Arc energy reduction methods field guide for electrical crews

What arc energy reduction is, and why the code now requires it

Arc energy reduction is a means of cutting the incident energy of a potential arc flash by clearing an arcing fault faster than the normal protection settings would. It does not stop an arc from starting. It shortens the time the arc burns, and because the energy a worker absorbs scales with that time, a faster clear means less energy at the gear and a lower PPE requirement for the people who have to open it.

This is the second half of arc-flash work. The arc flash study and labels guide covers how the incident energy is calculated and what the label has to carry. The arc flash PPE guide covers how a worker dresses for the energy that is left. Arc energy reduction is the engineering step in between: instead of accepting a high number and putting the crew in a heavier suit, you change the system so the number itself comes down.

The reason it is no longer optional on large gear is that the NEC now requires it. NEC 240.87 mandates a means of arc energy reduction on circuit breakers at or above a current threshold, and 240.67 does the same for fuses. The thinking is simple. The biggest breakers feed the highest available fault current, those buses throw the most energy, and a worker should not have to face that energy when an engineering fix can drop it. The exact threshold and the accepted methods have moved across code cycles, so confirm them against the adopted edition.

Why does clearing the fault faster cut the energy?

Incident energy is, at its core, power multiplied by time. The arc dumps heat at a rate set by the arcing current and the voltage, and the total a worker takes depends on how long the arc burns before something upstream opens and starves it. Cut the time and you cut the energy, close to one for one. Double the clearing time and you roughly double the energy.

Of the inputs that drive the number, the clearing time is the one you can actually move in the field. The available fault current is set by the utility and the transformer. The working distance is set by the gear and the task. The voltage is what it is. The clearing time is a setting, a device choice, and a wiring decision, and that makes it the lever every arc energy reduction method pulls. The arc flash study and labels guide walks through the rest of the inputs and how IEEE 1584 turns them into a cal/cm2 number.

So every method in this guide is doing the same thing by a different route. Zone-selective interlocking, a maintenance switch, an optical relay, a current-limiting fuse: each one finds a way to open the circuit sooner during a fault. The mechanism differs. The physics does not. Less time on the arc, less energy on the worker.

What does NEC 240.87 require?

NEC 240.87 requires a means of arc energy reduction where a circuit breaker is rated or can be adjusted to a continuous current trip setting at or above a threshold, commonly 1200 A in recent editions. It also requires that documentation be available to the people who design, install, operate, or inspect the gear, showing the method used and where it is applied. The article has been expanded across code cycles, so verify the threshold and the wording against the edition the jurisdiction has adopted.

The article does not let you pick a number off a chart and call it done. It hands you a list of accepted methods and tells you to provide one of them, set to operate at less than the available arcing current so it actually clears the fault during the dangerous window. That last condition is the part that gets skipped. A method that exists in the gear but is set above the arcing current never trips on the arc, and the energy reduction is on paper only.

Read 240.87 as a floor, not a ceiling. It forces arc energy reduction onto the largest breakers, where the hazard is worst. It does not stop you from applying the same methods to smaller gear where a study shows a high incident energy. Plenty of dangerous buses sit below the threshold, and the code requirement is the minimum, not the whole risk picture.

The 1200 A threshold, and what counts

The trigger in 240.87 is commonly stated as a circuit breaker rated or able to be adjusted to 1200 A or higher. The phrase rated or able to be adjusted is the part crews miss. A breaker does not have to be set at 1200 A to fall under the rule. If the frame and trip unit can be turned up to 1200 A, the requirement applies even when the breaker is dialed lower today, because the next person can change the setting.

That catches a lot of gear that looks like it should be exempt. An 800 A setting on a breaker whose trip unit goes to 1600 A is still in scope. The question is the capability of the device, not the dial position on the day of the inspection. This is exactly the kind of detail an inspector checks, and it is a common finding on a service that was sized with headroom.

Confirm the current threshold against the adopted NEC edition before you cite a number on a submittal. The figure has held at 1200 A through recent cycles, but the scope language and the exceptions around it have been revised, and the jurisdiction may be on an older or amended edition. The principle is stable even when the number is not: the largest devices get arc energy reduction.

