Here is my situation. We use G460 instruments for testing the atmosphere in an underground mine. Our instruments are set up to measure O2, LEL, CO, H2S and SO2. I did air sampling with one of our GfG G460 instruments. At the end of the sampling period (about 10 hours), when I picked it up, it displayed a -4 ppm value for carbon monoxide. I’ve never seen negative readings on the CO sensor.

My question: What could cause this to happen?

It was 114 degrees Fahrenheit in the workplace where the GfG instrument was being used. Could this be a factor? The G460 had been bump tested prior to use and is less than two years old.


Thanks very much for your question.

The ambient temperature should not be an issue.  G460 instruments are CSA® Certified for use in ambient temperatures up to 55°C  (131°F).  The age of the sensor should not be an issue either.  The dual-channel sensor used to measure CO and H2S in your instrument is warranted for three years, but expected to last even longer.

Carbon monoxide sensors are normally very stable.  They are cross sensitive to a couple of potential interfering gases, however.

Usually negative readings are the result of the CO sensor being fresh air adjusted while in the presence of a detectable interfering gas, or when the sensor is fresh air adjusted before it has completed recovering from a prior exposure to an interfering gas.

The CO sensor is always producing a signal as long as the instrument is turned on.  In the presence of CO the signal goes up.  In fresh air, you still get a signal, but the value is lower.  When you fresh air zero the instrument you are telling the instrument to use the signal from the sensor at that moment as the point of comparison.

You always want to make sure that the instrument is located in an area where the air is known to be fresh before making a fresh air “zero” calibration adjustment.  If you can’t be sure that the air is fresh you may need to use air from a cylinder of “fresh air” calibration gas.   If the instrument uses the signal of the sensor while it is in the presence of a detectable gas, later on, when the sensor is located in an area where there is no detectable gas, the signal will be lower, causing a negative reading.

Sometimes the problem is due to the presence of small concentrations of CO being present in the area where the instruments are being calibrated.  Make sure that any possible sources of CO are eliminated.  Also, make sure that the calibration gas that used to span adjust the CO sensor is not allowed to accumulate in the area where you fresh air adjust your instruments.  Even though four or five ppm of CO is not harmful or dangerous, the presence of the gas can have an effect on the accuracy of the fresh air zero procedure.

The most common interfering gas to have an effect on CO sensors is hydrogen.  I am sure you have many intrinsically safe, rechargeable battery equipped devices, such as cap lamps, tractors, etc. around the mine.  Try to make sure that the instruments are not stored or calibrated in areas where you are recharging large numbers of batteries, especially the lead acid batteries that are used to power larger types of equipment such as tractors and carts.

The hydrogen does not harm the CO sensor.  However, the CO sensors in your instruments will show a slight response.  If you fresh air zero adjust the instruments in an area that has 10 or 20 ppm of hydrogen around due to a nearby battery charging station the CO reading can wind up being off by a few ppm later on.

Another type of interfering gas that can have an effect on CO sensors is the vapor produced by solvents.  Many paints and degreasers contain solvents such as methanol and toluene.  CO sensors have an internal organic vapor filter that is able to remove small concentrations before the sensor is affected.  However, in higher concentrations, if the filter becomes saturated, the solvent can fully penetrate the sensor, in which case the reading may start to become affected.  Typically, when the CO sensor is exposed to alcohol or solvent vapor, the readings slowly climb for a period of time even after exposure to the solvent vapor has stopped.  In the case of heavy exposure it can sometimes take hours, or even a full day (or more) for the sensor to recover completely, and for the fresh air reading to stabilize completely back on zero.  The acetylene gas used in welding and hot work procedures has a similar effect, but can take even longer for the sensor to clear.

If the CO sensor is fresh air adjusted before it has completed recovering, say when it is still showing a reading of four or five ppm; later on, when it has finished recovering, it will show four or five ppm negative.

It is important to wait until the sensor has stabilized completely before making a fresh air zero adjustment.  If you see  the sensor reading is continuing to drop, even slowly, try to wait before performing the fresh air adjustment.  If you don’t you may need to readjust the fresh air reading to correct the response.

I hope this information is useful.  Thanks again for the question.

Do you have another question? Simply ask Bob! Contact us