Dual-LV Solar Transformers: Specify Every Pairwise Impedance and Its Base

A dual-LV solar transformer needs a pairwise impedance schedule, not one percentage copied from a nameplate. Specify HV–LV1, HV–LV2 and LV1–LV2 results, the MVA and voltage base for each, tap position and the state of the unused winding during measurement. Then require a model that reproduces those results on a common base before the collector protection and load-flow studies are released.

1. Identify the electrical arrangement first

Dual LV can mean electrically separate windings with a deliberate phase relationship, not interchangeable parallel terminals. Obtain the winding diagram, vector groups, terminal identification, rated power allocation, neutral treatment and restrictions on simultaneous loading. Ask whether each inverter group can operate when the other group is unavailable.

The broad renewable-transformer guide discusses topology selection. This article instead prevents an ambiguous impedance guarantee from entering the contract. IEC 60076-1:2011 supplies general transformer scope; pairwise project requirements must be agreed in the detailed specification.

2. Write a test matrix before requesting bids

Also request measured load loss or resistance information so the model can separate resistive and reactive effects. Identify the reference temperature and tested tap. A percentage without its base is not a complete result. The Eaton transformer fundamentals describes the purpose of impedance and resistance testing, rather than this project's particular matrix.

3. Worked illustrative base conversion

Assume a 10/5/5 MVA transformer with 33 kV HV and two 0.69 kV LV windings. Suppose both HV–LV pair impedances are 8% on a 5 MVA base, and LV1–LV2 is 14% on the same 5 MVA base. These are illustrative values, not recommended universal impedances.

For unchanged voltage bases, conversion to a 10 MVA common base multiplies each per-unit impedance by 10/5. Thus the pair results become 16%, 16% and 28%. If voltage bases differ, their squared ratio must also be included; simply doubling is then insufficient.

This scalar construction treats the illustrative pair values as reactances and neglects resistance. On that basis, the star-equivalent branches reproduce the three pairwise values. Arbitrary impedance magnitudes cannot be added and subtracted to obtain exact complex branches; use complex impedances when loss and phase-angle data are available. These are mathematical model parameters, not physical independent coils. A negative branch in another equivalent is not automatically a manufacturing defect; investigate base consistency and the validity of the chosen representation.

4. Show why the base error matters

At 0.69 kV, the 10 MVA impedance base is 0.04761 ohm. The correct pair impedance magnitude is 0.0076176 ohm, whereas the incorrect one is 0.0038088 ohm. The mistaken model can materially distort voltage-drop and fault calculations. It does not follow that a real inverter fault current doubles: converter limits, external grid strength, winding topology and fault location remain essential.

5. Study the cross-coupling cases

Run balanced full operation, one LV group disconnected, unequal inverter output, reactive support and the permitted charging case if storage shares the installation. Review voltage regulation at both LV terminals and current sharing only where parallel operation is expressly designed.

Fault studies must distinguish an HV-fed LV fault from inverter contribution across the LV pair. Do not apply the HV–LV percentage to every path. Harmonic behaviour also requires frequency-dependent data; a power-frequency pairwise matrix is not a complete harmonic model. The collector harmonic-scan guide explains why cable and converter equivalents matter. AEMO's harmonic guidance supports detailed reticulation modelling in its jurisdiction.

6. Accept evidence against the exact design

The RFQ should require guaranteed pair values, agreed tolerances, voltage/power bases, temperature, tap schedule and a test diagram. At factory acceptance, compare all reported pairs with the agreed schedule and verify that the study model reproduces measured values. Resolve discrepancies before shipment or explicitly revise the approved design.

Archive the winding diagram, serial-numbered test report, conversion sheet and model revision together. Confirm what routine tests cover and contract any additional pair measurements; do not invent a universal standard clause from a public abstract. Operational restrictions must reach plant controls and protection settings, not remain only in the purchasing correspondence.

7. Frequently asked questions

Can one uk% describe every fault path? No. A multiwinding transformer has distinct pairwise paths.

May separate LV windings be paralleled freely? No. Phase displacement, ratios, impedance and manufacturer restrictions must permit it.

Does the matrix establish harmonic performance? No. It provides power-frequency data; frequency-dependent modelling is separate.

8. References

• IEC: general transformer scope
• Eaton: transformer tests
• AEMO: detailed harmonic model requirements