Two tests, taken at the same terminals, minutes apart — and the correct method for one is the exact opposite of the correct method for the other. That is what makes Ze and prospective fault current a reliable source of exam questions and site mistakes alike.
The tests in order
Both of these are live tests. They come after the dead sequence is complete and the installation has been energised deliberately. The order in the BS 7671 test sequence is not decorative: continuity of protective conductors, then insulation resistance, then polarity, before anything is energised. You do not measure Ze on an installation whose insulation resistance you have not yet proved.
At the origin, the sequence within the live tests runs:
- Ze — external earth fault loop impedance, main earthing conductor disconnected
- PEFC — prospective earth fault current, line to earth, earthing conductor reconnected
- PSCC — prospective short-circuit current, line to neutral
- Take the higher of PEFC and PSCC as Ipf and compare against device breaking capacity
Ze: why the earthing conductor comes off
Regulation 643.7.3 requires the earth fault loop impedance to be measured, and Ze is specifically the impedance external to the installation — the supply transformer winding, the line conductor back to it, and the distributor’s means of earth.
If you leave the main earthing conductor connected, you are not measuring that. You are measuring it in parallel with every extraneous-conductive-part bonded at the main earthing terminal. The video makes the point squarely:
“If I would have left it connected, maybe there might have been a gas pipe, water pipe… then we might actually get an acceptable reading. However that could actually mean that effectively we’re using our incoming services such as gas and water as our means of earthing, which would not be a good idea.”
That is the whole risk. A gas or water service can be replaced with plastic tomorrow morning. A reading that only passes because of a parallel path through it is a reading that expires without anybody being told.
Isolate before you disconnect
This is not optional and it is worth being blunt about. The moment the main earthing conductor leaves the main earthing terminal, nothing in the installation is earthed. Every exposed-conductive-part in the building — every metal faceplate, every appliance casing — has lost its connection to earth while the installation is still capable of being live.
So: the installation is isolated first, using safe isolation to the standard set out in GS38 and the prove–test–prove routine, locked off, and only then is the earthing conductor removed. The video demonstrates this correctly — the board is switched off before the conductor is lifted, and the presenter says why:
“If I was to leave this live and then there was a problem, let’s say the metal face plate, then that could cause issues and danger to the customer or people in the area.”
Regulation 643.1 also requires that testing be carried out so as not to cause danger to persons or property. Lifting an earthing conductor on a live installation fails that on its face. If the supply genuinely cannot be isolated — an occupied premises, a tenant on oxygen, a server room — then you do not improvise. You either arrange the isolation properly or you use the distributor’s declared maximum Ze and state on the certificate that the figure is quoted rather than measured.
Tester settings, and the RCD problem
Loop testers offer a high-current test (typically two-lead) and a no-trip low-current test. The high-current test injects a substantial current down the protective conductor for a few cycles, which is why it gives a fast, stable reading — and also why an RCD downstream of the test point will disconnect.
| Test | Leads / setting | Where it is valid |
|---|---|---|
| Ze | L–PE, two-lead high current | At the origin, upstream of any RCD — no residual current path to trip |
| Zs on an RCD-protected circuit | Three-lead or two-lead no-trip | Downstream of an RCD; slower, and readings can be less stable |
| PEFC | L–PE, high current | At the origin, all parallel paths reconnected |
| PSCC | L–N, high current | At the origin; test lead moved from the PE to the N terminal |
The reading in the video came back as 0.23 Ω, and notably faster than the earlier no-trip measurement. That speed difference is the high-current test doing its job — but it is only available to you where there is no RCD in the path. Choosing the wrong setting downstream of an RCBO does not give you a wrong number; it gives you a tripped circuit and an unhappy customer.
PEFC and PSCC: everything back on
Now the logic inverts. Having taken Ze in isolation, the earthing conductor goes back into the main earthing terminal before prospective fault current is measured.
