AJJewsbury:

There's another failure mode for TN-S - these days with even domestic installations having many electronic appliances it's common for the standing earth leakage (sorry, protective conductor current) to reach dangerous levels (e.g. 50mA or more) - so a broken PE in the supply can make the installation's metalwork hazardous in much the same way as a broken PEN conductor in a TN-C-S system, if with somewhat lower currents involved.


If you want a less hazardous earthing system - there was one approach suggested (I think it was in one of the Cahiers Techniques) that introduced a deliberate resistance between the supply star point and Earth (like an IT system) but with the consumers' earthing system directly connected to the supply earth electrode (as in TN-S). The result of an L-PE fault is then that the voltage on the insulated live conductors goes somewhat awry w.r.t. true Earth but the earthing system itself remains at pretty much true earth potential. Earth fault currents are very low, so very negligible potential differences between even widely spaced parts even during a fault. RCDs are used automatic disconnection - but only to provide reliability - as they're not needed for ADS to provide shock protection.


   - Andy.

Well, true, however a 25 ohm earth electrode might be able to bring that voltage down- 2,400 ohm resistor in series with a 25 ohm resistor- 2.4 volts to remote earth.



The thing about IT systems is that more often than not they end up with a standing earth fault, making them a TN system. During the second fault however, the loop impedance doubles in that the fault current must travel through one circuit, to the MET, leave the MET onto the second circuit and then return via the 2nd phase conductor.


Of course you bring up a good point, having an RCD prevents this. And even if one fails, the one on the second faulted circuit would trip instantly.


In any case all the earths at each consumer should be connected together.


It is interesting: Norway used to use IT earthing extensivelly. They however made the mistake of not connecting all the METs together, just a local earth rod. This was ok until phase A faulted in one building, then phase B in another building. This resulted in 230 volts potential between structures which caused in many fires. Thankfully RCDs in existing buildings are correcting this.  

AJJewsbury:

There's another failure mode for TN-S - these days with even domestic installations having many electronic appliances it's common for the standing earth leakage (sorry, protective conductor current) to reach dangerous levels (e.g. 50mA or more) - so a broken PE in the supply can make the installation's metalwork hazardous in much the same way as a broken PEN conductor in a TN-C-S system, if with somewhat lower currents involved.


If you want a less hazardous earthing system - there was one approach suggested (I think it was in one of the Cahiers Techniques) that introduced a deliberate resistance between the supply star point and Earth (like an IT system) but with the consumers' earthing system directly connected to the supply earth electrode (as in TN-S). The result of an L-PE fault is then that the voltage on the insulated live conductors goes somewhat awry w.r.t. true Earth but the earthing system itself remains at pretty much true earth potential. Earth fault currents are very low, so very negligible potential differences between even widely spaced parts even during a fault. RCDs are used automatic disconnection - but only to provide reliability - as they're not needed for ADS to provide shock protection.


   - Andy.

Well, true, however a 25 ohm earth electrode might be able to bring that voltage down- 2,400 ohm resistor in series with a 25 ohm resistor- 2.4 volts to remote earth.



The thing about IT systems is that more often than not they end up with a standing earth fault, making them a TN system. During the second fault however, the loop impedance doubles in that the fault current must travel through one circuit, to the MET, leave the MET onto the second circuit and then return via the 2nd phase conductor.


Of course you bring up a good point, having an RCD prevents this. And even if one fails, the one on the second faulted circuit would trip instantly.


In any case all the earths at each consumer should be connected together.


It is interesting: Norway used to use IT earthing extensivelly. They however made the mistake of not connecting all the METs together, just a local earth rod. This was ok until phase A faulted in one building, then phase B in another building. This resulted in 230 volts potential between structures which caused in many fires. Thankfully RCDs in existing buildings are correcting this.