Buyer's guide

What drives the cost of a current transformer

11 August 2026 11 minute read

Current transformer price is set by roughly eight independent variables — ratio, accuracy class, rated burden, accuracy limit factor or C-class, insulation level, construction type, enclosure rating, and order quantity. Two quotations for "the same current" can differ several-fold if any one of these is left unstated in the enquiry.

Ratio and rated primary current set the starting point

A current transformer (CT) is defined first by its ratio: rated primary current over rated secondary current, with secondary current standardised at 1 A or 5 A. Rated primary current comes from a published series of standard values, not an arbitrary label, so two suppliers quoting the same nameplate current work from the same reference table.

That ratio is not cost-free — winding a core for an unusual ratio changes how much copper and core cross-section is needed to stay inside the design's accuracy limits.

The choice of 1 A versus 5 A secondary matters too: a 1 A secondary carries far less resistive loss over a long cable run, because loss scales with current squared. That feeds into rated burden, and into the worked example below.

Metering class or protection class: the decision that splits a CT in two

IEC accuracy classes

Under IEC, metering accuracy classes are numbered by the maximum ratio error permitted at rated current — class 0.5 permits up to ±0.5%. Tighter numbers (0.2, 0.1) demand tighter manufacturing tolerance, which is what a buyer pays for.

The IEEE and ANSI numbering is different

IEEE and ANSI instead define metering classes around a transformer correction factor band — a different, non-interchangeable numbering convention, so "class 0.3" is not the same tolerance under each scheme. Comparing a quotation written to one against a quotation written to the other requires converting the requirement, not the label.

Why one CT rarely serves both duties

Metering CTs are optimised for accuracy at and below rated current, where billing happens. Protection CTs are optimised to stay usable at many multiples of rated current during a fault. Because the two duties pull core design in opposite directions, a panel commonly needs two separate CT cores on one primary — doubling core and winding content, and cost, per measurement point.

Accuracy limit factor: how far above rated current the core must hold

IEC protection classes carry a "P" marking — 5P10, for example. The number before the P is the maximum permitted composite error at the accuracy-limit current; the number after it is the accuracy limit factor (ALF), the multiple of rated current the core must still hold that error at. Standard ALF values step up in a fixed series (5, 10, 15, 20, 30), and a higher ALF needs a larger core cross-section to postpone saturation.

The IEEE C-class rating serves a comparable purpose, and a higher class number again implies more iron and more cost. The exact numeric definition belongs to IEEE Std C57.13 and is not reproduced here.

Rated burden: the load the CT must be guaranteed to drive

Rated burden is the maximum secondary-circuit load, in volt-amperes, the CT is guaranteed to drive while meeting its stated accuracy class. Wiring resistance, the connected meter or relay's input impedance, and secondary cable length all add to it.

Both IEC and IEEE define burden in discrete standard steps, so a specifier rounds up to the nearest standard burden rather than an arbitrary VA value. Accuracy is only guaranteed across roughly a quarter to the full value of rated burden, so running a CT light does not automatically help, and can push some designs outside their error band.

Specifying more burden than an installation needs raises core material cost for no benefit. Over-specified burden, often driven by caution rather than a calculated circuit, is one of the most common ways two "equivalent" quotations diverge in price.

Working the numbers: secondary lead burden at 1 A against 5 A

Take a pilot cable of 2.5 mm² copper, run 50 m out and 50 m back — a 100 m loop. Copper resistivity is about 0.0175 Ω·mm²/m, so the loop resistance is:

R = ρ × L / A = 0.0175 × 100 / 2.5 = 0.7 Ω

The burden dissipated in the loop is I²R. At a 5 A secondary:

Burden(5 A) = 5² × 0.7 = 17.5 VA

At a 1 A secondary, same cable, same run:

Burden(1 A) = 1² × 0.7 = 0.7 VA

The identical cable and run produces 17.5 VA of burden at 5 A and 0.7 VA at 1 A — a 25-fold difference, matching the currents squared (5² ÷ 1² = 25). The method carries over to any cable size or run length: change the cross-section or the loop length and only R moves.

Practically: on a long secondary run, a 1 A secondary lets the CT's rated burden be specified much lower, since the cable consumes far less of the guaranteed VA. A lower rated burden generally means a smaller, cheaper core — which is why long pilot-wire runs favour a 1 A secondary.

Insulation level and system voltage

Instrument transformers carry a rated insulation level tied to the system's highest equipment voltage, expressed as withstand test voltages — power-frequency and impulse. This is what the CT's insulation system, not its core, is built to survive.

Insulation coordination groups system voltage into standard steps, and the required dielectric test battery — routine power-frequency withstand, lightning-impulse withstand, and at the highest steps switching-impulse withstand too — grows with that step, adding testing and material cost.

For outdoor apparatus, required creepage distance is specified per kV and scales further with site pollution severity: a coastal or industrial site needs more creepage per kV than a clean inland one.

