Dado 3 compartment or Maxi trunking - home office - use singles?

[SimonH2]

Yes, but just being a ring doesn't actually meet the requirements for a high-integrity earthing system.

No, you need 2 cpc rings.

Where does that come from? It would seem that a 'standard' ring final circuit (with a single, usually 1.5mm², 'CPC ring') would satisfy 543.7, by satisfying 543.7.1.203(iii) and 543.7.2.201(i), so long as it has the 'separate terminations' required by 543.7.1.204.

That was my understanding from when I looked at this <some years ago> - one CPC ring, separate terminations at each point (including at the DB/CU). The point being that by doing that, failure of any single connection could not leave a device (or group of devices) without a CPC.

I've witnessed office rings done using separate terminals at the sockets - but then (AFAICS) using one terminal at the board. I say AFAICS because I've never seen any markings used on the tails to show which circuit they belong to - so it's not impossible that they've "crossed" the tails, eg terminal 1 has ends from ring 1 and ring 2, terminal 2 has the other ends from both rings. I guess I work in a different environment - wherever I see work being done, it's seems to be the "speed conscious" end of the trade they are employing and no board ever looks anything like the good examples RF posts here.
I've never come across a board with an extra earth block to take separate terminations without "crossing the tails" - I know it's an option on many models.

 

[ban-all-sheds]

What, if anything, am I missing?

One of the shortest words in the English language, it seems, even though you used it.

It would seem that a 'standard' ring final circuit (with A single, usually 1.5mm², 'CPC ring') would satisfy 543.7, by satisfying 543.7.1.203(iii) and 543.7.2.201(i), so long as it has the 'separate terminations' required by 543.7.1.204.

543.7.1.203(iii) requires two individual cpcs. A ring is one cpc. So you need two.

If you want only one then it has to comply with 543.7.1.203(i) or (ii).

 

[JohnW2]

Where does that come from? It would seem that a 'standard' ring final circuit (with a single, usually 1.5mm², 'CPC ring') would satisfy 543.7, by satisfying 543.7.1.203(iii) and 543.7.2.201(i), so long as it has the 'separate terminations' required by 543.7.1.204.

That was my understanding from when I looked at this <some years ago> - one CPC ring, separate terminations at each point (including at the DB/CU). The point being that by doing that, failure of any single connection could not leave a device (or group of devices) without a CPC.

Indeed - that always has been (and remains, despite BAS) my understanding of the regs.

I've witnessed office rings done using separate terminals at the sockets - but then (AFAICS) using one terminal at the board. ...

Fair enough, but that's simply non-compliant with the reg, rather than any problem with the reg.

Kind Regards, John

 

[JohnW2]

It would seem that a 'standard' ring final circuit (with A single, usually 1.5mm², 'CPC ring') would satisfy 543.7, by satisfying 543.7.1.203(iii) and 543.7.2.201(i), so long as it has the 'separate terminations' required by 543.7.1.204.

543.7.1.203(iii) requires two individual cpcs. A ring is one cpc. So you need two. If you want only one then it has to comply with 543.7.1.203(i) or (ii).

That is not my interpretation and nor do I think that it is many people's interpretation. 543.7.2.201(i) says that one of the "acceptable arrangements" is "a ring final with A ring protective conductor". As for 543.7.1.203(iii), in the case of a (single) ring, each and every point in the circuit is connected to the MET via two separate CPCs (the two arms of the ring). Indeed, how are you suggesting that one could connect a socket to the MET by two parallel CPCs (which you accept is compliant) without creating "A (single) ring" (which you claim is not compliant)??

Have you ever seen a circuit with two CPC rings (i.e. 4 connections from each socket to MET)??

I am as sure as I can be that my interpretation of the regs is correct, and that most people other than you believe and do likewise.

Kind Regards, John

 

[Iggifer]

This link, albeit a bit old, agrees with John, last page is the clearest page

 

[SimonH2]

A ring is one cpc. So you need two.

Yes, I've seen that failed argument a few times.
To be really pedantic, a ring only has one CPC if it is a single cable from CU to CU with no joints - and that makes it a rather useless circuit unless you use "vampire taps" for all connections Once you have a join, there are two or more CPCs in the circuit.

A better way to look at it is not consider "quantity of bits of wire" but "quantity of current paths". So take a typical ring with a few sockets on it, and assume it's been properly wired for hi-integrity earthing. Pick a socket at random and consider the paths back to the CU/DB. You will have two paths, the CPCs in the two cables connected to the socket which make their way separately via a number of other accessories (except for the end socket) to separate terminals on the CU/DB earth bar.
Thus, for the purposes of the reg, TWO CPCs connecting the socket back to the CU/DB earth bar.

