Gas detector calibration: bump tests, span gas and why sensors are consumables
A bump test proves the alarm works; a calibration sets what it reads. The difference matters because gas detection sensors degrade continuously and fail in a direction that does not announce itself.
Gas detection has a feature that distinguishes it from most calibration: the instrument is often the last line of defence for somebody's life, and it fails in a quiet direction. A detector reading low does not alarm, does not complain, and looks exactly like a detector in clean air.
Bump test and calibration are different things
| Bump test | Calibration | |
|---|---|---|
| What it does | Applies gas above the alarm point and confirms the alarm fires | Adjusts zero and span against certified gas |
| Question answered | Does the whole chain work? | Is what it reads correct? |
| Output | Pass or fail | As-found and as-left readings |
| Typical frequency | Before each day of use | Periodically, and whenever a bump test fails |
| Covers | Sensor, electronics, alarm, horn, lights, vibration | Sensor response magnitude |
The bump test is the more frequently valuable of the two, because it tests the whole chain including the parts a calibration does not touch — a blocked sensor port, a dead horn, a sounder set to silent. A detector that is perfectly calibrated and cannot be heard is not protecting anyone.
Zero and span
Calibration sets two points. Zero is established in clean air, or in a certified zero gas where the ambient air cannot be trusted — a point worth care, because zeroing a detector in an atmosphere that already contains the target gas sets the baseline wrong in the dangerous direction. Span is set by applying a certified concentration and adjusting the reading to match.
Zeroing an oxygen sensor in a room where oxygen is depleted, or zeroing a toxic gas sensor where a trace of the gas is present, teaches the instrument that the hazardous condition is normal. Where there is any doubt, use certified zero air rather than the room.
The gas is a reference material
The whole calibration rests on the span gas being what the cylinder says it is. Certified gas mixtures carry an assigned concentration and an uncertainty, and that uncertainty belongs in any budget built on the detector. They also carry an expiry date which is not a formality: reactive components adsorb onto cylinder walls, and a cylinder of a low-concentration reactive gas can be meaningfully weaker than its label long before anyone notices.
- Check the certificate and the expiry before use, not after an unexpected result.
- Use the flow rate and regulator the sensor expects; too much flow can pressurise the sensor, too little starves it.
- Apply gas long enough to reach a stable reading — response time is a sensor characteristic, not a delay to be short-circuited.
- Keep a record of as-found values, not just as-left, because as-found is what shows the sensor declining.
Sensors are consumables
Electrochemical cells contain a finite quantity of electrolyte and reactive material, and they are consumed by exposure and simply by time. Catalytic bead sensors for flammable gas lose activity, and can be poisoned outright by silicones, sulphides or lead compounds — sometimes in a single exposure, with no external sign.
The practical consequence is that span drift is the expected behaviour of a healthy sensor approaching the end of its life, rather than an anomaly. Tracking the as-found span reading over successive calibrations turns sensor replacement into something planned, which is exactly the use the calibration interval tool was built for — the history is the evidence.
Oxygen deficiency and the catalytic trap
A catalytic bead sensor burns the gas it is detecting, so it needs oxygen to work. In an oxygen-deficient atmosphere it under-reads, and at high concentrations of flammable gas it can saturate and come back down the curve, reading a low value for a dangerously high concentration. Infrared sensors avoid both behaviours, which is part of why they are specified for confined-space and high-concentration work despite costing more.
Frequently asked questions
- What is the difference between a bump test and a calibration?
- A bump test applies a gas concentration above the alarm threshold and confirms the detector responds and alarms. It is a functional check of the entire chain — sensor, electronics, horn, lights, sensor port — and it returns a pass or a fail. A calibration adjusts the zero and span against certified gas so the reading is correct, and returns as-found and as-left values. Most guidance expects a bump test before each day of use and calibration periodically, or whenever a bump test fails. A perfectly calibrated detector with a blocked port or a silent sounder protects nobody, and only the bump test finds that.
- Can I zero a gas detector in ambient air?
- Only if the ambient air can be trusted to be clean, and that is exactly the assumption worth questioning. Zeroing a toxic gas sensor where a trace of the target gas is present, or an oxygen sensor in a partially depleted atmosphere, teaches the instrument that the hazardous condition is normal — an error in the dangerous direction that produces no symptom. Where there is any doubt, use certified zero air.
- Does span gas expire?
- Yes, and the expiry is not a formality. Reactive components adsorb onto the cylinder walls over time, so a cylinder of low-concentration reactive gas can be meaningfully weaker than its label states while looking perfectly normal. Since the entire calibration rests on the gas being what the certificate says, an out-of-date cylinder means an adjustment made to the wrong value. Certified mixtures also carry an assigned uncertainty, which belongs in any uncertainty budget built on the detector.
- Why do gas sensors need replacing?
- Because they are consumables rather than permanent components. Electrochemical cells hold a finite quantity of electrolyte and reactive material that is used up by exposure and by time alone. Catalytic bead sensors lose activity and can be poisoned outright by silicones, sulphides or lead compounds, sometimes in a single exposure with no outward sign. Span drift is therefore the expected behaviour of a healthy sensor approaching end of life, which is why tracking the as-found span across successive calibrations turns replacement into something planned rather than something discovered.
- Why do catalytic sensors under-read in low oxygen?
- Because they work by burning the target gas on a heated bead, and combustion needs oxygen. In an oxygen-deficient atmosphere there is not enough to support the reaction and the sensor reads low. They have a second trap at the other end: at very high flammable gas concentrations the sensor can saturate and the reading come back down the curve, so a dangerously high concentration can display as a low one. Infrared sensors are immune to both effects, which is why they are specified for confined-space and high-concentration work.
References
- [1]ISO/IEC 17025:2017 — General requirements for the competence of testing and calibration laboratories
- [2]JCGM 100:2008 — Evaluation of measurement data: Guide to the expression of uncertainty in measurement (GUM)
- [3]JCGM 200:2012 — International vocabulary of metrology (VIM), 3rd edition
General technical guidance written against the cited sources. It is not regulatory or legal advice and does not replace the applicable standard, guideline or a qualified reviewer's judgement.
Have a question on this topic?
Ask ValiTrac AI and see the evidence and calculation behind the answer.
Ask ValiTrac AIFree calculator: Calibration interval
Paste an instrument's as-found history and see how fast it is actually drifting, how much of its tolerance that consumes, and how long the evidence supports trusting it. The strongest input to an interval is the instrument's own record — this reads it.
Open the calculatorRelated articles
How to set calibration intervals
There is no universal interval. Set it from risk, manufacturer data and — above all — the instrument's own as-found history.
As-found vs as-left
Why both sets of readings matter, and what to do when the as-found data are out of tolerance.
Measurement uncertainty explained
What an uncertainty statement means, how it is built up, and why a coverage factor matters.