---
title: "Selectivity Calculator: Circuit Breaker Discrimination to AS/NZS 3000:2018 Cl 2.5.7.1"
url: "https://elecas.com.au/calculator/selectivity"
description: "Free circuit breaker selectivity calculator: check discrimination for every pair in a protection path, read the selectivity limit current Is, and see what to change."
updated: "2026-09-08"
standards:
  - "AS/NZS 3000:2018 (Wiring Rules, Clause 2.5.7 discrimination and coordination)"
  - "IEC 60898-1 (AS/NZS 60898.1) MCB tripping characteristics, curves B, C and D"
  - "IEC 60947-2 MCCB and ACB tripping characteristics, LSI trip units"
  - "IEC 60269-1 gG HRC fuse-link time-current characteristics"
author: "Wisam Tozah (B.Eng (Electrical), MIEAust, CPEng, NER, NSW DBP, NSW PRE, APEC, IntPE(Aus))"
site: "ElecAS"
license: "https://elecas.com.au/terms"
usage: "Cite and link. Do not reproduce or train on this content."
---

# Selectivity Calculator: Circuit Breaker Discrimination to AS/NZS 3000:2018 Cl 2.5.7.1

Source: https://elecas.com.au/calculator/selectivity

A free, browser-based selectivity (discrimination) calculator for Australian and New Zealand low-voltage installations. Describe the protective devices in a protection path - picked from the Schneider and NHP catalogue with their real trip-unit settings, or typed straight off any manufacturer's data sheet - and every neighbouring pair is checked against AS/NZS 3000:2018 Clause 2.5.7: total selectivity, partial selectivity with the selectivity limit current Is, or none at all. It handles MCBs to IEC 60898-1 (curves B, C and D), thermal-magnetic MCCBs, electronic LSI trip units on MCCBs and ACBs, and gG HRC fuse-links to IEC 60269-1, so a mixed path of fuse, MCCB and MCB grades in one pass. The path can be as long as the installation is - main switchboard, sub-main, distribution board, final subcircuit - every adjacent pair carries its own verdict, and the governing pair names the single weakest link the whole path stands on. Where a pair falls short the design notes name the setting to change (an upstream instantaneous stage to switch off or raise, a short-time delay to lengthen) instead of only reporting the failure, and the shortfall is quantified so a 5 ms miss is not read as the same problem as a two-second one. Enter the prospective fault current at the board and the result is annotated with whether the pair is selective at that site in practice. Everything is plotted on one time-current coordination chart with Is marked across the trip bands, and exports to a branded PDF coordination report.

## Key facts

- Selectivity holds while the downstream device's slowest clearing time stays below the upstream device's fastest tripping time: worst case against worst case, tolerance band against tolerance band.
- The selectivity limit current Is is the lowest current at which that stops being true. Below Is the pair discriminates; at or above it both devices can trip. A partial result is only meaningful once Is is compared against the prospective fault current at that board.
- A path is checked pair by pair and the headline is its weakest link: the governing pair is the one that gives up at the lowest current, and that is where a fix starts. Grading the whole path is not the same as grading one pair at a time.
- AS/NZS 3000:2018 Clause 2.5.7.1 requires the installation to be designed so the loss of supply resulting from a fault is minimised, and requires the selection and settings to be verified by inspection. Clause 2.5.7.2.1 pairs discrimination with backup (cascading) protection.
- Two MCBs in series rarely discriminate above the overload region, because neither has an intentional delay: grading in the short-circuit region needs a device with an adjustable short-time delay (tsd) upstream.
- The single most common cause of a lost grade is the upstream device's instantaneous stage (Ii). Switching it OFF, or raising it above the downstream device's magnetic band, is usually what turns partial selectivity into total.
- Manufacturer-verified discrimination tables can certify a pair beyond the point the published curves cross, because a device's real let-through energy is lower than its published envelope. Curve comparison is the design tool; the tested table is the confirmation.
- Selectivity and breaking capacity are separate questions. A pair can discriminate perfectly and still be non-compliant if the downstream device's Icu is below the prospective fault current, which is what backup (cascading) protection exists to solve.

