TN-S vs TN-C-S vs TT Earthing: Systems Explained (Plus IT)

The earthing system decides how much current flows when a live conductor touches a metal case — and therefore whether an ordinary breaker can clear the fault or you need an RCD. Get it wrong and every disconnection-time check on the project is built on sand.

This guide explains TN-S, TN-C-S, TT and IT in plain terms, shows what each means for your protective devices, and works through the same circuit on TN-C-S and TT so you can see the difference in numbers.

Reading the letters

First letter — the source: T (terra) means the transformer star point is connected directly to earth; I means it is isolated from earth or earthed through a high impedance.

Second letter — the installation’s exposed metal parts: N means they are connected to the earthed point of the source through a protective conductor supplied with the network; T means they go to a local earth electrode, independent of the source earth.

After the dash (TN only): S = separate neutral (N) and protective (PE) conductors; C = combined in one PEN conductor; C-S = combined in the supply network, separated at the origin of the installation.

The four systems in practice

TN-S — separate PE from the source

The supply provides a separate protective conductor (often the cable sheath or a dedicated core) all the way from the transformer. Fault loops are low impedance, and there is no PEN to break. You get TN-S inside any building with its own transformer: the MV/LV substation is earthed, and from the main LV board onwards you run five wires (L1, L2, L3, N, PE). It is the preferred system for commercial and industrial buildings.

TN-C-S — combined in the network, separated in the building

The utility runs a combined PEN conductor, earthed at many points along the network (protective multiple earthing, PME). At the origin of your installation the PEN is split into N and PE, and from there on they must never be joined again. It is the most common public supply in many countries because it gives low loop impedance cheaply.

The weak point is a broken PEN upstream: the installation’s metalwork can rise towards line voltage, and everything bonded to the MET rises with it. That is why many codes restrict PME for supplies to outdoor locations, EV charge points, caravans, boats, petrol stations and construction sites, and why main bonding conductors on PME supplies are sized to the supply neutral. The usual answers are an open-PEN detection device or a TT island with its own electrode.

TT — local electrode, RCDs essential

The installation has no protective conductor from the supply; exposed metal parts are connected to a local earth electrode. The fault current must pass through two electrodes and the soil, so it is only a few amps — far too little to trip an MCB magnetically. Protection relies on RCDs. TT is common in rural networks, for outbuildings and as an island on PME supplies.

IT — isolated source, keeps running

The source is not earthed (or earthed through a high impedance). A first fault to earth produces almost no current, so nothing disconnects; an insulation monitoring device raises an alarm and maintenance finds the fault. A second fault on another phase must disconnect, like TN or TT. IT is used where an unplanned shutdown is dangerous or expensive: operating theatres and intensive care (through a medical isolating transformer), process plants and some marine installations.

TN-C — the one to avoid in buildings

A PEN conductor throughout the installation mixes load current with the protective path. It is generally permitted only in fixed cables of 10 mm² copper or larger, so building designers separate N and PE at the origin and stay TN-S from there.

System	Typical external loop impedance Ze	Earth-fault current	Main protection for shock	Typical use
TN-S	≈ 0.8 Ω	Hundreds of A	MCB/fuse (Zs check), RCD for sockets	Own transformer, commercial, industrial
TN-C-S (PME)	≈ 0.35 Ω	Hundreds of A	MCB/fuse (Zs check), RCD for sockets	Public LV supply to homes and shops
TT	≈ 21 Ω plus your electrode	A few A	RCD (RA × IΔn ≤ 50 V)	Rural supplies, outbuildings, EV/outdoor islands
IT	—	First fault: mA	Insulation monitoring + second-fault disconnection	Operating theatres, critical processes

The Z_e values, disconnection times, conductor sizes and electrode limits in this article are typical (common utility-declared maxima in BS 7671 practice) — verify with the edition adopted locally and with the values declared by your utility.

Disconnection times and the two key checks

Automatic disconnection of supply (ADS) is the main protection against indirect contact. For 230 V systems, typical maximum times are:

• Final circuits up to 32 A: 0.4 s on TN, 0.2 s on TT (some editions extend this to 63 A for socket-outlet circuits).
• Distribution circuits and larger final circuits: 5 s on TN, 1 s on TT.

