Supplementary Bonding - my questionable conclusions?

[JohnW2]

Very odd. Original post and third attempt still awaiting moderation. Second post which I tried and deleted, indeed replacing with dots, has this note. ..... As you can see from the screenshot now posted - which would not originally work - contains very few symbols; certainly nothing out of the ordinary.

I've just had the same problem again, although this time almost immediately 'released' by mods (after I had posted the usual screenshots). It had a good few "Ω" symbols, but that's about all. Maybe I'll have to revert to typing "ohms"

Kind Regards, John

 

[JohnW2]

Yes, but I meant an actually negligible value - as in the recognised value of 0.05Ω. .... If the pipes were such that the impedance between them were 0.05Ω then SB would not be necessary.

I'm not sure where that 'recognised value' comes from - I only recall that figure being mentioned (probably in OSG!) as 'guidance' in relation to main bonding conductors. In any event, the figure (no matter how low) makes absolutely no difference difference to the generality of what I was saying ... <explanation and example> ...

I should perhaps have added ...

... quite apart from the question of where your 'recognised value' (of 0.05Ω) comes from, you need to remember that this whole discussion is about a regulation which says that SB may be omitted (in a bathroom) if it is RCD protected etc. and if the impedance between extraneous-c-p and MET is less than 1,666Ω - so it would definitely say that SB was unnecessary with the 2Ω I used for my illustration

In any event, as I illustrated, my argument remains unchanged for any credible value of that impedance.

Kind Regards, John

 

[JohnW2]

What's wrong with this little one, I wonder?

 

[JohnW2]

But surely there would be no current in the dead-end CPC/exposed-c-p - until you touched it and something earthed. That would be just like receiving a shock from a live conductor while earthed; not really what we are talking about, is it? In the case of a pipe enlivened by a boiler fault then surely all earthed parts and extraneous-c-ps in the room will be similarly enlivened to a similar voltage.

I forgot to respond to the above bit previously ...

If I understand your diagram correctly, that scenario is very different from the 'simple' situation we have been discussing.

We have been talking about a scenario in which an exposed-c-p, connected via a CPC to the MET, becomes 'live' as a result of a fault and where there is an extraneous-c-p which is also connected to the MET (via bonding and/or earthing). When the fault arises, the potential of the exposed-c-p rises to considerably greater than MET potential, whereas the extraneous-c-p remains at MET potential - hence a considerable PD between the two parts.

However, if I understand your diagram correctly, the extraneous-c-ps (pipes) are connected directly to the exposed-c-p which is 'live' because of a fault (i.e. the boiler). In that situation, the exposed- and extraneous-c-ps will obviously at the same potential (connected by copper pipe, through which no current is flowing), and hence no significant PD between them - so, from that POV, there would clearly be no point of/need for SB.

However, in that situation there will be dangerous PDs between both extraneous- and exposed-c-ps and the MET (due to the 'VD' in the CPC from exposed-c-p to MET, due to very high fault current), so would represent a hazard for anyone touching either of those parts simultaneously with touching something connected directly to the MET. That means that, getting 'more complicated', if there were another extraneous-c-p, connected directly to MET but (unlike the ones already considered) NOT connected to the exposed-c-p with a fault, OR if they were an exposed-c-p on some other circuit (again, connected directly to the MET, but to connected to part with fault), then there would (during fault) be a dangerous PD between that part and the other two (both 'live') ones - and the only thing that could be done to avoid that would be SB.

Kind Regards, John

 

[EFLImpudence]

If I understand your diagram correctly, that scenario is very different from the 'simple' situation we have been discussing.

We have been talking about a scenario in which an exposed-c-p, connected via a CPC to the MET, becomes 'live' as a result of a fault and where there is an extraneous-c-p which is also connected to the MET (via bonding and/or earthing). When the fault arises, the potential of the exposed-c-p rises to considerably greater than MET potential, whereas the extraneous-c-p remains at MET potential - hence a considerable PD between the two parts.

Originally yes, but then you introduced a CPC acting like an extraneous-c-p because of a fault elsewhere on the circuit.
Is what you had in mind not like my diagram?

