Power System Protection & Switchgear

Chapter 7
Differential and Unit Protection

Bachelors Program in Electrical Engineering
Power System Protection & Switchgear   •   Bachelor

Differential and Unit Protection

Introduction

Differential protection is one of the most selective and fastest forms of protection used in electric power systems. It operates on a simple physical principle: for a protected zone, the current entering must equal the current leaving, apart from transformer ratio, phase-shift, and measurement errors. A fault within the zone produces an imbalance, whereas a fault outside the zone should produce no operating current.

This principle is applied to generators, transformers, motors, busbars, reactors, and transmission lines. The protected zone is defined by the locations of current transformers (CTs). Because protection depends on comparing currents at different points, CT performance, wiring, relay characteristics, and communication channels are as important as the relay algorithm itself.

The central engineering challenge is dependability for internal faults and security for external faults, especially high-current through-faults that may drive one or more CTs into saturation.

Learning objectives

After studying this chapter, the reader should be able to:

  1. Explain the operating principle of current differential and unit protection.
  2. Distinguish circulating-current, percentage-biased, high-impedance, and low-impedance schemes.
  3. Analyze stability during external faults and CT saturation.
  4. Explain busbar, transformer, line, and restricted earth-fault protection.
  5. Select appropriate schemes for practical power-system applications.
  6. Interpret relay settings, CT requirements, communication channels, and testing needs.

1. Fundamental principle of differential protection

Consider a protected zone with currents referred to a common base and polarity convention. The differential, or operating, current is the phasor sum of currents entering the zone:

Iop = |∑ Ik|

Here, Iop is the operating current in amperes and Ik is the current contributed by terminal k, expressed with a consistent reference direction. For a two-terminal element, this commonly becomes:

Iop = |I1 + I2|

For an external fault, current enters the zone from one terminal and leaves through another. With ideal CTs, the currents cancel and Iop = 0. For an internal fault, both terminal currents normally contribute to the fault, so their sum is substantial.

A restraining quantity is often used to prevent unwanted operation during heavy external faults. A common definition for a two-terminal relay is:

Irest = (|I1| + |I2|) / 2

The exact definition varies among manufacturers. The relay operates when the differential current exceeds both a minimum pickup and a proportion of the restraining current:

Iop ≥ Ipickup + S × Irest

where Ipickup is the minimum operating current in amperes and S is the bias slope, dimensionless. This equation is a simplified representation of a percentage-biased characteristic; actual relays may use multiple slopes, breakpoint currents, or a polygonal characteristic.

Protected-zone boundaries

The CTs define the electrical limits of a unit-protection zone. If a fault occurs between the CTs, it is an internal fault. If it occurs outside them, it is an external or through-fault. CT placement therefore determines whether a device is protected by differential protection and whether adjacent zones overlap.

Overlapping zones are intentional. For example, a busbar zone may overlap transformer or feeder zones so that no section of conductor is left unprotected. Correct polarity, phase identification, ratio matching, and wiring are essential because a wiring error can create a false differential current during normal load.

2. Circulating-current and current-balance schemes

In the classical circulating-current arrangement, CT secondaries are connected so that, under healthy or external-fault conditions, their secondary currents circulate between CTs rather than through the relay. For an internal fault, the currents no longer balance and the relay receives operating current.

An ideal two-terminal external fault has:

Is1 = Is2

with opposing relay contributions. Consequently, the relay current is zero. For an internal fault, the relay current is approximately the sum of the secondary fault-current contributions, subject to source contributions and CT errors.

The ideal scheme is selective and fast, but it is not truly ideal in practice. CT ratios may differ, magnetizing currents are not identical, lead resistances vary, and CTs can saturate asymmetrically. These errors create spill current. A relay must therefore be designed to tolerate normal mismatch while remaining sensitive to internal faults.

3. Percentage-biased differential protection

Percentage bias, also called restrained differential protection, improves security during high-current external faults. The relay characteristic is generally represented on an operating-current versus restraining-current graph. A typical two-slope characteristic has a low-current region for sensitivity and a high-current region with increased restraint.

A simplified two-slope equation is:

Iop ≥ I0 + S1 Irest, for Irest ≤ Ib
Iop ≥ I0 + S1 Ib + S2(Irest − Ib), for Irest > Ib

I0 is the pickup current, S1 and S2 are low- and high-current slopes, and Ib is the breakpoint restraining current. All currents are in amperes; slopes are dimensionless.

The second slope is larger because CT errors and saturation become more significant as through-fault current increases. The relay thus remains sensitive at low fault levels while becoming more secure at high current.

