Strange fault, some clarification needed..

[flyingsparks]

Had a job today, customer reported sparks from an earth wire to the water pipe. When I got there, the main bond to the water had been cut off as there were building works and plumbing going on, and the earth was arcing across a bit to the copper pipe in the ground.

Its a PME supply, but even with the main switch off at the DB still got the arcing. I insulation tested the circuits and had one ring down to 0.03M neutraul to earth. After getting everything unplugged and switched off the fault had cleared , but on plugging things back in, the fault didnt return.. however in the mean time I had my apprentice re do the main bond to the small bit of copper in the ground and run a new 10mm to the new stopcock position (conneted to the old pipe in plastic.)

Have I missed something here a bit obvious, or is it OK. The fact I still had arcing with the main switch off is a bit odd?? It stopped when I temporarily disconnected the main earth from the cut out..

 

[flyingsparks]

Forgot to mention the faulty ring circuit was on RCD, and the RCD had tripped and reset about 10 mins before I got there..

 

[rebuke]

I should point out I'm not an expert here, but it sounds dodgy to me - if there was visible arcing then obviously a not insignificant amount of current is flowing between the PME earth and the copper pipe. If it's happening with the main switch off then it implies that it isn't being caused by a fault making the copper pipe live, so must mean that the PME earth is at a significantly raised potential vs the copper pipe - i.e. perhaps the PME isn't tied down properly.

I assume you're sure it really is PME / TN-C-S, and not e.g. a DIY PME conversion from a TT / TN-S head?

 

[seasickstevie]

Why should it be a 'not insignificant amount of current'?

It's the voltage that does the arcing.

 

[rebuke]

OK yes, my bad - it's having a significant potential difference (voltage) between the PME earth and the copper pipe, such that some (probably quite small) current flows as it is able to overcome the high resistance joint by arcing...

Having said that, if there is such a significant PD, then presumably when it's not arcing and in good contact with a low resistance joint, a not insignificant current will flow...

 

[Stoday]

Why should it be a 'not insignificant amount of current'?

It's the voltage that does the arcing.

Not always.

If the arcing's due to a loose connection, the current, interrupted when flowing through a partly inductive load, will cause high voltage transients and sparks.

 

[flyingsparks]

Pretty sure it is PME and not a DIY TT, however not been in contact with the DNO to confirm supply type. Any more ideas??

 

[rebuke]

Might be worth measuring what the voltage is between the MET and the copper pipe (obviously with the power off as you'll have to disconnect the bonding to do so) - might help to try and identify what's going on...

 

[MalcofArabia]

If you have the double pole switch isolated at your CU then can not see there is a problem with the installation.

You could make doubly sure by removing the L/N from the supply side of the switch and the main earth, that will completely isolate your installation. If you get the same situation then there is a possible fault or bad earth in another installation on the common run of the DNO supply.

You did the right thing by informing the DNO but if it is a fault from another property, then there is not a lot they can do, they certainly won't go to every house and check the installation. If it is a bad earth somewhere then hopefully they may be able to correct that themselves.

 

[flyingsparks]

Thanks for all your replys. Been back today and checked the Ze, which was 0.19, and Zs is 0.12 so that seems ok. I was wondering if it could be a problem with another property. The property is litterally right next to a substation, which probabily has burried earth tapes etc, wether this could be anything to do with it??

 

[Spark123]

How big were the sparks?
It is quite normal to have a voltage between a piece of metal buried in the ground and a neutral conductor (combined neutral and earth), the voltage depending on how much current is being drawn anywhere on the system.

 

[electronicsuk]

The property is litterally right next to a substation, which probabily has burried earth tapes etc, wether this could be anything to do with it??

Quite possibly, if the problem is down to circulating currents in the network as RF describes above. Really you need to disconnect the PEB and measure the voltage between it (the cable returning to the MET) and the pipework, which is at true earth potential. The P.D. between the two should not be particularly high, but there is certainly the possibility for not insignificant amounts of current to flow if there is a unusually low resistance path between the pipework and substation earth tape, which could explain the sparks you're seeing.

If you really are seeing 'arcing' and not just sparks when the conductors meet, i.e. a continuous spark between a discontinuous conductor, then I would be a little more worried. It takes a fair bit of voltage to cause this.

 

[Lectrician]

Probably a high resistance neutral in the supply - do a loop L to N with main switches OFF and earths disconnected from service head.

Then report to DNO.

 

[flyingsparks]

It wasnt like a proper arc, more like sparking as the earth cable was touching the pipe. Unfortunatly I am not going to be there again now for a while, so taking a L-N loop reading isnt going to be possible, unless you think its really that important to be done immediatly. Or do I just get the DNO in now to check it out?

 

[Spark123]

Aren't you in effect doing a L-N impedance check of the supply when doing a Ze check anyway? The only difference is the meter tail vs the earthing conductor however the neutral meter tail shouldn't cause much voltage to be apparent on the MET.