However, if I understand your diagram correctly, the extraneous-c-ps (pipes) are connected directly to the exposed-c-p which is 'live' because of a fault (i.e. the boiler). In that situation, the exposed- and extraneous-c-ps will obviously at the same potential (connected by copper pipe, through which no current is flowing), and hence no significant PD between them - so, from that POV, there would clearly be no point of/need for SB.

OK. I got something right.

Is that (my diagram) not what you meant by an exposed-c-p acting like an extraneous-c-p?

However, in that situation there will be dangerous PDs between both extraneous- and exposed-c-ps and the MET (due to the 'VD' in the CPC from exposed-c-p to MET, due to very high fault current), so would represent a hazard for anyone touching either of those parts simultaneously with touching something connected directly to the MET. That means that, getting 'more complicated', if there were another extraneous-c-p, connected directly to MET but (unlike the ones already considered) NOT connected to the exposed-c-p with a fault, OR if they were an exposed-c-p on some other circuit (again, connected directly to the MET, but to connected to part with fault), then there would (during fault) be a dangerous PD between that part and the other two (both 'live') ones - and the only thing that could be done to avoid that would be SB.

I'll have to study that later.

 

[JohnW2]

Originally yes, but then you introduced a CPC acting like an extraneous-c-p because of a fault elsewhere on the circuit.

I think you must be referring to my saying that the CPC (hence exposed-c-p) of some other circuit (say, 'Circuit B') acting as an extraneous-c-p in relation to an L-CPC fault on 'Circuit A'.

That will be true of anything entering the room which is connected to MET and is not carrying any current - whether a conventional extraneous-c-p (e.g. pipe) or the CPC/exposed-c-p of some other circuit

Is what you had in mind not like my diagram?

Yes, it is, provided that the thing labelled "CPC" at the bottom is the CPC of some circuit other than the one on which there is a fault.

What is confusing things a bit is that we (at least, I) have hitherto always been considering fault between L and exposed-c-p of some item in the bathroom, and the 'touch voltage' we have been concerned about has been the PD between that exposed-c-p and some extraneous-c-p (or an exposed-c-p on some other circuit).

However, in your diagram, the fault is outside of the bathroom, and probably on a circuit which does not supply the bathroom, so a person in the bathroom cannot touch the 'live' exposed-c-p directly - so you have then used the pipes not so much as a conventional extraneous-c-p (liable to introduce earth/MET potential into bathroom) but, rather, simply as an extension of the exposed-c-p in some other room (liable to introduce the high potential into the bathroom due to the {boiler} fault elsewhere).

Is that (my diagram) not what you meant by an exposed-c-p acting like an extraneous-c-p?

No - because, as I said above, your diagram, relating to a situation I hadn't discussed (with the fault outside of the room) is essentially doing the opposite of that - because what you are calling an extraneous-c-p is, as above, really simply an 'extension' of a ('live' during fault) exposed-c-p which is in some other room.

As above, for an example of "an exposed-c-p acting like an extraneous-c-p" you would have to introduce another circuit supplying something with a touchable exposed-c-p. For example (adding to your diagram - but see note below about 'bonding conductors') ...

However, for the purpose of my point, the 'extraneous-c-ps' (being used in your diagram as 'extensions' of an exposed-c-p in another room) are not needed, so I have adapted your diagram to have the item with the fault in the bathroom, and with the CPC of a shaver socket circuit effectively being an 'extraneous-c-p' in relation to the fault, with no conventional extraneous-c-ps at all ...

One other (important) thing about your diagram. If I understand it correctly, you appear to be showing 'bonds' (green lines) between the pipes and the MET (close to the boiler). If so, that represents "SB" (since they connect to your 'CPC' at the MET) and that would move the goalposts a long way. In fact, if those bonding conductors were hefty enough, and close enough to the bathroom, they would largely eliminate the "danger" you indicate in the bathroom

I'll have to study that later.

Fair enough.

Kind Regards, John

 

[JohnW2]

@EFLImpudence ... your introduction of the 'boiler fault' example/diagram has made me realise that, at least in the context we are discussing, the attempt at a distinction between "exposed-c-ps" and "extraneous-c-ps" can be potentially confusing and misleading, since, in functional terms (in relation to faults), there roles are essentially 'reversible'.

