The great advantage of infrared thermography on electrical equipment is distance. The camera measures radiated energy, so the thermographer never has to touch anything, never has to place an instrument inside an enclosure, and can stand outside the arc flash boundary while the asset carries full load.
That distance is also where the difficulty starts. Everything the camera reports has travelled from a surface, through air, to a detector — and each step can distort it.
What the camera is actually measuring
An infrared camera does not measure temperature. It measures radiated infrared energy arriving at its detector and converts that into a temperature, using assumptions the thermographer supplies.
The energy arriving at the lens is a mixture of three things:
- Energy emitted by the target because of its own temperature
- Energy reflected off the target from everything else in the room
- Energy transmitted through the target from behind it
For opaque electrical components the third is negligible. The first two are the whole problem.
Emissivity: why shiny metal lies
Emissivity is how efficiently a surface radiates energy compared with a perfect emitter. It runs from 0 to 1.
Painted steel, plastic, rubber and heat-shrink sit around 0.90–0.95. They are honest surfaces and the camera reads them accurately with default settings.
Bare copper busbar, polished aluminium and plated hardware are a different matter. A clean copper surface can have an emissivity as low as 0.05 — it radiates almost nothing of its own and mirrors nearly everything around it. Point a camera at a busbar running at 90 °C with the emissivity set to 0.95 and it may report 35 °C.
This is the single most consequential error in electrical thermography, and it always fails in the dangerous direction: a hot component reads cool.
The practical answers are:
- Correct the emissivity setting for the material being measured
- Measure at a nearby point with known, high emissivity — the heat-shrink on the cable, the painted lug body, the terminal insulation — rather than the shiny metal itself
- Where a site is inspected repeatedly, ask for high-emissivity targets to be applied to key measurement points during a shutdown
Reflection: the thermographer in the picture
A low-emissivity surface is, by definition, a good mirror in the infrared. It will reflect the ceiling lights, the surrounding cubicles, warm plant nearby, and the person holding the camera.
A "hot spot" that moves when the thermographer moves is a reflection. A hot spot that stays put when viewed from two different angles is real. Checking a finding from a second position costs a few seconds and eliminates a whole class of false positives.
Reflected apparent temperature is a camera setting for exactly this reason: it tells the camera what the surroundings are contributing so it can subtract them.
Spot size: the fault too small to see
Every camera has a measurement spot — the smallest area from which it can return a reliable temperature. It is set by the detector resolution and the lens field of view, and it grows with distance.
If the component you are measuring is smaller than that spot, the camera averages the component with everything around it. A small, very hot termination surrounded by cool metal will report as warm rather than hot.
The rule of thumb is that the target should be at least three times the measurement spot for a trustworthy figure. In practice that means getting closer, or fitting a telephoto lens, when the target is a single terminal rather than a whole board. This is covered in more detail in spatial resolution explained.
Why comparison beats absolute temperature
Because all three effects above push readings low, and because ambient conditions vary from site to site and season to season, a single absolute temperature is weak evidence.
The strong measurement is a comparison between two components doing identical work. Phase B against phase A on the same board, under the same load, made of the same material, with the same surface finish. Both are subject to the same errors, so the difference between them survives when the absolute figures do not.
This is why every finding in a competent report quotes both the anomaly and its reference, and why a report that gives you one number per photograph has not really told you anything.
Working safely
Thermography is a low-risk inspection method precisely because it is non-contact, but "low risk" is not "no risk":
- Covers are removed by a licensed electrician, not the thermographer
- The thermographer works outside the arc flash boundary and wears appropriate PPE for the task
- Boards remain live and loaded — this is a working environment, not an isolated one
- Live work permits and site inductions apply as they would for any other live electrical activity
The plant keeps running throughout. That is the point of the method, and it is why an annual survey does not need to be scheduled around an outage.
In short
The camera is the easy part. What makes a thermographic finding defensible is knowing what the surface is doing to the measurement, comparing like with like, and being honest about the readings that cannot be trusted.
If you want to see what that looks like on a real board, there are worked findings with both frames and the measured values on the electrical thermography page.