Earthing Systems Explained: TN-S, TN-C-S (PME), TT & IT

Whether a circuit disconnects safely in a fault, what protective device you need, and how much bonding the job requires all depend on one thing decided before you draw a single cable: the earthing system. BS 7671 describes it with a short letter code — TN-S, TN-C-S, TN-C, TT, IT — and getting it right is the foundation of a safe design. Here is what each code means, how to tell which arrangement you have on site, and what changes in your calculations as a result.

Try it: the earthing system explorer

Pick a system, then break something. The diagram redraws, the earth-fault loop is recalculated against the real BS 7671 limits, and the open-PEN scenario shows exactly how far the earthed metalwork can rise.

The naming code

The letters come from IEC 60364 and read left to right. Note that "earthing" and "grounding" are the same thing — the codes are international, so a TT grounding system and a TT earthing system are one and the same arrangement.

• First letter — the supply's relationship to earth: T (terra) = one point, the transformer star point, is directly earthed; I = isolated from earth, or earthed only through a deliberate impedance.
• Second letter — how your exposed metalwork is earthed: T = to your own local earth electrode; N = to the supply's earthed point, via a protective conductor supplied with the electricity.
• Further letters (S / C) — how neutral and earth are arranged in the supply: S = separate conductors; C = combined into one conductor, the PEN.

What is a TN system?

TN is the umbrella. In any TN system of earthing, your exposed conductive parts are connected back to the supply's earthed point through a conductor that comes with the supply — not through an electrode of your own. That gives a metallic, low-impedance fault path, which is why TN systems can normally rely on an ordinary overcurrent device to disconnect an earth fault fast enough.

The letters that follow describe how neutral and earth are arranged inside that supply, and there are three arrangements: TN-S keeps them separate throughout, TN-C combines them throughout, and TN-C-S combines them in the distributor's network then separates them at the origin of your installation. Almost every modern UK supply is the last of those.

TN-S earthing system — separate earth throughout

Neutral (N) and protective earth (PE) are separate conductors all the way from the transformer to the installation. The earth path is traditionally the metallic sheath or armour of the service cable. It gives a reliable, low-impedance earth and avoids the combined-conductor risk of PME entirely. Common on older urban supplies, and the typical declared maximum Ze is 0.8 Ω.

The one thing worth checking on an older TN-S service is that the earth really is still sound. Lead-sheathed cables corrode, and joints degrade; a supply that was TN-S when it was installed can behave very differently decades later. Measure Ze rather than trusting the label.

TN-C-S (PME) earthing system — combined, then separated

The supply brings a single PEN conductor — combined protective earth and neutral — to the property, and it is split into separate N and PE at the origin. This is Protective Multiple Earthing (PME), the standard modern UK arrangement. The "multiple" refers to the PEN being earthed repeatedly along the distributor's network, which is what keeps it close to earth potential in normal service.

It usually has the lowest external loop impedance — a typical declared maximum Ze of 0.35 Ω — so faults clear fast and cable sizing is easier. But it carries the open-PEN risk, and that single risk drives the strict main-bonding requirements on PME supplies and the special rules for outdoor EV charging.

The open-PEN fault

If the combined conductor breaks out in the network, the installation's load current has no low-impedance route back to the transformer. It returns instead through whatever earth path the installation has — the main bonding, buried metal services, an electrode — and the whole earthed metalwork of the installation rises above true Earth by the voltage that current drops across that path. With a poor earth path, that voltage approaches the full supply voltage.

The maths behind the explorer's open-PEN mode is the condition in BS 7671 Annex A722.3, which underpins Regulation 722.411.4.1(b). It comes in two forms. For a single-phase installation the electrode resistance must satisfy Ra ≤ 70 × U0·Cmax / (Iinst × (U0·Cmax − 70)), where Cmax is 1.1. At a 60 A demand that is about 1.6 Ω; at 100 A it is under 1 Ω. A driven rod is typically tens of ohms, which is precisely why an electrode on its own almost never satisfies the requirement and why PEN-fault detection is the practical route.

For a three-phase installation the condition becomes Ra ≤ 70 × U0·Cmax / (Im × U0·Cmax − 70 × (IL1 + IL2 + IL3)), where Im is the neutral current from Annex A722.1: Im = √(IL1² + IL2² + IL3² − IL1IL2 − IL1IL3 − IL2IL3). Put Im = ΣIL = Iinst and it collapses back to the single-phase form, so they are one condition, not two.

The three-phase case carries a genuinely useful insight. A balanced three-phase load has almost no neutral current, so there is little to divert when the PEN breaks and the metalwork barely moves — at perfect balance, Im is zero and it does not move at all. That is why NOTE 1 lets you take 200 Ω whenever Im ≤ 70 × ΣIL / (U0·Cmax), and why the hazard is really a story about imbalance rather than about total demand. Switch the explorer's load to three-phase and drag the line currents together to watch it happen. Note also that these are all magnitudes, not phasors (NOTE 2).

TN-C earthing system — and why you will not see it

In a TN-C system the neutral and earth stay combined as a PEN conductor right through the installation — the split never happens. Two things follow. First, there is no separate protective conductor, so an RCD cannot work: an RCD detects the imbalance between line and neutral, and here the earth return is the neutral. Second, a break in the PEN anywhere makes every connected earthed part live.

BS 7671 does not permit TN-C for a general consumer installation. If you find a combined conductor continuing past the origin, that is a defect to be corrected — splitting it into separate N and PE at the origin is what turns it into TN-C-S. TN-C survives only in specific, controlled situations such as parts of a distributor's own network.