Not the least because of the BS7671 definitions, we tend to think of an "exposed-c-p" as the touchable metal parts of a Class I item which "... is not normally live, but which can become live under fault conditions" and an extraneous-c-p as touchable metal parts which are "... liable to introduce a potential, generally Earth potential" ("... and not forming part of the electrical installation).

In other words, we usually think of an "exposed-c-p" as the touchable metal parts of an item in the room which can become 'live' (high PD relative to MET) in the event of an L-CPC fault in that item (OR anywhere else on the same circuit) and think of "extraneous-c-ps" as pipes etc. which are "liable to introduce roughly MET potential" because of deliberate or 'incidental' connections to the MET. Under fault conditions, there can therefore be dangerous PDs between those exposed- and extraneous-c-ps. It is also important to realise that the CPC of a different circuit is at least as "liable to introduce roughly MET potential" as is a pipe, and so (in the presence of a fault) presents exactly the same potential hazard as does the ('earthed') pipe.

However, as I explained/illustrated in my previous post, the opposite can happen. If a metal pipe is in electrical continuity with touchable metal parts ("exposed-c-ps) of something Class I, then, if there is an L-CPC fault which makes the exposed-c-p 'live' (high potential relative to MET), then the pipe is really just an extension of that exposed-c-p, and will 'become live' (high potential relative to MET) if/when the exposed-c-p does. If that pipe enters some room, it is therefore (despite it being called an "extraneous-c-p") that pipe which may, under some fault conditions, 'become live' (high potential relative to MET). On the other hand, if CPCs (hence exposed-c-ps) of any other circuits (other than the one with a fault) enter the room, then it is they which will remain at roughly MET potential during a fault, thereby resulting in a dangerous PD between them and 'the pipe' (but 'the other way around' in comparison with the previous paragraph's scenario).

Hence, if we continue to use "exposed-c-p" to refer to the touchable metal parts of a Class I item and "extraneous-c-p" to refer to things like pipes, then I think we need to try to get it out of our heads that it is always the former which can 'become live' (high PD relative to MET) and the latter which is liable to be at something like MET potential - since, as above, it can easily be the other way around. However, no matter which 'way around' it is, the PD between the two 'parts' can be equally dangerous.

Kind Regards, John

 

[EFLImpudence]

I have been busy. I will comment (was going to say answer ) as soon as I can.

Just to explain my diagram:
I did draw them ridiculously close to the boiler but the green lines are boiler CPC and bonds or other CPCs to which the pipes might be connected before they get to the bathroom; not supplementary bonding as they go to the MET.

The purpose of it was to illustrate that none of the fault current would be flowing in the pipes nor CPCs within the bathroom.


Just had a thought: do you think the reason for R≤50/Ia is because someone thought the MCB/RCD limited the magnitude of the fault current? It still doesn't seem relevant after what you have realised but it might be an explanation.
For example for just a light and a B6A MCB, R≤1.66Ω might just be possible - even though now redundant.

 

[JohnW2]

I have been busy. I will comment (was going to say answer ) as soon as I can.

Fair enough.

Just to explain my diagram: I did draw them ridiculously close to the boiler but the green lines are boiler CPC and bonds or other CPCs to which the pipes might be connected before they get to the bathroom; not supplementary bonding as they go to the MET.

Yes, I more-or-less realised that.

I very nearly re-drew your diagram with all those green lines to MET much longer because I think that, as you drew it, there is a psychological issue which possibly 'impairs understanding' ... they are so short that there is a temptation (at least, on the part of my brain!) to think that they are all 'joined together and to the MET', and hence will all be at similar (MET) potential - but, of course, the truth is that the very high fault current travelling through those (significantly long) 'green conductors' (all essentially in parallel) will result in a large potential difference between their ends (i.e. between boiler/pipes and MET).

The purpose of it was to illustrate that none of the fault current would be flowing in the pipes nor CPCs within the bathroom.

Fair enough. However, the point I was trying to make (perhaps not very well) is that I think the only sense in which those green conductors (close to boiler) from pipes to MET (hence also to the CPC going to bathroom) may not qualify as effectively being "SB" (even if not explicitly installed as such) is because of their probable remoteness from the bathroom. If, hypothetically, the boiler were within the bathroom, then I think you might find it difficult to argue that conductors connecting (within the bathroom) the pipes to the incoming CPC (connected to MET) were not "SB conductors", wouldn't you?

