Site cabin EICR

What is lowest acceptable resistance we are looking for when conducting an earth electrode test.

Presuming the generator is the source of a local TN-S system, then (from memory) I think BS 7430 suggests around 20Ω.


There doesn't appear to be much maths behind that figure though - and it's the same as DNOs use for MVA systems. So perhaps it's not the end of the world if it's somewhat higher for a ditty generator.


For conventional ADS (fault L-c.p.c) the electrode won't make any difference since the earth fault loop is all metallic on a TN-S system and doesn't involve the electrode. Ditto for additional protection within the unit (presuming extraneous-conductive-parts are bonded to the MET as normal). When it comes to additional protection from direct contact where the victim is in contact with true earth (e.g. where a circuit extends outside of the unit) a high resistance electrode cuts two ways - it'll reduce the fault/shock current which might prevent a protective device from operating, but it'll also limit the current through the victim and voltage across the victim - in the extreme case where the electrode had an almost infinite resistance the system almost becomes a separated system and the shock current becomes negligible. As usual it's the middle cases that are the most troublesome.


Section 717's diagrams make some interesting reading for the various different source arrangements - often 30mA RCD protection is only required for sockets supplying things outside of the unit (not ignoring the general requirements for sockets <=32A to have additional protection).


  - Andy.

It's about the same principle I explained in previous posts

There is a hidden assumption there that smaller gensets are not likely to be supplying more than one load, or at least no more than one burger van or caravan or whatever, and that the whole thing can be seen as a self contained equipotential zone. Larger gensets tend to sit at the centre of a web of outbound cables, and you do not want a fault on one load to make something on the far side of site suddenly dangerous.

I must admit I see the 20 ohms in '7430 as being intended for the case of larger fixed private gensets, usually 3 phase, more like fixed 'plant' than tow along, and the more ' can you trip the RCD?'  based advice of BS7909 a better fit for portable units.

Actually most gensets and inverters below about 3kVA do not  bond the N-E either so the RCD  also becomes pointless, and it has more in common with an isolated supply like a shaver socket.

Thanks you guys for all of your answers.

The site cabin is a self contained construction site welfare unit, powered by an on board 6 kVA generator. The cabin is metallic in construction and sat on timber sleepers. The circuits inside are as usual, lights, few sockets, water heaters and general heaters, but also charging a battery.

Currently there is an electrode installed into the ground connected to the generator via a 35mm2 conductor.

When we undertake the EICR soon I requested that we also undertake an earth electrode test, and I wondered what reading would be deemed acceptable.


I also am wondering what this electrode is actually achieving, seeing as the fault path on any of the circuits will not actually leave the cabin itself??

Hi Paul,


The on-site guide may be of help here. See pg 24-29.

Thanks chaps for all your input. I've got the answer I need now.

OSG 2.4.3  <200 ohms