Symmetrical Components and Fault Analysis

Introduction

Power-system faults produce currents and voltages that determine equipment damage, breaker duties, grounding requirements, and protective-relay settings. A balanced three-phase fault can be represented by a single-phase equivalent, but unbalanced faults couple the phase quantities. Direct phase-domain analysis is possible, yet it often obscures the physical roles of grounding, machine impedances, and transformer connections.

The method of symmetrical components resolves any unbalanced set of three-phase phasors into positive-, negative-, and zero-sequence sets. In networks that are balanced before the fault, each sequence can be represented by a separate network. Fault boundary conditions then determine how these networks are interconnected.

All phasors in this chapter are RMS quantities at the fundamental frequency unless stated otherwise. The phase order is assumed to be a–b–c.

Learning Objectives

After studying this chapter, students should be able to:

• transform between phase and sequence quantities;
• construct positive-, negative-, and zero-sequence networks;
• derive and apply equations for balanced and unbalanced shunt faults;
• calculate fault currents and post-fault bus voltages;
• explain the influence of grounding and transformer connections;
• convert per-unit fault results into amperes and equipment duties;
• assess breaker interrupting and making duties;
• use fault results in grounding and protection design; and
• recognize the assumptions and limitations of conventional fault studies.

1. Fault Types and Engineering Consequences

Common shunt faults are classified as follows:

Faults may be bolted, meaning negligible fault impedance, or may include arc, conductor, tower-footing, soil, or neutral impedance. A fault study can be required to determine:

• thermal and electrodynamic equipment stress;
• circuit-breaker interrupting and peak-making duties;
• current-transformer saturation exposure;
• relay pickup, reach, sensitivity, and coordination;
• ground-potential rise and touch or step voltage;
• voltage depressions affecting customers and generators; and
• negative-sequence heating in rotating machines.

2. Symmetrical Components

2.1 Sequence sets

An arbitrary three-phase phasor set can be represented as the sum of three balanced sets:

• Positive sequence: equal magnitudes separated by 120°, with phase order a–b–c.
• Negative sequence: equal magnitudes separated by 120°, with phase order a–c–b.
• Zero sequence: equal phasors in all three phases.

Define the unit phasor operator

The sequence-to-phase transformation for current is

The inverse transformation is

The same transformations apply to voltages. Subscripts 0, 1, and 2 denote zero-, positive-, and negative-sequence quantities, respectively.

An immediate consequence is

Thus, residual or ground current is three times the zero-sequence current. This relationship is fundamental to earth-fault protection.

2.2 Illustrative decomposition

Suppose I_a=6∠0° kA, I_b=3∠-120° kA, and I_c=3∠120° kA. Using the inverse transformation gives

The phase currents therefore contain positive and negative sequence but no zero sequence. Their sum is zero, as expected for a three-wire fault without a ground-return path.

3. Sequence Networks

3.1 Positive-sequence network

The positive-sequence network represents balanced normal operation. Rotating machines contribute positive-sequence internal emfs. At a faulted bus, the pre-fault network is commonly replaced by a Thévenin source E₁ behind the positive-sequence impedance Z₁.

Figure: Complete positive-sequence Thévenin network. The polarized source has both terminals connected, including the reference return path.

With current defined from the source toward the fault, the bus voltage is

3.2 Negative-sequence network

A balanced system contains no negative-sequence source. Consequently,

Negative-sequence impedance Z₂ is often approximately equal to subtransient positive-sequence impedance for static components and round-rotor machines, but manufacturer or study data should be used where available.

3.3 Zero-sequence network

The zero-sequence network has no source in a balanced pre-fault system:

Its topology may differ radically from the positive-sequence network. Zero-sequence current requires a return path through neutrals, grounding impedances, earth, cable sheaths, or grounded transformer windings.

A neutral impedance Z_n appears as 3Z_n in the zero-sequence network because the neutral carries I_a+I_b+I_c=3I₀. The neutral voltage drop is therefore 3I₀Z_n.

Transformer connections are especially important:

• a grounded-wye winding permits zero-sequence current if its neutral has a return path;
• an ungrounded-wye winding blocks external zero-sequence current;
• a delta winding can provide an internal circulating path but prevents zero-sequence line current from passing through its terminals;
• grounding transformers create a controlled zero-sequence path.

