Three-phase testing is single-phase testing done line by line, plus a phase rotation check that Regulation 643.9 of BS 7671 makes compulsory. Line-to-earth voltage (Uo) is still 230 V on a 400/230 V supply, so the maximum Zs values in Table 41.3 don't change: a 32 A Type B is 1.37 Ω whether it's on a single-phase board or one leg of a TP&N board. What does change is the number of insulation resistance pairs, a prospective fault current that can be roughly double the single-phase figure, 400 V between lines that will punish a slipped probe, and a neutral that carries real current. Here's how to work through it properly.
- Uo is 230 V line to earth, so Table 41.3 maximum Zs values are identical to single-phase; measure each line and record the highest.
- Regulation 643.9 requires phase sequence to be verified on multiphase circuits, at the origin and at every three-phase outlet.
- Insulation resistance has ten conductor pairs with a neutral; GN3 accepts six line/neutral tests plus one linked-lives-to-earth test, at 500 V DC, 1 MΩ minimum.
- Three-phase PFC is roughly the line-to-line reading ÷ 0.87, or about twice the line-to-neutral value as the On-Site Guide shortcut.
- RCDs are verified at IΔn since Amendment 2; test from each line and record the longest time.
- Pre-2006 work uses red, yellow, blue and black; mixed colours need the 514.14.1 notice.
What changes on a three-phase supply?
Three lines, L1, L2 and L3, sit at 230 V to neutral and 400 V to each other, 120° apart. Single-phase loads go line to neutral and are (ideally) spread across the three; true three-phase loads such as motors, large heat pumps, some EV chargers and commercial kitchen kit connect across all three lines. For the person holding the probes, four things matter: correct rotation, more tests, higher fault levels, and a neutral that carries the imbalance.
Domestic three-phase supplies used to be rare. They aren't any more, with larger heat pumps and 22 kW chargers pushing customers to ask the DNO for a three-phase upgrade, so you'll meet TP&N boards in houses as well as workshops.
| Quantity | Single-phase domestic | Three-phase 400/230 V |
|---|---|---|
| Line to neutral / earth voltage (Uo) | 230 V | 230 V |
| Line to line voltage | n/a | 400 V |
| Max Zs, 32 A Type B (Table 41.3) | 1.37 Ω | 1.37 Ω (each line) |
| Max Zs, 63 A Type C (Table 41.3) | 0.35 Ω | 0.35 Ω (each line) |
| IR test voltage (Table 64) | 500 V DC | 500 V DC (circuits up to 500 V) |
| IR minimum (Table 64) | 1 MΩ | 1 MΩ |
| Prospective fault current | L-N and L-E | L-N, L-E and L-L, three-phase value is highest |
| Phase sequence check | n/a | Required, Regulation 643.9 |
The 400 V circuit is still tested at 500 V DC because Table 64 groups everything up to and including 500 V nominal together; you only step up to 1,000 V DC above 500 V.
What should I check before any testing?
Start with a live voltage check at the origin, because a three-phase board will tell you a lot before you've disconnected anything. Every reading should sit within the UK supply tolerance of 230 V +10%/−6% line to neutral, which is roughly 216 V to 253 V.
| Measurement | Expect | If not, suspect |
|---|---|---|
| L1-N, L2-N, L3-N | ~230 V each | One low or zero: lost line, blown DNO fuse |
| L1-L2, L2-L3, L3-L1 | ~400 V each | ~230 V between two "lines": both on the same phase |
| N-E | Close to 0 V | Several volts: high neutral current or a poor neutral/PEN |
| Line voltages very unequal under load | Within a few volts | Severe imbalance or a deteriorating neutral |
A lost DNO fuse or a supply problem is the DNO's job, not yours. A rising N-E voltage on a TN-C-S supply deserves attention too: a deteriorating PEN on a three-phase supply can push single-phase loads well above 230 V on the lightly loaded lines, which is how appliances get cooked when a neutral fails.
How do I test phase sequence (Regulation 643.9)?
Regulation 643.9 is short: for multiphase circuits, verify that the phase sequence is maintained. Use a phase rotation instrument to BS EN 61557-7 (a dedicated three-lead indicator, or the rotation function on an instrument that's specified for it). Connect L1, L2 and L3 in order at the origin and you should see clockwise or "1-2-3". Then repeat at each distribution board, at every three-phase socket-outlet, and at motor isolators.
