Earth Fault Loop Impedance Testing: Ze, Zs and Ipf Explained

Earth fault loop impedance testing is the live test that proves a line-to-earth fault will draw enough current to disconnect the supply in time. You record two figures: Ze, the external loop impedance at the origin, and Zs, the total loop impedance at the far end of each circuit. Zs must not exceed the maximum for the protective device in Tables 41.2 to 41.4 of BS 7671:2018+A4:2026, compared at 80 per cent for a cold site measurement: 1.10 Ω for a 32 A Type B circuit-breaker against the tabulated 1.37 Ω. The same instrument gives prospective fault current, Ipf, which must not exceed the breaking capacity of the devices on the board.

What is the difference between Ze and Zs?

The earth fault loop on a TN system runs from the point of fault along the cpc to the main earthing terminal, through the earthing conductor to the supply earth (the cable sheath on TN-S, the combined neutral and earth on TN-C-S), back through the transformer winding and out along the line conductor to the fault. On a TT system the return path is the installation's earth electrode and the general mass of earth, covered in Earthing Arrangements Explained: TN-S, TN-C-S and TT.

Ze is the part of the loop that belongs to the supply: transformer winding, service cable and the distributor's earth. Zs is the whole loop seen from a point on a circuit: Ze plus the line conductor resistance R1 plus the cpc resistance R2 up to that point.

Ze goes on the certificate once, with the earthing arrangement and Ipf, in the supply characteristics section. Zs goes on the schedule of test results for every circuit, next to the maximum permitted value for the device. The short definitions are at What is Ze? External earth fault loop impedance and What is Zs? Earth fault loop impedance explained.

Regulation 313.1 of BS 7671 lets the supply characteristics be determined by calculation, measurement, enquiry or inspection, so Ze on a new job can come from the distributor before the supply is live. The figures distributors have traditionally quoted come from Energy Networks Association Engineering Recommendation P23:

These are typical maxima for supplies up to 100 A, not guarantees. The 2018 revision, P23/2, replaced the single TN figures with percentiles (90 per cent of premises below 0.34 Ω, 98 per cent below 0.64 Ω), and the IET's Wiring Matters article on the revision states that distributors are under no obligation to maintain a particular Ze. That is why the measured figure, not the quoted one, is what you certify.

What limits does a loop reading have to meet?

A fault has to clear within the time in Table 41.1. The maximum Zs for each device is simply the impedance that still lets the disconnecting current flow at 230 V × 0.95.

For circuit-breakers and RCBOs to BS EN 60898 and BS EN 61009 the disconnecting current is the top of the instantaneous band (5, 10 or 20 × In for Types B, C and D), so the 0.4 s and 5 s values are the same. The ones you meet daily:

The full set, with fuses and Type D, is in Maximum Zs Values BS 7671: Tables, 80% Rule and Ze. Check any figure against the printed table in the edition you are certifying to; rounding can differ in the last digit.

The 0.8 comes from temperature. The tables assume conductors at their 70 °C operating temperature, where copper has about 1.20 times its 20 °C resistance, and 1 ÷ 1.20 is 0.83, rounded down. IET Guidance Note 3 gives two acceptable routes: compare the cold measurement with 0.8 × the tabulated value, or multiply R1+R2 by 1.20, add Ze and compare with the full value. The second is more exact because Ze does not warm up with the circuit.

Where an RCD provides the disconnection, Table 41.5 applies instead: Zs × IΔn must not exceed 50 V, which gives 1667 Ω for 30 mA, 500 Ω for 100 mA, 167 Ω for 300 mA and 100 Ω for 500 mA. Regulation 411.4.204 allows those values on TN circuits as well as TT, with the overcurrent device still protecting the cable under Chapter 43.

How do I measure Ze at the origin safely?

The test measures the supply's earth path alone, so the parallel paths inside the installation (main bonding to gas, water and structural steel) must be taken out of the loop. That means disconnecting the earthing conductor from the main earthing terminal, and while it is off the installation has no earth at all.

1. Isolate the whole installation at the main switch, lock off and prove dead on the load side. Isolate any other source too: solar PV, battery storage or a generator.
2. Disconnect the earthing conductor from the main earthing terminal.
3. Connect the instrument between the incoming line at the supply side of the main switch and the disconnected earthing conductor, using leads and fused probes that meet HSE Guidance Note GS38. This is live working at the highest fault level in the building.
4. Run the high-current loop test. Read Ze and the prospective earth fault current shown with it. On a three-phase supply test each line and record the highest Ze.
5. Reconnect the earthing conductor, tighten and tug-test it, then remove the lock-off and re-energise.

