How to Read a 4-Gas Monitor: O2, LEL, CO and H2S Readings Explained

How do you read a 4-gas monitor?

Short answer: A 4-gas monitor shows four channels at once, in three different units — oxygen as a percentage by volume, combustible gas as a percentage of its lower explosive limit (%LEL), and carbon monoxide and hydrogen sulfide in parts per million (ppm). In clean air the display should read close to O₂ 20.9%, LEL 0%, CO 0 ppm, H₂S 0 ppm. Anything else is the instrument telling you the atmosphere has changed, and the unit on each channel is what tells you how alarmed to be.

Learning to read a 4-gas monitor is mostly learning that the four numbers are not comparable to each other. Reading 10 on the LEL channel and reading 10 on the CO channel describe completely different situations — one is a fraction of an explosive mixture, the other is a concentration measured in millionths. This guide walks the display field by field: what each channel measures, what a normal reading looks like, what the LOW, HIGH, TWA, STEL and PEAK labels mean when they appear, and which hazards the four channels cannot see at all. The numeric thresholds referenced throughout come from OSHA's permit-required confined spaces standard at 29 CFR 1910.146 and the exposure limits in 29 CFR 1910.1000.

It is written for the people who actually hold the instrument — confined-space entrants and attendants, utility and wastewater crews, oil and gas field staff, and the safety managers who have to explain a datalog after the fact. If you are still choosing an instrument rather than reading one, start with the best 4-gas monitor guide or the full 4-gas monitors shelf.

Why this matters.
A 4-gas monitor is a decision instrument, not a dashboard. Under 29 CFR 1910.146 a permit space may only be entered when the atmosphere is inside stated limits — oxygen at or above 19.5% and at or below 23.5%, flammable gas below 10% of its lower explosive limit, and any toxic contaminant below its permissible exposure limit. Those are numbers a worker has to read off a display correctly, under time pressure, often through a faceshield. Misreading %LEL as a percentage of gas in air is the single most common interpretation error, and it understates the hazard by more than an order of magnitude.

Part 1 — The four channels and the three units

Every standard 4-gas instrument reads the same four things. What changes between models is the housing, the pump, the datalogging and the sensor service model — not the channels. The units are the first thing to internalise, because they are not interchangeable.

Why three different units on one screen

Each unit is chosen to match the size of the hazard. Oxygen is roughly a fifth of the air, so a percentage is the natural scale — a change of one whole percentage point is significant. Carbon monoxide and hydrogen sulfide become dangerous at concentrations far too small to express as a percentage; hydrogen sulfide is life-threatening at around 0.01% of the atmosphere, which is why the toxic channels use parts per million instead. Combustible gas gets its own relative scale because the concentration that matters differs by gas, and %LEL normalises that into one number the wearer can act on.

Part 2 — What a normal 4-gas monitor reading looks like

The baseline is the reading the instrument gives in known-clean air, and it is worth memorising because every judgement you make is a comparison against it.

Where the baseline comes from

The instrument establishes its own zero when it powers up, and it assumes the air around it at that moment is clean. Powering on inside a truck cab, next to running plant, or over an open manhole bakes whatever is present into the baseline — the display will read 0 ppm in an atmosphere that is not at 0 ppm. Turn the instrument on in known-fresh air, outdoors and upwind of the work, and let it complete its warm-up before it goes anywhere near the space. The daily function check that confirms the sensors still respond is covered in bump test vs calibration for gas detectors.

Part 3 — Reading the oxygen channel

Oxygen is the channel people find most intuitive and misread most often, because the dangerous direction is both ways.

• 20.9% — the normal atmospheric value and the number the sensor is calibrated against.
• Below 19.5% — oxygen-deficient under 1910.146. This is an entry-prohibiting condition, not a caution.
• Above 23.5% — oxygen-enriched, also entry-prohibiting. Enrichment dramatically increases fire risk; materials that merely smoulder in normal air can burn fiercely.

