EN

Short circuit in DC and AC

What happens in the first milliseconds and which instant the worst case depends on

A short circuit is a very low-impedance path between points at different voltages. The current is limited only by what lies in that path: source, transformer and cables. What matters is not only its final value but how it gets there, because each check on the protective device uses a different value.

You can see everything below in the short-circuit transient simulator.

The circuit that sums it all up

For the first instants a simple model is enough: a source with a resistance R and an inductance L in series. The inductance prevents the current from jumping suddenly. The whole transient comes from this, in both DC and AC.

In direct current

\[ U = R\,i + L\frac{di}{dt} \quad\Rightarrow\quad i(t) = I_k + (I_0 - I_k)\,e^{-t/\tau} \]
\[ I_k = \frac{U}{R} \qquad \tau = \frac{L}{R} \qquad \left.\frac{di}{dt}\right|_{t=0} = \frac{U - R\,I_0}{L} \]

The current rises exponentially from its prior value to Ik = U/R. It never exceeds it and, as the source does not change, the moment of the fault does not matter: in DC no t₀ is worse than another.

That does not make it easier. The current keeps flowing in the same direction and never passes through zero, so the device has to force it down. With batteries, the resistance is so small that the current can be huge even at a modest voltage; a long line or a coil lengthens τ and gives more energy to dissipate on opening.

Example. U = 110 V, R = 20 mΩ, L = 200 µH, no prior current. Ik = 110/0.020 = 5.5 kA; τ = 200 µH/20 mΩ = 10 ms; the initial rise is U/L = 550 A/ms. After 2 ms there is already 5.5·(1 − e−0.2) ≈ 1 kA.

In alternating current

\[ e(t) = \sqrt{2}\,E\sin\omega t \qquad I_m = \frac{\sqrt{2}\,E}{Z} \qquad \varphi = \arctan\frac{X}{R} \qquad \tau = \frac{X}{\omega R} \]
\[ i(t) = \underbrace{I_m\sin(\omega t - \varphi)}_{\text{symmetrical}} \;\underbrace{-\,I_m\sin(\omega t_0 - \varphi)\,e^{-(t-t_0)/\tau}}_{\text{DC component}} \]

There are two parts. The symmetrical part is the steady-state current, with the value given by phasors. The DC component is what is needed for the current to start at zero: it has the opposite value to the symmetrical part at that instant and dies away with τ. So its size depends on t₀, the instant of the fault measured from when the voltage passes through zero.

Maxima according to t₀

It is often said that "the worst case is when the voltage passes through zero", but you have to distinguish what you want to maximise:

InstantWhat happensWhere it matters
\(t_0 = \varphi/\omega\)No DC component: the current is symmetrical from the start.The best case, and what a synchronised operation seeks.
\(t_0 = (\varphi - 90^\circ)/\omega\)Largest DC component, equal to Im.Asymmetry on opening and stress on the breaker in the first cycles.
\(t_0 = 0\)Largest first peak, about half a cycle later.Electrodynamic forces and making capacity.

In highly inductive networks φ approaches 90° and the two instants almost coincide, which is why they are often confused. They separate as R/X increases. The table comes from scanning t₀ numerically in 0.25° steps, and the maximum peak always appeared at voltage zero:

R/XφPeak with t₀ = 0Peak with maximum DCκ from the IEC 60909 formula
0.0587.1°1.8571.8561.864
0.184.3°1.7371.7331.746
0.278.7°1.5501.5391.558
0.373.3°1.4151.3971.418
0.563.4°1.2421.2141.239
145.0°1.0691.0441.069
\[ i_p = \kappa\,\sqrt{2}\,I_k \qquad \kappa \approx 1.02 + 0.98\,e^{-3R/X} \]

The peaks in the table are in multiples of √2·Ik. The approximate κ formula comes very close, and the standard uses it to avoid having to simulate.

Example. E = 230 V, R = 5 mΩ, X = 20 mΩ. Z = 20.6 mΩ and Ik = 11.2 kA. R/X = 0.25, so κ ≈ 1.48 and ip ≈ 1.48·√2·11.2 ≈ 23.4 kA. τ = 20 mΩ/(314 rad/s · 5 mΩ) = 12.7 ms: after 100 ms no DC component remains.

In three-phase systems

Each phase has its voltage displaced by 120°, so a simultaneous three-phase fault gives each phase a different t₀. The DC components of the three add up to zero. As some phase voltage passes through zero every 60°, there is always a phase near its worst case: that is why the maximum peak is calculated as if it fell on one of them. A fault between two phases is √3/2 of the three-phase value and its worst instant is when the voltage between those two phases passes through zero.

Close to an alternator

Far from generators, the AC component stays constant. Next to an alternator it does not: it starts at the subtransient value, falls to the transient value and ends at the steady-state value, sometimes below rated current. This matters because a delayed protection may find less current than expected. Running motors also contribute current in the first cycles, which IEC 60909 takes into account.

Which value checks what

Each effect depends on a different quantity, so a single "short-circuit current" figure is not enough:

EffectQuantityCheck
Force between conductors\(i_p\)Grows with the square of the peak. Busbars, making capacity Icm.
Heating\(\int i^2 dt\)Energy let through by the device against the cable's k²S².
Opening\(I_k\), \(I_b\)Breaking capacity Icu or Icn, according to the current at opening.
Tripping\(I_{k,\min}\)The weakest, furthest fault must operate the protection.

Continue with: Breaking capacity and opening, Short-circuit current calculator, Protection curves and Protective devices. The κ factor is quoted from IEC 60909; a formal calculation must follow the standard.