Emissivity, load banks and reference components: why reading a busbar in thermal imaging takes more than pointing the camera at the brightest area.
Hand a new thermographer a camera and point them at a live busbar chamber, and they will almost always call the brightest area as the fault. Sometimes they are right. Often they are not. Busbars are one of the places where the naive reading, hottest equals worst, breaks down most reliably, and understanding why is a good test of whether a survey is being run on physics or on colour.
Emissivity and the copper problem
The first trap is the material itself, and it is a lesson in emissivity. Bare, polished copper is a poor emitter of infrared. Its emissivity can be very low, which means a radiometric camera, left on a default setting, will badly under-read the true temperature of a clean copper bar and instead show you reflections of whatever is warm nearby. A shiny busbar can look deceptively cool while a dull, oxidised, or taped section beside it looks alarmingly hot, purely because of how each surface emits, not because of any real temperature difference.
This is why experienced thermographers pay attention to surface condition before they trust a reading. Where a length of busbar carries electrical tape, a label, or a matte painted finish, that higher-emissivity surface gives a far more honest temperature than the bare metal next to it. Where nothing suitable exists and the reading matters, the emissivity setting has to be considered deliberately rather than left on a generic default. Get this wrong and you will either miss a genuinely hot bar or chase a reflection that was never hot at all.
Expected heat versus fault heat
The second trap is assuming that heat means fault. Plenty of equipment is supposed to run hot. A load bank under test is the obvious example. It exists to dissipate energy, and it will show large, bright thermal signatures that are entirely by design. Reading those as faults is a rookie error that erodes a client's confidence fast.
The skill is separating heat that belongs from heat that does not, and you cannot do that on the thermal image alone. You do it by knowing what the asset is, what it is rated to do, and what it was doing when you looked at it. A busbar warming evenly along its length under heavy load is behaving. A single joint or connection running hotter than the bar either side of it, under the same current, is not. The pattern matters more than the peak.
Reference components: the method that keeps you honest
The technique that turns this from opinion into assessment is the use of reference components. Rather than judging a component against an absolute number, you compare it against something that should be operating under similar conditions. On three-phase equipment, the phases carrying similar load are natural references for each other. The line and load sides of a device, or identical connections along the same bar, serve the same purpose.
When one phase runs warmer than its neighbours under comparable load, that differential is a genuine signal, and it holds even when the absolute temperatures are unremarkable. It is also far more robust than a fixed threshold, because both the reference and the suspect component share the same ambient, the same load pattern and the same surface conditions. The comparison cancels out most of the noise that trips up single-point readings.
Bringing it together at the panel
A sound busbar assessment therefore layers several things that no single image contains. Surface condition and emissivity, so the reading is honest. Load, so expected heat can be separated from fault heat. Reference comparison, so a differential can be seen rather than an absolute guessed at. And where a real anomaly is found, a load-corrected figure, so its severity means something regardless of how loaded the system was on the day.
Note how little of that is about the camera. A capable radiometric camera is necessary, but the difference between a busbar survey that finds the right fault and one that chases a reflection is method, not megapixels. The camera shows you the scene. Deciding what in that scene is actually a problem is the thermographer's job.
The takeaway
On a busbar, the hottest spot on the image is a starting point, not a conclusion. Before you call it, ask what the surface is made of and how it emits, what the asset is rated to do and how hard it was working, and how the suspect compares against a component that should be behaving the same way. Skip those and you will confidently annotate reflections and miss real differentials. Work through them and the busbar stops being the confusing part of the survey and becomes one of the most reliable places to find a fault before it finds you.