Fuses and NEC 240.67

NEC 240.67 is the fuse counterpart to 240.87, and it works the same way at the same kind of threshold, commonly fuses rated 1200 A or higher. The structure is slightly different because a fuse clears on its own curve. The article gives an out: if the fuse clears the available arcing current fast enough on its time-current curve, a value commonly cited around 0.07 seconds, no further means is required, because the fuse is already fast. Confirm that clearing-time figure and the threshold against the adopted edition.

Where the fuse does not clear that fast, 240.67 requires a method, and the list overlaps with 240.87 but is shaped for fuses. It commonly includes differential relaying, energy-reducing maintenance switching, an energy-reducing active arc-flash mitigation system, current-limiting electronically actuated fuses, or an approved equivalent. The same documentation requirement applies: the people who work the gear have to be able to find what method is in place.

The practical line for a fuse job is to start with the curve. Pull the fuse manufacturer's time-current data, find the available arcing current at the bus from the study, and see where the fuse clears it. A current-limiting fuse on a high-fault bus often already clears in the first half-cycle and satisfies the requirement on its own. A slower fuse, or one on a bus where the arcing current lands in the slow part of the curve, needs one of the listed methods added.

The methods NEC 240.87 accepts

NEC 240.87 lists the accepted means, and the list has grown across editions. In recent editions it commonly includes zone-selective interlocking, differential relaying, an energy-reducing maintenance switch with a local status indicator, an energy-reducing active arc-flash mitigation system, an instantaneous trip setting, an instantaneous override, and an approved equivalent means. Each one shortens the clearing time during a fault, and each one trades cost, complexity, and coordination differently.

The table below is the field summary of how each method gets the energy down and where it fits. The sections that follow take each one in turn. Whatever you pick, the controlling condition from the article still applies: it has to operate at less than the available arcing current, or it does nothing on the arc it is supposed to clear.

Confirm the current list against the adopted NEC edition. Methods have been added as the technology matured, and an approved equivalent always requires the authority having jurisdiction to sign off on it, so it is not a blanket permission to invent a method.

What is zone-selective interlocking?

Zone-selective interlocking, ZSI, is a scheme where breakers talk to each other so an upstream breaker can trip fast for a fault in its own zone while still holding a delay for a fault further downstream. The downstream breakers send a restraining signal up. If the upstream device sees a fault and no downstream breaker is signaling that the fault is below it, the upstream breaker clears with no intentional delay. If a downstream breaker is signaling, the upstream device waits its coordinated time and lets the downstream breaker clear its own fault first.

The win is that it gives you fast clearing and selective coordination at the same time, which are usually at odds. Normal coordination buys selectivity by delaying upstream devices, and that delay is exactly what drives the incident energy up at the main. ZSI breaks the trade: the upstream breaker only takes the delay when the delay is actually needed to coordinate, and it clears fast for the worst case, a fault right on its own bus. That worst case is where the energy is highest, so this is the fault you most want cleared quickly.

ZSI lives in the trip units of modern breakers, so on new switchgear it is often a setting and a small interlock wiring run rather than added hardware. The catch is the wiring. ZSI depends on the signal between breakers, and if it is mis-wired or never commissioned, the interlock does not work and the breaker quietly reverts to its coordinated delay. A ZSI scheme that was never tested is a ZSI scheme you cannot count on.

Differential relaying

Differential relaying compares the current going into a protected zone with the current coming out of it. In normal operation, what goes in comes out and the difference is near zero. When a fault opens up inside the zone, current flows to the fault instead of out the far side, the in and the out no longer match, and the relay trips fast on that difference. Because it only responds to faults inside its own zone, it can be set to clear quickly without waiting on anything downstream.

That makes it both fast and inherently selective, which is why it shows up on switchgear, large feeders, and bus protection where the available fault current is high. It does not borrow time from a coordination scheme the way a delayed overcurrent device does. It defines a zone with current transformers at each boundary and protects what is inside.

The cost is the hardware. Differential protection means current transformers at every zone boundary, a relay, and the wiring and commissioning to go with it. On a major bus or a critical feeder that cost is justified by the speed and the selectivity. On a small breaker it is overkill, and a maintenance switch or ZSI does the job for far less.

What is an arc flash maintenance switch?

