BS 7671 Guide · Section 9

Inspection and testing

Testing is how you prove the installation does what the design says it will. This section walks through Part 6 in the order you actually work on site, from proving dead to signing the schedule.

Initial verification and periodic inspection

Initial verification (Chapter 64) applies to new installations, additions and alterations. You inspect and test while the work goes in and again when it is finished, before anyone uses it. The result is an Electrical Installation Certificate or a Minor Works Certificate.

Periodic inspection and testing (Chapter 65) checks whether an existing installation is still safe to use. You agree extent and limitations with the client first, then report on condition in an Electrical Installation Condition Report (EICR).

A4 tightens the reporting side. Regulation 653.1 now says the EICR must follow the Appendix 6 model and its notes. Regulation 653.2 says the report must include guidance for whoever receives it. A note also allows photos and thermal images to be attached. See certificates and EICR codes for the paperwork in detail.

Regulation references: Chapter 64, Chapter 65, 641.1, 653.1, 653.2

Safe isolation first

Every dead test assumes the circuit really is dead. Get this wrong and nothing else matters. Keep the routine short and identical every time:

1. Identify the circuit and the correct point of isolation.
2. Switch off and isolate with a device suitable for isolation.
3. Lock it off, keep the key, and fix a warning notice.
4. Check your approved voltage indicator on a proving unit.
5. Test every combination at the point of work: L–N, L–E, N–E (and all phases on three-phase).
6. Prove the indicator again on the proving unit.

HSE guidance GS38 sets the standard for test leads and probes. Probe tips should expose no more than 4 mm of metal, and ideally 2 mm or less. Leads need finger guards, sound insulation and suitable fusing or current limiting. Inspect them before every use.

Inspect before you test

Inspection often finds more than the meter does. Check what a test cannot see: device ratings, cable zones, fire barriers, IP ratings, labels, bonding clamps and tight, enclosed connections.

Dead tests

Continuity of protective conductors

Every circuit protective conductor (cpc) and bonding conductor must be continuous with a resistance you have actually measured. Null your leads first.

• R₁+R₂ method: link line to cpc at the board, then measure line to earth at each point on the circuit. The highest reading, usually at the far end, is the R₁+R₂ you record. This also proves polarity at each point you test.
• R₂ (wander lead) method: measure from the main earthing terminal to each exposed-conductive-part with a long lead. Useful for bonding and for metalwork with no line conductor alongside.

Ring final circuits

A ring must be a genuine ring, with no breaks and no interconnections. Test it in three stages:

1. End to end. Measure each conductor loop at the board: line (r₁), neutral (r_n) and cpc (r₂). Line and neutral should agree closely. With 2.5/1.5 mm² twin and earth, r₂ will be about 1.67 times r₁.
2. Line to neutral cross-connected. Join the line of one end to the neutral of the other, and vice versa. At each socket, L–N should read about a quarter of (r₁+r_n), and roughly the same everywhere.
3. Line to cpc cross-connected. Repeat with line and cpc. Each socket should read about a quarter of (r₁+r₂). The highest value is the circuit’s R₁+R₂.

Rising readings at some sockets point to a spur or break; a much lower one suggests an interconnection.

Insulation resistance

Test between live conductors, and between live conductors and earth. Disconnect anything that could skew the result or be damaged, such as SPDs, dimmers, timers and electronic controls. The test voltages and minimums in Table 64 are:

Where equipment such as an SPD cannot reasonably be disconnected, you may test that circuit at 250 V DC, but the reading must still be at least 1 MΩ. Record the test voltage you used and label vulnerable circuits at the board.

Polarity

Check that every fuse and single-pole switch is in the line conductor, that ES lampholder centre contacts are on line, and that sockets are wired correctly. Your R₁+R₂ readings cover most of this. Confirm polarity at the origin again once the supply is on.

Earth electrode resistance

On TT systems, measure the electrode, either with a loop tester at the origin (main earthing conductor disconnected) or a dedicated earth tester. With RCD protection, R_A × IΔn must not exceed 50 V. Guidance treats values above about 200 Ω as possibly unstable through the seasons.

Regulation references: 643.2.1, 643.3.1, 643.3.2, Table 64, 411.5.3, 514.9.1

Live tests

External loop impedance (Z_e)

Measure at the origin with the main switch off and the earthing conductor disconnected from the main earthing terminal, so parallel paths through bonding do not flatter the reading. Compare it with what the network operator declares. Reconnect straight away.