4. Per-Unit Quantities and Study Assumptions

For a three-phase base power S_B and line-to-line base voltage V_LL,B,

Here, Z_B is in ohms when volts and volt-amperes are used, and I_B is in amperes. Sequence quantities use the same three-phase power base and corresponding voltage base.

A conventional short-circuit study generally assumes:

1. the system was balanced before the fault;
2. network elements are linear over the interval analyzed;
3. frequency remains nominal;
4. loads and shunts are represented consistently with the study standard;
5. pre-fault bus voltage is known or approximated;
6. sequence impedances correspond to the required time frame, such as subtransient or transient conditions.

5. Fault Boundary Conditions and Network Connections

Let phase a be the reference phase, and let currents be positive from the network into the fault. The Thévenin equations at the faulted bus are

The formulas below apply to shunt faults at one bus. Z_f denotes the physical fault-path impedance specified for the fault under consideration.

5.1 Balanced three-phase fault

Only the positive-sequence network is involved. For equal per-phase fault impedance Z_f,

For a bolted fault, Z_f=0. The three phase currents are balanced and equal in magnitude to I₁.

5.2 Single-line-to-ground fault

For an a-to-ground fault,

From the component transformation, I₀=I₁=I₂=I_a/3. Therefore, the three sequence networks are connected in series:

and

The factor 3Z_f arises because the same sequence current flows through all three sequence networks while the physical fault current is 3I₀.

5.3 Line-to-line fault

For a b–c fault through impedance Z_f, zero-sequence current is absent:

The sequence current is

The faulted-phase currents satisfy I_b=-I_c and

5.4 Double-line-to-ground fault

For a b–c-to-ground fault in which both faulted phases meet at a common point connected to ground through Z_f,

The other sequence currents are

The total ground current is I_g=I_b+I_c=3I₀. These expressions assume that the unfaulted phase carries no fault current and that both faulted phases share the stated common ground impedance.

6. Calculating Bus Voltages

After sequence currents have been found, sequence voltages follow from the Thévenin equations. Phase voltages are reconstructed using

Voltages at other buses require voltage drops through the sequence networks. The impedance between the source and the bus of interest—not automatically the complete fault-point Thévenin impedance—must be used.

Worked Example 1: SLG current and healthy-phase voltages

A 13.8 kV bus uses a 100 MVA base. The pre-fault phase voltage is E₁=1∠0° pu. At the fault location,

• Z₁=j0.20 pu,
• Z₂=j0.20 pu,
• Z₀=j0.05 pu,
• Z_f=0.

For a bolted a-to-ground fault,

Thus,

The base current is

Therefore,

The sequence voltages are

Their sum gives V_a=0, satisfying the bolted-fault boundary condition. Transformation gives

Hence |V_b|=|V_c|=0.882 pu. The example shows that healthy-phase voltage depends strongly on zero-sequence impedance; it does not necessarily rise to the line-to-line voltage.

Worked Example 2: Line-to-line fault

Let E₁=1.0 pu, Z₁=j0.25 pu, Z₂=j0.20 pu, and Z_f=0. Then

The current in either faulted phase is

Zero-sequence current is zero because no ground-return path is involved.

7. Grounding Effects

Grounding chiefly changes the zero-sequence network. It can therefore alter SLG and DLG currents without significantly changing a three-phase fault current.

7.1 Solid, impedance, and resonant grounding

• Solid grounding usually produces substantial earth-fault current and straightforward relay detection.
• Resistance grounding limits damage and arc energy while maintaining detectable current.
• Reactance grounding limits current but affects transient behavior and relay quantities.
• Resonant grounding uses a Petersen coil to compensate network capacitive earth-fault current near the operating frequency.
• Ungrounded systems may continue operating temporarily after one earth fault, but healthy-phase-to-ground voltages can approach \sqrt{3} times normal and transient overvoltages require attention.

For a generator or transformer neutral grounded through Z_n, its contribution to zero-sequence impedance is 3Z_n. Omitting this multiplier is a common and consequential modeling error.

