The industry standard accuracy requirement for a flammable gas sensor is +/-10%. Response factors are therefore one of the most significant influences on accuracy, and can easily introduce large errors. Let's examine this concept when comparing FTA vs IR for process applications.
Response Factors
Response factors show how a sensor that is calibrated for one particular gas, usually a "reference" gas, will read when exposed to other gases. If two gases both have the same response factor, they respond equally, unless the sensor is non-linear (infrared), in which case linearization of the signal can introduce more error, or less, depending on the individual characteristics of the sensor for each gas.
Response factors are obtained by testing. The response factors should be obtained from the manufacturer of the sensor. However, the manufacturer's response factors should have independent verification as part of the third-party approval process (FM, CSA, ATEX, CENELEC).
Calibration
For many process monitoring applications, the sensor must be calibrated so that all gases to be detected read the actual concentration or higher, but do not under-report the actual concentration. Therefore the sensor is calibrated for the gas with the lowest response factor.
The amount of error that results from attempting to measure two different gases that have different response factors can be understood by taking the ratio of the two factors. Thus an attempt to measure two gases, one with a response factor to 0.5, and another with a response factor of 1.5, could yield a reading one third (0.5/1.5) or three times (1.5/0.5) the actual concentration.
The flame temperature type sensor is said to have a "universal calibration" for common solvent vapors, because the response factors for common solvent vapors are in the range from 0.9 to 1.1. By contrast, factors for infrared sensors can easily reach 0.25 to 2.0 (an eight-to-one ratio).
The National Fire Protection Association’s NFPA-86 2003 Standard for Safe Operation of Class A Ovens and Furnaces, has concluded that, “...Infrared calibration can vary considerably for various solvent types. Its area of application is for single solvent systems...” (Annex E.1). The same document (Annex E.1) describes Flame Temperature systems, “...calibration is relatively constant for various solvents...”
Add new comment