Circuit loop impedance (Z_s)

Measure at the furthest point of each circuit, or calculate it as Z_e + (R₁+R₂). A measurement also proves the earth path is really there. Use a no-trip setting on RCD-protected circuits and a plug-in lead where you can.

Prospective fault current

Measure prospective short-circuit and earth fault current at the origin and record the higher. Protective devices must be able to break it, so check high readings against the board’s rated or conditional short-circuit capacity.

RCDs

Since A2:2022, and unchanged in A4, every RCD is tested with a single AC test at its rated residual operating current, whatever its type. A general non-delay device must trip within 300 ms. A time-delayed S-type must trip between 130 ms and 500 ms. The 5× and ½× tests are no longer required, though ½× can still help with fault-finding.

Then press the test button. It proves the mechanism, but never replaces the instrument test.

AFDDs and SPDs

For AFDDs, operate the test facility as the manufacturer instructs and record the result. For SPDs, check the status indicator shows healthy and record the device details on the schedule. Both have their own fields on the Appendix 6 forms.

Functional checks

Operate switchgear, isolators, interlocks and controls to prove they work. On three-phase supplies, confirm phase sequence.

Regulation references: 643.7.1, 643.7.3, 643.7.3.201, 643.8, 434.5.1, 531.3.3

Comparing measured Z_s with the maximum

The maximums in Tables 41.2 to 41.4 assume conductors at operating temperature, 70 °C for thermoplastic cable. You test cold, when resistance is lower, so a raw comparison can pass a circuit that fails under load.

The quick method: compare your reading with 80% of the table value. If it fails that, do the full calculation before deciding.

Worked example

A ring final on a TN-C-S supply, protected by a 32 A Type B MCB. The device trips magnetically at 5 × 32 = 160 A. Using 230 V and the minimum voltage factor of 0.95, the tabulated maximum Z_s works out at 230 × 0.95 ÷ 160 ≈ 1.37 Ω. Check this against Table 41.3 in your copy.

• Quick check: 80% of 1.37 Ω is about 1.10 Ω. A measured Z_s of 0.87 Ω passes comfortably.
• Borderline case: a second ring measures 1.15 Ω, which fails the quick check. Measured Z_e is 0.35 Ω and R₁+R₂ is 0.80 Ω at about 20 °C.
• Full calculation: raising copper from 20 °C to 70 °C increases resistance by roughly 20%, so corrected R₁+R₂ is 0.80 × 1.20 = 0.96 Ω. Corrected Z_s is 0.35 + 0.96 = 1.31 Ω. That is below 1.37 Ω, so the circuit complies. Note the method on the schedule.

Regulation references: 411.3.2, 411.4, Tables 41.2, 41.3, 41.4, 41.5, Appendix 3

When a result fails

1. Stop. Do not energise or hand over a circuit you know is unsafe.
2. Retest to rule out your own error: leads nulled, correct range, loads disconnected.
3. Find the cause: loose terminations, an unconnected cpc, damaged insulation, moisture, or a wrong device rating.
4. Fix it, then repeat every test the repair could have affected, not just the one that failed.
5. On a periodic inspection, code the defect and tell the person who ordered the report. A danger needing immediate action must be reported straight away, in writing.

Recording results and instrument calibration

Fill in the schedule as you go, not from memory at the van. Record actual values, the insulation test voltage, the source of your maximum Z_s figures, and any circuit you could not test, with the reason.

Use instruments made to BS EN 61557 and record their serial numbers. Keep them in calibration, typically yearly, and check them in between against a check box or known reference. A result from an out-of-calibration instrument is hard to defend.

The practical sequence

1. Agree the scope and, for periodic work, the limitations.
2. Isolate safely and prove dead.
3. Carry out the visual inspection.
4. Continuity of protective and bonding conductors (R₁+R₂ or R₂).
5. Ring final continuity: end to end, then both cross-connection checks.
6. Insulation resistance at the Table 64 voltage.
7. Polarity, dead.
8. Earth electrode resistance on TT systems.
9. Energise. Check polarity and voltage at the origin.
10. Z_e and prospective fault current at the origin.
11. Z_s on each circuit, then corrected comparison.
12. RCD tests at IΔn, then the test button.
13. AFDD and SPD checks, then functional and phase sequence checks.
14. Complete the schedule, certificate or report, and hand it over.