## Who this page is for

Electrical engineers, electrical designers, switchboard builders and contractors grading protective devices and running discrimination studies on Australian and New Zealand low-voltage installations to AS/NZS 3000:2018.

## Relevant standards

- AS/NZS 3000:2018 (Wiring Rules, Clause 2.5.7 discrimination and coordination)
- IEC 60898-1 (AS/NZS 60898.1) MCB tripping characteristics, curves B, C and D
- IEC 60947-2 MCCB and ACB tripping characteristics, LSI trip units
- IEC 60269-1 gG HRC fuse-link time-current characteristics

## What this tool does

- Total, partial or no selectivity for every adjacent pair in the protection path, with the selectivity limit current Is quoted for every partial result.
- Devices from the Schneider and NHP catalogue with their real trip-unit settings, or any make typed in from its data sheet.
- MCBs (IEC 60898-1 curves B, C and D), thermal-magnetic MCCBs, electronic LSI MCCB and ACB trip units and gG HRC fuse-links, so a mixed fuse / MCCB / MCB path grades in one pass.
- A protection path of any length - main, sub-main, distribution board, final subcircuit - with the governing pair named as the weakest link.
- Design notes that name the setting to change - an upstream instantaneous stage or a short-time delay - rather than just reporting the failure.
- How far a crossing pair misses by, so a 5 ms shortfall is not read as the same problem as a two-second one.
- Enter the prospective fault current at the board and each verdict is annotated with whether the pair is selective at that site in practice.
- One time-current coordination plot with every device band drawn and Is marked, plus a branded PDF report with every device setting and the pair-by-pair result.

## How to check selectivity between two circuit breakers

1. **Describe the upstream device** — Pick it from the Schneider or NHP catalogue and dial in its trip-unit settings, or switch to Generic and enter the rating, curve class or pickup multiples, delays and breaking capacity from the data sheet.
2. **Describe the downstream device** — Same again for the device closer to the load. The list is ordered supply side first, so position in the list is the electrical hierarchy; use the arrows to reorder if you entered them the other way round.
3. **Add the rest of the path** — Keep adding devices at the load end for as many levels as the installation has - main, sub-main, distribution board, final subcircuit. The copy button clones a device you have already set up, and the arrows reorder the path if you entered it the other way round.
4. **Read the verdict and the limit** — The chip above the plot gives total, partial or no selectivity with the selectivity limit current Is, and the plot marks Is across the trip bands. The panel under the plot says what that means for the installation and names the governing pair; on the path itself every pair carries its own verdict.
5. **Act on the advice** — Where the pair falls short, the panel under the plot names the change that would move it - usually an upstream instantaneous stage that should be switched off or raised, or a short-time delay that should be longer - before you go looking for a different device.

## Circuit breaker selectivity: how discrimination is read off the trip curves

### What selectivity is, and what it costs when it is missing

Selectivity - used interchangeably with discrimination in Australian practice - is the arrangement of protective devices in series so that a fault is cleared by the device closest to it, and only that device. A fault on one final subcircuit should trip that subcircuit's breaker and leave the distribution board, the sub-main and the main switch closed. When selectivity is absent, the same fault opens two or three devices at once and takes out everything they supply.

The consequence is rarely a safety failure, because the fault is still cleared. It is an availability failure, and it is expensive in exactly the buildings where it is least acceptable: a faulty appliance on a tenancy circuit blacking out a whole floor, a hospital department losing supply because a lighting circuit faulted, a data hall dropping because a socket outlet did. This is why AS/NZS 3000:2018 Clause 2.5.7.1 frames the requirement as minimising the loss of supply resulting from a fault, rather than as a fixed numerical rule.

### How a verdict is reached from two tolerance bands

A protective device is not published as a line. It is published as a band, because manufacturing tolerance means two units of the same catalogue number will not trip at the same instant. Every device on the plot is drawn as an envelope: the fastest it might trip on one edge, the slowest it might clear on the other.

