9.2 Transformer and Switchgear Thermography
Key Takeaways
- Inspect transformer bushings, tank/cooling surfaces, radiators/fans, load-tap equipment, and external connections as a system—not as isolated “hot spots”
- External connection heating (pads, spades, bus links) is interpreted differently from internal winding or oil-circuit clues seen only as tank/radiator pattern changes
- Switchgear focus points include bus joints, breaker stabs/primary disconnects, cable lugs, fuse clips, and CT/PT secondary paths where accessible and safe
- Slight bushing warmth from dielectric losses can be normal; asymmetric bushing heating or rapid change versus peers is not automatically normal
- Safety first: respect approach boundaries, prefer IR windows, and never trade a thermogram for an arc-flash incident
Scope of Transformer and Switchgear IR at Level II
Transformers and switchgear concentrate energy, risk, and diagnostic value. A Level II thermographer is expected to know where to look, what patterns mean, and what must never be sacrificed for a better camera angle. This section is about pattern interpretation and survey strategy; absolute IEEE/NEMA temperature limits and formal NETA table language are reinforced in the severity chapters, while approach and PPE details continue in §8.4.
Transformer Survey Map
Think of a liquid-filled power transformer as several thermal systems sharing one nameplate:
| Zone | What you can usually see with IR | What anomalies often mean |
|---|---|---|
| HV/LV bushings | Porcelain/polymer surface and terminal caps | Connection heating at pads/spades; abnormal dielectric loss heating; tracking contamination patterns |
| Tank walls | Surface temperature distribution | Load heating, internal fault heat migration, circulating current effects, cooling inefficiency |
| Radiators / coolers | Vertical temperature gradient, fan influence | Plugged radiators, failed fans/pumps, oil flow issues |
| Conservator / accessories | Localized external hardware | External connection or accessory issues (context-specific) |
| Cable boxes / bus ducts | Joints and terminations | Classic I²R connection faults |
| LTC / DETC areas (external) | Compartment exteriors, linkages if visible | Abnormal heating may warrant specialist electrical follow-up—do not over-claim internal LTC diagnosis from one exterior spot alone |
Load and time matter. Transformer thermal mass is large. Capture under representative load after the unit has approached thermal steady state when possible. A snapshot five minutes after a major load step can understate severity.
Bushings: Normal vs Abnormal
Condenser (capacitor-type) and other high-voltage bushings can show slight elevation relative to tank steel even at light load because of dielectric losses related to applied voltage (loss mechanisms scale with voltage and insulation dissipation factor, not only with load current). That mild, relatively symmetric warmth can be expected behavior—not an automatic P1.
Treat as abnormal when:
- One bushing is substantially hotter than its phase peers under the same voltage and similar load.
- Temperature is trending upward rapidly across inspections without a matching load/ambient explanation.
- Terminal cap / pad connections are hotter than the bushing body in a classic connection pattern (heat localized at the bolted interface).
- Visual signs accompany heat: oil weeping, cracked porcelain, heavy contamination, tracking paths.
| Pattern | Prefer interpretation | Follow-up |
|---|---|---|
| All bushing tops slightly > tank, similar to each other | Possible normal dielectric heating | Baseline and trend; no panic language |
| One bushing 20+ °C above peers | Asymmetric anomaly — connection or insulation concern | Elevate priority; electrical testing / outage planning |
| Hot only at bolted pad, body cooler | External connection I²R | Torque/repair connection; NETA ΔT vs sister phases |
| Hot bushing + oil leak / damaged insulator | Compound failure risk | Immediate engineering/operations attention |
Emissivity caution: Glazed porcelain, silicone sheds, and metal caps differ. Use consistent methods (reference emitters on metal hardware when critical) so “hot bushing” is not a coatings artifact.
Cooling System Patterns
Radiators and fans convert internal losses into rejectable heat. Useful observations:
- Even, progressive cooling along radiator panels under load often indicates flowing oil and working surface area.
- Cold radiator sections while the tank is hot can indicate blocked valves, sludge, closed radiator valves left after maintenance, or lost flow paths.
- Failed fans (forced-air units) produce warmer tank/oil indications at the same load versus historical baselines.
- Compare like load and like ambient before declaring a cooling defect; a 10 °C ambient swing can move surface temperatures without a new internal fault.
Document fan stage (ON/OFF), oil temperature gauges if visible, and load MVA or percent. Cooling findings often need operations confirmation—not just a red pixel.
Load Tap Changers and External Clues
Load tap changing (LTC) equipment can develop contact wear and mechanism issues that produce heating. From the IR camera’s external viewpoint you may only see compartment exterior or accessible linkage/cable interfaces. Professional reporting language:
- Describe what was measured (location, ΔT, load, ambient).
- Recommend qualified electrical investigation rather than claiming a specific internal contact ID you cannot see.
- Note whether the unit was tapping or holding a fixed tap during the survey if known.
