PART XIII Measurement & Instrumentation
§39 / 42

Instrument Transformers & Metering

CT · PT · IEEE C57.13 · burden · accuracy class · revenue vs protection

Every meter on a switchgear cubicle, every relay watching a feeder, every utility revenue billing — they all see the system through instrument transformers. Get the CT and PT wrong and downstream metering reads garbage or, worse, a relay misses the fault it was supposed to clear.

Why You Need Instrument Transformers

Direct measurement at MV is impossible. A 12.47 kV, 2,000 A feeder needs to feed a 5 A relay input and a 120 V meter — that's a current step-down of 400:1 and a voltage step-down of 60:1 with galvanic isolation between the primary system and the secondary instrument loop. Instrument transformers do exactly this, accurately enough that the secondary readings are a faithful scaled image of the primary quantities.

TypeSteps downStandard secondaryPrimary use
Current Transformer (CT)Primary current → secondary current5 A (US standard) · 1 A (IEC / large substation)Ammeters, kWh meters, overcurrent / differential / distance relays
Voltage / Potential Transformer (VT, PT)Primary voltage → secondary voltage120 V L-N or 115 V depending on PT ratioVoltmeters, kWh meters, undervoltage / sync / distance relays
Capacitor Voltage Transformer (CVT)Same role as PT but uses a capacitor divider120 VEHV / UHV systems where a wound PT becomes uneconomical (≥ 138 kV)
Optical / electronic transducerUses fiber-optic sensors (Faraday/Pockels effect) or Rogowski coilDigital signal (IEC 61850-9-2 sample values)GIS substations, digital substations, retrofit

How a CT Actually Works

A CT is a step-up transformer with a single primary turn (the conductor itself passing through the window) and many secondary turns. By transformer action, secondary current = primary current ÷ turns ratio. The defining constraint: the secondary must always have a low-impedance path. An open-circuited CT secondary develops dangerous voltages — the magnetizing flux is no longer constrained by the secondary ampere-turns and rises until the core saturates, with peak voltages of thousands of volts. Always short the CT secondary before disconnecting downstream meters or relays.

Window-type CT single primary turn through toroid · many secondary turns on core Primary I₁ (e.g., 2000 A) N secondary turns X1 X2 → relay / meter (5 A) Polarity dot: H1↔X1 enter together. Reverses polarity → relay sees opposite-direction current. CT Equivalent (Secondary Side) ideal source · magnetizing branch · burden · saturation curve I₂ = I₁ / N Xₘ magnetizing reactance Burden Z₋ leads + relay + meter V₂ Iₘ (excitation) knee Above the knee, core saturates → secondary no longer scales.

Accuracy Class — IEEE C57.13

The accuracy class on the nameplate tells you exactly how much error to expect at rated burden. Two families: metering class (small errors at rated load, undefined behavior at fault levels) and relaying / protection class (large errors allowed at rated load, but guaranteed not to saturate up to a fault current many times rated).

ClassApplicationRatio error limitPhase angle limit
0.1 / 0.15 / 0.2Revenue metering (utility billing, transmission interchange)± 0.1% to ± 0.2% at 100% rated I± 5 to ± 15 minutes
0.3Standard revenue metering — most utility installations± 0.3% at 100% rated, ± 0.6% at 10% rated± 15 minutes
0.6Industrial metering, sub-billing± 0.6% at 100% rated, ± 1.2% at 10% rated± 30 minutes
1.2Indication only (panel ammeter)± 1.2% at 100% rated, ± 2.4% at 10% rated± 60 minutes
C100 / C200 / C400 / C800Protection class — burden voltage rating at 20× rated I≤ 10% ratio error at 20× rated up to the C-class voltage at standard burden
T-class (e.g., T100)Wound CT — same as C-class but accuracy verified by test, not calculation≤ 10% at 20× rated

Many CTs carry dual ratings — e.g., 0.3B0.5 + C400 means metering-class 0.3 at 0.5 Ω burden, plus protection-class C400 (will not saturate before 400 V secondary at 20× rated, ≈ 80 A through standard 5 Ω burden = 400 V). One CT serves both meter and relay through separate cores or separate windings on the same core.

Burden — The Most-Misunderstood Spec

"Burden" is the impedance presented to the CT secondary, in ohms or volt-amperes at rated secondary current. It's the sum of: lead-wire resistance (round trip from CT to switchgear), the relay or meter input impedance, and connector/terminal contact resistance. If the actual installed burden exceeds the CT's rated burden, the CT may saturate under fault current and the relay misses the fault.

