Wind Farm Step-Up Transformer: How to Size It (0.69 kV → 33 kV), Why Turbine Harmonics Drive the Spec, and What Changes Onshore vs Offshore

The answer first: a wind farm steps voltage up in two stages — size the turbine unit on generator kVA and harmonic duty, and the farm unit on total coincident output and grid-code requirements

A wind farm steps voltage up in two stages. The turbine step-up transformer (also called the WTG or unit transformer) raises the generator's low voltage — usually 0.69 kV, or 0.9–1.14 kV on the largest machines — up to the farm's collector voltage of 20 kV, 33 kV or 35 kV, and it is sized on the generator's rated apparent power in kVA (not its active power in kW), with a small margin for ambient and harmonic heating. The farm (substation) transformer then lifts the whole collector bus from 33 kV to the grid-connection level of 110 kV, 132 kV or 220 kV, sized on total farm output corrected for diversity rather than the sum of every turbine nameplate. Three things separate a wind transformer from a plain distribution transformer: converter harmonics that add winding losses, cyclic wind-driven duty that makes no-load loss dominate lifetime cost, and harsh siting — offshore salt fog, onshore altitude, and wide thermal cycling. Specify those three correctly and the transformer survives the turbine's 20-year design life.

Key takeaways

• Two stages, two sizing rules. The WTG transformer is sized on generator apparent power (typically 1.0–1.1× the generator kVA); the farm transformer is sized on total farm output using a diversity/coincidence factor, not the arithmetic sum of all turbine ratings.
• Harmonics set the thermal derate. Both full-converter (Type 4) and doubly-fed (Type 3) turbines inject current distortion; specify a K-factor rating or a documented IEEE C57.110 derate, plus an electrostatic shield between the LV generator winding and the MV collector winding.
• Wind duty is cyclic, not steady. Output follows the wind, so no-load (iron) loss runs thousands of hours a year while load loss runs only at high wind — a better loss grade (S13/S20) or an amorphous core is often justified on lifetime cost.
• Offshore changes the type, not only the enclosure. Nacelle-mounted units trend toward dry-type or ester-filled designs with C5-M corrosion protection and sealed, IP-rated enclosures; onshore collector transformers are usually oil-immersed pad-mounted units.
• Grid-code compliance is a transformer requirement. Low-voltage ride-through, reactive-power capability, harmonic limits and fault-level coordination all flow back into impedance, tap range and thermal design.

Why a wind turbine transformer is not a standard distribution transformer

A distribution transformer steps a stable low-voltage network down to a customer and runs near a steady load. A wind turbine transformer does almost the opposite. It takes a power-electronics generator's low-voltage output and raises it to the collector voltage under three conditions that reshape the design.

First, the source is harmonic-rich. A utility-scale turbine is either a doubly-fed induction generator (Type 3), whose rotor-side converter injects low-order harmonics, or a full-converter machine (Type 4), whose stator is fed entirely through a converter that switches at high frequency. Both are held to defined power-quality limits — current distortion at the point of connection is governed by IEEE 519 and IEC 61000-3-6, and turbine characteristics are measured to IEC 61400-21 — but even compliant machines leave switching ripple and low-order harmonics (5th, 7th, 11th, 13th) that raise eddy-current and stray losses in the transformer windings. That is the same mechanism behind transformer harmonic heating and K-factor selection.

Second, the load follows the wind. A turbine transformer is energised whenever the machine is available — often 8,000+ hours a year — but reaches full load only during strong wind. So its no-load loss accrues almost every hour of the year, while its load loss accrues only part of the time. That asymmetry reshapes the efficiency argument, exactly as it does for solar PV step-up transformers.

Third, the environment is harsh. Offshore units face salt-laden air, high humidity and restricted maintenance; onshore units sit in steppe, desert, mountain and coastal sites with high or very low ambient, intense solar gain and wide daily temperature swings. A transformer designed for a 25 °C indoor room is not automatically right for a 45 °C outdoor pad or a −30 °C cold-start site.

How to size a wind turbine (WTG) step-up transformer: a three-step method

The arithmetic is simple; choosing the right inputs is not.

Step 1 — Start from the generator's rated apparent power (kVA)

Use the generator's rated apparent power, not its active-power rating. A 5 MW turbine operating at 0.95 power factor presents about 5.26 MVA to its transformer; a 6 MW machine at the same power factor presents about 6.32 MVA. Sizing on MW alone understates the current the winding must carry.

Step 2 — Add a margin for ambient, altitude and harmonics

Apply a design margin — commonly 5–15% as a starting point — for the site's hot-day ambient, altitude (which reduces air cooling, see the high-altitude derating guide) and the harmonic content of the machine. Round up to the next standard rating, for example 5.5 MVA or 6.3 MVA for a 5 MW class turbine. Oversizing beyond about 15% mostly adds no-load loss and cost.