CT saturation and stability

A CT reproduces primary current only while its magnetic core remains within an appropriate operating range. During a severe external fault, the current may contain a decaying direct-current component. This produces an asymmetric flux excursion and can saturate a CT. A saturated CT delivers a distorted secondary current, often with missing portions of the waveform.

If one CT saturates more than another during an external fault, the relay may see a false differential current:

Iop,error = |Is1,error + Is2,error|

The scheme is stable if this false operating current remains below the relay characteristic. Stability is not a property of the relay alone; it depends on CT excitation capability, secondary resistance, lead resistance, fault current, remanent flux, relay burden, and the selected characteristic.

A useful engineering check compares the voltage required to force the necessary secondary current through the CT circuit with the CT excitation capability. If the required voltage is too high, saturation is likely. For an approximate loop:

Vreq ≈ Is(Rct + Rlead + Rrelay)

where Is is the secondary current, Rct is CT secondary winding resistance, Rlead is total lead resistance, and Rrelay is relay burden resistance. This is an approximate resistive check and does not replace a complete transient CT analysis.

Worked example: biased stability check

A two-terminal differential relay has Ipickup = 0.20 A and a bias slope of 40%. During an external fault, the measured currents are 8 A and 7.6 A on a common secondary base. Using:

Iop = |8 − 7.6| = 0.4 A

and:

Irest = (8 + 7.6)/2 = 7.8 A

The operating threshold is:

Ithreshold = 0.20 + 0.40 × 7.8 = 3.32 A

Since 0.4 A < 3.32 A, the relay remains stable. The 0.4 A spill current represents measurement mismatch rather than an internal fault.

4. High-impedance differential protection

A high-impedance differential scheme connects CT secondaries in parallel and places a high-impedance relay element in the differential path. A stabilizing resistor is commonly included. During an external fault, the scheme is designed so that the voltage required to drive current through the relay is high enough that CT secondary currents circulate through the CT circuits rather than operating the relay.

For a simplified worst-case external-fault condition, the stabilizing voltage may be estimated by:

Vs ≥ If(Rct + Rlead)

where If is the maximum secondary through-fault current that may be forced through the saturated CT branch, Rct is the relevant CT secondary resistance, and Rlead is the resistance of the associated lead loop. The exact design equation depends on the CT arrangement and the fault model.

The relay operating current is approximately:

Ir = Vs / (Rs + Rr)

where Rs is the stabilizing resistance and Rr is the relay-element resistance. A high-impedance scheme usually requires matched CT ratios and similar CT characteristics. It is particularly attractive for busbar and restricted earth-fault protection because it can achieve high sensitivity and excellent stability.

A major limitation is that the CT circuit can develop a dangerous voltage during an external fault. CT secondary circuits must never be open-circuited while energized. Proper shorting links, insulation coordination, supervision, and maintenance procedures are mandatory.

5. Low-impedance differential protection

Low-impedance, or numerical biased differential protection measures individual CT currents electronically and calculates operating and restraining quantities. It does not require all CT secondaries to be connected in a single high-impedance circuit. This permits flexible ratio compensation, vector-group compensation, phase correction, multiple slopes, and selective treatment of CT saturation.

Advantages include:

  • tolerance of moderate CT-ratio differences;
  • compensation for transformer ratio and phase displacement;
  • multiple restraint slopes;
  • disturbance recording and event analysis;
  • integration with breaker-failure and communication functions.

However, low impedance does not eliminate CT saturation. The relay algorithm must distinguish internal faults from transient spill current using waveform analysis, restraint logic, saturation detection, or time-domain methods. Settings must also account for instrument-transformer accuracy, wiring, communications, and system fault levels.

6. Transformer differential protection

Transformer differential protection requires more compensation than a simple line or busbar scheme. The transformer changes current magnitude and phase angle, and transformer winding connections introduce phase displacement. Numerical relays generally perform these corrections internally using transformer ratings, winding voltages, vector group, CT ratios, and grounding connections.

Important sources of false differential current include:

  1. CT ratio mismatch.
  2. Transformer tap-changer movement.
  3. Vector-group phase displacement.
  4. Magnetizing inrush.
  5. Overexcitation.
  6. Zero-sequence current behavior.
  7. CT saturation during external faults.

Magnetizing inrush restraint

When a transformer is energized, the core flux can become highly asymmetric. The resulting magnetizing current flows mainly from the source side and appears as differential current, even though there is no internal winding fault. Traditional relays used second-harmonic restraint; modern relays may use waveform and flux-based methods. The restraint principle must remain secure for energization but must not block an internal fault whose current contains harmonic components.

