TNC-S Loss of Neutral

[JohnW2]

There is no gas pipe to the bugalow, just water and oil. The water pipe is plastic and the oil pipe is copper and just goes to my own oil tank about 10m away for heating purposes. The oil pipe is earthed as normal where it enters the bungalow and is run in a protective ducting as well as being plastic coated. I don't expect that the copper of this oil feed to ever touch ground and therefore should not accidently make an intended TT system to be TNC-S. ... Does that make sense so far?

Yep, and fair enough.

I am more convinced now that I need to use one or more earth rods to provide an earth if the incomming TN is broken.

That's you're decision - but it's perhaps worth pointing out that there are obviously millions of TN--C-S installations out there whose owners are not even aware of the potential issue, let alone done anything about it.

At the moment I am thinking about using a number of earth rods to connect to the incomming TN connection. That way I get a fairly low earth impedance (to be measured and I would be surprised if it was anywhere near 0.35-ohm), which in normal operation will provide a good fault path under fault conditions to ensure that the RCBO's trip etc. .... Any ideas on what a good target earth rod impedance should be would be appreciated.

For a start, it's incredibly improbable that an earth fault in your installation would arise at the very same time as you had 'lost the TN neutral'. However, even in the event of that almost unbelievable co-incidence, it's not really the operation of RCDs/RCBOs (under fault conditions) that you need to worry about, since that only requires an earth impedance under about 7,667 Ω, which it would be very difficult NOT to achieve with any sort of earth rod (2" nail? ).

... so I should be able to get a fairly low impedance. But I don't want it to be too low since the incomming TN connection may vary relative to true ground (depending on my loading and other users loading) which could mean a continuous earth current would flow through my earth rods. If for example TN was say 10V above true ground and the earth rod impedance was 100-ohm, this would be a current of 100mA continuously.

There's nothing wrong with your arithmetic, but I don't see that as a problem. If you had an incoming metal water pipe (and bonded it, as required), it's impedance to earth could be very much lower than 100 Ω, hence the continuous current very much above 100 mA, without your being able to do anything about that.

Of coarse, if TN was losted, TN could be 240V (presumably) with an earth rod current of 2.4A.

All true, but you would probably be aiming for an earth electrode of much lower impedance than 100 Ω, in which case the current under such fault conditions would be much higher than 2.4A. Indeed, if the earth impedance got anywhere near the 0.35 Ω, then that would represent an 'adequate' return path to the transformer, in which case your installation would function as 'normal', with all of what would normally be 'neutral current' flowing through the earth rod(s)!!

Kind Regards, John

 

[bernardgreen]

The orange path is the only way the Neutral currents from the loads in the houses can return to the sub-station. The purple link Rod to CPC in the house is the critical weak point.

 

[JohnW2]

The orange path is the only way the Neutral currents from the loads in the houses can return to the sub-station. The purple link Rod to CPC in the house is the critical weak point.

That corresponds with what I wrote.

However, as I said, and unless the CNE fault is at or close to the transformer (e.g. due to 'metal theft', seemingly fairly improbable with 'a transformer in someone's garden') in the OP's situation it would seem that any such fault would be most likely to arise downstream of the last connection to another installation, in which case the only current which had to return to the transformer would be that due to the OP's installation - and, in that situation, even if the earth rod (or a bonded extraneous-c-p) had a (very improbable) extremely low impedance to earth, it's unlikely that a TN-C-S-compliant earthing/bonding conductor (i.e. ≥10 mm²) (your 'purple link') would come to any harm with a domestic supply.

However, as I also said, literally millions of people have TN-C-S installations without even being aware of this potential issue, let alone 'doing anything about it'. Even if a local earth rod becomes a requirement for TN installations, the likely impedance of any real-world domestic rod is unlikley to have appreciable impact on the situation, or be asked to carry any large current. The greatest potential risk exists if there are very-low-impedance (bonded) extraneous-c-ps, in which case addition of an earth rod would essentially make no difference at all.

Kind Regards, John

 

[flameport]

The water pipe is plastic and the oil pipe is copper and just goes to my own oil tank about 10m away for heating purposes.

With that and the other description of the property, making the whole thing TT and not using the TNCS earth at all would be a far better option.

a good target earth rod impedance should be

Under 100 ohms would be desirable, and certainly easily achievable in most situations.

earth rods

Earth rods are one cheap and not particularly good option.
There are much better ways.

 

[JohnW2]

...Under 100 ohms would be desirable, and certainly easily achievable in most situations.

That's certainly the 'traditional wisdom', and I think probably relates to the fact that anything higher might indicate an iffy situation which might progressively deteriorate with time.

From the point of view of protection (by 'ADS') against L-E faults (of the usual 'negligible impedance' persuasion), I can't see that there is really anything to choose between, say 5,000 Ω and 10 Ω, since neither would facilitate operation of even a B6 MCB, whilst even the former would facilitate operation of a 30mA RCD/RCBO.

Kind Regards, John

 

[mattylad]

As Flameport was suggesting, something far far better than earth rods would be a condudisc and some conducrete over it.
Look on Youtube for reviews and info about them.

IMO well worth the extra cost.

 

[JohnW2]

As Flameport was suggesting, something far far better than earth rods would be a condudisc and some conducrete over it. Look on Youtube for reviews and info about them. IMO well worth the extra cost.

There are certainly alternatives to rods that would/could result in much lower impedances to earth.

