Why do LEL sensors drift so much?

Why do LEL sensors drift so much?

Dave Wagner | Friday, September 20, 2013

In recent weeks, I have addressed many questions as to why catalytic bead LEL sensors seem to drift so much, particularly in the hot, humid summer weather. To understand the reason for the drift, you must understand the operating principle behind the sensors themselves.

Catalytic bead LEL sensors are made up of two resistive fine-wire elements, one a detector and the other a reference. As the detector element encounters combustible gas, the temperature of the bead increases and the resistance increases accordingly. The difference in the resistance between the detector and the reference elements is the signal representing the concentration of gas.  

Theoretically, in a “clean-air” atmosphere, the resistance of the elements does not change and the sensor signal remains at zero, but in reality, this is not the case. The resistance of the elements will also change as the thermal properties of the atmosphere change. Again, in theory, if the resistance of the two elements is equal, any change in the thermal conductivity of the atmosphere not related to the presence of combustible gas would cause the resistance of each element to change at the same rate and the signal from the sensor would remain at zero. But this is only theory and the resistance of the two elements is very rarely exact. Thus, any change in the thermal conductivity of the atmosphere such as the change caused by an increase in water vapor, ie. relative humidity, will cause the resistance of the elements to change at different rates creating a signal from the sensor that appears as zero drift. The drift may be positive or negative and is solely dependent on the resistance of one element relative to the other.

Instrument manufacturers and sensor users both have different ways of dealing with the drift. Some manufacturers will mask any drift in the negative direction. Others will mask a certain amount of drift in both directions by applying a “dead-band” to the sensor readings where any reading within the limits of the dead-band will appear as zero. Other manufacturers minimize the amount of drift by using more complex software filtering algorithms. Many sensor users will simply ignore the drift as it is rarely more than just a few percent LEL and well below the typical LEL alarm points at 10 percent. Others that use very low LEL alarm point thresholds find the drift much more troublesome. In these cases, the best way to compensate for the drift in the sensor caused by the changes in the atmosphere is to allow the sensor elements to stabilize in the thermal environment in which they are used and then zero the sensors in that environment.

The good news about LEL sensor drift is that more comfortable fall weather in the northern hemisphere is at hand and the drift issue will typically take care of itself. Until then, and again when the hot-humid weather of summer returns, my recommendation is to keep the LEL alarm thresholds set at reasonable levels and zero the LEL sensors in the outdoor environments in which they are used.

Enjoy the fall and keep safe!

Dave