Melted on earth sleeving on a ring circuit

[JohnW2]

CPCs can over heat due to high currents created by the CPC connecting two different "Earths" together when those two "Earths" are at different potentials. A cable bonding a metallic supply pipe to the MET of a PME system can carry very high currents when the PME Neutral is pulled a few volts above local ground potential by un-balanced loading on the three phases in the local netwrork. ... Connecting the CPC of a PME system to a boiler, electric shower or other device that has metallic connection to a low impedance Ground will result in current in the the CPC when the Neutral is not at local Ground potential.

An interesting thought.

However, I would have thought that, in the absence of a 'lost neutral' in the supply network (which would only be brief, until it was rectified), and in the presence of (deemed to be) 'adequate' main bonding, it would take a fairly extraordinary set of circumstances for the current through a (say) 1.5mm² CPC to be high enough to raise its temperature enough to melt PVC. After all, the path through that CPC to some 'true earth' would, in such a situation, be in parallel with both the supply neutral and the main bonding, so only a fairly small fraction of the total current would flow through the CPC.

Kind Regards, John

 

[bernardgreen]

I did differentiate between a bonding cable between MET and a metallic supply pipe and a CPC . If that bond cable was in place then as you say the CPC would not be carrying the entire Ground to Neutral current.

"Lost Neutrals" are not the only incidents that drive the supply Neutral and hence also the MET to a potential several volts above local Ground potential. Un-balnaced loads on the three phases in the local network can drive the Neutral away from true ground potential.

If the impedance of the copper ( CPC and/or bond ) between MET and true Ground is for example 0.1 ohms then a 3 volt difference between MET and true Ground CPC will create a 30 amp current through that copper. A bond 6mm² or 10mm² should be OK with 30 amps but a 1.5mm² CPC on it's own could be at risk if the voltage increased to 5 volts

 

[AndyPRK]

So that’s why earth cables have to be so thick these days. So they don’t melt !

 

[reds42]

There was a problem with a ring final at my workplace the other month tripping out.

A socket had overheated due to a bad conection and the high load of a tea urn running off it. The heat was on the neutral connection and the cable got hot enough that it melted through its insulation and the earth sleeving of the cpc pressed up against it in the backbox.

It wasn't a dead short and caused intermitant tripping of the RCBO.

 

[Harry Bloomfield]

I'm not sure, but mention of current running through the cpc makes me think that maybe there is a fault somewhere which is insufficient to trip but is heating up the conductors.

Just a look-see at some sockets and the inside of the board would be interesting.

A neutral acidently swapped with an earth at a (twin?) socket and a substantial load on the socket or sockets, would load up the CPC and maybe cause some heating. The result, could then be damage to the live conductor, within the T&E, enough to cause the MCB to trip.

Every socket on the circuit needs to be removed and inspected, plus the insulation resistance checked.

 

[JohnW2]

I did differentiate between a bonding cable between MET and a metallic supply pipe and a CPC . If that bond cable was in place then as you say the CPC would not be carrying the entire Ground to Neutral current.

Exactly my point.

"Lost Neutrals" are not the only incidents that drive the supply Neutral and hence also the MET to a potential several volts above local Ground potential. Un-balnaced loads on the three phases in the local network can drive the Neutral away from true ground potential.

Indeed so. However, taht's why I said that in the absence of a 'lost neutral' and if (deemed to be) 'adequate' main bonding was in place, then it would probably take a fairly extraordinary set of circumstances for the current through a (say) 1.5mm² CPC to be high enough to raise its temperature enough to melt PVC ....

.... If the impedance of the copper ( CPC and/or bond ) between MET and true Ground is for example 0.1 ohms then a 3 volt difference between MET and true Ground CPC will create a 30 amp current through that copper. A bond 6mm² or 10mm² should be OK with 30 amps but a 1.5mm² CPC on it's own could be at risk if the voltage increased to 5 volts

... again, not really an issue if correct 'standard' main bonding was in place. Even if the supply neutral was 10V above true earth (probably unlikely in the absence of a network fault), the current flowing from neutral to true earth (100A with your figures) would be split roughly in the ratio 1.5:10 between a 1.5mm² CPC and a 10mm² main bonding conductor - so something like 13A through the CPC, which is well within its 'normal limits'. Indeed, even if we got silly and postulated an N-E pd of 20V, the current through the CPC would still only be of the order of 26A, which surely would not result in melting insulation, given that a ('Methoid C') 1.5mm² conductor is deemed to be able to safely carry 29A for at least an hour.

Kind Regards, John

 

[securespark]

Not quite the same, but I came across this a few years ago.
https://www.diynot.com/diy/threads/guess-the-fault-competition.156922/#post-1059369

 

[sparkwright]

All wiring done in 2007, T+E cables. No 30mA RCD protection anywhere.

Not my job, another electrician's job.

Will let you know how he gets on.