9.2 Conducting and Evaluating Thermographic Surveys (NETA Section 9, Table 100.18)

Conducting and Evaluating Thermographic Surveys (NETA Section 9, Table 100.18)

Quick Answer: NETA Table 100.18, "Thermographic Survey Suggested Actions Based on Temperature Rise," grades every finding two ways — ΔT against a similar component under similar loading, and ΔT against ambient air temperature. The similar-component comparison is the more reliable of the two because it automatically cancels out load, ambient temperature, and time of day.

---

1. The severity criteria

Two points about the structure of this table that questions are built on:

• The two columns are not interchangeable. A 12 °C rise is a probable deficiency when measured against a similar component, but only a possible deficiency when measured against ambient. Reporting which comparison basis was used is therefore mandatory, not optional.
• The similar-component column has no "monitor" row. That intermediate action level exists only in the ambient comparison, because ambient comparison is the weaker method and the standard provides a graded response before jumping to immediate repair.

NETA's own note qualifies the table: temperature specifications vary by equipment type, heating is generally related to the square of the current so load has a major impact on ΔT, and "in the absence of consensus standards for ΔT, the values in this table will provide reasonable guidelines." It is guidance, not an absolute pass/fail — the equipment's own temperature rating always governs where it is known.

2. Choosing the comparison basis

Similar component under similar loading is preferred whenever it is available:

• Three phases of the same bus, the same breaker, or the same cable set.
• Identical parallel units — two transformers on the same bus, two identical feeders.
• The same connection on an adjacent identical cubicle.

It works because ambient temperature, solar loading, wind, and camera settings are common to both the target and the reference, so they subtract out. This is why the three-phase comparison is the workhorse of electrical thermography — the reference is right there in the same image.

Ambient comparison is used when no similar component exists — a single transformer bushing of its type, a unique connection, an isolated device. It is weaker because ambient is not a like-for-like reference and the reading now depends on load, cooling, and the equipment's own normal operating temperature.

3. Load correction

The severity ladder assumes the equipment is at representative load. Because rise scales with the square of current, a finding taken at partial load understates what will occur at full load. The scaling relationship is:

$\Delta T_{\text{full}} \approx \Delta T_{\text{measured}} \times \left(\frac{I_{\text{rated}}}{I_{\text{measured}}}\right)^{2}$

Worked example. A connection shows a 5 °C rise over its sister phases while the feeder carries 300 A on a 600 A circuit.

$\Delta T_{\text{full}} \approx 5 \times (600/300)^2 = 5 \times 4 = 20\ ^\circ\text{C}$

At the measured load the finding sits in the "probable deficiency, repair as time permits" band. Projected to full load it exceeds 15 °C and becomes a major discrepancy requiring immediate repair. Recording load current at the time of the survey is what makes this correction possible, and it is why NETA lists load conditions among the mandatory report contents.

Treat the projection as an estimate, not a measurement — but a finding at light load on a circuit whose load is expected to grow deserves elevated attention, not a lower priority.

4. Survey workflow

1. Plan the route. List the equipment, the specific inspection points at each, and the access method — door open, infrared window, or existing viewport. Sequence by criticality and by access.
2. Confirm conditions. Equipment energized, load at or above the practical minimum, and load current recorded. Note wind, solar exposure, and indoor ventilation.
3. Establish the safety condition. Opening an energized enclosure is energized work: arc flash PPE per the equipment label, an energized work permit where the facility requires one, and a second person where policy dictates. Infrared windows exist to remove this exposure and should be used where installed.
4. Set the camera: emissivity per the target surface, reflected background temperature, distance, and — above all — focus.
5. Scan systematically. Work the same order every time so the survey is repeatable and comparable year over year.
6. Capture radiometric images, not screenshots. A radiometric file stores the per-pixel data and permits re-analysis with corrected emissivity and background later; a plain image does not.
7. Capture a matching visible-light photograph of each finding. A thermogram alone rarely lets a repair crew identify which of six identical lugs is the problem.
8. Record the reference. Whether the comparison is a sister phase or ambient, the reference temperature is part of the finding.
9. Note every area you could not see. Inaccessible and unobservable areas are explicitly required to be identified, because a survey report that is silent about them implies coverage that did not occur.

5. The report — NETA Section 9 required content

A NETA-compliant thermographic report identifies:

1. The area of concern — precisely enough for a crew to find it.
2. The reference area used for comparison.
3. The temperature difference between the area of concern and the reference.
4. The probable cause of the temperature difference.
5. The areas inspected, identifying inaccessible and unobservable areas and equipment.
6. The load conditions at the time of inspection.
7. Photographs and thermograms of the deficient area.
8. A recommended action for repair.

Items 4 and 8 are where technician judgement shows. "Hot spot at 22 °C ΔT" is data. "22 °C ΔT on the B-phase load-side lug of feeder breaker 12, referenced to A and C phases at identical load; probable cause is a loose or corroded compression lug; recommend de-energize, inspect, clean, and re-torque to manufacturer's value, then re-survey under load" is a finding a maintenance planner can act on.

6. Follow-up and closing the loop

• Re-survey after repair, under load. A repair is not verified until the thermal signature has returned to normal. This is the step most often skipped.
• Cross-check with contact resistance. A DLRO measurement on the repaired joint during the outage quantifies what the camera qualified.
• Trend across surveys. A connection that appears at 3 °C one year and 8 °C the next is deteriorating even though neither reading alone is alarming. Year-over-year comparison at similar load is the strongest evidence thermography produces.
• Feed the findings into the maintenance program. A recurring hot spot at the same location after repeated re-torquing points to a design or material problem — aluminium-to-copper interface without the correct compound or connector, an undersized lug, a thermal cycling issue — not a workmanship problem. Recognizing that pattern is a Level IV analysis skill.

Exam trap: A question reports a 12 °C ΔT and asks for the recommended action, giving both comparison bases among the answers. Against a similar component 12 °C is "probable deficiency; repair as time permits." Against ambient it is only "possible deficiency; warrants investigation." Without knowing the comparison basis the question cannot be answered — which is exactly why NETA requires the reference area to be reported.