Overexcitation can produce a high fifth-harmonic component and excessive core flux. Many transformer relays use harmonic supervision or a separate volts-per-hertz function, although harmonic content alone should not be treated as infallible under all system conditions.

7. Busbar differential protection

A busbar fault is especially severe because many sources can feed the fault simultaneously. Busbar protection must therefore be fast, secure, and capable of accommodating changing breaker and disconnect-switch configurations.

In a simple single-bus arrangement, feeder CT currents are summed. Under normal operation or an external feeder fault, the sum should be zero. For a bus internal fault, current from multiple circuits flows into the fault and produces differential current.

Modern busbar systems commonly use:

  • low-impedance numerical differential protection;
  • zone selection based on disconnector status;
  • check zones for security;
  • breaker-failure initiation;
  • redundant main protection systems.

The check zone is an independent busbar-wide measurement used to confirm that a selected local zone is genuinely faulted. It reduces the risk of incorrect tripping caused by faulty switch-position indication or wiring errors. The protection design must address evolving bus configurations, transfer buses, bus couplers, and maintenance conditions.

8. Restricted earth-fault protection

Restricted earth-fault (REF) protection is a unit scheme for detecting earth faults within a defined portion of a transformer or generator winding. It compares the residual current from phase CTs with the neutral CT current.

The residual phase current is:

I0phase = IA + IB + IC

where IA, IB, and IC are phase currents. For a correctly wired system, this residual current balances the neutral current for external earth faults. An earth fault inside the protected zone produces spill current and operates the relay.

REF is more sensitive than phase differential protection for faults close to a grounded neutral, because the fault current may be too small to create strong phase-differential current. It may be implemented as high impedance or low impedance. High-impedance REF demands careful CT matching and a stable secondary circuit; low-impedance REF offers greater setting flexibility and numerical compensation.

REF is restricted: an earth fault outside the CT boundaries should not operate it. The grounding arrangement, neutral CT location, transformer vector group, and zero-sequence current paths must be included in the design.

9. Line differential protection

Line differential protection compares currents at two or more line terminals. Because the terminals are geographically separated, a communication channel is required. The relay may exchange sampled values, phasors, or processed operating quantities through fiber-optic, microwave, pilot-wire, or other suitable channels.

A line differential scheme provides excellent selectivity and fast clearing, but performance depends on channel availability, synchronization, latency, bandwidth, and security against communication errors. Modern relays include channel supervision, echo logic, permissive fallback functions, and a backup distance or overcurrent element.

For a multi-terminal line, the general operating quantity remains:

Iop = |I1 + I2 + ... + In|

All currents must be referred to a consistent phase and polarity convention. A channel failure should not produce an unjustified trip; conversely, a brief channel interruption should not unnecessarily disable all protection without alarm or backup.

10. Stability, dependability, and CT design

Two protection attributes must be balanced:

  • Dependability: the ability to trip for all faults within the zone.
  • Security: the ability to avoid tripping for faults outside the zone and for non-fault disturbances.

The principal CT-related design checks are:

  1. Maximum internal-fault current and minimum internal-fault current.
  2. Maximum external through-fault current.
  3. CT ratio and accuracy class.
  4. CT excitation curve and knee-point voltage where applicable.
  5. DC offset and remanent flux.
  6. Total secondary resistance and relay burden.
  7. CT saturation effects on both internal and external faults.
  8. Thermal and mechanical withstand.

For high-impedance schemes, the CTs should generally have closely matched ratios and excitation characteristics. For low-impedance numerical schemes, mismatch can be compensated to a degree, but the allowable error must be verified against the relay characteristic and fault studies.

Protection engineers should not rely solely on nominal CT accuracy class. The actual secondary circuit burden, cable length, terminal connections, saturation behavior, and transient fault current are decisive.

11. Illustrative case study: transformer differential misoperation

Background

An industrial plant experiences an unwanted trip of a 33/11 kV transformer differential relay during a fault on an 11 kV outgoing feeder. The feeder fault is outside the transformer differential zone.

Problem

The relay record shows substantial differential current during the external fault. The transformer was correctly energized, and no internal damage was found.

Analysis

The 11 kV-side CT associated with the faulted feeder saturated because of high asymmetric fault current. The opposite transformer-side CT remained approximately linear. The unequal secondary currents created spill current. The relay had a single low bias slope that was insufficient for the measured transient mismatch. In addition, the CT secondary leads were longer than assumed in the original design.

Approach and solution

The protection team:

  1. Verified CT polarity and ratios by primary injection.
  2. Measured CT secondary resistance and lead resistance.
  3. Recalculated the maximum through-fault duty.
  4. Reviewed the relay’s two-slope characteristic and saturation security.
  5. Introduced a higher high-current bias slope and appropriate transient restraint.
  6. Confirmed transformer vector-group and tap-range compensation.
  7. Tested internal and external faults using a relay test set and recorded waveforms.

