In switchgear, the insulation usually is not what fails first — the current-carrying connection is. A bolted busbar joint that slowly loses contact pressure heats up, accelerates its own degradation, and can end in a flashover inside the cubicle. The good news: that process announces itself as temperature, months ahead, if you are measuring the joint.
1. Why busbar joints overheat
Current through a bolted joint dissipates heat in proportion to the joint's resistance:
A sound joint has very low, stable R_joint. Several mechanisms drive it up over time:
- Bolt-tension relaxation & creep: thermal cycling and metal creep loosen the joint, reducing contact pressure and the real metal-to-metal contact area.
- Oxidation: aluminium in particular forms a hard, insulating oxide layer; copper oxidises more slowly but still raises resistance, especially when hot.
- Fretting: micro-movement under load/thermal cycling abrades plating and creates oxide debris at the interface — a known killer of aluminium joints.
- Al–Cu intermetallics: dissimilar-metal joints without proper bimetallic transition form brittle, resistive intermetallic compounds.
- Workmanship: incorrect torque (under or over), missing anti-oxidant compound, missing Belleville/spring washers, contamination.
2. From hot spot to arc-flash
Left unmanaged, a degrading joint progresses: elevated ΔT → insulation/support degradation nearby and conductor annealing → partial discharge or tracking → arcing → arc-flash inside the enclosure, with the associated safety and asset-loss consequences. Because much of this happens behind closed panels, it is invisible to routine visual inspection — which is exactly why continuous, in-enclosure temperature measurement matters.
3. Detection methods — what works inside switchgear
The hard constraint in switchgear is access: the critical joints are energised and enclosed. Methods differ mainly in whether they can see inside, and whether they are continuous.
| Method | Continuous? | Sees inside enclosure? | Notes |
|---|---|---|---|
| IR thermography via IR windows | No (periodic) | Only what the window frames | Useful for surveys; limited field of view; needs scheduled access and a trained thermographer |
| Thermal indicator labels | Visual only | Yes, at the dot | Cheap one-shot; shows a threshold was crossed, no trend, no remote alarm |
| Fixed IR sensors | Yes | Line-of-sight only | Cannot see behind barriers/insulation boots |
| Wireless contact temperature sensors | Yes | Yes — mounted on the joint | Direct measurement, per-joint, remote alarm/trend; needs HV-rated, EMI-immune design |
For the bolted joints, breaker contacts and cable terminations that actually fail — and that sit behind panels and insulation boots — direct-contact wireless sensors are the only option that is both continuous and able to measure the real point. The historical objections are power (batteries across many points) and EMI; energy-harvesting designs remove the battery, and purpose-built RF handles the field.
4. Reading the data: when is a joint actually in trouble?
Absolute temperature alone misleads. Apply the same discipline used across thermal CBM (full method in our CBM & RBM guide):
- Temperature rise vs limit: compare ΔT above ambient to the standard rise limit for the joint material/coating — IEC 62271-1 / IEEE C37.20.x (commonly ~50 K rise for bare copper/aluminium, higher for silver/nickel-coated; confirm per equipment).
- Normalise to load: rise scales with I², so judge a joint at comparable current, or normalise — a "warm" joint at 40% load may be critical at full load.
- Phase comparison: on a balanced feeder the three homologous joints should track within a few kelvin; one consistently hotter phase is a near-certain defect, independent of calibration.
- Rate-of-rise: a load-normalised ΔT climbing week-over-week signals degradation even below any absolute limit — often the earliest actionable flag.
5. Prevention & corrective action
When a hot spot is confirmed, the fix is usually at the joint — but do it right:
- De-energise and inspect the flagged joint (the monitoring tells you which one, saving blind panel-by-panel work).
- Clean and restore surfaces: remove oxide, apply the correct anti-oxidant/joint compound on aluminium.
- Re-torque to specification with a calibrated wrench; fit Belleville/spring washers to hold contact pressure through thermal cycling.
- Repair dissimilar-metal joints with proper bimetallic transition washers/plates.
- Verify the fix: after re-energising, confirm ΔT returns to baseline at load — proof the repair worked, and a new reference for trending.
This closed loop — detect, locate, fix, verify — is what converts raw temperature into reliability, and it underpins both condition-based and risk-based maintenance programs.
6. Where VTI fits
VTI self-powered wireless temperature sensors are designed for exactly this duty: clamp-mounted on busbar joints, isolator contacts and cable lugs inside MV/HV switchgear, harvesting energy from the conductor (no battery), installed live-line, and transmitting over an EMI-immune link to a gateway that feeds your monitoring software, SCADA and CMMS.
See every joint before it fails
Continuous, in-enclosure, battery-free temperature monitoring for switchgear — installed live-line, integrated to your systems.
Request the technical datasheetFrequently asked questions
What causes busbar joints to overheat?
Rising contact resistance from bolt-tension relaxation, oxidation (especially aluminium), fretting, dissimilar-metal intermetallics, and workmanship issues such as incorrect torque or missing joint compound. Higher resistance raises heating (P = I²R), which accelerates the degradation in a self-reinforcing loop.
Can you detect a hot spot inside closed switchgear?
Periodic IR thermography only sees what an IR window frames and requires scheduled access; fixed IR needs line of sight. Direct-contact wireless sensors mounted on the joint measure the real temperature continuously, even behind panels and insulation boots, and raise remote alarms.
What ΔT should trigger action?
Frame thresholds from the standard rise limit for the joint material/coating (IEC 62271-1, IEEE C37.20.x) as a rise above ambient, normalised to load, then refine against the commissioning baseline. Combine with phase comparison and rate-of-rise rather than a single absolute number.
How do you fix a hot busbar joint?
De-energise, clean surfaces, apply anti-oxidant on aluminium, re-torque to specification with a calibrated wrench, fit spring washers, and repair dissimilar-metal joints with proper bimetallic hardware. Then verify ΔT returns to baseline at load.