Short circuit and available fault current study field guide

What the short-circuit study is, and why it exists

A short-circuit study calculates the available fault current at every point in the distribution system, from the service down to the last panel that matters, so each piece of gear can be checked against the amount of current a bolted fault at its terminals would actually deliver. The output is a number at each bus in amps, and that number is the question every breaker, fuse, and enclosure has to answer: can you handle this when it lets go.

The reason the study exists is brutal and physical. When a fault throws thousands of amps into a breaker that was only built to interrupt a fraction of it, the breaker does not trip cleanly. It can blow apart, sustain the arc instead of clearing it, and turn the enclosure into the thing the arc-flash study warns you about. Gear rated below the available fault current is not a code paperwork problem. It is a bomb waiting for the right short.

This study is the foundation under two others. The arc-flash study uses these fault currents to compute incident energy, and the coordination study uses them to set where the time-current curves cross. Get the available fault current wrong here and both downstream studies are wrong too. The arc-flash labels and the breaker settings inherit the error. See the companion guides on the arc-flash study and on selective coordination for how each one takes this number and runs with it.

What an under-rated breaker does on a fault

A breaker has two jobs on a fault. Sense it, and interrupt it. The interrupting rating is the larger fault it can break and survive. Push more through it than that and the contacts can weld, the arc can restrike across the opening gap, and the case can rupture before the current ever stops.

The failure is fast and loud. A molded-case breaker asked to clear 40,000 A when it was listed for 22,000 A can come apart inside the panel, throw the cover, and feed the fault instead of clearing it. Now the next device upstream has to take the full fault for longer than the coordination ever assumed, and the energy that lands on anyone standing in front of that gear is far past what the label said.

The quiet version is worse because nobody sees it coming. A panel sits for years rated below the available fault current and works fine, because it never sees a real fault. The day a lug fails or a rodent bridges the bus, the gear that should have contained the event instead becomes the event. The study is how you find that mismatch on paper, while it is still cheap to fix.

What is available fault current vs an interrupting rating?

Available fault current is what the system can deliver into a bolted short at a given point. The interrupting rating and the short-circuit current rating are what the equipment can survive. The whole study comes down to one inequality holding at every point: the available fault current must be at or below the rating of the gear installed there.

Keep three terms straight, because they are not interchangeable. Available fault current, sometimes written AFC or called the available short-circuit current, is the supply side, set by the utility, the transformer, the conductors, and the motors. AIC, the ampere interrupting capacity or interrupting rating, belongs to a single overcurrent device, a breaker or a fuse, and is the largest fault it can clear. SCCR, the short-circuit current rating, belongs to an assembly, a panelboard, a switchboard, an industrial control panel, and is the largest fault the whole assembly can withstand without a hazard.

The trap that catches people is assuming a high-AIC breaker makes a high-SCCR panel. It does not. The assembly's SCCR is set by its weakest component. Drop a control panel that has a contactor rated 5,000 A, a terminal block rated 10,000 A, and a breaker rated 65,000 A into a 25,000 A spot, and the panel is a 5,000 A assembly sitting in a 25,000 A location. The breaker rating tells you nothing about that. NEC 110.9 governs the device interrupting rating, and the assembly rules at 409.22 for industrial control panels, 670.5 for industrial machinery, and 440.10 for HVAC equipment require the SCCR to be at or above the available fault current.

What is NEC 110.24 and the marking it requires?

NEC 110.24 requires service equipment in other than dwelling units to be legibly field marked with the maximum available fault current, along with the date the calculation was performed, in a marking durable enough for the environment. The marking exists so the next person who touches the service, designer, installer, inspector, or maintainer, can confirm at a glance that the gear is rated for what the system can deliver.

The marking ties back to two other sections. NEC 110.9 requires every overcurrent device to have an interrupting rating at least equal to the available fault current at its line terminals. NEC 110.10 requires the overcurrent device, the impedance of the circuit, and the component short-circuit ratings to be selected so a fault is cleared without extensive damage to the components of the circuit. The 110.24 marking is the field record that someone actually checked 110.9 and 110.10 against a real number.

Two things people miss. First, 110.24 also addresses modifications: when a change affects the maximum available fault current at the service, the value has to be verified or recalculated so the equipment ratings still hold. Second, there is an exception for industrial installations under qualified maintenance and supervision, where the marking is not required. The marking requirement has been in the code since the 2011 edition, but the exact wording shifts between cycles, so confirm it against the edition the jurisdiction has adopted. The engineer of record and the AHJ control how the calculation is documented.

Where the fault current comes from

Fault current is not one source. It is everything connected that can dump current into the short at the instant it happens, added together. Get the source list wrong and the total is wrong.

