CT Ratio & Burden Calculator

Pick the ratio for the load you have

The commonest field mistake comes from good intentions: fitting a 1000/5 CT on a circuit drawing 150 A “for future expansion”. That CT now runs at 15 % of its rating, which for a plain class 0.5 core is below the range its accuracy is specified over. The readings are not slightly worse — they are undefined.

Choose the ratio so normal running current lands between 40 % and 100 % of the primary rating. For a 150 A circuit that is a 200/5, not a 1000/5. A CT is a cheap component; change it when the load actually grows, rather than accepting years of poor readings in between.

Burden is a budget, and cable spends most of it

A CT is a current source. It drives its secondary current through whatever impedance you connect, developing whatever voltage that takes — until the core saturates and it cannot. Burden is that total impedance, expressed in VA at rated secondary current.

Lead loop  = 2 × ρ × L / A          ρ = 0.01851 Ω·mm²/m copper
Lead VA    = I² × R
Total VA   = lead VA + meter VA + terminals

Modern electronic meters draw a few tenths of a VA. The 5 VA instruments that made large CTs normal are long gone. What is left is almost entirely cable — and cable burden goes with the square of the secondary current.

Which is the whole argument for a 1 A secondary. Same cable, same meter, same distance, and the lead burden falls twenty-five fold. Use 5 A where the CT is close to the meter; above about 15 m, 1 A is usually the right answer, and above 30 m a 5 A secondary needs uncomfortably fat cable to work at all.

What over-burdening actually does

Under-burdening is harmless — a 15 VA CT running at 2 VA does not mind. Over-burdening saturates the core, so the CT cannot deliver its rated secondary current and the meter reads low.

The dangerous part is the shape of the error: it grows with current. The readings are worst exactly when the energy matters most, and they look entirely plausible at every point. Nothing alarms, nothing trips, and you find out when somebody reconciles submeter totals against the main meter and finds several per cent missing.

Accuracy class, and the S

The S is worth more than most specifications acknowledge. A plain class 0.5 CT holds its accuracy from 20 % of rating; a class 0.5S holds it from 1 %. On a plant that idles overnight, that difference is a third of every day spent outside specification. If the metering allocates cost or hunts for savings, specify 0.5S — and see what it is worth with the metering accuracy calculator.

Protection cores are a different device. 5P10 means 5 % accuracy at ten times rated current: designed to stay linear during a fault and deliberately inaccurate at normal load. Never meter from one.

Wiring, and the one that hurts

Never open-circuit an energised CT secondary. With no burden to drive, the core saturates and develops hundreds or thousands of volts across the open terminals — enough to kill somebody and enough to destroy the CT's insulation. Short the secondary before disconnecting a meter on a live circuit. That is what the shorting links on a test block are for, and it is why a test block is not an optional extra.

• Polarity. P1/P2 primary, S1/S2 secondary. A reversed CT shows negative power on that phase, and a three-phase total that is low by roughly a third rather than obviously wrong. Check each phase against a clamp meter at commissioning.
• Earth the secondary at one point only, normally S2 at the meter. Two earths make a loop.
• Split-core CTs must close fully on clean mating faces. A gapped or dirty joint adds reluctance and the CT reads low — check this before doubting anything else on a retrofit.

A specification that will read correctly

1. Ratio such that normal load is 40–100 % of primary.
2. Class 0.5S for anything used to allocate or analyse energy.
3. Burden calculated — meter plus lead loop plus terminals — then rounded up to the next standard VA.
4. 1 A secondary if the run exceeds about 15 m.
5. FS5 or FS10 instrument security factor on metering cores.
6. Shorting-type test block at the meter.
7. Polarity verified per phase at commissioning.

Metering cores and protection cores want opposite things

A metering CT is designed to be accurate around normal load and to give up gracefully above it. A protection CT is designed to stay linear far past full load, because the current it has to reproduce faithfully is the fault current. The same core cannot do both well, which is why larger installations specify multi-core CTs with separate secondaries.

Metering classes carry an instrument security factor — an FS5 CT saturates at five times rated current, deliberately, so that a downstream meter or ammeter is not destroyed by a fault. Protection classes carry the opposite specification: the accuracy limit factor. A 5P20 core promises 5 per cent composite error up to twenty times rated current, and a 10P10 promises 10 per cent up to ten times.

Put a meter on a protection core and it reads acceptably but the meter is exposed to fault current. Put a relay on a metering core and the core saturates at the exact moment the relay needs the reading, so the relay sees a collapsing secondary current and either delays or fails to operate. The second mistake is the dangerous one and it tests perfectly during commissioning, because nobody injects twenty times rated current into a working panel.

Polarity, and the readings it quietly inverts

Every CT has a defined direction. P1 and P2 mark the primary, S1 and S2 the secondary, and the convention is that current entering P1 leaves S1. Reverse the primary bar or swap the secondary pair and the measured current is 180 degrees out.

An ammeter will not notice — it reads magnitude and nothing else, which is why a polarity error can sit in a panel for years without being found. Anything that measures power will. A single reversed phase on a three-phase energy meter produces a reading roughly one-third of the true value, and on a plant with any export capability it can show power flowing the wrong way entirely.

Differential and directional protection is less forgiving still: a reversed CT makes a healthy circuit look like an internal fault, and the relay operates correctly on information that is wrong. Check polarity at commissioning with a known single-phase load and compare the sign of the measured power against what the load is actually doing. It takes minutes, and it is far cheaper than discovering it from a disputed bill or an unexplained trip.