Just had a thought: do you think the reason for R≤50/Ia is because someone thought the MCB/RCD limited the magnitude of the fault current? It still doesn't seem relevant after what you have realised but it might be an explanation.
For example for just a light and a B6A MCB, R≤1.66Ω might just be possible - even though now redundant.

I've been giving a lot of thought to the question of "who may have thought what" to result in the appearance of that regulation - but so far without much success, beyond wondering whether they were actually 'thinking clearly' at all!

I'm still thinking about this, so more about it soon!

Kind Regards, John

 

[EFLImpudence]

@EFLImpudence ... your introduction of the 'boiler fault' example/diagram has made me realise that, at least in the context we are discussing, the attempt at a distinction between "exposed-c-ps" and "extraneous-c-ps" can be potentially confusing and misleading, since, in functional terms (in relation to faults), there roles are essentially 'reversible'.

Yes. Their roles reversible but not their definitions

Not the least because of the BS7671 definitions, we tend to think of an "exposed-c-p" as the touchable metal parts of a Class I item which "... is not normally live, but which can become live under fault conditions" and an extraneous-c-p as touchable metal parts which are "... liable to introduce a potential, generally Earth potential" ("... and not forming part of the electrical installation).
In other words, we usually think of an "exposed-c-p" as the touchable metal parts of an item in the room which can become 'live' (high PD relative to MET) in the event of an L-CPC fault in that item (OR anywhere else on the same circuit) and think of "extraneous-c-ps" as pipes etc. which are "liable to introduce roughly MET potential" because of deliberate or 'incidental' connections to the MET.

Yes, but as we are discussing - only exposed-c-ps can become live by direct contact with a line conductor.
That is 'taken care of' by ADS, the simultaneous touching of those parts is 'taken care of' by the SB.

Under fault conditions, there can therefore be dangerous PDs between those exposed- and extraneous-c-ps. It is also important to realise that the CPC of a different circuit is at least as "liable to introduce roughly MET potential" as is a pipe, and so (in the presence of a fault) presents exactly the same potential hazard as does the ('earthed') pipe.

By "the CPC of a different circuit", don't you really mean another exposed-c-p (on a different circuit)? In which case, I don't think that is anything unexpected and is why the CPCs of different items have to be connected by SB.
(Presumably subject to the always forgotten 'simultaneously accessible.)

However, as I explained/illustrated in my previous post, the opposite can happen. If a metal pipe is in electrical continuity with touchable metal parts ("exposed-c-ps) of something Class I, then, if there is an L-CPC fault which makes the exposed-c-p 'live' (high potential relative to MET), then the pipe is really just an extension of that exposed-c-p, and will 'become live' (high potential relative to MET) if/when the exposed-c-p does.

Well yes, but isn't it always stated that in an L-E fault, all items connected to the MET become live - for an instant until ADS? That's one of the reasons for Class II items and why they should(must) not be earthed.

I quite often feel when discussing with you that you come up with things that are so obvious that it is difficult to be sure of why you wrote it - possibly having an as yet hidden reason for writing it.

If that pipe enters some room, it is therefore (despite it being called an "extraneous-c-p") that pipe which may, under some fault conditions, 'become live' (high potential relative to MET).

Yes, but that is why it says generally earth potential.

On the other hand, if CPCs (hence exposed-c-ps) of any other circuits (other than the one with a fault) enter the room, then it is they which will remain at roughly MET potential during a fault, thereby resulting in a dangerous PD between them and 'the pipe' (but 'the other way around' in comparison with the previous paragraph's scenario).

Yes, as above.

Hence, if we continue to use "exposed-c-p" to refer to the touchable metal parts of a Class I item and "extraneous-c-p" to refer to things like pipes,

Yes, depending on whether they are part of the electrical installation or not.

then I think we need to try to get it out of our heads that it is always the former which can 'become live' (high PD relative to MET) and the latter which is liable to be at something like MET potential - since, as above, it can easily be the other way around. However, no matter which 'way around' it is, the PD between the two 'parts' can be equally dangerous.