“I’ve reconnected the earthing conductor into the MET and the board, and this is to make sure all parallel earth paths are in place.”
The reason is that Regulation 643.7.3.201 and the assessment of prospective fault current under Regulation 434.1 are about the worst case. Parallel paths lower the loop impedance, and lower impedance means higher fault current. For Ze you want the honest, unassisted figure. For PFC you want the largest current the installation could ever be asked to interrupt.
The measurements in the video:
- PEFC (line to earth): 799 A
- PSCC (line to neutral): 759 A
Taking PSCC means physically moving the test lead from the protective earth terminal on the instrument to the neutral terminal and selecting the L–N loop range. Same probes, same terminals in the board, different pair of conductors.
Which figure is your Ipf
The higher one. Here that is the PEFC at 799 A, so 799 A is the prospective fault current recorded on the certificate.
On a three-phase supply there is a further step that the video does not cover: the line-to-line prospective short-circuit current is conventionally taken as twice the measured line-to-neutral value, and that doubled figure is usually the one that governs. On the single-phase board demonstrated, the straightforward comparison of the two measured values is the whole job.
Both PEFC and PSCC have their own boxes in the supply characteristics section of the EIC and the EICR, and the higher figure carries through as Ipf. This is one of those places where blank cells get read as “not tested” — see our note on what belongs on each certificate for why an incomplete schedule is an incomplete document.
The point of the number: breaking capacity
Measuring Ipf is not an academic exercise. Regulation 434.5.1 requires that the rated breaking capacity of every protective device be not less than the prospective fault current at the point where it is installed.
In the video the devices are marked 6000 A against a measured 799 A, which is comfortably adequate. That comfortable margin is typical of a domestic supply well down a street, and it is exactly why the check gets skipped — until you are working at a large TN-C-S intake, close to a substation, or on a three-phase supply where 16 kA is not an unusual figure and a 6 kA device on the board is a genuine finding.
Where you do find a device with a breaking capacity below the measured Ipf, that is not a paperwork note. An overcurrent device that cannot clear the fault it is asked to clear is a real danger, and on an EICR it belongs in the C2 territory — potentially dangerous — rather than being noted as an improvement recommendation. The exception is the backup protection arrangement described in Regulation 434.5.1 itself, where an upstream device with adequate capacity protects the downstream one; if that arrangement is present and verifiable, record it.
The details that get marked
Under 2391-52 conditions, the things people lose marks on here are consistent:
- Direction of the earthing conductor. Off for Ze, on for PFC. Reversing the two is the single most common error on this topic.
- Isolation before disconnection. If the answer describes lifting the earthing conductor without stating the installation is isolated, it is wrong regardless of the numbers.
- Recording a quoted value as measured. If you used the distributor’s declared maximum Ze, say so on the certificate.
- Comparing Ipf against breaking capacity. Measuring it and not acting on it is only half the test.
Because the exam is open book, the useful preparation is knowing where things are rather than memorising them: Part 6 for the test requirements, Chapter 43 for the breaking capacity duty, and the maximum Zs tables in Chapter 41 for the disconnection-time checks that follow on from your loop readings. Two minutes per question does not leave room for hunting.
How 2391 Prep Fits Into This
2391 Inspection & Testing Pro is built around the same working detail. The City & Guilds 2391-52 is 60 questions in 120 minutes at a 75% pass mark, and it is an open book exam — so the app is organised for fast lookup as much as for revision.
- 556 practice questions across 7 subject areas, including the test procedures and certification areas that carry Ze, PEFC, PSCC and breaking-capacity questions
- 4 working calculators — including loop impedance and prospective fault current work, so you can check a measured Ze against the maximum Zs for the device on the board
- 15 quick reference cards, covering the test sequence and supply characteristics you need to reach for under time pressure
- 6 field tools for the on-site side of the same job
If you want a broader treatment of where each of these values is measured and why, our guide to Ze, Zs, PFC and PSCC covers the full picture.
