As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
A gas detector showing zero is making one of two very different statements, and nothing on the screen distinguishes them. Either the air is clean, or the sensor has stopped responding and will go on reading zero in an atmosphere that would hurt you. Sensors fail silently: a catalytic bead can be poisoned, an electrochemical cell can dry out or expire, and a sinter can block with dust or paint, and in each case the instrument looks healthy, passes its own start up routine and says nothing. A bump test, which simply exposes the instrument to a gas concentration above its alarm point and confirms that it alarms, is the only cheap way to turn that silence from an assumption into evidence.
Silence from a detector is an ambiguous reading
Every other instrument on a van gives some feedback that it is working. A tape measure is obviously a tape measure. A multimeter reading zero volts on a dead circuit will read two hundred and thirty on a live one a minute later, so its silence is constantly contradicted by use. A gas detector carried through a normal working week is silent the entire time, and that silence is indistinguishable from the silence of a broken instrument.
The asymmetry of the consequences is what makes this worth attention. A detector that false alarms is annoying and gets investigated within minutes. A detector that has quietly stopped detecting is comfortable, invisible, and still clipped to a collar months later. The failure that produces no symptoms is the one that survives longest in service.
That is the whole argument for a deliberate exposure. It is not a ritual and it is not paperwork. It is the only moment in the life of the instrument when the thing it was bought for is actually demonstrated. The related background on combustible gas detectors covers what the sensing element is doing when it works.
What a self test on power up actually checks
A start up routine typically lights every display segment so missing pixels show, sounds the buzzer and flashes the indicators so the output stages are proven, checks battery voltage against a threshold, and performs an electrical check that a sensor is connected and within an expected impedance window. Some instruments add a check of the clock and the calibration due date.
Notice what is missing from that list. Nothing in it applies gas. An electrochemical cell whose electrolyte has dried out still presents a plausible impedance. A catalytic bead whose catalyst has been poisoned still forms a balanced bridge and still draws its normal heater current. A pellistor behind a sinter blocked solid with paint overspray is electrically perfect. In every one of those cases the self test passes and the instrument is dead to gas.
Some instruments go further and monitor sensor health continuously, flagging an end of life warning when a cell drifts beyond a window. That is a genuine improvement and it is still an inference from electrical behaviour rather than a response to a real sample. It narrows the gap and does not close it.
How a catalytic bead loses sensitivity without saying so
A catalytic bead works by oxidising fuel on a heated catalytic surface and reporting the resulting temperature rise. Anything that coats or chemically occupies that surface reduces the reaction without affecting the electrical circuit around it. Silicone vapours, which come off sealants, greases, release agents and a surprising number of personal care products, deposit silica on the catalyst and the loss is largely permanent. Lead, sulphur compounds and halogenated solvents inhibit the surface too, sometimes reversibly in clean air and sometimes not.
Electrochemical cells degrade differently and just as quietly. They consume a reagent and an electrolyte over their life, they dry out faster in hot low humidity storage, and they simply expire whether or not they have been used. A cell past its life reads low or reads nothing, and the first symptom is usually the absence of an alarm that should have happened.
Then there is the mechanical path. Gas has to physically arrive at the element, which means passing through a sinter, a membrane or a filter. Dust, water, paint and oil mist all block those, and a blocked path produces slow response first and no response later. The sensor underneath may be in perfect health, which is of no comfort at all. Pages on multi gas instruments and the general question of how long detectors last cover the ageing side of this in more depth.
What a bump test proves and what it leaves open
| Link in the chain | Covered by a bump test | Covered by a power on self test |
|---|---|---|
| Gas can reach the element through sinter and filter | Yes | No |
| Sensing element produces a response to fuel | Yes | No |
| Reading crosses the configured alarm threshold | Yes | No |
| Buzzer, light and vibration outputs operate | Yes | Usually |
| Pump or aspirator draws the expected flow | Yes, where fitted | Sometimes |
| Displayed concentration is close to the true value | No | No |
| Span has not drifted since the last adjustment | No | No |
| Channels not exposed to the test gas are healthy | No | No |
The bottom three rows are the reason a bump test never replaces a calibration. They are also the reason the procedure stays cheap and quick enough to do daily, which is the trade that makes it worth having. A separate page sets out how a calibration is performed and why its interval is so much longer.