TN check: Z_s × I_a ≤ U₀. Z_s is the fault-loop impedance (Ω) = Z_e + R₁ + R₂, where R₁ and R₂ are the line and protective conductor resistances of the circuit at operating temperature; I_a is the current that operates the device within the required time (A) — for an MCB, its upper magnetic threshold (5 × I_n for type B, 10 × I_n for type C); U₀ is the line-to-earth voltage, 230 V. Many designers apply a factor of 0.95 to U₀ to allow for supply voltage tolerance.

TT check: R_A × I_Δn ≤ 50 V. R_A is the resistance of the local electrode plus protective conductor (Ω), and I_Δn the rated residual operating current of the RCD (A). With a 30 mA RCD the limit is 50 / 0.03 ≈ 1667 Ω, but an electrode above about 200 Ω is usually considered unstable (it changes with season and moisture), so design for well below that.

Worked example: the same circuit on TN-C-S and TT

Which system will you have?

Usually the supply decides: a public LV supply is mostly TN-C-S (sometimes TT), and a building with its own MV/LV substation is TN-S from the main board. You choose actively in three cases:

• TT island on a PME supply for EV chargers, outdoor equipment or outbuildings where a broken PEN would be dangerous.
• IT for medical group 2 locations and processes that must ride through a first fault.
• TN-S instead of TN-C-S inside your own network: always separate N and PE at the main board.

Once the system is known, size cables and protective devices to match: the fault-loop check sets the maximum length of each MCB-protected circuit (see our MCB selection and discrimination guide), and the same lengths drive voltage drop. On unbalanced boards the neutral also carries the difference current, so keep boards balanced as shown in phase balancing in panel schedules.

How it looks on the drawing

• The single-line diagram states the earthing system and Z_e at the origin (for example “TN-C-S, Z_e 0.35 Ω”).
• The N–PE link is shown at the main board only, with the main earthing terminal and main bonding conductors to water, gas and structural steel.
• Each circuit on the schedule lists its protective conductor size and the calculated Z_s or maximum length.

Common mistakes

• Re-joining N and PE downstream of the origin (in a sub-board or at an appliance), which puts load current on the protective conductors and blinds RCDs.
• Assuming TN values on a TT supply and relying on MCBs for shock protection.
• Using the 20 °C conductor resistance in the Z_s check instead of the operating-temperature value.
• Supplying EV chargers or outdoor equipment from a PME earth without the required measures.
• Designing a TT electrode to the 1667 Ω arithmetic limit instead of a stable value well under 200 Ω.
• Not stating the earthing system and Z_e on the single-line diagram, so the reviewer cannot check any disconnection time.

Quick checklist

• Earthing system and Z_e confirmed with the utility (or calculated for your own transformer).
• N and PE separated at the origin and never joined again.
• Every MCB/fuse circuit: Z_s ≤ Z_s,max for the required disconnection time.
• TT: RCD on every circuit, R_A × I_Δn ≤ 50 V, electrode resistance stable and recorded.
• Special locations (EV, outdoor, medical) checked against PME and IT requirements.
• Main bonding conductors sized and shown on the drawings.

FAQ

What is the difference between TN-S and TN-C-S?

In TN-S the supply provides a separate protective conductor all the way from the transformer. In TN-C-S the network uses one combined PEN conductor, which is split into neutral and protective conductors at the origin of the installation. Both give low fault-loop impedance, but TN-C-S carries a broken-PEN risk.

Why does a TT system need RCDs?

In a TT system the fault current returns through the soil between two earth electrodes, so it is typically only a few amps. That is far below the current needed to trip an MCB magnetically, so an RCD is the device that disconnects the fault in time.

What earth electrode resistance is acceptable for TT?

The condition is electrode resistance times the RCD rated residual current not exceeding 50 V, which allows about 1667 ohms with a 30 mA RCD. In practice an electrode above about 200 ohms is considered unstable, so aim well below that.

Where is an IT earthing system used?

IT is used where a first earth fault must not cause a shutdown, such as operating theatres, intensive care units and critical industrial processes. An insulation monitoring device raises an alarm on the first fault, and a second fault must disconnect.

Can I supply an EV charger from a PME (TN-C-S) earth?

Often only with extra measures, such as an open-PEN detection device built into the charger or a separate TT earth electrode for the charger. Check the rules of the edition adopted locally before deciding.