Ring Final Circuit Testing: r1, rn, r2 and R1+R2 Step by Step

A ring final circuit leaves the board, visits every socket and comes back to the same 30 A or 32 A device, so each outlet is fed from two directions. Proving it means the three-step dead test in IET Guidance Note 3: measure r1, rn and r2 end to end at the board, cross-connect line and neutral and read at every socket, then cross-connect line and cpc and read again. On 2.5/1.5 mm² twin and earth, r1 equals rn, r2 is about 1.63 times r1 (12.10 ÷ 7.41 mΩ/m from On-Site Guide Table I1), and the step 3 reading at every socket sits close to (r1+r2)/4. The highest step 3 figure is the circuit's R1+R2 and goes in column 21 of the BS 7671:2018+A4:2026 schedule.

Why does a ring final circuit need a three-step test?

Regulation 433.1.204 of BS 7671:2018+A4:2026 lets a ring wired in 2.5 mm² copper sit behind a 30 A or 32 A device even though clipped-direct 2.5 mm² T&E is only rated at 27 A (Table 4D5, Reference Method C). The relaxation works because the load is shared between two legs; the regulation requires a current-carrying capacity (Iz) of at least 20 A and a load that is unlikely to exceed Iz in any part of the ring for long periods. Break one leg and the sockets beyond the break become a 2.5 mm² radial behind a 32 A breaker. A 3 kW heater and a 3 kW kettle on that section pull about 26 A through cable that may only be rated for 21 A in a ceiling (Method 100) and nothing trips. A broken cpc is quieter still: everything works, the RCD does not care, and the protective conductor to half the sockets is relying on one path or none.

A single continuity reading to the furthest socket misses both faults, because a broken ring still gives continuity by the other leg. Only the three-step method proves every conductor is continuous all the way round, that the legs are the right way round, and that nobody has joined two points of the ring with an extra cable (an interconnection, which makes a parallel section that can carry more than its share).

Regulation 643.2.1 requires the continuity of every conductor of every ring final circuit, including the cpc, to be verified at initial verification, which is why the EIC schedule has dedicated columns for it. GN3 (10th Edition, 2026, for A4) describes the three-step method in its section on continuity of ring final circuit conductors; if you still work from a 2022 first printing of the 9th Edition, note that the IET issued an errata in November 2024 correcting two equations in that section (2.6.6), so check your copy against the reprint. For how the numbers are laid out on the form, see How to Fill In a Schedule of Test Results.

What do I need before I start?

• The circuit isolated, locked off and proved dead, with both legs of the ring disconnected from the breaker, the neutral bar and the earth bar so you have six free conductor ends at the board.
• A continuity instrument to BS EN 61557-4, which means a test current of at least 200 mA into resistances up to 2 Ω. The Megger MFT-X1, for example, pushes more than 200 mA from 0 to 2 Ω and reads down to 0.01 Ω (tester.co.uk listing, £1,319 ex VAT, checked 21 October 2026).
• Leads nulled, or the lead resistance noted and subtracted. A 0.05 Ω lead error is 25 per cent of a typical step 3 reading.
• A way to tell leg A from leg B. Mark them with tape at the board before you touch anything, because steps 2 and 3 depend on pairing the two legs correctly.
• Access to every socket-outlet, fused connection unit and junction box on the circuit. If you cannot get to some of them on a periodic, that is a limitation to record, not something to guess round; see EICR Sampling and Limitations Explained.

Step 1: how do I measure r1, rn and r2?

With the six ends free, measure between the two line ends (r1), the two neutral ends (rn) and the two cpc ends (r2). Each figure is the resistance of that conductor all the way round the ring. An open-circuit reading on any of the three means that conductor is broken somewhere, and there is no point going further until it is found and repaired.

The ratio comes straight from conductor resistance. On-Site Guide Table I1 gives 7.41 mΩ/m for 2.5 mm² copper and 12.10 mΩ/m for 1.5 mm² at 20 °C, so r2 ÷ r1 is 1.63 in theory and 1.67 is the rule of thumb most of us carry round. A 40 m ring therefore gives r1 ≈ 0.30 Ω and r2 ≈ 0.48 Ω. For a 4/1.5 mm² ring the ratio is about 2.6 (12.10 ÷ 4.61); for singles in conduit with a 2.5 mm² cpc it is 1.0.

