Home › Guides › Transformer Vector Groups

Transformer Vector Groups Explained: Letters, Clock Numbers and Why They Matter

Every transformer nameplate carries a short code that looks like alphabet soup to the uninitiated: Dyn11, YNd11, YNyn0, Dyn5. That code is the vector group, and it is one of the most important pieces of information on the whole nameplate. It tells you how the windings are connected, whether a neutral is available, and — most critically — the phase relationship between the primary and secondary voltages. Get the vector group wrong when connecting, paralleling or replacing a transformer and the consequences range from circulating currents and overheating to a catastrophic fault on energisation. This guide decodes the letters and numbers, explains why phase displacement matters, and sets out the rules for parallel operation.

Vector group verification is also a standard part of commissioning, covered in our power transformer testing and commissioning guide. The theory here and the test procedure there belong together: this guide explains what you are proving, the testing guide explains how you prove it. For the voltage and ratio calculations that go with vector group work, use our free transformer calculators.

The letters: how each winding is connected

A vector group code starts with two letters, one for each winding. Capital letters describe the high-voltage winding and lowercase letters the low-voltage winding. The letter D or d means delta: the three winding ends are connected in a triangle, with no neutral point. The letter Y or y means star (also called wye): one end of each winding is joined at a common neutral point, and the other ends form the three phases. The letter Z or z means zigzag (also called interconnected star): each phase winding is split into two halves wound on different core limbs and connected in a way that gives a star-like output with special properties for unbalanced loads and earth faults.

An additional letter n or N after a star letter means the neutral point is brought out to a terminal and is available for use — for earthing, for single-phase loads, or for connecting a neutral conductor. So in Dyn11, the capital D tells you the high-voltage winding is delta, the lowercase y tells you the low-voltage winding is star, and the n tells you the low-voltage neutral is brought out. In YNd11, the high-voltage winding is star with its neutral brought out, and the low-voltage winding is delta with no neutral. Reading the letters in order, high-voltage first and low-voltage second, is the first step in decoding any vector group.

The number: the clock-face notation

The number at the end of the code describes the phase displacement between the high-voltage and low-voltage voltages, using the analogy of a clock face. Imagine the high-voltage phase voltage pointing at 12 o'clock, as the minute hand of a clock. The low-voltage phase voltage points at the hour indicated by the number. Each hour on the clock represents 30 degrees, because a full circle is 360 degrees divided by 12 hours. So a vector group number of 0 means the two voltages are in phase, pointing the same way. A number of 6 means they are 180 degrees apart, pointing in opposite directions. And the very common number 11 means the low-voltage voltage lags the high-voltage voltage by 330 degrees — or, equivalently, leads it by 30 degrees, since 330 degrees of lag is the same position on the circle as 30 degrees of lead.

That last point causes more confusion than any other part of vector groups, so it is worth being precise. Displacement is conventionally stated as the angle by which the low-voltage voltage lags the high-voltage voltage, measured in the direction of rotation. For Dyn11, the low-voltage voltage lags by 11 times 30 degrees, which is 330 degrees. On a phasor diagram that puts the low-voltage voltage 30 degrees ahead of the high-voltage voltage if you measure the smaller angle, but the code always uses the lag convention, so 11 it is. The practical consequence is what matters: two transformers can only be paralleled if their low-voltage voltages are in phase with each other, which means their vector group numbers must be compatible, as the parallel operation section below explains.

Dyn11: the distribution workhorse

Dyn11 is the most common vector group for distribution transformers, and understanding why explains a lot about how distribution networks are designed. The delta high-voltage winding has no neutral, which is fine because high-voltage distribution networks are normally three-wire systems with no neutral loads. More importantly, a delta winding provides a closed path for third-harmonic currents: the magnetising current of a transformer contains harmonics, and the third harmonic and its multiples circulate inside the delta instead of flowing into the supply and distorting the voltage waveform. Without that delta path, those harmonics would appear in the supply voltage and could disturb other equipment.

The star low-voltage winding, with its neutral brought out, is exactly what a low-voltage distribution network needs: three phases plus a neutral, so that single-phase loads can be connected between any phase and neutral while three-phase loads use all three phases. The 11 displacement is a consequence of the delta-star connection geometry and is consistent across all Dyn11 units, which is what makes them interchangeable. When a distribution transformer fails, the replacement is almost always specified as Dyn11, because the whole network — protection settings, phase relationships and paralleling arrangements — was designed around it.

