6. Protection Coordination and Settings Engineering

Introduction and context

Protection coordination is the engineering process by which protective devices are selected and adjusted so that the device nearest to a fault operates first, while upstream devices remain available as backup. Settings engineering converts system studies, equipment ratings, protection philosophies, and operating constraints into numerical relay, fuse, and circuit-breaker settings. It is therefore both an analytical discipline and a practical design activity.

A protection system must satisfy two apparently competing requirements. It must operate rapidly enough to limit damage, instability, and arc-flash energy, yet it must not trip unnecessarily for load current, transformer energisation, motor starting, transient swings, or faults outside its assigned zone. Coordination is achieved through discrimination in time, current, direction, impedance, or a combination of these quantities.

The subject is closely connected with short-circuit analysis, equipment specifications, system grounding, power-system operation, and maintenance. A setting that is satisfactory in one network configuration may become insecure or excessively slow after a line is opened, a transformer is removed, or distributed generation changes the fault-current contribution. Consequently, coordination is not a one-time calculation; it is a controlled engineering process supported by studies, field verification, and periodic review.

Learning objectives

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

• explain the principles of selectivity, sensitivity, speed, dependability, and security;
• coordinate fuses, circuit breakers, overcurrent relays, and directional relays;
• calculate pickup current, time-dial settings, grading margins, and approximate reach;
• distinguish phase and earth-fault coordination requirements;
• assess the effects of topology changes, distributed generation, and transformer inrush;
• use time-current curves and coordination software critically rather than mechanically;
• identify common settings errors and develop defensible engineering recommendations.

The coordination problem: selectivity under uncertainty

Protection coordination begins by defining the protection zones and the preferred sequence of operation. For a radial feeder, a fault should normally be cleared by the feeder relay or fuse. The upstream transformer or incomer protection should operate only if the downstream device fails or the fault lies outside the feeder zone. This arrangement provides service continuity and limits the number of disconnected customers.

The principal performance attributes are:

Selectivity (discrimination) is the ability of the correct device to operate while healthy sections remain energised. Sensitivity is the ability to detect the minimum fault current for which operation is required. Speed is the time taken to issue and complete the trip. Dependability is the probability of operating when required, whereas security is the ability to avoid operation when operation is not required. These attributes cannot always be maximised simultaneously. Increasing pickup sensitivity, for example, may reduce security during heavy loading or transformer energisation.

Coordination is usually established for the credible system states that have been defined by the planning and operations teams. These may include maximum and minimum generation, normal and contingency topology, transformer in or out of service, bus-section breaker open or closed, and credible distributed-generation dispatch. A setting group may be required for materially different states.

Essential quantities and terminology

The pickup current is the minimum measured current at which an overcurrent element is intended to start or operate, depending on relay design. The plug setting multiplier (PSM) is commonly expressed as

PSM = fault current seen by the relay / relay pickup current.

The time multiplier setting (TMS) or time dial scales the operating time of an inverse-time characteristic. A grading margin is the time interval allowed between a primary device and its backup to accommodate breaker interruption, relay overtravel, current-transformer errors, relay tolerances, and safety allowance.

The reach of a distance element is the apparent impedance boundary within which the relay operates. For a line protection zone, reach is normally specified as a percentage of the protected line impedance. Underreach means that the relay fails to cover the intended boundary; overreach means that it operates beyond it. Zone 1 is commonly instantaneous and set below the remote-bus boundary, while delayed Zones 2 and 3 provide remote backup.

Other important terms include through-fault current, fault contribution, end-of-line fault, inrush current, cold-load pickup, breaker clearing time, CT saturation, residual current, zero-sequence current, directional polarisation, and setting group. Each must be interpreted in the context of the actual protection element and its measurement circuit.

A systematic settings-engineering workflow

Establish the engineering basis

The first step is to collect a reliable single-line diagram, equipment ratings, conductor data, transformer impedances, grounding arrangements, CT and VT ratios, breaker interrupting ratings, relay models, communication channels, and operating configurations. The study must state the voltage bases, frequency, fault assumptions, source impedances, and whether minimum or maximum fault levels are being considered.

Data quality is fundamental. An incorrect CT ratio can make a numerically precise setting practically wrong. Likewise, neglecting a transformer neutral impedance or an inverter fault-current limit can distort earth-fault sensitivity and breaker-duty conclusions.

Define protection objectives and zones

Each device should have a clearly stated primary zone and backup responsibility. The engineer should identify which faults must be cleared instantaneously, which may be delayed, and which devices are permitted to trip for abnormal but non-fault conditions. Protection objectives should be agreed with system operations because the fastest possible trip is not always the preferred operational response.