Why it matters on site: a pump, compressor, extract fan or lift motor that runs backwards ranges from noisy to genuinely damaging. Reverse rotation at an outlet means two lines are crossed somewhere between the origin and that point, which also tells you the identification is wrong somewhere in the run. Find it and correct it rather than swapping two cores at the plug to "make it work".
Record the result on the certificate. Note the number of phases and wires under supply characteristics (for example three-phase, four-wire) and confirm phase sequence in the circuit results or remarks so the next inspector knows it was checked.
How do I do insulation resistance on a three-phase circuit?
Same principle as single-phase: isolate, prove dead, then remove or disconnect anything that won't survive 500 V DC. On three-phase boards that list is longer. Variable speed drives, soft starters, electronic controllers, SPDs and some contactor coils will all object. A 500 V test pushed into a drive's input can destroy it in an instant, so disconnect at the drive input or leave that circuit's IR test out and record the limitation.
With a neutral there are ten conductor pairs:
| Group | Pairs | Count |
|---|---|---|
| Line to line | L1-L2, L2-L3, L3-L1 | 3 |
| Line to neutral | L1-N, L2-N, L3-N | 3 |
| To earth | L1-E, L2-E, L3-E, N-E | 4 |
Guidance Note 3 accepts linking all live conductors (lines and neutral) and testing them together to earth, which turns the last four into one test. On a big board that's what keeps the job sane: six tests between live conductors, then one linked test to earth per circuit or group of circuits.
Table 64 sets the minimum at 1 MΩ, but a healthy modern circuit should read hundreds of megohms or your instrument's maximum. GN3 suggests treating anything under 2 MΩ as a sign of a possible latent defect worth investigating. Motor windings will always read lower than the cabling, so disconnect the motor at its isolator or contactor and test the wiring on its own; test the machine separately if you've been asked to.
For low readings, see low insulation resistance troubleshooting, and for the general method insulation resistance testing.
How do I measure Zs on each line?
An earth fault on a three-phase circuit is still one line to earth at Uo of 230 V, so the disconnection time and maximum Zs are exactly those for a single-phase circuit with the same device. The difference is you now have three lines to prove.
Measure Zs from each line to earth at the furthest point of the circuit and record the highest of the three. Don't average: a spread between the lines is the useful information. If L1 and L3 read 0.42 Ω and L2 reads 0.71 Ω on the same cable run, that's a loose or burnt termination on L2 waiting to be found.
Compare cold measured values with roughly 80% of the Table 41.3 figure to allow for conductor temperature rise in service, so a 32 A Type B circuit should measure about 1.10 Ω or less against the 1.37 Ω tabulated value. Full tables are in maximum Zs values, and if a line reads high work through high Zs troubleshooting.
At the origin, measure Ze from each line to earth with the main earthing conductor disconnected (and the installation isolated), then reconnect before you do anything else. Record the highest of the three as Ze.
How do I measure and record three-phase PFC?
This is the bit most often done wrong. Regulation 434.5.1 says every device must have a breaking capacity at least equal to the maximum prospective fault current where it's installed, and on a three-phase board the worst case isn't the line-to-neutral figure you'd record on a house.
Measure at the origin, with the earthing and bonding connected:
- line to neutral on each line (prospective short-circuit current, PSCC)
- line to earth on each line (prospective earth fault current, PEFC)
- line to line on each pair
Then work out the three-phase symmetrical fault value, which is the one that matters for breaking capacity. Two accepted methods:
| Method | Calculation | Comment |
|---|---|---|
| From line-to-line reading | Highest L-L reading ÷ 0.87 | More accurate; uses a measured L-L loop |
| On-Site Guide shortcut | Highest L-N reading × 2 | Quick, errs on the high side because it assumes the neutral impedance equals the line |
Worked example: a TP&N board reads 2.4 kA line to neutral (highest of three) and 4.0 kA line to line. The line-to-line method gives 4.0 ÷ 0.87 ≈ 4.6 kA. The doubling shortcut gives 4.8 kA. Either way, 6 kA-rated MCBs on that board are fine. Had the readings come out at 3.6 kA L-N and 6.2 kA L-L, the three-phase value would be over 7 kA and those 6 kA devices would need checking for back-up protection from the upstream device, confirmed by the manufacturer's data. Without it, inadequate breaking capacity is a potentially dangerous condition.
