Calibration for Gas Detectors
Gas detection calibration may feel like a black art to those who have not been trained to undertake it. Unfortunately there are also far too many who do it with no real understanding of what is taking place and so have no explanation if things go wrong. The reality is that all events are explainable and of course so is how the units arrive at this information. That though does not mean that you will always arrive at the answer you wanted, just that you will get an answer.
So let us firstly look at calibration of flammable gases. What are we trying to accomplish when we do this? The answer to this is quite straight forward. We should have initially conducted a risk assessment which would have identified a particular target gas as being the most likely to be found where we are working. In the case of working in sewers this would most likely be methane / CH4. If this is the case then your gas detector would need to be calibrated for methane. The values at which methane explode are very specific as are all flammable gases. So because we know that methane needs 5% by volume of for instance the room you are in, we know that this amount in any space now offers the potential for explosion. So this is then translated in to a percentage of LEL / Lower Explosive Level. Because we know that 5% is our first potential explosion this is now called 100% of LEL. This is what you will see on the screen of your gas detector when using it.
So when we are calibrating your unit we will deliver an exact percentage of methane to the sensor, for instance 50% of LEL. This should then give us a reading of 50% LEL on the screen of your gas detector. So how does this reading take place? To try to put this very simply, if the detection cell is a pellistor then it will have 2 beads internally inside it. These beads are made up of very small coils of wire which are heated to 500 degrees plus. One bead is coated so that nothing can get to it and the other is open. The beads now glowing cause internal explosions within the sensor housing. These explosions can only happen around the exposed bead, these explosions are then converted into an electrical current. The coated bead sees nothing so is used as a reference point and the reading sent to the screen is a direct relationship as to how much gas the sensor is seeing. So our 50% LEL gas should cause the correct number of internal explosions to give us the 50% LEL reading on the screen. If this does not happen then an adjustment is made so that this is the reading that is detected. Once this has been done then the gas detector will be set so that the unit shows that it has been calibrated with a verified gas and all of the details regarding calibration gas trace-ability values used will be logged. This is then related onto the test certificate for your gas detector. This can now be done electronically with some units using calibration gas that has RFID tags so that all of the information is stored within the units own memory at time of calibration.
So how is this done if your gas detector has no screen, a fixed system for instance. Well because the voltage output is a know value for the percentage of calibration gas used, then an AVO / Amps, Volts, Ohms, meter, can be used to verify the electrical output of the cell. For instance with a flammable cell set for LEL then the full range of the cell would be 100% and if this used a standard 4-20 mv output cell the cell would read 20 mv for the full 100%. The standard cells for fixed systems use this range and 4 mv is usually equal to zero. So if we now put a 50% CH4 calibration gas over this cell we should get a reading of 12 mv, this is because the usable range is 16 mv. So half of this value is 8 mv, add that to the 4 mv that we require for a reading of zero and we end up at the 12 mv. So if we don't see this reading then we make an adjustment is made so the cell shows the exact gas value. This is then recorded again and the certificate issued.
Calibration gases are supplied with very strict documentation and some of these gases like H2S are reactive. This means that the gas will leach through the walls of the cylinder over a certain period of time. So they will have a use by date on them. This ensures that the readings you are taking are stable and can then be used for calibration. Whilst flammables are generally stable they can only be supplied one at a time. So what do we mean by this, well if you have a cylinder of calibration gas with a cocktail mix of; 02, H2S, CO and 2.2% CH4 you can not have another flammable in the mix. This is because your gas detector can not discern which is which and you will simply get the reading from the gas that ignites earliest.This I find is generally not understood, like the question I have several flammables that I am looking to detect and I need an accurate alarm for them all. Unfortunately this shows a lack of understanding as to what the cell can actually do, it after all only provides an electrical pulse and doesn't truly know what it is recording. It also demonstrates a lack of understanding as to how gas detectors are calibrated. As I stated at the start of this you do need to ascertain the most likely gas you are trying to detect. The flammable sensor may well see any of the flammables you have,(I will cover possible problems on this in a bit) but it will only alarm accurately for the gas you have calibrated the unit for. So if it sees a flammable other than the target gas then that alarm may be either too soon or too late. It is of course always better that it alarms too soon rather than too late, but only if you have a genuine possibility of the gas you are looking for, otherwise you may get consistent false alarms. This could be because you have asked for your unit to be calibrated for a gas that explodes earlier than CH4 but that you are never likely to find. This is a pointless exercise and often then leads to the users loss of trust in the unit. This is not a fault of the unit and most units may suffer the same problem, simply because the target gas is incorrect.
We can then also see other problems where the target gas is not within the range of the sensor. This could be that the flammable required to be detected is in actual fact a vapor rather than a gas or a VOC / volitile organic compound. In these instances an NDIR sensor rather than a pellistor can prove more useful as this may well pick up these but then you may loose sensitivity on small chain hydrocarbons like hydrogen.
Toxic cells derive there information by the target gas reacting with chemicals inside the sensor. So in the case of H2S / hydrogen sulphide, if the target gas enters the sensor it will create and electronic response which gives a stable electrical voltage indicating the ppm / parts per million detected. So again when it comes to calibration this can be formulated fairly easily if you know what you are looking for. The easiest to explain would again be a fixed system sensor that uses a 4 to 20 mv output. The first thing we need to know is the gas range the cell has been made to detect. So it may be a 0 to 50 ppm H2S cell, so 20 mv is equal to the full range of 50 ppm H2S. So divide 16 by 50 and that equals 0.32 mv. This means that every 1 ppm is equal to 0.32 mv as only 16 mv of the range gives a reading. So if we are using 20 ppm H2S calibration gas then we would need to see a reading of 20 x 0.32 = 6.4 mv + 4 for the zero reading, so 10.4 mv to set the calibration level.
All catalytic cells work in this way and this is why you always have to replace oxygen sensors at their warranty dates. The chemicals inside the oxygen sensors get used up over a period in time and this is a very finite calculation. So as there is always oxygen in the atmosphere your O2 / oxygen cell will start using up its expected life. So even if your gas detector is stored in the cupboard from the day you bought it the oxygen sensor will start depleting its life expectancy. This is why you always need to replace these sensors at their stated time. The other toxic sensors come into contact with there target gases less so their life expectancy can be a lot longer than the warranty period.
There are many things that can effect your calibration, dirty sensor filters, the cell not having been made correctly, seeing too much of the target gas which can over range the sensor along with a whole load more things. So when manufacturers state that they have a gas detector that doesn't need calibration, this can in fact be true. But be vary wary as all manufacturers state that you should bump or function test your unit regularly. If you bump test your gas detector regularly it will prove that all of you sensors are working and have not been affected by any of the things that can stop them from sensing correctly. If the bump test that you do does then not give a reading that is within the units preset range the unit will go straight into calibration mode to rectify the fault. It will then either pass or fail this calibration. Either way the user will be safe, but only if you are bump testing regularly. The problem is that many will read no calibration required and believe they don't have to do anything else as the manufacturer will not identify the bump test requirements other than in the user manual. So when marketed like this it can lead to decreased safety. The way I always phrase it to clients who have sent in units for servicing and require new cells fitted because they have stopped working is "Did it stop a few days after its last calibration or a few days before you sent it to us". Unfortunately the answer is you just don't know.
So please remember that a gas detector that says it doesn't need calibrating will still require bump testing although you may not be told that. A gas detector that doesn't detect what it was bought for is PPE that is not fit for purpose and that presents a massive problem, not only after an accident but you may have supplied your staff with an item that could lead to their death.