Maximum Zs Values Explained: The 80% Rule and Zs Tables

Quick answer

Maximum Zs is the highest earth fault loop impedance a circuit can have while still guaranteeing its protective device disconnects fast enough under an earth fault. BS 7671 tabulates maximum Zs values for each device type and rating (Tables 41.2 to 41.4), calculated at conductor operating temperature. Because you test at ambient temperature, you compare your measured reading against 80% of the tabulated figure. For the classic example, a 32 A Type B MCB or RCBO, the design maximum is 1.37 Ω and your on-site measured limit is 1.10 Ω.

What earth fault loop impedance actually is

Strip away the jargon and Zs is just the total resistance (strictly, impedance) of the path a fault current takes when a line conductor touches earth. Picture a live conductor chafing through on a pump motor casing in a washdown area, which I have seen more times than I would like. The fault current flows out along the line conductor, through the fault, back along the CPC, through your main earthing terminal, and back to the transformer winding via the supply earth. That whole round trip is the earth fault loop, and Zs is its impedance in ohms.

It breaks down into two parts:

• Ze - the external loop impedance, everything on the supply side of your installation
• R1 + R2 - the resistance of the line conductor and CPC of the circuit itself

So Zs = Ze + (R1 + R2). That is why a long lighting circuit in 1.0 mm² twin and earth has a much higher Zs at the far point than a short 6 mm² cooker circuit two metres from the board.

Why maximum Zs matters

Ohm's law is doing all the work here. Fault current = voltage divided by loop impedance. Low Zs means a big fault current, which slams the MCB open in milliseconds. High Zs means a limp fault current that might sit there cooking the cable and holding the casing of that pump motor at a dangerous voltage while the breaker does nothing.

BS 7671 sets maximum disconnection times in Table 41.1. On a TN system at 230 V, final circuits up to 32 A must disconnect within 0.4 seconds, and distribution circuits and sub-mains within 5 seconds. The maximum Zs value is simply the highest impedance at which the device is guaranteed to meet that time.

If your measured Zs is over the limit, the circuit does not comply for fault protection by automatic disconnection. On an EICR that is a genuine defect, not a paperwork quibble. If you want the coding logic for that scenario, see our guide to EICR codes.

Where the numbers come from

The formula behind every value in the tables is:

Zs = (Cmin × U0) / Ia

Where U0 is 230 V, Cmin is 0.95 (a factor allowing for supply voltage dipping below nominal), and Ia is the current that trips the device within the required time.

For MCBs and RCBOs, Ia is the instantaneous trip threshold, taken at the worst case of the band:

• Type B: trips at 3 to 5 × In, so Ia = 5 × In
• Type C: trips at 5 to 10 × In, so Ia = 10 × In
• Type D: trips at 10 to 20 × In, so Ia = 20 × In

That is why a Type C breaker of the same rating always has half the maximum Zs of a Type B, and a Type D half again. Every time someone swaps a tripping Type B for a Type C "because it holds in", they have just halved the Zs headroom on that circuit. Check it still complies before you walk away.

For fuses (BS 88-2, BS 88-3, BS 3036 and the old BS 1361 cartridges) there is no clean multiplier. Ia comes off the time-current curves in Appendix 3 and differs for 0.4 s and 5 s disconnection, which is why fuse tables carry separate columns for each.

The 80% rule of thumb

This is the bit that catches people out on assessments. The tabulated values in Tables 41.2 to 41.4 assume the conductors are at their maximum operating temperature, 70°C for standard thermoplastic insulation. When you test, the circuit is usually unloaded and the copper is at ambient, maybe 10 to 20°C. Cold copper has lower resistance, so your meter reads lower than the impedance the loop would actually present mid-fault, when everything is hot.

The fix is the 80% rule: your measured Zs must not exceed 0.8 × the tabulated maximum. Multiply the book value by 0.8, or equivalently divide your reading by 0.8 and compare with the table. Same arithmetic either way.

Two important wrinkles:

1. A reading between 80% and 100% of the tabulated value is not an automatic fail. It is a flag to investigate. You can do a proper temperature-corrected calculation (Appendix 14 territory) instead of condemning the circuit. If the conductors genuinely were cold at test, the correction may still pass it. Record what you did in the remarks on the schedule.
2. Manufacturer data can trump the generic tables. Appendix 3 encourages using manufacturer-specific Zs figures where available, and these are often more generous than the BS 7671 tables. If you use them, say so in the remarks column. On industrial MCCBs to BS EN 60947-2 you have no choice: BS 7671 publishes no Zs tables for them at all, so it is manufacturer data or calculation from the time-current curve.

The On-Site Guide also publishes alternative values calculated at a 10°C conductor temperature, which sit slightly above the plain 80% figures. Fine to use if that assumption genuinely holds, and again, note it.

Zs tables for Type B and Type C MCBs and RCBOs

These are the values you will use daily, from Table 41.3, for devices to BS EN 60898 with the 80% measured limits alongside. RCBOs to BS EN 61009-1 use exactly the same values as an MCB of the same curve and rating, because the overcurrent element is the same. The RCD part is irrelevant to these numbers.

