Supplementary Bonding Revisited
- Thread starter JohnW2
- Start date
Sponsored Links
OK. Start with the simplest situation, which I presume is what you are thinking of, namely ...
If that's all there is, then, obviously, the resistance you measure between A and C will be R1+R2 and the voltage you measure between A and C will be zero, regardless of the resistance.
Now find a hypothetical 144V battery, man enough to push a current of 144/R2 amps through R2 (and if R2 is 'extremely small', then the battery will have to be 'extremely large'
).Since the voltage across R2 is now 144V, you will hopefully agree that the voltage you now measure between A and C will now be 144V, despite the fact that the measured resistance between A and C (obviously without the battery connected!) could be anything, including 'incredibly low'. Do you agree? Translating that into the real-life scenario I have been talking about, we get ...
You will hopefully agree that the voltage measured between A (now the extraneous-c-p) and C (now the exposed-c-p) will again be about 144V, and that will remain the case no matter how small the (R1+R2) you measured
If you agree with the above, it takes me back to the question I asked above - what is the point/significance in 'passing' the 415.,2.2 'test' if it still allows a potentially lethal PD to exist between (potentially) simultaneously touchable parts??
I think your difficulty probably arose because of the word "negligible" (resistance). In order for the voltage between ends of some path to be negligible, the resistance of that path has to be 'negligible' in the context of the magnitude of currents which may flow through parts (or all) of that path. In other words, if (as in the situation I'm describing) 'incredibly large' currents may flow through part of the path, then the resistance has to be 'incredibly low' for there to be negligible voltage between the ends of that path.
Kind Regards, John
Yes, that's alright as you have drawn it, but isn't the last statement agreeing with what I am saying - and the regulations?
Surely with a bond between your A and C like this:
would mean a person (assuming not earthed) touching Y and Z would not result in a shock any more than if touching W and X.
Surely with a bond between your A and C like this:
would mean a person (assuming not earthed) touching Y and Z would not result in a shock any more than if touching W and X.
Yes, of course it wouldn't.
However, surely the point is that if (without SB) you had measured a very small resistance between your Z and Y (or between Z and MET), you would have concluded that the bond was not required, wouldn't you? If that were the situation, and your conclusion, such that you did not install any SB, you would have left the door open to a potentially lethal PD arising between your Z and Y (which might be 'simultaneously touchable'), wouldn't you?
Whatever, my question about the 'test' of 415.2.2 remains - what on earth is it attempting to achieve? As a test of the adequacy of SB which one had installed (which is seemingly what it is intended for) it is surely superfluous - since it's inconceivable that any bit of G/Y SB conductor would not satisfy the test, isn't it? - i.e. all one would be testing would be how well one had connected the G/Y.
Kind Regards, John
Sponsored Links
Oh, right. Have I been looking at the wrong bit?
Yes. Edit - Z and Y; not Z and MET.
I suppose so.
Then what would be the difference between touching two points at the same potential on a single conductor and your drawing?
Maybe 49V instead of 1V - Is that the point?
Well, yes, but as I said before; you cannot dismiss a test as superfluous because it verifies a satisfactory result.
Last edited:
Have made a small edit.
You tell me - I'm not sure which 'bit' you have been looking at
Only one of my diagrams (the last one) is relevant. The fundamental difference between mine and your modification of it is that my diagram (and everything I have written) relates to the situation in which there is NO supplementary bonding in place. In your modification, you have added SB (i.e. a short bit of G/Y, usually of of ≥4mm² CSA, between extraneous-c-p and exposed-c-p) - which, as you say, would obviously ensure that the potential difference between those two parts would be close to zero under any credible circumstances.
It doesn't really matter to this discussion, but I'm a bit confused by your 'edit' - I thought that you regarded a very low resistance between a 'part' and the MET as indicating that it did not need to be bonded?
I'm afraid that I don't understand those questions. Could you perhaps clarify in a way that even I can understand?
As a test of the adequacy of SB which has been installed, I would say that it is not merely 'superfluous' but is actually potentially dangerous. In a situation in which two nearby conductive parts have theoretically been electrically connected by a short length of G/Y (usually of of ≥4mm² CSA) that 'test' would seemingly 'verify that the connection was satisfactory' if the measured resistance between the parts was as high as ~1,667Ω if RCD-protected and ~0.31Ω if only protected by a B32> The former would be totally ridiculous, and even the latter not a lot better (given that, say, 5m of 4mm² G/Y should have a resistance of only about 0.028Ω). Resistances so much higher than one would expect would presumably be indicative of very iffy connections which could well deteriorate with time.
Kind Regards, John
Ok.
Good.
No, that means it is an extraneous-c-p to the bathroom.
For supplementary bonding I have always said measure between the exposed- and extraneous-c-p.
