Earth Rod Resistance (ARW?)

[Dave Liquorice]

What is considered "good" or "low" for the resistance of an earth rod? Rod to ground as tested using multiple temporary test rods.

[dennis]

It needs to be low enough for a live to earth fault to trip the breaker in the maximum time allowed. If I had the figures to hand it would be easy to work out. I would be aiming for low milliohms rather than ohms.

[A.Lee]

100 Ohms is just about the limit for a stable resistance, so you really want to get below that. In real life, you can probably get less then 20 ohms with 2 rods either joined together going 2 metres deep, or in separate areas. Of course, it all depends on what your ground is like around the rods.

And if Dennis can get his down to milliohms, then he really should be lecturing on how to do it.

[dennis]

!00ohm will give you a fault current of about 2.5A!

20 ohms is going to give a fault current of about 12.5A!

Is it worth having an earth that poor? The breaker will never blow and the "earthed" cases will be live at near mains voltage.

You can always change it.

[Cursitor Doom]

This chap here has made a very good video on the subject:

[Tim Watts]

Sorry to interject, but that is completely incorrect.

a) You will not get an earth rod Ze to low milliohms, even if you bury it in marconite;

b) The aim is to keep the touch voltage down to safe limits up until the RCD (which there must be) trip current is reached.

The exact allowable earth loop impedance is a matter of debate - there is an upper bound which is supply-voltage/RCD-tri-current but a lot of leeway needs to be added as the ground in which the rod is buried is an unstable and slightly unpredictable medium.

Here is a bit of reasoning:

[Tim Watts]

Dennis - RCD. Or ELCB before that.

No TT installation in the history of man has ever been able to achieve disconnect times with fuses/MCBs along.

Not unless you maybe live in a salt water marsh...

[ARW]

Why ask me when Alan knows the answer?

I would say 200 Ohms is the max but this could depend on the ground dampness when the tests are done. So a 100 Ohm limit when the ground is wet is a good limit -or to use Alan's word "stable resistance".

And of course dennis is talking s**te.

[ARW]

Who bought you a Ladybird book of maths?

[dennis]

Yet they existed before elcb or rcds.

[John Rumm]

In your dreams...

Sub 10 ohms would be "good" in my book. However even a few hundred will still be adequate to trip a RCD and have some "spare" capacity left over.

[tony sayer]

Christsakes Den where is the world you live in eh:?..

Milliohms;?...

[Tim Watts]

Oh do stop talking bollocks.

[tony sayer]

One of a series that Dennis will be very well advised to study very carefully!...

[JimK]

/>> 100 Ohms is just about the limit for a stable resistance, so you really

Who bought you a Ladybird book of maths? /q

Was it a genuine one?

Jim K

[Dave Liquorice]

Great, so 20 ohms with a single rod (unknown length) is better than "good".

'cause you 'ave bits of paper saying you should know the correct answer (maybe Alan does as well?).

Bit of context this rod is for the ground connection to a mobile generator that is powering mobile Outside Broadcast equipment. At one frequently visited location we suffer hum on audio and video if the OB equipment chassis makes electrical contact with the nearly 100 year old steel structure. This hum problem has got worse since a lot of new lights and steel work was added to the old structure. One would assume that the new stuff has been installed to the latest regs and thus all the new steel work should be bonded to the grid supply MET with its own bond(s) and/or via the PEs of all the new lights.

As the tech earth rod is 20 ohms that seems OK (and is the same at other locations where we don't get hum). So that leaves the old steel struture probably "floating" possibly with leakage from old wiring or faults or equipment or the grid supply MET not being up to scratch. The grid supply could well be (trying..) to use the lead sheath of a

100 year old incomer.

Or not making the distinction between an earth rod and MET? Doesn't the ground impedance of a MET have to be very low so that decent sized fault currents can flow back to the (remote) power source?

Define "remote power source", is that the central start point on the secondary side of the 11 kV to 415 V transformer or further up the chain of supply?

[dennis]

Old buildings can have all sorts of corrosion at the joints in the frame. They can even rectify radio signals that are picked up by the frame.

When you say touch the frame, why aren't they insulated? It would solve the problem.

[Andy Wade]

Actually he's right, they (TT systems) did /exist/. The trouble was that they didn't actually /work/ until ELCBs came along.

[Tim Watts]

Well, I don't dispute that back in 1880, they did dodgy stuff. I was answering more in recent contexts :)

[Andy Wade]

However perfect (or not) the building's installation there is bound to be a voltage between the mains earth at the MET (bonded to the structure), and your clean earth. There's always leakage present putting current into the mains earth - cable capacitance, EMC filters and so on - and not just from 'your' building, from everything that shares the same mains cable. There's also diverted neutral current in that cable, even if the supply is notionally TN-S - wherever new connections and repairs are made, the DNOs usually bond neutral and earth. The cable in the street is PME, even if your supply isn't, and the incoming earth is usually a few volts above the local ground.

Any connection between the metalwork of equipment earthed 'locally' and the structure will allow current to pass, and it will go through the chassis, internal wiring, PCB groundplanes - whatever routes it can find, in accordance with the laws of Ohm and Kirchhoff. It's a big hum loop - as you've found. Ironically the better your earth rod, the worse it will be, since more current can flow.

I suggest you try a hefty bond (16 or 25 mm^2, say) between the MET of your gen-set and the building steelwork (or the MET at the intake). Either that, or maintain 'galvanic separation', which might not be easy.

MET is just an abbreviation for main earth terminal, the point where the earth conductor (connected to the means of earthing), the circuit protective conductors and the main bonding all meet. The earth fault loop impedance (Ze) will depend on the type of earthing, being relatively high for a TT system, compared to the TN systems

For the present discussion, yes.