If you’ve worked around switchboards, transformers or large electrical plant in Australia, you’ll have seen the phrase temperature rise test on spec sheets and test reports. It sounds technical, but at heart the test is simple: we want to know how hot parts of the equipment run when carrying the currents they’re meant to carry. That matters because heat is the enemy of insulation, metal strength and long-term reliability — and because a surprisingly small thermal fault (a loose bolt, a poorly seated lug, restricted ventilation) is often the earliest warning sign of something that could become dangerous.

This article walks through the rules you’ll meet in Australia, the practical steps test labs use, the instruments and tricks of the trade, how to interpret results and the common traps I see on site. I’ll keep it practical — the sort of thing an electrician, switchboard assembler, site manager or design engineer can use straight away.

The basic idea: temperature rise vs absolute temperature

There are two related but different ways to talk about heat in electrical equipment:

• – Absolute temperature — the actual temperature measured at a point (e.g. 140 °C).
• – Temperature rise (ΔT) — the difference between that point and the ambient air around the equipment (e.g. 140 °C − 35 °C = 105 K).

Standards and declarations usually use temperature rise as the formal measurement because it factors out local ambient. In many switchboard and distribution contexts the key number you’ll see is a ΔT limit, rather than a bare maximum temperature.

What standards apply in Australia (short answer)

In Australia we work to the AS/NZS 61439 family for low-voltage switchgear and controlgear assemblies — this is the local adoption of IEC 61439 with Australian amendments. That standard sets the framework for temperature rise verification: where you place thermocouples, what counts as acceptable verification, and the three matched pathways for proving compliance (test, comparison to a reference design, or calculation under prescribed rules).

For transformers the primary international reference is IEC 60076-2 (temperature rise tests for liquid-immersed transformers), which is the method most Australian spec writers and labs use when testing power transformers. It defines acceptable measurement points and test methods (direct loading, back-to-back, simulated loss injection, etc.).

The two routes to compliance: test, comparison, or calculation

AS/NZS 61439 lays out three valid ways to demonstrate an assembly meets temperature rise requirements:

1. 1. By test — the most certain path: put the complete assembly (with covers and internal components in their final positions) on a test rig, run the specified current until temperatures stabilise, and measure with calibrated sensors.
2. 2. By comparison — show your design is essentially identical to a previously tested reference design (same materials, geometry, enclosure). This is common for modular systems.
3. 3. By calculation (assessment) — permitted in limited circumstances (for example certain ratings or compartment sizes) where validated thermal models and conservative assumptions are acceptable. Many assemblers will only use calculation for boards below specific current thresholds because modelling can be fragile in real life.

If you buy a pre-verified modular system from a major manufacturer you’ll often get “partially type tested assembly” (PTTA) documentation that makes life simpler — but beware: any modification to layout, ventilation, or components can invalidate the comparison and force a retest.

Typical numeric limits you’ll encounter (practical rules of thumb)

A few numbers come up repeatedly in Australian practice and manufacturer guidance — they’re worth committing to memory:

• – 105 K (ΔT) for copper busbars is the familiar design/limit figure (often quoted as “busbar permitted to operate at 105 °C above ambient” or “105 °C absolute at 35 °C ambient = 140 °C absolute”).
• – Where standards quote temperature rise limits relative to ambient you’ll also see the 70 K and 55 K practical rise checks referenced in industry materials. These are useful quick checks when sizing busbars or confirming vendor figures.

Important: those numbers are not “magic” — project specifications can and do require different limits (for example where life-cycle or derating needs are tighter). Always check the declaration of compliance and the assembler’s supporting test/calculation paperwork.

How a temperature rise test is actually done (step-by-step)

Below is a practical walk through a standard test sequence you’d find in a test lab running a switchboard or transformer test.

1. Preparation

• – The assembly must be complete: all internal wiring, covers, doors, busbar shrouds and accessories in place.
• – Select and label thermocouple positions per the standard or the test plan (main bus, joints, terminal connections, device terminals, and any manufacturer specified “hot spots”).
• – Calibrate thermocouples/RTDs and log calibration certificates. Record ambient conditions (room temp, any forced ventilation, barometric pressure if relevant).

2. Instrumentation

Thermocouples (type and calibration traceable to national standards) are the formal measurement method used for acceptance testing. Place them at standard positions — the standard tells you acceptable locations so results are comparable and repeatable.

3. Loading

Apply the rated or specified test current. For switchboards this is typically the rated continuous current for the bus or the current agreed in the test plan. For transformers it’s often rated current using direct or simulated methods. Run the current until readings reach steady-state (which could be hours for thermal mass and oil-cooled equipment).

4. Stabilisation & recording

• – Wait for steady-state: temperatures stabilise (small change over a set time). Log temperatures, currents and ambient continuously.
• – Calculate ΔT = measured temperature − ambient for each thermocouple and compare to limits.

5. Report

• – The test report should include thermocouple map, calibration certificates, ambient conditions, time-to-steady-state, raw data logs and a clear pass/fail statement tied to the applicable standard clause or project requirement.
• – If IR thermography was used it should be marked as supplementary diagnostic evidence, not a substitute for contact sensors.