The subtlety is what a small oxygen drop implies. Oxygen does not usually disappear on its own — it is displaced by something else or consumed by something. A reading of 20.1% means roughly 4% of the atmosphere in front of the sensor is now something that was not there before. If the LEL channel is still reading zero, that displacing gas is not a combustible the sensor recognises: it could be nitrogen, carbon dioxide, argon or a purge gas, none of which the standard four channels identify. A falling oxygen reading with a silent LEL channel is one of the strongest arguments for bringing a CO2 detector or a dedicated oxygen detector into the assessment.

Which oxygen readings are entry-prohibiting: 19.5% and 23.5%

Under 29 CFR 1910.146 a permit space atmosphere is acceptable only when oxygen is at or above 19.5% and at or below 23.5% by volume. Both are hard limits and both prohibit entry when crossed. Deficiency below 19.5% is the failure most people anticipate; enrichment above 23.5% is the one that gets underestimated, because an oxygen-rich atmosphere makes materials ignite far more readily and burn far more fiercely than the same materials in normal air. Neither figure is a target to work toward — a reading drifting toward either is a reason to ventilate and re-test.

Why oxygen is read first

OSHA's testing sequence for permit spaces is oxygen, then combustible gases and vapours, then toxics. The order is not arbitrary. The catalytic bead sensor used on most LEL channels burns the sample gas on a heated element, and that reaction needs oxygen to proceed — in an oxygen-deficient atmosphere the combustible channel can under-read or fail to respond at all. Reading oxygen first tells you whether the LEL number underneath it can be trusted.

Part 4 — Reading the %LEL channel (the one that gets misread)

The lower explosive limit of a gas is the lowest concentration in air that will support ignition. For methane that is about 5% by volume. Below the LEL the mixture is too lean to burn; above it, it will.

The combustible channel does not display the gas concentration. It displays what percentage of the way to the LEL the atmosphere has travelled. So on a monitor configured for methane:

This is why a display showing 10 on the LEL channel is a serious reading and a display showing 10 ppm of CO is not an emergency. It is also why alarm set points are conventionally placed at 10% and 20% LEL — those trip at a tenth and a fifth of an ignitable mixture, leaving margin to evacuate. The %LEL scale is relative to a threshold; the ppm and %volume scales are absolute measures of how much gas is present. Mixing the two up is what makes a 10 on one channel read as equivalent to a 10 on another.

The calibration gas caveat

An LEL channel is calibrated against one specific gas — commonly methane, sometimes pentane or propane depending on the industry. Faced with a different combustible, the sensor still responds, but the number it reports is scaled to the calibration gas rather than to what is actually present, and it can read high or low. If your hazard is not the gas the instrument was calibrated on, the manufacturer's correlation factor for that gas has to be applied before the reading means what it appears to mean. Calibration gas selection is covered on the gas detector calibration gas and accessories shelf.

Part 5 — Reading the CO and H₂S channels in ppm

One part per million is one volume of gas in a million volumes of air. The toxic channels use it because the numbers that matter are small.

The two rows disagree with each other on purpose, and that disagreement is the most important thing on this page for anyone setting alarms. The OSHA permissible exposure limit is the enforceable federal value. The NIOSH recommended exposure limit and the ACGIH threshold limit value are professional recommendations, are frequently lower, and carry no force of law on their own. For hydrogen sulfide the gap is twentyfold. Which number your instrument alarms at is a decision your safety program makes, not something the standard settles — the distinction is unpacked in ACGIH TLV vs OSHA PEL.

Do not use smell as a cross-check on H₂S

Hydrogen sulfide is detectable by smell at concentrations far below the exposure limit, which invites workers to treat their nose as a backup sensor. It is not one. At higher concentrations H₂S rapidly deadens the sense of smell, so the odour fades exactly as the hazard becomes acute — a documented characteristic described in OSHA's hydrogen sulfide hazard guidance. A rising then vanishing smell is a reason to leave, not a reason to relax. Dedicated single-gas options are on the hydrogen sulfide detectors shelf and compared in the best H2S monitor guide.

