High integrity earthing

[NotHimAgain]

If I measure the continuity of the end of a ring final circuit I might get exactly the same result as I would obtain with two radial circuits connected in parallel. So your point is?

Two radials in parallel would be a speciial case of a ring final circuit - that is one with just one sink (and of course one source).

Now BAS we could continue this all night - I am well aware of the topology of electrical networks having taught the subject at advance level. You have not managed to introduce any new arguments to support your case and I am afraid this is just becoming an exercise in trying to avoid defeat at all costs - not very interesting.

 

[SeanT]

I'm still wondering how you are going to explain the topology argument - a tree structure vs a ring?

 

[NotHimAgain]

Seant
The topology of the ring main is very simple. It is a loop with at least two nodes. One is a node to allow energy delivery (a source) and the other is a node to allow energy to be utilised in some appliance (a sink).

Most RFC have more than one sink - i.e Most RFC have a number of socket outlets.

BAS is trying to contend that a high integrity ring main requires a cpc greater than 1.5mm². He bases this on his definition of a ring main.

Now I sat on the committee that wrote section 607 for the 16 th edition in the early 90's. This of course does not make me definitely right about the subject but I do have some insight.

IMO the minimum conductor size for the cpc is clearly defined (see 543.7.1.3 (iii) pargraph two in the 17th for details) as 1.5mm² and BAS chooses to disagree.

However, he has not put forward any credible argument to support his assertions.

Now you must decide who is right here - because I am finding the argument tiresome and of little value.

 

[Spark123]

(iii) 2nd para implies one multicore cable to me, a RFC has more than one multicore cable

 

[ban-all-sheds]

BAS, you seem to be missing the point here.

No - I'm not missing the point. I know exactly what point people are trying to make - they are saying that a ring circuit has two of everything - two phase, two neutral and two cpcs.

They are wrong - it does not - it has one of each. The fact that at any point on the ring current can travel along two paths to get there does not mean that the circuit has two of each conductor.

I'd like to see anybody who believes that a ring circuit has two of everything draw the circuit, indicate clearly which is cpc 1, and which is cpc 2, and state a logical and consistent rule, or set of rules, which means that on any ring circuit, with any indeterminate number of outlets, cpc 1 can always be defined, and cpc 2 can always be defined, and why if a ring with no outlets has 1 cpc, and one with 1 outlet has 2 cpcs, one with N outlets still only has 2, and not N-1.

As I'm sure we are all aware, the purpose of high integrity earthing is to ensure that circuits designed to supply equipment with high leakage currents do not lose their connection to earth, as this could lead to all metalwork connected to the CPC rising to a dangerous potential.

A ring circuit is ok to use for this purpose providing both protective conductors are terminated independantly of each other throughout the entire circuit.

A ring circuit is OK for this purpose, but not one cabled in 2.5/1.5 T&E, because that does not comply with the cross-sectional requirements of 607-02-04

And you can't say "both protective conductors" because there is only one.

This is so if a screw is left loose, or works itself loose, there is still an earth connection to all the appliances on that circuit.

That may well have been the only intention of the people who wrote the regulations, but that is not what they wrote, and if they did not write what they meant then that's because they too could not successfully count beyond one or grasp the topology of a ring.

 

[securespark]

(iii) 2nd para implies one multicore cable to me, a RFC has more than one multicore cable

"Where the two individual protective conductors are both incorporated in a multicore cable..."

That to me means where they are both part of a multicore cable, but not necessarily the same cable...

 

[ban-all-sheds]

If I measure the continuity of the end of a ring final circuit I might get exactly the same result as I would obtain with two radial circuits connected in parallel. So your point is?

My point is that I wasn't comparing a ring final with two radials connected in parallel. My description was abundantly clear - two conductor ends in an MCB, removed, continuity measured between the two, and the question asked - what order of magnitude of resistance would your tester display if the two were the ends of a ring, and what would it display if the two were the ends of two radials. I said nothing about parallel radials.

Two radials in parallel would be a speciial case of a ring final circuit - that is one with just one sink (and of course one source).

Indeed they would, but I never said I was comparing a ring to two radials in parallel, just two radials.

Surely I don't have to draw it out for you to understand that?

Now BAS we could continue this all night - I am well aware of the topology of electrical networks having taught the subject at advance level. You have not managed to introduce any new arguments to support your case

I have, but you are perversely ignoring them, and refusing to answer the questions I asked.

and I am afraid this is just becoming an exercise in trying to avoid defeat at all costs - not very interesting.

Well stop doing that then...

A single conductor with two ends, bent round so that the two ends touch is still a single conductor, not two.

 

[ban-all-sheds]

Seant
The topology of the ring main is very simple. It is a loop with at least two nodes. One is a node to allow energy delivery (a source) and the other is a node to allow energy to be utilised in some appliance (a sink).

Most RFC have more than one sink - i.e Most RFC have a number of socket outlets.

BAS is trying to contend that a high integrity ring main requires a cpc greater than 1.5mm². He bases this on his definition of a ring main.