Well, I am not sure. Just trying to work out the regulation requirements.
Are you looking at the situation in a normal situation, rather than the split second (or 5) during a fault?
What happens to the e-c-p potential during that fault?
With your hypothesis it would appear that SB is always required because, without it, the e-c-p is always going to be a 0V.
Yes, but you know as well as I that the 1,667Ω (actually 1,666.66.; 1,667 is too high), like the 23,000kΩ, is just a nominal figure for 30mA and 10mA respectively.
Fair enough - thanks for clarifying.
The resistance measurement obviously has to be undertaken when there is no fault. However, the rest of what I'm talking about (and depicting in diagrams) obviously relates to that "split second (or 5)" before the fault is cleared (assuming it is) by a protective device, since it's only during that very brief period that there is any possible 'dangerous potential' around to consider or worry about.
In the scenario I've been talking about, it would be at exactly MET potential (since no current flowing through the extraneous-c-p to MET, hence no VD). However, the exposed-c-p would be considerably above MET potential (144V above MET in my diagram) - hence a large potential difference between the two parts.
There is a 'converse' sort of fault (maybe somewhat bernard-like) in which (perhaps due to a "frayed vacuum cleaner lead" touching it) it is the extraneous-c-p, rather than the exposed-c-p which 'becomes live'. However, exactly the same argument applies, other than that it is now the extraneous-c-p which rises to a proportion of supply potential above MET, with the exposed-c-p remaining at MET potential. However, there is again (very briefly) a large potential difference between the extraneous- and exposed-c-ps.
[well, 0V above MET potential]
Indeed, that is what I am saying/suggesting - unless/until someone can find a flaw in my argument. With either of the types of fault I have mentioned, it seems to me that the only way to avoid what could be a very high (albeit brief) potential difference between extraneous- and exposed-c-ps (IF one feels the need to avoid that) would be to join them with an SB conductor - regardless of the results of any of the measurements we've been talking about.
As you've said, these potentially lethal PDs which I'm suggesting could only be prevented/avoided by SB (regardless of anything) should always be very brief, so maybe one could argue that we don't need to take steps to prevent/avoid them? However, as always, a potentially more worrying scenario might arise (but only in the absence of RCD protection) with faults of more than 'negligible' impedance, hence much longer disconnection times if the only fault protection were that provided by an OPD.
However, I don't think any of this helps me to make much sense of the regulations - what about you?
I don't think that makes 415.2.2 any less 'potentially dangerous' - because it could ('understandably) lead people to believe that ('required', for whatever reason, and present) SB was adequate when there was a "far too high" measured resistance between the extraneous- and exposed- parts, suggesting very poor connections.
Kind Regards, John
I don't really know any more, but -
Surely, during the period of the fault with many Amps flowing, the PD between the exposed- and extraneous-c-p is dependent on the impedance between them.
Surely, during the period of the fault with many Amps flowing, the PD between the exposed- and extraneous-c-p is dependent on the impedance between them.
No - what makes you think that?
Look back at the middle one of the three diagrams I posted (the one with the 144V battery). Do you not accept that you would measure 144V between my A and C, despite the fact that (if your meter had 'infinite' input impedance) no current would be flowing through "R1" (which is analogous to path from extraneous-c-p to MET).
Similarly (again assuming that your meter has 'infinite' input impedance), do you not accept that you would always measure 144V between A and C, regardless of the value of R1 - i.e. it you would measure 144V if R1 were zero (hence the total path impedance be as low as possible, namely R2) and would also measure 144V if R1 (hence total path impedance) were extremely high? In other words, the PD between A and C remains the same (144V) regardless of the impedance of the path between them.
If your meter is drawing no current ('infinite input impedance'), then R1 is really nothing more than an extension of your meter lead, so that what your meter is measuring is the voltage between B and C (i.e. across the 144V battery)
You still seem to be thinking of the first of the three diagrams I posted, in which no 'external influences' are injecting voltage into part of the path between A and C.
Kind Regards, John
Last edited:
Oh, I give up.
You know voltage drop is not my favourite subject.
You know voltage drop is not my favourite subject.
I suspect you may be trying to make it more complicated than it actually is - since it's really little more than Ohm's Law.
There is only a voltage drop when current flows through a resistance/impedance, and the magnitude of the VD is then simply down to Ohm's Law.
Hence, if one has a conductor through which no (or virtually no) current is flowing, then there will be no (or virtually no) voltage drop along it's length - i.e. the potential (relative to anything) will be the same at both ends of the conductor. That is the case with meter leads, the R1s in my diagram and the path via extraneous-c-ps to the MET.