Step 3 — Fix the harmonic, thermal and tap specification

Set the K-factor (or a documented IEEE C57.110 derate), the winding temperature rise and insulation class for the design ambient, the cooling class (ONAN for oil-immersed, AN for dry-type), and the off-circuit tap range — typically ±2 × 2.5% — so the collector voltage can be trimmed as the grid condition changes. Add an electrostatic shield between LV and MV windings to keep converter-side and grid-side disturbances apart. General requirements are set by IEC 60076-16 (transformers for wind turbine applications), with IEC 60076-1 and IEEE C57.12.00 as the parent standards.

How to size the wind farm (substation) transformer

Step 1 — Add up the farm's AC output, then apply diversity

Sum the farm's total AC capacity at the point of common coupling, then apply a coincidence (diversity) factor: not every turbine produces nameplate output at the same instant, even in a well-sited array. A farm of twenty 5 MW turbines (100 MW installed) may reach a coincident output of roughly 90–97% of installed capacity under ideal wind, so the farm transformer's design basis is the site's studied peak export, not the arithmetic sum of every turbine's transformer rating.

Step 2 — Convert to kVA and round to a standard size

Convert the coincident AC MW to MVA using the farm's contracted power factor (often 0.95 lagging to 0.95 leading under modern grid codes), then round up to a standard rating — for example a 100 MW farm at 0.95 pf gives about 105 MVA, so a 120 MVA or 125 MVA unit is a typical choice.

Step 3 — Specify impedance, taps and cooling for grid-code compliance

Set impedance voltage (Uk%) to balance fault-level limitation against voltage regulation and ride-through (typically 10–14% for a 110–220 kV class unit); choose on-load or off-load tap changing according to the grid code's voltage-range and reactive-power requirements; and confirm the cooling class and overload profile against the required reactive capability. The transformer must not undermine the farm's fault ride-through or reactive-power compliance — a system issue, not a catalogue choice.

The voltage architecture: from 0.69 kV to 220 kV in two stages

Where each transformer sits fixes its ratio and its rating. Four arrangements cover almost every onshore and offshore wind farm.

Onshore collector voltages cluster around 33 kV (Europe and much of Asia), 34.5 kV (North America), 35 kV (China) and 20 kV; the grid-connection level is chosen by the local transmission operator. Offshore, the collector voltage often rises to 66 kV to reduce cable losses, and the export voltage is set by the cable length and the onshore grid.

Converter harmonics and K-factor: why modern turbines heat the transformer

Turbine type decides the harmonic spectrum the transformer must tolerate.

• Type 3 — doubly-fed induction generator (DFIG). The stator is directly connected, so the transformer sees mainly low-order harmonics (5th, 7th, 11th, 13th) from the partially rated rotor converter. Distortion is moderate but persistent.
• Type 4 — full converter. The generator output is fully rectified and re-inverted, so the transformer sees switching-frequency ripple plus low-order sidebands whose magnitude depends on the converter topology and its filter. Total distortion is often lower, but the high-frequency content is richer.

Eddy-current and stray losses rise with the square of the harmonic current and with frequency, so a transformer rated for a pure sine wave runs hotter than its nameplate suggests under converter load. Two specification options handle this: a K-factor rating (K-4 to K-9 covers most modern turbines) or a standard unit with a documented IEEE C57.110 derate. Both should be paired with an electrostatic shield between the LV and MV windings, which also protects the turbine converter from grid-side transient overvoltages. For the underlying methods see the harmonics guide and the K-factor guide.

Cyclic wind duty and loss economics: which loss grade pays back

A turbine transformer's duty cycle is dominated by hours, not by load. It is energised whenever the machine is available, and its no-load loss (P0) accrues every one of those hours, while its load loss (Pk) accrues only during generation at load. On a site with a capacity factor of, say, 35%, the transformer may run near full load only a fraction of the year. That is why lifetime-cost calculations for wind transformers weight no-load loss more heavily than for an industrial transformer — and why a better loss grade (S13 or S20), or an amorphous-metal core, can pay back within a few years despite a higher first cost. The method is the same one set out in the TCO guide, and the trade-off between loss grades is covered in the S11/S13/S20 comparison and the amorphous-core guide.

Two thermal consequences follow from cyclic duty: daily thermal cycling (expansion and contraction of windings and insulation, which ages the paper insulation and can loosen clamping over time) and fan or forced-cooling switching if ONAF is used. Both argue for robust clamping, a conservative winding temperature rise, and a cooling design that does not depend on frequent switching. Insulation-life fundamentals are covered in the insulation class guide.

