An earth electrode on a TT system passes when RA × IΔn is no more than 50 V (Regulation 411.5.3 of BS 7671): on paper that is 1667 Ω for a 30 mA RCD, 500 Ω for 100 mA and 167 Ω for 300 mA. Note 2 to Table 41.5 adds that the resistance should be as low as practicable and that a value above 200 Ω may not be stable, so 200 Ω is the working ceiling. Regulation 643.7.2 requires the resistance to be measured, either with a dedicated earth electrode instrument by the fall-of-potential method or with an earth fault loop impedance instrument at the origin. Both are done with the installation isolated and the earthing conductor disconnected.
- On TT the installation's only earth is its own electrode. Every earth fault current returns through the ground, so the loop is tens or hundreds of ohms, not fractions of one.
- The rule is RA × IΔn ≤ 50 V (411.5.3). The 200 Ω figure is a stability note in BS 7671 itself (Table 41.5, Note 2), repeated in IET Guidance Note 3 and the On-Site Guide.
- Fall of potential measures the electrode directly. The loop impedance method is accepted by GN3 for RCD-protected installations and reads slightly high, which is the safe side.
- Isolate before you disconnect the earthing conductor. While it is off the installation has no earth at all.
- A 1.2 m rod in ordinary soil lands around 70 Ω; the same rod in dry sand can be over 350 Ω. Depth is the cheapest fix.
- On an EICR, no effective earth and a water or gas pipe used as the electrode are both C2 in Electrical Safety First's Best Practice Guide 4.
Why does a TT installation depend on its electrode?
In a TT system the supply neutral is earthed at the transformer and the installation is earthed separately through its own electrode, with no metallic path between the two. The distributor's earth terminal is either not provided, as on many rural overhead supplies, or deliberately not used. The three arrangements are compared at Earthing Arrangements Explained: TN-S, TN-C-S and TT.
The consequence is fault current. On a TN-C-S supply with Ze of 0.35 Ω, a line-to-earth fault draws several hundred amps and a 32 A Type B circuit-breaker trips in milliseconds. On a TT supply with a 100 Ω electrode the same fault draws about 2.3 A. No fuse or circuit-breaker will ever see that as a fault, and exposed metalwork sits near 230 V until something else disconnects it. Regulation 411.5.2 therefore makes RCDs the preferred fault protection on TT, and allows overcurrent devices only where a suitably low Zs is "permanently and reliably assured". A B32 needs 1.37 Ω; a rod in the garden will not give you that.
So on TT two things have to work together: the RCD, and an electrode good enough to let the RCD see the fault while keeping the touch voltage down. The test proves the second half.
What resistance is acceptable for a TT earth electrode?
| RCD rated residual current (IΔn) | Maximum RA for 50 V (Table 41.5) | Touch voltage at 200 Ω | Verdict at 200 Ω |
|---|---|---|---|
| 30 mA | 1667 Ω | 6 V | Meets 411.5.3 |
| 100 mA | 500 Ω | 20 V | Meets 411.5.3 |
| 300 mA | 167 Ω | 60 V | Does not meet 411.5.3: 167 Ω is the limit |
| 500 mA | 100 Ω | 100 V | Does not meet 411.5.3: 100 Ω is the limit |
RA is the sum of the electrode resistance and the protective conductor connecting it to the exposed-conductive-parts. In practice the conductor is a fraction of an ohm and the electrode is the number that matters.
Use the RCD that provides fault protection for the part of the installation concerned. With a 100 mA time-delayed device at the origin and 30 mA devices on the final circuits, the distribution side is designed to the 100 mA device (500 Ω) and the 200 Ω stability note governs anyway. With a 300 mA origin device, which is common on farms and larger TT installations, 167 Ω becomes a hard limit and 200 Ω is a fail.
The 200 Ω figure is often quoted as "GN3 guidance", but it sits in BS 7671 itself. Note 2 to Table 41.5 says the resistance of the installation earth electrode should be as low as practicable and that a value exceeding 200 Ω may not be stable, and it points to Regulation 542.2.4, which requires the type and embedded depth of the electrode to be such that soil drying and freezing will not push its resistance above the required value. A rod reading 180 Ω in a wet March can read 400 Ω in a dry August. A reading over about 100 Ω is worth improving before you sign anything.
