Steel wire armour as a protective conductor — why the armour's cross-section must be verified in its own right

The guide on single-core cables and eddy currents in a steel gland plate covers why single-core armoured cable is discouraged for AC because of eddy-current heating in the steel armour. This article covers a different, frequently asked question about the same armour: may the steel armour of a multicore armoured cable serve as a protective conductor, and if so, under what conditions.

Armour as a protective conductor is permitted, but not automatic

Multicore armoured cable (for example SWA cable) has a steel wire or tape armour that surrounds the whole cable. With multicore cable — unlike single-core cable — there is no eddy-current problem, because the armour encloses all line conductors and the neutral together, and the net magnetic flux through the armour under balanced load is close to zero. The armour may therefore in principle serve as the protective conductor (PE) or part of it, similar to a cable screen used as a combined screen-and-protective-conductor (see also the guide on cable screen bonding for a related question with screens rather than armour). This is not automatic, however: §543.1 (IEC 60364-5-54) requires that any conductor serving as a protective conductor be verified, either by calculation using the adiabatic method or by test, to have adequate cross-section for the expected earth fault current and disconnection time.

Why steel's k-factor is the issue

The adiabatic formula (§543.1.2) is S = √(I²·t) / k, where k is a material constant representing the conductor's heat capacity and permissible temperature rise. For copper protective conductors, k typically lies between roughly 115 and 143, depending on the insulation type and whether the conductor is in contact with insulating material. For steel, the k-factor is considerably lower — steel has a lower volumetric heat capacity relative to its conductivity, and a higher resistivity than copper. The result is that a steel armour with an apparently generous cross-section still has a considerably smaller effective protective capacity than a copper conductor of comparable cross-section, for the same earth fault current and disconnection time. The armour's cross-section must therefore be explicitly calculated or tested — not assumed from the apparent size of the armour wires.

The gland: a mechanical and electrical joint at once

A second condition is that the armour is properly terminated at each cable entry. A suitable armoured-cable gland (for example a gland with a clamping ring and sealing washer specifically designed for armoured cable) mechanically clamps the armour wires between an inner and outer component, and simultaneously provides the electrical contact between the armour and the metal enclosure or earthing terminal. If an ordinary cable gland is used instead, which only seals the outer sheath, the armour is mechanically present but not reliably bonded electrically — with the result that the armour does not actually form a usable protective conductor, regardless of its calculated cross-section. At a joint or branch, the same continuity must be preserved, for example with an earth braid bridging the armour on both sides of the joint.

Practical relevance

When assessing an existing installation where the armour serves as the sole protective conductor — without a separate green-and-yellow core — it must be verified that (1) the glands at both ends are suitable for armoured cable and actually clamp the armour, (2) the armour's cross-section has been checked against the earth fault current and disconnection time of the circuit concerned, using the correct k-factor for steel (see also the guide on short-circuit protection for the relationship between disconnection time and protective device characteristic), and (3) continuity is present at every joint or branch. If any of these elements is missing, a separate protective conductor must be installed instead of relying on the armour (see also the guide on sizing the protective conductor for the general sizing rules).

Common mistakes

Assuming the armour is suitable as a protective conductor without calculating its cross-section using the k-factor for steel — the nominal cross-section of the armour wires says nothing by itself about the protective capacity.
Using an ordinary cable gland instead of an armoured-cable gland, so that the armour is mechanically present but not reliably bonded electrically.
Not providing continuity at a joint or branch, so that the protective conductor is effectively interrupted at that point.
Treating the armour of single-core cable the same as that of multicore cable, while for single-core cable eddy-current heating is the dominant problem instead (see the guide on single-core cables and eddy currents).