9.3 Capacitor, Bus, and Connection Anomalies
Key Takeaways
- Most electrical IR “findings” are connection problems: loose lugs, oxidized joints, poor crimps, fretting hardware, and weak fuse clips—driven by I²R heating
- Always compute ΔT to a similar phase or similar joint under similar load before applying NETA P1–P4 priorities
- Capacitor bank surveys compare cans/units and fuses/phases; a failed or open unit often runs cooler than healthy peers under voltage, while connection heating still looks hot at interfaces
- Fuse clips and disconnect contacts are high-value targets because spring tension and plating wear create progressive resistance
- Correct emissivity on shiny bus and hardware or use reference methods—false ΔT from mixed surfaces is a classic Level II catch
Connection Anomalies: The Core of Electrical Thermography
If you only mastered one electrical IR skill, master connection assessment. Industry experience and exam scenarios alike concentrate on bolted joints, compression lugs, bus splices, fuse clips, and stab interfaces because they are numerous, accessible (relatively), and fail by increasing contact resistance.
Physics You Must Keep Visible: I²R
Power dissipated in a resistive interface is P = I²R. Implications for Level II:
- Heating is highly load-dependent. Double the current and, for the same resistance, heating power quadruples.
- A defect that is invisible at 15% load may be obvious at 60% load—and severe at 100%.
- Reporting a ΔT without load context understates or overstates risk. Note percent of rated or measured amperes.
- After repair (re-torque, clean, replace lug), re-inspect under comparable load to verify the hotspot is gone—not merely that the panel door is closed again.
| Current (relative) | Relative I² heating (same R) | Practical note |
|---|---|---|
| 0.4 × rated | 0.16 × | Near minimum survey guidance; small faults may hide |
| 0.7 × rated | 0.49 × | Common productive survey region |
| 1.0 × rated | 1.00 × | Highest sensitivity; watch absolute temperature limits |
| 1.2 × rated | 1.44 × | Overload territory—operations issue as well as IR |
Loose Lugs and Bolted Joints
Thermal signature: Heat localized at the lug barrel, pad, or bolt head; conductor a short distance away is cooler if the wire itself is sound. Sister phases at the same load are cooler when their joints are healthy.
Common mechanisms: under-torque, thermal cycling fretting, improper washer stack, aluminum-copper transition issues, strand breakage under the lug, contamination under the pad.
Field method:
- Identify the hottest phase joint and the best similar reference joint.
- Measure ΔT with correct ε/RAT/focus.
- Confirm load balance with current measurement when available.
- Assign P1–P4 from this guide’s bands (P4 1–10 °C, P3 >10–20 °C, P2 >20–40 °C, P1 >40 °C).
- Recommend mechanical/electrical remediation (de-energized torque to manufacturer values, replace damaged hardware, infrared recheck).
Do not recommend “just tighten hot” on energized gear as a thermographer improvisation—remediation is a controlled electrical maintenance task.
Oxidized and Contaminated Joints
Oxidation and fretting films raise resistance without the joint looking dramatically “loose” to a casual glance. IR often finds these before a shutdown inspection would. Patterns resemble loose lugs: interface-centric heat. Visual follow-up under outage may show discoloration, pitting, or powdery oxide.
Humidity, chemical plants, and outdoor gear accelerate this class of faults. Trend ΔT over routes; a joint climbing from 6 °C to 18 °C ΔT across quarters at similar load is a P3 escalation story even if absolute temperatures still look “survivable.”
Fuse Clips and Fuse Bodies
Fuse clips lose spring tension and plating over time. Signatures:
- Hot clip fingers with a relatively cooler fuse center (contact problem).
- Entire fuse body hot versus peers (element or connection issues—correlate with load and fuse type).
- One phase fuse hot, others normal → single-phase contact or loading issue.
Compare like fuse types and like load. A larger ampere-rated fuse on one pole is not a similar component for casual ΔT.
Bus Bars, Bus Duct, and Flexible Links
Long bus runs allow both joint anomalies and section heating:
| Pattern | Interpretation focus |
|---|---|
| Hot only at joint plates / splice | Connection resistance |
| Entire bus section elevated vs parallel bus | Overload, harmonics, or insufficient bus rating/cooling |
| Hot flexible braid or expansion joint | Broken strands, poor termination of braid |
| Periodic hot spots along bolted covers | Cover ground/bond or hardware issues (context-specific) |
Shiny plated bus is a radiometric trap. Left at ε = 0.95, polished metal often reads incorrectly and reflections from nearby hot gear can paint false hotspots. Use high-ε reference targets on non-critical areas when authorized, measure RAT carefully, and trust relative phase comparison only when surfaces are comparable—or after correction.