The settings were accepted only after confirming that the minimum internal earth fault remained detectable and that the external-fault case remained stable.

Outcome

The revised scheme remained stable during simulated through-faults and operated correctly for injected internal-fault conditions. The investigation also identified inadequate documentation of CT lead resistance.

Learning points

Differential security cannot be established from relay settings alone. CT saturation, wiring, transformer data, and realistic transient testing must be considered together. A change that increases security must always be checked against dependability for low-current internal faults.

12. Testing, commissioning, and maintenance

Commissioning should include verification of:

  • CT ratio, polarity, phase identification, and insulation;
  • secondary-loop continuity and resistance;
  • relay configuration and compensation;
  • trip outputs and breaker interfaces;
  • communication channel performance;
  • busbar zone selection and isolator logic;
  • internal-fault operation and external-fault stability;
  • disturbance-recording triggers and time synchronization.

Primary injection is the most comprehensive method because it tests the complete current path. Secondary injection is efficient for relay logic and characteristic verification but may not reveal CT wiring or polarity errors. End-to-end testing is essential for line differential schemes.

Protection settings and logic should be version-controlled. Any modification to CT circuits, transformer taps, bus arrangements, communication equipment, or breaker wiring should trigger a protection-impact assessment.

13. Standards, safety, and professional responsibility

Differential protection must be designed and tested in accordance with the applicable utility rules, national standards, manufacturer instructions, and relevant IEC or IEEE practices. Exact requirements differ by jurisdiction and equipment class.

CT secondary circuits present serious hazards. An open CT secondary can develop a dangerous voltage, damage insulation, and endanger personnel. Isolation procedures must short the CT before disconnecting a relay or test instrument. Trip circuits and busbar protection require controlled work permits because an incorrect test connection can trip multiple breakers or leave a critical zone unprotected.

Protection engineers have a professional duty to document assumptions, identify uncertainty, preserve test records, and communicate temporary impairments. Cybersecurity is also relevant to numerical relays and communication-assisted protection: access control, secure configuration, event logging, and controlled firmware management are necessary safeguards.

14. Contemporary developments

Digital substations increasingly use process-bus sampled values and digitally communicated status signals. This can reduce copper wiring and improve observability, but it introduces requirements for time synchronization, network redundancy, traffic management, interoperability, and cybersecurity.

Adaptive protection can modify settings according to network topology or operating condition. Such systems may improve sensitivity and security, but their logic must fail safely when communications or topology information is unavailable. Artificial intelligence is being investigated for disturbance classification and CT-saturation recognition; however, protective tripping should remain explainable, validated, deterministic where required, and supported by conventional backup functions.

Chapter Summary

Differential and unit protection compare currents at the boundaries of a defined electrical zone. Internal faults create differential current, while external faults should be rejected. Percentage bias improves security by increasing restraint as through-fault current rises. High-impedance schemes depend on a carefully designed CT circuit and are widely used for busbar and REF applications. Low-impedance numerical schemes provide flexible compensation, multiple slopes, event recording, and sophisticated saturation treatment.

CT saturation is the principal threat to stability during external faults. It is influenced by fault asymmetry, remanent flux, CT excitation capability, secondary resistance, and relay burden. Transformer differential protection must additionally address ratio, phase displacement, tap changes, magnetizing inrush, and overexcitation. Busbar protection requires fast operation and reliable zone selection, while line differential protection depends on secure communications. Correct design combines fault studies, CT analysis, relay settings, wiring verification, commissioning tests, safety procedures, and ongoing maintenance.

Key Terms

Differential protection: Protection that operates from the difference or sum of currents entering a defined zone.

Unit protection: Protection restricted to a specific electrical zone using measurements from its boundaries.

Protected zone: Portion of the power system between the relevant CT locations.

Operating current: Current quantity that drives a differential relay toward operation.

Restraining current: Current quantity used to improve security during heavy through-faults.

Percentage bias: Restraint that increases in proportion to measured current.

Through-fault: A fault outside the protected zone that causes large current to pass through it.

CT saturation: Nonlinear failure of a current transformer to reproduce primary current accurately.

Spill current: Residual current caused by imperfect current cancellation.

High-impedance differential: Scheme using a high-impedance relay circuit and stabilizing resistor.

Low-impedance differential: Numerical or electromechanical scheme that calculates differential current with comparatively low circuit impedance.

Busbar protection: Protection for faults on a substation bus section.

