Earth rod required

[Mark McGee]

Hi

I'm running some cable to a consumer unit in some outbuildings, and I vaguely recall reading something about needing an earthing rod.

Should I earth the consumer unit to an earth rod? If so, what do I do with the earth on the supply cable?

Do I just bash the earth rod in to the ground outside, and run some (6mm?) earth wire in to the consumer unit from there?

Thanks Mark

[Wheelbarrowbob]

It is a good idea to make your outbuildings a TT system which as you say requires an earth electrode but use at least 16mm. As it is a TT system your soket circuits need to be RCD protected to limit fault current although it is not necsarry to rcd protect your lighting circuit. You need to test your spike and it must be below 200 ohms, if not stick 2 in and retest. ( 200 ohms earth fault loop impedence can result in high fault current before a mcb interupted supply hence an RCD ) >Should I earth the consumer unit to an earth rod? If so, what do I do

If you are using SWA just gland off the armour at each end and earth the cu to the rod. How far is the cable run and how are you getting the supply to the outbuilding, as you may need some cable calcs for volt drop

[Lurch]

On 26 Apr 2004 05:36:58 -0700, in uk.d-i-y [e-mail] (Mark McGee) strung together this:

You don't 'need' an earth rod, but you can use one, recommended if the outbuildings are some distance from the house.

Yes, run an earth wire of 4mm cable if mechanically protected or the same size as the supply cable live conductors if not. Connect it to the earth bar in the consumer unit.

Terminate it in a plastic adaptable box if it is an SWA and earth it at the house end. If it is T&E make the end safe within the CU.

Pretty much, you'll need an earth loop impedance tester to test the effectiveness of the rod and add more or extend the one that's there until the Ze is low enough to comply with the relevant regulations.

[John Rumm]

You probably won't need 16mm for the earth conductor (the earth loop impedance of the spike is unlikely to be low enough to pass that much current through it).

I think you have your "high fault current" and "high earth impedance confused".

A high earth loop impedance (i.e. a "poor" earth) will result in LOW fault currents - that is the problem. In a setup with a good earth (typically one provided by the electricity supplier), should a fault occur that allows the exposed metalwork of an appliance come into contact with live, the current that flows to earth will be massive - this is a "good thing" because it will cause the protective device (fuse, MCB etc) to operate and cut off the current.

With a TT system that has a high earth loop impedance you can not rely on this form of protection because the earth impedance may limit the fault current to a level where it will not activate the protective device.

For example with a fault loop impedance of 100 ohms, your maximum fault current on a live to earth short will be 2.3A - not even enough to trip a typical lighting circuit breaker. This is why TT systems will need an RCD in addition to fuses / breakers.

or use an insulated (i.e. plastic) consumer unit.

[Mark McGee]

Here's a bit more info (and some more questions!);

I'm actually running two separate cables, one to an outbuilding about

20 feet from the home CU (cable runs above ground, along outside of house - this cable was already installed - 6mm T&E, I think). The other is a 10mm 3-core SWA to a workshop, which I just laid, the cable run is about 100-120 ft from the home CU, and runs underground for the majority of the run.

I've obtained some earth rods for both installations.

I'm still not 100% sure what to do with the earth wire on the cables. Do I earth them at the house end, and just ignore them at the w/s or outbuilding end, and just hook-up the workshop CU earth to the earth rod?

Currently, in a special (and very expensive!) junction box, I've joined the earth core wire with the outer cable shielding. It's not live yet.

Can I not test the earth impedance with a multimeter?

Thanks, Mark

[John Rumm]

Yup that is about it. In many cases you would use 2 core SWA and just earth the screen. Terminating in an insulated CU at the workshop end would then neatly drop the connection to house earth leaving you free to connect your earth rod.

(the setup in my house (came with the house - not installed by me) actually does the reverse - connects the screen of the SWA to the workshop earth rod via a metal clad CU and then isolates it from the house at the house end of the cable)

You can either earth the third wire or ignore it. The main purpose of the earth in the SWA is to give a fault current return path should someone manage to get a spade through it.

Alas no. The test needs to check the impedance when connected to AC mains voltage to get a true picture of how it will behave in real life.

[Wheelbarrowbob]

Sorry , let me clarify my thoughts ( I am at moment doing year 2 of 2361, so am trying to get my head around a lot of issues ) I understand the maths and recently done earth loop testing and with a high earth fault loop impedance on TT system if a fault occurs as you say with 100 ohms impedance 2-3 amps will flow and protection will not trip, but as a 15 mA shock across the chest could be fatal I am basing a 3A fault on metalwork or pipes as a dangerous current flow. As for the 16 mmsq earth cable i thought the earth cable on TT system had to be a min of half the supply size cable but I may be confused by main equipotential bonding cable to pipes. But thanks for the reply.

[John Rumm]

Not sure I follow that. The current flowing through exposed metalwork is not in itself relevant from a shock point of view. The voltage present on it however may pose a risk. To mitigate this there is a requirement that your RCD trip current is low enough to ensure that a voltage of no more that 50V can appear on an earth conductor. So in the case of a

100mA RCD with your 100 ohm fault loop impedance, you could get a maximum voltage drop of 10V drop. Put it in other words, the earth conductor could rise to 10V above ground potential before the breaker tripped. Hence that setup would be safe. However a 330mA RCD with a 190 ohm earth fault loop would give a voltage on exposed metal work of over 60V

It is a dangerous current flow in the sense that it is too small and will not trip a breaker - not that it is a shock hazard.

Have a look at table 10C in the on site guide. That seems to suggest a conductor size of as little as 2.5mm sq for situations where the wire is not burried and is protected against mechanical damage, but upto 25mm sq for a burried unprotected conductor.

[Mark McGee]

Thanks John. I have an electrician coming over soon to install a new CU indoors and a few other jobs. I'll get him to do the earth loop impedence test.

Thanks again Mark

[Wheelbarrowbob]

I did not know that, but am assuming if there is a potential difference on a cpc there will be a current flow.>However a 330mA RCD with a 190

A 330mA RCD seems quite high to me are you refering to the old ELCB that used to fitted ?>Have a look at table 10C in the on site guide.

OSG is in van will look at 10c 2morrow, Once again thanks for the reply

[John Rumm]

I was using 330mA to exagerate the point that either a high leakage current or a high earth impedance could result in a dangerous voltage being present on the earthed metalwork. Having said that you can get

300mA+ RCDs:-