Selectivity between two devices holds while the downstream device's slowest total clearing time stays below the upstream device's fastest tripping time, checked at every current. Worst case against worst case. If the two envelopes never touch anywhere up to the downstream device's breaking capacity, the pair has total selectivity. If they meet at some current, the pair has partial selectivity and that current is the selectivity limit current Is. If they overlap from the start, there is no selectivity at all.

That definition is deliberately conservative, and it is worth knowing why. Comparing published envelopes ignores the fact that a real breaker's let-through energy on a high fault is lower than its published envelope implies, and that the downstream device is often current-limiting. A manufacturer's tested discrimination table for a specific pair can therefore certify selectivity past the current at which the drawn curves cross. Curve comparison is the right tool for designing and for finding the problem; the tested table for that exact pair is the right reference before a design is changed.

### Why Is only means something next to a fault level

A partial result is not a failure on its own. It says the pair discriminates below Is and may not above it. Whether that matters depends entirely on whether the installation can actually deliver a fault current that high at that point.

A pair with Is of 4 kA on a distribution board whose prospective fault current is 2.6 kA is selective in practice: nothing above Is can happen there. The same pair on a board fed directly from a substation with 25 kA available is not. This is why the calculator takes the prospective fault current at the board as an input and annotates each verdict with whether the pair is selective at that site. The verdict itself never moves - the check still runs against the full envelope - because the fault level is a property of the installation as built, and a supply upgrade or a shorter run can raise it later.

### Why two MCBs in series almost never grade

Above its magnetic pickup, an MCB to IEC 60898-1 opens in a few milliseconds. It has no intentional delay, and neither does the MCB upstream of it. Once a fault current is high enough to sit inside both devices' magnetic bands, the two are racing, and which one opens first is not something published curves can settle.

Selectivity between two MCBs is therefore generally limited to the overload region, where the thermal elements are separated by rating. Above that it depends entirely on the manufacturer's tested pairs, and many are only certified with a substantial ratio between ratings. Grading properly in the short-circuit region needs a device with an adjustable short-time delay upstream: a thermal-magnetic MCCB with a delayed magnetic element, or an electronic LSI trip unit on an MCCB or ACB.

The corollary matters for switchboard design: if a board needs genuine discrimination down to the final subcircuit, that requirement has to be in the specification before the board is built, because it constrains what the incomer can be.

### The settings that decide the outcome

On an electronic LSI trip unit three settings do almost all the work. The long-time pickup Ir and delay tr set the overload region and are usually decided by the cable, not by grading. The short-time pickup Isd and delay tsd are the grading controls: raising tsd on the upstream device holds it closed long enough for the downstream device to clear, and it is what buys selectivity between two MCCBs. The instantaneous stage Ii is the one that breaks grades, because above its pickup the upstream device has no intentional delay left and the downstream device has nothing to be faster than.

Switching the instantaneous stage OFF, or raising it above the downstream device's magnetic band, is the single most common fix and is usually what turns a partial verdict into a total one. It is not free: the upstream device then holds a fault on its own busbar for the short-time delay instead of clearing it immediately, which raises the incident energy at that board and should be checked against the arc flash study. Many trip units also offer an I²t short-time characteristic, which slopes the short-time region and can slide a marginal pair apart without lengthening the delay.

### Discrimination and backup protection are two different requirements

AS/NZS 3000:2018 Clause 2.5.7.2.1 treats coordination as two things at once, and they are easy to confuse. Discrimination is about which device opens. Backup (also called cascading) protection is about whether the downstream device survives at all: it lets a device whose breaking capacity Icu is below the prospective fault current be used, because an upstream current-limiting device cuts the let-through energy to something the downstream device can handle.

A pair can discriminate perfectly and still be non-compliant on breaking capacity, and a backed-up pair is by definition not selective at the currents where the backup acts, because both devices operate. Backup combinations come only from the manufacturer's tested tables - they cannot be read off published curves at all - so a design that relies on cascading needs the table for that exact combination on file.