OK - consider this.

I take a length of twin and earth cable. How many conductors do I have in total?

Through some undefined molecular welding technique I join the ends of the conductors so the cable forms a completely unbroken ring. How many conductors do I have in total?

At an indeterminate number of points on the ring I remove the sheathing and sleeving etc, but I do not cut any of the cores. How many conductors do I have in total?

At one of those points I connect a current source. How many conductors do I have in total?

At the other points I connect current sinks. How many conductors do I have in total?

If the numbers you give in answer to each of those questions is not the same all the way through, please explain the logic of how and why it changes, and explain how that logic, consistently applied no matter how many points on the ring have the conductors exposed and/or things connected to them, results in the answers you give.

Now I sat on the committee that wrote section 607 for the 16 th edition in the early 90's. This of course does not make me definitely right about the subject but I do have some insight.

It could also mean that you and the other committee members were as unclear then as you are now on how many conductors a ring final has.

IMO the minimum conductor size for the cpc is clearly defined (see 543.7.1.3 (iii) pargraph two in the 17th for details) as 1.5mm² and BAS chooses to disagree.

543.7.1.3 (iii) para 2 (in the DPC at least) says the same as 607-02-04 (iii) para 2 in the 16th, and it only works for 2.5/1.5 T/E if you decide that, against the evidence of your own eyes, this drawing:

has 6 lines on it.

Also I would draw to your attention the last sentence in that paragraph. If the requirement for two individual protective conductors is met by a cable with one protective conductor bent round so that the ends touch, what is the logical explanation for the possible existence of yet another individual protective conductor formed by the protective sheath, armour or wire braid of the cable? Surely it can't mean that the conductor in the cable is one cpc, and the armour etc is the 2nd?

Or if it does mean that, when the cable is formed into a ring, are there then 3 cpcs? Or 4? Or do we then go back to just counting the core as 1, in order to remain in compliance with (iii), which only allows for 2 cpcs?

A logical and consistent explanation would be appreciated.

However, he has not put forward any credible argument to support his assertions.

Other than the fact that the drawing above clearly has only 3 lines on it, not 6....

Let's try this. The drawing below shows, at A, a length of copper wire.

At B, C, D etc it shows the wire cut in one or two places, and joined or not in one or two places with a piece of choc-block. For each of B - H could you please state how many conductors there are, and why.

 

[NotHimAgain]

Taking your last drawing first - now my PC skills do not come near yours so I can't (or may be that should be - can't be bothered to try) produce diagrams like yours. So what to do - well I just go to the CCU terminations of my RFC and count how many conductors ends I can pull out. Next I could go to each socket outlet and do the same.

Now its a great pity BAS that you were not on the 607 committees that met after the one I sat on - if you had been you could have helped them out. You see on the committee I attended we actually spelled out in plain English that 2.5mm² / 1.5mm² was adequate when used to form a RFC with a high integrity earth (see my earlier post). Since then they have used a form of words that is not so starkly obvious.

Now the fact of the matter is that my view is supported by the industry in general - BTW some of the large national contractors that have installed these circuits in the way I have described - may be just a little irritated by the fact that you are implying that they have not complied with BS 7671, and are therefore, in breach of contract with many major clients .

Just as an aside BAS - can you explain why high integrity earthing is sometimes required and the basic principles behind the protection it offers? If you can you may come to realise that, just in this case - size does not matter .

 

[equitum]

........we actually spelled out in plain English that 2.5mm² / 1.5mm² was adequate when used to form a RFC with a high integrity earth (see my earlier post). Since then they have used a form of words that is not so starkly obvious..

Please post your version of para 2 of 543.7.3 as the version I have does not state in plain english that 2.5/1.5 is adequate when used to form a RFC with a high integrity earth.

Also diasgree that if you can count 2 cpc's in a socket or in a consumer unit that they are separate conductors. Its the same conductor joined together ! If I can count 2 CPC's in a radial lighting circuit in a JB or ceiling rose, does that mean they are separate conductors ?

 

[tim west]

This is so if a screw is left loose, or works itself loose, there is still an earth connection to all the appliances on that circuit.

If both ends of the ring were terminated under the same screw, then one screw coming loose could lead to some or all of the appliances loosing their connection to earth.

Of course they would have to be terminated at the cu apart in two earth terminals for this to work ie the terminal screw could in theory loosen either end of the IE RFC and I agree with Notimagain that the RFC is not a circle but two seperate conductors from the source, to be a single conductor the CSA would have to be increased, one reason for designing them in the first place, to be able to use smaller size cables yet still allow a 30A supply

 

[ban-all-sheds]

Taking your last drawing first - now my PC skills do not come near yours so I can't (or may be that should be - can't be bothered to try) produce diagrams like yours. So what to do - well I just go to the CCU terminations of my RFC and count how many conductors ends I can pull out. Next I could go to each socket outlet and do the same.