I don't think that has been in our heads but as I tried to show in my diagram, it depends where the fault is.
A pipe in a bathroom cannot become live in the way that an exposed-c-p can.

However, what you are saying seems to be the realm of ADS.
Supplementary Bonding - the original subject - is how to reduce the potential between those parts to reduce the touch voltage until ADS has disconnected the supply if one part is at mains potential with high current flowing through it and all the others at MET potential with no current flowing through them.
The SB obviously causes current to be shared and flow through all connected parts between the point of connection and the MET - but not the 'rest of those parts' - but reduces the PD between them. So, what calculation needs to be done to ensure this PD is <50V - or is it too dependent on all the various circumstances and position of the parts as to be virtually impossible?

This might seem a daft question but - were you to take a live conductor and touch the MET with it, at what potential would the MET be bearing in mind that the actual earthing point might be at some distance from the MET?

 

[JohnW2]

Yes. Their roles reversible but not their definitions

Agreed, but the 'reversibility of roles' can cause confusion (at least for me!) in these discussions. Throughout this thread (and the one that spawned it), I had been talking (and illustrating) exclusively about the situation (which I think is the one most people think of) in which it was the exposed-c-p which, due to a fault, became 'live' (high potential wrt MET),whilst the extraneous-c-p remained at MET potential. However, in post #37 of this thread you introduced the opposite scenario (your 'boiler fault'), in which it was the extraneous-c-p (per definition) which became 'live', whilst the exposed-c-p remained at MET potential. Maybe it's just me, but that sudden reversal of what was being discussed initially 'threw my mind' a bit

Yes, but as we are discussing - only exposed-c-ps can become live by direct contact with a line conductor.

Bernard would presumably disagree, since a 'frayed vacuum cleaner lead' could theoretically come into contact with a pipe. Even with your 'boiler fault', the distinction is almost semantic - whilst (with your fault) it is literally true that the pipe is not direct contact with a line conductor, the pipe is in direct (electrical) contact with a piece of metal which, in turn, is in direct contact with a line conductor - so electrically no different.

By "the CPC of a different circuit", don't you really mean another exposed-c-p (on a different circuit)?

Yes, of course I meant that - I assumed that was obvious. I used that form of words since I was talking about the 'role equivalence' to an extraneous-c-p "which entered the room" (and hence was 'liable to introduce a potential' into that location) - but it would not have 'read very well' if I had tried talking about an "exposed-c-p which entered the room", would it?

In which case, I don't think that is anything unexpected and is why the CPCs of different items have to be connected by SB. (Presumably subject to the always forgotten 'simultaneously accessible.)

Agreed, and I don't think I suggested (certainly didn't mean to!) that it was "unexpected". I was just underlining the fact that when there is fault causing the exposed-c-ps of item(s) on one circuit to become 'live', then exposed-c-ps of things on other circuits had to be considered 'in the same breath' (i.e. for exactly the same electrical reason) as extraneous-c-ps - hence, as you say, the need for SB, if implemented, to join all the ('simultaneously touchable') extraneous- AND exposed-c-ps.

Well yes, but isn't it always stated that in an L-E fault, all items connected to the MET become live - for an instant until ADS?

Ah - this may be part of the crux of the problem (maybe "your problem"?) with a lot of this discussion.

Assuming that the building has (as required) been constituted as an equipotential zone (by main bonding, if required), then the only thing within that building which can meaningfully be described as "live" is something at a potential different from (usually 'much higher than') MET potential. Hence, during an L-E fault on one circuit, the MET, and everything connected to the MET via paths through which no current was flowing (all extraneous-c-ps and all exposed-c-ps of all other circuits) would remain at MET potential- hence "not live".

Let's face it, if what you wrote were true, then we would have had nothing to discuss. If, prior to clearing of the fault, everything connected to the MET were "live" (hence all at roughly the same potential), then there would be no potential differences between any two extraneous- and/or exposed-c-ps- and hence no risks/hazards to consider or talk about!

Of course, in practice (since Ze will always be above zero), the potential of MET relative to true earth will rise - but, assuming the building is an equipotential zone, that is of no relevance or importance. The potential of the MET could be 5,000V above true earth potential and no-one in the building would be in any way aware of it, and nor would it even be apparent to any measurements taken wholly within the house.