What four gas instrument listings say about verification
Run the question across real listings and the gap is stark. The Testo gas leak detector appears with no feature bullets at all, so sensor type, measurement scale, response time and any verification routine are Not stated in the listing bullets. The Inficon combustible gas detector publishes its category, its country of manufacture, the manufacturer name and package dimensions of fifteen inches by three point nine inches by eight point four inches, and nothing at all about how its response might be confirmed.
The plug in natural gas and carbon monoxide alarm states a ten year sealed battery life, audible and visual alerts, installation by plugging into a standard outlet, and a sensitive sensor designed to detect trace amounts. A ten year service life is a long time for a device whose only feedback is silence, and no bump test route appears in the bullets. The Forensics Detectors chlorine detector is the outlier: it publishes a range of zero to fifty parts per million with a resolution of a tenth of a part per million, an electrochemical cell, an accuracy of five percent of full scale, a response time under thirty seconds, factory calibration, and both zero and span calibration options.
Only the last of those four gives an owner a documented route to verify the instrument, and it does so by publishing a calibration interface rather than a bump procedure. That is the realistic picture of this market: verification is treated as a professional feature, and the devices most likely to be installed and forgotten are the ones that publish the least about it.
Building a bump test into the way you start work
Attach the check to an event rather than to a date, because events happen and dates get postponed. Before entry into a confined space, before a shift on a site where fuel gas is present, and after any drop, soaking or exposure to solvents are the natural triggers. The procedure itself is short: apply test gas at a concentration comfortably above the alarm threshold, confirm the reading rises and the alarm activates on every channel the gas covers, then let the instrument return to clean air and confirm it settles back.
Record the result even when it passes, because the value of the record is in the pattern rather than the individual line. Instruments that start needing two attempts, or that respond more slowly month by month, are telling you about a sensor nearing the end of its life before it fails outright. And when one fails, take it out of service rather than retrying: a detector that needed persuading is a detector that will be silent at the wrong moment. The distinction between an instrument and a household warning device matters here too, which is why a gas detector and a smoke detector are maintained on entirely different principles.
Frequently Asked Questions
How is a bump test different from switching the instrument on?
Switching on exercises the electronics. A bump test exercises the entire measurement path from the outside air to the alarm. Gas has to get through the sinter and any filter, reach the sensing element, produce a response, drive the reading past the threshold and trigger the indicators. Only the last two of those links are even partly covered by a start up routine.
What concentration should a bump test use?
Enough to take the instrument clearly past its alarm point, which means the test gas concentration has to be chosen against the thresholds actually configured. A gas mixture that only just reaches the alarm level turns a pass or fail check into a judgement call. The purpose is a decisive result, not a measurement, so comfortable headroom above the threshold is the right choice.
Does a passed bump test mean the readings are accurate?
No, and this is the most common misunderstanding about the procedure. A bump test proves the instrument responds and alarms. It says nothing about whether a displayed number is close to the true concentration, because a sensor that has lost a third of its sensitivity can still cross an alarm threshold when a generous test concentration is applied. Accuracy is the job of a calibration.
Can a multi gas instrument be bump tested one channel at a time?
It can, and a mixed test gas that covers every channel at once is the usual approach precisely because single channel testing is easy to leave half finished. Whichever route is used, a channel that was not exposed to anything has not been tested, and an instrument where three of four channels responded is an instrument with one unverified sensor.
What poisons a catalytic sensor?
Silicones are the classic offenders, and they turn up in sealants, greases, release agents and personal care products. Lead compounds, sulphur species and halogenated vapours also degrade or inhibit the catalyst. Some of that damage is permanent and some recovers partially in clean air, and none of it announces itself on the display, which is the reason the failure mode is dangerous rather than merely inconvenient.
What should happen when an instrument fails a bump test?
Take it out of service immediately rather than retrying until it passes. A full calibration is the next step, and if the sensor cannot be brought back inside tolerance it needs replacement. Logging the failure matters as much as fixing it, because a pattern of marginal results across several instruments usually points at storage conditions or a contaminated environment rather than at bad luck.