Shakey

depends what you mean by 'do away with'

No value would allow you to do away with an unreliable earth

Or are you refering to the value of 200 ohms for TT earth electrodes?

This is the max value at which it could be condidered a 'stable reading', but is not a max value as far as the RCD is concerned

there is no 'leeway' for an unreliable earth

Dinosaur

No Earth path - No electric shock???????????

tony.towa

Technically an RCD will operate without any form of earth path being installed. It operates. as stated in other answers, by monitoring the phase and neutral and reacting to an imbalance between them. Normally this imbalance is as a result of an "earth fault" be it to a back box, cpc, or as a result of external influences like moisture. It is possible to have a created imbalance, for example where wall lights have been installed and the live is taken from the lighting circuit and the neutral is taken from a socket on the final ring circuit protected by an RCD (16th edition style split load board with protected and unprotected ways). By switching the wall lights on an imbalance is created on the final ring circuit which will trip the RCD. Whilst not a common fault it is a typical "lash up" job by a "Barry the Bodger".

Shakey

yup, done a few like that Tony

"just installed a new circuit, RCD trips as soon as i switch it on"

Investigation finds that the guy didnt know there was seperate neutral bars

Doh!

greekislandlover

Not sure certain CU manufacturers are aware of this fact either! I've found several new CU's to have the neutrals connected arbitrarily to whatever bar suits at the time of manufacturer.

Carter

Modern (current controlled) RCDs make the assumption that any imbalance in line and neutral currents must be flowing to earth (because generally there isn't anywhere else for it to go) they don't actually measure anything with reference to the main or cct cpc's. However it still doesn't remove the fact an earth path must be present or there would be no residual current and the device would not operate.
I think the older voltage controlled ELCBs sensed voltage on the cpc but they were effectively proscribed in 1981 as it was realised that a parallel earth path would render them inoperable under true fault conditons.

Ask your man to draw out the FULL fault path and ask him were exactly does he thinks that inconvenient stray, 'residual' "I delta n" tripping current is going.

weyboy

what about disconnection times?