The main feature being that you can disconnect any single connection or cut any single CPC conductor in the circuit and the socket is still connected to the earth bar.

 

[JohnW2]

A ring is one cpc. So you need two.

Yes, I've seen that failed argument a few times.

Indeed - as I said, it is really not a viable argument. As I wrote to BAS, running two parallel CPCs from a point (e.g. socket) back to the CU/MET actually inevitably creates 'a ring' (in the usual sense) - which makes BAS's argument self-defeating.

The regulations, as interpreted by you and I (and probably most other people other than BAS) seem to make good sense - so I am very happy to accept that interpretation!

Kind Regards, John

 

[ban-all-sheds]

That is not my interpretation and nor do I think that it is many people's interpretation.

Everyone who says otherwise is not looking at what the regulations actually SAY.

Anybody who publishes "interpretations" or "guidance" which shows a single cpc ring when the csa is not 10mm² or 4mm² as per 543.7.1.203(i) or (ii) is publishing something which contradicts what the regulations actually SAY.

Look at what 543.7.1.203(i), (ii) & (iii) talk about.

(i) A single protective conductor having...

(ii) A single copper protective conductor having...

(iii) Two individual protective conductors, each complying with the requirements of Section 543...

With the latter, it is important to consider 543.2.9:

Except where the circuit protective conductor is formed by a metal covering or enclosure containing all of the conductors of the the ring, the circuit protective conductor of every ring final circuit shall be run in the form of a ring having both ends connected to the earthing terminal at the origin of the circuit.​

So what the regulations say when you read 543.7.1.203(iii) and 543.2.9 in conjunction is

Two individual protective conductors, each run in the form of a ring and each having both ends connected to the earthing terminal at the origin of the circuit.

Taking a ring cpc and putting the ends into separate terminals does not create two individual rings.

As for 543.7.1.203(iii), in the case of a (single) ring, each and every point in the circuit is connected to the MET via two separate CPCs (the two arms of the ring).

But two arms of the ring are not each a ring themselves.

Forget circuits, cpcs, rings etc etc for a minute, and just think of the general situation where you have two individual {things}, each of which having {some sort of property}.

As you have two individual {things} you must be able to take one {thing} away and be left with the one remaining {thing}. Since each separate {thing} had (some kind of property} your one remaining {thing} has {some kind of property}.

If you cannot take one away as described then you cannot have had two in the first place.

Indeed, how are you suggesting that one could connect a socket to the MET by two parallel CPCs (which you accept is compliant) without creating "A (single) ring" (which you claim is not compliant)??

If I may be allowed, for the purposes of illustration, to use non-standard colours, like this:

As you can see, I can remove either of the two individual rings and be left with one remaining ring

Have you ever seen a circuit with two CPC rings (i.e. 4 connections from each socket to MET)??

No, but then I'm not likely to have seen any circuits except the ones in my house.

I am as sure as I can be that my interpretation of the regs is correct, and that most people other than you believe and do likewise.

You are not alone in refusing, for some reason I cannot fathom, to simply read and think about regulations actually SAY, or what the implications are if your theories on topology are correct.

If you believe that a different arrangement complies with 543.7.1.203(iii) and 543.2.9 in conjunction, then please post a drawing which shows what you consider to be two individual cpcs each run in the form of a ring and each having both ends connected to the earthing terminal at the origin of the circuit. Please use a different colour for each individual cpc in the form of a ring, and then post a drawing which shows one of those two individual cpcs entirely removed to leave one remaining cpc run in the form of a ring having both ends connected to the earthing terminal at the origin of the circuit.

 

[JohnW2]

That is not my interpretation and nor do I think that it is many people's interpretation.

Everyone who says otherwise is not looking at what the regulations actually SAY.

I am perfectly comfortable with my interpretation of the regulation, which is believe is also shared by most people other than you - not the least because it actually makes sense.

In contrast, you appear to be trying to say that, taking the simplest example of one-socket circuit, if one takes two separate CPCs (different terminals at both ends) from the socket to the CU earth bar, if one calls that "two parallel CPCs" it is compliant, but if one calls it "a (single) CPC ring" it is non-complaint - which does not make sense.

Kind Regards, John

 

[ban-all-sheds]

A ring is one cpc. So you need two.

Yes, I've seen that failed argument a few times.
To be really pedantic, a ring only has one CPC if it is a single cable from CU to CU with no joints - and that makes it a rather useless circuit unless you use "vampire taps" for all connections Once you have a join, there are two or more CPCs in the circuit.

If you believe that then you also believe that no circuit will ever comply with what the regulations say is required.

If you believe that then you believe that there are an awful lot of places in the regulations where they do not mean what they say.

A better way to look at it is not consider "quantity of bits of wire" but "quantity of current paths".