However, in situations like the OP's what is the attraction of a very low impedance earth, particularly given that they would make the high currents (during rare fault conditions) which he and Bernard are concerned about even higher?

Kind Regards, John

 

[flameport]

Rods corrode and rot away, often in a very short time.

This is an example after less than 10 years. Was a 4ft long copper clad rod.

 

[JohnW2]

Rods corrode and rot away, often in a very short time. This is an example after less than 10 years. Was a 4ft long copper clad rod.

I don't doubt that they (particularly fairly 'modern' ones) can, if (assuming they are steel) the copper cladding/plating is not very substantial, although I have to say that my 'original' rod (which I suspect is best part of 100 years old) still has a very respectable impedance - but I suspect that "they don't make them like wot they did back then"!

I do agree that one should use something which is likely to last (physically) for a long time, and that one should look for a n impedance no greater (ideally appreciably less) than 100 Ω, but as I recent wrote, particularly in the context of discussions like the one in this thread, I don't see why people should be wanting 'very low' impedances, particularly give that such will increase even further the potentially very high currents, under certain fault conditions, which they appear to fear.

Kind Regards, John

 

[ericmark]

I know when working at Point of Aye gas terminal there was a massive resistor about the size of a small van to connect DNO earth to site earth to limit current flow.

I also came across a radio ham who had a very good TT earth and the PEN was lost causing the earth wire to the radio shack to melt.

However in a farm it is likely a small step down transformer either supplying just the farm or one or two extras, and often split phase rather than 3 phase, so an earth rod of even 20 Ω is unlikely to cause a problem, and most earth rods are more like 60 Ω, to get 23 amp at 230 volt it needs to be 10 Ω and 6 mm² will have no problem with 23 amp.

However if the cow shed has the re-bar as an earth in the concrete slab then we are looking at 1 ohm or less, so then we are looking at high currents likely limited by the DNO earth at transformer rather than the consumers earth.

Where everyone in a housing estate has their gas and water pipes earthed with the loss of the PEN the current is shared between each home, the danger is when all but one home have the bonding after the insulating block in the gas supply, and just one home has the bond the wrong side.

I know my parents house before I left home around mid 60's had an earth good enough to rupture a 13 amp fuse, as I did it, but in 2004 when I came to work on the house the only earth rod was for the GPO party line, and they did not have TN supply, it was turned into a TN-C-S supply, but there was no sign of how the house had been earthed, I would assume water pipes, and street had been turned into a plastic pipe supply. Gas was added after I had left home, so would not have been earthed with the gas. House built 1954. I do wonder how long the house had been without an earth?

 

[FarmerJohn]

Wow that has given me a lot to think about. Many thanks for everyone who has taken the time to help. Much appreciated.

I think that perhaps instead of using earth rods I will choose a more robust harder wearing alternative since I already have the pavement removed from all away around the bungalow for some ground works that were needed about 3 years ago.

I think I am still tending towards the TNC-S system with a good local earth connected to the incomming TN.

Regards
FarmerJo

 

[JohnW2]

I know when working at Point of Aye gas terminal there was a massive resistor about the size of a small van to connect DNO earth to site earth to limit current flow. .... I also came across a radio ham who had a very good TT earth and the PEN was lost causing the earth wire to the radio shack to melt.

I think those illustrate the reason why, as I suggested, at least in some senses (and particularly if TN-C-S in involved) it is not only unnecessary but probably undesirable to have very low impedance paths to earth through earth electrodes (or other extraneous-c-ps).

Kind Regards, John

 

[JohnW2]

I think that perhaps instead of using earth rods I will choose a more robust harder wearing alternative since I already have the pavement removed from all away around the bungalow for some ground works that were needed about 3 years ago.

Again, it's your decision. As I said, it's clearly desirable to have something which you believe will last' for a long time, although, despite flameport's anecdote/photo, earth rods generally seem to be fine for decades.

I think I am still tending towards the TNC-S system with a good local earth connected to the incomming TN.

Again, it's your decision but, as flameport said, in your situation it could well be more sensible to forget the TN-C-S and just have a TT installation.

However, if you do decide to add the local earth to a TN-C-S one, bear in mind what I said about the (at least in my opinion) undesirability of having a very low impedance local earth - since, not only is that unnecessary but it would also increase the very risks you are seeking to address, by possibly considerably further increasing the (already very high) currents that could flow through some conductors (potentially causing fires) in the event of the (very rare) supply-side fault scenarios you are thinking about.

Also, in an attempt top keep this discussion's feet on the ground (i.e. 'in some perspective'), dare I again repeat (I think for the third time ) that literally millions of people are living with TN-C-S installations without any awareness of the 'risk' that concerns you, and hence obviously not 'doing anything about it)!

Kind Regards, John

 

[mattylad]

So you'd put a very rare condition over normal safety use of an earth....

Isn't it more likely that someone would need to be protected by a good earth than the DNO's fail?
I'd have thought so.

 

[bernardgreen]

Isn't it more likely that someone would need to be protected by a good earth than the DNO's fail?

If the DNO's fail is an open circuit Neutral in a TN-C-S service then all the electrical wiring in the house including the CPC will be pulled up to Live potential. The path for this pull up is through any appliance or lamp that is switched on. Gas meters and other external metal works bonded to the CPC / MET will also be pulled up to Live potential.

To the people in the house it will appear to be a power cut, nothing electrical is working, they will see it as annoying but not dangerous . They will be unaware that almost all Earthed metal will be at Live potential.