The utility is the first source, and on most projects the biggest. The power coming into the service can drive enormous current into a fault, limited mainly by the source impedance the utility provides and the service transformer. The transformer is the second source and usually the gatekeeper, because its impedance sets a ceiling on the secondary fault current no matter how stiff the utility is upstream. The conductors are the third factor, but they subtract rather than add, dropping the available fault current as you move downstream away from the transformer.

Then there are the motors, and they are the source crews forget. A running motor is a spinning machine with stored energy, and for the first cycles after a fault it acts like a generator, feeding current backward into the short. On a facility with a lot of connected horsepower, that motor contribution can push the first-cycle fault current meaningfully above what the utility and transformer alone would give. Leave it out and the study under-rates the gear nearest the motors.

The infinite bus and the transformer impedance

The standard worst case for sizing gear is the infinite bus assumption: treat the utility as a source with zero impedance that can supply unlimited current, so the only thing limiting the secondary fault is the transformer itself. It is conservative on purpose. It gives the highest secondary fault current the transformer can produce, which is exactly what you want when the goal is to make sure the gear is rated high enough.

What limits the fault under that assumption is the transformer's percent impedance, the %Z stamped on the nameplate. Percent impedance is the percentage of rated primary voltage you have to apply to drive full-load current through the secondary when the secondary is bolted shut. A 5.75 percent transformer reaches full-load secondary current at 5.75 percent of rated voltage into a short, so at full voltage it can push roughly 100 divided by 5.75, about 17 times, full-load current into that fault.

This is why lower impedance means higher fault current, and it surprises people who think a low-impedance transformer is the better one. A tight, efficient transformer with a low %Z delivers a bigger secondary fault than a higher-impedance unit of the same kVA. Swap a 5.75 percent transformer for a 4 percent unit of the same size and the available fault current on the secondary jumps. If nobody re-checks the downstream gear after that swap, the panels that were fine are now under-rated.

How is the transformer secondary fault current calculated?

Start with the secondary full-load current, then divide by the per-unit impedance. For a three-phase transformer, full-load secondary current is the kVA times 1000, divided by the line-to-line secondary voltage times the square root of three. The infinite-bus secondary fault current is that full-load current times 100, divided by the percent impedance.

Worked through: a 1000 kVA transformer at 480Y/277 V has a full-load secondary current of 1,000,000 divided by 1.732 times 480, about 1,203 A. At 5.75 percent impedance, the infinite-bus secondary fault current is 1,203 times 100 divided by 5.75, about 20,900 A. That is the symmetrical bolted fault right at the transformer secondary terminals, before any conductor takes it down and before any motor adds to it.

From the transformer out, the fault current only drops, and the point-to-point method tracks that drop. Each run of conductor adds impedance, so you compute an f-factor for the run from its length, the starting fault current, and the conductor constant, then convert it to a multiplier M equal to 1 divided by 1 plus f. Multiply the upstream fault current by M to get the fault current at the downstream point. Walk that calculation bus by bus through the whole one-line and you have the available fault current everywhere it matters.

%Z

Transformer percent impedance from the nameplate, the percent of rated voltage that drives full-load current into a shorted secondary

I_FLA

Transformer secondary full-load current in amps

L

One-way conductor length in feet for the run between the two points

C

Conductor constant from the published point-to-point tables, larger for larger and lower-impedance conductors

n

Number of conductors per phase in parallel for the run

M

Multiplier that reduces upstream fault current to the downstream point, 1 / (1 + f)

How does the point-to-point method work?

The point-to-point method walks the fault current down the system one segment at a time. You start with a known fault current at one point, usually the transformer secondary from the infinite-bus calculation, then apply the impedance of the conductor or busway to the next point to get the lower fault current there. Repeat to the end of the line.

The mechanics are the f-factor and the M multiplier, the approach published in the Bussmann and Eaton short-circuit references and consistent with the IEEE analysis practices. The f-factor captures the impedance of the run between the two points, built from the conductor length, the starting fault current, and the conductor constant C that comes from the published tables. The C value rolls up the conductor's resistance and reactance per foot. The longer the run and the smaller the conductor, the bigger the f, the smaller the M, and the more the fault current falls.

Take the 20,900 A secondary fault and run it through 100 ft of a single set of 500 kcmil copper, where C is roughly 22,000. The f comes out near 0.34, M is about 0.75, and the fault current at the downstream panel lands near 15,600 A. Same start, run it through 200 ft of #2 copper feeding a small panel, and the fault current can fall to a few thousand amps. That spread is the whole point: the gear near the transformer needs a high rating, and the gear far out the line can use less, but you have to run the number to know which is which.