### What this page produces, and what it does not

The calculator takes a protection path of any length and grades every adjacent pair in it: main switchboard to sub-main, sub-main to distribution board, distribution board to final subcircuit, as many levels as the installation has. Each pair carries its own verdict and its own Is, the whole path is summarised by its governing pair - the one that gives up at the lowest current - and the shortfall on a crossing pair is quantified, so a 5 ms miss is not presented as the same problem as a two-second one. Everything is drawn on one time-current coordination plot with Is marked across the bands, and exports to a branded PDF report carrying every device setting and the pair-by-pair result.

What it does not do is replace the manufacturer's tested discrimination table, the fault level study that supplies the prospective fault current, or the engineer of record. It compares published characteristics, which is the conservative side to be on: a pair this page passes will pass a tested table, and a pair it reports a limit on may still be certified by one. It is a design and review tool, not a certification.

## Key terms

- **Discrimination (selectivity)** — Coordination between two protective devices in series such that a fault downstream is cleared by the downstream device alone, leaving the upstream device closed and the rest of the installation supplied.
- **Selectivity limit current (Is)** — The lowest fault current at which the upstream device can begin to trip before the downstream device has finished clearing. Below Is the pair discriminates; at or above it, both may open.
- **Time-current curve (TCC)** — The log-log plot of tripping time against current that every protective device is published on. Devices are published as bands rather than lines, because manufacturing tolerance means two units of the same catalogue number will not trip at the same instant.
- **Governing pair** — The adjacent pair in a protection path that gives up at the lowest current, or that has no discriminating region at all. It sets the headline verdict for the whole path, and it is where a fix starts.
- **LSI trip unit** — An electronic trip unit with three independently adjustable stages: Long-time (Ir, tr) for overload, Short-time (Isd, tsd) for delayed short circuit, and Instantaneous (Ii) for undelayed short circuit. The short-time delay is what makes grading between MCCBs possible.
- **Breaking capacity (Icu / Ics)** — The fault current a device can interrupt: Icu the ultimate rated capacity, Ics the service capacity it can interrupt and remain in service after. It is a separate compliance question from selectivity, and it is what backup protection addresses.
- **I²t short-time characteristic** — An optional sloped short-time region on an electronic trip unit, where clearing time falls as current rises rather than staying flat. It can separate a marginal pair without lengthening the short-time delay.
- **gG fuse-link** — A general-purpose HRC fuse to IEC 60269-1, protecting against both overload and short circuit. Fuses grade well against each other and against breakers because their characteristics are steep and their let-through energy is low.
- **Total selectivity** — No crossing anywhere up to the downstream device's breaking capacity: the downstream device clears alone at every fault current it is rated to interrupt.
- **Partial selectivity** — Selectivity up to Is and not beyond. Whether that is acceptable is a design decision to record, and one a manufacturer's tested table may settle in your favour.
- **Backup (cascading) protection** — The other half of coordination: an upstream device limiting the let-through energy so a downstream device can be used on a board whose fault level exceeds its own breaking capacity. It comes from the manufacturer's tested combinations, not from the curves.
- **Prospective fault current** — The current a bolted fault at that point in the installation would draw, set by the supply impedance and the cable run. It is what decides whether a partial selectivity limit matters in practice.
- **Instantaneous stage (Ii)** — The undelayed element of a trip unit. Above its pickup the upstream device has no intentional delay left, which is the single most common reason an otherwise well-graded pair loses selectivity.
- **Short-time delay (tsd)** — An intentional delay on the upstream device's short-circuit element, held long enough for the downstream device to clear first. It is what buys selectivity between two MCCBs.

## Frequently asked questions

### Is partial selectivity a problem?

Not necessarily. Partial selectivity means the two devices can both trip above the selectivity limit current Is; below Is the downstream device clears alone. Whether that is acceptable is a design decision: compare Is against the prospective fault current your fault study gives for that board, and where the installation cannot reach Is the pair is selective in practice. Where it can, either record the decision, change a setting or a device, or check the manufacturer's tested discrimination table for the pair - a tested pair can be certified beyond the point the published curves cross.

