Temperature is defined as a measure of the relative hotness or coolness of a material compared to a known reference. We use relative temperature scales such as Fahrenheit and Celsius in most day-to-day temperature measurements. They are called relative because they are relative to the boiling and freezing points of water (Fahrenheit used a brine solution). That is a common reference. We also have absolute temperature scales, Kelvin, and Rankine, which are called absolute because they start at absolute zero and that is our reference, the point where molecular motion stops. However, in thermography, it is not just a question of which scale to use because once you decide on Celsius or Fahrenheit (usually), you must determine which temperature to report. Rise over ambient? Rise over adjacent/normal phase (in electrical applications)? Apparent temperature? Which one do you use in your report when required to report the “temperature of a suspected fault?” should you use a different one for trended temperature data?

For many applications, qualitative data may be adequate. Ideally, we may avoid reporting any temperatures due to the potential limitations and potential inaccuracy of radiometric measurements. But, when required by the end-user (person or persons that we provide the report to) to report temperatures, we must consider the pros and cons of each one.

The temperature value that your camera provides you when you view a surface is the apparent temperature of that surface. This would be the quickest, easiest temperature to report. However, we must take a few factors into account. Even when we use high emissivity targets with little or no convection losses across the surface of the target, the apparent temperature still may not be correct. Most thermal imagers have an accuracy of +/- 2°C or +/-2% (whichever is greater at 30°C) in the best-case scenario. For precise temperatures, we must also know the exact emissivity and background (reflected temperatures) of the surface we are viewing and make the necessary adjustments to the camera. This may not be practical in real-world situations. Without knowing the exact emissivity and background, we can never get an exact radiometric temperature. If we use high emissivity targets, the apparent temperature should be close to the actual temperature, but it will not be exact. With low emissivity surfaces, our apparent temperature can be off by a wide margin.

Rise over ambient temperature is another possible temperature to report. To use this method, we must ask, what is ambient? And how did we determine the ambient temperature? Did we use the weather forecast? A personal weather station? An app? Or, did we use our camera to get an apparent temperature from a wall or floor (which from the above paragraph we know will not be exact)? Whatever method is used must be documented and we must consider the reliability of the method. Another concern with this method involves distance. If we are looking at a splice on an overhead line, 100 feet in the air, are the ambient conditions there going to be the same as where I am standing? Probably not. Another example might include the ambient near a belt-driven asset or a motor will be significantly different than where the thermographer is standing.

We can report the temperature rise above the maximum operating temperature of the equipment (if we have that data). If we use this method of reporting, we must bear in mind that the maximum operating temperature of the device is dependent on the device being under maximum load as well. How will loads other than the maximum affect the operating temperature of the device? If the device is only operating at 75% load it may very well be running under the maximum operating temperature, leading us to believe that there is no issue. What will happen to the temperature of the device if load increases in the near future? Load readings, cycle times must be taken on each device at the time of discovery which will greatly increase the time it takes to complete your inspection. So, we need to consider environmental and operational conditions in each of these situations. Another consideration when comparing to maximum operating temperature is the fact that this approach does not support condition monitoring. But instead, may indicate functional failure has already occurred and catastrophic failure may be imminent.

Of these various methods, probably the most useful temperature to report is rise over adjacent/normal phase or similar component. Most of what we do as thermographers is to look for differences qualitatively. When comparing similar components or adjacent phases it is easy to see apparent temperature differences as the similar components should be near the same temperature. That is the beauty of IR, it gives us a qualitative image that displays an apparent temperature difference So, even though we know that the apparent temperature of the fault and the apparent temperature of the similar component are not entirely accurate, we can still visually see the difference between them. The apparent Delta T between two similar components may not be as large as the actual Delta T, meaning the values we get from our camera should be considered minimum readings, if we set our collection parameters conservatively.

For thermographers that usually work as outside contractors, you may want to include a disclaimer in your reports explaining what temperature(s) you are reporting, how you determined those temperatures, the limitations of the technology, and factors such as emissivity.

Without proper training, many people do not understand that infrared cameras are not just expensive thermometers. The camera measures infrared radiation and from that data calculates or infers a temperature for us, which is not always correct or exact. If we understand the limitations of the technology, basic heat transfer theory, and the pros and cons of what data to report we can avoid some of the pitfalls that beset the untrained. Remember, you are only as good as your reports so make sure you report what is relevant.