Motor contribution to the fault

Running motors feed the fault. The moment a short happens and the voltage on the bus collapses, every spinning motor becomes a temporary generator, dumping current backward into the fault from the energy stored in its rotation and magnetic field. For the first cycles, that backfeed runs roughly 4 to 6 times the motor's full-load current.

The practical method when detailed motor data is not in hand is to add up the full-load amps of all the connected motors and multiply by a factor in that 4 to 6 range, commonly 5, then add that to the fault current from the utility and transformer at the point in question. It matters most at the buses close to large motor loads, where the contribution is a real fraction of the total first-cycle current. The contribution does not last. Induction motor contribution typically decays within about one to four cycles, while synchronous machines can feed the fault for six to eight cycles, which is why first-cycle and interrupting-duty calculations treat them differently.

The reason this earns its own step is that motor contribution is exactly the part a quick utility-plus-transformer estimate leaves out, and it is the part that pushes the fault current at a motor control center above what a careless study assumed. Ignore it and the gear nearest the horsepower is the gear most likely to be under-rated.

Symmetrical, asymmetrical, and the X/R ratio

The fault current the study reports first is the symmetrical RMS value, a clean sine wave centered on zero. The real fault in the first cycles is not centered. It carries a DC offset that pushes the actual peak well above the symmetrical value, and how big that offset gets depends on the X/R ratio of the system at the fault point.

X/R is the ratio of system reactance to resistance. A high X/R, common close to a large transformer, means the DC offset decays slowly and the asymmetrical peak runs high. A low X/R, common far out on resistive conductor runs, means the offset dies fast and the peak is closer to the symmetrical value. The asymmetrical first-cycle current is what the gear's closing and momentary withstand has to survive, while the lower symmetrical value at the time the device opens is the interrupting duty.

This is where a real study earns its keep over a back-of-envelope number. Equipment is tested and listed at a standard X/R. When the actual system X/R is higher than the test value, the equipment's published rating effectively comes down, and a fault current that looked safe against the symmetrical rating can exceed what the gear can take once the asymmetry is accounted for. The software handles the multiplying factors, but the engineer has to confirm the device is applied within its tested X/R, not just below its nameplate amps.

What equipment ratings have to clear the available fault current?

Every device and assembly in the fault path needs a rating at or above the available fault current at its location. For overcurrent devices, that is the interrupting rating, the AIC. For assemblies, it is the short-circuit current rating, the SCCR. For the bus and structure of a switchboard or panelboard, it is the short-circuit withstand rating, the current the bracing survives mechanically while the fault is cleared.

Check them as a set, not one at a time. A breaker can be rated 65,000 A and still sit in a panelboard whose bus bracing is only good for 25,000 A, and the bracing loses at 40,000 A while the breaker is busy proving it could have interrupted it. Fuses, because they are current-limiting, often carry very high interrupting ratings, 200,000 A or 300,000 A, and on high-fault services they are a straightforward way to get a device rated above the available current without exotic gear.

The number you check against is the available fault current at that exact point from the study, including the motor contribution where it applies, not the service number applied blindly everywhere. The gear at the service sees the most. The gear at the end of a long feeder sees far less. Rating everything for the service number is safe but expensive, and rating everything for the smallest number is how panels end up under-rated. The study is what lets you rate each point for what it actually sees.

What is a series rating?

A series rating lets a downstream breaker with an interrupting rating below the available fault current be installed legally, because an upstream device is tested to help it clear the fault. The upstream device, a fuse or breaker, is rated at or above the available fault current. The downstream breaker is rated below it. The pair is listed by a testing lab as a tested combination that works together, and that test is the only thing that makes it legal under NEC 240.86.

The hard rule is that the combination has to be a recognized tested pairing, not two devices an engineer thinks should get along. You cannot calculate your way to a series rating. It exists only where the manufacturer and the testing lab have actually run the combination and published it, or where a licensed professional engineer applies the engineering-supervision route the code allows for existing installations. The gear also has to be field marked that it is part of a series combination, because the next person needs to know not to swap the upstream device for something untested.

The limit that bites is motor contribution. NEC 240.86(C) prohibits a series-rated combination where motors connected between the two devices have a combined full-load current exceeding 1 percent of the interrupting rating of the lower-rated breaker. The reason is that UL 489 series-combination testing does not include motor contribution. The motors backfeed the fault between the two devices, where the test never accounted for it, and that extra current can push the downstream breaker past what the tested combination can handle. On a lighting panel with no motors it is a non-issue. On a power panel it can rule the series rating out entirely.

