RCDs in TT, TN and IT Earthing Systems
How does the earthing system change RCD requirements? On a TT system the fault-loop impedance back to the source is too high for a fuse or MCB to clear a line-to-earth fault fast enough, so IEC 60364-4-41 makes a 30 mA RCD mandatory on every final circuit, sized against the earth electrode resistance RA. On TN systems (TN-S, TN-C-S) the low-impedance metallic earth return usually lets the MCB or MCCB clear a bolted fault within the required disconnection time on its own, so 30 mA RCDs are added mainly for socket-outlets and additional protection rather than for basic fault clearance, while IT systems rely on a continuous insulation monitoring device to flag the first fault and reserve RCDs for sub-circuits with a local earth reference. This article covers the RA x IΔn ≤ 50 V rule for TT, where TN still needs 30 mA devices, how IT insulation monitoring changes the logic, and how to size sensitivity and poles once the earthing system is known.
Why the Earthing System Decides Whether You Need an RCD
Every low-voltage installation has an earthing arrangement that defines two things: how the neutral point is grounded, and how exposed metal parts are connected to earth. IEC 60364-4-41 (and its national derivatives, BS 7671 and NF C 15-100 among them) names three arrangements with a two-letter code: TT, TN, IT. The first letter is the supply earthing (T = neutral point earthed directly, I = isolated or high-impedance earthed), the second is how exposed conductive parts are earthed at the installation (T = local electrode, N = connected to the supply earth conductor).
That code is not academic. It sets the fault-loop impedance, and fault-loop impedance is what decides whether an overcurrent device can clear an earth fault inside the disconnection time the standard requires (0.4 s for 230 V final circuits, longer for distribution circuits). Low impedance means a fuse or MCB sees enough fault current to trip fast. High impedance means it doesn't, and something else has to do the job. That something is the RCD protection guide's starting premise: an RCD trips on the leakage current itself, not on the fault current reaching the source, so it works regardless of loop impedance.
TT Systems: RCDs Are Not Optional
The Loop Impedance Problem
On a TT system the installation has its own earth electrode, physically separate from the supply neutral's earth. The fault-loop impedance for a line-to-earth fault is the sum of the source earth resistance, the earth electrode resistance at the installation, and the conductor impedances between them. That electrode resistance, RA, is usually tens of ohms, sometimes over a hundred in dry or rocky ground. Ohm's law does the rest: with 230 V driving the fault and tens of ohms in the loop, the fault current is a few amps at most, nowhere near what a 16 A or 32 A MCB needs to see to trip on its magnetic instantaneous element. The breaker sits there while a live enclosure stays live.
Formula: Maximum Earth Electrode Resistance for TT Systems — Source: IEC 60364-4-41, Clause 411.5.3
RA x IΔn ≤ 50 V
| Symbol | Description | Unit |
|---|---|---|
| RA | Resistance of the earth electrode and associated protective conductor at the installation | Ω |
| IΔn | Rated residual operating current of the RCD protecting the circuit | A |
| 50 V | Conventional touch-voltage limit for normal, dry locations (25 V where risk is increased, e.g. agricultural or medical) | V |
Rearrange it and the sizing logic is obvious: at 30 mA, RA can run up to roughly 1667 Ω and still hold touch voltage under 50 V. That is a very forgiving number, which is exactly why 30 mA is the default sensitivity for TT final circuits rather than something tighter. What we see in the field is installers testing RA once at commissioning and never again; electrode resistance drifts with soil moisture and corrosion, so a site that passed in a wet winter can be marginal by a dry summer, and periodic retesting matters more on TT than on TN.
TN Systems: TN-S and TN-C-S, Where RCDs Still Apply
A TN system earths exposed metal parts through the same conductor system that carries the supply neutral back to source, either as a separate protective conductor throughout (TN-S) or as a combined neutral-and-earth conductor for part of the run (TN-C-S, also called PME in UK terminology). That metallic path has an impedance in the fraction of an ohm to low single-digit ohms, so a line-to-earth fault produces hundreds of amps, enough for a standard MCB to trip on its magnetic element well inside the required disconnection time. In theory, the overcurrent device alone satisfies basic fault protection.
In practice, most codes still require 30 mA RCDs on socket-outlet circuits and other higher-risk final circuits (bathrooms, outdoor supplies, cables buried without earthed metallic sheath) as additional protection against direct contact — protecting a person who touches a live conductor directly, which no overcurrent device can do. TN-C-S carries one extra hazard worth flagging: if the combined PEN conductor breaks upstream of the split point, exposed metalwork throughout the installation can rise to near line voltage with no fault current flowing at all, so no RCD or MCB sees anything to trip on. That failure mode is why some jurisdictions restrict PME earthing for certain outdoor and agricultural installations and specify a TT arrangement with its own electrode instead.
IT Systems: Insulation Monitoring First, RCDs on Sub-Circuits
An IT system has no direct connection between the supply neutral and earth; it is either fully isolated or earthed through a high impedance (often over 1000 Ω). A single line-to-earth fault on an IT system does not close a low-impedance loop, so fault current stays negligible, typically well under an ampere, and nothing trips. That's the point: a healthy circuit keeps running through a first fault, which is why IT earthing shows up in operating theatres, offshore platforms, and continuous process lines where an unplanned shutdown is itself a hazard.