If you have a DVM with a very high input impedance, you ought to be able to put a very large resistor in series with one if its leads (or substitute leads which are hundreds of metres long) without noticeably altering the voltage measurement - since, with virtually no current flowing, there will be virtually no voltage drop, even if the resistance is high.
However, although you may have 'given up', I am left with what (unless someone can correct me) seems to be a compelling argument that nothing but local SB can restrict potential differences between extraneous - and exposed-c-ps to a safe level under certain fault conditions, and also regulations relating to SB which I find difficult to rationalise, and the 'test' of 415.2.2 the point of which I can understand even less.
Kind Regards, John
One more try.
Ignoring the extraneous-cp, what would happen if you could touch the exposed-c-p and the MET when the fault occurred?
Wouldn't the Touch Voltage be 160 x R of CPC? So the lower the resistance, the lower the TV.
160 x 0.3 = 48V would be ok.
160 x 1.0 = 160V not ok, so how to reduce? Reduce the resistance with bonding.
So, isn't that the same, allowing for greater resistance, when instead touching the extraneous-c-p?
So, won't it happen in the CPC during the fault with, say, 160A flowing to the MET to which the extraneous-c-p is connected.
Yes, but we are talking about fault currents.
Ignoring the extraneous-cp, what would happen if you could touch the exposed-c-p and the MET when the fault occurred?
Wouldn't the Touch Voltage be 160 x R of CPC? So the lower the resistance, the lower the TV.
160 x 0.3 = 48V would be ok.
160 x 1.0 = 160V not ok, so how to reduce? Reduce the resistance with bonding.
So, isn't that the same, allowing for greater resistance, when instead touching the extraneous-c-p?
In the fault situation I was talking about (L-E fault in a Class I item), the extraneous-c-p (or anything else connected to the MET) is irrelevant as far as the fault current is concerned. In the absence of SB, all of the fault current (160A or whatever) travels through the circuit's CPC to the MET, resulting in a large voltage drop along the length of that CPC (144V VD in my diagram, for a 2.5/1.5 mm² cable) - the voltage at the downstream end of that CPC (i.e. at the exposed-c-p) is therefor 144V (or whatever) above MET potential.
However, the extraneous-c-p has nothing to do with the fault, fault path or fault current. No current ('fault current' or otherwise) will be flowing through it, so there will be no VD along it's length - so the end at the bathroom will be at MET potential.
With the exposed-c-p at 144V (or whatever) above MET potential and the extraneous-c-p at 0V above MET potential, there is clearly a 144V (or whatever) potential difference between those two parts.
Ah - you just beat me to it, since I was about to throw that very scenario at you!
As you say, imagine that the situation was such that you could touch both the exposed-c-p and the MET simultaneously whilst there was a fault. The exposed-c-p is clearly 'live' (the 144V above MET potential in my example) - and the MET is clearly "at MET potential". So if one of your hands touched something (the MET) which was AT MET potential, and your other hand touched something (the exposed-c-p) which was 144V above MET potential, then you would clearly have 144V 'across your body' - not pleasant
Now, imagine that (more realistically!) the MET was too far from the bathroom for you to be able to touch it whilst touching the exposed-c-p in the bathroom. Determined not to be beaten by that, you reached for a few metres of cable, connected one end to the MET and held the other end in your hand (whilst your other hand was holding the exposed-c-p). I hope you are agreed that, electrically speaking, that's no different from touching the MET directly - so, again 144V between your hands, so, again, not a very nice experience!
Finally, being one for realism, you decide not to use that piece of wire to connect your hand (electrically) to the MET but, instead, created a length of copper pipe (with appropriate bends/couplings etc.) - with one end connected to the MET and the other end held in your hand (whilst your other hand was holding the exposed-c-p). Again, electrically no different - still 144V between your hands, but this time with a newly created "extraneous-c-p" as the path between your hand and the MET.
One little pedantic issue to get out of the way .... we've previously never talked about current going through a person (merely about a potential difference between two things which a person might touch simultaneously). If you did as above, during a fault, and first got hold of the exposed-c-p and then, fractionally later, got hold of the wire or pipe (providing a path to the MET) then, immediately before you got hold of the wire/pipe (c.f. the extraneous-c-p) there would be NO current flowing through it, hence no VD along it's length, and hence the end you were about to touch would be AT MET potential.
When you did touch it (during the fault) a current would flow through your body and through that wire/pipe - but, in terms of the 'big picture' it would be a very low current (e.g. 144 mA if the 'touch voltage' were 144V and your body resistance 1,000Ω). There then would (for the pedants!) be a small voltage drop in that path between your hand and the MET (via the extraneous-c-p or equivalent), but it would be tiny. With, say, that 144 mA through your body (and the 'extraneous-cp'), even if the resistance of the path to MET via the e-c-p were 10Ω (fairly unlikley for copper pipework), the VD would only be 1.44V - so the potential at the 'bathroom end' of the the e-c-p would be fractionally above the potential of the MET, but as close to MET potential "as to make no difference".
