What is short-circuit power, and how do I get it?

What it actually is

Short-circuit power is the apparent power that would flow if you placed a dead short at a particular point on the network. Nobody intends to do that — the figure is a way of describing how stiff the supply is at that point.

A useful way to hold it: imagine the whole network upstream of you collapsed into a single ideal source behind a single impedance. Short-circuit power is another way of stating that impedance. Small impedance, large short-circuit power, stiff supply. Large impedance, small short-circuit power, soft supply.

It is a property of the network at a place, not of your equipment. The same motor on two different sites has two different answers.

Why every assessment divides by it

When you switch a load on, current flows through that upstream impedance, and the voltage at your connection point drops by the product of the two. P28's Equation 4 expresses it with the impedance already folded in:

%ΔV  =  ( Sswitched / Ssc ) × 100

The step voltage change is inversely proportional to short-circuit power. Double the short-circuit power and the step halves. That single relationship is why the number matters so much, and why "where can I connect?" is often a better question than "what can I install?".

G5 uses it too. Your harmonic currents produce harmonic voltages across the same impedance, so the same stiffness governs how much distortion your equipment causes. One request covers both assessments.

What the numbers feel like

The inverse relationship is easier to hold with one piece of equipment held constant. A 22 kW motor started direct-on-line switches about 0.155 MVA. Against a range of supplies:

The same motor is a non-issue on one supply and a refused application on another. Nothing about the equipment changed — which is why this number, rather than the equipment schedule, is usually the first thing worth establishing on a new site.

Fault level, MVA, kA and impedance

The same quantity travels under several names, which is a common source of confusion in correspondence.

To convert a three-phase fault current to short-circuit power:

Ssc (MVA)  =  √3  ×  V (kV)  ×  Ifault (kA)

So 16 kA at 11 kV is √3 × 11 × 16 = 304.8 MVA.

How to get it

Three routes, in order of how much you should trust the answer.

1. Ask your distribution network operator

The authoritative source, and normally free. Every DNO publishes a route for network data enquiries — usually an online form under a heading like "budget estimates", "connection enquiries" or "network information". You will need the site address, the MPAN if there is an existing supply, and some idea of what you intend to connect.

Turnaround is typically days to a few weeks. That lead time is the single best argument for asking early: it is free, it is not committing, and it removes the largest unknown from your design.

2. Read it from an existing study

Sites that have had connection work done often already hold the figure — in a previous connection offer, a protection study, or an arc-flash assessment. Check the date and whether the network has been reinforced since, but it saves a wait.

3. Estimate from generic values

P28's Table 7 gives generic supply impedances for LV metered connections, which our calculator can use when you have nothing better. Good enough to find out whether you have a problem; not good enough to submit.

What to ask for, precisely

A vague request gets a vague answer. Four things worth naming explicitly:

1. The short-circuit level at the point of common coupling — the point you share with other customers, not your distribution board. Say where you mean.
2. The minimum-plant value, not just the maximum. This is the one people miss. Fault level varies with which network plant is in service, and assessments must use the worst case: minimum plant, maximum impedance, smallest short-circuit power. A single "fault level" figure is usually the maximum, which is the one used for switchgear duty — the opposite end from what you need.
3. The R and X components, or the X/R ratio, if they have them. The simplified equation does not need them, but a more accurate assessment does.
4. The nominal voltage at that point, so there is no ambiguity converting between kA and MVA.

What to use while you wait

You do not have to stop. For LV metered connections, Table 7's generic impedances are enough to tell you which of three situations you are in:

• Comfortably inside the limit — carry on designing; the real figure will not change the answer.
• Comfortably outside — start considering starting methods or connection points now, rather than after the DNO replies.
• Marginal — wait for the real number before committing, because this is exactly the case where generic values mislead.

Being able to distinguish those three, on the day, is worth a great deal — and it is the actual purpose of a generic value.

It is not one number

Treating short-circuit power as a fixed property of a site is the most common conceptual error here. It varies, sometimes by a factor of two, and which value you use decides whether your assessment means anything.

Hence the phrase minimum plant. Assessments use the network in its weakest credible configuration, because that is when your equipment causes the largest disturbance — and a maintenance outage is not a rare event to be discounted, it is a normal operating state.

Why bigger is not simply better

For voltage fluctuation, a stiffer supply is unambiguously better: bigger S_sc, smaller step, easier compliance.

But short-circuit power is also the energy available into a fault. A higher fault level raises the duty on your switchgear, the required breaking capacity, and the arc-flash energy someone would be exposed to working on it.

So the two considerations pull in opposite directions, and the network you would choose for easy P28 compliance is not the network you would choose for cheap switchgear. Worth knowing before asking for a stiffer connection: you may get one, and it may cost you somewhere else.