Record the highest value as Ipf on the certificate. On the current model forms the note against Ipf allows it to be determined by enquiry or by measurement, so if the instrument won't read reliably at low impedance (high fault levels give tiny loop readings, where instrument resolution matters), use the DNO's declared figure and say so. More detail is in prospective fault current explained.
How do I test RCDs on a three-phase board?
Most testers derive the test current from line to neutral or line to earth, so a three- or four-pole RCD or RCBO is tested from each line in turn. Test at the rated residual current (IΔn) and record the longest time.
Since Amendment 2, Regulation 643.8 verifies RCD effectiveness at IΔn. The usual limits are 300 ms for a general-type RCD to BS EN 61008 or BS EN 61009, and 130 to 500 ms for a time-delayed S type. The old routine 5× IΔn test is no longer the verification test, so don't treat a 5× time as the pass criterion.
Where a four-pole RCD protects a three-phase load with no neutral connected downstream (a motor, for example), you may not be able to test line to neutral on the load side. Use the instrument's line-to-earth mode or test at the RCD outgoing terminals, following the instrument maker's instructions. Then press the test button as the last check, not the first. See RCD testing explained.
Why does the neutral matter more on three-phase?
On a perfectly balanced linear load the neutral carries nothing, which is why it gets ignored. Real boards are never balanced. A board with lighting on L1, a kitchen on L2 and offices on L3 has neutral current equal to the vector imbalance, and third-harmonic currents from LED drivers, IT power supplies and drives add together in the neutral instead of cancelling. Appendix 4 of BS 7671 covers derating for triple-N harmonics, and on a heavily loaded commercial board the neutral can carry more than any of the lines.
What to check: neutral terminations as carefully as the lines (torque and discolouration), neutral csa against the lines on harmonic-heavy loads, and never disconnect a shared or common neutral with any of the lines still loaded. Borrowed and shared neutrals are their own problem; see borrowed neutrals.
Which conductor colours will I find?
Harmonised colours were introduced from 31 March 2004 and became mandatory for new work from 31 March 2006. That gives you two schemes in the field:
| Conductor | Pre-2004/06 | Current |
|---|---|---|
| L1 | Red | Brown |
| L2 | Yellow | Black |
| L3 | Blue | Grey |
| Neutral | Black | Blue |
| Protective | Green/yellow | Green/yellow |
Note the trap: black is the neutral in old work and L2 in new; blue is L3 in old work and the neutral in new. Regulation 514.14.1 requires a notice at the distribution board where an installation contains both schemes. Never trust colour alone at an interface: prove which conductor is which with your tester. See polarity testing explained.
How do typical three-phase findings get coded on an EICR?
Coding always depends on what you find on the day. As a starting point, consistent with Electrical Safety First's Best Practice Guide 4:
| Finding | Typical code |
|---|---|
| Exposed live parts at a three-phase board (missing blanks, cover removed) | C1 |
| Zs on any line exceeds the limit and no RCD meets Table 41.5 | C2 |
| Breaking capacity below the three-phase PFC with no confirmed back-up | C2 |
| Reverse rotation at a three-phase outlet serving a motor | C2 |
| Mixed old/new colours with no 514.14.1 notice | C3 |
| Neutral termination discoloured, cause not established in the time available | FI |
Remember that only C1 and C2 make the report unsatisfactory; FI is a recommendation to investigate. For wording, see how to write EICR observations and the C1, C2, C3 and FI guide.
What about safety at 400 V?
A dropped probe across two bus bars at 400 V, with several kiloamps behind it, is an arc flash. Use GS38-compliant fused leads with shrouded, finger-guarded probes and no more than 4 mm of exposed tip, with the instrument and leads rated CAT III 600 V or better. Prove your voltage indicator on a proving unit before and after each test for dead. Keep covers on for live tests where the board allows it, and plan your probe route so you're never reaching across bars.
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 £4.99 a month, charged when you sign up.