CT Selection for Numerical Protection Relays: Ratio, Class, Burden, Knee Point and Saturation

A protection CT is acceptable only if the complete CT–lead–test-block–relay chain reproduces the current needed by the protection algorithm for the required fault and time window. Ratio, “5P20,” burden or knee-point voltage alone cannot prove that. Load resolution, maximum asymmetric fault, remanence, secondary resistance and the relay’s dynamic behaviour must be evaluated together.

This engineering guide provides a practical workflow for selecting and validating conventional current transformers for numerical MV protection. It covers IEC and IEEE specification approaches, 1 A versus 5 A, P/PX/PR/transient classes, burden and excitation calculations, saturation, high- and low-impedance schemes, worked examples, testing and documentation.

Executive rules

Choose ratio from both maximum legitimate current and minimum/maximum faults—not bus rating alone.
Specify the CT class/parameters that match the scheme: overcurrent, biased differential, high-impedance REF/bus or transient-duty protection are different problems.
Calculate external burden using hot lead resistance and the actual fault-current return path; include test switches, terminals and relay inputs.
Include CT secondary winding resistance when calculating excitation voltage and saturation, even though it is not external burden.
Apply maximum primary fault, X/R, inception angle, duty cycle and remanence to the time-domain performance requirement.
Do not assume a lightly burdened CT is automatically satisfactory; ratio, core area, flux history and relay algorithm still matter.
Never open-circuit an energised CT secondary. Provide shorting/test facilities and a controlled procedure.
Validate nameplate assumptions with certificates, design calculation, field tests and dynamic relay testing.

1. Start from the protection duty

Application	CT performance that matters most	Typical specification route
50/51 phase overcurrent	Pickup resolution, maximum fault range, operating time before harmful saturation	IEC P class or IEEE relaying class plus application check
Sensitive earth fault	Low-current ratio/phase error and phase-fault spill	Dedicated CBCT or suitable core; complete residual-chain test
Low-impedance differential	Matched transient reproduction, saturation/restraint algorithm and through-fault stability	IEC PX/P or transient parameters as scheme study requires
High-impedance differential/REF	Knee point, excitation current, winding/lead resistance and stabilising voltage	IEC PX-style specified parameters and matched CTs
Distance/directional	Phase accuracy and transient reproduction during close faults	Application-specific P/PX/transient study
Metering + protection	Conflicting low-current accuracy and fault saturation/security objectives	Separate cores unless a combined core is explicitly validated

2. Standards and class systems are not interchangeable

IEC 61869-2: additional requirements for inductive CTs, including protection classes, transient performance and tests.
IEC 61869-1:2023: current general instrument-transformer framework; check contractual compatibility because IEC 61869-2:2012 was published with the earlier general part.
IEEE C57.13-2016: active IEEE requirements and test framework for conventional instrument transformers.
IEEE C57.13.1-2017: active field-testing guide for relaying CT connection, ratio, polarity, condition, burden and excitation.

An IEC 5P20 designation and an IEEE C-class designation express performance differently. Do not convert them with a one-line equivalence table. State the governing standard/edition and all project parameters. A relay manual’s CT recommendation may be more restrictive than the generic class because it is tied to a particular algorithm and application.

3. Ratio selection: create a feasible window

For ratio CTR = I_p,rated/I_s,rated, the ideal secondary current is I_s = I_p/CTR. Evaluate:

Upper normal duty: maximum continuous/emergency load, transformer overload, motor start, capacitor/reactor current and future growth must be inside CT thermal and relay continuous ranges.
Minimum protection duty: minimum remote/high-resistance fault must produce adequate relay operating multiple above CT/relay error and pickup.
Maximum fault: close-in asymmetrical current must be within CT short-time/dynamic and relay input withstand; protection must remain dependable for the required time.
Resolution: excessively high ratio can make sensitive earth, unbalance and low-set elements depend on a few ADC counts/noise.
Standard ratio and taps: document which physical tap is used, nameplate class on that tap and how unused secondary terminals are treated by the manufacturer.

Do not select a 2,000/1 A CT merely because the switchgear is rated 2,000 A if the feeder can never carry that current and its minimum fault is low. Conversely, do not select a low ratio that overloads the core/secondary or relay under contingency load and maximum fault.

4. 1 A or 5 A secondary?

Criterion	1 A	5 A
Lead burden at same impedance	1/25 of 5 A burden because VA = I²Z	High on long runs
Secondary voltage for same impedance	Lower at rated current	Five times higher
Legacy compatibility	Common in modern numerical schemes	Common in existing North American/legacy installations
Wiring error/noise consideration	Lower signal current, so complete-chain EMC and input specification matter	Higher signal but much higher I²R loss
Decision	Use one controlled project basis and verify CT, relay hardware, test switch and settings agree

5. External burden calculation

For an essentially resistive secondary circuit, Z_burden = R_leads,hot + R_{terminals/test} + Z_relay and VA = I_s,rated²|Z_burden|. For reactive burdens, retain complex impedance and power factor. Determine which conductors carry current for each fault: a star-connected CT circuit may use different phase/neutral paths for phase and earth faults.