### Does AS/NZS 3000 require full discrimination?

Clause 2.5.7.1 requires the installation to be divided into circuits and its protective devices selected with appropriate discrimination so that the loss of supply resulting from a fault is minimised, and it requires the selection and settings to be verified by inspection. Clause 2.5.7.2.1 adds that coordination means both discrimination and backup (cascading) protection. Safety service circuits are dealt with separately in Clause 2.5.7.2.2 and general supply circuits in Clause 2.5.7.2.3, so what counts as good enough depends on which the circuit is.

### Why do two MCBs in series rarely discriminate?

Because neither has an intentional delay. Above the magnetic pickup an MCB opens in a few milliseconds, so once a fault current is high enough to be inside both devices' magnetic bands the two race, and which one opens is not something the curves can settle. Selectivity between MCBs is generally limited to the overload region, and above that it depends on the manufacturer's tested pairs. Grading properly needs a device with a short-time delay upstream.

### Why does my upstream breaker's instantaneous setting matter so much?

The instantaneous stage removes the upstream device's delay entirely above its pickup, which leaves the downstream device nothing to be faster than. On an LSI electronic trip unit switching the instantaneous OFF - or raising it above the downstream device's magnetic band - is usually what turns partial selectivity into total, at the cost of a longer let-through on a fault on the upstream device's own busbar.

### Can I trust this instead of the manufacturer's discrimination table?

Use it to design, and the table to confirm. This page compares published tolerance envelopes, which is deliberately conservative: a real breaker's let-through energy is lower than its published envelope, so a manufacturer's tested table can certify a pair as selective past the current at which the drawn curves cross. Where this page reports a selectivity limit, the tested table for that exact pair is the reference before a design changes.

### What is selectivity in an electrical installation?

Selectivity, also called discrimination, is the arrangement of protective devices in series so that a fault is cleared by the device closest to it and only that device. A fault on a final subcircuit should open that subcircuit's breaker and leave the distribution board, the sub-main and the main switch closed. Without it, one faulty appliance can black out a whole floor or building. It is achieved by separating the devices' time-current characteristics so that at every fault current the downstream device finishes clearing before the upstream device begins to trip.

### What is the difference between selectivity and discrimination?

Nothing. They are two names for the same thing. "Discrimination" is the traditional British and Australian term and is the word AS/NZS 3000:2018 uses in Clause 2.5.7; "selectivity" is the IEC term used in IEC 60947-2 and by most manufacturers' international literature. Australian drawings and specifications use both, often in the same document, and a "discrimination study" and a "selectivity study" are the same deliverable.

### How do you calculate selectivity between two circuit breakers?

Plot both devices on the same time-current axes as bands rather than lines, because published characteristics carry manufacturing tolerance. Then compare, at every current up to the downstream device's breaking capacity, the downstream device's slowest total clearing time against the upstream device's fastest tripping time. If the downstream band stays entirely below and to the left of the upstream band, the pair has total selectivity. If they meet, the current at which they meet is the selectivity limit current Is and the pair is partially selective. If they overlap from the start, there is no selectivity. This calculator does that comparison from the devices' published data and returns the verdict and Is directly.

### Is the 2:1 rating ratio rule reliable for discrimination?

It is a rule of thumb for the overload region only, and it does not settle the short-circuit region at all. A 2:1 or 1.6:1 ratio between rated currents separates the thermal elements, so it usually holds for slow overloads. It says nothing about what happens once a fault current reaches both devices' magnetic or instantaneous bands, which is where most grades are actually lost: two MCBs at a 3:1 ratio can still both trip on a short circuit. Check the curves, then check the manufacturer's tested table.

### Can a fuse and a circuit breaker discriminate?

Yes, and often better than two breakers. A gG HRC fuse-link to IEC 60269-1 has a steep characteristic and very low let-through energy, so a fuse upstream of a breaker grades well provided the ratings are separated, and a fuse downstream of a breaker clears fast enough that the breaker never starts. This calculator handles gG fuse-links alongside MCBs, thermal-magnetic MCCBs and electronic LSI trip units, so a mixed fuse, MCCB and MCB path is graded in one pass. For fuse-to-fuse pairs the manufacturers publish I²t let-through and pre-arcing data that will settle marginal cases more precisely than a curve comparison.