Burden in ohms vs VA
VA = Isec2 · Zburden
At rated 5 A: Zburden (Ω) = VA / 25 · 1 VA ≈ 0.04 Ω · 25 VA ≈ 1 Ω · 50 VA ≈ 2 Ω
Burden sourceTypical valueNotes
kWh meter (electromechanical)1–2 VA0.04–0.08 Ω at 5 A
Solid-state meter (e.g., ION, EPM)0.05–0.5 VANegligible
Electromechanical overcurrent relay (51)2–8 VASignificant at fault levels
Microprocessor relay (SEL, GE Multilin)< 0.5 VANegligible
Lead wire (one-way, #10 AWG, 100 ft)0.10 Ω → 2.5 VARound trip doubles — most installations dominate by leads
Lead wire (#10 AWG, 1 m)0.0033 Ω/m round tripSI scale

Polarity

The dot or H1/X1 marking defines instantaneous polarity. Current entering H1 on the primary side comes out X1 on the secondary side at the same instant. If polarity is reversed (X1 and X2 swapped), a relay measuring directional current (67) reverses sense — what looked like a forward fault now reads as reverse, and the relay either trips for through-current or fails to trip for an actual fault. Polarity errors are the most common installation defect on differential schemes (87) and the cause of nuisance tripping or missed trips on day-one energizations.

PT / VT Selection

PTs follow a similar scheme but the constraint is opposite: never short-circuit a PT secondary (it's a step-down voltage transformer; shorting the secondary creates very high primary-equivalent current). Standard ratios produce 120 V L-N or 115 V at full primary voltage. PT accuracy classes follow the same 0.3 / 0.6 / 1.2 metering scheme.

System voltageTypical PT ratioSecondary at full primary
480 V (L-L)4:1120 V L-L · 69.3 V L-N
4.16 kV35:1120 V L-L · 69.3 V L-N
12.47 kV104:1120 V L-L · 69.3 V L-N
13.8 kV115:1120 V L-L · 69.3 V L-N
34.5 kV287:1120 V L-L
69 kV and aboveUse CVT instead of wound PT120 V L-L

Worked Example 1 — Sizing a CT for Atlas DC1 Main Breaker

Example 01 · CT selection4000 A, 480 V main breaker on Atlas DC1 SWGR-A. Both revenue metering AND overcurrent protection.
  1. Choose ratio. Pick a CT ratio so that full-load primary current ≈ 100% of CT primary rating (linear region) and short-circuit primary stays below saturation. For a 4000 A breaker → CT ratio 4000:5 (800:1). Full load gives 5 A secondary; a 65 kA fault gives 81 A secondary.
  2. Pick metering class. Atlas DC1 uses customer-side revenue metering on the main service, but the utility owns the actual revenue CT upstream. The customer meter is sub-billing across data halls — class 0.3 is sufficient (utility revenue would be 0.15 or 0.2).
  3. Pick protection class. Available fault current at the bus is 65 kA. Through CT: 65,000 / 800 = 81 A secondary. Need C-class such that core does not saturate at 81 A through standard 5 Ω burden:
    Vknee = 81 A × 5 Ω = 405 V → choose C400 (knee voltage ≥ 400 V at 20× rated). Conservative: C800 if leads and relay add significant burden.
  4. Compute actual burden. Lead resistance: 80 ft round-trip × 0.001 Ω/ft (#10 AWG) = 0.08 Ω. Microprocessor relay: 0.1 Ω. Meter: 0.05 Ω. Total ≈ 0.23 Ω.
  5. Verify against C-class. At 81 A secondary, voltage across burden = 81 × 0.23 = 18.6 V — well below C400 capability. Even at 20× rated (100 A), voltage = 23 V. C400 is more than enough. C200 would have been adequate; designer specified C400 for margin against future relay additions.
  6. Final spec: 4000:5 multi-ratio CT, dual core: 0.3B0.5 metering + C400 protection. Polarity: H1 toward source. Mounting: window-type around the bus stab.
i
Multi-ratio CTs
Many breakers use multi-ratio CTs (e.g., 4000:5 with taps at 600, 1200, 2000, 3000, 4000). The user picks the active ratio by tapping the right secondary terminals. This lets one CT serve a 1500 A feeder today and a 3000 A feeder after upgrade — without replacing hardware.

