Power & Distribution

Sizing transformer overcurrent protection (NEC 450.3(B) & CEC 26)

Entogo

Industrial plant power distribution — sizing transformer primary and secondary overcurrent protection and available fault current to NEC 450.3(B) and CEC Section 26

A transformer’s overcurrent protection is the one part of the installation that has to satisfy two unforgiving documents at once: the code rule that caps the device size, and the physics of the fault current the gear downstream has to survive. Get the first wrong and an inspector rejects the job — or worse, the primary fuse nuisance-trips on every energization. Get the second wrong and a panelboard is asked to interrupt a fault it was never rated for. This is how both numbers are found, for the United States (NEC) and Canada (CEC), with the worked examples that show where the two codes quietly disagree.

Everything below is the rule set behind the transformer sizing & overcurrent protection calculator, which runs these tables live; this is the explanation of what it computes and why. It assumes the kVA is already chosen — if it is not, start with how to size a transformer.

First principle: 450.3(B) protects the transformer, not the wire

The single most common misreading of transformer protection is to treat the primary device as if it protected the secondary conductors. It does not. NEC 450.3(B) and CEC Section 26 size the device to protect the transformer winding — the conductors on either side are a separate calculation. With primary-only protection in particular, the secondary conductors and the downstream panel must be protected by other means, typically the secondary-conductor tap rules of NEC 240.21(C). Keeping those two jobs separate in your head is what makes the rest of the table make sense.

The percentages are also maximums, not targets. The code tells you the largest device you may use; coordination, inrush and the load decide how close to that ceiling you actually sit.

The NEC rule: Table 450.3(B), 1000 V and less

For transformers rated 1000 V or less, the maximum overcurrent device is a percentage of the winding’s full-load (rated) current. The percentage depends on whether you protect the primary only or both sides, and on how large the current is.

ProtectionWinding & currentMax OCPD (% of rated)
Primary onlyPrimary ≥ 9 A125% (round up*)
Primary 2–9 A167%
Primary < 2 A300%
Primary and secondaryPrimary (any)250%
Secondary ≥ 9 A125% (round up*)
Secondary < 9 A167%

*Note 1 to Table 450.3(B): where 125% of the rated current does not land on a standard device rating, the next higher standard rating is permitted. The standard ratings are the NEC 240.6(A) ladder — 15, 20, 25, 30 … 6000 A, plus the small fuse ratings 1, 3, 6 and 10 A that protect control and lighting transformers.

The two columns answer different design intents. Primary-only is the minimum-hardware path for a smaller transformer: one device on the line side at ≤ 125%. Primary-and-secondary adds a secondary main at ≤ 125% — which buys the primary a much larger ceiling (250%) so it can ride through inrush without nuisance-tripping. The 167% and 300% rows exist for small transformers whose full-load current is so low that a 125% device would trip on the magnetizing inrush alone.

A NEC worked example

A 150 kVA, 480 V → 208Y/120 V, three-phase dry-type unit, protected on both sides:

  1. Primary full-load current: 150 000 ÷ (480 × √3) = 180 A.
  2. Primary device: 250% × 180 = 451 A → the largest standard rating not above that is 450 A.
  3. Secondary full-load current: 150 000 ÷ (208 × √3) = 416 A.
  4. Secondary device: 125% × 416 = 520 A → round up under Note 1 to 600 A.

Drop the secondary main and the same transformer on primary-only protection takes a primary device of 125% × 180 = 225 A — at which point the secondary conductors need protecting under 240.21(C) instead.

The CEC rule: Section 26, and where Canada differs

Canada sizes the same protection under CSA C22.1 (the Canadian Electrical Code) Section 26. For dry-type transformers 750 V and under, the governing rule is 26-254:

  • 26-254(1): the primary overcurrent device is set at ≤ 125% of rated primary current (next standard size up permitted, as in the NEC).
  • 26-254(2): where a secondary device is set at ≤ 125% of rated secondary current, the primary feeder may be set at ≤ 300% of rated primary current.

That 300% is the quiet but real divergence from the NEC. On the same both-protected transformer, the US caps the primary at 250% and Canada at 300%. Take the 150 kVA, 600 V → 208 V Canadian equivalent: primary full-load current is 144 A, so the NEC’s 250% gives 361 A → a 350 A device, while the CEC’s 300% feeder allowance gives 433 A → a 400 A device. The secondary stays at 125% either way. Liquid-filled and over-750 V power transformers fall under Rules 26-250 / 26-252 instead, and the exact subrule should always be confirmed against CSA C22.1 and the local AHJ / ESA.

This US-vs-Canada split is exactly why a North American supplier’s tool has to carry both codes — switch the calculator between NEC and CEC and the primary device changes for the very same transformer.

The second number: available fault current and AIC / SCCR

Sizing the overcurrent device tells you nothing about whether the gear can interrupt a fault — that is a separate question answered by the available fault current at the transformer secondary. The textbook first-pass uses the infinite-source method: the available fault current equals the secondary full-load current divided by the transformer’s per-unit impedance.

Available fault current ≈ secondary FLA ÷ (%Z ÷ 100)

A 500 kVA, 480 V → 208 V transformer draws 1388 A on the secondary; at a nameplate 5% impedance that is 1388 ÷ 0.05 ≈ 27.8 kA of available fault current at the secondary terminals. Every breaker and fuse just downstream must have an interrupting rating (AIC) at least that high, and every panelboard and switchgear assembly a short-circuit current rating (SCCR) that meets or exceeds it — mandated by NEC 110.9 / 110.10 and CSA C22.1 Rule 14-012. 27.8 kA rounds up to a 35 kA rated device.