Restricted earth fault: Sensitive unit protection for earth faults within a defined grounded winding zone.

Magnetizing inrush: Transient current drawn when a transformer is energized.

Vector-group compensation: Correction for transformer phase displacement and winding connection.

Dependability: Ability to operate for faults within the intended zone.

Security: Ability to remain stable for external faults and non-fault conditions.

Check zone: Independent busbar-wide supervision used to confirm a local busbar trip decision.

Primary injection: Testing by applying current through the primary circuit or CT primary.

Review Questions

Short Answer Questions

  1. What is the operating principle of unit differential protection?
  2. Why is percentage bias used in a differential relay?
  3. Define CT saturation and identify two causes of it during an external fault.
  4. What is the purpose of a check zone in busbar protection?
  5. Why is REF generally more sensitive than ordinary phase differential protection for earth faults near a transformer neutral?

Long Answer Questions

  1. Explain the construction, operation, advantages, and limitations of high-impedance differential protection.
  2. Describe how transformer ratio, vector group, tap position, magnetizing inrush, and overexcitation affect transformer differential protection.
  3. Compare high-impedance and low-impedance differential protection for busbar and REF applications.
  4. Explain the causes of false differential current during an external fault and describe methods for improving stability.
  5. Discuss the testing and commissioning requirements for a line differential protection scheme.

Analytical / Application Questions

  1. A relay has Ipickup = 0.15 A, bias slope 50%, and measured currents of 6.0 A and 5.7 A during an external fault. Calculate Iop, Irest, and the operating threshold. Decide whether the relay should operate under the stated simplified characteristic.
  2. A transformer differential relay trips during a feeder fault outside its zone. Develop a systematic investigation covering CT performance, polarity, wiring resistance, relay settings, transformer compensation, and event records.
  3. Select a suitable protection arrangement for a double-bus substation with multiple feeders and a bus coupler. Justify the use of main zones, a check zone, zone-selection logic, breaker-failure protection, and communication or redundancy requirements.

Multiple Choice Questions with Answers

Question 1

For an ideal external fault in a two-terminal differential zone, what is the expected differential current?

A. Approximately zero

B. Equal to the full through-fault current

C. Equal to twice the load current

D. Equal to the CT excitation current only

Correct Answer: A. Approximately zero

Question 2

What is the main purpose of percentage bias in differential protection?

A. To increase transformer magnetizing current

B. To improve security during heavy through-faults

C. To eliminate the need for CTs

D. To provide breaker-failure protection

Correct Answer: B. To improve security during heavy through-faults

Question 3

A relay has a pickup of 0.20 A and a 40% bias slope. If the restraining current is 5 A, what is the simplified operating threshold?

A. 0.40 A

B. 1.20 A

C. 2.20 A

D. 5.20 A

Correct Answer: C. 2.20 A

Question 4

Which condition most directly causes a false differential current during an external fault?

A. Equal CT excitation characteristics

B. Asymmetric saturation of one CT

C. Correct polarity wiring

D. A low relay burden

Correct Answer: B. Asymmetric saturation of one CT

Question 5

Which protection scheme commonly uses a stabilizing resistor and a high-impedance relay element?

A. High-impedance busbar differential

B. Distance protection

C. Overfrequency protection

D. Under-voltage protection

Correct Answer: A. High-impedance busbar differential

Question 6

Why is transformer differential protection compensated for vector group?

A. To increase system frequency

B. To correct winding phase displacement

C. To prevent breaker contact wear

D. To measure insulation resistance

Correct Answer: B. To correct winding phase displacement

Question 7

What is the principal function of a busbar check zone?

A. To replace all feeder CTs

B. To confirm a local busbar differential trip decision

C. To regulate transformer tap position

D. To provide generator excitation

Correct Answer: B. To confirm a local busbar differential trip decision

Question 8

Restricted earth-fault protection compares which quantities?

A. Voltage and frequency at one bus

B. Phase residual current and neutral current

C. Active power and reactive power

D. Positive- and negative-sequence voltages only

Correct Answer: B. Phase residual current and neutral current

Question 9

Which testing method best checks the complete CT primary-to-relay current path?

A. Visual inspection only

B. Secondary injection only

C. Primary injection

D. Frequency-response analysis only

Correct Answer: C. Primary injection

Question 10

Why is an open-circuited CT secondary hazardous?

A. It can produce a dangerously high secondary voltage

B. It always reduces primary fault current to zero

C. It disables all transformer cooling

D. It causes system frequency to rise

Correct Answer: A. It can produce a dangerously high secondary voltage

Power System Protection & Switchgear
Chapter 7 of 15