Thermocouples vs infrared thermography — when to use which

Both measurement methods are valuable — but they’re different tools.

• – Thermocouples / RTDs measure contact/point temperatures and are the required method for formal acceptance testing. They give traceable, repeatable point readings at specified locations.
• – Infrared (IR) thermography is excellent for diagnostics — a quick survey to spot hot spots, poor contacts, or load imbalance before you attach sensors. But IR measures surface radiance, which depends on emissivity, viewing angle and surface condition; it often reports higher or different readings than a contact sensor.

In short: use IR for inspection, thermocouples for acceptance.

Transformer specifics — special considerations

Transformers are a bit of a different world because of oil, top-oil temperatures and winding hot-spot concerns.

• – IEC 60076-2 defines how to measure winding and oil temperature rise and gives several recognised test techniques: direct loading at rated current, back-to-back testing (where supply capacity is limited), and simulated loss injection (where losses are induced to reproduce heating). Each method has trade-offs in duration, accuracy and supply requirements.
• – For transformers you’ll commonly see hot-spot temperature calculations: the hottest point inside the winding is typically inferred from top-oil readings plus a manufacturer hot-spot factor (and sometimes by embedded fibre optic sensors in research or critical units).

Remember: transformer tests can take many hours or even days to reach true thermal equilibrium because of oil thermal inertia.

Common pitfalls and practical tips (what I tell clients)

1. 1. Don’t trust the “eye test.” A panel that looks neat can still have a marginal joint that runs hot under load. Always check documentation or insist on a thermal survey.
2. 2. Torque matters. Loose lug bolts and poorly prepared copper faces are a leading cause of localised hot spots. Confirm correct torque and joint preparation before testing.
3. 3. Ambient & ventilation control. Tests are highly sensitive to ambient and to airflow. If the test room has fans, ducts or other equipment nearby, make sure the test ambient and any “micro-ambient” around the busbars are recorded — sometimes an enclosed room creates a hotter microclimate that changes pass/fail outcomes.
4. 4. Harmonics and waveform shape. Non-sinusoidal currents increase heating (I²R losses) — if the field installation is expected to see high harmonic content, the test current should reflect that, or an allowance must be made.
5. 5. Documentation is king. AS/NZS 61439 requires a declaration of compliance and supporting evidence (test logs, calibration certificates, assumptions for calculations). If you’re buying equipment, ask to see the declaration and supporting test/calc paperwork.

Interpreting borderline results — what to consider

If a measured ΔT is close to the limit (say you measure 68 K when the reference is 70 K) don’t immediately panic — first:

• – Check sensor placement and calibration. Misplaced thermocouples or a broken TC often explain odd readings.
• – Verify ambient measurement and whether a local micro-ambient (inside the enclosure) might be higher than the room ambient recorded.
• – Re-check mechanical joints (torque, surface cleanliness). A small mechanical fix may drop a local hotspot several degrees.
• – Consider duty cycle and life-expectancy. Some projects choose to derate continuous currents or specify a lower allowable ΔT to improve insulation life.

A simple worked example

Say you test a main copper busbar. Ambient = 35 °C. Measured thermocouple on the bus = 140 °C.

• – ΔT = 140 − 35 = 105 K.
• – If the project busbar limit is 105 K, the board is at the limit and you should document this as “at limit” and investigate whether long-term operation at that level is acceptable or whether derating or design change is desirable.

What to expect in a test report (must-have items)

A robust temperature rise test report should include:

• – Test plan reference (standard clause or project spec), date and test lab details.
• – Full thermocouple map and sensor IDs plus calibration certificates.
• – Measured ambient (and how it was measured), test current, time to steady-state and raw logged data.
• – Calculated ΔT for each sensor and clear pass/fail statements against the specified limits.
• – Any IR images used for diagnostics (labelled and marked as supplementary).
• – Commentary on anything unusual and recommended follow-up actions.

If a supplier gives you only a short “pass” certificate without the supporting logs, ask for the full report — it’s important evidence for safety and warranty.

Practical checklist before you call a lab

• – Ensure the assembly is complete and in its final configuration (doors, covers, control wiring).
• – Clean and torque all bolted connections to manufacturer torque figures.
• – Agree test ambient and ventilation conditions with the lab.
• – Provide the lab with the applicable standard clause or project limit (e.g. AS/NZS 61439 clause X, busbar ΔT = 105 K).
• – Request the lab include raw data logs and calibration certificates in the report.

Designing for the test, not just to pass it

A final tip from experience: design and fabrication that only “passes a test” tends to be brittle. If you design for comfortable margins — cleaner jointing, slightly larger conductor size, better compartment cooling — you’ll reduce field issues, extend equipment life and avoid repeat testing. Where possible, choose modular, pre-verified building blocks from reputable manufacturers and keep good paperwork. AS/NZS 61439 gives you the framework — but the practical reliability comes from attention to mechanical joints, ventilation and honest verification.