Determine operating and fault currents

Calculate maximum load, emergency load, motor-starting current, transformer inrush, minimum and maximum short-circuit current, earth-fault current, and contribution from all relevant sources. Both symmetrical and asymmetrical effects may be required for breaker duty. For inverter-based resources, fault current may be controlled, limited, or phase-shifted and should not automatically be treated as a conventional synchronous contribution.

Select pickup values

For a phase overcurrent element, pickup must exceed the highest legitimate current, including permissible overload and transient effects, while remaining below the minimum fault current in the protected zone. A simplified design inequality is

I_load,max × margin < I_pickup < I_fault,min.

The margin is not a universal constant; it depends on load uncertainty, relay accuracy, CT performance, emergency policy, and transient conditions. Earth-fault pickup can often be lower, but residual measurement errors, load unbalance, CT mismatch, and grounding practices must be considered.

For fuses, the continuous rating, minimum-melting curve, total-clearing curve, cable damage curve, and transformer inrush must all be examined. Fuse coordination cannot be inferred from nominal ampere ratings alone.

Select the operating characteristic

Inverse-time overcurrent relays provide shorter operating times at higher fault currents and are useful in radial systems with varying fault levels. A generic inverse characteristic may be written as

t = TMS × K / (PSM^α − 1),

where K and α depend on the selected IEC or IEEE curve. The exact manufacturer equation must be used in a study. Extremely inverse curves may coordinate well with transformer damage and fuse characteristics, but can become slow for low-level faults. Definite-time elements offer predictable delays but may require more stages or communication.

Directional elements are necessary where current can flow from multiple sources, such as parallel feeders, ring networks, and systems with distributed generation. Directional polarisation must remain reliable during close-in faults and voltage depression.

Apply grading margins

If device B is primary and device A is its backup, the backup operating time should satisfy approximately

t_A ≥ t_B + Δt,

where Δt includes breaker opening and arc extinction, relay overtravel and reset, CT error and saturation effects, relay and timer tolerances, and a design allowance. Modern numerical relays may permit smaller margins than electromechanical relays, but the margin must be supported by documented device performance and testing. A nominal curve separation that ignores breaker clearing time is not a valid coordination study.

Verify sensitivity, speed, and equipment withstand

The final settings must be tested at minimum fault current, maximum load, and relevant abnormal states. Clearing time should be compared with conductor thermal limits, transformer through-fault withstand, generator stability requirements, and arc-flash objectives. A setting that coordinates perfectly on a graph may still be unacceptable if it allows damaging energy to persist.

Coordination of radial overcurrent protection

Consider a radial source feeding a transformer, a medium-voltage feeder, and a downstream distribution panel. The downstream fuse or relay is the primary device for faults on its outgoing circuit. The feeder relay is backup, and the transformer or incomer relay is further backup. Coordination is examined from the load toward the source, because downstream settings constrain upstream settings.

The downstream device is first selected to carry load and withstand inrush. The upstream relay is then set above the downstream operating curve by the required grading margin over the fault-current range of interest. The process is repeated upstream. At high fault currents, fuse and breaker total-clearing curves must be used rather than relay pickup curves. At low fault currents, the coordination boundary may be determined by the minimum fault level, not by the maximum short-circuit case.

A common mistake is to coordinate only at the maximum fault current. This can leave inadequate discrimination for a remote, low-current fault. Another is to coordinate only phase elements and assume earth-fault elements will follow the same relationship. Earth-fault current may be limited by grounding impedance and may have a different CT connection, so it requires a separate study.

Distance protection and reach coordination

Distance protection measures apparent impedance, approximately

Z_app = V / I,

with compensation as required for phase-to-earth faults. For a line of impedance Z_L, a first-zone reach may be selected below the remote terminal, often to avoid overreach caused by measurement errors, infeed, mutual coupling, and transients. The exact percentage is a design choice governed by the line and relay application, not a universal rule.

Zone 2 extends beyond the protected line with a time delay and provides remote-end backup. Its reach must account for remote infeed and outfeed because the current seen by the relay may not be proportional to the current in the protected line. Zone 3, where used, must be secured against load encroachment, power swings, and faults on adjacent lines. Loadability checks are therefore essential.

Reach coordination also involves resistive coverage. High-resistance faults may produce an impedance that lies outside a simple mho characteristic. Quadrilateral characteristics can improve resistive coverage but may be more vulnerable to load encroachment or power swings if not properly constrained.

Distributed generation and changing topology

Distributed generation changes both the magnitude and direction of fault current. A downstream generator can cause a feeder relay to see current flowing outward toward the source, reduce the current seen by an upstream relay, or create an unintentional island. Inverter-based resources may contribute only a limited current and may cease contribution rapidly under voltage disturbance. Traditional non-directional coordination may consequently lose selectivity or sensitivity.

The engineering response may include directional overcurrent elements, adaptive setting groups, transfer-trip schemes, permissive communications, revised grounding protection, or controlled generator disconnection. Settings should be checked for minimum generation as well as maximum generation. A relay that is sensitive with all sources connected may fail to detect a remote fault when one source is unavailable.

Topology changes also affect grading. Opening a tie can convert a network into a radial system; closing it can create parallel paths and reverse current. Breaker status, automatic transfer schemes, and bus couplers must therefore be included in the settings philosophy and in commissioning tests.

Coordination software and engineering judgement

Modern coordination software can import network models, calculate fault currents, plot time-current curves, check relay settings, and compare multiple operating scenarios. Useful outputs include one-line diagrams, fault-current reports, device curves, equipment-damage overlays, grading tables, and setting-group comparisons.

Software does not replace engineering judgement. The user must verify model topology, CT and VT ratios, relay curve definitions, breaker times, cable thermal data, and the interpretation of manufacturer curves. Particular care is needed when software uses idealised relay characteristics but the field relay includes definite-time overrides, logic blocking, harmonic restraint, cold-load pickup, or communications-assisted tripping.

A defensible digital workflow includes version-controlled input data, named study cases, documented assumptions, independent review, and export of approved settings. Any change to a relay setting should have a reason, authorisation, effective date, and rollback record. Automated optimisation can minimise operating time or maximise margin, but its objective function must include sensitivity, security, equipment limits, and operational priorities. The mathematically fastest solution may be operationally unsafe.

Illustrative numerical application

Suppose a feeder relay has a CT ratio of 400/1 A and a calculated maximum feeder load of 260 A. The minimum fault current at the remote end is 1.8 kA. If the selected secondary pickup is 0.8 A, the primary pickup is 320 A. The load-to-pickup ratio is 260/320 = 0.8125, providing a reasonable load margin in this illustrative case. At the remote fault, the relay secondary current is 1800/400 = 4.5 A, giving PSM = 4.5/0.8 = 5.625.

The operating time must then be obtained from the selected manufacturer or standard inverse curve using PSM and TMS. If the downstream relay operates in 0.42 s at the same fault and the required grading margin is 0.30 s, the upstream relay should operate no faster than approximately 0.72 s for that coordination point. The result must be checked at other fault levels because inverse curves do not maintain a constant time difference.

This example demonstrates why secondary settings cannot be selected without reference to CT ratio, load, minimum fault current, and the downstream device. It also illustrates that a single coordination point is insufficient for approving a complete setting.

Illustrative case study: industrial plant with embedded generation

Background

An industrial plant is supplied from a utility substation through two parallel 11 kV feeders. The plant contains a 5 MW synchronous generator and several large motors. A tie breaker allows either feeder to supply the complete plant during maintenance.

Problem

After the generator was commissioned, nuisance trips occurred during faults on one feeder. A review found that the original non-directional feeder relays had been coordinated for a single-source radial configuration. With the generator connected, current flowed toward the utility and toward the fault from different locations. In one scenario, the feeder relay nearest the fault did not trip first, while the incomer relay tripped and disconnected the entire plant.

Analysis

Short-circuit studies were performed for maximum and minimum utility strength, generator in service and out of service, tie open and tie closed, and motor contribution included and excluded. The study showed that generator contribution reduced the fault current measured at one feeder relay for certain remote faults. It also caused reverse current through an incomer relay during faults on the adjacent feeder. Existing earth-fault pickup was too high for the minimum utility-source case, while a low pickup used during generator operation could respond to residual load unbalance.

Approach and solution

Directional phase and earth-fault elements were enabled on the incomers and feeders. Forward directions were defined toward the protected feeder sections, and reverse elements were applied where required for backup blocking. Separate setting groups were created for the tie-open and tie-closed states, with interlocking logic supervised by breaker status. Pickup values were reduced only after checking CT saturation, load unbalance, motor starting, and generator grounding. Communication-assisted tripping was added for selected high-speed feeder faults, while time-delayed overcurrent remained as backup.

The settings were tested by secondary injection and by verifying the actual breaker and interlock logic. Operating procedures were revised so that the control room could identify the active setting group before changing the network configuration.

Outcome

The revised scheme restored feeder selectivity for the studied operating states and reduced the need to trip the complete plant for a single feeder fault. The principal improvement was not a single numerical setting but the integration of directional protection, topology awareness, documented setting groups, and commissioning verification.

Learning points

Protection coordination must reflect actual power-flow direction and source availability. Distributed generation can reduce sensitivity as well as increase fault current. Finally, a setting-group design is incomplete unless the logic that selects each group is secure, visible to operators, and tested in the field.

Common errors, risks, and limitations

Important errors include using maximum load instead of emergency load, ignoring motor starting, applying a relay curve incorrectly, confusing primary and secondary current, neglecting CT saturation, treating instantaneous pickup as total clearing time, and failing to include breaker interruption time in the grading margin. Settings copied from a similar installation may be inappropriate because source impedance, grounding, CT class, and topology differ.

Overcoordination is also a risk. Excessive delays improve selectivity on paper but increase equipment damage, arc-flash energy, voltage depression, and the possibility of generator instability. Conversely, aggressive instantaneous settings may cause unwanted tripping for external faults or transformer inrush.

A further limitation is model uncertainty. Fault resistance, inverter controls, arc behaviour, instrument-transformer performance, and future network changes may not be represented precisely. Engineers should therefore state assumptions, examine sensitivity to uncertain parameters, and avoid presenting calculated settings with unjustified precision.

Professional, regulatory, and governance considerations

Protection settings are safety-critical configuration data. They should be approved by competent engineers, independently reviewed where required, and controlled through a formal management-of-change process. Field settings must match approved study files, and discrepancies should be treated as technical findings rather than administrative imperfections.

Relevant grid codes, utility requirements, equipment standards, workplace electrical-safety rules, and applicable IEC or IEEE relay standards must be consulted. Engineers must preserve records of calculations, software versions, test results, as-left settings, and any temporary changes. Cybersecurity is increasingly relevant because remotely accessible relays and coordination databases can be altered or disrupted. Access control, authentication, backups, and audit trails are therefore part of protection governance.

Emerging developments and best practices

Adaptive protection uses real-time topology, measurements, and communications to select settings suited to the current network state. Wide-area measurements can support remedial action and system-wide protection, although communication delay, data quality, and failure modes must be carefully analysed. Digital substations enable sampled-value measurements, process-bus architectures, and automated testing, but they introduce interoperability and cybersecurity considerations.

Data analytics can identify setting drift, repeated nuisance trips, relay self-monitoring alarms, and patterns in fault records. Machine learning may assist event classification or anomaly detection, but protection decisions should remain explainable, validated, and bounded by deterministic security requirements. AI-generated recommendations must not be accepted without engineering review and testing.

Best practice is to maintain an auditable model, study all credible operating states, coordinate from load toward source, check both maximum and minimum fault currents, use manufacturer data, verify settings by independent review, test relay logic and breaker operation, and reassess coordination after every material system change.

Chapter Summary

Protection coordination ensures that the device nearest a fault operates first while upstream devices provide dependable backup. Settings engineering requires the integration of load and fault studies, pickup selection, inverse-time or definite-time characteristics, grading margins, distance reach, CT and breaker performance, and operating topology. Radial systems are commonly coordinated by time-current curves, whereas multidirectional systems require directional, distance, or communications-assisted protection.

Distributed generation and topology changes can reverse current, alter fault levels, and reduce sensitivity, making adaptive settings and setting groups increasingly important. Coordination software improves analysis but cannot replace validation of data, assumptions, device characteristics, and field logic. Safe practice depends on documented engineering judgement, independent review, commissioning, change control, and periodic reassessment.

Key Terms

Selectivity: Operation of the intended primary device while healthy network sections remain energised.

Sensitivity: Ability of protection to detect the minimum fault current within its assigned zone.

Dependability: Probability that a protection system operates when required.

Security: Ability to prevent unwanted operation for external faults or normal conditions.

Pickup current: Current threshold at which an overcurrent element starts or operates.

Plug setting multiplier: Ratio of relay fault current to relay pickup current.

Time multiplier setting: Scaling factor applied to an inverse-time relay characteristic.

Grading margin: Time interval separating primary and backup operation.

Reach: Apparent impedance boundary of a distance-protection element.

Underreach: Failure of a protection element to cover its intended zone boundary.

Overreach: Operation of a protection element beyond its intended zone.

Directional protection: Protection that determines both current magnitude and fault direction.

CT saturation: Distortion of current-transformer output when its magnetic core cannot reproduce primary current accurately.