Worked Example 2 — Burden Calculation Forces a CT Upgrade

Example 02 · Burden trapExisting 1200:5 C100 CT feeding a 1950s-vintage electromechanical relay 200 ft away. Adding a new revenue meter requires 2 more leads.
  1. Existing burden: 200 ft round-trip × 0.001 Ω/ft = 0.20 Ω leads. Electromechanical 51 relay = 6 VA at 5 A → 6/25 = 0.24 Ω. Total = 0.44 Ω.
  2. Existing margin to C100: At 20× (100 A secondary), Vburden = 100 × 0.44 = 44 V. Well within C100 (≥ 100 V at 20× rated). System works.
  3. Add the meter. Solid-state meter: 0.4 VA = 0.016 Ω. Trivial. New total ≈ 0.46 Ω. Still well within C100.
  4. The actual hazard: The new meter cable is 350 ft round-trip in #14 AWG (someone reused a spare). #14 AWG = 0.0025 Ω/ft → 350 × 0.0025 = 0.88 Ω of additional lead resistance — added in series with the meter, in parallel with the relay.
  5. Compute composite burden. Two parallel branches: relay branch = 0.20 Ω leads + 0.24 Ω relay = 0.44 Ω. Meter branch = 0.88 Ω leads + 0.016 Ω meter = 0.90 Ω. Parallel: (0.44 × 0.90) / (0.44 + 0.90) = 0.30 Ω. CT now drives this parallel pair through its own internal secondary winding resistance (~ 0.1 Ω).
  6. Re-check at 20× rated: Vburden = 100 × 0.30 = 30 V. Still inside C100. However, the high-impedance #14 AWG meter run drops 0.88 V × secondary current — at 5 A normal, that's 4.4 V across just the leads. The meter sees only 0.08 V at full load — meter reads near zero. Meter installation works electrically but reads garbage.
  7. Fix: Pull #10 AWG meter cable (0.001 Ω/ft × 350 ft = 0.35 Ω round-trip), or install a dedicated metering core CT (separate winding for metering, sized to the meter's actual lead path). Latter is cleaner — keeps relay performance untouched and gives the meter its own properly-burdened source.
What you can do after this section
  1. Pick a CT ratio and accuracy class for any breaker, justifying both metering and protection requirements.
  2. Compute installed burden including lead resistance and verify it against the CT's C-class rating.
  3. Identify polarity errors on a one-line diagram and explain why they cause directional / differential misoperation.
  4. Translate between burden in VA and burden in Ω at rated secondary.
  5. Choose between a wound PT and a CVT based on system voltage and economics.

Drill — Quick Self-Check

Work each problem mentally; reveal to check.

Drill 1 · CT secondary

Standard CT secondary current rating in the US is __ A.

Drill 2 · Open CT

Why must a CT secondary never be left open?

Drill 3 · Class C400

What does "C400" tell you about a CT?

Drill 4 · Burden conversion

25 VA at 5 A rated = ___ Ω.

Drill 5 · CVT vs PT

CVT preferred over wound PT above what voltage class?

If You See THIS, Think THAT

If you see…Think / use…
"4000:5 multi-ratio"CT with selectable taps; pick the ratio matching the active load.
"0.3B0.5"Metering class 0.3 at 0.5 Ω burden — revenue / sub-billing acceptable.
"C400" or "C800"Protection class CT with knee voltage ≥ 400 V (or 800 V) at 20× rated.
"H1, H2, X1, X2"Polarity terminals — H1 and X1 enter at the same instant.
"Burden in VA"Convert with VA = I²Z; at 5 A rated, Z (Ω) = VA / 25.
"Knee voltage"Where saturation begins; above this the secondary stops scaling linearly.
"CVT"Capacitor voltage transformer — used at ≥ 138 kV instead of wound PT.
"Open CT secondary"Hazard — short the CT secondary before disconnecting downstream devices.
"Polarity reversal"Directional (67) and differential (87) relays misoperate. Recheck H1↔X1 alignment.
"IEEE C57.13"The standard governing instrument transformer accuracy classes.
"Rogowski coil"Air-cored CT alternative — no saturation, output is mV-level signal proportional to dI/dt.
"Metering core" + "Relay core"Two separate CTs (or two windings on the same core) — meter sees clean low-burden, relay sees C-class.
Also see