Two cautions make this a screen, not a study. The infinite-source assumption is conservative-high — a finite utility source and the impedance of downstream conductors both reduce it — but connected motors feed into a fault and push it back up. And a lower actual impedance yields a higher fault current: small liquid-filled units can run near 2% %Z, where the same kVA produces far more fault current than a 5.75% dry-type. The nameplate %Z governs; the IEEE 242 (Buff Book) point-to-point method is the basis for the full study that confirms it.

A protection specification checklist

The two numbers above are the headline, but a complete protection scheme also settles:

  • Winding full-load current — primary and secondary (the basis for every percentage above)
  • Protection scheme — primary-only vs primary-and-secondary, and therefore which table rows apply
  • Standard device ratings — fuses vs inverse-time breakers (NEC 240.6(A) / CSA Section 14)
  • Inrush coordination — a time-delay device that rides 8–12× rated for a few cycles without tripping
  • Secondary-conductor protection — 240.21(C) tap rules where the primary device does not cover the secondary
  • Available fault current and the downstream AIC / SCCR that must exceed it
  • Medium-voltage primaries — above 1000 V the NEC switches to Table 450.3(A) (by %Z and supervised location), a different table not covered here

Where Entogo fits

Entogo manufactures liquid-immersed, dry-type and pad-mounted transformers, switchgear and distribution assemblies and prefabricated substations in its own source factory — designed and built to ANSI/IEEE C57 or IEC 60076, UL/CSA certifiable on request. Because the same house builds the transformer and the switchgear it feeds, the protection and the SCCR are coordinated as one package rather than bolted together on site.

Run your numbers first in the transformer sizing & overcurrent protection calculator — full-load current, NEC 450.3(B) / CEC 26-254 OCPD, available fault current and enclosure in one pass — then turn the result into a specification with the transformer configurator or browse the transformer & substation range. For the sizing decision that comes before protection, see how to size a transformer.

Transformer protection is not a hard calculation, but it is an exacting one: size the device to the winding, keep the conductors a separate job, ride the inrush, and make sure everything downstream can interrupt the fault the transformer can deliver. Name each number for the rule behind it, and the scheme will pass the inspector and survive the fault.

FAQ

Common questions

How do you size transformer overcurrent protection under NEC 450.3(B)?
For a transformer rated 1000 V or less, NEC Table 450.3(B) sets the maximum overcurrent device as a percentage of the winding's rated (full-load) current. With primary protection only, the maximum is 125% where the primary current is 9 A or more (the next standard size up is permitted under Note 1), 167% for 2–9 A, and 300% below 2 A. With both primary and secondary protection, the primary may go up to 250% and the secondary is held to 125% (167% below 9 A). Apply the percentage to the winding's full-load current, then pick the device from the NEC 240.6(A) standard ratings.
What is the difference between primary-only and primary-and-secondary transformer protection?
Primary-only protection uses a single overcurrent device on the primary (line) side, sized at up to 125% of primary full-load current. It is common for smaller transformers, but the secondary conductors must then be protected by other means under the tap rules (NEC 240.21(C)). Primary-and-secondary protection adds a secondary device sized at up to 125% of secondary current, which in turn lets the primary device go higher — up to 250% under the NEC — to ride through inrush. NEC 450.3(B) protects the transformer; it never protects the conductors, which are a separate calculation.
How does the Canadian Electrical Code (CEC Section 26) differ from the NEC for transformer protection?
The CEC governs dry-type transformers 750 V and under in Rule 26-254. Rule 26-254(1) limits the primary overcurrent device to 125% of rated primary current (next standard size up permitted). Rule 26-254(2) is where Canada differs from the US — when a secondary device is set at not more than 125% of rated secondary current, the primary feeder may be set as high as 300% of rated primary current, versus the NEC's 250%. Liquid-filled and over-750 V power transformers follow Rules 26-250 and 26-252. Always confirm the exact subrule against CSA C22.1 and your AHJ / ESA.
How do you calculate the available fault current at a transformer secondary?
The quick infinite-source estimate is the secondary full-load current divided by the per-unit impedance — available fault current = secondary FLA ÷ (%Z ÷ 100). A 500 kVA, 208 V secondary draws about 1388 A; at 5% nameplate impedance that is roughly 27.8 kA of available fault current. This assumes an unlimited utility source, so it is deliberately conservative (high). A finite utility source and downstream conductors lower it, while connected motors raise it — a full short-circuit study refines the number before it is relied on.
What AIC or SCCR rating do I need downstream of a transformer?
The breakers and fuses just downstream of the transformer must have an interrupting rating (AIC) at least equal to the available fault current at that point, and panelboards and switchgear must carry a short-circuit current rating (SCCR) that meets or exceeds it — required by NEC 110.9 / 110.10 and CSA C22.1 Rule 14-012. Take the available fault current at the secondary terminals and select the next standard interrupting rating at or above it (for example, 27.8 kA rounds up to a 35 kA device). A lower actual transformer impedance produces a higher fault current, so verify the nameplate %Z before committing to an AIC.
Why does a transformer primary breaker trip when the transformer is energized?
Transformer magnetizing inrush can reach roughly 8 to 12 times rated current for a few cycles at energization. A primary device sized at the bottom of the allowable range will nuisance-trip on that inrush — which is exactly why NEC Table 450.3(B) and CEC Rule 26-254 permit the device to be sized well above 100% (up to 250–300% in the both-protected case). Use a time-delay (dual-element) fuse or an inverse-time breaker and coordinate it with the transformer's inrush and damage curve.

Keep reading

Related insights

Project inquiry

Turn the brief into a specified system.