Summary

Earth fault loop impedance testing is one of the dead-and-live tests every electrician carries out to prove a circuit is safe. The loop impedance (Zs) is the total resistance of the path a fault current would take if a line conductor touched earth — out through the supply line, back through the earth path, and round to the source. If that path has too much impedance, fault current is too low to trip the protective device fast enough, and exposed metalwork can sit at a dangerous touch voltage long enough to kill.

The whole point of measuring Zs and comparing it to a maximum is automatic disconnection of supply (ADS) — the primary protective measure in BS 7671 for most installations. Get the loop impedance low enough and the MCB, RCBO or RCD operates inside the required disconnection time. This is core knowledge for anyone doing an EICR, an initial verification, or signing an Electrical Installation Certificate. It is also one of the most commonly misunderstood areas on site, because the tabulated maxima are stated at 70°C/conductor operating temperature while you measure on a cold installation.

The common misconception is that the table value is the number you compare your meter reading against. It is not. The table values are the design maxima at full operating temperature. Your measured (cold) value must be lower — by convention, no more than 80% of the tabulated figure (the Cmin / rule-of-thumb 0.8 factor). The other frequent error is forgetting that on a TT system you almost never rely on the overcurrent device for ADS at all — the RCD does the job, and the Zs limit is governed by the RCD's rated residual operating current, not the table.

Key Facts

Quick Reference Table

Quoting an electrical job? Describe the work and squote handles the pricing.

Try squote free →

Indicative maximum Zs at the stated disconnection time — tabulated (operating temperature) values, 230V. Apply the 0.8 factor before comparing to a cold measured reading. Always confirm against the current BS 7671 tables; figures marked.

Device Rating Max Zs (0.4s, tabulated) × 0.8 (measured limit)
Type B MCB 6A 7.28Ω 5.82Ω
Type B MCB 16A 2.73Ω 2.18Ω
Type B MCB 20A 2.19Ω 1.75Ω
Type B MCB 32A 1.37Ω 1.09Ω
Type C MCB 6A 3.64Ω 2.91Ω
Type C MCB 16A 1.37Ω 1.09Ω
Type C MCB 20A 1.09Ω 0.87Ω
Type C MCB 32A 0.68Ω 0.55Ω
Type D MCB 32A 0.34Ω 0.27Ω
30mA RCD (any rating) — 50V ÷ 0.03 ≈ 1667Ω use Table 41.5 / ≤200Ω guidance

The MCB values follow the relationship Max Zs ≈ U0 ÷ (trip multiple × In), e.g. Type B 32A: 230 ÷ (5 × 32) = 1.44Ω before the 0.95 Cmin factor is applied. Use this to sanity-check a table figure but always cite the printed table.

Detailed Guidance

Why the 80% rule exists

The maximum Zs values in Tables 41.2–41.4 are calculated for conductors at their normal operating temperature (typically 70°C for thermoplastic/PVC insulation). At that temperature copper has a higher resistance than when cold. When you test, the circuit is usually near ambient (20°C-ish) and unloaded, so its actual resistance — and therefore measured Zs — is lower than it will be in service.

If you measured a value that was right on the tabulated limit while cold, the circuit would exceed that limit once it warmed up under load, and disconnection could become too slow. To prevent that, the accepted approach is to compare your measured value against 80% of the tabulated maximum (the rule-of-thumb 0.8 factor). The rigorous method in Guidance Note 3 / the On-Site Guide uses the 0.95 (Cmin) factor combined with a temperature-correction factor; the simplified 0.8 figure rolls these together for site use and is conservative.

Is measured Zs ≤ (0.8 × tabulated max Zs)?
         │
   ┌─────┴─────┐
  YES         NO
   │           │
 PASS    Investigate: high Ze? long cable run?
         undersized CPC? loose connection?
         → reduce R1+R2, improve earthing,
           or upsize CPC. Re-test.

Working out Zs: measured vs calculated

You can establish Zs two ways and they should agree:

On a final circuit protected by a 30mA RCD, ADS for very small fault currents is provided by the RCD regardless of the overcurrent device's Zs — but you still record Zs to confirm the circuit's earth fault path is sound and within limits.

TT systems and RCDs

On a TT system the earth return is through an electrode in the ground, so Ze is high and the overcurrent device will not reliably disconnect on an earth fault. ADS is provided by an RCD. The maximum Zs is then governed by Reg 411.5.3:

Common reasons Zs is too high

Frequently Asked Questions

Do I compare my meter reading to the table value or to 80% of it?

To 80% of it. The printed maximum Zs values assume conductors at operating temperature. Your test is on a cold circuit, so apply the 0.8 rule-of-thumb factor (or the formal Cmin/temperature method in GN3) and compare against that lower figure. If your meter has a built-in "pass/fail" function it usually applies the correction for you, but always know which value it is using.

What disconnection time applies — 0.4s or 5s?

0.4s for final circuits up to 63A on a TN system supplying socket-outlets, or fixed equipment up to 32A (Reg 411.3.2.2). 5s for distribution circuits and final circuits over 63A (Reg 411.3.2.3). On TT systems use the times in Table 41.1, with the RCD providing disconnection.

My RCD-protected circuit has a high Zs but the RCD trips fine — is it a pass?

The 30mA RCD provides ADS for earth faults, so the circuit can satisfy the disconnection requirement via the RCD even with a higher loop impedance. But a high Zs still points to a problem — undersized CPC, long run, or a loose connection — and you should record it and investigate. Don't treat the RCD as a licence to ignore an out-of-spec earth path.

Is loop impedance testing notifiable under Part P?

The testing itself isn't notifiable, but the work that requires it often is. In England and Wales, new circuits and certain alterations in dwellings are notifiable under Part P (Approved Document P). Loop testing is part of the verification you document on the certificate that supports the Building Control notification or competent-person scheme registration.

Why did my loop tester trip the RCD?

A standard high-current loop test injects enough current to operate a 30mA RCD. Use the instrument's no-trip / low-current loop range on RCD-protected circuits, or test upstream of the RCD where practical. The trade-off is slightly lower accuracy on the no-trip range.

Regulations & Standards