Two checks before you move on. If r1 and rn differ by more than a few hundredths of an ohm, suspect a loose termination, a section in a different cable size, or two legs that belong to different circuits (a classic in a board that has been "tidied"). If r2 is much higher than 1.67 × r1, look for a loose cpc terminal; if it is much lower, the cpc has a parallel path through metal trunking, conduit or back boxes, which is not a fault but means the r2 figure is not the cable alone.

Step 2: how do I do the line-to-neutral cross-connection test?

Join the line of leg A to the neutral of leg B, and the neutral of leg A to the line of leg B. The two loops now form a figure of eight. Measure between the two joined pairs at the board: you should get (r1+rn)/4, which for equal conductors is r1/2. For the 40 m example, 0.15 Ω.

Now go to every socket on the circuit and measure between line and neutral. Because any point on a cross-connected ring has two paths back to the board whose resistances always add up to the same total, the reading is the same at every socket: roughly (r1+rn)/4, give or take a few hundredths. GN3 calls this "substantially the same", and for r1 = rn it is exactly the same in theory.

What the odd readings mean:

• One socket reads higher by a fixed amount: it is on a spur. The extra is the line plus neutral of the spur cable, so a 5 m 2.5 mm² spur adds about 0.07 Ω (5 × 2 × 7.41 mΩ). Note it; you will see the same socket stand out in step 3.
• Readings climb steadily in one direction round the ring: either you have paired leg A line with leg A neutral (the most common mistake), or one leg is open and you are reading along a radial. Go back to the board and check the pairing before you condemn the cable.
• Readings roughly equal to r1 at most sockets rather than half of it: the conductors are not cross-connected at all, usually because one joint at the board has fallen off.
• A reading noticeably lower at one group of sockets: an interconnection bridging two parts of the ring. A kitchen extension joined into the old ring at two points is the usual cause.

Step 3: how do I do the line-to-cpc test and find R1+R2?

Separate the neutrals, then join the line of leg A to the cpc of leg B and the line of leg B to the cpc of leg A. At the board, between the pairs, expect (r1+r2)/4: for the 40 m ring, (0.30 + 0.48)/4 = 0.195 Ω.

Go round every socket again, this time line to earth. On a 2.5/1.5 mm² ring the readings cluster round (r1+r2)/4; the lowest is at the sockets nearest the board, the highest at the socket electrically midway round the ring, and the spread is only a few per cent because the two conductors are close in size. The highest reading on the ring is the circuit's R1+R2. It goes in column 21 of the schedule, with r1, rn and r2 in columns 18 to 20.

Step 3 does two more jobs for free. Because you are reading between the line and earth terminals of each socket, a correct value proves the line conductor is landed on the L terminal, so polarity at each outlet is confirmed as you go, as GN3 notes, and Polarity Testing Explained covers the rest. And R1+R2 added to Ze gives the calculated Zs for the circuit, which you compare with the live loop test later and with the 80 per cent figures in Maximum Zs Values BS 7671. Remember the Table I1 values are at 20 °C; multiply by 1.20 to estimate the conductor resistance at 70 °C before comparing directly with the Table 41 maxima.

A socket that reads open in step 3 but normal in step 2 has no cpc connected at that outlet. On an EICR that is "absence of earthing at a socket-outlet", C2 in Best Practice Guide 4 Issue 7.4; on an EIC it is a fail until it is connected. If a socket is high in step 3 by more than its step 2 excess, the spur's cpc or the socket's earth terminal is suspect.

Why do the readings vary on a 4 mm² ring?

The "substantially the same" rule assumes the line and cpc are similar sizes. Wiring Matters 89 (March 2022) worked through a 60 m ring in 4/1.5 mm² cable with sockets every 10 m: r1 = 0.277 Ω, r2 = 0.726 Ω, and the step 3 readings ran 0.20 Ω at the origin, 0.23, 0.24, 0.25 at the mid-point, then back down. That is a 20 per cent spread on a perfectly good ring, and the article notes that the figure-of-eight reading at the board is not the maximum. The rule of thumb (r1+r2)/4 still predicts the highest reading, 0.25 Ω, but the sockets near the board will sit below it.

So on 4/1.5 mm², 6/2.5 mm² or any ring where the cpc is much smaller than the line, expect a smooth rise towards the middle and a fall after it. What you are looking for is a step, not a slope: a socket that jumps above its neighbours is on a spur or has a bad joint, while a steady slope on a mixed-size ring is just physics. The IET added this guidance to GN3 from the 9th Edition onwards.

