See this video from Chris Kitcher.

www.youtube.com/watch

AJJewsbury said:

Disconnecting the boiler c.p.c. might be an option, but there might be other parallel paths (e.g. from a heated towel rail) back to the MET. Another option is to disconnect the entire installation from its means of earthing - but again there could be parallel paths (e.g. via other extraneous-conductive-parts).

One of my relatives had a water leak last christmas which tripped an RCD on their split board.

I went and cleaned all the water - the short was on an appliance but I wanted to be sure everything was okay and like an idiot I tested IR on each circuit to well and truly open pandora's box.

Turned out there was an N-CPC fault on a ring final which also served the boiler. It was inside the wall so I assume the plasterer did it. What recklessly dangerous step had the electrician taken to prevent? Disconnected the TN-S earth from the MET. So how did the flood trip the RCD, and why were I getting Zs readings of around 3ohms [not acceptable but not open circuit either]? The entire installation was earthed via the 1.5mm CPC in the boilers flex, down the gas pipe, through the gas meter and down the street.

Parallel paths can mask all kinds of issues if we’re not careful.

like the non existent earth electrodes, there are the non-existent insulating joints in the gas network to guard against that.

The problem in all cases is legacy - about half the houses in the UK are pre-war, but it is not an even mix - you get whole estates with lead water pipes and black iron gas mains, and former TNS with the jacket rotted off the PILC, and then whole streets of new houses all with plastic utilities and TNC-s

Even so, if each of the 50- 70 houses per phase on a substation managed a 100 to 150 ohm electrode (the ground is stony round here, if you have London clay, halve that), then we'd be looking at  total electrode resistance of about an ohm. Depending where the break is, there is risk that the substation earth voltage gets pulled more off true terra-firma potential than the broken off section of the network if the latter has a lower electrode resistance.

Mike

Actually some of the clearest pics that I have seen in the wild of a diverted current problem came from John Peckham 

It happens, but I think here at least it is often only partly recognized for what it is, and then silently fixed very quickly.

Mike

mapj1 said:

Even so, if each of the 50- 70 houses per phase on a substation managed a 100 to 150 ohm electrode (the ground is stony round here, if you have London clay, halve that), then we'd be looking at  total electrode resistance of about an ohm.

Hence in an ideal world we should be getting newbuilds to install parameter tape electrodes - down in Greece nearly everything is 3 phase TT. The standard of work is at first glance horrific for me, but then the TT is done so well with Type S RCCB's at the incommer that it's very difficult for a fault to become a thermal hazard - even if the electrician was completely incompetent.

mapj1 said:

Depending where the break is, there is risk that the substation earth voltage gets pulled more off true terra-firma potential than the broken off section of the network if the latter has a lower electrode resistance.

Should note that the substation's earth should be sized to accomodate a fault on the HV side. So it's often huge. However in rural locations I've seen many examples of a 6.6kV pole mounted transformer feeding a single property with the earthing electrode at the property itself providing protection to the pole mounted transformer - my family are originally from cumbria where in fairness to ENW and their predecessors the ground is 5cm of soil followed by rock. An L-E fault on the HV side of that transformer would be very bad news, even though the properties typically don't have any extranous conductive parts except maybe a propane tank.

Round here (SSE) where the transformer is a small pole pig to a few houses or a farm, the HV earth to the transformer core and steels, goes down the pole with the transformer on it, and the next wooden pole along has the earth electrode for the LV/neutral side of things, so they are more then 10m apart, and the LV electrode at least can measure 10 plus  ohms. I don't know, but I'd expect the HV earth to be similar resistance given the construction. Ground mounted substations on housing estates do rather better and have buried mesh and all sorts to get sub-ohm and combined HV  LV earth.  I can well imagine the local soil will have a significant effect ;)

PNB can be a bit prickly and there one does see shared HV-LV earths  on single user supplies.

Interesting to know what happens elsewhere.

Mike

We need to know that in the event of a fault the potential will not exceed 50V under fault conditions save for ADS timeframe when not in a special location.

Although the 50V is only for supplementary bonding - BS 7671 has no requirement at all for what main bonding has to achieve - just a c.s.a. - nothing limits bond length (and so impedance) or indeed specifies the current it needs to cope with. Basically chunk in 10mm² (or whatever's tabulated) and hope for the best.

   - Andy.

Should note that the substation's earth should be sized to accomodate a fault on the HV side.

Indeed - but the HV side is often not a simple TN or TT arrangement, but impedance earthed (with N not distributed) which yields surprisingly modest earth fault currents - sometimes only a few tens of amps - with disconnection then done by a HV device similar to an RCD (but typically directly monitors current returning to the star point rather than having to infer it from the difference between live currents). Where  a HV/LV transformer doesn't have metallic earth connection back to the source (e.g. if there's any section of wooden pole overhead line anywhere upstream), the LV earth tends to be separated from HV earth ("hot" vs "cold" sites in old school terminology, but driven by calculation these days I gather, if with usually similar results).

   - Andy.

AJJewsbury said:

impedance earthed (with N not distributed) which yields surprisingly modest earth fault currents

Good point. They use Z wound transformers, the symbol for which looks like a very troubling political emblem.