Disconnection time is rarely the problem. Table 41.1 gives 0.2 s for TT final circuits up to 63 A with socket-outlets, or 32 A supplying fixed equipment, and Regulation 411.3.2.4 allows 1 s for distribution circuits. At 200 Ω the fault current is about 1.15 A, which is 38 times the rating of a 30 mA RCD and over 11 times that of a 100 mA device, so both operate in their fastest region.
What reading should you expect from a rod?
Electrode resistance depends on the soil far more than on the rod. BS 7430, the code of practice for protective earthing, gives the resistance of a vertical rod as R = ρ/(2πL) × [ln(8L/d) − 1], where ρ is soil resistivity in Ω·m, L the driven length and d the rod diameter. Worked through for a standard 5/8 inch (about 14 mm) copper-bonded rod:
| Soil resistivity (indicative ground) | 1.2 m rod | 2.4 m rod | 3.6 m rod |
|---|---|---|---|
| 25 Ω·m (wet clay or loam) | 18 Ω | 10 Ω | 7 Ω |
| 100 Ω·m (ordinary mixed soil) | 73 Ω | 41 Ω | 29 Ω |
| 500 Ω·m (dry sand, gravel or stony ground) | 366 Ω | 206 Ω | 146 Ω |
These are calculated values for uniform soil, so treat them as a sense-check, not a prediction. Three things follow from the table. Doubling the depth cuts the resistance by 40 to 45 per cent, and usually by more in practice because deeper soil is wetter. A single 1.2 m rod is marginal anywhere the ground drains freely. And a reading that looks too good to be true, 5 Ω on one short rod, usually means the electrode is not on its own: something bonded is still in parallel with it.
How do you test an earth electrode?
Regulation 643.7.2 requires the electrode resistance to earth to be measured where the earthing system incorporates an electrode. GN3 describes two practical methods for a single TT electrode.
| Method | Instrument | What it measures | Best used for |
|---|---|---|---|
| Fall of potential | Three- or four-terminal earth electrode instrument (BS EN 61557-5) with two auxiliary spikes | The electrode alone, independent of the supply | New work, EICs, anything borderline, electrodes for generators or with no supply yet |
| Loop impedance at the origin | Earth fault loop impedance instrument (BS EN 61557-3) | The whole loop: electrode, ground path, the distributor's electrode and the line conductor | RCD-protected installations, periodic inspection, sites with no room for spikes |
Whichever you use, the installation must be isolated first. The electrode has to be disconnected to be measured on its own, and disconnecting the earthing conductor on a live TT installation removes the only earth it has.
Fall-of-potential method step by step
- Isolate the installation at the main switch and prove dead. Disconnect the earthing conductor from the electrode, or at the main earthing terminal.
- Connect the instrument's E terminal (C1 and P1 linked on a four-terminal set) to the electrode with the short lead.
- Drive the current spike (C2) into the ground in a straight line away from the electrode. GN3 expects reliable results where it is at least ten times the largest dimension of the electrode away: about 12 m for a 1.2 m rod, 24 m for a 2.4 m rod, 30 m for a 3 m rod.
- Drive the potential spike (P2) midway between the two, in line. Take a reading.
- Move the potential spike 10 per cent of the electrode-to-C2 distance towards the electrode and read again, then the same distance the other side of the midpoint and read again.
- Average the three. GN3's check is that the largest deviation from the average, as a percentage and multiplied by 1.2, should not exceed 5 per cent. If it does, the spikes are inside the electrode's resistance area: move C2 further out and repeat.
- Reconnect the earthing conductor, check the termination is tight, then re-energise. Record the value, the method and the ground conditions.
Instrument makers often quote the "62 per cent rule" instead, with the potential spike at 61.8 per cent of the distance to C2 and the current spike 30 to 50 m out. It is the same physics. What matters is a flat part of the curve, where moving the potential spike barely changes the reading. Buried metal pipes or cables running parallel to the test line distort the result, so run the leads at right angles to known services where you can, and wet the ground around the auxiliary spikes if the instrument flags high spike resistance.