7.2 Grounding design trade-offs

Increasing neutral resistance reduces fault current and equipment damage, but may reduce relay sensitivity. Grounding design must coordinate:

• fault-current magnitude;
• insulation stress;
• touch and step voltage;
• arc-flash exposure;
• continuity requirements; and
• dependable fault detection.

8. Equipment Duty and Protection Application

8.1 Symmetrical and asymmetrical current

Sequence calculations normally provide the symmetrical RMS AC component. The actual first-cycle current may include a decaying DC offset determined by fault inception angle and system X/R ratio.

Under the IEC 60909 peak-current method, one commonly used approximation for a single equivalent series path is

Here, I_k'' is the initial symmetrical RMS short-circuit current, i_p is peak current, and R/X belongs to the applicable equivalent fault path. The precise standard method, network frequency, meshed-network correction, and equipment rating convention must be observed; this approximation must not replace a standards-compliant study.

A breaker assessment should check at least:

• rated maximum voltage;
• symmetrical interrupting current at contact separation;
• short-time withstand current and duration;
• peak or making-current capability;
• operating sequence; and
• transient recovery voltage.

8.2 Protection implications

Fault results support selection and coordination of:

• phase overcurrent elements;
• ground or residual overcurrent elements;
• negative-sequence elements;
• distance protection;
• differential protection; and
• breaker-failure schemes.

Maximum fault current is used for equipment duty and CT saturation assessment. Minimum fault current, including credible high fault resistance and weak-source conditions, is vital for relay sensitivity. Settings must not be based only on a bolted maximum-current fault.

9. Applied Case: 11 kV Industrial Substation

This is an illustrative engineering case.

Background and problem

An industrial facility is adding generation to an 11 kV bus. Engineers must verify the bus breaker and determine whether ground protection remains suitable. On a 100 MVA base, the equivalent impedances and voltage are

• E₁=1.05 pu;
• Z₁=0.08+j0.32 pu;
• Z₂=0.08+j0.32 pu;
• Z₀=0.04+j0.12 pu;
• SLG fault impedance Z_f=0.02 pu.

The breaker is rated 25 kA symmetrical RMS and 65 kA peak making current at the applicable voltage.

Analysis

For an a-to-ground fault,

Its magnitude is 0.8032 pu. Therefore,

and

The current base is

Thus, the SLG current is

For comparison, the bolted three-phase current is

The ground fault is more severe because the relatively low zero-sequence impedance creates a strong return path.

Using the SLG equivalent X/R=0.76/0.26=2.923,

and the approximate peak duty is

Approach and outcome

The calculated 20.58 kA symmetrical current is below the breaker's 25 kA rating, and the approximate 39.9 kA peak is below its 65 kA making rating. These comparisons are necessary but not sufficient: the engineer must also verify contact-separation current, transient recovery voltage, short-time withstand, study-standard corrections, and rating margins.

Protection engineers should use the 20.58 kA result for maximum ground-fault duty and conduct a separate minimum-fault study including source outages and higher fault resistance. Residual CT performance must be checked because DC offset can promote saturation.

Learning points

1. An SLG fault can exceed a three-phase fault when Z₀ is sufficiently low.
2. Fault impedance belongs in the SLG denominator as 3Z_f.
3. RMS interrupting and peak-making duties are distinct.
4. Maximum and minimum system configurations serve different protection purposes.

10. Study Workflow and Best Practices

A defensible fault study should follow a controlled process:

1. Establish scope, operating configurations, study standard, and time frame.
2. Select consistent power and voltage bases.
3. Validate utility, generator, motor, transformer, cable, and grounding data.
4. Construct all three sequence networks, including transformer zero-sequence paths.
5. Calculate maximum and minimum credible faults.
6. Reconstruct bus voltages and branch currents where protection requires them.
7. Evaluate interrupting, momentary, thermal, and mechanical duties.
8. Check relay sensitivity, selectivity, speed, CT performance, and backup.
9. Perform independent review and document assumptions and data provenance.
10. Update the model after network, grounding, or generation changes.

