Low-Emissivity Busbar Anomaly in a 13.8 kV Substation
A Level III reviews an electrical survey at a 13.8 kV substation. A technician imaged bare copper busbar connections with the camera emissivity set to 0.95 and reported a 47 °C apparent temperature rise on one phase relative to the other two. Ambient air was 31 °C, the sky was clear, and the oxidized copper had a measured emissivity of 0.62 in the 8–14 µm band. The technician measured reflected apparent temperature as 18 °C using a diffuse reflector. The camera has a 320×240 detector, a 24° lens, and an IFOV of 1.3 mrad; the connection was 7 m away and the spotmeter was placed on a 25 mm wide lug. All three phases carried 420 A on a 600 A rated bar, within 2% of each other. A white painted enclosure wall at 44 °C is visible directly behind the connection. The technician did not use a contact probe or change viewing angle.
1. Is the reported 47 °C rise a valid basis for severity classification? Justify your answer using emissivity, reflected apparent temperature, spot size, and load.
Read the worked answer
Not as reported. The 47 °C value was produced with ε = 0.95 on a surface whose emissivity is about 0.62. Because the camera assumed a higher emissivity than the target actually has, it attributed too much of the received radiance to target emission and will generally understate the true surface temperature for a target warmer than the reflected background. The reported rise is therefore not a conservative value. The reflected apparent temperature of 18 °C is lower than the expected busbar temperature, so sky reflection is not the dominant error, but the 44 °C painted wall behind the connection can add reflected radiance if viewed at the wrong angle. Spot size is also marginal: 1.3 mrad at 7 m gives a 9.1 mm spot, so a 25 mm lug is only about 2.7 pixels across. That is below a reliable 3×3 measurement field of view and allows background and adjacent metal to influence the reading. Load is reasonably balanced, so phase comparison is useful, but the absolute rise must be recalculated with correct emissivity, reflected temperature, viewing angle, and adequate pixel coverage before it can drive a severity class.
2. Outline the measurement procedure the technician should have used before reporting the anomaly.
Read the worked answer
The technician should first verify the target material and emissivity, using a contact thermometer or a known reference for the oxidized copper, and enter the correct ε rather than a default 0.95. Reflected apparent temperature should be measured with a diffuse reflector at the same angle and vicinity, not assumed. The camera should be positioned to avoid the 44 °C wall in the reflected direction, or the angle should be changed and the result compared. The 25 mm lug should fill at least a 3×3 pixel array; that means getting closer, using a telephoto lens, or using a close-up lens so the spot is no more than about one-third of the target width. Focus must be correct and the image should not rely on digital zoom for spatial resolution. The technician should confirm steady load and compare the same phase conductors, same material, and same viewing geometry. A contact probe or clamp meter reading should corroborate any rise that will be classified. Only after these steps should the corrected ΔT be compared with program criteria and reported with the measurement conditions.
3. How should the final report express the result and its uncertainty?
Read the worked answer
The report should not present 47 °C as a definitive temperature rise. It should state the measured apparent rise, the emissivity used, the reflected apparent temperature, the distance, lens, IFOV, spot size relative to the lug, and the fact that pixel coverage was marginal. It should state that the value was corrected or re-measured, and give the corrected rise with an uncertainty statement, for example ±2 °C or ±2% of reading, whichever is greater, plus the environmental and emissivity contributions. If a contact reading was obtained, report it separately and explain any difference. The report should describe the load at the time of inspection and whether it was stable. It should classify the anomaly only after correction, and should recommend a repeat measurement with adequate spatial resolution and a changed viewing angle. The conclusion should distinguish confirmed overheating from measurement uncertainty, and should note that the white wall was a potential reflection source. This gives the client an auditable basis for maintenance priority rather than a single unsupported absolute temperature.