Loop impedance method step by step
- Isolate the installation at the main switch. Disconnect the earthing conductor from the main earthing terminal so that bonded pipework is no longer in parallel with the electrode.
- Connect the instrument between the incoming line on the supply side of the main switch and the disconnected earthing conductor. Those terminals are live: use fused, GS 38 leads.
- Take the reading. If an RCD sits upstream of your connection point, use the no-trip range.
- Record the result as the electrode resistance and as Ze, noting the method.
- Reconnect the earthing conductor, check the termination, then re-energise.
The reading includes the distributor's electrode at the transformer and the line conductor back to it, typically no more than 20 Ω or so between them, so it is slightly higher than RA. GN3 accepts the method where fault protection is by RCD. The Ze and Zs procedures are covered in full at Earth Fault Loop Impedance Testing: Ze, Zs and Ipf Explained and the meaning of Ze at What is Ze? External earth fault loop impedance.
A Zs reading taken at a socket on a final circuit is not an electrode test. It will pass 1667 Ω comfortably and tell you nothing, because the bonded gas and water pipework is in parallel with the rod. Clamp-on "stakeless" instruments have the same weakness in reverse: they need a parallel earth path to work, so they suit sites with several interconnected electrodes and do not suit a single TT rod.
Worked example
A detached cottage on an overhead supply has a single 2.4 m rod, a 100 mA time-delayed RCD at the origin and 30 mA RCBOs on every final circuit. With the installation isolated and the earthing conductor off the main earthing terminal, the loop reading at the origin is 84 Ω.
- Origin RCD: 0.1 A × 84 Ω = 8.4 V, well under 50 V.
- Final circuits: 0.03 A × 84 Ω = 2.5 V.
- Earth fault current: 230 V ÷ 84 Ω = 2.7 A, which is 27 times the 100 mA rating and 91 times the 30 mA rating.
- 84 Ω is under 200 Ω with margin for a dry summer.
The electrode is satisfactory. Zs on each circuit will be about 84 Ω plus that circuit's R1+R2, and that is what goes in the schedule; the layout is at How to Fill In a Schedule of Test Results, Column by Column. Then test every RCD, as described at RCD Testing Explained: Types, Test Currents and Trip Times.
Why is my electrode reading high, and how do I bring it down?
| Symptom | Likely cause | What to do |
|---|---|---|
| Over 200 Ω on a single short rod | High-resistivity or free-draining soil | Couple on another 1.2 m section and re-test; depth is the cheapest fix |
| Rod will not drive past about 1 m | Rock or rubble | Parallel rods spaced at least their driven depth apart, or bare copper tape or conductor in a trench at 600 mm or deeper |
| Reading much higher than on the last certificate | Dry or frozen ground, or a corroded termination | Inspect and remake the clamp first, then re-test; if it is the soil, go deeper |
| Two rods in parallel give little improvement | Rods too close, so their resistance areas overlap | Space them at least one rod length apart; expect roughly 60 per cent of a single rod, not 50 |
| Fall-of-potential readings do not agree | Spikes inside the resistance area, or buried metal along the test line | Move C2 further out, or turn the test line through 90 degrees |
| Reading is implausibly low | Bonding or SWA armour still in parallel | Check the earthing conductor really is disconnected from everything else |
The IET's guidance is that an electrode should reach soil that stays damp, with buried conductors usually at least 600 mm down so they sit below the layer that dries out and freezes. Chemical treatment of the soil gives a temporary improvement only. For the wider picture of high loop readings on any system, see High Zs Readings: Causes, Checks and How to Fix Them.