Use loop length, not one-way length, unless the actual common return is explicitly modelled.
Correct copper resistance to maximum expected conductor/terminal temperature.
Include marshalling kiosks, multicore cable, test blocks, shorting links, interposing CTs and all series terminals.
Use the relay’s burden at the selected current input/frequency; do not substitute auxiliary-power consumption.
For multiple relays/meters on one core, add the actual series burden and review reliability/maintenance risk.
Measure operating burden in the as-built circuit because cable length and terminal changes are common.

6. Worked lead-burden example

Illustrative only. A CT circuit has an 80 m one-way run of 2.5 mm² copper. Using an assumed hot resistivity ρ = 0.021 Ω·mm²/m, loop resistance is 0.021 × 160 / 2.5 = 1.344 Ω. Relay, test switch and terminals add 0.150 Ω, so external burden is approximately 1.494 Ω.

At 1 A rated secondary, burden ≈ 1² × 1.494 = 1.49 VA.
At 5 A rated secondary with the same wiring, burden ≈ 5² × 1.494 = 37.35 VA.
The CT winding resistance is not added to the external VA figure, but it is added when calculating the secondary excitation voltage and internal loss.
Final design uses actual cable data/temperature, common-return topology, relay impedance and applicable power factor.

7. IEC P class and accuracy limit factor

A designation such as 5P20 at a stated rated burden describes composite-error performance at the rated accuracy-limit current under specified conditions. It does not promise an undistorted secondary waveform at 20 times rated current for every X/R, remanence or burden.

A commonly used preliminary estimate adjusts actual accuracy-limit factor with the ratio of total secondary impedance at rated versus actual burden, including CT secondary winding resistance. Treat it as a screening calculation only: core nonlinearity, power factor, temperature, remanence and transient duty require manufacturer data or an excitation/time-domain model.

Confirm rated burden on the nameplate/certificate and the class validity on the selected tap.
Use hot R_ct for worst-case voltage/loss.
Compare protection-required current/time with CT performance—not simply I_fault/I_rated ≤ ALF.
Check relay saturation detection/restraint and transient overreach/underreach with representative waveforms.

8. PX, PR and transient classes

Class/parameter route	What is specified	Typical reason
PX	Turns ratio, knee-point voltage, secondary winding resistance and excitation current at stated voltage(s), as specified	High-impedance differential/REF and applications needing an equivalent-circuit design
PR	Protection performance with low remanence characteristics	Duty where residual flux control improves repeated-fault response
TPX/TPY/TPZ (and applicable transient parameters)	Specified transient reproduction/secondary time-constant behaviour and error limits	High X/R, fast protection and defined auto-reclose/fault duty cycles

For a common IEC PX excitation characteristic, the knee point is associated with the region where a 10% voltage increase causes a 50% increase in excitation current. Use the exact IEC edition, certificate and manufacturer test definition. Knee-point voltage is a core characteristic, not the maximum safe continuous secondary terminal voltage.

9. Secondary excitation voltage and saturation

Before substantial saturation, a screening secondary voltage requirement is:

V_s,req ≈ I_s,fault(R_ct,hot + Z_external).

This sinusoidal RMS relation does not capture the decaying DC offset that drives unidirectional flux. The asymmetrical primary waveform depends on system X/R and fault inception angle. Flux can accumulate until the core saturates; remanence and rapid reclose/repeated faults can start the next event closer to saturation.

Use maximum X/R at the CT location, not a generic value.
Apply the actual protection operating time: some high-speed elements need only an accurate early window, while time-delayed functions may need longer reproduction.
Model maximum through fault for differential stability and minimum internal fault for dependability.
Include breaker operating/clearing and BF time if current must be reproduced until backup.
Test the relay with saturated waveforms because numerical filters/restraint can tolerate or respond to distortion differently.

10. High-impedance schemes

High-impedance differential or REF protection intentionally develops a voltage across the relay/stabilising branch during internal spill current while remaining stable for an external fault with one CT heavily saturated. The design usually requires matched CT ratios/characteristics, maximum through-fault current, the worst CT secondary-plus-lead resistance to the junction point, stabilising resistor, relay impedance, minimum fault current and a nonlinear resistor/MOV voltage-energy check.