When you did touch it (during the fault) a current would flow through your body and through that wire/pipe - but, in terms of the 'big picture' it would be a very low current (e.g. 144 mA if the 'touch voltage' were 144V and your body resistance 1,000Ω). There then would (for the pedants!) be a small voltage drop in that path between your hand and the MET (via the extraneous-c-p or equivalent), but it would be tiny. With, say, that 144 mA through your body (and the 'extraneous-cp'), even if the resistance of the path to MET via the e-c-p were 10Ω (fairly unlikley for copper pipework), the VD would only be 1.44V - so the potential at the 'bathroom end' of the the e-c-p would be fractionally above the potential of the MET, but as close to MET potential "as to make no difference".
I assume that by "touch voltage" you're referring to the potential difference between the extraneous- and exposed- parts. Is that correct?
If so, then, as I explained/illustrated, in the absence of SB it is essentially determined by the ratio of CSAs of the L-conductor and CPC of the cable of the circuit with a fault - the 144V I've been talking about is roughly what it would be with 2.5/1.5mm² cable (144V being 230V x 5/8) but, for example, it would be around 115V with 1.0/1.0mm² cable.
This is where I might well 'lose you' even more, because if one does add a SB conductor, then things get quite a bit more complicated, and one can probably look at things in terms of earthing as well as bonding!
Thought of in terms of bonding, if one connects the two parts with a relatively short length of an adequate conductor, then one more-or-less ensures that there can never be much of a potential difference between the two connected parts (the two ends of the SB conductor) - since it will have a very low resistance/impedance, such that no appreciable voltage will be developed across the length of the conductor with any credible current flowing through it.
However, by adding that SB conductor, what one has actually done is to place an additional conductor in parallel with the circuit's CPC, thereby reducing the effective "R2" of that circuit, hence also reducing its Zs (and hence increasing PEFC). The total fault current will therefore increase - if it were your 160A without SB, it could be appreciably, perhaps considerably, more than 160A after you added the SB. In that sense, therefore, one has 'improved' the earthing of the circuit, by effectively reducing the resistance of its CPC.
However, that (increased) fault current due to the SB will now be shared between two parallel paths back to the MET - one via the circuit's CPC and the other through SB conductor and the extraneous-c-p. How the fault current is shared between those two paths will obviously depend upon the relative resistances/impedances of the two paths.
As a result of the effective reduction (due to SB) of the CPC resistance (R2) of the circuit, the potential of the exposed-c-p (relative to MET) during a fault will reduce - e.g. if the cable is 2.5/1.5 mm², the voltage of exposed-c-p during a fault will reduce from ~144V (230V X 5/8) without SB to some lower (but probably not dramatically lower) figure with SB.
So, it's quite likely that, even with SB, the potential of the exposed-c-p during a fault will still be 'dangerously high' relative to the MET. However, because there is now a proportion (possibly a substantial proportion) of the fault current going (via SB) through the extraneous-c-p, the potential of its bathroom end will also rise considerably above MET potential. Without SB there was no current through the extraneous-c-p, hence no VD along it, hence the bathroom end of the extraneous-c-p was AT MET potential. However, with the SB, which results in an appreciable proportion of fault current flowing through the extraneous-c-p, there will be a considerable VD in that path, and hence the bathroom end of extraneous-c-p (as well as the exposed-c-p) will now be at a potential considerably above that of the MET.
Without SB the exposed-c-p was (during fault), say, about 144V above MET potential and the extraneous-c-p at 0V above MET potential - so a difference between them ('touch voltage'?) of 144V. However, just pulling numbers out of the air for illustration, with SB the exposed-c-p may have fallen to, say, 90V above MET potential, but the extraneous CP risen to, say, 85V above MET potential - hence now a difference ('touch voltage') of only 5V.
However, what I've just written illustrates one important point namely if one applies ANY supplementary bonds (between extraneous-and exposed- parts), then it's probably essential that one bonds together ALL extraneous and exposed parts in that location - since, during a fault, bonded extraneous-c-ps could be at a high potential relative to MET, hence a serious risk if simultaneously touchable with something unbonded which was connected to the MET.
I suspect you will (to use the language of my daughters) by now be 'well confused' - but I am trying to be as clear as I can
Kind Regards, John
DIYnot Local
Staff member
If you need to find a tradesperson to get your job done, please try our local search below, or if you are doing it yourself you can find suppliers local to you.
Select the supplier or trade you require, enter your location to begin your search.
Are you a trade or supplier? You can create your listing free at DIYnot Local
Sponsored Links
