EV Charging Station Transformer: How to Size It, Why Harmonics Decide the Spec, and When a Standard Distribution Unit Falls Short

The answer first: an EV charging transformer is sized on coincidence, harmonics and fault current — not on the chargers' connected kVA

Almost every EV charging site looks overloaded on paper before it is overloaded in reality, because the connected kVA of the chargers is not the load the transformer actually carries. To specify the right unit you need three numbers: your coincidence (diversity) factor — how many chargers draw full power at the same moment — the harmonic content of the charger front-ends, which heats a transformer well beyond its nameplate rating, and the available fault current the transformer will let through. For an AC Level 2 fleet depot that usually means sizing on roughly 50–70% coincidence and buying a K-4 to K-9 unit; for a DC fast-charging hub it means assuming near-total coincidence, specifying K-13 (or derating 15–20%), and often a dedicated 400 V / 480 V step-down unit fed from the medium-voltage grid. Size on connected kW alone and the transformer becomes the weakest link — it will run hotter than its insulation class allows long before the chargers reach their limit.

Key takeaways

• EV chargers are power-electronics loads, not resistive loads. Their harmonic currents add eddy-current and stray losses a standard transformer was never rated to carry — which is why K-factor, not just kVA, belongs in the specification.
• Size on coincidence, not connected power. A site full of chargers almost never draws its nameplate sum; how much of that sum is simultaneous is the single biggest lever on transformer size.
• K-factor and IEEE C57.110 derating set the real thermal rating. K-4 for light, mixed office/EV load; K-9 to K-13 for regular AC charging; K-13 — or a hard derate of a standard unit — for DC fast charging.
• Fault current and inrush are transformer decisions. The impedance voltage (Uk%) caps the fault current the chargers must clear, and energising a site full of chargers draws an inrush that must not trip the upstream breaker.
• EV hubs run 24/7 at part load, so no-load loss matters. A lower-loss grade or an amorphous core pays back over years of continuous standby power — the economics are quantified below.

Why an EV charging load is not a normal transformer load

An AC Level 2 charger and a DC fast charger both convert AC into DC through a rectifier, and that rectified front-end is a nonlinear load: it draws current in short pulses rather than as a smooth sine wave. The result is a rich harmonic spectrum — dominated by the 5th, 7th, 11th and 13th harmonics — that flows through the transformer and produces extra eddy-current and stray losses in the windings and core. A transformer rated for a linear (sinusoidal) load can overheat under this harmonic load even while it carries less than its nameplate current, because the additional losses turn into heat exactly where the insulation is. That physics is the reason the K-factor rating exists; the mechanism is explained in detail in our guides to transformer harmonics and K-factor transformer selection.

The second non-standard feature is the load profile. A factory loads its transformer gradually; an EV hub swings from near-idle standby to a heavy, power-dense load the moment a vehicle plugs in. That cycling matters twice over: it subjects the insulation and oil to repeated thermal cycling (which accelerates ageing), and it makes the peak — not the average — the design case. A transformer sized on the daily average will be thermally stressed every time the site fills up.

How to size an EV charging transformer: a three-step method

The method is deliberately simple, because the mistakes happen at the inputs, not the arithmetic.

Step 1 — Add up the connected load

Sum the chargers' nameplate power. Twenty-four 22 kW AC chargers are 528 kW connected; six 180 kW DC fast chargers are 1,080 kW connected. For DC units the nameplate is the output power, so add roughly 5–10% for the charger's own conversion losses to reach the input kW the transformer must supply.

Step 2 — Apply the coincidence (diversity) factor

Multiply the connected load by the fraction of chargers you expect to run at full power simultaneously. Long-dwell public AC charging is staggered (typically 0.3–0.5); a fleet depot that turns vehicles around back-to-back is more coincident (about 0.6–0.7); a DC fast hub with short, high-power sessions and peak-hour queues approaches 0.7–0.9. Where a load-management or smart-charging controller is installed — increasingly standard practice — the site demand is capped by that controller, and the transformer should be sized to the cap, not to the raw sum. The same load-factor thinking that governs any transformer applies here; see our guide to load factor and overload capacity.

Step 3 — Convert to kVA and choose a harmonic rating

Convert the coincident kW to kVA using the charger input power factor (about 0.95–0.99 for modern units), then select a K-factor that matches the harmonic profile — or, if a standard distribution transformer must be used, apply the IEEE C57.110 derating described below. Finally add a margin of 10–15% for ambient temperature and future growth. The result is the transformer kVA, and it is usually smaller than the naive sum of the chargers — which is exactly why an EV site can be specified correctly instead of over-bought by guesswork.

AC Level 2 depot vs DC fast-charging hub: two different transformer specifications

The same physics produces two quite different specifications. The table below summarises the typical design envelope for each (values are industry experience for mainstream equipment, not catalogue limits).

K-factor, harmonic derating and the IEEE C57.110 rule

K-factor is a single number that describes how well a transformer tolerates harmonic load current without exceeding its rated temperature rise. The standard series is K-1 (linear loads only), K-4, K-9, K-13 and K-20. A K-13 transformer is built to carry its full nameplate current with a harmonic load whose eddy-current losses are up to thirteen times those of a pure sine wave, which covers most mixed EV installations; K-4 suits light loads dominated by resistive and lighting circuits with only a little charging; a heavy DC fast hub is best served by K-13 or better.