How do I read the anomalies?

Interconnections are the one that gets past people, because every socket works and the end-to-end readings look believable. The give-away is a cross-connected reading that falls over part of the ring rather than rising. They also raise the fault current in the bridged section, so the circuit may be fine on paper and still overload one run of cable.

If the ring shares a neutral with another circuit, the step 1 rn reading can come back through the other circuit's neutral and look plausible; Borrowed Neutrals: Find, Test and Code covers that separately.

What are the limits for a ring final circuit?

Regulation 433.1.204 says accessories to BS 1363 may be supplied through a ring final circuit, with or without unfused spurs, protected by a 30 A or 32 A device, wired in copper of at least 2.5 mm² (1.5 mm² for two-core mineral insulated cable). The circuit is deemed to satisfy the overload rule in 433.1.1 if Iz is not less than 20 A and the load in any part of the ring is unlikely to exceed Iz for long periods. Appendix 15 (informative) then adds the conventions designers work to.

The Iz figure is where thermal insulation catches rings out. Table 4D5 rates 2.5 mm² T&E at 27 A clipped direct (Method C), 21 A above a plasterboard ceiling under up to 100 mm of insulation (Method 100), 17 A under more than 100 mm (Method 101), 21 A in an insulated stud wall touching the inner skin (Method 102) and 13.5 A buried in the insulation with no contact (Method 103). Methods 101 and 103 put 2.5 mm² below the 20 A floor, so a ring that runs through a well-insulated loft or inside a new stud wall is not a 433.1.204 ring at all and needs a bigger conductor or a different design; Cable Sizing and Volt Drop Basics has the method. NICEIC's explanation of why an unfused 2.5 mm² spur is allowed (Professional Electrician, 15 January 2025) is the same logic: the spur must also have Iz of at least 20 A, it is unlikely to be overloaded for more than an hour because it feeds one accessory, and the adiabatic check under 434.5.2 and 543.1.3 shows any 30 A or 32 A device protects the 2.5 mm² live conductors and the 1.5 mm² cpc against fault current provided Zs is within the 0.4 s limit.

The Amendment 4 changes you will notice on a ring are on the form side: the schedule layout and the advisory FI code.

What goes on the certificate?

On both the A4 EIC and EICR schedules the ring block sits under "Continuity (Ω)": column 18 r1 (line), 19 rn (neutral), 20 r2 (cpc), 21 (R1+R2) and 22 R2, with the figures in ohms to two decimal places. Column 22 stays blank for a ring where you have an R1+R2; it is for the alternative R2 method on radials. Then column 26 for polarity and column 27 for the measured Zs, which you will take live afterwards and compare with Ze + R1+R2.

On a periodic, the ring test is still the test; GN3 does not offer a lighter version. What changes is the extent you agree beforehand. If the client will not have the board off or the sockets disturbed, you cannot produce r1, rn and r2, and the schedule entry is "LIM" with the limitation recorded in Section D of the report, not an invented figure. A ring with a known open conductor that the client declines to have repaired is still a C2 and the report is unsatisfactory; EICR Observation Codes C1, C2, C3 and FI covers how that reads on the A4 form.

Remarks worth writing in column 31: which socket the R1+R2 was taken at, that a spur exists and what it feeds, that r2 includes parallel paths through containment, and any socket not reached.

What are the common mistakes?

• Testing through the breaker. Leaving one leg on the MCB gives you r1 through the device contacts, and leaving the cpcs on the earth bar gives a parallel path through every other circuit. Six free ends, every time.
• Pairing the same leg. Line A to neutral A looks like a cross-connection but is not; the readings climb round the ring and people go hunting for a break that does not exist. Tape the legs before you start.
• Forgetting the lead null. A couple of hundredths of an ohm is a real share of a 0.20 Ω reading, and an un-nulled instrument makes every socket look like a weak spur.
• Reading at a socket on a fused spur and recording it as the ring's R1+R2. The FCU and the spur cable add to the figure; the ring's R1+R2 is the highest reading on the ring or its unfused spurs.
• Not reinstating properly. After step 3 the lines are joined to cpcs. Land both lines on the breaker, both neutrals on the bar, both cpcs on the earth bar, then confirm polarity live before you hand it back.
• Treating a sloped set of step 3 readings on a 4/1.5 mm² ring as a fault. Check the cable size before you code it.

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.

Written by the Certio team. Last checked 21 October 2026.