Use of ELCB-v with diffrent earth systems.

[ericmark]

From the electric vehicle thread there it seems is a problem where TN-C-S and TT earth systems are close together and some one can be subjected to a steep voltage gradient or even two separately bonded devices of differing voltages in excess of safe parameters under fault conditions.

To isolate under these conditions could a variant of the old ELCB-v be used?

My problem is it is the earth connection which is the danger and to isolate the line, neutral and earth of the PME supply would go against the idea of not having an earth through any breaker.

So thoughts please. It could just switch of supplies leaving earth wires connected but would this remove the danger? It could just ring a bell or light a light to warn.

So where supplies to caravans, cars, or boats require a TT supply but it's close to a TN-C-S supply can some warning or disconnection device be used to stop or reduce the danger of earthing systems of different voltage?

This includes the idea of using a ELCB-v with the two connections across the earth diodes so if the diodes fail the supply would trip.

 

[bernardgreen]

Measuring the voltage difference between local true ground and an imported "earth" will show that there are times when there is a significant difference.

If the local true ground is the earth CPC of a TT system then that CPC will be pulled towards the voltage of the imported "earth". The voltage the CPC reaches will depend on the ratio of the impedance of the ground rod and the impedance of the cable importing the other "earth". This is not going to create a hazard if the TT installation is fully compliant as regards the equipotential zone.

The hazard is when the ground rod is very low impedance. In this case very high currents will flow along the cable from imported "earth" to true ground when the imported "earth" is not at true ground potential.


Provided both CPCs remain connected to an effective "earth" and equipotential zones are inplace then disconnecting the link between the two earths may be needed to reduce the risk of fire from an over heating earth ( bond ) cable when the two CPC are at different potentials. Measuring the amperage flowing in the bond wire would be easier than trying to measure the voltage drop along the bond wire.

 

[JohnW2]

From the electric vehicle thread there it seems is a problem where TN-C-S and TT earth systems are close together and some one can be subjected to a steep voltage gradient or even two separately bonded devices of differing voltages in excess of safe parameters under fault conditions.

I would say the the theoretical problem is more general than that - not a problem of "having a TN-C-S and TT earth systems in close proximity" but, rather,of "having a TN-C-S earth in close proximity to true earth" - whether that 'earth' be part of a TT electrical installation or merely a conductive object in contact with the ground (you often mention metal rainwater/waste pipes yourself).

To isolate under these conditions could a variant of the old ELCB-v be used? My problem is it is the earth connection which is the danger and to isolate the line, neutral and earth of the PME supply would go against the idea of not having an earth through any breaker.

Precisely. The regs would have to be changed dramatically to make that permissible. These days, it would be very easy to produce a nearly-perfect 'true VO-ELCB' (i.e. one which was truly voltage-operated, not reliant on current through relatively-high-impedance coil in the earth loop) which sensed the pd between the installation (e.g. TN-C-S) 'earth and some 'true earth' reference (e.g. a rod). However, it would only be permissible for that device to interrupt the L and N conductors (after splitting of the PEN conductor) - which, as you say, would leave the primary hazard (the 'earth') still connected.

So thoughts please. It could just switch of supplies leaving earth wires connected but would this remove the danger? It could just ring a bell or light a light to warn.

As above, if the problem were a substantial rise in PEN, hence 'E' as well as N, potentials above true earth, then I can't see that having a device which interrupted just L and N would really help to remove that hazard (in situations when something close to 'true earth' was in proximity), but to have the E interruped by the device would be seriously contrary to current regs. The device could, indeed, as you say, be used to activate some warning alarm (without interrupting the earth connection), which would not be a problem in terms of regs - but an alarm is clearly not as idea as 'removing the hazard'.

I can't see regs about not having any 'switch' in the earth loop being in any way 'relaxed' any time soon. A highly reliable 3-pole (L, N & E) device or even a DNO-installed 2-pole (L & PEN) one near the origin of an installation should theoretically be harmless enough provided that the E (or PEN) path through them (when not operated) could be made incredibly reliable - but that's likley to be the sticking point.

