What calibration gas a detector needs, and how often it is due

As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.

Matching the cylinder to the sensors you actually have

A cylinder spans the sensors whose gases it contains, and it is silent about the rest. This sounds obvious until you own two instruments that share a model name and differ in their fitted cells. The sensor list printed on the instrument is the thing that selects the cylinder, and the common four gas arrangement is a useful illustration of how a mixture is put together. The Forensics Detectors four gas monitor states that it senses oxygen, carbon monoxide, hydrogen sulfide and combustibles as a percentage of the lower explosive limit, which is exactly the set a four gas mix is built for.

In such a mixture the combustible component is usually methane at half the lower explosive limit, the toxic components sit at concentrations comfortably above the usual alarm thresholds, oxygen is reduced below the ambient figure, and nitrogen makes up the balance. Oxygen is reduced rather than raised because ambient air already supplies the upper reference at close to twenty one percent, so the cylinder has to supply a point lower down the scale for the span to mean anything.

The combustible channel carries a complication the others do not. A catalytic bead sensor calibrated on methane does not give the same reading in pentane, propane or hexane, because each fuel oxidises on the bead with a different efficiency. Manufacturers publish a correlation factor for that purpose, and the cylinder you calibrate with determines which factor applies afterwards. Choosing a cylinder is therefore also choosing the scale the combustible reading will be expressed on, and that choice has to match the hazard on site rather than whatever cylinder was nearest. Our guide to how many detectors a site actually needs covers the companion question of coverage.

Zero gas and span gas do two different jobs

A calibration sets two points on a line. The zero point tells the instrument what nothing looks like and the span point tells it what a known quantity looks like, and everything the display reports in between is interpolation. For most toxic channels the zero can be taken in clean ambient air, because the ambient concentration of carbon monoxide or hydrogen sulfide outdoors is close enough to nothing. Where the air is not clean, or where the sensor responds to something present everywhere, a cylinder of zero air or nitrogen supplies the lower point instead.

Oxygen inverts the arrangement. Ambient air is the well known point at roughly twenty one percent, so ambient air serves as the span and the cylinder supplies the depleted reference. Instruments that offer both operations separately make this visible: the same listing above states zero and span calibration options as distinct functions rather than a single button, which is the honest way to present a two point procedure.

Carbon dioxide channels on infrared benches behave differently again, because the zero cannot be taken in ambient air at all. Ambient carbon dioxide is several hundred parts per million, which is a large fraction of the working range, so the low point has to come from a cylinder of nitrogen or from a soda lime scrubber. That difference between an electrochemical channel and an optical one explains why one instrument offers a one button field routine and another insists on a cylinder.

Shelf life, cylinder material and why reactive gases expire first

A certified mixture is a reference only while the certificate holds. Two things erode it. The first is adsorption: molecules stick to the inside wall of the cylinder, so the concentration in the gas phase falls over time. The second is reaction, either with trace moisture or with the wall itself. Inert components such as methane, carbon dioxide and nitrogen are barely affected and their mixtures carry long certified windows. Hydrogen sulfide, chlorine, ammonia and sulfur dioxide are affected strongly, which is why those mixtures are supplied in treated aluminium cylinders and carry a shorter stated life.

Pressure matters for the same reason. As a cylinder empties, the ratio of wall surface to gas volume rises and the adsorbed fraction becomes a larger share of what is left, so the delivered concentration drifts low near the bottom of the cylinder. Manufacturers publish a minimum usable pressure for this reason, and a cylinder below it should be treated as empty even though gas still comes out. Temperature history matters too, because a cylinder left in a hot vehicle or allowed to freeze has been outside the conditions the certificate assumed.

Bump test daily, calibrate on a schedule, and the exceptions

The widely published practice for portable instruments is a bump test before each day of use, with a full calibration at the manufacturer interval, which is commonly six months. The daily check exists because the failures that matter most are blockages and dead cells, and both are invisible on a display that still shows a plausible number. Our explainer on why detectors need bump tests and calibration sets out the reasoning in more depth, and how often to test a gas detector deals with fixed installations.

The exceptions are where the calendar stops being the authority. Calibrate after a bump test fails. Calibrate after any sensor is replaced. Calibrate after an exposure that drove a channel over its range, because a saturated cell recovers with a changed baseline. Calibrate after physical shock, submersion or a battery removal that lost the configuration. Calibrate after a long period in storage, because electrochemical cells age on the shelf as well as in service. None of these events is on a schedule, which is why the schedule alone is never the whole policy.

Fixed domestic alarms sit outside this entirely. A plug in unit like the sealed natural gas and carbon monoxide alarm states a ten year sealed battery and a plug in installation with no setup, and nothing in that design is adjustable. Its test button checks the sounder and the electronics, not the sensor, and the maintenance model is replacement at end of life rather than calibration. Our note on whether a natural gas detector is worth fitting deals with where those units belong.

What these four gas listings publish about calibration

The blanks are not all the same kind of blank. For the sealed domestic alarm the absence is consistent with the design, because there is no adjustment to make. For the Inficon combustible gas detector the bullets supplied with the listing describe the category, the manufacturer and the package dimensions and say nothing about the sensor or its maintenance, so a buyer has to go to the manual before planning a cylinder around it. That is a gap in the listing rather than a judgement on the instrument.

Writing your own calibration schedule

Start from the instruments rather than the calendar. List every portable instrument, and against each one write its fitted sensors, the mixture that spans all of them, the regulator type it needs and the minimum cylinder pressure. That list is the shopping list, and it usually shrinks the number of distinct cylinders a site thought it needed. Then write the daily bump test into the start of shift routine rather than into a separate maintenance plan, because a check that lives in a maintenance plan is the first thing a busy week removes.

Put the calibration interval and the cylinder certificate window on the same page, since a calibration performed with an expired mixture is worse than no calibration at all: it leaves a confident record behind an uncorrected instrument. Finally, record every bump and every calibration, including the failures, because an instrument that cannot show when it was last proven is an instrument whose readings cannot be defended later. For the hardware side of this, our overview of calibration kit components explains what a complete kit contains.

Frequently Asked Questions