Stray currents through parallel N-PE paths — why a second earth point on the neutral is dangerous

The guide on §543.4 — minimum cross-section and switching prohibition of the PEN conductor and the guide on PEN conductor breaks in TN-C-S cover two well-known PEN-related risks: a PEN conductor that is too thin, and a complete interruption of it. This article covers a third, less well-known but in practice regularly occurring problem: what happens when the neutral (N), after the splitting point, is nevertheless connected at a second location to earth or to an extraneous-conductive-part — not through a break, but through an extra, unintended connection.

The principle: one splitting point, then strictly separated

In a TN-C-S installation, the combined PEN conductor is split at exactly one point — usually at the main incoming connection — into a separate neutral conductor (N) and protective conductor (PE), as covered in the PEN conductor guide. From that splitting point onward, N and PE must remain strictly separated: the neutral conductor must from then on never again be connected to the protective conductor, to the main earthing busbar, or to any other earthed or extraneous-conductive-part.

How a second, unintended N-PE connection arises

In practice, such an unwanted second connection rarely arises deliberately, but through a roundabout route:

An electrician mistakenly connects the neutral bar in a subdistribution board to the earth bar of that same enclosure (an "erroneous re-bonding").
A building is effectively earthed at more than one point via structural steel that is in contact at multiple points with both the main PE earthing busbar and a neutral point.
Two buildings fed from the same source share a common metal pipe earthed at both locations — for example a water pipe or a cable tray — creating a parallel path between the two buildings alongside the actual N conductor in the supply cable.
A data cable screen between two buildings that is connected to the local earth bar at both ends (see also the guide on cable screen earthing) forms a comparable extra path if both earth bars are also connected to the neutral conductor of their own installation via the network.

Why this causes stray currents

As soon as such a second N-PE or N-earth path exists, the normal operating current of the neutral conductor is no longer forced to flow entirely back to the source via the intended neutral conductor: part of that current — the greater the asymmetry between the phases and the lower the impedance of the alternative path, the greater that part — flows via the unintended parallel path: through structural steel, a water pipe, a cable screen, or another extraneous-conductive-part that was never intended to carry operating current. This stray current (also called an objectionable current in the English-language literature) can have various different consequences:

Nuisance RCD tripping: an RCD measures the difference between the current flowing out via the phase and the current returning via the neutral; if part of that neutral current flows via a parallel path instead of via the device's own neutral conductor through the RCD, the RCD detects this as an apparent residual current and may trip incorrectly, without there being an actual earth fault.
Corrosion: a persistent DC or asymmetric AC component through an underground metal pipe (for example a water pipe acting as an unintended return path) can accelerate galvanic corrosion at the point where that current leaves the pipe again.
Elevated touch voltage during normal operation: extraneous conductive parts that are normally supposed to be current-free can, as a result of the stray current, acquire a small but, during peak load, sometimes noticeable voltage relative to earth, even without an actual insulation fault anywhere.

The difference from a PEN break

This mechanism is expressly different from a PEN conductor break: with a break, a return path is precisely missing, resulting in a dangerous voltage rise; with a parallel N-PE path, there is instead an extra, unintended return path alongside the intended path — the neutral conductor itself remains intact, but unwillingly shares its function with another conductive part not designed for that purpose.

Practical relevance

When assessing an existing TN-C-S installation, particularly for buildings interconnected via structural steel, piping, or data cables, it is important to explicitly check whether the neutral conductor is in fact not connected a second time to earth or to an extraneous-conductive-part after the splitting point — measuring the current through an extraneous-conductive-part that is normally supposed to be current-free (for example with a clamp meter around a water pipe or around a cable screen) can reveal such a parallel path.

Common mistakes

Treating a stray-current problem as a PEN break, while the underlying mechanism — an extra, parallel path instead of a missing path — and therefore also the solution, is fundamentally different.
Mistakenly connecting the neutral bar of a subdistribution board to the earth bar of that same enclosure, on the assumption that this is "extra safe", while this in fact creates a prohibited second N-PE connection.
Not accounting for structural steel, piping, or cable screens between buildings as possible unintended parallel paths, and looking for the problem exclusively within the installation cabling itself.
Attributing nuisance RCD tripping to a faulty RCD, without investigating whether a parallel N-PE path elsewhere in the installation is the actual cause.