Determining Measurement Uncertainty, PLUS: Your Favorite Quotations

This week, we answer a question from a customer who asked how to determine measurement uncertainty when using a humidity sensor (i.e. in an environmental chamber). He says that although measurement uncertainty is not always requested by his clients, who usually accept the unit calibration certificate as sufficient, it is something he wants to be able to provide if asked.

From this customer's email:

"I'm more concerned with the uncertainty of measurement associated with the results obtained from the units, than the calibration of the unit itself. Do you ever take into account other factors that could influence the measurements obtained from the humidity logger during use, i.e. the stability of the unit itself between calibrations, the ambient temperature, errors associated with the positioning of the device, sufficient flow across the sensor? Or are such factors considered insignificant considering the overall accuracy and do you consider this sufficient justification for ignoring them? "

It sounds like you are trying to determine the uncertainty of a measurement you are making with one of our instruments. It is important when using an RH device to fully understand all the components that contribute to the overall measurement uncertainty. The performance and calibration uncertainty of a measuring device are just two factors that influence the total measurement uncertainty. Remember, each and every measurement has an associated uncertainty. A device used to make a measurement in one specific application may yield a different uncertainty when used in a different application.

In practice, measurements made outside of a calibration laboratory are subject to a huge variety of variables that are not easily quantified. Factors such as air flow, temperature gradients in the chamber, temperature stability, and radiant heat sources may all contribute significant uncertainty to a measured value. These might be obvious, but also consider factors such as the proper use of the product, knowledge of the product and its applications, operator competence and alertness, unnoticed damage to the measuring device, environmental conditions outside of the chamber, and spurious electromagnetic signals (large electric motors, walkie-talkies, etc.).

Coming back to the original question of “are these factors significant sources of error?” there is no single correct answer. Specific circumstances and an understanding of the purpose of the measurement will go a long way toward answering this question. For example, a standard bathroom scale is adequate for weighing yourself, but definitely inadequate for weighing 100 mg of an active pharmaceutical ingredient. This is why metrology cannot be 100% proceduralized and why we have metrologists.

For example, if the measurement instrument is in the direct airflow of an HVAC outlet, it may see a very different temperature (and therefore relative humidity) than if the sensor is in a sheltered space. Heat generating equipment nearby or even people being located immediately next to a sensor can cause changes in readings. Actually, people tend to have a large effect, especially in low relative humidity environments because we are humidity sources (just breath on an RH sensor to see a rapid change). All of these outside influences should be considered when taking a measurement to ensure that the effects are minimized or at least understood.

Your Quotes on Science, Metrology & Quality

Thank you to those who responded to last week's blog post on "The Most Inspiring Quotations on Science, Metrology & Quality"!
Here are the best quotations sent in by blog readers: