Tiza Group

Home / Guides / Protection, control and communications

Protection Coordination and Setting Criteria

What protection coordination is, selectivity methods, setting criteria for 50/51, 21 and 87, and why settings must be reviewed when the network changes.

By Shadai García, CEO and protection engineer at Tiza Group.

ATMT515151ATincomerfeederfuseeach step waits for the downstream onea margin of 0.3 s or morenormal loadfault downstream of the fusethe fuse blows firstthe three 51s resetfault on the feederthe feeder 51 tripsincomer and HV resetrestoration1101000.1110I / Int (s)fusefeederincomerATI faultΔtΔtCascadeFusePTOC feederPTOC incomerPTOC AT© tizagrp.comATMT515151ATincomerfeederfuseeach step waits for the downstream onea margin of 0.3 s or morenormal loadfault downstream of the fusethe fuse blows firstthe three 51s resetfault on the feederthe feeder 51 tripsincomer and HV resetrestoration1101000.1110I / Int (s)fusefeederincomerATI faultΔtΔtCascadeFusePTOC feederPTOC incomerPTOC AT© tizagrp.com
Setting criteria: cascaded coordination of fuse and 51 relays with time margins. Location in the system and operating principle. Illustrative: real values come from the nameplate and the manufacturer's manual.
Short answer

Coordinating protection means setting it so that, on a fault, only the nearest breaker trips and the others act as backup with a controlled delay. Setting criteria balance selectivity, speed and sensitivity: detect the minimum fault in the zone without tripping on maximum load, energization or faults elsewhere, in every credible network topology.

Objective: understand what it means for protections to be coordinated, which criteria are used to set each function and why a setting that is correct today can stop being so when the network changes.

A system is selective when, on a fault, only the breaker closest to it trips and the rest stay in service. Coordination is the setting of currents, curves and times that makes it possible. They are not two types of protection: selectivity is the goal and coordination the means. If the closest protection fails, the next one must operate with a controlled delay: that is backup protection.

Every setting balances three competing requirements:

  • Selectivity: trip only what is necessary.
  • Speed: clear before damage and loss of stability.
  • Sensitivity: detect the minimum fault in the zone, including high-impedance faults.

A good setting detects that minimum fault, does not trip on maximum load, energization (inrush), motor starting or faults elsewhere, and clears before the equipment damage limit.

Selectivity methods

MethodHow it discriminatesTypical use
Current gradingEach protection picks up at a different thresholdInstantaneous 50 with appreciable impedance between relays
Time gradingThe upstream protection waits longerDefinite-time 51
Current and timeInverse curves separated over the whole fault rangeRadial medium-voltage networks
LogicThe downstream relay blocks the upstream one, by wire or by IEC 61850 GOOSEMedium-voltage busbars
DirectionalTrips only in the set directionRings and networks with several sources
DifferentialCompares currents at the zone boundariesTransformers, busbars, lines
DistanceStepped impedance zonesTransmission lines

Differential protections are selective by principle and do not need time coordination, but their backup does. In low voltage the problem is solved with another technology (trip units, manufacturer tables); we explain it in low-voltage selectivity.

The coordination margin

Between two protections in series, the upstream one must wait for the downstream one to operate and its breaker to open. That margin adds the interrupting time of the downstream breaker, the relay timing error, the overshoot in electromechanical relays and a safety margin. As a reference from the technical literature, between numerical relays it is typically on the order of 0.2 to 0.3 s, and 0.3 to 0.4 s with electromechanical ones. With a fuse downstream it is measured on its total clearing curve. The final value depends on the breaker and on each utility’s policy. This is complemented by ABB SPAJ, SPAM and REF54x legacy relays.

Curves must not cross within the range of possible faults. When the chain of steps leaves excessive clearing time at the source, it is better to change method (logic selectivity, differential) than to add delays.

Data a setting needs

  • Updated single-line diagram with impedances and network equivalents.
  • Ratios and class of the current transformers (CTs) and voltage transformers (VTs).
  • Maximum and minimum load flow and short circuit (three-phase, phase-to-phase, single-phase and with fault resistance), in several topologies and dispatches.
  • Damage curves: transformer through-fault withstand (IEEE C57.109) and cable thermal capacity.
  • Settings of neighboring protections.

Gathering that data, simulating the cases and delivering the time-current plots with the settings table is the protection coordination study, a service in itself. Here we stay with the criteria.

Criteria by function

Time-delayed phase overcurrent (51). The pickup goes above maximum load with a margin and below the minimum fault at the end of its backup zone. The standard curve (IEC 60255-151 or IEEE C37.112: moderately, very or extremely inverse) is chosen to fit the neighbors and stay under the damage curve. See phase and ground overcurrent.

