Transformer Core Engineering Notes · CF-N253
A short load peak is not thermally identical to the same load maintained for hours. Nevertheless, calling a peak “short” does not establish that it is acceptable. Its effect depends on the starting temperatures, duration, cooling response and temperature limits of the actual transformer, including local components that a simplified bulk-oil model may not represent.
Interpret a time constant correctly
For a simple first-order model following a step, one time constant corresponds to about 63% of the change toward the new equilibrium, not complete thermal stabilization. A real transformer can require several interacting thermal responses rather than one constant. Distinguish the winding response from the bulk-liquid response and state the model used. A generic number copied from another unit is not evidence for the transformer’s actual transient performance.
Include the initial thermal state
The same peak can start from a cool, lightly loaded unit or from a transformer already near its operating limit. These are different events. Retain the preceding load and ambient history or use a justified conservative initial condition. A calculation initialized at ambient temperature after continuous operation can understate the predicted temperatures. The input record should show why the starting state is credible.
Resolve repeated peaks and cooling delays
Repeated peaks may arrive before the transformer has cooled from the previous event. Cooling controls can also introduce delay between a load change and the establishment of the assumed airflow or oil flow. Preserve the event sequence in the simulation instead of replacing it with unrelated average values. Check that the available time resolution captures the relevant changes without implying that every small sampling fluctuation is a physical load event.
Evaluate more than one thermal output
Report maximum winding hot spot, relevant liquid temperatures and the chosen ageing assessment, together with any other component limits identified by the OEM. State uncertainty in model parameters and loading measurements. A peak that looks acceptable in a top-oil trace can still require review elsewhere. The result should be an approved operating envelope for a defined scenario, not a universal rule that all peaks shorter than a chosen duration are harmless.
| Review item | Evidence to request | Boundary to preserve |
|---|---|---|
| Response model | Separate winding and liquid dynamics | One time constant does not mean full stabilization |
| Starting state | Preceding load and ambient history | A hot transformer cannot be initialized as cold |
| Peak sequence | Duration, repetition and cooling transitions | Averages can conceal critical sequences |
References and scope
- IEEE C57.91-2025 — Loading Mineral-Oil-Immersed Transformers and Step-Voltage RegulatorsPublic standard abstract · Mineral-oil loading, thermal duty and loading beyond nameplate; do not transfer the scope automatically to ester or dry-type equipment.
- IEC 60076-7:2018 — Loading guide for mineral-oil-immersed power transformersPublic standard abstract · Mineral-oil transformer loading and thermal-ageing framework, not a universal asset-life prediction.
- IEEE C57.165-2024 — Temperature Monitoring of Liquid-Immersed TransformersPublic standard abstract · Temperature-monitoring methods for liquid-immersed transformers and their measurement/model boundaries.