An energy-reducing maintenance switch, often called ERMS or maintenance mode, is a switch a worker flips before energized work that puts the breaker's electronic trip unit into a faster-tripping state for the duration of the task. In maintenance mode the trip unit drops its instantaneous pickup to a low preset value, so any fault during the work clears with no intentional delay and the incident energy at the gear falls, often enough to drop the PPE category. When the work is done, the worker flips it back to normal and coordination returns.

It is the cheap, popular retrofit, and for good reason. Most modern electronic trip units already have the function built in or can take a module, so adding ERMS to existing gear is far cheaper than differential relaying or new switchgear. On a typical electronic trip unit the maintenance setting drops the instantaneous pickup to a low multiple of the breaker's current rating, which clears an arcing fault in the breaker's own fast region. The exact behavior is set in the trip unit per the manufacturer, so confirm the pickup and the response against the device documentation.

The code calls for a local status indicator, and the indicator is not decoration. It is the thing that tells a worker the mode is actually engaged and tells the next person it is still on. Many trip units light a clear indicator, often a bright color on the switch and the display, so the state is obvious from the front of the gear. A maintenance switch with a dead or missing indicator is a maintenance switch nobody can trust the state of.

Using maintenance mode on the job

The whole value of maintenance mode is in the two flips, and both have to happen. You flip it on before the energized work begins, while you are about to be inside the arc-flash boundary, so a fault during the exposure clears fast and the energy you face is the reduced number. You flip it off after the work is done, so the breaker returns to its coordinated settings and stops tripping on currents it should ride through.

Leave it off during the work and there is no protection. The worker is standing in front of gear at the full, un-reduced incident energy, wearing PPE that may have been selected for the lower maintenance-mode number. That is the failure that turns a safety feature into a trap: the second label says the energy is low, but only if the switch is on, and the switch is off. Engage it before you open the gear, every time, and confirm the indicator shows it engaged.

Leave it on after the work and it nuisance-trips. With the instantaneous pickup dropped low, normal load steps, motor starts, and inrush that the breaker should ride through can now trip it, and worse, the low setting can clip the coordination so a downstream fault takes out the main instead of the branch. That is why the indicator exists and why the procedure has to include the restore step. The arc-flash label and the PPE selection both assume the mode is in the right state for the task, so the state has to be verified, not assumed.

The instantaneous trip setting

An instantaneous trip is a trip with no intentional time delay: when the current crosses the instantaneous pickup, the breaker opens as fast as it mechanically can. A breaker sitting in its instantaneous region clears an arcing fault in a few cycles, which keeps the energy low. NEC 240.87 lists a permanent instantaneous trip setting, and an instantaneous override, among the accepted methods, because a device that always clears fast at the arcing current already does what the article is asking for.

The reason it is not the universal answer is coordination. A permanent instantaneous on an upstream breaker will often clear for a downstream fault that the downstream device should have cleared by itself, which defeats selective coordination. On a service that has to coordinate, you cannot just turn on instantaneous everywhere. This is the exact conflict that ZSI and the maintenance switch exist to resolve: they give you the fast instantaneous behavior only when it is needed or only during the work, instead of all the time.

Where it does fit is gear that does not need to coordinate below it, or where the instantaneous can be set above the load inrush but below the arcing current. If the arcing current lands above the instantaneous pickup and below the point where coordination breaks, a permanent instantaneous is the simplest method there is. Run the coordination study before you set it, because an instantaneous that clips a downstream device trades one problem for another.

The optical arc-flash relay

An optical arc-flash relay, sometimes called an arc-flash detection relay or an active arc-flash mitigation system, senses the light of an arc with optical sensors and confirms it with a rise in current, then trips the breaker in milliseconds. The light is what makes it fast. An arc throws a flash of light the instant it strikes, well before an overcurrent device would see enough current to act, and the relay reacts to that light almost immediately.

It is the fastest method on the list. The relay can pulse its trip output in around a millisecond, so the total arcing time collapses to roughly the mechanical opening time of the breaker, commonly in the range of 30 to 75 milliseconds. The current check is there to keep ambient light, a camera flash, or a lamp from tripping the gear, so it trips only when light and current both say arc. Sensors come as point sensors that watch a spot and fiber-optic loops that watch a whole compartment.

