Low-Emissivity Busbar Joint: The 61 °C That Wasn't
A 400 V distribution board is scanned at 62% of rated load. One phase of a 630 A bolted busbar joint reads 61.0 °C; the two adjacent phase joints read 33.0 °C. The busbars are bare tinned copper, ε ≈ 0.10. The enclosure interior is 30 °C and the camera's reflected apparent temperature is set to 30 °C. Ambient is 28 °C. The client's acceptance limit is 70 °C absolute for busbar joints. The camera is a 320 × 240 uncooled microbolometer (8–14 µm) with a 25° lens at 1.2 m; each joint is about 40 mm wide. The thermographer left the camera emissivity at the default 0.95 and concludes that 61 °C is below 70 °C, so the board passes.
1. Is the thermographer's pass conclusion defensible? Correct the 61.0 °C reading using the camera's assumed emissivity and the actual surface emissivity, and state the assumptions.
Read the worked answer
No. With ε set to 0.95 but an actual ε of about 0.10, the camera assumes most of the received radiance is emitted by the metal. In reality, about 90% is reflected enclosure background. Using the camera's grey-body inversion in the 8–14 µm band as a σT⁴ approximation:
L/σ = ε_assumed·T_disp⁴ + (1 − ε_assumed)·T_refl⁴ T_disp = 61 + 273 = 334 K; T_refl = 303 K L/σ = 0.95(334⁴) + 0.05(303⁴) = 0.95(1.244 × 10¹⁰) + 0.05(8.43 × 10⁹) = 1.224 × 10¹⁰ K⁴
Actual surface: 0.10·T_true⁴ + 0.90(303⁴) = 1.224 × 10¹⁰ 0.10·T_true⁴ = 1.224 × 10¹⁰ − 7.59 × 10⁹ = 4.65 × 10⁹ T_true⁴ = 4.65 × 10¹⁰ T_true ≈ 464 K ≈ 191 °C
For an adjacent joint at 33 °C displayed: T_true ≈ 328 K ≈ 55 °C. So the apparent 28 K difference corresponds to a true difference of roughly 135 K - far above the 70 °C limit. Assumptions: uniform ε, correct reflected temperature, negligible atmospheric/optics effects at 1.2 m. Even with large uncertainty, the joint cannot be passed on an uncorrected 61 °C reading.
2. Before any pass/fail decision is issued, what measurement steps and report wording would you require?
Read the worked answer
Do not issue a pass. If safe and production allows:
- Apply a thin patch of high-emissivity tape or matte paint to the suspect joint and to one adjacent joint. Set camera ε to the patch value and re-scan at the same 62% load.
- Measure reflected apparent temperature properly, e.g. with a crumpled aluminium-foil target, rather than assuming it.
- Verify load current, ambient temperature, distance, lens and spot size; record all of them.
- Confirm with a contact probe on the joint if accessible.
If access is not possible, report the joint as an unconfirmed anomaly requiring immediate investigation, with the corrected estimate, its uncertainty and the limitation that the absolute 70 °C limit is not meaningful for bare low-ε metal without correction. Recommend shutdown inspection and re-torque.
3. In plain language, explain to the client why a shiny busbar can read cooler than it is.
Read the worked answer
The camera does not measure temperature directly; it measures infrared radiance. Bare tinned copper emits very little infrared and reflects most of its surroundings like a mirror. With the camera set for a high-emissivity surface, it assumes most of the signal is emitted by the metal and reports a temperature far below the truth. The 61 °C reading is mostly a reflection of the 30 °C enclosure; the joint is actually close to 190 °C. That is why a shiny conductor needs a high-emissivity target or a contact measurement before any absolute temperature judgement is made.