Thermographic examination of electrical equipment

Practical advice, limit values, estimation methods for limit values

The thermographic condition assessment of electrical equipment is based on the fact that undersized or damaged conductors, poor connections (due to their increased transitional resistance), and in most cases, electrically faulty devices heat up to higher temperatures than usual (than allowed).

The biggest advantage of thermography is that measurements can be carried out from a safe distance - even on equipment operating at several kV - without influencing the operation of the device under test. Typical application areas include condition assessment of electrical equipment in the energy industry

• generators, transformers, insulators, switches, conductors, cables, connections, distributors...
• (in high, medium, and low voltage networks alike)
• Inspection of electrical equipment in production plants
• feed-in transformers, main and sub-distributors, switchgear cabinets, individual connections...
• Electrical machines, equipment in industry
• electric motors, phase correctors, frequency converters...

Electrical faults detectable by thermography

• undersized (overloaded) cables, switchgear, connections, transformers
• faulty (loose, corroded) contacts on any electrical equipment or device
• damaged cable heads, rings, surge protectors
• poor contacts on fuse springs, knife switches, conductor splices, cable connections, motor starters
• damaged circuit breakers and switchgear
• phase imbalances in three-phase supplies
• overheating due to overload or harmonics
• overheated (damaged or incorrectly adjusted) motor and generator brush assemblies
• clogged transformer and generator cooling systems
• faulty components, such as capacitors and power semiconductors

Important advice

• For high-voltage equipment, safety distance must be strictly observed - danger to life!
• Outdoor measurements (e.g., transformers and overhead lines) should be performed early in the morning or late in the evening (during periods without sunlight) to eliminate heat reflections and interfering background radiation.
• Especially with new equipment (as highly polished or finely machined metallic surfaces have very low emissivity), heat reflection may occur. To eliminate such measurement errors, it is advisable to perform measurements from multiple positions.
• If reflections make it impossible to obtain evaluable measurements, the surfaces to be measured can be dulled (with matte paint) or adhesive tape with known emissive properties can be applied - all of this can only be done in a de-energized state.
• In order to detect fault locations that become dangerous under nominal (or operational or "full") load during measurements, the measurements should be carried out under operational load. The power load of electrical components increases with the square of the current, leading to higher temperatures.
• The enclosed parts of switchgear and distribution equipment cannot be measured directly. Thermographic measurements cannot be performed through plexiglass or other plastic windows. If possible, these should be removed (in a de-energized state!) before the measurement.
• When measuring small objects or from a long distance (e.g., on overhead lines), the geometric resolution provided by the thermographic device (thermal camera) must be taken into account. At least three pixels should fall on each measurement surface for accurate temperature data evaluation. If necessary, the geometric resolution capability should be adjusted to the object size/distance by using appropriate optics (special telephoto lens).

Figure: Fault under cover [source: PIM]

Figure: Contact fault [source: PIM]

Figure: Faulty cable lug [source: PIM]

Figure: Loose contact [source: PIM]

Practical limit values

The question often arises where to draw the line between the perceived heating being considered faulty or even dangerous. It is particularly difficult to decide on this matter when the thermographic survey was conducted at a current significantly lower than the maximum load. To avoid this problem, it is advisable to accept that thermographic condition assessments of electrical equipment should only be carried out when at least 50% of the nominal load is present. Only very severe faults can be detected even at 30% load. Accepted limit values and decision rules (at least 75% load) Limit values compared to ambient temperature Heating* <20°C: okay Heating* <40°C: check Heating* >40°C: urgent check Heating* >60°C: critical (* above ambient temperature) Limit values for differences between phases Deviation** <5°C: okay Deviation** <20°C: check Deviation** >20°C: urgent check Deviation** >40°C: critical (** between phases) Limit values depending on insulation material Rubber-insulated cables: max. 60°C *** PVC-insulated cables: max. 70°C *** Silicone-insulated cables: max. 180°C *** (*** absolute temperature values) Other limit values Electric motors (measured on cooling fins): depending on the type, max. 60 ...80°C *** Plastic covers: depending on the material: max. 50 ... 75°C *** Magnetic switches: typically max. 85°C *** Transformers: typically max. 85°C *** (*** absolute temperature values) Note: Under lower loads, the limits below all the above values are valid.

Figure: Unevenly loaded busbars (reflection on non-dusty surfaces!) [source: PIM]

Estimation of heating under nominal load

Heating occurs due to the transient resistance of the tested electrical device (cable, busbar, etc.) or contact, resulting in energy loss in the form of heat. The device then dissipates this power through thermal radiation, convection (towards the air), and heat conduction towards the connected elements. Since the expected heating under higher loads (either current or voltage) at the time of measurement is influenced by the load- and temperature-dependent resistance change of the device or contact in question, as well as the expected stronger heat conduction, thermal radiation, and convective heat dissipation, determining the resulting heating would require very complex mathematical relationships. In a steady state in a rail or conductor assumed to be infinitely long, the expected temperature can be calculated based on the following equation:

Since this equation is quite complex for practical application due to the difficult-to-access material properties, we recommend using the following estimation instead:

Simplified heating estimation for nominal load (Eric Rahne's estimation) Assuming that the temperature increase between the measured load and the nominal load is so small that the material-specific resistance temperature change (increase) can be neglected, and neither the current crowding factor nor the heat transfer factors change significantly due to the observed device temperature increase, then all these factors can be considered constant in the previous equation. (This can be applied in cases of temperature increases of a few tens of degrees Celsius. However, for larger temperature changes, these simplifications can lead to significant errors, so the equations below CANNOT be applied.) By applying the above simplifications, the following equation is obtained:

The estimated absolute temperatures can be compared with the limit values listed on the previous page.

Figure: Transformer [source: InfraTec]

Figure: Substation inspection [source: InfraTec]

Figure: Overhead line with fault [source: InfraTec]

Important: When inspecting substations and overhead lines, all conditions for outdoor measurements must be strictly followed. Measurements can be conducted at night (preferably under a cloudy sky) or during the day in very thick, completely closed - but rain-free - cloud cover. (Based on experience, we prefer conducting measurements at night.) Also, remember that we mostly aim to determine the temperature of metallic objects, which have significant heat radiation-reflection capabilities, without physical contact. To minimize the effects of interfering radiation, it is best to measure when such radiation is absent. (For example, helicopter inspections of overhead lines conducted under conditions requiring clear daytime visibility - i.e., sunny weather - are pointless unless the goal was sightseeing.)

From a measurement technology perspective, geometric resolution is critical: for measuring the transient resistance increase of even 18 mm diameter ropes at a height of 30 m and their connections, a geometric resolution of 0.2 mrad (or even better) is required, hence a large telephoto lens is necessary. Some faults in insulators may still be detectable with a weaker geometric resolution.

Rahne Eric (PIM Ltd.) pim-kft.hu, gepszakerto.hu

Contact

The content of this publication is protected by copyright. Any (even partial) use, electronic or printed republication is permitted only with the indication of the source and the author's name, and with the author's prior written permission. Violation of copyright will result in legal consequences.