STAR-STAR VTs (also known as Y-Y), when fed from a Delta source (specifically when no star-point is made available on the secondary), should be treated as a Delta and the instrument set-up accordingly.
At a small, but relevant, tangent; The reason the transformers are wound star-star is two fold. The first is economics - simply put, less copper is used. We'll use example figures of 11kV to 110V (a standard sub-station VT).
If the primary were to handle 11kV on each winding the manufacturer of the transformer is faced with two challenges. The first is insulation. The entire winding must be insulated from all structures to 11kV. This then imposes a minimum insulation material which forces a larger overall diameter of the wire. Although the wire may be thinner (less current), the overall length will be longer to make way for the insulation.
If in Star mode, the insulation requirement at the inner of the windings (closest to the core) may effectively be zero as the transformer will create a "virtual neutral" (this is sometimes made available for common mode fault analysis). However, each winding is now only required to be insulated to 6.35kV, and that is only at the supply end. Although the copper is thicker in star-star, less is used in this mode because of thinner insulation.
The second, but less known reason for star-star transformers is because of thinner insulation, thicker copper, less turns (through lower voltage), and overall less space required, the coupling of the windings to the core is higher. This results in a lower transfer impedance which means the output is less affected by varying loads within the substation i.e. this is more ideal for instrumentation. The phase angle is also less affected because of this coupling. This lower transfer impedance has a superb by-product, it allows better transfer of harmonics i.e. better suited to ensuring G5/4 compliance!
Back to the main subject. If we look at the accompanying diagram it will be noticed that, although the voltage between R&Y is split over two windings, the same two windings are directly transformed to the two that exist between A&B. The voltage that exists between A&B will therefore be directly related (by the ratio of the transformer) to what exists between R&Y (and the same for the rest).
Hence, the fact that the feed is delta, we set the instrument to delta, the nominal voltage to the secondary output (in our example, 110V), and the multiplier to the transformer ratio (in our example, 100).
If you have a star-star feed-load configuration then having a star-star VT without grounded star points is exactly the same as a delta-delta and will unfortunately leave you without being able to determine imbalance and harmonics accurately (because of measuring the summing of two phases). You also totally lose any recovery of common mode voltages so creating a virtual neutral is of no use whatsoever.
The source may well be a star secondary with the star-point brought to ground, but this must not be confused with safety requirements. In this case should any one phase go to ground the protection can detect this and trip out.
If there are any questions please email
NOTE: This webpage does not form part of any official documentation.
© M.T.P. - 04.12.01