If a K-rated unit cannot be sourced and a standard distribution transformer must be used, IEEE C57.110 defines how much to derate it. The derating depends on the harmonic spectrum and the transformer's eddy-loss factor: as a rule of thumb, a load with roughly 20% current THD typically forces an order-of-magnitude 10–20% kVA derate, and harmonic-rich DC charging can push it further. The practical consequence is blunt — a transformer labelled 1,000 kVA may behave as an 800–900 kVA unit once EV harmonics are accounted for. Specifying a harmonic-rated transformer is almost always cheaper than discovering this after the site is built.

Voltage class, oil-immersed vs dry-type, and where the transformer sits

EV charging is a low-voltage load: AC chargers take 230 V single-phase or 400/480 V three-phase, and DC fast chargers conventionally take 400 V (EU) or 480 V (US) three-phase. The transformer is therefore almost always a step-down unit from the site's medium-voltage supply — commonly 10, 15, 20 or 33 kV, or 13.8 kV in North America — to 400 V or 480 V. Two placement choices dominate. Inside a parking structure or charging hall, a dry-type (cast-resin) transformer is the usual pick: no oil to leak or burn, and smaller clearances, as compared in our oil-immersed vs dry-type guide. Outdoors, at a pad or inside a compact substation, an oil-immersed transformer is more compact and cheaper per kVA and tolerates high ambient and solar gain more gracefully — which matters because charging hubs frequently sit in full sun. Whichever you choose, match the cooling class to the site's real ambient, because a unit that is correct at 25 °C can be overloaded at 45 °C.

Protection, inrush and fault current: the specification points buyers forget

Three points are routinely missed when a transformer is ordered for EV duty. First, fault current: the transformer's impedance voltage (Uk%) sets how much short-circuit current reaches the chargers and their breakers. A low-Uk% unit gives tighter voltage regulation but a higher fault level, so the LV protection and the chargers' interrupting ratings must be coordinated with the actual available fault current — worked through in our short-circuit current guide. Second, inrush: energising the transformer magnetises its core and draws a brief inrush current far above rated, and a hub that energises the transformer and then brings many chargers online at once can trip an upstream device sized only for steady-state — the 5–12× rule is explained in our inrush current guide. Third, voltage regulation: chargers derate or drop out on undervoltage, and the voltage each charger actually sees is set by the transformer's regulation plus the LV cable drop, as covered in our voltage regulation guide. Ordering the transformer without checking these three is how a site that "works" on commissioning fails at peak a year later.

Efficiency and standby loss: the 24/7 economics of an EV hub

Unlike a factory transformer that idles a few hours a night, the transformer at a public charging hub is energised around the clock and runs at part load for most of it. The loss that dominates in that regime is no-load (iron) loss, which accrues 8,760 hours a year whether or not a vehicle is plugged in. That makes the higher efficiency loss grades — S13/S20 or an amorphous-core unit — economically interesting: the extra first cost is recovered from lower standby loss over the life of the hub. Add the load-loss component and you have the total-cost-of-ownership argument; our loss-grade comparison and amorphous-core guide quantify the trade-off, and the total cost of ownership method turns it into a number you can put in a tender. To size the load side in one step, the engineering toolbox calculator takes the site data and returns equipment and budget together.

EV charging transformer specification checklist

• Connected kW and coincidence factor documented — including any load-management or smart-charging demand cap.
• Harmonic profile / expected current THD stated, with the resulting K-factor (or IEEE C57.110 derate) fixed.
• Secondary voltage and configuration — 400 V or 480 V, delta-wye, vector group matched to the site's earthing.
• Impedance voltage (Uk%) chosen for fault-current coordination, not only for voltage regulation.
• Available fault current calculated and coordinated with the chargers' breaker interrupting ratings.
• Cooling class and design ambient temperature matched to the installation (outdoor hubs in hot climates de-rate).
• Loss grade or amorphous core selected for 24/7 part-load economics.
• Inrush margin verified against the upstream protection scheme.
• Oil-immersed or dry-type chosen for the installation location and fire code, with the right IP and corrosion class for outdoor or coastal sites.

Sources / 资料来源

• Source: IEC 61851 series — Electric vehicle conductive charging system
• Source: IEC 62196 — Plugs, socket-outlets, vehicle connectors and vehicle inlets for electric vehicles
• Source: IEC 60364-7-722 — Low-voltage electrical installations — Supply of electric vehicles
• Source: IEEE C57.110 — Recommended practice for establishing transformer capability when supplying nonsinusoidal load currents
• Source: IEEE C57.12.01 — General requirements for dry-type distribution and power transformers
• Source: IEEE 519 — Recommended practice and requirements for harmonic control in electric power systems
• Source: NEC Article 625 (NFPA 70) — Electric Vehicle Power Transfer System
• Source: GB/T 18487.1 — Electric vehicle conductive charging system — General requirements
• Source: QDTB Engineering Team — industry experience for typical coincidence factors, charger ratings and K-factor selection