Two transformers that are each perfectly healthy can still be impossible to parallel. If their vector groups differ, or their ratios do not correspond across the tap range, the closed loop between them carries a circulating current the moment they are connected, and neither unit has a fault. Transformer vector group testing before parallel operation is the check that prevents that situation being discovered at energisation.

The verification has four parts: ratio across the tap range, vector group and phase displacement, phase rotation, and an assessment of how load will divide between the units. The first three are measured; the fourth is calculated from the equipment data.

Why parallel operation needs verification

When two transformers are operated in parallel, their secondary windings are connected together. Any difference between the voltages they produce drives a current around the loop formed by the two secondaries, limited only by the combined impedance of the two units.

A small ratio difference produces a circulating current proportional to that difference, which consumes capacity and heats the windings without delivering any load. A vector group difference produces a much larger effective voltage difference, because the phase displacement adds a component that no ratio adjustment can remove.

Neither condition shows up in a single-unit test. Each transformer can pass every routine measurement and still be unsuitable for paralleling with the other, which is why the verification is a comparison exercise rather than a set of individual measurements.

Ratio verification across the tap range

The ratio comparison has to be made at every tap position that the pair will use, not only at the nominal tap. The tap changers of two units are normally operated together in parallel, so a correspondence has to exist between their positions.

The practical output is a table for each transformer, with the measured ratio at each tap position and the resulting deviation from the declared value. The two tables are then compared tap by tap, and the correspondence between the tap positions established.

Where the two transformers have different numbers of taps, the comparison covers the range over which both can operate. Where the tap changer direction convention differs between the units, that has to be established explicitly, because a pair operated with inverted conventions will diverge rather than track.

Vector group confirmation

The vector group defines the phase displacement between the primary and secondary windings and the internal connection arrangement that produces it. It is established during manufacture and recorded on the nameplate, and the test confirms that the unit delivered corresponds to the nameplate it carries.

The confirmation is a measurement of the phase relationship, not a reading of the nameplate. A transformer that has been rewound, reconnected after a repair, or fitted with replacement bushings can carry a nameplate that no longer describes its connections, and the test is what makes that visible.

For a pair intended for parallel operation, the two units must have the same vector group. Where they do not, the options are to operate them separately, to reconfigure one internally where the design allows it, or to install an appropriate phase-shifting arrangement. The test establishes which situation applies before the decision is made rather than after.

Polarity and phase rotation checks

Polarity confirms the direction convention between the windings, and phase rotation confirms the sequence in which the phases reach their peaks. Both affect the connection of a parallel pair, and both are checks that are frequently assumed rather than measured.

Rotation deserves particular attention where the two transformers are supplied from different sources, or where one has been in service and the other is new. The rotation at the point of connection is what matters, and it can differ from the rotation at the source if the intervening arrangement changes it.

The measurement is straightforward: a phase sequence indicator, or a set of phase-to-phase and phase-to-neutral voltage measurements taken together, establishes the sequence at the point where the parallel connection will be made.

Matching impedances and load sharing

The impedance parameters do not prevent paralleling; they determine how the load divides. A pair with markedly different impedances shares load in inverse proportion to those impedances, so the lower-impedance unit carries more and reaches its rating first. The assessment establishes how much of the combined capacity is actually usable.

Documenting results before energisation

The documentation should contain the ratio table for each unit across the tap range, the vector group and phase displacement measurement for each, the rotation check at the point of connection, the tap correspondence table and the impedance data used for the load-sharing assessment.

It should also record the instrument used and its calibration status, the test voltage, and the connection arrangement for each measurement. Those fields are what allow a reviewer to reproduce the test and to check that the comparison was made on a consistent basis.

Where the two units are of different manufacture or different vintage, the documentation should note that explicitly, because the comparison is then between units whose design data may not be directly comparable. Manufacturing tolerances that are unremarkable within one design can produce measurable differences between two.

Common connection errors found at this stage

The commonest finding is a vector group that does not match the other unit, either because the units are of different types or because one has been reconnected in a way the nameplate does not reflect.

The second is a phase rotation mismatch, usually found where the two units are supplied through different arrangements or where one has been connected after a cable or busbar modification.

The third is a tap correspondence error, where the tap changers of the two units operate in opposite directions or where one unit’s tap numbering does not follow the same convention. This is discovered by comparing the ratio tables, and it is corrected in the control arrangement rather than in the transformer.

The fourth is a polarity error on a single-phase unit or on one winding of a three-phase unit, which produces a phase displacement that does not correspond to any nameplate vector group.

Sign-off for parallel operation

The sign-off should state what was verified, with the measurements attached, and identify any condition attached to the paralleling decision. Where the load-sharing assessment shows that only part of the combined capacity is usable, that limitation belongs in the sign-off rather than in a separate analysis.

Where the two units are to be operated together on an ongoing basis, the sign-off should also state the tap positions at which the pair may be operated and the control arrangement that keeps them corresponding. That is an operating instruction derived from the test results, and it belongs with them.

Where the decision is to operate the units separately, the sign-off should record why, because the reason is what prevents the question being reopened without evidence. The transformer characteristics that the comparison is made against are defined in IEC 60076-1, the instrument transformer requirements that apply where the comparison involves measurement transformers are in IEC 61869-2, tap changer mechanism documentation is published by the changer manufacturers such as Maschinenfabrik Reinhausen, method documentation is available from instrument suppliers such as OMICRON, and utility practice is coordinated through CIGRE study committees with the research background published by EPRI. The ratio test range is grouped under transformer ratio test equipment.