Type B (Ia = 5 × In)

Rating	Max Zs, design (Ω)	Max Zs, measured 80% (Ω)
6 A	7.28	5.82
10 A	4.37	3.50
16 A	2.73	2.18
20 A	2.19	1.75
32 A	1.37	1.10
40 A	1.09	0.87
50 A	0.87	0.70

Type C (Ia = 10 × In)

Rating	Max Zs, design (Ω)	Max Zs, measured 80% (Ω)
6 A	3.64	2.91
10 A	2.19	1.75
16 A	1.37	1.10
20 A	1.09	0.87
32 A	0.68	0.54
40 A	0.55	0.44

A pattern worth memorising: a Type C at any rating has the same Zs limit as a Type B at double that rating. A 16 A Type C and a 32 A Type B both come out at 1.37 Ω design, 1.10 Ω measured. Once that clicks, you can sanity-check readings in your head at the board.

Worked example: 32 A Type B RCBO on a ring final

Say you are testing a kitchen ring final on a 32 A Type B RCBO, TN-C-S supply.

Step 1 - the limit. Ia = 5 × 32 = 160 A. Maximum Zs = (0.95 × 230) / 160 = 218.5 / 160 = 1.37 Ω. That is the design figure for the schedule. Your on-site measured limit is 1.37 × 0.8 = 1.10 Ω.

Step 2 - the reading. Ze at the origin measures 0.28 Ω. Your R1 + R2 from the ring tests came out at 0.36 Ω at the furthest point. Expected Zs is roughly 0.28 + 0.36 = 0.64 Ω. Your loop tester at the worst socket reads 0.66 Ω. That tallies, and 0.66 Ω is comfortably under 1.10 Ω. Pass, record it, move on.

Step 3 - the borderline case. Suppose instead the meter read 1.18 Ω. That is over the 1.10 Ω measured limit but under the 1.37 Ω design maximum. Do not fail it on the spot. Check the reading is real (test lead nulled, clean connection, try adjacent sockets), check the conductor temperature assumption, and run the temperature-corrected calculation. If it still will not comply, then you are into fixing the circuit: usually a poor joint, an undersized or damaged CPC, or simply a circuit that was always too long for the device.

One practical note on measurement: on RCBO-protected circuits you need a tester with a proper no-trip loop test, and cheap implementations can be wildly optimistic on low readings. A serious MFT such as the Megger MFT1741+ or the Fluke 1664 FC will do accurate no-trip loop tests without firing the RCD, and the Kewtech KT65DL does the same job at friendlier money for a domestic-focused outfit. If you do a lot of fault-finding and quick verification rather than full certification, a dedicated two-wire loop tester like the Megger LTW335 lives in a pocket and gives fast loop readings without dragging the full MFT out. For a proper comparison of the big testers, see our best multifunction tester guide.

What about TT systems and RCDs?

On a TT supply the earth return path runs through an earth electrode in the dirt, and the loop impedance is usually far too high for an MCB ever to see enough fault current. Fault protection comes from an RCD instead, and the test changes from "is Zs below the Table 41.3 value" to Ra × IΔn ≤ 50 V (Regulation 411.5.3). With a 100 mA RCD that allows up to 500 Ω; even a 30 mA device allows 1667 Ω. Those numbers look enormous next to an MCB table, which is exactly the point: the RCD trips on milliamps of imbalance, not hundreds of amps of fault current.

On TN systems, an RCD can legitimately be used for fault protection where the loop impedance is too high for the overcurrent device, but that should be a documented design decision, not a shrug because the Zs failed. And it does nothing for the short-circuit withstand side of the equation. This comes up constantly on EV installs, where long supply runs push Zs up; our EV charger installation and testing guide covers it, and if you are condemning an old board over it, the consumer unit upgrade guide is the next read.

FAQ

What is the maximum Zs for a 32 A Type B MCB?

1.37 Ω is the design maximum from Table 41.3, and that is the figure that goes in the maximum Zs column on your schedule of test results. Your measured reading on site should not exceed 80% of it, which is 1.10 Ω, unless you temperature-correct or use manufacturer data.

Do RCBOs have different maximum Zs values from MCBs?

No. An RCBO to BS EN 61009-1 uses the same maximum Zs as an MCB of the same type and rating, so a 32 A Type B RCBO is 1.37 Ω design, 1.10 Ω measured. The residual current element is a separate function and does not change the overcurrent trip characteristic.

Do I record the 100% or the 80% value on the certificate?

Record the 100% design value from the tables as the maximum permitted Zs, and record your actual measured reading against it. The 80% comparison is your on-site acceptance check, not the figure that goes in the maximum Zs column. If you used manufacturer data or a corrected calculation instead, note the source in the remarks.

My measured Zs is just over the 80% value. Is that a fail?

Not automatically. Between 80% and 100% of the tabulated value, investigate before condemning. Verify the reading, consider the actual conductor temperature, and do the temperature-corrected calculation. If it exceeds the 100% design value even after that, the circuit does not comply and needs remedial work.

Why is my measured Zs lower than Ze plus R1+R2?

Usually parallel earth paths. On TN-C-S supplies especially, bonded services and structural steel give fault current extra routes back, so the measured loop reads lower than the arithmetic sum. It is normal, and it is one reason the measured value alone should not be your only check that the CPC is intact. Your continuity tests prove the CPC; the loop test proves disconnection.