The protection duty shifts from overcurrent devices to a permanently connected insulation monitoring device (IMD) that measures the leakage resistance of the whole system to earth and alarms when it drops below a threshold, before a second fault can turn the isolated system into an effective TN or TT fault loop with real fault current flowing. RCDs still have a role on IT systems, but downstream, on sub-distribution boards or final circuits where equipment is deliberately earthed locally — a printer or a workstation fed from an isolating transformer, for example — and even there, Type selection needs care, since some IMD injection signals can desensitize a poorly chosen RCD.
TT vs TN vs IT: Side-by-Side
| Criteria | TT | TN (TN-S / TN-C-S) | IT |
|---|---|---|---|
| Earth reference | Local electrode, independent of the supply earth | Metallic connection to the supply neutral/PEN | Isolated or high-impedance grounded supply |
| Fault-loop impedance | High (tens of ohms, electrode-dominated) | Low (fraction of an ohm) | Very high on first fault, no return path |
| First-fault behavior | MCB may not trip; RCD required for basic protection | MCB/MCCB usually clears within disconnection time | No trip; IMD alarms, supply continues |
| Primary RCD role | Mandatory, 30 mA per final circuit, basic fault protection | Additional protection on sockets and higher-risk circuits | Sub-circuits with a local earth, after IMD |
| Typical use | Rural supplies, standalone generators, sites without low-impedance PEN | Most urban and utility LV supplies | Operating theatres, offshore platforms, process plants |
Selecting RCD Sensitivity, Type and Poles by Earthing System
Earthing system tells you whether an RCD is mandatory and roughly where; it does not tell you sensitivity or type — that's a separate exercise covered in full in RCD sensitivity and pole selection, and the mA thresholds themselves are broken down in RCD sensitivity ratings. What the earthing system does change is the sizing math and the failure mode you're guarding against: on TT, RA x IΔn ≤ 50 V drives the maximum acceptable sensitivity; on TN, sensitivity is chosen for personal protection on sockets (30 mA) independent of loop impedance; on IT, the RCD sits behind an IMD and its sensitivity is set by the sub-circuit's own leakage budget, not the earthing arrangement upstream.
Poles follow the supply, not the earthing system: 2P (1-phase) RCCBs and RCBOs for single-phase final circuits, 4P for three-phase feeders, on any of TT, TN or IT. Type selection (AC/A/F/B) is driven by load waveform, covered separately, though IT systems with variable-speed drives or UPS front ends push toward Type B more often, since the isolated neutral doesn't filter DC leakage components the way a solidly earthed TN system's return path does. Across brands, Schneider's Acti9 range, ABB's F200 and DS201, and Siemens' 5SV cover 2P and 4P residual current devices for TT and TN work, with add-on and integrated RCBOs where a combined overload-plus-leakage module suits the board layout better than a separate RCCB and MCB pair — the choice between those two approaches is its own topic, laid out in MCB, RCBO, RCD and RCCB differences. For readers new to the device itself, how an RCD works covers the toroidal-core detection principle this whole earthing discussion assumes.
Frequently Asked Questions
Do TN systems need RCDs if the MCB already clears the fault?
Yes, on socket-outlets and most higher-risk final circuits. The MCB clears the fault current fast enough on TN, but it can't detect the smaller leakage current that flows through a person touching a live conductor directly. A 30 mA RCD covers that separate failure mode.
Why is a TT system considered more dangerous without an RCD?
The earth electrode resistance limits fault current to a few amps at most, far below what an MCB needs to trip on its instantaneous element. Without an RCD, a line-to-earth fault can leave an enclosure live indefinitely with no automatic disconnection.
Can a Type AC RCD be used on a TT system?
Type is about the waveform of the residual current, not the earthing system, so Type AC is technically permitted on TT if the connected loads only produce pure sinusoidal leakage. Most modern installations carry electronic loads that push the choice to Type A or higher regardless of earthing arrangement.
Does an IT system ever actually need an RCD?
Yes, on sub-circuits and sub-distribution boards where equipment has its own local earth reference, such as loads fed through an isolating transformer. The main isolated network itself relies on an insulation monitoring device rather than an RCD for first-fault detection.
What happens if the earth electrode resistance is too high for a 30 mA RCD on a TT system?
The touch voltage during a fault exceeds the 50 V limit even with the RCD tripping correctly at 30 mA, since RA x IΔn no longer satisfies the standard. The fix is either lowering RA by improving the earth electrode, or using a more sensitive RCD if the leakage budget of the circuit allows it.
Is TN-C-S (PME) treated differently from TN-S for RCD selection?
The RCD selection logic is the same, but TN-C-S carries an added risk: a broken combined neutral-earth conductor upstream can raise exposed metalwork to near line voltage without producing fault current for any device to detect. Some codes restrict PME for outdoor, agricultural or marina installations for this reason and specify TT with a local electrode instead.
Conclusion
The earthing system is the first question to answer before specifying RCD protection, not an afterthought. TT systems make a 30 mA RCD mandatory on every final circuit because the electrode resistance defeats overcurrent devices; TN systems let the MCB clear the fault current and use RCDs for additional protection on sockets and higher-risk circuits; IT systems put the first line of defense on an insulation monitoring device and confine RCDs to sub-circuits with their own local earth. Get the earthing system right first, and the sensitivity, type and pole count decisions that follow have a much narrower, more defensible set of correct answers.
