Two windings can have the same outside diameter and different leakage impedance. One may contain a different radial conductor distribution, a changed interwinding gap or shorter active height. The external envelope is only a partial description of the field that stores leakage energy. [1]
For a rewind, preserve the original impedance evidence before comparing geometry. Record the tap, frequency, reference temperature, rated power basis and whether the report separates resistance and reactance. A change in percentage impedance can otherwise reflect a reporting basis rather than a changed winding.
Map changes where leakage energy is stored
Leakage flux does not link every winding turn equally. It occupies the winding region and surrounding gaps, with end effects determined by the actual geometry. Changes in relative position, current distribution and active height can alter that field even when turns ratio is restored.
Create a sensitivity register before releasing the replacement winding drawing. It identifies the change and its likely mechanism; it should not assign a universal percentage correction. The actual magnitude requires a suitable electromagnetic model or a justified comparison supported by measurements.
| Geometry change | Leakage-field question | Companion effect |
|---|---|---|
| Interwinding radial gap | Has the principal field region widened? | Dielectric and cooling interface |
| Winding active height | Has the axial current distribution changed? | End forces and heat flow |
| Axial displacement | Are end fields and symmetry altered? | Unequal mechanical forces |
| Disc/section spacing | Is current distribution redistributed? | Local electric stress |
| Lead and tap routing | Are extra field paths introduced? | Local heating and support loads |
| Conductor arrangement | Does effective current distribution remain comparable? | AC resistance and loss |
Use a common impedance basis
For a three-phase transformer, a convenient base impedance is Zb = VLL²/S, with line-to-line voltage VLL in volts and three-phase apparent power S in volt-amperes. Per-unit impedance is the measured equivalent impedance divided by this base, using a consistent winding reference. A changed declared rating changes the percentage value even if the physical ohmic impedance is unchanged.
Separate the resistive component from reactance where the evidence supports doing so. A new conductor material or joint arrangement can shift resistance without the same change in leakage geometry. Conversely, a displacement can alter reactance while a resistance check remains satisfactory.
Compare ohmic impedance referred to the same winding before comparing percentages. Keep the resistance/reactance convention with the turns-ratio reference so a changed rating base does not appear as a changed leakage field. [2]
Use one base for leakage-impedance comparison
Assumptions: a three-phase 10 MVA, 11 kV line-to-line base gives Zb = 11,000²/10,000,000 = 12.1 Ω. Original equivalent impedance 0.726 Ω is 6.0% on this base, but would be reported as 7.5% on a 12.5 MVA base at the same voltage without any physical change. Suppose the actual rewind result, referred to the original winding/tap/frequency, is 0.750 Ω. That is 6.198% on the original base and a physical increase of 3.306% over 0.726 Ω. Against an assumed project ±2% relative impedance-change gate, the upper bound is 0.74052 Ω and the rewind fails by 0.00948 Ω. Hold acceptance and investigate the component split and winding geometry; correcting the reporting base does not erase this residual change. [1][2]
References
- U.S. Bureau of Reclamation, FIST Volume 3-31, Transformer Diagnostics, June 2003.
- U.S. Department of Energy, DOE-HDBK-1011/4-92, Electrical Science, Volume 4, 1992.