High-fault services and data centers

Some facilities sit on a service stiff enough that the available fault current swamps ordinary gear. Large transformers, low impedance, a strong utility, and parallel sources all push the number up, and data centers manage to combine most of those at once.

A data center pairs a high-capacity utility service with banks of paralleled generators and UPS modules, and a lot of connected motor and fan load. Each source adds to the fault. Multiple transformers feeding a common bus, or generators paralleling onto it, can drive the available fault current well past what a standard 65,000 A panelboard can take, and into territory where current-limiting fuses, high-AIC breakers, or a deliberately higher-impedance transformer become the design decision rather than an afterthought.

These are also the facilities where the short-circuit study, the arc-flash study, and the coordination study have to be done as one coherent package, because the same high fault currents that strain the equipment ratings also drive the incident energy up and make coordination harder to achieve. The fault numbers from this study set the stage for both of the others.

What the study delivers

The deliverable is a calculated available fault current at every significant bus, backed by a one-line diagram, the source data, and a report that an engineer signs. The one-line is the spine of the whole thing: it shows the utility source, the transformers with their kVA and %Z, the conductors with size and length, the protective devices, and the major motor loads, all the way down to the buses being evaluated.

The calculation almost always runs in software now. ETAP and SKM Power Tools are the two names you will see most on commercial and industrial work, and they model the system per the IEEE recommended practices, including IEEE 3002.3 for short-circuit studies, with the older IEEE 141 Red Book and IEEE 242 Buff Book still cited for the underlying methods. The point-to-point method by hand is fine for a small system or a sanity check, but a real facility with parallel sources, motor contribution, and dozens of buses gets modeled.

The report lists each bus, its available fault current, the gear installed there, that gear's rating, and a pass or fail on whether the rating clears the fault. That table is the answer the whole study exists to produce. It is also the input file for the arc-flash study, which takes these fault currents and clearing times to compute incident energy, and for the coordination study, which uses them to place the device curves. One set of fault numbers, three studies built on it.

The field data and the utility letter

The study is only as good as the data fed into it, and the single most important input comes from the utility. Request the available fault current at the service in writing, the utility letter, because the utility's contribution sets the top of the whole calculation and only they know their source impedance and how it changes as their system grows.

From the utility number down, the rest is field data you collect or pull from the drawings. Transformer nameplates give the kVA and the %Z, and the nameplate beats any assumed value, because two transformers of the same size can have different impedance. Conductor data, size, material, length, and the number of parallel sets, sets how fast the fault current drops downstream. Motor schedules and connected horsepower set the contribution. Service and main gear nameplates give the ratings you are checking against.

The number that quietly wrecks studies is conductor length. Estimated lengths off a plan understate the real routed run, and since longer conductors lower the downstream fault current, an underestimated length makes the downstream fault look higher than it is, which is conservative, while an overestimated length makes it look lower than it is, which is not. When the result is close to a rating limit, measure the run instead of guessing it.

When the study has to be redone

The available fault current is not fixed for the life of the building. It moves when the system upstream of the gear changes, and the most common mover is the utility. Utilities upgrade their distribution, add capacity, and stiffen the source over time, and a stiffer source raises the available fault current at your service without anyone on site touching anything. The gear that was rated correctly at install can quietly fall below the new available current.

Inside the fence, a transformer swap is the classic trigger. Replace a failed transformer with one of lower impedance or higher kVA and the secondary fault current rises, and everything downstream that was checked against the old number needs to be checked again. Adding a generator or a second utility feed, paralleling sources, or adding significant motor load all push the fault current up too. NEC 110.24(B) addresses exactly this: when a modification affects the maximum available fault current at the service, the value has to be verified or recalculated and the equipment ratings rechecked.

A common cadence on facilities that maintain an electrical safety program is to revisit the studies on a multi-year interval, often tied to the arc-flash review cycle, and any time a known change hits the source. The re-study is cheap next to the gear failure it prevents. Skip it after a utility upgrade and you are betting the building on a number that is no longer true.

Verifying the gear at commissioning and re-marking after

Commissioning is where the study stops being paper and becomes the gear on the wall. The commissioning agent's job here is direct: confirm that every device and assembly installed is actually rated at or above the available fault current the study calculated for its location, and that what got installed matches the one-line the study was built on.

Field substitutions are where this falls apart. The study assumed a 65,000 A panel and the contractor installed a 22,000 A panel because that is what was on the shelf, and unless someone checks the nameplate against the study, it ships under-rated. Confirm the transformer that arrived has the %Z the study assumed, because a substituted transformer with lower impedance raises the fault current the downstream gear has to take. Verify the AIC and SCCR nameplates, not the catalog numbers from the submittal that may not match what was delivered.