YNd11 and the other common groups

YNd11 is the mirror image in some respects: a star high-voltage winding with a neutral brought out, and a delta low-voltage winding. It is common on generator step-up transformers and on transmission transformers where the high-voltage side needs an earthed neutral for protection and insulation coordination, while the delta low-voltage side traps harmonics and provides a stable connection for the generator or the lower-voltage network. The 11 displacement is the same as Dyn11, because both are delta-star combinations with the same winding geometry — only the roles of high and low voltage are swapped.

Other groups appear where their properties are needed. YNyn0, with both windings in star and both neutrals brought out and zero displacement, is used where the primary and secondary must be in phase and both sides need neutrals, for example in some autotransformer and interconnection duties. Dyn5, with a 150-degree displacement, appears in older installations and in some industrial applications. Zigzag groups such as Yzn11 are used for earthing transformers and for supplies with heavily unbalanced single-phase loads, because the zigzag winding limits the voltage shift of the neutral under unbalanced conditions. The lesson is not to memorise every group, but to read any group you meet: letters first, then the clock number, then ask what that combination gives you — a neutral or not, harmonic trapping or not, and what phase relationship.

Why phase displacement matters so much

Phase displacement matters because voltages that are not in phase cannot simply be connected together. If two transformer secondaries with different phase displacements are paralleled, their voltages do not cancel even when their magnitudes are equal: the difference between them drives a circulating current around the loop formed by the two transformers, limited only by their impedances. That current flows continuously, heats both transformers, wastes capacity and can be large enough to trip protection or damage windings. The same logic applies to energisation against a network: connecting a transformer whose secondary is out of phase with the busbar it is being connected to is effectively a short circuit through a phase difference, and the resulting current can be destructive.

Displacement also matters for protection and metering. Differential protection compares currents on the two sides of a transformer, and if the vector group is not accounted for, the phase shift looks like a fault current and the protection will trip on load. Modern numerical relays handle this with vector group compensation settings, but those settings must match the actual vector group of the transformer — a relay set for Dyn11 protecting a YNd11 transformer will misbehave. Current transformer connections for metering and protection must likewise be chosen with the vector group in mind. This is why vector group verification is a commissioning test, not a paperwork exercise: the nameplate says what the group should be, and the test proves what it actually is.

Parallel operation: the four rules

Transformers are paralleled to increase capacity, to provide redundancy, or to allow one unit to be taken out of service without interrupting supply. Four conditions must be met for satisfactory parallel operation. First, the vector groups must be compatible — in practice, identical group numbers, so that the secondary voltages are in phase. Groups with the same clock number but different letter combinations can sometimes be paralleled with appropriate connection changes, but that is specialist work and should never be improvised on site. Second, the voltage ratios must be equal on the tap in use: a ratio difference drives a circulating current even when the groups match, because the two secondaries produce different voltages.

Third, the percentage impedances must be reasonably close, because load shares between paralleled transformers in inverse proportion to their impedances. If one transformer has a much lower impedance than the other, it will take more than its share of the load and may overload while its partner is lightly loaded. As a practical guide, impedances within about ten percent of each other give acceptable sharing, though the exact tolerance depends on the application and the manufacturer's guidance. Fourth, the phase sequence must be the same on both sides, and the transformers should be of similar capacity — paralleling a very small transformer with a very large one gives poor load sharing and makes protection coordination difficult. When all four conditions are met, paralleled transformers share load predictably and the arrangement is robust. When any one is missed, the symptoms — overheating, circulating current, unexplained trips — can be very difficult to diagnose after the event.

How vector group is verified on site

Vector group verification on site is done by applying a known three-phase voltage to one winding and measuring the voltages between terminals on both windings, then comparing the pattern of measured voltages with the pattern predicted for the nameplate group. The classic method uses a low test voltage, applied with the transformer isolated and earthed under permit, and a voltmeter to record the terminal-to-terminal voltages. Each vector group produces a distinctive pattern of voltage magnitudes, so a mismatch between the measured pattern and the expected pattern reveals a wrong connection, a swapped terminal or a manufacturing error before the transformer is ever energised against the network.

Modern test sets automate much of this: a transformer turns ratio set with vector group capability applies the voltage, measures the responses and reports the group directly, and the result is recorded alongside the ratio results for every tap. Whichever method is used, the check should be done on the tap that will be used in service, and the result should be compared with the nameplate and with the factory test certificate. A vector group error found at this stage costs an afternoon to correct. The same error found after energisation can cost a transformer. That asymmetry is the whole reason the test exists, and it is worth remembering every time the test feels like a formality. For the companion electrical tests, see our insulation resistance and PI test guide, our transformer oil testing guide and our clamp meter buying guide.

Frequently asked questions

← Back to all guides · Transformer testing guide · Transformer calculators · Electrical tools guide