Ring Final Circuits Explained: Design, Testing and Common Faults

Quick Answer: A ring final circuit is a UK socket-outlet arrangement where the cable leaves the consumer unit, loops around the sockets and returns to the same protective device, forming a ring. Typically wired in 2.5mm² twin-and-earth and protected by a 32A overcurrent device, BS 7671 Appendix 15 permits an unlimited number of BS 1363 socket-outlets serving a floor area up to 100m². Correct testing requires the three-step ring continuity check (end-to-end r1, rn, r2, then cross-connected readings), insulation resistance and polarity verification.

Summary

The ring final is a peculiarly British solution, born of post-war copper shortages. By feeding sockets from both ends of a ring, the current in any one leg is roughly halved, so 2.5mm² conductors can serve a 32A circuit that would otherwise need a heavier radial cable. Both legs of the ring connect to the same 32A protective device at the board and to both terminals of each socket, so every accessory is fed from two directions. That is the whole idea — and also why testing it correctly matters so much, because a break in the ring is invisible in normal use until the surviving leg overheats.

BS 7671 Appendix 15 sets out the standard arrangement: 2.5mm² copper twin-and-earth (with a 1.5mm² CPC) on a 32A device, serving an unlimited number of socket-outlets over a floor area not exceeding 100m². Spurs are permitted — one unfused spur may be taken from the ring to feed a single twin socket or one fused connection unit, and fused spurs (via an FCU) can feed multiple points behind a 13A fuse. The "100m²" is a floor-area guide, not a cable-length limit; for longer rings you still verify voltage drop and that disconnection times are met.

The two faults that catch people out are the broken ring and the borrowed neutral. A broken ring (where the loop is opened somewhere, often because someone wired it as two radials joined in the middle — a "figure-of-eight" or interconnected ring) leaves part of the circuit fed by a single 2.5mm² leg carrying up to 32A, which is a fire risk. A borrowed neutral happens when a neutral from one circuit is used to complete another — it defeats RCD protection and is dangerous. Both are found by methodical continuity testing, which is why the three-step procedure below is non-negotiable.

Key Facts

• 2.5mm² twin-and-earth — the standard ring conductor size (1.5mm² CPC in standard T&E), protected by a 32A device.
• 32A — the standard overcurrent rating for a ring final (BS EN 60898 MCB, BS EN 61009 RCBO, or BS 88 fuse).
• 100m² — maximum floor area served by one ring final per Appendix 15 guidance.
• Unlimited socket-outlets — no fixed maximum number of BS 1363 sockets on a ring within the 100m² area.
• Both legs to the device — the two ends of the ring connect to the same line, neutral and earth terminals at the board.
• Unfused spur — one non-fused spur may feed a single OR twin socket, or one FCU; taken from a socket terminal or junction box on the ring.
• Fused spur — via a fused connection unit (FCU) with a BS 1362 fuse rated ≤13A; can supply multiple outlets downstream of the fuse.
• Don't spur off a spur — a non-fused spur must come from the ring itself, not from another spur.
• r1 — end-to-end resistance of the line conductor (loop).
• rn — end-to-end resistance of the neutral conductor (loop).
• r2 — end-to-end resistance of the CPC (loop); higher than r1/rn because the CPC is usually 1.5mm² in 2.5mm² T&E.
• (r1 + r2)/4 — the expected resistance at each socket measured line-to-CPC when the ring is cross-connected and healthy.
• Insulation resistance — minimum 1.0 MΩ at 500V DC for a 230V circuit (test value; 2 MΩ is a common "investigate" threshold).
• Common faults — broken ring, interconnected/figure-of-eight ring, spur wired as part of the ring, borrowed neutrals, reversed polarity at an accessory.
• Voltage drop — limit of 5% for socket circuits (≈11.5V on 230V); long rings must be checked even within 100m².

Quick Reference Table

Detailed Guidance

Why a ring and not a radial — design context

A 32A ring in 2.5mm² and a 20A radial in 2.5mm² (or 32A radial in 4mm²) both have a place. Choose a ring where you want many sockets over a larger area with economical cable; choose a radial where the layout is simple, the run is short, or you want to avoid the testing/maintenance burden of a ring. The ring's advantage — shared loading across two legs — is also its liability: it depends entirely on the loop being intact, and a hidden break converts it into an overloaded radial. If you cannot guarantee the ring will be properly tested and maintained, a radial may be the safer specification.