Step 5 is the one that gets forgotten when the phone rings. If a later loop test at a socket shows no earth, look at the main earthing terminal before anything else.

On a periodic inspection where the client cannot accept a shutdown, do not disconnect the earth on a live installation. Take the reading at the board with everything connected, record it as Zs at the distribution board (Zdb) rather than Ze, and write the limitation on the report. The figure will be lower than true Ze because of the parallel paths through bonding. Agreeing limitations is covered in EICR Sampling and Limitations Explained: Extent, LIM and N/V.

How do I measure Zs on each circuit?

Zs is measured live at the point electrically furthest from the board: the last socket on a radial, the midpoint of a ring final, the last luminaire on a lighting circuit, or the terminals of fixed equipment. On three-phase circuits test each line to earth and record the highest.

Regulation 643.7.3 requires the relevant impedances to be measured or determined by an alternative method, which gives three legitimate ways of arriving at Zs:

• Direct measurement at the far point with the loop function.
• Calculation: Ze (or Zdb for a sub-board) plus the R1+R2 obtained during dead testing. See Continuity Testing: R1+R2 Method, Wander Lead and Values.
• Measurement at the board plus R1+R2, which is the same sum with a measured starting point and is the normal method for circuits fed from distribution boards.

Whichever you use, compare the result with the limit before moving on. If the measured value fails, add Ze to the measured R1+R2. If the sum passes and the reading fails, suspect the measurement first. If both fail, the circuit is too long for the device, has a high-resistance joint or sits on a high Ze, and High Zs Readings: Causes, Checks and How to Fix Them takes it from there.

Where the far point is downstream of an RCD and you have already confirmed the loop at one point, the note to Regulation 643.7.1 accepts continuity of the protective conductors as proof for the rest of that circuit. That saves loop testing every spur behind a device that keeps tripping.

Worked examples: does this circuit pass?

A socket radial that passes. 20 A Type B RCBO, 2.5/1.5 mm² twin and earth, 28 m to the last socket, TN-C-S with Ze measured at 0.24 Ω. From the On-Site Guide resistance table, 2.5 mm² with a 1.5 mm² cpc is 19.51 mΩ/m at 20 °C, so R1+R2 should be 28 × 0.01951 = 0.55 Ω, and the dead test gives 0.56 Ω. Expected Zs is 0.24 + 0.56 = 0.80 Ω. The no-trip measurement at the last socket is 0.86 Ω. The limit is 2.19 × 0.8 = 1.75 Ω. Pass, and the measured and calculated figures agree within what a low-current test can resolve.

A commercial radial that fails. 32 A Type C circuit-breaker with no RCD, 4.0/1.5 mm² twin and earth, 22 m, Ze 0.31 Ω. R1+R2 at 16.71 mΩ/m is 0.37 Ω, so calculated Zs is 0.68 Ω, and the high-current test reads 0.70 Ω. The cold limit is 0.68 × 0.8 = 0.55 Ω. Fail. Corrected properly, 0.31 + (0.37 × 1.20) = 0.75 Ω against a tabulated 0.68 Ω, so it still fails. The options are a Type B device if the load's inrush allows (limit 1.10 Ω), a larger cpc, or an RCD providing disconnection under Regulation 411.4.204.

The second example is why a 1.5 mm² cpc on a Type C or D device deserves a calculation at design stage, not a surprise at test. Sizing is covered in Cable Sizing and Volt Drop: Ib, In, Iz and Appendix 4 Explained.

High-current or no-trip, 2-lead or 3-lead: which test mode?

A conventional loop test draws a substantial current between line and earth for a couple of cycles and works out the impedance from the voltage drop. That current is a genuine earth fault as far as a 30 mA RCD is concerned, so the device trips. No-trip modes avoid that by using a test current of about 15 mA, the level a healthy 30 mA device is required to ignore, over a much longer test.

Practical rules that hold on most instruments:

• Use the high-current test wherever nothing will trip. It is faster and more repeatable.
• Use 3-lead no-trip wherever a neutral is available; the neutral reference improves the measurement and confirms polarity at the same time.
• If an RCD still trips in no-trip mode, the circuit probably has standing earth leakage that the test current is adding to. Calculate Zs and note it instead of tripping the device repeatedly.
• Never link out or bypass an RCD on a live installation to get a high-current reading.
• Zero the test leads in loop mode if the instrument requires it, and re-zero after swapping a plug-in lead for probes.