Inrush current: Transient current, commonly associated with transformer energisation, that may resemble a fault.

Setting group: A coordinated collection of relay settings selected for a defined operating state.

Cold-load pickup: Temporary increase in load current when a supply is restored after an outage.

Remote infeed: Fault-current contribution entering a protected line from the remote terminal.

Adaptive protection: Protection whose settings or logic change in response to network conditions.

Review Questions

Short Answer Questions

1. Define selectivity and explain why it is important in a radial feeder.

2. What is the purpose of a grading margin?

3. Distinguish between relay pickup current and plug setting multiplier.

4. Why are directional elements useful in networks with distributed generation?

5. State two factors that must be included when selecting an earth-fault pickup.

Long Answer Questions

1. Describe a complete workflow for developing and approving overcurrent relay settings.

2. Explain how distance-protection Zones 1, 2, and 3 are coordinated and what causes overreach.

3. Discuss the coordination challenges created by topology changes and embedded generation.

4. Explain how CT errors, breaker time, relay tolerance, and relay overtravel influence grading margins.

5. Describe the role and limitations of coordination software in professional protection studies.

Analytical / Application Questions

1. A feeder has a maximum emergency load of 310 A and a minimum remote fault current of 1.6 kA. Propose the considerations governing pickup selection and explain what additional data are needed before approving the setting.

2. A downstream relay operates in 0.38 s and the required grading margin is 0.28 s. Determine the minimum approximate operating time for its upstream backup and discuss why the calculation must be repeated at other fault levels.

3. An industrial plant changes from a single utility source to a utility-plus-generator configuration. Develop a study plan identifying the operating cases, protection changes, and commissioning checks required to preserve coordination.

22. Multiple Choice Questions with Answers

Question 1

Which condition best represents successful protection coordination in a radial feeder?

A. All upstream and downstream breakers trip simultaneously

B. The upstream incomer always trips first to ensure fault clearance

C. The device nearest the fault trips while upstream devices remain available as backup

D. Only the device with the highest current rating is permitted to trip

Correct Answer: C. The device nearest the fault trips while upstream devices remain available as backup

Question 2

A relay has a 400/1 A CT and a secondary pickup of 0.8 A. What is the corresponding primary pickup current?

A. 200 A

B. 320 A

C. 400 A

D. 500 A

Correct Answer: B. 320 A

Question 3

If a relay sees 4.5 A secondary current and its pickup is 0.8 A, its approximate plug setting multiplier is:

A. 0.18

B. 2.8

C. 4.5

D. 5.625

Correct Answer: D. 5.625

Question 4

Which factor is normally included in a grading margin?

A. Only the nominal load current

B. Breaker clearing time and relay tolerances

C. The transformer turns ratio only

D. The cable length without its impedance

Correct Answer: B. Breaker clearing time and relay tolerances

Question 5

Why can a distributed generator cause a conventional non-directional scheme to lose selectivity?

A. It always eliminates all fault current

B. It can change fault-current magnitude and direction

C. It makes CT ratios irrelevant

D. It prevents all earth faults from occurring

Correct Answer: B. It can change fault-current magnitude and direction

Question 6

For an inverse-time relay, increasing the time multiplier setting generally:

A. Decreases operating time at all fault levels

B. Increases operating time at a given plug setting multiplier

C. Changes the CT ratio automatically

D. Removes the need for a grading margin

Correct Answer: B. Increases operating time at a given plug setting multiplier

Question 7

A distance relay operates beyond the intended protected line boundary. This condition is called:

A. Underreach

B. Security

C. Overreach

D. Reset

Correct Answer: C. Overreach

Question 8

Which study case is especially important for checking overcurrent sensitivity?

A. Maximum fault current only

B. Minimum fault current at the remote end of the protected zone

C. Rated transformer voltage only

D. Normal load with all breakers open

Correct Answer: B. Minimum fault current at the remote end of the protected zone

Question 9

What is a major limitation of coordination software?

A. It cannot plot time-current curves

B. It cannot represent any relay characteristic

C. Its results depend on the accuracy of data, topology, and assumptions

D. It prevents independent engineering review

Correct Answer: C. Its results depend on the accuracy of data, topology, and assumptions

Question 10

In the illustrative industrial case, which combination best addressed reverse fault-current flow and changing topology?

A. Higher non-directional pickups and removal of backup protection

B. Directional elements, setting groups, communications, and tested logic

C. Replacing all relays with fuses without further study

D. Using only the maximum utility fault level

Correct Answer: B. Directional elements, setting groups, communications, and tested logic