Part 6 — LOW, HIGH, TWA, STEL and PEAK: the five labels

When a 4-gas monitor annunciates, a word usually appears alongside the number. Each one describes a different kind of exposure, and they call for different responses.

TWA and STEL are the two that surprise people. Both are computed by the instrument from its own logged history, so a TWA alarm can annunciate in air that is currently clean — it is reporting where the shift has been, not where it is. That is also why leaving a personal monitor running through breaks and transit matters: the average is only meaningful if the instrument saw the whole shift. Wearable options are on the personal gas detectors shelf and compared in the best personal gas detector guide.

What TWA (time-weighted average) means on a gas monitor

Time-weighted average — the instrument's running average concentration for that gas over an 8-hour reference period, calculated from its own logged readings. It exists because some harm depends on total dose across a shift rather than on any single peak. A TWA alarm therefore reports where the shift has been, not where the atmosphere is now, and it can annunciate while the live reading sits at zero. It is only meaningful if the instrument saw the whole shift, which is the practical argument for leaving a personal monitor running through breaks and transit.

What STEL (short-term exposure limit) means on a gas monitor

Short-term exposure limit — the average concentration over a 15-minute window. It catches a sustained burst that would be diluted to invisibility in an 8-hour figure: fifteen minutes at a damaging concentration barely moves the TWA but is exactly the exposure a STEL is defined to flag. An instrument can therefore alarm on STEL while its TWA remains well inside limits, and that combination is a real finding rather than a contradiction.

Instrument alarm setpoint is not an exposure limit

These are four different things and they are routinely spoken about as one:

• OSHA PEL — the enforceable federal limit.
• NIOSH REL and ACGIH TLV — professional recommendations, frequently lower than the PEL, carrying no force of law on their own.
• IDLH — the concentration immediately dangerous to life or health; an escape threshold, not a working limit.
• The instrument's alarm setpoint — a number your safety program chooses, which may sit at or below any of the above.

For hydrogen sulfide the gap is twentyfold: a 20 ppm OSHA ceiling against a 1 ppm ACGIH TLV. An instrument set to alarm at the PEL and one set to alarm at the TLV are both defensible and will behave completely differently on the same job. What is not defensible is treating the four as interchangeable, or raising a setpoint to stop nuisance alarms. The general comparison of the exposure-limit systems is covered in ACGIH TLVs vs OSHA PELs.

Part 7 — Why the monitor alarms when nothing seems wrong

An alarm with no obvious source is the most common reason a reading gets doubted, and doubting a reading is how people get hurt. The honest answer is that several of these have benign explanations and none of them can be diagnosed from inside the space.

• A real source you have not found yet. The default assumption. Exhaust drifting from plant running upwind, a vehicle idling near an intake, hot work in an adjacent bay.
• A zero taken in air that was not clean. The baseline is wrong, so every subsequent reading is offset.
• Another gas responding on a sensor that was not meant for it. Electrochemical sensors are not perfectly selective; hydrogen in particular is a well-documented responder on carbon monoxide cells. This is cross-sensitivity, and the interference table for your specific sensor and model is published by its manufacturer.
• A sensor at end of life. Cells lose sensitivity with age and exposure and can behave erratically before they fail outright.
• Rapid environmental change. Moving a cold instrument into a warm humid space can disturb readings until it equilibrates.

The response is the same in every case: treat the alarm as real, leave, and diagnose from clean air. An instrument that alarms without an explanation you have confirmed is an instrument that goes for a bump test before it goes back to work.

Part 8 — What the four channels do not cover

A 4-gas monitor covers the atmospheric hazard classes a confined-space entry has to rule out — oxygen level, flammability, and the two toxics most commonly encountered in industry. It does not cover everything a space can hold, and the sensor set a given space requires depends on its hazard assessment rather than on convention.

Contaminants a standard O₂/LEL/CO/H₂S instrument will not report include volatile organic compounds, carbon dioxide, ammonia, chlorine, sulfur dioxide, nitrogen dioxide and hydrogen cyanide. Several of these will not move any of the four channels at all; carbon dioxide and nitrogen show up only indirectly, as a falling oxygen number. Where the assessment names one of them, the answer is an added sensor or a second instrument — a VOC detector for solvent vapours, or the relevant single-gas unit from the gas detector hub. The trade-off between one multi-gas instrument and dedicated single-gas units is worked through in 4-gas monitor vs single-gas detector.