How accurate is a loop impedance reading?

Less accurate than the two decimal places suggest. Loop instruments are built to BS EN 61557-3, which requires the manufacturer to declare a range over which the operating error stays within 30 per cent under worst-case conditions. Basic accuracy on the bench is much better, often around 5 per cent plus a few digits, but site conditions are not a bench.

Three things move the reading:

• Supply noise and switching loads. A low-current test is measuring a tiny voltage change against a background of everything else on the network. Manufacturers' application guides recommend taking more than one measurement for exactly this reason.
• The RCD itself. A 15 mA test current may not saturate the sensing coil of some RCDs and RCBOs, and the coil's impedance then appears in the result. A reading taken upstream of the RCD will be lower than one taken just after it.
• Low impedances near a transformer. Close to the substation the loop is mostly reactance. Instruments that measure resistance only can under-read, and at 0.05 Ω a resolution of 0.01 Ω is already a 20 per cent step.

The practical consequence: a reading within a tenth of an ohm or so of the limit, taken in no-trip mode, has not proved anything on its own. Ze + (R1+R2), both obtained with better methods, is the tie-breaker.

Why does measured Zs not match Ze + (R1+R2)?

How do I check prospective fault current against breaking capacity?

Prospective fault current is the current that would flow into a fault of negligible impedance at the point of measurement. Two values matter:

• PSCC, prospective short-circuit current: line to neutral, measured with the instrument's L-N function.
• PEFC, prospective earth fault current: line to earth, equal to U0 ÷ Ze at the origin. A Ze of 0.24 Ω gives 230 ÷ 0.24 = 958 A, or 0.96 kA.

Ipf is the higher of the two. On TN-C-S they are close because neutral and earth share a conductor upstream of the cut-out. On TT the PEFC may be only a few amps and the PSCC is the figure that matters. On a three-phase supply the worst case is a three-phase fault, about 1.73 times the single-phase line-to-neutral value; doubling the line-to-neutral reading is the common rule of thumb because it errs on the safe side. More on three-phase work is in Three-Phase Testing Basics for Domestic Electricians.

Measure at the origin and at every distribution board, then compare with the kA marking on each device there. Circuit-breakers to BS EN 60898 are commonly 6 kA in consumer units and 10 kA in commercial boards; BS 88 fuses are typically 80 kA.

A reading above 6 kA in a dwelling is not automatically a failure. The usual declared maximum for a single-phase service up to 100 A is 16 kA at the distributor's end of the service cable, and consumer units to BS EN 61439-3 are tested to a 16 kA conditional rating when backed by a service fuse of up to 100 A. Chapter 43 accepts a lower-rated device where an upstream device gives back-up protection and the manufacturer confirms the combination. Record the breaking capacity in the device columns so the comparison is visible. The definition is at What is prospective fault current (PFC)? Explained.

How should loop test results be recorded and coded on an EICR?

Enter the measured (or calculated) Zs for each circuit beside the maximum permitted value, and say in the remarks column where a value was calculated or taken at the board. Column-by-column guidance is in How to Fill In a Schedule of Test Results, Column by Column.

Coding is the inspector's judgement. The classifications below are the ones most inspectors would reach, in line with Electrical Safety First Best Practice Guide 4:

Write the observation with the numbers in it: circuit, device, measured Zs, the limit and Ze. Wording examples are in How to Write EICR Observations That Stand Up and the codes themselves in EICR Observation Codes Explained: C1, C2, C3 and FI.

Amendment 4 was published on 15 April 2026. The published summaries of it from NICEIC and SELECT do not list a change to the loop impedance test in Regulation 643.7; the visible changes for inspectors are in how the model forms set out observations. Check the tables in your own copy before relying on any figure.

Where does loop testing fit in the test sequence?

Loop impedance and prospective fault current are the first live tests. They follow the dead tests (continuity, ring final circuit continuity, insulation resistance and polarity) and come before RCD testing, because you want to know the earth path is sound before deliberately putting a fault current down it. Functional testing comes last.

Related tests are covered in Insulation Resistance Testing Guide: Voltages, Minimums, Method, Ring Final Circuit Testing: r1, rn, r2 and R1+R2 Explained, Polarity Testing Explained: Checks, Methods and EICR Codes and RCD Testing Explained: Types, Test Currents and Trip Times. The whole inspection, start to finish, is in How to Do an EICR Step by Step: Inspector's Workflow.

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 free for 7 days, then £4.99 a month.

Written by the Certio team. Last checked 4 October 2026.