What does BS 7671 require of the electrode itself?
| Item | Requirement | Reference |
|---|---|---|
| Recognised types | Earth rods or pipes, earth tapes or wires, earth plates, underground structural metalwork in foundations, welded reinforcement of concrete, lead sheaths and other metal cable coverings (with conditions), other suitable underground metalwork | 542.2.2 |
| Not permitted | Metallic pipes for gas or flammable liquids; the water utility's supply pipe | 542.2.6 |
| Corrosion and strength | Type, material and dimensions chosen to withstand corrosion and to have adequate mechanical strength | 542.2.3 |
| Soil drying and freezing | Type and embedded depth such that drying and freezing will not raise the resistance above the required value | 542.2.4 |
| Buried earthing conductor | 16 mm² copper if sheathed against corrosion but not mechanically protected; 25 mm² copper if bare; where protected against both, sized as a protective conductor under 543.1 | Table 54.1 |
| Connection | Soundly made, electrically and mechanically satisfactory, and labelled | 542.3.2 |
| Label | "Safety Electrical Connection – Do Not Remove", durable, at the electrode connection | 514.13.1 |
With an earth fault current of two or three amps the adiabatic calculation is satisfied by almost anything, so on TT it is the mechanical and corrosion minimums in Table 54.1 that set the earthing conductor size. Main protective bonding still applies exactly as on any other system; see What is main protective bonding? Explained.
Put the rod top in a lidded inspection pit so the connection can be inspected and disconnected for the next test. A rod buried under a patio is the single most common reason an inspector cannot verify a TT earth.
How should RCDs be arranged on a TT installation?
Because RCDs are the fault protection, their arrangement is a design decision, not an afterthought.
- Origin device. A 100 mA or 300 mA time-delayed (Type S) RCD protects the tails, the main switch and any distribution circuits. A Type S device must trip between 130 and 500 ms at its rated current and within 150 ms at five times it, inside the 1 s allowed for distribution circuits.
- Final circuits. 30 mA RCBOs on every circuit give additional protection and fault protection together, and a fault on one circuit trips only that circuit. The trade-offs against a dual-RCD board are at RCBO vs RCD Split-Load Boards: Which to Fit and Why.
- Selectivity. Regulation 536.4.1.4 asks for the upstream device to be the time-delayed type with a rated residual current at least three times that of the downstream device: 100 mA Type S over 30 mA meets it.
- Switch both poles where you can. A neutral-to-earth fault downstream of a single-pole RCBO stays connected after the device has tripped and can hold the upstream RCD out. Double-pole RCBOs avoid the problem.
- Metal consumer units. A fault from the tails to a metal enclosure upstream of every RCD has nothing to disconnect it, because a couple of amps will never blow the distributor's fuse. Either keep the tails double-insulated right up to the main switch with a proper tail gland, or fit the time-delayed RCD in its own insulated enclosure ahead of the board. This comes up on every board change; see Consumer Unit Replacement: Which Certificate and What to Check.
RCD types and what each detects are at What is an RCD? Types AC, A, F, B and RCBOs explained.
How do you record the electrode on an EIC or EICR?
On the model forms in Appendix 6 of BS 7671 the electrode appears in three places. Under supply characteristics and earthing arrangements, tick TT. Under particulars of the installation, tick "installation earth electrode" as the means of earthing and complete the type (rod, tape and so on), the location and the measured resistance to earth. Ze goes in as measured, and on TT it is the same loop reading if you used the loop impedance method. Note the method used, because 84 Ω by fall of potential and 84 Ω by loop impedance are not quite the same statement.
Record the origin RCD's rated residual current, time delay and measured operating time against the main switch details. In the schedule, the maximum permitted Zs for each circuit on TT is the Table 41.5 value for its RCD, with the 200 Ω stability note in mind, and the measured Zs will be dominated by the electrode. The certificate itself is explained at Electrical Installation Certificate (EIC) Explained.