TT earthing system — your own earth electrode

The distributor provides no earth; the installation makes its own connection to earth with a local electrode, usually one or more rods. Common in rural areas and wherever the supply is overhead. Because an electrode's resistance is high compared with a metallic supply earth, the earth-fault loop impedance is high — commonly a declared Ze of 21 Ω, and often much more in practice.

That is too high for an overcurrent device to clear an earth fault in time, so a TT system needs an RCD for automatic disconnection, sized so that Ra × IΔn ≤ 50 V (Regulation 411.5.3). For a 30 mA RCD that permits an electrode resistance up to 1,667 Ω, which is why a 30 mA device passes comfortably on almost any real electrode — but a 500 mA time-delayed device used upstream is limited to 100 Ω, and that one does bite.

IT earthing system — rare in the UK

The supply is unearthed or earthed only through a high impedance, and exposed metalwork is earthed locally. A first fault does not cause a large fault current, so supply continuity is maintained and an insulation monitoring device raises an alarm instead of tripping. That is why IT is used where losing the supply is itself the hazard — some medical locations, industrial process plant and safety-critical installations. A second fault on a different line becomes a line-to-line fault and must be disconnected, so the protection is designed around that case. IT is uncommon in general UK distribution.

Which earthing system do I have?

Work from what you can see at the intake, then confirm it by measurement. The identifier below walks the same logic an engineer would use on site.

At a glance

Ze figures are the DNO's typical declared maxima, not measured values — always use the figure measured or declared for the actual supply. "Not for ADS" means an RCD is not needed to achieve automatic disconnection; in practice RCDs are near-universal anyway because of Regulation 411.3.3 (socket-outlets up to 32 A) and 522.6.202 (cables concealed in walls).

Maximum Zs by protective device

On a TN system, the check is simply whether your measured Zs is below the maximum for the device. For circuit-breakers the figure is Zs = U0 × Cmin / Ia with U0 = 230 V and Cmin = 0.95, where Ia is 5 × In for a Type B, 10 × In for a Type C and 20 × In for a Type D. Fuses have no such formula and are read from the tables.

Maximum Zs for a 0.4 s disconnection time at U0 230 V — BS 7671 Table 41.3 for circuit-breakers and Table 41.2 for BS 88-2 fuses. Values apply at the conductor operating temperatures in the table notes; adjust a cold test reading accordingly. Indicative subset — verify against your own copy of BS 7671. On a TT system the limit is not from this table at all: it is 50 / IΔn.

Maximum disconnection times

AC systems with U0 between 120 V and 230 V. Note the exception under Table 41.1: where disconnection in a TT system is achieved by an overcurrent device and protective equipotential bonding is connected to all extraneous-conductive-parts per Regulation 411.3.1.2, the TN times (0.4 s / 5 s) may be used instead.

Main protective bonding on a PME supply

Bonding matters on every system, but on PME it is doing safety-critical work: it is what keeps everything in the building at the same potential when the PEN misbehaves. The minimum size is set by the copper-equivalent CSA of the supply PEN conductor, not by your main switch.

BS 7671 Table 54.8. The distributor's local network conditions may require a larger conductor — ask. Where an installation has more than one PME source, size from the largest supply PEN.

Where PME must not be used

A PME earth is excellent right up until the PEN fails, and the situations where that failure would be lethal are precisely the ones BS 7671 carves out:

• Outdoor EV charge points (Section 722). A PME earthing facility must not be used for a charge point outdoors, or one that could reasonably be used to charge a vehicle outdoors, unless Regulation 722.411.4.1(b)–(e) is met.
• Caravans and motor caravans (Section 708), boats and marinas (Section 709). The person is in good contact with true earth and holding metalwork bonded to the PEN — the worst possible combination.
• Petrol filling stations and similar hazardous areas. Potential differences between the PME earth and true earth in a zone containing flammable vapour are an ignition risk.
• Swimming pools and other special locations where Part 7 imposes its own earthing and bonding requirements.

Converting PME to TT for an EV charge point

Where the electrode route is chosen, you are not converting the whole installation — you are creating a separate TT island for the charge point. In outline: the charge-point circuit gets its own earth electrode; that electrode's earth is kept separate from the PME main earthing terminal and far enough from any PME-earthed metalwork that the two earth zones do not interact; the circuit is protected by its own RCD; and the electrode resistance must satisfy the Annex A722.3 condition above, not merely the 50 V TT rule.

In practice this is hard to achieve and awkward to prove, which is why most installations use a charge point with built-in PEN-fault detection instead — a device meeting indent (c), which disconnects on a cpc-to-Earth rise above 70 V, or indent (d), which disconnects when the line-neutral voltage leaves the 207–253 V window. Read the manufacturer's declaration rather than assuming a unit has it.

Why the system governs your design

The earthing system is not a label you record at the end of the job — it is an input to almost every calculation you make:

• It sets the Ze you start from, and therefore every Zs down the installation and whether your device disconnects in time or you need an RCD. Try it in the free BS 7671 cable sizing calculator, where the earthing system is one of the first board settings.
• It sets the prospective fault current, which drives breaking capacity and the adiabatic check on your cpc — see the fault level guide.
• It sets the main protective bonding requirements, heaviest on PME.
• It constrains specific applications — EV charging, caravans, marinas, pools.
• On HV sites the earthing arrangement is a study in its own right — see earth potential rise and our earthing study design service.

Glossary

Frequently Asked Questions

Sources and verification. BS 7671:2018+A4:2026 Chapter 41 (Table 41.1 disconnection times, Tables 41.2–41.4 maximum Zs, Reg 411.5.3 for TT), Chapter 54 (Table 54.8 main protective bonding), Section 722 and Annex A722 (EV charging on PME). The figures on this page are an indicative subset for explanation — verify every value against your own copy of BS 7671 and the figures measured or declared for the actual supply. For wider background see the IET's Wiring Matters series.