§411.3.2 — Automatic disconnection of supply (AUV)

Automatic disconnection of supply (AUV, ADS) is the most widely used measure for fault protection (formerly "protection against indirect contact"): if an insulation fault can bring a dangerous voltage onto an accessible conductive part, the supply must be disconnected quickly enough that the touch voltage is not present long enough to be dangerous. It complements basic protection (insulation, barriers; see the guide on IP2X) and only works together with sound earthing and main equipotential bonding.

Operating principle

In the event of an insulation fault between a live conductor and a conductive part connected to PE, a fault current flows through the fault loop: source, phase conductor, the fault itself and the protective conductor (PE) back to the source. This current must be large enough to make the protective device (fuse, MCB, RCD) disconnect within the permitted time.

The core formula: Zs × Ia ≤ U₀

For TN systems the condition is:

Zs × Ia ≤ U₀

Zs = impedance of the fault loop (source + phase conductor + PE conductor), in ohms;
Ia = the current that makes the protective device disconnect within the required time, in amperes;
U₀ = nominal voltage to earth (230 V in the Dutch network).

Rearranged: Zs ≤ U₀ / Ia. The lower the loop impedance, the larger the fault current and the faster the protection responds.

Worked example

A final circuit with an MCB B16 (U₀ = 230 V):

A B-curve trips magnetically between 3 and 5 × In. Using the upper bound: Ia = 5 × 16 A = 80 A.
Zs ≤ 230 V / 80 A = 2.87 Ω.

For an MCB C16 (magnetic trip 5 to 10 × In, upper bound 10 × 16 A = 160 A) the limit is Zs ≤ 230 / 160 ≈ 1.44 Ω. A C-curve breaker therefore requires a lower loop impedance than a B-curve breaker of the same rating. For fuses, Ia is read from the time-current characteristic at the required disconnection time.

Measuring Zs

Loop impedance is measured with a loop impedance tester (installation tester), preferably at the far end of the circuit (the worst-case point); see the guide on loop impedance measurement (Zs). The measurement is made at ambient temperature, whereas conductors get warmer under load and then have a higher resistance. In practice a margin is therefore often applied: measured Zs ≤ 0.8 × the calculated or tabulated value.

Disconnection times (Table 41.1)

System	U₀ ≤ 120 V	120 < U₀ ≤ 230 V	230 < U₀ ≤ 400 V	U₀ > 400 V
TN	0.8 s	0.4 s	0.2 s	0.1 s
TT	0.3 s	0.2 s	0.07 s	0.04 s

In practice

Final circuits ≤ 32 A: see the table above (so 0.4 s at 230 V in a TN system).
Distribution circuits and fixed equipment > 32 A: max. 5 s (TN) or 1 s (TT).
For 30 mA RCD: tripping time ≤ 0.3 s at 1×IΔn and ≤ 0.04 s at 5×IΔn.
In a TT system the condition is formulated differently (earth electrode resistance × RCD operating current ≤ 50 V); see the guide on the TT system.
In an IT system a separate disconnection condition applies to a second fault, see IT system — second fault.

Common mistakes

Taking Ia = In instead of the real tripping current (the magnetic threshold of the MCB or the value from the fuse characteristic).
Measuring Zs without allowing for temperature and poor connections: a cold measurement is too optimistic; a loose terminal or oxidation can raise Zs considerably under load.
Measuring only Ze (external loop) or measuring Zs at the wrong place (in the meter cupboard instead of at the worst-case point of the circuit).
Assuming an RCD replaces the Zs check in TN. A 30 mA RCD is additional protection; the check Zs × Ia ≤ U₀ is still required, and a too-high Zs points to a physical problem (cable too long or too thin, poor connection) that must be fixed.
Broken or unconnected PE (for example a missing protective conductor to a Class I luminaire): no fault current flows and nothing trips.