A CT rated for a higher voltage class is not the same core in a bigger box. The added creepage, clearance and impulse-withstand requirements change the insulation design outright, making voltage level a cost driver separate from ratio, burden or accuracy class.

Construction type: wound, bar, window and split core

Window (ring-core) and bar-type

Window-type, or ring-core, construction uses only a secondary winding on a toroidal core, with the primary conductor passed through the window rather than built in — removing the primary winding's copper and insulation cost. Toroidal geometry is generally preferred for accuracy because its closed, symmetric magnetic path avoids the leakage-flux errors of non-toroidal designs. Bar-type, or bushing, construction uses a single rigid conductor bar as the primary; needing no separate winding, it is mechanically simpler at the cost of a fixed, low turns ratio.

Split core for retrofit

A split-core design cuts a ring core into two hinged halves so it can close around a live conductor without disconnecting it. That cut introduces a mechanical joint into the magnetic path, and the accuracy the design can hold depends on how precisely the cut faces mate. The machining that buys that precision is part of what a split core costs over a solid ring — the premium is paid for not having to de-energise the circuit, and it is worth asking whether the outage is really unavailable before paying it.

Wound primary for low currents

A wound-primary CT forms the primary conductor into one or more turns through the core, raising effective ampere-turns for low primary currents. It needs more copper and assembly labour than a single-pass bar or window design at the same current, which is why wound-primary CTs cost more at the low end of the range even when the core is no larger.

Enclosure and environment rating

Enclosure protection against dust and water is classified by IP code under IEC 60529, in two digits — solid-object ingress, then water ingress. Outdoor or washdown duty needs a materially higher IP rating than indoor switchgear mounting.

IEC 60529 covers equipment up to 72.5 kV and excludes corrosion resistance, which sits under separate standards or under the North American NEMA 250 system, a different classification scheme. Because IP and NEMA ratings are not a direct conversion, an enquiry that just says "outdoor" leaves the enclosure specification open to interpretation, and suppliers will price to different assumptions.

Order quantity and tooling: why unit price falls with volume

Fixed costs that do not depend on unit count — tooling, jigs, first-article testing, drawing approval — are spread across the order, so their per-unit share falls as quantity rises while material and labour per unit stay comparatively flat.

This is why two or three units of a non-catalogue combination can carry a much higher unit price than the same specification ordered fifty or a hundred at a time. A like-for-like price comparison has to hold quantity constant, not only the electrical rating.

Current transformer cost drivers at a glance

What to state in an enquiry for a comparable price

A measuring CT's approval record requires nameplate information covering ratio, rated burden or burdens, accuracy class per burden, voltage classification and thermal current rating — close to the minimum data set a buyer needs for a comparable price.

A CT can carry more than one accuracy class on the same nameplate, one per standard burden, so "class 0.5" with no burden stated is incomplete: the standards define accuracy as meaningless without the burden it was measured at. The same logic applies to protection duty — omitting the ALF or C-class leaves out the current multiple the CT must hold its accuracy across.

System voltage must be one of the standard steps, not a rounded figure, since the dielectric test regime is selected by step. Construction type is a buyer decision driven by installation constraints — whether the circuit can be de-energised, whether retrofit onto a live conductor is required — and stating it stops a supplier defaulting to whichever construction suits their own tooling. Quantity and batching belong on the same line as the electrical specification, since price per unit is a function of the specification together with order quantity.

Why quotations for the same rating come back different

An unexplained price gap between suppliers quoting "the same" CT almost always traces back to one of the drivers above being interpreted differently, rather than one supplier simply being cheaper. Left to guess, a supplier defaults to whichever construction, burden or batch size suits them, not necessarily what the installation needs.

The fixed and variable cost split is the most common reason two otherwise-identical quotations diverge sharply at low volume, since a small order carries the same fixed tooling and testing cost as a larger one spread over far fewer units. The fix is an enquiry that states every driver above, including quantity, so every supplier prices the same thing.

Frequently asked questions

Why do two quotations for the same current rating come back at different prices?

"Same current rating" only fixes the ratio. Accuracy class, burden, ALF or C-class, insulation level, construction type, enclosure rating and quantity are all still open, and each of them moves the price.

Does a higher accuracy class always cost more?

Generally yes — tighter classes need tighter manufacturing tolerance. But the burden that class is rated at matters too; a tight class at a low burden is not comparable to the same class number at a higher one.

Why is a small order sometimes priced far higher per unit than a large one?

Fixed costs — tooling, first-article testing, drawing approval — are spread across the batch. A batch of two carries the same fixed cost as a batch of fifty, over far fewer units.

Is a 1 A secondary always cheaper than a 5 A secondary?

Not in the core itself, but 1 A drives far less resistive loss over long secondary cable runs, which can let a lower-burden, smaller core serve the same installation, as the worked example above shows.

What is the minimum an enquiry needs to state for a comparable price?

Ratio, accuracy class per burden, rated burden in VA, ALF or C-class, system voltage step, construction type, enclosure rating, and quantity.