Part 9 — Worked example: reading a 4-gas monitor before a confined-space entry

To make the sequence concrete, here is how the display is read on a pre-entry check of a below-grade vault, using an instrument like the Forensics 4 Gas Meter for the survey and a wearable such as the Honeywell BW Clip4 4-Gas Detector on the entrant:

1. Power up in clean air and confirm the baseline. Outdoors, upwind of the work. Let the warm-up finish and confirm the display settles at O₂ 20.9, LEL 0, CO 0, H₂S 0. A baseline that will not settle is a bump test, not an entry.
2. Bump test and confirm all four channels respond. Apply known gas and watch each channel move and each alarm annunciate. A channel that does not respond is a failed check — see bump test vs calibration for gas detectors.
3. Sample remotely, before anyone leans in. With a pump and sample line the atmosphere is drawn to you rather than you going to it. Allow the full response time for the length of tubing in use — the pump-versus-diffusion trade-off is covered in diffusion vs pump gas detector.
4. Read the channels in order: oxygen, then combustible, then toxic. Oxygen first because the LEL channel depends on it. Confirm oxygen is between 19.5% and 23.5%, LEL is below 10%, and both toxic channels sit below the limits your program has adopted.
5. Sample the full depth, not just the opening. Lower the probe in stages and read at each. Stratification depends on how a gas was released, on temperature and on ventilation, so the reading at the lip tells you very little about the reading at the floor.
6. Keep a monitor on the entrant for the duration. The pre-entry survey describes one moment. Conditions change, so a personal instrument stays in the breathing zone throughout, per the requirements summarised in permit-required confined space requirements.

The same reading discipline applies across the rest of the range we cover — the portable gas detectors used for survey work and the area gas monitors that cover a perimeter all present the same four channels in the same units. Supporting entry equipment is on the confined space equipment shelf.

Frequently asked questions

What should a 4-gas monitor read in normal air?

Oxygen 20.9% by volume, LEL 0%, carbon monoxide 0 ppm and hydrogen sulfide 0 ppm. Small oxygen variation of a tenth or two either side of 20.9 is normal sensor tolerance. Any sustained reading away from that baseline means the instrument is seeing something.

What does LEL mean on a gas monitor?

LEL is the lower explosive limit — the leanest mixture of a combustible gas in air that will ignite. The channel displays the percentage of that limit the atmosphere has reached, so 10% LEL means one tenth of the way to an ignitable mixture, not 10% gas.

Is 10% LEL the same as 10% gas in air?

No, and this is the most consequential misreading on the display. For methane, whose LEL is about 5% by volume, a 10% LEL reading corresponds to roughly 0.5% methane in air — a tenth of the concentration the number appears to suggest.

What does ppm mean on a 4-gas monitor?

Parts per million — one volume of the gas in a million volumes of air. Carbon monoxide and hydrogen sulfide use it because they are hazardous at concentrations far too small to express usefully as a percentage.

Why does my monitor show oxygen at 20.9%?

Because that is oxygen's share of normal dry air, and it is the value the sensor is calibrated against. Treat it as the reference point: a reading below it means oxygen is being displaced or consumed, and a reading above it means the atmosphere is being enriched.

What oxygen reading is dangerous on a gas monitor?

Under 29 CFR 1910.146 an atmosphere below 19.5% oxygen is oxygen-deficient and one above 23.5% is oxygen-enriched; both prohibit entry. Enrichment is the one people underestimate — it sharply increases how readily materials ignite and burn.

What does TWA mean on a gas detector?

Time-weighted average — the instrument's running 8-hour average for that gas, computed from its own logged readings. A TWA alarm reports accumulated exposure across the shift and can trigger when the current live reading looks unremarkable.

What is the difference between TWA and STEL?

