S3][S5]. This is also why "explosion-proof Ex d" is sometimes specified instead of Ex ia: a Zone 1 fixed monitor like the Sensepoint XCD is Ex d-rated and uses magnetic non-intrusive calibration (no hot work permit needed), which is operationally different from a body-worn Ex ia portable [S4].
Alarms, Bump Test, and Calibration Intervals
CSE detectors have three alarm layers: low, high, and TWA/STEL (time-weighted average / short-term exposure limit), with TWA being the rolling 8-hour toxic exposure metric and STEL typically a 15-minute window [S3]. A confined space unit without TWA/STEL computation is a survey tool, not an occupational-hygiene monitor, and the difference matters for hot-work permits in adjacent spaces.
Bump test (functional gas exposure) is the daily field check; full calibration (span adjustment to a certified gas concentration) is typically a 30–180 day interval depending on sensor and manufacturer [S3]. MSA's XCell sensors, for example, are marketed around extended calibration intervals and longer sensor life relative to older electrochemical generations, which lowers total cost of ownership in fleet use [S5]. A practical engineering rule is to bump-test before every CSE entry and run a documented full calibration at the manufacturer's stated interval or after any failed bump test.
Motionless / SOS detection — present on the Dräger X-am 5800 — is the dead-worker alarm: if the entrant stops moving, the device escalates an alarm without manual input [S3]. For lone-worker CSE in pits and rail tank cars, that feature is a specification line item, not a luxury, and should be specified alongside a combustible gas detector channel that responds to methane at ppm resolution for natural-gas-leak scenarios [S3].
Use-Case Match: Who Needs What
For municipal sewer and utility-vault entry, a pumped 4-gas (LEL/O2/CO/H2S) diffusion-when-needed portable is the standard ticket, with the diffusion mode worn by the entrant after pre-entry clearance [S2][S3]. For chemical-plant vessel entry, a 5- to 7-gas pumped monitor with IR LEL, O2, CO, H2S, and at least one additional toxic (SO2, NH3, or Cl2 depending on the last cargo) is closer to the minimum [S3][S5].
For welding inside a confined space, the portable gas detector selection for welding operations map overlaps heavily: CO and NO2 are the welding-specific toxics, so a CSE welding entry needs CO and NO2 channels on top of the standard four. For warehouse or general industry where CO is the dominant concern, the warehouse portable gas detector selection 4-gas vs single-gas spec map frames the same stack differently — single-gas CO is acceptable for forklift-area ambient monitoring but fails CSE.
For chemical plants the spec drifts to IR LEL and a wider toxic library; the portable gas detector spec map for chemical plants goes deeper on IR vs pellistor trade-offs. The right tool for a given entry is decided by the pre-entry risk assessment, not by the model on the shelf — but the shelf must contain at minimum an ATEX/IECEx 4-gas pumped portable with TWA/STEL and a motion/SOS alarm. Anything less is a paper compliance exercise.
Verifiable signals to track between now and the next quarterly review: ATEX/IECEx certificate dates on the actual unit (not the catalogue), bump-test gas cylinder expiry, and a documented TWA/STEL log per entrant. A CSE unit whose certificate is older than its firmware, or whose calibration gas expired last quarter, fails the same audit regardless of brand.