Bit more detail needed. Are you testing the RCD on its own, or are various outbound circuits connected at that point. An RCD will not trip if there is no supply between it's incoming L and N, but I assume you would have noticed that. There may be a large Dc blinding the RCD.

But before we worry about the clever stuff,Does the RCD trip OK on its own test button,  and if so how does the connection of your test meter differ. I presume the tester is not accidentally plugged into a circuit that is not isolated by that RCD ;-)

Mike.

If you're saying the RCD in the left hand position fails (regardless of which RCD it is) - my first guess would be a N-PE fault on that side (the fault then allows a proportion of the PE test current to return to the load N conductor on the load side of the RCD - so the RCD doesn't see the imbalance). Theoretically that should have shown up on the insulation test before energisation - but in practice such faults can occur in appliances or extension leads that were unplugged or switched off (via DP switching) during the test.

      - Andy.

Thank you Andy I tested the rcd with the circuit breakers off then put the live proton the bus bar yhen earth clipped on cpc bar and neutral on its bar 

Have you tried testing the RCD using the upstream/downstream method (Test Method 2) in the 9th Edition (2022) of IET Guidance Note 3 Inspection & Testing ?

That will help identify if it's an issue with the earth fault path in some way ... or alternatively an issue with protective conductor currents, etc.

Assuming, of course, this is from a grid supply, not a generator less than 10 kVA?

Paul Berryman said:

Its a tt system earth is 205 ohms 

The earth electrode resistance is on the high side (Table 41.5, Note says that readings > 200 ohms may not be stable.?

Hmm, if the circuit breakers are single pole, then the load-side neutral and CPC are still connected during that test, and it is possible that is confusing matters.

Against an electrode of 200 ohms, you could have even a 10 ohm NE fault  perhaps in say an immersion heater or something, and that might not trip the RCD in normal operation, (depends on N-E offset voltage) but would also mean that during the test, 200/210 of your test current would pop down the CPC of the circuit with the fault and come back up the neutral and home via the RCD so it would not register as out of balance.... so No Trip.

Try un-shipping the neutrals as well as the lives (at the breakers), retesting, and then, assuming that then trips OK, then reconnect neutral for each circuit covered by that RCD in turn and retest. The result may be revealing, let us know what you find.

Mike.

As we are in the middle of the hottest June since 1880 odd and have had little to no rain for ages i would think it's fine.  :)

The 200 ohms is not a target for a wet autumn. Guidance would be to aim much lower under those conditions.

Plus, not all the country has a hoespipe ban ... it's not the same conditions everywhere.

There are 2 problems, one that prevents the RCD from operating, the other that is the electrode may or may not really be adequate for continued service. But lowering the electrode resistance may mask the problem that causes the RCD not to trip, as it should operate in the most onerous condition, which is a poor earth, so fix that first. Then go and look at the electrode and the wiring to it. Round here at least (sandy gravelly ground) a figure of 100 ohms or so is quite normal for a 4 ft rod,*

However, it is very possible that the rod or more often the clamp arrangement to connect it is corroded - really you want bright metal on all contacting surfaces, and perhaps a smear of vaseline or similar over the top once completed. (a quick way to do a large area of that is to part dissolve some in white spirit or petrol and paint it on with a brush as a cream/slurry. But be aware that it is not a 'fire safe' technique so outdoors only - even before you add the flammable liquid, a dob of vaseline makes a great emergency  fire lighter, especially on cotton tee shirt type material)

* Note that the RCD tester does not need the electrode to be perfect, and even a Zs of over 1000 ohms, the sort of thing that has a human in the loop, should still trip it - eventually some testers refuse to apply the fault current if they decide the earth voltage rise is excessive.

Mike

Thank you Andy I tested the rcd with the circuit breakers off

As Mike mentioned, that would leave the downstream N and PE connected - so a N-PE fault still seems like the most likely explanation. An insulation test (at 250V if you like) under the same conditions of things being plugged in/switched on as during your RCD test, but with the RCD off of course, will likely show the problem.

The current earthing arrangement shouldn't have any detrimental effect on the test - for a 30mA test it would only raise the touch voltage by around 6V - and just over 30V for a 5x test - both well within the 50V limit most testers accept.

   - Andy.

That wasn't what i meant. If the measurement was taken during a wet autumn i doubt we would be having this conversation. I take the point that not everywhere is bone dry at the moment though.

I'll add that i always presumed that it possibly being unstable meant if the ground were to get very dry as it is round here at the moment. Am i wrong about that?  Putting aside the obvious of checking it is connected properly. 

Gary