### What is the difference between selectivity and cascading (backup) protection?

They answer different questions and AS/NZS 3000:2018 Clause 2.5.7.2.1 requires both. Selectivity is about which device opens: only the one nearest the fault should. Backup or cascading protection is about whether the downstream device survives: it allows a device whose breaking capacity Icu is below the prospective fault current to be used, because an upstream current-limiting device reduces the let-through energy to something it can handle. They pull in opposite directions - a backed-up pair is by definition not selective at the currents where the backup acts, because both devices operate - and backup combinations come only from the manufacturer's tested tables, never from the published curves.

### Does selectivity apply to RCDs as well as circuit breakers?

It does, but by a different mechanism, and it is outside what this calculator checks. RCD selectivity is achieved with a time-delayed S-type (selective) RCD upstream of instantaneous general-type RCDs, with the upstream device also having a higher residual current rating - typically 300 mA upstream of 30 mA. That is a matter of matching device types and rated residual currents, not of comparing overcurrent time-current curves. This page grades overcurrent protection: MCBs, MCCBs, ACBs and fuses.

### Do I need a discrimination study, and is this calculator free?

A discrimination study is normally required wherever continuity of supply is specified: hospitals and health facilities, data centres, essential and safety services, and most commercial and institutional projects where a specification calls for coordination. AS/NZS 3000:2018 Clause 2.5.7.1 also requires the selection and settings of protective devices to be verified by inspection on any installation. This calculator is free to use in the browser with no installation and no sign-up to calculate; downloading the branded PDF coordination report needs only a free account.

### What do I do when the calculator reports no selectivity?

Work through it in this order. First check the upstream device's instantaneous stage: switching it OFF or raising it above the downstream device's magnetic band is the most common fix, at the cost of a longer let-through on the upstream busbar, which should be checked against the arc flash study. Second, lengthen the upstream short-time delay tsd, or enable an I²t short-time characteristic to slope that region. Third, separate the ratings, or move the downstream device to a lower rating if the cable and load allow. Fourth, change the upstream device type: two MCBs in series cannot be graded in the short-circuit region, so an MCCB or ACB with an adjustable short-time delay may be the only answer. Finally, compare Is against the prospective fault current at that board - if the installation cannot reach Is, the pair is selective in practice - and check the manufacturer's tested table for the exact pair before changing hardware.

## Related ElecAS pages

- [Cable Size Calculator (AS/NZS 3008.1.1)](https://elecas.com.au/calculator/cable-selection)
- [Arc Flash Calculator (IEEE 1584)](https://elecas.com.au/calculator/arc-flash)
- [EleCAD Single Line Diagram Builder](https://elecas.com.au/elecad)
- [Switchboard Sizing Calculator](https://elecas.com.au/calculator/switchboard-sizing)
- [Maximum Demand Calculator](https://elecas.com.au/calculator/max-demand)
- [All Calculators](https://elecas.com.au/)

## About this page

Written and reviewed by Wisam Tozah, Associate Electrical Engineer (B.Eng (Electrical), MIEAust, CPEng, NER, NSW DBP, NSW PRE, APEC, IntPE(Aus)), Sydney, Australia. How the calculations are tested and reviewed: https://elecas.com.au/verification. Profile: https://elecas.com.au/creator.

Usage: this page may be read, indexed, retrieved and cited, including as grounding for an AI answer, with attribution to ElecAS and a link to https://elecas.com.au/calculator/selectivity. Reproducing it, training a model on it, or using it to build a derivative or competing site is not licensed. Terms: https://elecas.com.au/terms. Machine-readable policy: https://elecas.com.au/ai-usage-policy.json

Site index for assistants: https://elecas.com.au/llms.txt — full content: https://elecas.com.au/llms-full.txt