This is the method for switchgear lineups and critical-power gear where the available fault current is high and the lowest possible energy is worth the hardware. It is more expensive than a maintenance switch and it is its own system to install and commission, but on a high-fault bus it produces the largest energy reduction of any of the methods, because it cuts the arc to the breaker's bare opening time.

Current-limiting fuses

A current-limiting fuse clears within the first half-cycle of a high fault, before the current ever reaches its peak. It does this by melting and interrupting so fast that it cuts off the rising fault current early, which both limits the let-through energy to the gear and clears the arc faster than any breaker can. On a bus with high available fault current, that first-half-cycle clearing is exactly the fast operation 240.67 is looking for.

This is the fuse side of arc energy reduction, and on the right bus it satisfies the requirement on its own. Where the available arcing current is high enough to drive the fuse into its current-limiting range, the fuse already clears fast and no added method is needed, which is the clearing-time exception in 240.67. The catch is that current limiting depends on the fault being big enough. At a lower arcing current the same fuse may operate on its slower thermal curve and not limit at all, so you have to check the fuse against the actual arcing current at the bus, not assume it limits in every case.

Pull the fuse's let-through and time-current data and compare it to the arcing current from the study. If the fuse clears the arcing current fast enough on its curve, you are done. If the arcing current lands in the slow part of the curve, the fuse is not limiting there and 240.67 sends you to one of the other methods, such as differential relaying or a maintenance switch.

How does faster tripping fight selective coordination?

Faster tripping and selective coordination pull against each other, and understanding that tension is most of the skill in picking a method. Selective coordination means a fault on one branch is cleared by the device right above it, while everything upstream stays closed, so only the faulted branch goes dark. You buy that selectivity with time: upstream devices are intentionally delayed so the downstream device clears first. That delay is the very thing that drives the incident energy up at the upstream bus.

So the naive fix, turn on fast instantaneous tripping everywhere, blows up coordination. Now an upstream breaker clears for a downstream fault, and a fault on one load takes out a whole bus instead of one branch. On systems that are required to coordinate, this is not allowed. The NEC requires selective coordination for emergency systems and legally required standby systems, commonly cited at 700.27 and 701.18, and similar requirements apply to other critical loads. Confirm the article numbers against the adopted edition.

ZSI and the maintenance switch are the two methods that resolve the conflict instead of choosing a side. ZSI clears the upstream breaker fast only for a fault in its own zone, and holds the coordinated delay when a downstream device is handling the fault, so you get both. The maintenance switch gives up coordination only while a worker is inside the boundary and the system is under a maintenance procedure, then restores it. Differential relaying sidesteps the conflict by being selective on its own, since it only sees faults in its zone. The method you pick depends on whether the gear has to coordinate and whether it has to do so all the time or only during normal operation.

Which arc energy reduction method should you pick?

Start with the cheapest method that solves the actual problem, then move up only when the gear or the coordination forces you. For existing gear with electronic trip units, the maintenance switch is usually the answer: it is the low-cost retrofit, it drops the energy during the work, and it does not touch coordination in normal operation. The price is procedure and discipline, because it depends on a worker flipping it on and off.

On new switchgear, ZSI is often already in the trip units, so it is a setting and an interlock wiring run rather than added hardware, and it keeps coordination automatically without anyone flipping a switch. Where you need the absolute lowest energy on a high-fault bus, the optical arc-flash relay is the fastest method and produces the largest reduction, at the highest cost and complexity. On fuse-protected gear at high available fault current, current-limiting fuses may already satisfy the requirement on their own. Differential relaying is the choice for a major bus where you want fast, selective protection built into the gear.

The decision is not abstract. It comes out of the coordination study and the arc-flash study together, run for the specific lineup. The study tells you the incident energy and the arcing current at each bus; the coordination study tells you what selectivity you have to preserve. The method is whatever drops the energy at the buses that need it while keeping the coordination the system is required to hold.

How the arc-flash study quantifies the reduction

The arc-flash study is what proves the method works, because it calculates the incident energy with the reduced clearing time and shows the number drop. A method that is not in the study is a claim. A method that is in the study, with the faster clearing time modeled and the lower cal/cm2 result calculated, is a quantified reduction you can put on a label. The arc flash study and labels guide covers how IEEE 1584 turns the clearing time into the energy number.

