Carbon Monoxide Measurement


Thread Starter

CESInstAnalyser HZ

Hi All,
We are using Luft Detector type Infra Red analyser to measure CO (0-200 ppm) in
following background gases. The normal concentration of the CO in the process is
around 10-12 ppm.

CO2 7.12 %
Ethylene 23.0 %
Ethane 0.48 %
O2 4.11 %
Argon 9.76 %
N2 2.53 %
Methane 51.4 %
Ethylene Oxide 2.07 %
Moisture 1.76 %
Acetaldehyde 0.002 %

We calibrate the analyser with following calibration gases.

Zero gas : N2 - 8.5%, Ethylene - 25%, CO2 - 5%, Argon - 6%, Ethane -
0.5%, Methane - Balance (55%)
Span gas : CO - 200 ppm balance above zero gas.

With the help of above calibration gases we are able to calibrate the analyser, but it shows higher value in the process sample viz. it shows 50 ppm CO instead of 12 ppm.

This means that there is a interference of some background gases. We want to know which of the above listed components interfere with CO measurement and to which extent. ( e.g. 20% CO2=2ppm CO etc.)

The reference cell is filled with the Zero gas and then sealed. We want to make it "FLOW THROUGH" reference with process sample itself less CO component, so that it will be TRUE CO measurement with interference of all the background gases nullified. To achieve this we have to remove the CO from the process sample. Does anybody know the method to remove the CO from the above process sample (without affecting the other background gases.) or convert it to CO2 or other gas which does not interfere with CO measurement.

Alternately, I would like to know which type of analyser will be best suitable for above application.

Thanks in advance.
Manoj Topiwala

The analyser supplier should be able to advise you as to the 'Interferance Ratio' of your background gases. It varies between different brands of analyser depending on the band(s) of Infra Red being used.

The N2, O2, and Ar will not affect the reading, however most of the other gases will - depending on the Interferance Ratio for your analyser.

If the concentration of these gases is constant - the error can be calibrated out. For the gases that vary their concentration - you need divide the variation by the interferance ratio to determine the possible error in your CO measurement.

eg: if a gas has in interferance ratio of 1:1000, and its concentration varies from 49 %vol to 50 % vol :
variation = 1 %vol = 10,000ppm
10,000 / 1000 (interferance ratio) = 10
therefor you would see a error of 10ppm in your measurement


Phil McKay
Silvan, Vic, Australia
[email protected]