Case study 02 · Example
Example Example – illustration built from open-licensed images and simulated data, not a real client.
01 / Starting point
A small production hall runs one or two shifts: a press, a conveyor, a compressor. The machines are fed from a row of switchgear cabinets along the back wall. One feeder carries 160 A per phase for about 3,000 hours a year (assumption).
Nothing looks wrong: no trips, no smell, everything runs. Maintenance is in next year’s plan. The question is whether the row contains a connection that still works but is already getting hot – and if so, where.
02 / Measurement
Electrical thermography only makes sense under load: without load there is no heating, so no fault shows.
We agree which cabinets are covered, when normal operating load is present and which authorised electrician will open the covers.
We measure during normal operation from a safe distance. Your authorised electrician opens the covers; we never touch live parts.
Per cabinet: a thermogram and an ordinary photo of the same view, load current and operating state, ambient temperature, the emissivity setting.
ΔT against similar components and against ambient, A/B/C priority, recommendation. If we find a deviation we say so on site instead of waiting for the report.
03 / Heat map
A simulation: the geometry is simplified and the temperatures are invented. Rotate the view, switch between thermal, visual and findings, and hover a surface for its temperature.
3D view is not available on this device – the image is a still of the simulated heat map.
Rotatable, with temperature read-out
04 / Finding
Real, open-licensed thermograms of loose or overloaded electrical connections. The images are unmodified; the temperatures come from each image’s own scale and markers. They were taken with different cameras under different conditions, so they are not comparable with one another.
A clear illustration of how a report is built: the visual photo sits next to the thermogram and the marker shows which component it is. Two spots around the faulty connection read 99–112 °C, while the third marker (A3) reads 36 °C.
Poor cable termination: the conductor and terminal reach about 108 °C while the neighbouring conductors stay around 35–60 °C.
One phase connection and its cable are visibly hotter than the other two – the textbook case for comparing with similar components.
A three-pole breaker mounted side by side without spacing dissipates heat poorly – not a loose connection but a heat-removal problem, yet it shows up just the same.
The typical picture of a loosely tightened or poorly crimped connection: one outlet is clearly warmer than the reference, the other is raised too.
A motor warms up in operation – that is normal (scale 80–133 °F, about 27–56 °C). The task is to tell expected warmth from a deviation.
| Comparison | Against what | Why it helps | Watch out for |
|---|---|---|---|
| ΔT vs a similar component | The neighbour with the same load and type – e.g. the terminals of the other two phases | Load and ambient largely cancel out; in practice the strongest indicator | If all three phases are equally faulty there is no good reference |
| ΔT vs ambient | The cabinet or hall air temperature | A simple, easily understood number | Load-dependent: at a lower current the fault looks smaller |
| Absolute temperature | The limit allowed by the component manufacturer | A temperature above the limit is a problem in itself | Apparent temperature also depends on surface emissivity and reflections |
In the illustrative example the loose terminal is +38 K above the neighbouring phase terminal and about +50 K above the hall air.
Our report gives A/B/C priorities: for an A item we tell you on site and the fix should not wait for the next maintenance; B goes into the next planned maintenance window; C means watch and re-scan. The band limit is not a single threshold but the combination of load, the manufacturer’s limit and the ΔT against similar components. ΔT-based classifications are common practice, but guidelines and manufacturers give different limits – so the report always names the basis of the decision.
05 / Engineering assessment
Two separate questions: how much energy does the faulty connection waste, and how large is the risk. The first can be calculated; the second cannot be priced.
P = n · I² · Rextra = 1 · (160 A)² · 0.5 mΩ = 12.8 W
| Case | Power | Energy | Cost, central (range) | Local rise (illustrative) |
|---|---|---|---|---|
| One phase | 12.8 W | 38 kWh/yr | HUF 5,100/yr (HUF 4,000–5,900) | 13–51 K |
| All three phases | 38.4 W | 115 kWh/yr | HUF 15,000/yr (HUF 12,000–18,000) | 38–154 K |
Estimate, not a guarantee. Net of VAT. Local rise = P · R_th with R_th = 1–4 K/W: an illustration, not a measurement.
Why does 12.8 W show on a thermogram? Because the heat is generated in a contact area of a few square centimetres and can hardly escape. At a thermal resistance of 1–4 K/W, 12.8 W means roughly 13–51 K of local rise. The simulated finding (about +50 K above hall air) sits at the top of that range – an advanced state. The heating may also speed up oxidation of the contact surface, so the fault can feed itself.
The wasted energy is worth a few thousand forints a year, so it is not the point. The point is that the terminal can get worse with time – and that a thermogram shows it before the consequences do.
06 / Recommended action
Priority A the loose terminal · Priority B the cable run
We report the A item to your electrician on the day of the measurement – it does not wait for the report.
An authorised electrician isolates the circuit, inspects the terminal, re-torques or replaces it, and checks the neighbouring terminals and the cable insulation.
The moderate warming along the cable leaving the terminal is also a reason to check the cable’s load rating and bundling (priority B).
After the fix we measure the same point at a comparable load again: the thermogram shows whether the ΔT is gone.
07 / What it means in money
| Item | Amount, central (range) |
|---|---|
| Wasted energy (one phase) | HUF 5,100/yr (HUF 4,000–5,900) |
| Terminal re-torque + checks | HUF 45,000 (HUF 25,000–90,000) |
| Replace terminal / contactor | HUF 120,000 (HUF 60,000–250,000) |
| Thermographic cabinet screening | HUF 12,000 per cabinet with a package; stand-alone HUF 18,000 per cabinet, minimum HUF 60,000 |
Estimate, not a guarantee. Repair costs are indicative, quotation-based assumptions; the actual price depends on the local contractor.
The real stake is failure, downtime and fire risk. We deliberately do not put a monetary figure on it: we have no reliable source for downtime cost and will not invent a number.
One sourced fact: of 736 fires investigated in Hungary in 2022, 73 (9.9%) had electrical energy as the cause. It is the national disaster-management (OKF) sample, all building types together – there is no breakdown for business premises. Source (OKF, PDF)
The point
An early warning, not a savings project.
A fault of a dozen watts shows on a thermogram at once – to the eye or to a breaker it does not yet. The fix is small next to the consequences – but we cannot say what the consequences are worth, and we promise no savings.
08 / Limits
This example consists only of open-licensed photographs, our own simulation and the data of the worked calculation example. CC BY-SA images are shown unmodified (only scaled for the web).
Available as a quick diagnosis or as part of a larger survey. Briefly tell us how many cabinets are involved – we reply by the next working day at the latest.