Interpreting ring final circuit results on a periodic inspection
On a periodic inspection of a 1980s three-bed semi you isolate the ground-floor ring final: 32 A Type B BS EN 60898 MCB, 2.5 mm² line and neutral, 1.5 mm² cpc, protected by a 30 mA RCBO. End-to-end readings are r1 = 0.40 Ω, rn = 0.41 Ω, r2 = 0.68 Ω. You then cross-connect line and cpc and take a reading at every socket outlet: 0.27, 0.27, 0.26, 0.53, 0.28, 0.27, 0.26 Ω. The 0.53 Ω is at a double socket in a conservatory added in 2015. Insulation resistance at 500 V is 0.6 MΩ line-to-neutral and >299 MΩ line-to-cpc and neutral-to-cpc.
1. Evaluate the end-to-end and cross-connected readings. Is the ring intact, and what does the 0.53 Ω socket tell you?
Read the worked answer
End-to-end
- r2/r1 = 0.68/0.40 = 1.70, against the expected 1.67 for 1.5 mm² cpc with 2.5 mm² live conductors. The cpc is correctly sized and continuous.
- rn ≈ r1 (0.41 vs 0.40), so line and neutral are both continuous and of similar length.
- From r1, ring length ≈ 0.40 / 0.00741 Ω/m ≈ 54 m - plausible for this property.
Cross-connected
- Expected (r1 + r2)/4 = (0.40 + 0.68)/4 = 0.27 Ω. Six sockets read 0.26–0.28 Ω, so the ring is intact and the readings are consistent around it.
- The 0.53 Ω socket is 0.26 Ω above the ring value. At roughly 19.5 mΩ/m for 2.5/1.5 mm² that is about 13 m of extra cable - consistent with a spur of the right length to the conservatory.
Judgement This is not automatically a defect. One twin socket on an unfused spur is permitted. What you must do is confirm visually that it is a single twin socket and not two, that the spur is 2.5 mm², and that the socket is not feeding further outlets - a second socket on the same unfused spur is a non-compliance and would need a fused connection unit or inclusion in the ring.
2. The 0.6 MΩ line-to-neutral insulation resistance reading: what do you do before you code it?
Read the worked answer
0.6 MΩ is below the 1 MΩ minimum for a 500 V test, so as it stands the circuit is unsatisfactory - but the reading is not yet attributable to the wiring.
Before coding, eliminate connected load
- 500 V applied to connected electronics (boiler, extractor, TV/phone amplifier, neons in FCUs, smart switches, dimmers) risks damaging them and gives false readings.
- Isolate the circuit and disconnect the accessories/appliances on it, or test in sections, then retest the wiring alone.
If it clears
- Record the condition of test honestly, e.g. '0.6 MΩ L–N with accessories connected; >200 MΩ L–N with accessories disconnected; no defect in the fixed wiring.' No code for the wiring, but note the connected load as the cause.
If it does not clear
- Split the ring at the mid-points to localise, and check for a shared/borrowed neutral from an adjacent circuit, moisture, or a damaged cable under a floor or behind a kitchen unit.
- A genuine 0.6 MΩ L–N in the fixed wiring is a fire risk and is coded C2 - deterioration between live conductors can break down under load.
The mistake to avoid is coding from the first number without establishing what produced it.
3. The client says 'anything over 1 Ω on those socket readings is a fail - my last electrician condemned it'. How do you answer?
Read the worked answer
- The 0.53 Ω figure is a resistance measurement of line plus cpc around the ring, not an earth fault loop impedance, and 1 Ω is not a threshold anywhere in BS 7671.
- The relevant comparison for this circuit is Zs against the maximum for a 32 A Type B device: 1.44 Ω from the table, derated by 0.8 for measurement at ambient to give about 1.15 Ω as the maximum acceptable measured value.
- With Ze of roughly 0.35 Ω and R1 + R2 at that spur of 0.53 Ω, the expected Zs is around 0.88 Ω - comfortably compliant. Confirm it by measuring Zs at the conservatory socket rather than calculating.
- The correct conclusion: the spur socket is compliant, and nothing is coded against it. A previous report that 'condemned' it on a 1 Ω rule of thumb was wrong, and it is worth saying so plainly, because the client has been given bad information before.
The judgement being tested is refusing to code to a number that has no basis in the standard, in either direction.