The TT system — why an installation's own earth electrode makes an RCD mandatory

The guide on protective earthing (§411) already briefly mentions the TT system in its earthing-system comparison table: an earth electrode local to the installation, independent of the supply network's earthing. This article goes deeper into why that one difference — an own electrode instead of a PEN or PE connection to the network — changes the entire fault-protection strategy of the installation.

What the TT system is

In the TT system, one point of the supply (generally the star point at the network's transformer) is directly earthed, and the parts of the installation liable to become live under a fault (the "exposed-conductive- parts") are earthed via an own earth electrode, independent of the supply's earthing — for example an earth rod or ring electrode at the building or the installation itself.

Why an ordinary overcurrent device is not sufficient here

In a TN system, the fault current from an earth fault returns to the source via a metallic PE or PEN connection, resulting in a low loop impedance — low enough for a fuse or circuit breaker to operate within the required time (see the guide on loop impedance testing). In a TT system, however, the fault current returns via the installation's earth electrode, through the soil, back to the network's earthing — a path with substantially higher and less predictable impedance (typically tens to several hundred ohms, depending on soil type and electrode configuration). That high loop impedance often limits the fault current to a value that an ordinary overcurrent device will not trip, or not within the required time. For that reason §411.5 mandates an RCD as the standard solution for fault protection in the TT system.

The calculation rule: RA × IΔn ≤ 50V

The test used to check whether the earth-electrode resistance, combined with the chosen RCD sensitivity, provides adequate protection is:

$$R_A \times I_{\Delta n} \le U_L$$

where RA is the resistance of the earth electrode plus the protective conductor to the exposed-conductive-parts (in ohms), IΔn is the rated residual operating current of the RCD, and UL is the conventional touch-voltage limit (generally 50V for dry, normal conditions — see the guide on touch voltage & UL limits).

Practical example: with a 30 mA RCD, an earth-electrode resistance up to roughly 1666 Ω (50V ÷ 0.03A) is sufficient — a margin that covers virtually every practical earth electrode. With a less sensitive RCD, for example a 300 mA RCD sometimes applied further upstream for selectivity or fire protection, the permitted RA is substantially lower (roughly 166 Ω) — which makes the actual earth-electrode resistance a relevant measurement, not just a formality.

Where the TT system occurs more often than expected

Greenhouses and agricultural premises: an outdoor installation of, for example, PV inverters is often supplied from PME (TN-C-S) or configured as a TT system, depending on what the grid operator offers and on the reliability of the PEN connection at that location.
Construction sites and temporary installations: where a reliable PEN connection cannot be guaranteed, a TT configuration with its own earth electrode and an RCD is the usual solution (see also the [guide on construction sites (§704)](/guides/nen-1010/704-bouwplaatsen)).
Sites without guaranteed PEN continuity: anywhere the grid operator does not deliver a guaranteed PEN connection, TT with an own earth electrode is the alternative.

Practical relevance

At handover or inspection of a TT installation, it is not enough to check only that an RCD is present — the actual earth-electrode resistance must be measured (see the guides on earth-electrode resistance measurement and the clamp-on method) and checked against the RA × IΔn ≤ 50V requirement for the RCD sensitivity actually installed. An RCD that is present but whose sensitivity does not match the measured RA does not provide the intended protection.

Common mistakes

Letting fault protection in a TT installation rely on the circuit breaker or fuse alone, as in a TN system — the loop impedance through the soil is generally too high to trip that within the required time.
Not measuring the earth-electrode resistance and only checking that an RCD is present — without that measurement it cannot be established whether RA × IΔn ≤ 50V is actually met.
Mixing a TT installation and a TN-C-S installation at the same site — for example connecting exposed-conductive-parts to both the PEN and an own earth electrode, which produces an unclear and potentially unsafe mixed system.
Applying a less sensitive RCD (100 mA/300 mA) without verifying that the measured RA is low enough for that specific IΔn — the calculation rule shifts with the RCD's sensitivity.