Then mark it. NEC 110.24 requires the service to be field marked with the maximum available fault current and the date, so the marking goes on at commissioning with the study's number. When a re-study later changes the available fault current, the label gets changed too, with the new value and new date, so the marking on the gear always reflects the current calculation. The arc-flash labels, which carry their own data from the companion study, get applied in the same pass. A label that no longer matches the system is worse than no label, because it tells the next person a number that is wrong.

What to document

The study's value lives in the record. Years out, when a transformer gets swapped or the utility upgrades, the next engineer needs to see what was assumed, what was calculated, and whether the gear cleared it, without re-doing the whole thing from scratch.

Capture each evaluated bus, the available fault current there, the device or assembly installed, its interrupting rating or SCCR, a clear pass or fail on whether the rating exceeds the available current, and the source of each input. Record the utility's available fault current and the date of their letter, every transformer kVA and %Z, the conductor sizes and lengths used, the motor contribution assumed, the software and version, and the engineer of record who sealed it. The service marking carries the headline number and date; the report carries everything behind it.

Common mistakes

• Installing gear rated below the available fault current at its location, the failure the whole study exists to catch.
• Skipping the NEC 110.24 service marking, or leaving an old marking up after a re-study changed the number.
• Leaving motor contribution out, so the buses near large motor loads come back under-rated for the first cycles.
• Assuming a high-AIC breaker gives the panel a high SCCR, when the assembly rating is set by its weakest component.
• Applying a series rating that is not a listed tested combination, or applying one where motor contribution exceeds the 240.86(C) limit.
• Not re-studying after a utility upgrade, a transformer swap, or an added source changed the available fault current.
• Building the study on the design one-line instead of the as-built, so substituted gear and routed lengths never get checked.
• Checking only the symmetrical rating and ignoring the X/R, so the asymmetrical first-cycle peak exceeds what the gear can take.

Field checklist

Standards and references

The NEC, NFPA 70, sets the requirements the study satisfies. NEC 110.9 requires interrupting ratings at or above the available fault current at the device line terminals. NEC 110.10 requires the overcurrent device, circuit impedance, and component ratings to clear a fault without extensive damage. NEC 110.24 requires the service to be field marked with the maximum available fault current and the calculation date, with verification on modifications under 110.24(B) and an exception for supervised industrial installations.

Assembly ratings carry their own sections: 409.22 for industrial control panels, 670.5 for industrial machinery, and 440.10 for air-conditioning and refrigerating equipment, each requiring the SCCR to equal or exceed the available fault current. Series ratings live at 240.86, including the 240.86(C) motor-contribution limit. The exact article and section numbers shift between code cycles, so confirm them against the adopted edition and local amendments before citing them on a submittal.

The calculation methods come from IEEE: the recommended practice for short-circuit studies, IEEE 3002.3, with the IEEE 141 Red Book and IEEE 242 Buff Book behind it, and the point-to-point approach published in the Eaton and Bussmann short-circuit references. The utility governs the available fault current at the service and supplies it in writing. UL listings set the AIC and SCCR of the equipment. The engineer of record owns the study and seals it. The companion arc-flash study, using IEEE 1584, and the selective-coordination study both build on the fault currents this study produces.

Units, terms, and synonyms

The same idea shows up under different names across a utility letter, a nameplate, a study report, and a code section, so the vocabulary is worth pinning down.

Available fault current is also called available short-circuit current, AFC, or the available short-circuit kA, and it is reported in amps or kiloamps RMS symmetrical. AIC, the interrupting rating, and SCCR, the short-circuit current rating, are both in amps or kiloamps but apply to different things, a device versus an assembly. Transformer impedance is %Z or just Z in percent. Fault current is reported as symmetrical RMS, with asymmetrical and peak values higher by a factor that depends on the X/R ratio.

Available fault current (AFC)

The current a bolted fault can draw at a point, set by the utility, transformer, conductors, and motors

AIC / interrupting rating

The largest fault a single overcurrent device can interrupt safely, in amps or kA

SCCR

Short-circuit current rating, the largest fault an assembly can withstand without a hazard

%Z

Transformer percent impedance, which sets the ceiling on secondary fault current

X/R ratio

System reactance over resistance, which sets how high the asymmetrical first-cycle peak runs

Symmetrical RMS

The fault current measured as a sine wave centered on zero, the base value the study reports

Infinite bus

The conservative assumption of a zero-impedance utility source, used for worst-case fault current

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 is selective coordination?
• What is a service entrance?
• What is document control in construction?
• How do you find an underground cable fault?