Ground (51N). Above the normal network unbalance and low enough to see faults with resistance. It depends a lot on how the system is grounded; see zero-sequence current.

Instantaneous (50). Above the maximum fault at the remote bus or on the low side of the transformer, with margin, and above the inrush. If the close-in fault and the remote fault are similar, the 50 adds no selectivity.

Distance (21). Zone 1 covers most of the line without delay (in common practice, 80 to 85 %) so as not to reach the remote bus. Zone 2 covers the whole line with margin, without exceeding zone 1 of the shortest adjacent line, and with a short delay. Zone 3 gives remote backup with more delay and must not encroach on the maximum load impedance. For ground faults, the zero-sequence compensation factor and the mutual coupling of parallel lines come in (IEEE C37.113).

Differential (87). In the 87T, the minimum pickup and the slopes must cover the CT error and the tap changer range, with second-harmonic restraint for inrush and fifth-harmonic for overexcitation. In the 87B, high impedance requires CTs of equal ratio and characteristics; low impedance works with dynamic zones according to the disconnectors. In the 87L, the line charging current and the channel delay count. Values depend on the manufacturer.

Setting for all topologies

Topology is how the network elements are connected at a given moment. Settings are calculated for certain topologies; when the network changes, so do the currents and impedances each relay sees.

Which changes matter: switching (lines out of service, bus coupler open or closed, transformers in parallel), N-1 contingencies and maintenance outages, expansions (new lines, substations, transformers), maximum and minimum generation dispatch, connection of solar or wind plants, and a new grounded neutral that modifies the zero-sequence network.

What effect they have:

  • Minimum sensitivity. With fewer sources the minimum fault drops and the protection may stop seeing it. With more sources the maximum fault rises and an instantaneous element may reach the next zone.
  • Coordination. If the current distribution changes, the margin between relays can vanish and the wrong one trips.
  • Change of direction. If the network goes from radial to meshed or generation appears downstream, the fault is fed from both sides and directional 67/67N functions are needed.
  • Intermediate infeed. A source between the distance relay and the fault makes the relay see the fault farther away and lose reach; an intermediate outfeed makes it see it closer.
  • Distributed generation. It can reduce the current seen by the head-end relay, cause sympathetic trips on healthy feeders and break recloser-fuse coordination. Inverters, in addition, limit their fault current by control. See protection requirements for distributed generation. This is complemented by recloser–fuse and recloser–relay coordination.

How it is solved: all credible topologies are studied and robust settings sought. When there are none, relay settings groups activated by breaker status or by logic are used, or functions less dependent on topology, such as line differential or teleprotection. So-called adaptive protection is applied in practice with settings groups and logic calculated and tested beforehand, not with relays that “learn” without a study.

Common mistakes

  • Setting with a single short-circuit case: it ends up selective but blind to high-impedance faults.
  • Coordinating only phases and leaving the ground protection without a study.
  • Using the same margin for numerical relays, electromechanical relays and fuses.
  • Forgetting inrush in the 50 and 87 functions.
  • Setting zone 3 without checking maximum load under contingency.
  • Not restudying after expansions or new generation, nor testing the logic that changes the settings groups.
  • Loading settings without an injection test that confirms the real curve.

How we help

At Tiza Group we carry out protection coordination studies and load and verify the settings, remotely with our remote relay configuration (worldwide) or on site in Colombia, or elsewhere on request, for medium-, high- and extra-high-voltage substations and industry. Learn about the protection coordination study or contact us.

References

  • IEC 60255-151 and IEEE C37.112, characteristics of inverse-time overcurrent protection.
  • IEEE C37.113-2025, guide for protective relay applications to transmission lines.
  • IEEE C57.109, guide for liquid-immersed transformer through-fault-current duration.
  • J. L. Blackburn and T. J. Domin, Protective Relaying: Principles and Applications.

Frequently asked questions

What is the difference between selectivity and coordination?

Selectivity is the goal: only the breaker closest to the fault trips. Coordination is the means: the setting of pickups, curves and times that makes it possible.

Why is distance zone 1 not set to 100 % of the line?

To avoid reaching the remote bus because of CT, VT and impedance data errors. In common practice it covers 80 to 85 % of the line, and the final margin is set by each utility or grid operator.

How often should protection settings be reviewed?

Every time the network changes: new lines or transformers, new generation, grounding changes or permanent switching. Also after an unwanted trip or a poorly cleared fault.

WhatsApp