Transformer Core Engineering Notes · CF-N217
A lightning impulse does not necessarily divide evenly along a transformer winding. The winding behaves as a distributed electrical network, with capacitance and inductance influencing the initial distribution and later oscillations. Dividing the terminal impulse voltage by the number of turns is therefore not a sufficient basis for checking local insulation stress.
Separate initial distribution from later response
At the beginning of a steep-front event, capacitive paths strongly influence how the voltage appears across winding sections. As the response develops, inductance and damping also shape the voltages. A design review should state which time-dependent quantities were assessed: terminal voltage, section voltage or turn-to-turn stress. A single static picture may miss the maximum value occurring later in the waveform. Model detail should be appropriate to the insulation feature being evaluated.
Connections are part of the network
Leads, taps, winding sections and nearby conductive structures influence the network that the impulse sees. Changes to section connections or grading features can therefore alter local voltage distribution while leaving the nominal turns ratio unchanged. Ask whether the model includes the connections actually used for the relevant service and test conditions. A reference winding with a similar rating may have different capacitances and cannot automatically establish the new winding’s transient behaviour.
Treat a diagram as a concept unless data are supplied
A curve labelled “impulse distribution” should identify its winding arrangement, waveform, time and normalization. Without these inputs, it is only a conceptual illustration. Do not use a generic exponential curve to specify insulation dimensions for a real transformer. The useful deliverable is a traceable analysis with its assumptions and applicable test evidence. This article’s review matrix is not a measured voltage profile or a substitute for the OEM’s transient calculation.
Make changes trigger a focused review
When taps, leads, shields or section arrangements change, identify which parameters of the original transient model have changed. Recheck the affected insulation regions and confirm the continuing applicability of test evidence. The agreed dielectric test programme should be based on the equipment requirements, not copied from a generic article. A routine resistance or ratio test can detect some manufacturing issues, but it does not directly establish impulse voltage distribution along the winding.
| Review item | Evidence to request | Boundary to preserve |
|---|---|---|
| Time response | Initial and later section-voltage behaviour | A static division by turns is insufficient |
| Network | Actual sections, taps, leads and screens | A similar rating does not imply the same capacitance |
| Evidence | Traceable model and applicable impulse tests | Conceptual curves are not measured data |
References and scope
- IEC 60076-3:2013+AMD1:2018 CSV — Insulation levels and dielectric testsPublic standard abstract · Insulation levels, dielectric testing and external clearances; the article does not reproduce test values or design distances.
- IEC 60076-5:2006 — Ability to withstand short circuitPublic standard abstract · Thermal and dynamic short-circuit capability; calculation and verification apply to the actual design.