Capacitor Banks
Power factor correction and harmonic filter banks add special patterns:
Healthy bank: Similar can/unit temperatures under the same voltage and environmental exposure; fuse indicators normal; phase currents reasonable for the bank design.
Failed / open capacitor unit: Frequently cooler than energized healthy peers because it is no longer drawing its share of reactive current (open fuse or internal open). A cold can among warm cans is a diagnostic clue—not automatically “best unit.”
High-resistance connection on a capacitor circuit: Still presents as a hot interface (lug, fuse clip, bus link) even though the can body patterns differ from motor terminals.
Unbalanced phase heating on bank feeders: Investigate blown fuses, open units, and connection faults together—capacitor banks often fail unit-by-unit and shift current among phases.
| Capacitor observation | Likely direction | Caution |
|---|---|---|
| One can much cooler than peers under voltage | Open/failed unit or open fuse | Confirm bank is energized; cool ambient wind can also cool edge cans |
| One can much hotter than peers | Internal unit stress, poor ventilation, or measurement artifact | Correlate with fuse status and manufacturer limits |
| Hot fuse clip on one unit | Connection / clip resistance | Classic I²R fix path |
| Whole bank hot enclosure | Ambient, sun, lost ventilation, or overload of duty | System-level, not single lug |
Safety note: capacitors store energy. Switching and discharge procedures belong to qualified electrical workers—IR survey planning must respect stored-energy rules even when the camera is non-contact.
Similar-Phase ΔT Discipline (Exam Gold)
When the question stem gives temperatures for A/B/C joints:
- Pick the reference (usually the coolest healthy peer or average of two healthy peers—be consistent and state it).
- Compute ΔT.
- Map to P1–P4.
- Only then discuss root cause hypotheses.
Example: Bus link temperatures at 62% load: A = 48 °C, B = 71 °C, C = 49 °C. ΔT_B ≈ 22–23 °C → P2. Currents balanced → connection on B. Currents show B low → possible open downstream with heating elsewhere—still investigate, but logic changes.
Mixed-Surface and Reflection Errors
Before writing a severe priority on bus or capacitor hardware:
- Are finishes matched (painted vs bare)?
- Is sun loading one side of an outdoor bank?
- Is a shiny surface reflecting a nearby hot transformer or heater?
- Is the camera reporting a reflection as a target temperature?
Level II reviewers reject findings that ignore these artifacts. A quick check: change viewing angle slightly; a true I²R hotspot usually stays on the metal, while many reflections move.
Documentation That Maintenance Can Use
Good findings packages include: clear photo pair (IR + visual), equipment and phase ID, load, ambient, parameter list (ε, RAT, distance, window transmittance if used), temperatures and ΔT, NETA priority, and a specific recommended action (“replace B-phase fuse clips and re-terminate lug; re-IR at ≥40% load”).
Avoid vague closings like “check connections someday.” Priorities exist so work orders can be scheduled.
Mini Case Set
Case 1 — Oxidized lug. ΔT = 16 °C at 50% load, balanced current → P3, clean/reterminate at next outage; consider earlier action if the circuit is critical.
Case 2 — Fuse clip. ΔT = 42 °C → P1, urgent risk discussion with operations; do not normalize because “fuses run hot.”
Case 3 — Capacitor. One can 12 °C cooler than bank average, matching blown blown-fuse indicator → failed/open unit replacement workflow, not a NETA hot-connection P-priority on the can body.
Case 4 — False ΔT. Bare copper reads 20 °C “hotter” only until ε is corrected with tape reference; true ΔT falls to 3 °C → P4 monitor, not emergency.
Quick Answer: Treat loose lugs, oxidized joints, bus splices, and fuse clips as I²R problems scored by ΔT to a similar phase under similar load (P4 1–10, P3 >10–20, P2 >20–40, P1 >40 C°). Capacitor banks add the twist that failed open units often run cooler, while bad clips and lugs still run hot. Fix emissivity and reflection errors before you escalate priorities.
A bolted bus joint on phase A is 19 °C hotter than identical joints on phases B and C at similar load. Per the Experience-Based criteria (Infraspection Standard §10.1), the priority is:
Why can a failed capacitor can appear cooler than neighboring energized cans on the same bank?
Resistive connection heating power scales with current approximately as:
Which practice BEST reduces false “hot bus” findings on shiny plated bus bars?