Wiring it correctly

Run the cable from the 32A device, visit each socket connecting line-to-line, neutral-to-neutral, CPC-to-CPC at the accessory terminals (two cables into each terminal at ring sockets), and bring both ends back to the same device. Spurs are taken from a socket terminal or a junction box on the ring — never daisy-chained off another spur. Keep a note of which sockets are on the ring and which are spurs; it makes the continuity test interpretation far easier.

The three-step ring continuity test — procedure

This is the test that proves the ring is a ring. Disconnect/isolate first, and identify both legs at the board.

1. End-to-end (Step 1). Measure each conductor loop on its own: line-to-line (r1), neutral-to-neutral (rn), CPC-to-CPC (r2). For 2.5/1.5 T&E, r1 ≈ rn, and r2 ≈ 1.67 × r1 because the CPC is thinner. Record all three. A very high or open reading means a break in that conductor.
2. Cross-connect line and neutral (Step 2). Join one leg's line to the other leg's neutral, and vice versa, at the board. Measure line-to-neutral at every socket. On a healthy ring every reading should be substantially the same, approximately (r1 + rn)/4. A reading that climbs as you move round, then falls, flags a spur or a break.
3. Cross-connect line and CPC (Step 3). Repeat the cross-connection with line and CPC. Measure line-to-CPC at every socket; each should read about (r1 + r2)/4 and be roughly constant. This value is also your R1+R2 for the furthest point, used to confirm Zs. Sockets that read noticeably higher are spurs (fed from one direction only); a wildly different reading indicates a fault.

Constant, near-equal readings around the ring = healthy ring. Readings that rise toward the middle and fall again = the ring is open somewhere (you are effectively measuring two radials). Use this pattern to locate the break.

Insulation resistance and polarity

After continuity, prove insulation resistance at 500V DC (line-neutral, line-earth, neutral-earth) — expect well above the 1.0 MΩ minimum; anything near it warrants investigation. Confirm polarity at every accessory (line to the correct terminal, single-pole switching in the line) and at the origin. Finally verify Zs (using R1+R2 from Step 3 plus Ze) meets the maximum disconnection time, and test the RCD providing additional protection.

Common faults and how the test reveals them

• Broken ring / figure-of-eight: Step 1 end-to-end reads open or far too high on one conductor; Step 2/3 readings vary around the loop instead of staying constant. Trace the open conductor.
• Spur mistaken for ring: a socket on a spur reads higher than its neighbours in Step 3 because it is fed from one direction only. Confirm against your wiring notes.
• Borrowed neutral: insulation resistance neutral-to-earth or between circuits is abnormal, and the neutral end-to-end may include conductors from another circuit. This defeats RCD operation — find and correct it.
• Reversed polarity: polarity test flags line and neutral swapped at an accessory; correct at the affected socket.

Frequently Asked Questions

How many sockets can I put on a ring final?

There is no fixed maximum number of socket-outlets. BS 7671 Appendix 15 allows an unlimited number of BS 1363 sockets on a 32A ring serving a floor area up to 100m². Beyond that area, or for very long runs, check voltage drop and disconnection times and consider a second circuit.

Is the 100m² a cable-length limit?

No, it is a floor-area guide. You still have to verify voltage drop (5% limit for sockets) and that Zs meets the required disconnection time. A compact 100m² area is fine; a long, thin run may hit voltage drop before area becomes the constraint.

Can I take a spur off a spur?

No. A non-fused spur must be taken from the ring itself (a socket terminal or junction box on the ring), not from another spur. If you need more outlets from one point, take a fused spur via an FCU, which can feed multiple points behind a 13A fuse.

What do the three continuity steps actually prove?

Step 1 proves each conductor loop is continuous and correctly sized (r2 > r1). Steps 2 and 3 prove the ring is genuinely a ring: constant readings at every socket confirm the loop is intact and let you spot spurs and breaks. Step 3 also gives you R1+R2 for the Zs calculation.

My readings rise toward the middle of the ring and fall again — what's wrong?

That is the classic signature of an open ring. You are measuring what is effectively two radials joined at the board, so resistance peaks at the electrical midpoint. Find and repair the break in the conductor identified in Step 1.

Regulations & Standards

• BS 7671:2018+A2:2022 (IET Wiring Regulations, 18th Edition) — Appendix 15 (ring and radial final circuit arrangements), Chapter 43 (overcurrent protection), Section 411 (ADS), Section 612 (testing).
• BS 1363 — 13A plugs, socket-outlets and connection units.
• BS 1362 — cartridge fuses for plugs and FCUs (≤13A).
• IET On-Site Guide — standard final circuit arrangements, spur rules and test procedures.
• IET Guidance Note 3 (Inspection & Testing) — ring continuity, insulation resistance and polarity test methods.