Software should be validated against benchmark cases or hand calculations. Automated results are not reliable when transformer vector groups, neutral impedances, or equipment status are modeled incorrectly.

11. Limitations and Contemporary Developments

Symmetrical-component analysis is exact for linear, sinusoidal, three-phase networks expressed through phase impedance matrices. The convenient use of three independent sequence networks additionally requires a cyclically symmetric network. Untransposed lines, asymmetrical mutual coupling, nonlinear arcs, saturation, and frequency-dependent behavior may require phase-domain or electromagnetic-transient models.

Inverter-based resources complicate traditional assumptions. Their fault current may be limited, controlled, and dependent on positive- and negative-sequence voltage, control software, grid code, and ride-through mode. A fixed voltage source behind reactance may therefore misrepresent their response. Manufacturer-validated dynamic or EMT models may be required.

Modern digital relays estimate sequence quantities in real time. Synchrophasor data, event records, and traveling-wave measurements can improve fault location and model validation. Machine-learning tools may help classify events or identify anomalous data, but they should not replace deterministic duty calculations, protection-security checks, or engineering accountability.

12. Governance, Safety, and Professional Responsibility

Fault studies directly affect public and worker safety. Applicable IEC, IEEE, ANSI, national wiring, grid-code, and occupational-safety requirements must be identified for each project. Engineers should preserve model versions, assumptions, equipment data, approval records, and change histories.

Uncertainty must not be hidden by reporting excessive numerical precision. Conservative assumptions should be technically justified rather than applied indiscriminately, because excessive protection sensitivity can cause unnecessary outages while insufficient sensitivity can leave dangerous faults uncleared. Final settings and equipment selections require authorized engineering review.

Chapter Summary

Symmetrical components transform unbalanced phase quantities into positive-, negative-, and zero-sequence sets. Balanced pre-fault systems can then be represented by sequence networks. The positive-sequence network contains the Thévenin source, while negative- and zero-sequence networks are passive. Fault boundary conditions determine their interconnection: only positive sequence is used for balanced faults; positive and negative sequence are used for line faults; all three sequences participate in ground faults.

Fault currents must be converted from per unit to physical units, and sequence voltages must be transformed back to phase quantities when insulation performance or relay behavior is assessed. Grounding and transformer connections control the zero-sequence network and can make an earth fault more severe than a three-phase fault. Engineering application requires assessment of symmetrical RMS current, asymmetrical peak duty, breaker capability, CT performance, and protection sensitivity under both maximum and minimum fault conditions.

Key Terms

Symmetrical components: Transformation of unbalanced phase phasors into three balanced sequence sets.

Positive sequence: Balanced set having the normal a–b–c phase order.

Negative sequence: Balanced set having the reverse a–c–b phase order.

Zero sequence: Three equal phasors with no angular displacement.

Sequence network: Equivalent circuit representing one sequence of voltage and current.

Thévenin impedance: Equivalent impedance viewed from the fault location.

Bolted fault: Fault having negligible fault-path impedance.

Single-line-to-ground fault: Connection of one phase to ground.

Line-to-line fault: Connection between two phases without ground involvement.

Double-line-to-ground fault: Connection of two phases to a common ground path.

Residual current: Sum of the three phase currents, equal to 3I₀.

Fault impedance: Impedance of the arc, conductor, earth, or other physical fault path.

Neutral grounding impedance: Impedance connecting a system neutral to ground.

Subtransient reactance: Initial machine reactance used for early short-circuit current.

Symmetrical RMS current: RMS value of the AC component excluding DC offset.

Peak making current: Highest instantaneous current a closing device must withstand.

Interrupting rating: Specified fault current a breaker can safely interrupt.

Transient recovery voltage: Voltage appearing across breaker contacts after interruption.

Petersen coil: Grounding reactor used to compensate capacitive earth-fault current.

Review Questions

Short Answer Questions

1. Define positive-, negative-, and zero-sequence phasor sets.
2. Why does a balanced three-phase fault use only the positive-sequence network?
3. Why is a physical neutral impedance represented as three times its value in the zero-sequence network?
4. What is the relationship between residual current and zero-sequence current?
5. Distinguish symmetrical interrupting current from peak making current.

