Short-Circuit Studies: What the Per-Unit Method Is Actually Calculating

"What's the available fault current at this panel?" is a question that every coordination study, arc-flash assessment, and equipment interrupting-rating decision ultimately depends on: and in a surprising number of projects, it is a question that gets answered with a guessed, inherited, or badly outdated number far more often than it should be, given how consequential the answer actually is. A short-circuit study is not a formality that precedes the studies clients actually care about. It is the foundational input that every downstream electrical safety and protection calculation is only ever as accurate as.

This article works through what the per-unit method is actually calculating, why the method's mathematical rigor does not protect a study from bad input data, and the specific places where short-circuit studies commonly go wrong even when the underlying computation is executed correctly.

The Per-Unit Method, Explained

Rather than carrying transformer, cable, generator, and motor impedances in a mix of ohms, percentages, and per-unit values referenced to different base quantities across a system, the per-unit method normalizes every impedance in the network to a common base (typically a system-wide MVA base and each piece of equipment's own nominal voltage base) so that impedances originating from wildly different equipment types can be summed directly using simple series and parallel combination rules, exactly as one would with resistors in a basic circuit.

The foundational conversion is:

Z(per-unit) = Z(actual, ohms) × [S(base) / V(base)²]

A transformer's nameplate impedance, which is given as a percentage referenced to the transformer's own kVA rating, has to be converted onto the system's chosen base before it can be combined with anything else in the network:

Z(per-unit, new base) = Z(per-unit, nameplate) × [S(base) / S(nameplate)]

Once every element in the fault current path (the utility source, the service transformer, feeder and branch cable impedance, and any motor contribution) has been expressed on the same common base, they combine directly according to the network topology, and the available fault current at any given bus follows from:

I(fault) = I(base) / Z(per-unit, total)

where the base current is calculated as:

I(base) = S(base) / (√3 × V(base))

This methodology is standard industry practice, documented extensively in IEEE Std 141 (commonly known as the "Red Book") and carried forward in the more recent IEEE 3002 series of recommended practices for industrial and commercial power systems analysis. It is also the calculation method underlying essentially every commercial short-circuit analysis software package on the market. But the fact that the method itself is standardized and well-understood does nothing to guarantee that the input data feeding that method is accurate: and in practice, input data quality is where the vast majority of short-circuit study errors actually originate, not in the arithmetic.

Where Studies Go Wrong Even When the Math Is Right

Utility source impedance treated as a placeholder value. A short-circuit study run against a generic textbook utility fault contribution figure, or against a value that was requested from the serving utility years earlier and never refreshed, can be meaningfully inaccurate compared to the utility's current available fault current at that specific point of interconnection. Utility system configurations change over time (new generation, transmission upgrades, or reconfiguration of the local distribution network can all shift available fault current at a given service point, in either direction. Understating available fault current from the utility results in under-rated switchgear and protective devices being specified for a system that can actually deliver more fault energy than the study assumed) a genuine safety hazard. Overstating it drives unnecessary equipment cost through over-specification that provides no real safety benefit. Either error traces back to the same root cause: treating a single utility-provided number as permanently valid rather than confirming it is current for the specific study being performed.

Motor contribution omitted or oversimplified. Induction motors do not simply consume power during normal operation (during the first several cycles immediately following a fault, they act as generators, contributing fault current back into the system as their stored rotational energy is converted back to electrical energy through the motor's own internal impedance. This contribution is typically modeled using a subtransient reactance value in the range of 16.7 to 25 percent, depending on motor type and size, per standard IEEE guidance on motor fault contribution modeling. A facility with significant motor load) a pump station, an HVAC central plant, a manufacturing line with large induction motor drives: that ignores this contribution in its short-circuit study will systematically understate fault current specifically during the first-cycle window that matters most for determining equipment momentary (closing and latching) and interrupting ratings, which is precisely the window in which those ratings are actually tested.