TWA averages exposure across a full 8-hour shift; STEL averages it across 15 minutes. STEL catches a sustained burst that would be diluted to invisibility in the 8-hour figure, which is why an instrument can annunciate STEL while its TWA remains well inside limits.

What does PEAK mean on a 4-gas monitor?

PEAK recalls the highest concentration the instrument has seen since it was last reset. It is a memory display rather than a live alarm, and it is the field most worth reading after an incident to reconstruct what the atmosphere actually did.

What is the difference between a LOW and a HIGH alarm?

They are two instantaneous set points on the same channel. LOW is the early warning that conditions are moving; HIGH is the second threshold, and under essentially every written program it means evacuate rather than silence and continue.

Why is my gas monitor alarming when I cannot smell anything?

Because smell is not a detection method. Carbon monoxide has no odour at all, and hydrogen sulfide deadens the sense of smell as concentrations climb, so the odour can fade exactly as the hazard becomes acute. Treat the instrument as correct and leave.

Why does my CO channel read when there is no carbon monoxide?

Most often a real source you have not located, a zero taken in air that was not clean, or another gas responding on the carbon monoxide cell — hydrogen is a well-documented responder. Check the cross-sensitivity table your sensor manufacturer publishes for that specific cell, and treat the alarm as real until you have confirmed otherwise from clean air.

Which gas should be tested first in a confined space?

Oxygen, then combustible gases and vapours, then toxics. The catalytic bead sensor on most LEL channels needs oxygen to react, so an oxygen-deficient atmosphere can make the combustible reading unreliable — reading oxygen first tells you whether to trust the number below it.

What gases does a 4-gas monitor not detect?

Volatile organic compounds, carbon dioxide, ammonia, chlorine, sulfur dioxide, nitrogen dioxide and hydrogen cyanide, among others. Carbon dioxide and nitrogen appear only indirectly, as a falling oxygen reading. Where the hazard assessment names one of these, it needs its own sensor.

Can I enter a space if the readings are inside limits?

Inside limits is one condition among several, not permission on its own. A permit space still requires the rest of the entry program — permit, attendant, ventilation, rescue provision and continuous monitoring — as summarised in permit-required confined space requirements.

Do I need to sample at more than one depth?

Yes where stratification is possible. Gases distribute according to how they were released, temperature and ventilation as well as density, so a reading at the opening can differ substantially from one at the floor. Lower the probe in stages and read at each.

What should I do if a channel reads negative or shows dashes?

Do not enter on it. A negative value or a dashed field usually means a faulted sensor, an incomplete warm-up, or a zero captured in contaminated air. Return to clean air, re-zero, and bump test before the instrument goes back into service.

What does STEL mean on a gas detector?

Short-term exposure limit - the instrument's average reading over a 15-minute window. It flags a sustained burst that would be diluted to invisibility in the 8-hour time-weighted average, so a monitor can alarm on STEL while its TWA stays well inside limits.

Is a gas detector alarm setpoint the same as the OSHA PEL?

No. The OSHA PEL is the enforceable limit, the NIOSH REL and ACGIH TLV are professional recommendations that are often lower, IDLH is an escape threshold, and the alarm setpoint is a number your safety program chooses. For hydrogen sulfide the OSHA ceiling is 20 ppm against a 1 ppm ACGIH TLV, so the choice materially changes instrument behaviour.

Further reading on this site

• Bump test vs calibration for gas detectors — the daily function check that makes any reading on this page trustworthy.
• Best 4-gas monitor guide — the buyer's-guide companion, with picks by use case and budget.
• 4-gas monitors — the full shelf of O₂/LEL/CO/H₂S instruments we cover.
• Personal gas detectors — wearable clips that track TWA and STEL in the breathing zone.
• Gas detection complete guide — the wider category overview covering sensor types and instrument formats.
• Combustible gas detector vs gas leak detector — why a %LEL instrument and a leak-pinpointing sniffer are not interchangeable.
• Gas detector calibration gas and accessories — cylinders, regulators and tubing for bump testing and calibration.