Long Answer Questions

1. Derive the symmetrical-component transformation and its inverse for an a–b–c phase sequence.
2. Derive the current equation for an a-to-ground fault through nonzero fault impedance.
3. Explain how grounded-wye, ungrounded-wye, and delta transformer windings affect zero-sequence current.
4. Discuss how fault studies support circuit-breaker selection and protective-relay coordination.
5. Critically evaluate the assumptions of conventional sequence-network fault analysis for inverter-dominated systems.

Analytical / Application Questions

1. A 50 MVA, 33 kV bus has E₁=1.0 pu, Z₁=j0.18 pu, Z₂=j0.20 pu, and Z₀=j0.12 pu. Calculate the bolted a-to-ground fault current in per unit and kiloamperes.
2. For E₁=1.05 pu, Z₁=j0.25 pu, Z₂=j0.20 pu, and Z_f=j0.05 pu, calculate the current magnitude in each faulted phase for a line-to-line fault.
3. Compare the protection consequences of changing a transformer neutral from solid grounding to resistance grounding. Address maximum duty, minimum relay sensitivity, healthy-phase voltage, and ground-potential rise.

Multiple Choice Questions with Answers

Question 1

For an a–b–c system, which expression gives the zero-sequence current?

A. I₀=(I_a+I_b+I_c)/3

B. I₀=I_a+aI_b+a²I_c

C. I₀=I_a-I_b-I_c

D. I₀=3(I_a+I_b+I_c)

Correct Answer: A. I₀=(I_a+I_b+I_c)/3

Question 2

Which sequence network contains the Thévenin voltage source for a balanced pre-fault system?

A. Zero-sequence network only

B. Positive-sequence network only

C. Negative-sequence network only

D. All three sequence networks

Correct Answer: B. Positive-sequence network only

Question 3

For a bolted single-line-to-ground fault, the three sequence networks are connected in what manner at the fault?

A. Positive and negative in parallel; zero disconnected

B. All three in parallel

C. All three in series

D. Positive only

Correct Answer: C. All three in series

Question 4

A generator neutral is grounded through Z_n=j0.04 pu. What impedance does this grounding element contribute to the zero-sequence network?

A. j0.0133 pu

B. j0.04 pu

C. j0.08 pu

D. j0.12 pu

Correct Answer: D. j0.12 pu

Question 5

For a line-to-line fault with E₁=1 pu, Z₁=j0.20 pu, Z₂=j0.30 pu, and Z_f=0, what is |I₁|?

A. 2.00 pu

B. 3.33 pu

C. 5.00 pu

D. 1.15 pu

Correct Answer: A. 2.00 pu

Question 6

Which condition is characteristic of a line-to-line fault without ground involvement?

A. I₀=I₁=I₂

B. I₀=0 and I₂=-I₁

C. I₁=0 and I₀=I₂

D. I₂=0 and I₀=-I₁

Correct Answer: B. I₀=0 and I₂=-I₁

Question 7

On a 100 MVA, 20 kV base, what is the three-phase base current?

A. 2.00 kA

B. 2.89 kA

C. 5.00 kA

D. 8.66 kA

Correct Answer: B. 2.89 kA

Question 8

Which change most directly alters the zero-sequence network while leaving the positive-sequence network largely unchanged?

A. Changing a neutral grounding resistor

B. Changing system frequency from 50 Hz to 60 Hz

C. Reversing the positive phase sequence

D. Increasing the pre-fault voltage angle uniformly

Correct Answer: A. Changing a neutral grounding resistor

Question 9

Why must minimum-fault conditions be studied for protective relaying?

A. They determine only breaker peak-making duty

B. They always produce the largest CT saturation

C. They test whether relays remain sensitive to credible faults

D. They eliminate the need to model fault resistance

Correct Answer: C. They test whether relays remain sensitive to credible faults

Question 10

For an SLG fault, E₁=1 pu, Z₁=j0.2 pu, Z₂=j0.2 pu, Z₀=j0.1 pu, and Z_f=0. What is the phase-a fault-current magnitude?

A. 2 pu

B. 5 pu

C. 6 pu

D. 15 pu

Correct Answer: C. 6 pu