Cable impedance calculated from assumed rather than as-built routing and length. Cable impedance is a direct function of actual physical routing length, conductor size and material, and installation configuration (cable installed in a metallic conduit behaves differently, in terms of reactance, than the same cable installed in free air or in a cable tray, because the surrounding conduit material affects the magnetic field geometry around the conductor. A short-circuit study built from design-drawing lengths and assumed installation configurations, rather than confirmed as-built lengths and configurations, is effectively calculating fault current for a system that may not be the one that was actually constructed) and the discrepancy grows with cable length, meaning the error is often largest on exactly the long feeder runs where accurate coordination matters most.

Three-phase bolted fault treated as the only scenario worth studying. A three-phase bolted fault produces the maximum symmetrical fault current in most system configurations and is the correct basis for verifying equipment interrupting and withstand ratings: this much is standard and well-established practice. But a three-phase bolted fault is not always the governing case for protective device coordination or for arc-flash incident energy calculations. Line-to-ground faults, which are statistically the most common fault type encountered in most real-world electrical systems, require their own separate analysis, particularly on solidly grounded systems where ground fault current magnitude and duration can differ substantially from the three-phase case in ways that materially affect both coordination and incident energy outcomes.

Data Collection: The Part of the Study That Actually Determines Accuracy

Given that the per-unit calculation methodology itself is standardized and essentially error-free when executed correctly in modern software, the real determinant of study accuracy is the rigor applied to data collection before the calculation is ever run. This includes confirming current utility source impedance directly with the serving utility rather than relying on historical figures, verifying transformer nameplate data against the actual installed unit rather than the originally specified unit (transformers do get swapped during construction or subsequent maintenance, and the replacement is not always identical to the original specification), confirming as-built cable routing and installation method through field verification where design records are uncertain, and building a complete and accurate motor load inventory rather than estimating aggregate motor contribution from a facility's overall connected horsepower.

None of this data collection work is glamorous, and none of it shows up as a visible line item in a typical study deliverable. But it is the single largest determinant of whether the resulting fault current numbers reflect the system that was actually built, or a idealized system that exists only in the original design intent.

Why This Feeds Directly Into Two Other Critical Studies

A short-circuit study is very rarely the final deliverable a client actually wants (it is the foundation that the protective device coordination study and the arc-flash hazard assessment are both built directly on top of. We covered how time-current curve coordination depends entirely on accurate, current fault current data) not assumed or placeholder values (in our earlier post on Selective Coordination: Why the Wrong Breaker Trips First, and that same dependency runs directly through our companion post on arc-flash studies as well) incident energy calculations performed under IEEE 1584 methodology are a direct mathematical function of the available fault current that the short-circuit study produces at each bus. Get the underlying source data wrong at this foundational stage, and every downstream number inherits that error, regardless of how carefully and correctly the later coordination and arc-flash calculations are themselves executed.

How We Approach This

Our electrical engineering team builds short-circuit models from utility-confirmed source impedance data and as-built cable routing information wherever field verification is available, specifically because these studies form the foundation for the coordination and arc-flash work that is typically delivered as part of the same engagement. We treat data collection rigor as inseparable from calculation accuracy, because a per-unit calculation executed flawlessly against stale or assumed input data produces a confident-looking result that is simply wrong. See our Electrical Engineering Services page for how we scope this work.

Conclusion

The per-unit method is a mature, well-validated calculation technique, and errors in its mechanical execution are rare in modern short-circuit analysis software. The real risk in a short-circuit study lies almost entirely in the data feeding that calculation: unconfirmed utility source impedance, omitted or oversimplified motor contribution, cable impedance based on design assumptions rather than field-verified as-built conditions, and a narrow focus on three-phase bolted faults that overlooks the ground fault scenarios governing many real-world coordination and safety outcomes. A study is only as trustworthy as the data collection effort that preceded it.

References

• IEEE Std 141: Recommended Practice for Electric Power Distribution for Industrial Plants (IEEE Red Book)
• IEEE Std 1584: Guide for Performing Arc-Flash Hazard Calculations
• IEEE 3002 Series: Recommended Practice for Industrial and Commercial Power Systems Analysis
• Related reading: Selective Coordination: Why the Wrong Breaker Trips First
• Related reading: Arc-Flash Studies Explained: What They Are and When You Need One
• Pathworks services: Electrical Engineering Services