Wye / star connection

What it is (precise)

Draw the three windings as the legs of a Y. Each winding sits between one line terminal and the shared neutral, so the voltage across a winding is the line-to-neutral voltage, and the voltage between any two line terminals is the difference of two winding phasors 120° apart — √3 times bigger, which is where the 1.732 comes from. Current has nowhere to divide: each line conductor feeds exactly one winding, so line current equals winding current.

The paired-voltage convention on European equipment reads straight off this geometry — 230/400 V is line-to-neutral over line-to-line, and 400/230 is √3 within the rounding of the published values. The same arithmetic runs at the PCS terminals: a 690 V AC bus, the common output class for grid-scale inverters, has about 398 V from any phase to the neutral point, whether or not that neutral is physically accessible.

The winding-level consequence is economy at high voltage. A star winding only ever sees 1/√3 — about 58% — of the line-to-line voltage, so it needs fewer turns and a lower insulation class than a delta winding on the same system, which must withstand the full line voltage.

IEC 60076-1's own worked generator step-up example, YNd11, stars the 20 kV network side — windings actually operating near 11.5 kV each — and puts the delta on the low-voltage machine side; the transformer-vector-group entry unpacks the notation itself. The trade is symmetrical: a delta buys √3 lower winding current instead, which is why the choice is made winding by winding rather than system by system.

The neutral point and what it enables

Only a star (or zigzag) connection has a neutral point at all, and the nameplate says whether you can reach it: IEC 60076-1 appends N or n only when the neutral is brought out to a terminal, so Dyn11 offers an accessible low-voltage neutral and Dy11 does not. That terminal is the site of two separate decisions.

The first is system earthing — solidly earthed, impedance-earthed, or left floating — which sets earth-fault current, insulation stress and protection philosophy for everything on that bus; the grounded-wye entry owns that choice. The second is distribution: a brought-out, distributed neutral turns the three-phase system into a four-wire one, and every line-to-neutral tap is a ready-made single-phase supply.

That second use is what runs a storage plant's balance of system. Auxiliary loads — control cabinets, BMS supplies, lighting, small heaters, communications — are largely single-phase, and they hang line-to-neutral off a four-wire wye secondary such as a 230/400 V auxiliary board, because a line-to-neutral supply needs a neutral and the wye is the connection that has one.

In a balanced system the neutral conductor carries nothing, since the three line currents sum to zero at every instant; single-phase auxiliaries are never perfectly balanced, so real current flows in the real neutral — the neutral-current entry covers how unbalance and triplen harmonics load it, and the phase-unbalance entry what an uneven auxiliary schedule does to the transformer above it.

What the neutral does not do by itself is provide a fault path. A wye whose neutral is left unearthed is an open circuit in the zero-sequence network, exactly as a delta is — the zero-sequence entry carries that machinery — so drawing a Y on the single-line diagram settles nothing about where earth-fault current will flow.

And a neutral that is not distributed cannot be tapped: a three-wire medium-voltage collection system at 34.5 kV may be star-connected at the transformer and still offer no conductor from which to take a single-phase supply, which is why station service comes from a dedicated auxiliary transformer rather than a tap off the collector.

Where the wye sits in a BESS plant

Read the plant's transformers by their letters. On a main step-up unit the classic arrangement stars the grid side — IEC's worked GSU example is YNd11, network windings in star with the neutral brought out, delta toward the machine — while many MV power-block transformers run the other way, Dy11, with the delta on the 34.5 kV collection side and the star on the inverter's low-voltage side; both arrangements ship in current product, and some inverter models require that low-voltage neutral wired back to them while others forbid earthing or even connecting it.

Which pairing is right is a function of the MV neutral treatment and the protection scheme, and that decision — along with the clock numbers — is the transformer-vector-group entry's subject. This page's job is the letter: wherever a winding must give the system an earth reference or a neutral, it is a wye.

The auxiliary chain is wye end to end. The station-service transformer's secondary is a four-wire star feeding the LV distribution boards, and the single-phase vs three-phase entry covers why the loads on it should be spread deliberately across the three phases rather than landed wherever is convenient. UPS-backed life-safety circuits, tripping supplies and SCADA all terminate line-to-neutral somewhere on that wye — so the integrity of one neutral conductor and its earthing sits underneath a surprising fraction of the plant's availability and its fire-detection duty.

How it shows up in specs, studies and contracts

On datasheets and in studies, the wye is a bookkeeping discipline. An AC voltage on a nameplate is line-to-line unless it says otherwise; a paired figure like 230/400 V is line-to-neutral over line-to-line; and every per-phase model in a load-flow or short-circuit study works in line-to-neutral volts and line amps — quantities that are literal winding quantities only for a wye.

Divide by √3 in the wrong place and the error is 73%, the same trap the single-phase vs three-phase entry documents from the other direction. Whenever a study, a relay-setting file and a datasheet exchange numbers, the first check is which of the two voltages every figure is.

Commissioning puts a meter on the geometry. A winding-resistance test on a star winding measured between two line terminals reads two windings in series — twice the per-winding value — so the test sheet must state its measurement points before anyone compares it against a factory record.

Phase-to-neutral voltage checks on the auxiliary boards verify the wye and its earthing at once. Contracts should say plainly who supplies, connects and earths each neutral: the EPC-supplier scope boundary runs straight through the auxiliary transformer, and an unearthed — or doubly earthed — neutral discovered at energization is a finding that stops the sequence.

Common pitfalls

The arithmetic traps come first. In a wye the √3 belongs to voltage only — line current equals winding current, and applying the factor to both double-counts it. The reverse error is quoting winding voltage where line voltage is meant: a 398 V figure taken from a 690 V bus is not a different system, just the other leg of the same conversion. And a wye drawn on a diagram does not promise a usable neutral — confirm the neutral is brought out (the n in the vector group), distributed, and earthed before planning any line-to-neutral supply from it.

The structural traps follow. Star is not shorthand for earthed: the connection creates the neutral, the earthing design disposes of it, and the two are specified in different documents by different parties. A transformer starred on both sides with no delta anywhere has known third-harmonic and zero-sequence problems — the transformer-vector-group entry covers why practice either adds a stabilizing delta or avoids the connection.

And never specify star or delta as loose adjectives in a procurement document: the binding string is the full vector group with its clock number, because two transformers both wye on the LV side can still be impossible to parallel if their displacements differ.