I really can't believe that you don't grasp the point I'm trying to make.

I'm not asking you how you would determine whether you had a ring or two radials.

I'm trying to explain to you, by getting you to think about continuity readings, the following:

1) If you get a reading of infinity, you have two separate conductors,

2) If you get a reading of a few hundred milliohms you have one separate conductor.

And you don't have to draw anything - you just have to answer my question about how many conductors there are at B, C, D etc, and the only PC skill you need to do that is the ability to type a letter, a space or other separator, and a number, on 7 consecutive lines.

Now its a great pity BAS that you were not on the 607 committees that met after the one I sat on - if you had been you could have helped them out. You see on the committee I attended we actually spelled out in plain English that 2.5mm² / 1.5mm² was adequate when used to form a RFC with a high integrity earth (see my earlier post). Since then they have used a form of words that is not so starkly obvious.

That must be the understatement of the year. They have used words so appallingly badly that they have written a requirement that is not what they mean, and a requirement which, in the way that it is written, does not allow a 2.5/1.5 ring if high integrity earthing is required.

The fact is, you cannot count the conductors in a ring twice - there are not two phase or neutral conductors, there is one of each. They are in the shape of a ring, and the fact that at any point of use current will flow in two directions along each segment of them does not make each two conductors. There are not two cpcs, there is one. It is in the shape of a ring, and the fact that at any point of use current could flow in two directions along each segment of it does not make it two conductors.

If you try to use the argument that it does, then how many cpcs does 10mm² have? Seven, because that's how many strands there are, and therefore that's how many paths a cpc current could take?

If you're using the braid of an SY cable as a cpc, how many cpcs do you count that as?

By what logically inconsistent and plain-English incorrectness do you decide that if you take one conductor, and bend it round so that the ends touch, you now have two conductors?

I don't believe that a single one of all the questions I've asked you can be logically and consistently answered and not show that a conductor in the shape of a ring is still a conductor. But I do believe that the realisation of that is why neither you nor anybody else has attempted to answer any of them.

Now the fact of the matter is that my view is supported by the industry in general - BTW some of the large national contractors that have installed these circuits in the way I have described - may be just a little irritated by the fact that you are implying that they have not complied with BS 7671, and are therefore, in breach of contract with many major clients .

Well I'm all broken up about that. But they undeniably are in breach of BS7671 as it is written.

Just as an aside BAS - can you explain why high integrity earthing is sometimes required and the basic principles behind the protection it offers? If you can you may come to realise that, just in this case - size does not matter .

Please don't think that I don't understand why it's needed, and why separate terminals are used etc - I'm simply pointing out that a ring final has only one cpc, and if that cpc is only 1.5mm², and if high-integrity earthing is required, then according to BS7671 size does matter, at least to the extent that one cpc of 1.5mm² is not deemed to be adequate.

 

[tim west]

A cable is surrounded by an insulation therefore a 10mm stranded would still be one cable no matter how many strands.

Whether you strip and bend or fold a conductor its still classed as two cables by electricians as they both go back to a single source and the current is split equally amongst them.

The term a "Ring" circuit is probably confusing you as you see a ring as a continous circle which it isnt it has tap off points.

 

[Spark123]

I can see what you are meaning, when does one CPC become 2 CPCs etc.
Taking into account 607-02-04 (i) and (ii) are both talking about larger single protective conductors, I took (iii) to mean if you need to have two single protective conductors (as in pieces of wire) which both need to comply with section 543 then it is also OK. A CPC can be sized using the adiabatic equation hence 1.5mm² can be acceptable (other factors still need to be considered).
Guidance note 8 says: "Ring final circuits provide duplication of the protective conductor, and if the ends of the protective conductor are separately terminated at the distribution board and at the socket outlets, the requirements of 607 will be met, as shown in figure (below). Socket outlets are available with two earth terminals for this purpose, both of which should be used to uphold this increased integrity of the CPC"

 

[ban-all-sheds]

A cable is surrounded by an insulation therefore a 10mm stranded would still be one cable no matter how many strands.

So where, how, and by what logical and consistent rule does it become two cables if I remove a section of insulation, bend it double, and insert it into an accessory terminal?

Whether you strip and bend or fold a conductor its still classed as two cables by electricians as they both go back to a single source and the current is split equally amongst them.

It may be so classed, but that is obviously wrong. If I take a length of cable, I have one conductor.

If I strip the ends, I have one conductor.

If I join the ends together, you're saying I now have two conductors. You must therefore be able to draw me a cable with the ends joined together and show me which is conductor #1 and which is conductor #2.

The term a "Ring" circuit is probably confusing you as you see a ring as a continous circle which it isnt it has tap off points.

It's not confusing me, but it seems to be confusing everyone else who can't see that whether it has zero, one or more than one tap off points it is still one conductor.

Please import the diagram that Spark123 has posted into the drawing tool of your choice, and mark it up to show which is cpc #1 and which is cpc #2, as per "two individual protective conductors each complying with the requirements of Section 543".