Calculate the external-fault stability voltage using the scheme manufacturer’s equation and worst lead/CT path.
Select operating voltage above stability voltage with margin, then verify sensitivity at minimum internal fault after CT excitation currents.
Specify adequate knee point and low excitation current for all parallel CTs.
Verify peak secondary voltage and energy; apply the specified nonlinear resistor and insulation/clearance.
Keep CT secondary star-point/junction and single earth location exactly as the approved scheme.
Do not add meters/test devices into the high-impedance differential path without re-engineering stability.

11. Low-impedance numerical differential schemes

Biased/percentage differential relays can remain stable through CT error using restraint, but “low impedance” does not mean CT selection is unimportant. Unequal saturation can create large false differential current and delay internal-fault operation.

Follow the relay’s CT dimensioning rule or validated time-domain model for maximum through fault, X/R and clearing time.
Check ratio matching and compensation range; avoid unnecessarily disparate ratios/core construction at zone ends.
Verify low-set sensitivity with load/tap/vector-group compensation and CT steady-state error.
Verify high-set/unrestrained element security for maximum external fault and dependability for severe internal fault.
Test external-fault CT saturation, CT open/short, energisation and evolving external-to-internal faults dynamically.

12. Thermal, dynamic and insulation duties

Rated short-time thermal current I_th and duration must exceed maximum primary duty using the applicable I²t basis.
Rated dynamic current I_dyn must withstand first-peak electromagnetic force.
Continuous thermal rating must cover emergency/overload current and ambient/enclosure conditions.
Insulation level, partial-discharge requirements, creepage and internal-arc installation arrangement must suit the switchgear.
Secondary terminal insulation and overvoltage limiting must suit open-circuit/test and high-impedance scheme duty.
Mechanical fit, primary conductor position, window fill and manufacturer mounting tolerances matter for ring/bushing CT accuracy.

13. Specification data sheet

governing IEC/IEEE standard and edition;
highest system voltage, insulation/PD/environment and frequency;
primary/secondary rated current, ratio/taps and continuous rating factor;
core number and exclusive protection/metering function;
class plus rated burden/ALF, or PX/transient parameters and limits;
maximum R_ct at stated temperature and excitation-current limits;
I_th, duration, I_dyn and duty cycle;
maximum system fault current/X/R, required accuracy duration and relay type/algorithm;
actual connected burden, lead size/length/temperature and test arrangement;
terminal markings, polarity, secondary earthing and nameplate diagram;
type/routine/special test certificates and excitation curve/data points;
interchangeability and dimensional constraints.

14. Factory and site verification

Review certificate/nameplate against the approved data sheet and each core/tap.
Inspect terminal markings, physical polarity, core allocation and secondary earth point.
Measure ratio and polarity through final wiring where possible.
Measure secondary winding resistance and correct to the reference temperature.
Measure excitation characteristic/knee point and excitation current where the class/application requires it.
Measure actual connected burden/loop resistance in the final path.
Demagnetise after excitation/resistance testing using the instrument procedure.
Primary-inject balanced and residual quantities to prove ratio, phase identification and complete relay path.
Secondary/dynamic-test relay dependability and CT-saturation security with approved settings.
Archive raw results, test connections, temperature, instrument calibration, curves and as-built values.

OMICRON’s CT Analyzer is an example of an instrument that derives equivalent-circuit parameters and evaluates ratio/phase accuracy, winding resistance, excitation/knee point, burden, composite error and transient parameters. OMICRON’s 2026 CMC CT-Check announcement describes guided functional checks with CMC 500, while explicitly positioning CT Analyzer for comprehensive standards-based assessment. Select test equipment by required voltage, class verification and safety—not brand familiarity.

15. Safety controls

Treat every CT secondary as capable of dangerous voltage while primary current flows.
Short the CT on approved shorting terminals before opening/removing a relay/test plug.
Never rely on a software block to make an energised CT circuit safe.
Use one controlled secondary earth point unless an engineered scheme states otherwise; verify no accidental multiple earths.
Discharge/demagnetise as required after excitation testing and confirm circuits restored before energisation.
Apply site isolation, test-permit, arc-flash and induced-voltage controls.

16. Frequent mistakes

Mistake	Consequence	Correction
Ratio = bus rating	Poor sensitivity or overload mismatch	Load/fault feasible window
Nameplate VA only	Hot as-built burden unknown	Calculate and measure full loop
ALF treated as transient multiple	Saturation missed	X/R, remanence and time-domain check
R_ct omitted	Excitation voltage understated	Use hot winding resistance
5P/PX/IEEE class equated	Specification ambiguity	One standard/edition and explicit parameters
Separate CTs tested, wiring not tested	Polarity/core/terminal errors latent	Complete-path primary/functional test
No demagnetisation	Remanence changes duty/test result	Controlled demagnetisation

References and further reading

Engineering note: Equations here are screening tools. Final CT acceptance requires the governing standard, manufacturer excitation/transient data, the exact relay application rule and approved fault/time-domain study.