As things stand, if there is a concern about the theoretical hazard, then the 'simple' solution is obviously simply to avoid exporting a TN-C-S 'earth' outside of its equipotential zone, certainly to any place where there are easily-touched routes to true earth.

Kind Regards, John

 

[bernardgreen]

the 'simple' solution is obviously simply to avoid exporting a TN-C-S 'earth' outside of its equipotential zone, certainly to any place where there are easily-touched routes to true earth

It is also necessary to avoid importing into the equipotential zone any low impedance conductiv paths to true ground.
The regulations require any extraneous conductors to be bonded. This is OK and works well for most extraneous conductors which are medium to high impedance to true ground. ( plastic water mains, copper pipes to a nearby oil tank etc ) The 10mm ( or 4 mm ) bond cable can cary the current without over heating.

When the water service main is metallic all the way to a metallic street main the impedance may be less than one tenth of an ohm. A one volt difference between the earth systems will create a 10 amp current in the bond. In extreme cases bond wires have overheated and even melted. ( metal theft open circuited the network neutral and the water mains were all metallic, bond wires melting in six kitchens started small fires, a gas pipe was also affected and an explosion occured )

 

[JohnW2]

It is also necessary to avoid importing into the equipotential zone any low impedance conductiv paths to true ground.

That's obviously a different matter, and is totally unrelated to the issue/problem which eric was discussing.

However, as you go on to imply, one cannot 'avoid' that in the way you suggest. If there is a very low impedance path to earth entering the property (e.g. a metal water supply pipe connected to a metal water main), hence the 'equipotential zone' of the property, then that is a fact which one can do nothing about (unless one can persuade the water supplier to put in an 'insulating insert' outside of the property) - since the regs would require it to be bonded. In that situation, the only way one could address the issue which concerns you is to get rid of the TN-C-S earth and instead convert the whole installation to TT (with a corresponding 'TT-referenced' equipoential zone).

In reality, a 10mm² main bonding cable will (as far as I can make out from the tables) carry at least 500A, probably more, before melting, so serious issue is only likely to arise in the sort of extreme situations you describe, in which the incoming neutral (PEN) potential rose to a very high potential above earth. If that arose, due to a network problem, it would usually affect several properties - and if several properties were each drawing extremely large currents of that order through their bonding conductors, I would hope that the DNO's OPD (at transformer or wherever) would operate before any bonding cables melted or caused fires.

Kind Regards, John

 

[plugwash]

hence the 'equipotential zone' of the property, then that is a fact which one can do nothing about (unless one can persuade the water supplier to put in an 'insulating insert' outside of the property)

I was under the impression that the water pipe from the outside stopcock to the house belonged to the householder not the water company. Is that impression wrong?

 

[JohnW2]

hence the 'equipotential zone' of the property, then that is a fact which one can do nothing about (unless one can persuade the water supplier to put in an 'insulating insert' outside of the property)

I was under the impression that the water pipe from the outside stopcock to the house belonged to the householder not the water company. Is that impression wrong?

It's confusing, at least to me. I think there may well be a difference between who the pipe 'belongs to' and who is responsible for it's repair/mainenance (and/or 'who is allowed to do things to it')..

I may be wrong, but I suspect that most water suppliers would be unhappy if a house owner dug up the pipe between an outside stopcock (or outside meter) and replaced it with their own pipe (plastic or otherwise) which they connected to the suppliers stopcock or meter.

However, I'm far from certain about this. Even with my own house, when I last had an Anglian Water man out here to try to resolve the question, he could not work out, or tell me, "who owned what" and "who was responsible for what"!

Whatever, in terms of the point I was making, one can only 'do domething about' a very low impedance extraneous-conductive-part if 'whoever is allowed to' changes it to plastic (or, at least, puts in an 'insulating insert') before it enters the property.

Kind Regards, John

Kind Regards, John