For a maintenance switch this commonly produces two labels, or one label with two numbers: the normal incident energy with the breaker in its coordinated state, and the reduced incident energy with maintenance mode engaged. The two numbers tell a worker exactly what the mode buys and exactly what they are facing if it is off. A worker who reads only the maintenance-mode number and does not engage the switch is dressed for an energy that is not present.

Get the study to model the reduced clearing time at the actual arcing current, not at the bolted fault current and not at an assumed time. The energy reduction is only real at the current where the method actually operates faster, and the study has to use that current. A study that claims a reduction the device does not deliver at the real arcing current is worse than no claim, because it produces a confident label that is wrong.

Lower energy, lower PPE, but never skip the PPE

Lower incident energy means a lower PPE requirement, and that is much of the point. Drop a bus from a high cal/cm2 number into a lower category and the work that was a full arc-flash suit becomes a lighter, more workable system, which means the task is more likely to get done correctly and the worker is less encumbered. The arc flash PPE guide covers how the categories and arc ratings track the energy.

What arc energy reduction does not do is remove the hazard. A reduced energy is still an energy, and the shock hazard, the blast pressure, and the shrapnel are all still there regardless of how fast the breaker clears. Maintenance mode that drops a bus to a lower category still leaves a worker who needs arc-rated PPE for that category, the right shock protection, and a verified justification for working it energized at all. De-energizing remains the only state with no arc-flash hazard.

Treat the reduction as a way to make energized work safer when it genuinely cannot be avoided, not as a reason to work hot more often. The hierarchy in NFPA 70E still puts de-energizing first. Arc energy reduction is an engineering control, which sits above PPE in that hierarchy but below elimination. It lowers the energy you have to dress for. It does not replace the decision to put the gear in an electrically safe work condition whenever the work allows it.

Testing, commissioning, and verification

An arc energy reduction method is only worth what commissioning can prove it does. The most common way these methods fail is not that the hardware is wrong. It is that the method was specified, installed, and never verified, so nobody knows whether it actually clears fast when it should. The settings in the gear have to match the study, and that match has to be tested, not assumed. Recent NEC editions make this a code requirement, not just good practice: 240.87 and 240.67 now call for the arc energy reduction system to be performance tested when first installed, by a qualified person using primary current injection or another method approved by the manufacturer, with the results documented and available to the authority having jurisdiction.

The verification depends on the method. A maintenance switch gets tested by engaging it and confirming the trip unit actually drops to the maintenance pickup and that the indicator shows the state. ZSI gets tested by injecting a fault condition and confirming the upstream breaker clears fast with no restraining signal and holds its delay when the downstream breaker signals, which means proving the interlock wiring works in both states. An optical relay gets tested with a light source and a current input to confirm it trips on both together and not on light alone. Differential relaying gets tested by proving the zone trips on an internal fault and stays put for a through fault.

This is acceptance and maintenance testing, and the NETA acceptance testing specifications are the common reference for how protective devices and their settings are verified in commissioning. The critical as-left check is that the protective settings in the gear match the coordination and arc-flash study, because a breaker left on factory defaults or a trip unit never set to the study values blows up both the coordination and the labeled energy. Confirm the testing scope against the project specification and the manufacturer's instructions.

AI loads and data-center gear

Data centers and large AI compute halls concentrate the exact conditions that make arc energy reduction necessary. The services are large and stiff, the transformers are big, parallel sources and generators feed the buses, and the result is high available fault current behind many breakers at or above the 1200 A threshold. High fault current plus a coordinated delay equals a large incident energy at precisely the gear that has to be maintained while the load stays up.

The design answer on these sites is the same pairing the arc flash study and labels guide describes: selective coordination for reliability, plus arc energy reduction so the coordinated delays do not leave the energy dangerously high. ZSI on the breakers and a maintenance switch on the mains and ties are the common combination, with optical relays where the lowest energy is worth the cost. The reduction is applied across many buses, and every one of them has to be commissioned and labeled, not just the service main.

What to document

Arc energy reduction that is not documented is arc energy reduction nobody can verify or maintain. Both 240.87 and 240.67 require documentation available to the people who design, install, operate, and inspect the gear, showing the method and where it is applied. Beyond the code minimum, the record is what lets the next person confirm the method still works and the study still matches the gear.

