earth for pond ?

[ac1951]

Friend of mine as a fairly large pond circa 2,000 gal. From an electrical point of view it has a double insulated submersible pump and a UV light that is housed in a plastic enclosure that the water is pumped through on its way to the filter before returning to the pond.

These are supplied on a separate RCBO circuit from his CU. A 2.5mm SWA twin cable runs from the CU to a junction box ajacent to the pond and whilst not used the earth (via the SWA) is available for connection. The house is supplied from a PME overhead supply.

He asked me if I thought the water should be earthed (the pond is lined with a PVC liner so insulted from earth) and if so should he dangle a length of copper wire into the pond with the other end either connected to the SWA earth terminal or drive an earth rod into the ground and connect it to that.

opinions most welcome.

thanks Andy .

[John Rumm]

The short answer is a definite "no". If you do need an earth for something, then you would need to make the installation TT at the pond end of things. Bad things can happen if you attempt to export a PME earth into a situation where you cannot also extend the equipotential zone that should go with it (in this case because you have ready access to an independent earth).

[The Medway Handyman]

I was going to say that :-)

[Andy Wade]

The liner might be insulting, but it will be fairly thin and PVC has a dielectric constant of around 4 at 50 Hz. The capacitance of the water mass to earth will be quite enough to stop the pond floating away, electrically speaking...

The former would be positively dangerous. The latter would be harmless, but is quite unnecessary. There's already Class 2 equipment and RCD protection.

[ac1951]

Thanks all, I will pass on your valuable comments.

[mark]

How rude of it!

mark

[Martin Bonner]

I don't believe you :-)

[John Rumm]

Why? ;-)

[geoff]

it gives his life a purpose

[The Medway Handyman]

Neither do I :-)

[funkmish]

For those of us "not in the know" - why would it be dangerous? What could happen?

[Tim S]

funkmish coughed up some electrons that declared:

Depending on the type of supply, some faults can cause the "earth" in the house to rise a considerable number of volts/tens of volts/hundreds of volts above true earth. In the former case, the pond water now has a significant potential difference with the surrounding ground.

Granted, such faults are probably not common, but in this case you have done nothing to improve the safety of the situation and everything to increase the potential danger.

Earthing is a hairy issue at the best of times and bringing one out of the house equipotential zone is even hairier.

Cheers

Tim

[John Rumm]

The supply to the house in this case is a TN-C-S one. This means that the earth and neutral are connected together where the supply enters the house. While this has the effect of providing a very good earth within the property, there is a failure mode you need to consider: what would happen if the combined neutral and earth conductor was disconnected outside of the property for any reason? This would leave the main earth terminal in the house "floating" - it would however be connected to the live of the supply via any appliances connected in the house. Since there is no neutral connection to pull the voltage on the earth down, all the earthed metalwork in the house would now be sat at 240V.

One of the ways that the dangers of this situation are mitigated is via equipotential bonding - anything that has the capability of introducing a potential into the house (and that includes anything independently connected to real earth like gas and water pipes), is electrically bonded together. That should mean that in the event of the fault described, the bonding will have two effects: firstly any exposed metalwork you can come into contact with will all be at the same potential - even though this may be a high voltage with respect to true earth, you can't touch anything at a significantly different potential and hence not receive a serious shock. Secondly, the fortuitous earthing gained from any of the bonded services will tend to reduce the actual voltage.