Most electrical maintenance is still done one of two ways: on a calendar, or after something breaks. Thermography sits between the two, and it changes the economics of both.
1. It finds faults while the plant is still running
The defining feature of an infrared survey is that nothing is switched off. The camera reads radiated energy from a distance, so the boards stay live, the site keeps producing, and the inspection happens around normal operations rather than instead of them.
That matters more than it first appears. The reason so much electrical equipment runs to failure is not that nobody cares — it is that the only way to inspect it properly used to involve an outage, and outages are expensive enough that they get deferred.
Remove the outage requirement and inspection becomes something you can do annually without a business case each time.
It also means the fault is inspected in the condition that produces it. A loose termination is only hot when current is flowing. Under load is the only state in which it is visible at all, which is why boards must be carrying at least 40% of normal load during a survey.
2. It produces measurements, so work can be ranked
A visual inspection produces opinions. A thermographic inspection produces numbers.
Every finding in a competent report carries the measured temperature of the anomaly, the temperature of a reference component doing identical work, and the difference between them. That difference maps onto a priority rating with a defined repair timeframe:
| Priority | ΔT vs reference | Recommended action |
|---|---|---|
| 1 — Critical | > 30 °C | Serious abnormality. Immediate repair recommended — failure may be imminent. |
| 2 — Significant | 20 – 30 °C | Should be corrected — could fail suddenly. Repair within one week. |
| 3 — Caution | 10 – 20 °C | Minor problem. Rectify when convenient, then re-inspect for the trend. |
| 4 — Checked | < 10 °C | No abnormality at inspection. Re-inspect at regular intervals. |
The value of that is not the individual finding. It is that twenty findings across a site arrive pre-sorted. Your electrician spends the budget on the connection that is 34 °C over its neighbour, not on the one that happens to be nearest the door.
It also names the location. The report identifies the board, the component and the terminal, so one connection gets opened rather than a whole panel.
3. Repeated, it becomes prediction rather than detection
A single survey finds faults. A programme finds patterns, and patterns are where the real money is.
Once you have two or three surveys of the same assets, several things become visible that no single inspection can show:
- A connection that is warming year on year is deteriorating, even if it has not yet crossed a threshold. That is a scheduled repair, booked at your convenience.
- A defect appearing repeatedly across a population — the same component type failing in different positions on different boards — is a batch problem, not a series of unrelated faults. It gets replaced as a batch, once.
- A board that stays clean for years earns a longer interval, and the budget moves to the assets that need it.
This is the difference between condition monitoring and inspection. The first survey is a snapshot. The third one is a maintenance strategy.
We have a site on a three-monthly thermographic and ultrasound round where exactly this happened: switched fuses running hot kept appearing in different positions across successive visits, with ultrasound confirming arcing behind them. Not one bad unit — a population of them, quietly cooking. They have since been replaced as a batch. A single inspection would have generated one work order.
Where it fits with everything else
Thermography is not a complete condition-monitoring programme on its own. It sees heat, which means it sees resistive faults extremely well and sees some other things not at all.
Airborne ultrasound hears what infrared cannot see — arcing, tracking and corona discharge on switchgear, and bearing and lubrication faults on rotating plant. The two run together on the same visit, and each covers the other's blind spot.
Making the case internally
If the argument you have to make is financial rather than technical, the cost of late detection calculator will do it with your own numbers: it prices the same defect at four points along its progression using your repair cost, your downtime rate and your hours lost, and shows what an inspection programme buys you against it.
For what the inspection itself involves, see the electrical thermography page.