Onshore vs offshore: what changes in the specification

Offshore units also carry seismic and vibration requirements, very low cold-start temperatures in some regions, and tighter fire-loading limits inside a nacelle. Onshore, the biggest practical issues are altitude derating, dust and insects in cooling ducts, and moisture and rodent ingress in pad-mounted designs. For oil systems, the choice between a conservator and a hermetically sealed tank is covered in the oil-preservation guide, and fluid selection in the transformer oil guide.

Wind farm transformer specification checklist

• Design basis is generator apparent power (kVA) at rated power factor, with the hot-day ambient and altitude margin stated separately.
• Voltage ratio fixed: 0.69 / 0.9 / 1.14 kV LV to 20 / 33 / 35 / 66 kV collector — one stage or two.
• K-factor or documented IEEE C57.110 derate matched to the turbine type (Type 3 or Type 4).
• Electrostatic shield between LV and MV windings.
• Vector group — Dyn11 is the common default; confirm against the farm's earthing and protection design (see the vector-group guide).
• Impedance voltage (Uk%) set to balance fault level, voltage regulation and ride-through.
• Off-circuit taps (±2 × 2.5%) for collector-voltage trimming; on-load tap changing only if the grid code requires it.
• Loss grade or amorphous core selected on lifetime no-load-loss cost, not first cost.
• Cooling class and design ambient matched to the site (for example 45–50 °C hot-day design).
• Corrosion class (C4 onshore / C5-M offshore per ISO 12944), IP rating and sealed design.
• Surge arresters and protection coordinated with the transformer fault level and turbine ratings (see the lightning protection guide and protection devices guide).
• Type and routine tests to IEC 60076-16 plus the parent IEC 60076 series, with a factory acceptance test before shipment (see the pre-shipment inspection checklist) and a verified manufacturer audit (see the supplier audit checklist).

Frequently asked questions about wind farm transformers

How do you calculate the kVA rating of a wind turbine transformer?

Divide the turbine's active-power rating by its power factor to get apparent power, then add a 5–15% margin for ambient, altitude and harmonics and round up to a standard size. A 5 MW turbine at 0.95 pf is about 5.26 MVA, so a 5.5 MVA or 6.3 MVA transformer is a typical choice.

Is a wind turbine transformer oil-immersed or dry-type?

Onshore, oil-immersed pad-mounted or tower-base transformers are the norm. Offshore, and especially inside a nacelle, dry-type or ester-filled units are preferred because they reduce fire load and the risk of an oil spill in a location that is hard to reach for maintenance.

Why do wind farms use two transformer stages?

The generator's voltage (about 0.69 kV) is far too low to transmit across the farm, so each turbine steps up to a 33 kV class collector and one central transformer lifts the collector to the 110–220 kV grid level. Transmitting at collector voltage keeps cable losses and cable size manageable.

What is the difference between a WTG transformer and a wind farm substation transformer?

The WTG (unit) transformer raises one turbine's output to the collector voltage and is sized on that turbine's apparent power. The wind farm substation transformer raises the whole collector bus to the grid voltage and is sized on the farm's total coincident output.

Sources / 资料来源

• Source: IEC 60076-16 — Power transformers — Part 16: Transformers for wind turbine applications
• Source: IEC 60076-1 / 60076-2 / 60076-3 — Power transformers (general requirements, temperature rise, insulation levels and dielectric tests)
• Source: IEC 60076-11 — Power transformers — Dry-type transformers
• Source: IEC 61400-1 — Wind energy generation systems — Design requirements
• Source: IEC 61400-21 — Wind energy generation systems — Measurement and assessment of power quality characteristics of grid-connected wind turbines
• Source: IEC 61000-3-6 — Electromagnetic compatibility — Assessment of emission limits for distorting loads in MV and HV power systems
• Source: IEEE C57.12.00 — General requirements for liquid-immersed distribution, power and regulating transformers
• Source: IEEE C57.12.01 — General requirements for dry-type distribution and power transformers
• Source: IEEE C57.110 — Recommended practice for establishing transformer capability when supplying nonsinusoidal load currents
• Source: IEEE 519 — Recommended practice and requirements for harmonic control in electric power systems
• Source: GB/T 1094 series — 电力变压器 (Power transformers)
• Source: GB/T 19963.1 — 风电场接入电力系统技术规定 第1部分：陆上风电 (Technical requirements for connecting wind farms to the power system — Part 1: Onshore wind)
• Source: VDE-AR-N 4120 — Technical requirements for the connection and operation of customer installations to the high-voltage network
• Source: ISO 12944 — Corrosion protection of steel structures by protective paint systems
• Source: DNV-ST-0145 — Offshore substations
• Source: QDTB Engineering Team — industry experience for typical WTG transformer sizing margins, K-factor selection and site-ambient design