Capture the method used at each applicable device, how it reduces the energy, the device and trip-unit settings that make it work, the normal and reduced incident energy from the study, and the verification that it was tested in commissioning. For a maintenance switch, record both the normal and the maintenance-mode numbers and the procedure for engaging it. Tie the as-left protective settings to the device so a future change can be traced and the labels can be trusted.

Field checklist

Common mistakes

• No arc energy reduction provided where 240.87 or 240.67 requires it, including breakers settable to the threshold but dialed lower.
• Maintenance mode left off during energized work, so the worker faces the full incident energy in PPE selected for the reduced number.
• Maintenance mode left on after the work, causing nuisance trips and clipping downstream coordination.
• ZSI mis-wired or never commissioned, so the interlock does not work and the breaker reverts to its coordinated delay.
• A method set above the available arcing current, so it never trips on the arc it is meant to clear.
• Turning on permanent instantaneous tripping that defeats required selective coordination.
• No second label for the maintenance-mode energy, so a worker cannot tell what the switch buys.
• No worker training on engaging and restoring the method, so the procedure depends on people who do not know it exists.
• Trusting a study reduction modeled at the bolted fault current instead of the actual arcing current where the method operates.

Standards and references

The NEC, NFPA 70, carries the installation requirement. NEC 240.87 requires a means of arc energy reduction on circuit breakers rated or settable at or above a current threshold, commonly 1200 A, and lists the accepted methods: zone-selective interlocking, differential relaying, an energy-reducing maintenance switch with a local status indicator, an energy-reducing active arc-flash mitigation system, an instantaneous trip setting, an instantaneous override, and an approved equivalent. NEC 240.67 places the same kind of requirement on fuses at a similar threshold, with a clearing-time exception commonly cited around 0.07 seconds and a method list shaped for fuses including current-limiting electronically actuated fuses. The thresholds, the method lists, and the exact wording have been revised across code cycles, so confirm them against the edition the jurisdiction has adopted and any local amendments.

Selective coordination is required for certain systems and shapes which method fits. The NEC requires it for emergency systems and legally required standby systems, commonly cited at 700.27 and 701.18, and the requirement is what makes faster tripping a coordination problem rather than a free win. Confirm those article numbers against the adopted edition.

NFPA 70E, the standard for electrical safety in the workplace, is the worker-side framework: the hierarchy of risk control that puts engineering controls like arc energy reduction above PPE and below elimination, the incident energy and the PPE selection, and the energized work permit. IEEE 1584 is the calculation method that quantifies the energy reduction. The NETA acceptance testing specifications are the common reference for verifying protective settings in commissioning. The gear and relay manufacturer's instructions govern the specific settings and behavior of any method, and the project specification and the adopted code edition control. The arc flash study and labels guide and the arc flash PPE guide cover the study and the protective equipment that go with this work.

Units, terms, and acronyms

Arc energy reduction carries a vocabulary that travels across a study report, a trip-unit manual, and the code, and the same idea sometimes wears more than one name. The terms below are the ones that show up across the whole subject.

Incident energy (cal/cm2)

Thermal energy a worker would receive at the working distance from an arc, in calories per square centimeter; what arc energy reduction lowers

Clearing time

The time from the start of an arcing fault to the device opening and starving the arc; the variable every method shortens

Arcing current

The current that actually flows in the arc, lower than the bolted fault current; a method must operate at less than this to work

ZSI

Zone-selective interlocking; breakers signal each other so the upstream device trips fast only for a fault in its own zone

ERMS / maintenance mode

Energy-reducing maintenance switch; a worker-flipped mode that drops the instantaneous pickup for faster clearing during energized work

Differential relaying

Protection that trips on the difference between current entering and leaving a zone; fast and selective on its own zone

Optical arc-flash relay

An active mitigation system that trips on the light of an arc confirmed by current, clearing in milliseconds

Current-limiting fuse

A fuse that interrupts within the first half-cycle of a high fault, limiting let-through energy and clearing the arc fast

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is incident energy in an arc-flash study?
• What are the arc flash PPE categories?
• What is NEC 110.26 working space?
• What is lockout/tagout?
• What is electrical submetering?