Switchgear and MCC Focus Points
Metal-clad and metal-enclosed switchgear, switchboards, and MCCs are connection-dense. Prioritize:
- Breaker primary disconnects / stabs — high-current interfaces; fretting and misalignment create classic hot stabs.
- Bus joints and splices — bolted and clamped joints; oxidized or under-torqued hardware.
- Cable lugs and terminations — including compression lug quality and strand damage.
- Fuse clips and fuse bodies — clip tension loss, single-phase fuse issues.
- Contactor and starter poles — contact wear patterns phase-to-phase.
- Neutral and ground connections where loaded or carrying imbalance/harmonic current.
Breaker stab pattern: A single hot stab finger or cluster with peers cool, under load, strongly suggests interface resistance. Confirm the breaker is closed and carrying current; an open racked breaker is not a valid “healthy cool” comparison for a loaded cubicle.
External vs Internal Clues
| Clue type | Examples | What Level II should say |
|---|---|---|
| Direct external connection | Hot lug, hot splice plate, hot stab visible through window | Location-specific I²R anomaly; NETA ΔT vs similar phase/joint |
| Indirect internal indicator | Unusual tank hotspot, asymmetric radiator, hot compartment exterior without visible joint | “Thermal anomaly consistent with internal heating; recommend electrical diagnostic testing / internal inspection under outage” |
| Environmental / installation | Solar loading on one outdoor cubicle face, process heat from nearby boiler | Control the survey (time of day, shade, open doors only if permitted) and note limitations |
Over-claiming “turn-to-turn short” from a single exterior tank image is an ethics and technical problem. Under-claiming a 40 °C stab ΔT as “slightly warm” is equally wrong.
Geometry, Distance, and IR Windows
Medium- and high-voltage gear often forces long working distances. Remember:
- IFOV / measurement spot size must fit the target (stab, pad, fuse clip). If the spot is larger than the target, temperature is diluted toward background.
- IR windows (viewing ports) allow closed-door surveys when transmittance is known and entered; Level II corrects for window transmittance rather than ignoring it.
- Angle matters on cylindrical bushings and shiny bus—favor near-normal viewing when safe.
Safety Distance and Energized Work Discipline
Thermography does not create an exception to electrical safety rules. High-level expectations for this chapter:
- Know limited / restricted / prohibited approach boundaries from the facility’s arc-flash study and NFPA 70E framework (exact distances are task- and equipment-specific—do not invent generic footages on the exam as universal constants).
- Prefer closed-door inspection through IR windows or mesh when the design allows meaningful data.
- If covers must be open, that is energized electrical work requiring qualified persons, appropriate PPE, permits as required, and a plan—not “just five seconds for a photo.”
- Maintain stable footing, one-hand discipline where required by procedure, and never use the camera as a reason to lean past the safe boundary.
Worked Scenarios
Scenario A — Hot stab. Through an IR window, B-phase primary disconnect on a loaded 4160 V breaker is 33 °C above A and C at 55% load. Classification: P2 by ΔT bands; recommend expedited outage for inspection/cleaning/refurbishment of the disconnect interface; do not wait for “next annual only” language if plant risk policy is stricter.
Scenario B — Bushing peers. All three HV bushings read 6–8 °C above tank, within 2 °C of each other, steady versus last year’s baseline at similar load. Interpretation: consistent with mild dielectric heating / normal pattern; document baseline, no emergency.
Scenario C — Cooling. Tank unusually hot for load; one radiator bank remains near ambient while others are warm; inlet valve found closed after maintenance. IR supports a cooling path finding, not a winding redesign.
Scenario D — Unsafe request. Supervisor asks you to remove an arc-resistant door alone, without PPE, to “get a better emissivity.” Decline, escalate through safety channels, propose window install or scheduled outage inspection.
Report Elements Specific to This Equipment
Include equipment ID (transformer serial / switchgear cubicle), voltage class, load (A or % or MVA), ambient, cooling stage, whether survey was through a window (with transmittance), phase comparison table, ΔT and priority, and clear separation of observed external heating versus suspected internal condition requiring electrical testing.
Quick Answer: Survey transformers for bushings, tank/cooling, external connections, and accessible tap equipment; survey switchgear for stabs, bus joints, lugs, and fuse clips. Symmetric mild bushing warmth can be normal dielectric behavior; asymmetric or connection-localized heat is not. External ΔT drives NETA priorities; internal clues drive cautious language plus electrical follow-up. Safety distance, IR windows, and PPE always outrank the perfect image.
Slight, nearly equal warming of all three HV bushing tops above the transformer tank under voltage, without rapid change from baseline, is MOST consistent with:
A loaded medium-voltage breaker shows one primary disconnect stab 48 °C hotter than the other phases through an IR window. Using the Experience-Based criteria (Infraspection Standard §10.1), the priority is:
Which statement BEST describes Level II language for an unusual hot region on a transformer tank wall with no accessible external joint at that location?
What is the PRIMARY reason to prefer permanently installed IR windows for energized switchgear surveys when they provide an adequate view?