That's one of the reasons for Class II items and why they should(must) not be earthed.

Well, as above, it would only be a ("live") issue if the Class II item were 'earthed' to the CPC of the circuit with the L-E fault. If it were 'earthed' to the CPC of any other circuit, the only theoretical 'downside' would be that it would be creating 'unnecessarily earthed metal' in the building.

Yes, but that is why it says generally earth potential.

Exactly - I was just underlining the 'reversibility of roles' - many people seem to think only of "introducing earth potential"..

I don't think that has been in our heads but as I tried to show in my diagram, it depends where the fault is. A pipe in a bathroom cannot become live in the way that an exposed-c-p can.

I think this is becoming almost 'semantic'.

Consider your diagram. Knock down the wall, so that the boiler (with an L-E fault) is now in the bathroom. Then ...

(1) I presume you will agree that the metalwork of the boiler constitutes an exposed-c-p, and that if you touch it, you will therefore be touching "live" exposed-c-p?

(2) Bolt a bit of metal (maybe a 'mounting bracket or somesuch) to the boiler. Do you agree that if you touch that 'bracket', you will still be touching a "live" exposed-c-p?

(3) Instead of the metal bracket, attach (with electrical continuity) a short piece of copper pipe to the boiler. If you touch the end of that bit of pipe, do you agree that you will still be touching a "live" exposed-c-p?

(4) Now re-build the wall, with that short piece of copper pipe sticking through the wall. If you touch that pipe (whilst in the bathroom), is it not still ('effectively') an exposed-c-p?
​

Supplementary Bonding - the original subject - is how to reduce the potential between those parts to reduce the touch voltage until ADS has disconnected the supply if one part is at mains potential with high current flowing through it and all the others at MET potential with no current flowing through them.

Indeed - and, as I keep saying (the whole point of these discussions) is that I still do not believe that anything other than local SB can achieve such a reduction in the "potential between those two parts".

'The SB obviously causes current to be shared and flow through all connected parts between the point of connection and the MET - but not the 'rest of those parts' - but reduces the PD between them. So, what calculation needs to be done to ensure this PD is <50V - or is it too dependent on all the various circumstances and position of the parts as to be virtually impossible?

As I wrote yesterday, if SB is already present, then passing the 415.2.2 test will ensure that, if the fault current is not quite high enough for 'immediate' ADS, then, in the worst-case scenario (e.g. 'broken CPC', hence all fault current goes through SB) the PD between the parts ('touch voltage') cannot exceed 50V. If the fault current is high enough for 'immediate ADS', then all bets are off as regards that 'touch voltage' (during the very brief period prior to ADS), since the fault current could be very high (for the duration of the fault), hence the touch voltage well above 50V, even if the 'test' had been passed.

However, that is NOT what we are discussing. We are talking about the situation in which SB is currently NOT present and we are discussing criteria for deciding whether it is needed. As I keep saying, as far as I can see, in that situation (without SB) the 'touch voltage' will always be roughly Vf x R2 / (R1+R2) [Where Vf is the supply voltage during the fault], regardless of anything else - in particular, regardless of any measurements of impedance/resistance between the parts, or between parts and the MET (including a situation in which impedance from extraneous-c-p to MET were 'zero'). Hence, as above, I see no way of reducing the 'touch voltage' (during the duration of a fault) other than by installing local SB.

Whether one is worried about the magnitude of the touch voltage during the very brief period before an RCD clears the fault is, of course, a different matter - and perhaps a reason why our discussion may be rather moot (and maybe one reason why 'they' have not bothered to re-visit the regs about SB?).

This might seem a daft question but - were you to take a live conductor and touch the MET with it, at what potential would the MET be bearing in mind that the actual earthing point might be at some distance from the MET?

As above, you have to remember that we are talking about an equipotential zone.

If you touch a live conductor onto the MET, then the MET will obviously (by definition) remain 'at MET potential', and that is all that matters to someone within the equipotential zone of the building. The potential of the MET relative to true earth will obviously rise (perhaps considerably so), due to the finite Ze - but, again, that is irrelevant to anyone/anything within the equipotential zone.

Kind Regards, John