How do you code TT earthing problems on an EICR?
| Finding | Typical classification | Why |
|---|---|---|
| No electrode, or the earthing conductor disconnected or broken | C2 | Best Practice Guide 4: absence of a reliable and effective means of earthing |
| Water utility pipe, gas pipe or oil pipe used as the means of earthing | C2 | Best Practice Guide 4; prohibited by 542.2.6 |
| No RCD, and Zs too high for the overcurrent devices to disconnect in time | C2 | Fault protection is not effective |
| RA × IΔn above 50 V, for example 190 Ω with a 300 mA origin device | C2 | 411.5.3 is not met |
| Meets the 50 V rule but above 200 Ω | C3, with a recommendation to improve | Meets the regulation today but may not be stable (Table 41.5, Note 2) |
| Clamp corroded or loose, reading high or unstable | C2 | The earth is not reliable |
| Electrode cannot be found or reached, loop reading at the origin satisfactory | Record as a limitation or C3 | Connection cannot be inspected; the earth itself is proven |
| "Safety Electrical Connection" label missing | Note only | Best Practice Guide 4 does not ask for a code |
The middle rows are the inspector's judgement and ours, not a quotation from the guide; write down your reasoning. Under the Amendment 4 model forms FI becomes advisory and no longer makes the report unsatisfactory on its own, which matters when you are tempted to use FI for "could not find the rod". The codes are explained at EICR Observation Codes Explained: C1, C2, C3 and FI and the wording of observations at How to Write EICR Observations That Stand Up.
When is TT the right choice?
TT is unavoidable where the distributor provides no earth terminal. It is chosen deliberately where a PME earth should not be exported, because under an open-circuit PEN conductor the PME earth can rise towards line voltage and every bonded metal part rises with it.
| Situation | Position under BS 7671 | More detail |
|---|---|---|
| EV charging points | 722.411.4.1: a PME earth may be used only if one of the listed conditions is met, such as an open-PEN protective device; otherwise TT the charge point | EV Charger Installation Certificate: Section 722 and the EIC |
| Caravan pitches and marina berths | Sections 708 and 709: the PME earth must not be connected to the caravan or boat; a local electrode is the norm | Caravan Parks and Marinas: Sections 708 and 709 Explained |
| Outbuildings, garden supplies | Not prohibited, but a TT island avoids exporting PME potential to metalwork in contact with the ground | Outbuilding and Garden Supply: SWA, TT or PME Earth? |
| Hot tubs and pools | Section 702; many installers TT an outdoor hot tub rather than export the PME earth to it | Hot Tub and Swimming Pool Wiring: Section 702 Explained |
| Farms | Section 705; TT with 300 mA fire-protection RCDs is common, so 167 Ω is the limit | Agricultural Installations: Section 705 Explained |
| Construction sites | Section 704 restricts PME for the site installation | Temporary Supplies on Construction Sites: Section 704 Explained |
Where part of an installation is TT and part TN, keep the two earths apart. Regulation 411.3.1.1 requires simultaneously accessible exposed-conductive-parts to be connected to the same earthing system, so the TT island's metalwork must not be within reach of anything on the PME earth, and an SWA armour earthed at the house end must be insulated from the TT earth at the far end. Keep the electrode clear of buried PME-bonded metalwork as well. Record the arrangement plainly on the certificate; the next inspector will thank you.
Has Amendment 4 changed anything for TT earthing?
BS 7671:2018+A4:2026 was published on 15 April 2026 and Amendment 2 is withdrawn on 15 October 2026. SELECT's published summary of the changes lists none to Regulation 411.5, Table 41.5, Regulation 542.2 or the electrode test in 643.7, so the 50 V rule, the 200 Ω note and the test methods above carry over. The changes it does list near this subject are a new protective neutral bonding (PNB) arrangement recognised under 312.2.1.1, a 25 V limit in place of 50 V for the TT RCD condition in medical locations (Section 710), and minor changes to the earthing particulars on the model forms. Check the standard itself before relying on any summary, this one included.
How Certio helps
Certio produces the EICR, EIC, Minor Works certificate and PAT records on the BS 7671 model forms and turns them into branded PDFs. Photograph the consumer unit and the circuit schedule drafts itself; point the phone camera at a Megger MFT-X1 and the reading goes straight into the schedule of test results, or dictate it while your hands stay on the probes. Suggested observation wording with the regulation reference is there for you to check and confirm. The judgement stays yours; the typing goes. Starter is free for 7 days, then £4.99 a month.