So why in the table below do some of the Zs add up, and some not?

My guess is "Zs" has been measured with a loop tester on a non-trip setting (while Ze was done on high current, and R1+R2 with a simple Ohm meter). My old megger installation tester very often reads a couple of a tenths of an Ohm high on no-trip (not fruit machine exactly, it's often reproducible, and within published specs of course).

  - Andy.

In a TN-C-S installation, neutral currents (in the network in general, but also in the installation) can affect Ze between measurements. There's no saying that Ze was 0.31 ohms at the time Zs was measured for any circuit.

Some instruments are better than others at nulling the effects of load currents and 'leakage' currents.

There will also be a small resistance (measured in a few 1/100ths of ohms, but perhaps upto 0.1 ohms in lower rated breakers) attributable to the contact resistances in each RCBO, which will differ from device to device. These resistances are typically less, the higher the breaker rating (or the higher the current of loads connected in use, especially if it's an existing installation) ... but I'm guessing it's anything above 0.08 to 0.1 ohms difference you're questioning (quite rightly)?

Might the apparent increase in the measured Zs also be due to the inductance of the RCBO coils; particularly if, as Mr. Jewsbury suggests, this has been measured using a low current (non-trip) instrument? Should this were the case, there remains an anomaly between the results for the two cooker circuits; are they both protected by RCBOs?

Geoff

Possibly, however if inductance were the principle issue, surely 'lights up' and 'lights down' would have a similar impedance delta?

Looking again, I think it's the R1+R2 numbers that worry me more ... why's the upstairs lighting almost twice that for downstairs? 1860 mΩ would imply over 50m of 1mm2 to the furthest point (about 24m further than downstairs). Likewise why downstairs sockets more than double upstairs?  - is there a very long spur out to an outbuilding or something or does it really have a 300m long ring?

   - Andy.

Agreed. There’s always repeatability of measurements and variability between devices to consider; 300 milliohm difference between (presumably) similar devices does seem a lot though. I’ve been trawling my archive of test results to see if I can see a pattern, or get a feel for typical uplift values; that hasn’t helped me much since, where I have a measured Ze (high current) and R₁+R₂, I normally state Zs as the sum of these rather than a directly measured result which is so often at odds, especially when a non-trip setting is used.

It would be interesting to know if anyone has measured inductance, or impedance (@ 50 Hz), results for a range of devices (or even just one).

Geoff

AJJewsbury said:

1860 mΩ would imply over 50m of 1mm2

2-way switching can add a lot (e.g. landing light). Believe it or not, it's possible to get over 50 m cable distance in (R1+R2) from the furthest loop-in luminaire on a lighting circuit in a 32' x 10' transportable office unit.

gkenyon said:

but I'm guessing it's anything above 0.08 to 0.1 ohms difference you're questioning (quite rightly)?

Thank you Graham (and others) - spot on!

gkenyon said:

Possibly, however if inductance were the principle issue, surely 'lights up' and 'lights down' would have a similar impedance delta?

There are one or two inconsistencies in the information that I have about occupancy of the house, but it seems to have been let.

Remember that existing let properties had to be inspected by 01/04/2021 and by coincidence this one was I&T'd on 06/11/2020 with a recommendation for testing 5 years later.

The first question is whether R1 + R2 was measured, and if so, why? I wouldn't normally expect it for an EICR.

The DBs (there is an off-peak one, but this appears not to have been connected at the time of the EICR) are barely accessible. They are very low down under the stairs, so it would be a case of lying on one's front with head and shoulders in the cupboard. I wouldn't pull the DB apart unless I had to.

The DBs look only a few years old - modern all RCBO at the budget end of the market. When I saw them, I assumed that they had been installed in 2020. Perhaps they were, but then there should be an EIC rather than an EICR.

I don't think that I would have submitted this report. I am happy to accept a few tens of milliohms discrepancy, but I would want to explain more even if it meant checking the calibration of my tester.