When you calibrate a gas detector, the certified concentration in the cylinder is only the starting point. Between the cylinder and the sensor lies a length of tubing, and that tubing can quietly steal gas and delay it, so the concentration arriving at the sensor is not the concentration on the label. This guide explains why calibration gas delivery tubing must be short, inert, and properly purged, how dead volume and adsorption on the tube walls suppress and slow the reading, especially for reactive gases like hydrogen sulfide, and how a rushed calibration lands the span low and leaves a detector that under-reports.
Cal-gas purge in one line: Cal-gas tubing purge refers to running calibration gas through the delivery tubing long enough to flush out the dead volume and saturate the tube walls before trusting the reading, and to keeping that tubing short and made of inert material. Dead volume dilutes the incoming gas until it is displaced, and adsorption of reactive gases like H2S onto the tube walls suppresses the concentration until the walls are saturated. If the tubing is long, reactive, or not purged long enough, the sensor sees a low concentration and the calibration sets the span too low.
Every length of tubing between the regulator and the sensor holds a volume of gas, and at the moment calibration begins that volume is full of whatever was there before, usually room air. When the cylinder gas starts to flow, it first has to push all of that old air out before the sensor can see pure calibration gas. Until the tubing is fully swept, the sensor is exposed to a mixture of fresh gas and residual air, a diluted concentration lower than the cylinder's certified value. This trapped, must-be-displaced volume is the tubing's dead volume, and the longer or larger the tube, the more of it there is and the longer it takes to clear.
Adsorption is the subtler and more damaging effect. The inner surface of the tubing is not perfectly inert; gas molecules can stick to it, and reactive gases stick readily. As calibration gas flows down the tube, some of its molecules are captured by the walls instead of continuing to the sensor, so the concentration reaching the sensor is suppressed below what entered the tube. The walls keep taking up gas until they are saturated, and only then does the full concentration pass through. Until saturation, the tubing is actively removing gas from the stream, and the sensor reads low as a direct result.
Hydrogen sulfide is the classic troublemaker because it is highly reactive and adsorbs strongly onto surfaces. A run of tubing that would be harmless for an inert gas can noticeably eat H2S, so the sensor sees far less than the cylinder holds until the walls have taken their fill. This is why calibrating a low-level H2S detector through a long or poorly chosen tube is notoriously prone to setting the span low: the tubing itself is a sink that competes with the sensor for the gas, and the effect grows with tube length and with how reactive the gas is.
The countermeasures follow directly from the causes. Keeping the tubing run short minimizes both problems at once: less tube means less dead volume to flush and less wall area to adsorb gas, so the delivered concentration is closer to the certified value and reaches full strength faster. This is why good practice puts the detector as close to the cylinder as the setup allows and avoids long, coiled delivery lines. Every extra length of tube is extra volume to purge and extra surface to strip gas from the stream.
Material choice attacks the adsorption problem specifically. Inert materials such as PTFE, or metal tubing that has been passivated to make its surface unreactive, adsorb far less gas than ordinary or reactive tubing, so much less of the calibration gas is lost to the walls. For reactive gases like H2S this choice is not a nicety; using a reactive tube can make an accurate low-level calibration effectively impossible because the tube consumes too much of the gas. Selecting tubing known to be compatible with the specific calibration gas is therefore part of getting a valid calibration, not an afterthought.
Even with short, inert tubing, purge time remains essential. Purging means letting the calibration gas flow long enough that the dead volume is fully displaced and the walls are saturated before the reading is trusted or the span adjusted. The reading rises toward its true value as the tube clears and the walls fill, and only when it has stabilized does it reflect the certified concentration. Cutting this short, adjusting the span while the reading is still climbing, is the single most common way to set a span low. The discipline is to wait for a stable reading, allowing extra time for reactive gases and longer lines, before touching the calibration.
The whole point of a gas detector feeding a monitoring system is that its reading can be trusted to warn of dangerous concentrations, and calibration is what earns that trust. When cal-gas tubing quietly suppresses or delays the delivered concentration, the calibration is corrupted in a way that does not show up as an obvious error. The technician sees a reading that has climbed to somewhere near the target and adjusts the span to the labelled value, not realizing the sensor never actually saw the full concentration because the tube stole part of it. The span is set low, and from then on the detector under-reports real gas by the same margin.
This is the dangerous outcome: a detector that passed calibration but reads low. In service it will show a smaller number than the true concentration, so it alarms late or fails to alarm at a level that should trigger it. For a cloud SCADA platform such as Merobix, where gas readings from field detectors drive alarms and are trended over time, a fleet of detectors calibrated through reactive or unpurged tubing would send systematically low values, giving false reassurance about a genuinely hazardous atmosphere. The distortion is invisible in the data itself; it is baked in at calibration and only revealed by comparison with a known reference.
The response times seen at calibration are also a clue to real-world performance. A detector that took a long time to reach its full reading on cal gas, because of dead volume and adsorption in the delivery tube, hints that its response is being slowed somewhere, and a sluggish detector responds late to real gas too. Good practice keeps delivery tubing short, inert, and adequately purged so that both the concentration and the response time seen at calibration reflect what the sensor is truly capable of. Only then does the value arriving in the monitoring system mean what it says, and only then can operators rely on it to protect people.
A short tube has less dead volume to flush and less wall area to adsorb gas, so the certified concentration reaches the sensor faster and closer to full strength. Inert materials like PTFE or passivated metal adsorb far less gas than reactive tubing, which matters greatly for gases like H2S that stick strongly to surfaces. Long or reactive tubing strips gas from the stream and delivers a diluted, delayed concentration, setting the calibration low.
Hydrogen sulfide is highly reactive and adsorbs strongly onto tube walls, so as it flows down the delivery line some of it sticks to the surface instead of reaching the sensor, suppressing the concentration until the walls saturate. If the technician adjusts the span before the walls are saturated and the reading has stabilized, the sensor is seeing less than the cylinder holds, so the span is set too low. The detector then under-reports H2S in real service.
Long enough for the reading to stabilize, meaning the dead volume in the tubing has been fully displaced and the walls have saturated so no more gas is being lost to them. The reading climbs toward its true value as the tube clears, and only a stable reading reflects the certified concentration. Reactive gases and longer tubing runs need more time; adjusting the span while the reading is still rising is the most common way to set a span low.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.