The regulations do not use that term.

So take a typical ring with a few sockets on it, and assume it's been properly wired for hi-integrity earthing.

Do you mean properly, i.e. in accordance with what the regulations say, which is two individual protective conductors, each run in the form of a ring and each having both ends connected to the earthing terminal at the origin of the circuit?

Or do you mean some other arrangement?

Since clarification of that is important, I'd like you too to draw it, before and after the complete removal of one of the two individual cpcs, as I asked John to do.

Pick a socket at random and consider the paths back to the CU/DB. You will have two paths, the CPCs in the two cables connected to the socket which make their way separately via a number of other accessories (except for the end socket) to separate terminals on the CU/DB earth bar.
Thus, for the purposes of the reg, TWO CPCs connecting the socket back to the CU/DB earth bar.

Even though the regulations do not use the term, or consider the concept of, "current paths", if you believe that's what they mean then can you show that each of your paths, which you consider to be equivalent to a cpc, is a ring? Can you show that there are two individual paths, and that each of them is a ring? Can you show that one of the two individual paths can be removed, leaving the one remaining individual path which is in the form of a ring?

The main feature being that you can disconnect any single connection or cut any single CPC conductor in the circuit and the socket is still connected to the earth bar.

That's no different to a non-high-integrity cpc scenario.

 

[JohnW2]

The main feature being that you can disconnect any single connection or cut any single CPC conductor in the circuit and the socket is still connected to the earth bar.

That's no different to a non-high-integrity cpc scenario.

Exactly. The regs are saying that, other than the additional 'separate terminals' requirement, a standard "non-high-integrity" ring final circuit is acceptable in high leakage situations.

Kind Regards, John

 

[ban-all-sheds]

I am perfectly comfortable with my interpretation of the regulation, which is believe is also shared by most people other than you - not the least because it actually makes sense.

Do you have any theories of why you, and "most other people" want to "interpret" regulations rather than do what they actually say?

What you believe complies with what the regulations actually say is required does not, in fact, comply with what they actually say is required. It only complies with what you think they mean once you start acting on the basis of "well, they don't actually mean what they say".

In contrast, you appear to be trying to say that, taking the simplest example of one-socket circuit, if one takes two separate CPCs (different terminals at both ends) from the socket to the CU earth bar, if one calls that "two parallel CPCs" it is compliant, but if one calls it "a (single) CPC ring" it is non-complaint - which does not make sense.

It would be compliant if in that situation the regulations did not require that each CPC be in the form of a ring. For example if the circuit was not a ring, or the cpcs were of a sufficient size.

It would not be compliant if in that situation the regulations did require that each CPC be in the form of a ring.

And that does make perfect sense - how can "a (single) CPC ring" be compliant with a requirement, clearly and unabmiguously stated in the regulations for two individual protective conductors, each complying with the requirements of Section 543?

 

[ban-all-sheds]

Exactly. The regs are saying that, other than the additional 'separate terminals' requirement, a standard "non-high-integrity" ring final circuit is acceptable in high leakage situations.

How does taking the topology of a single ring make it two rings if you use separate terminals?

Please draw it, using different colours to indicate which is ring A and which is ring B.

543.7.1.203(iii) and 543.2.9 in conjunction requires two individual protective conductors, each run in the form of a ring and each having both ends connected to the earthing terminal at the origin of the circuit.

Unless you want to extend your "well, they don't actually mean what they say" concept to the meaning of everyday English words, you are left with that problem of "two individual...".

If you have two individual [circuit protective conductor of every ring final circuit shall be run in the form of a ring having both ends connected to the earthing terminal at the origin of the circuit]s then you must be able to remove one of the two individual [circuit protective conductor of every ring final circuit shall be run in the form of a ring having both ends connected to the earthing terminal at the origin of the circuit]s leaving one individual [circuit protective conductor of every ring final circuit shall be run in the form of a ring having both ends connected to the earthing terminal at the origin of the circuit].

Can you do that?

 

[JohnW2]

How does taking the topology of a single ring make it two rings if you use separate terminals?

It doesn't. However, can you point me to the regulation which requires two CPC rings? All I can see is a requirement for two separate CPCs from a socket back to the earthing terminal at the origin of the circuit - and a single ring satisfies that.

Although it might conceivably happen in some specialised environments, I have never seen or heard of a sockets circuit being wired with two CPC rings - has anyone? ... and don't forget that this discussion is all about an ordinary domestic situation - and I really cannot believe that anyone has ever seen (or installed) a ring final with two CPC rings (i.e. 4 CPCs from each socket to origin) in an ordinary domestic environment.

Kind Regards, John

 

[ban-all-sheds]

It doesn't. However, can you point me to the regulation which requires two CPC rings?

I have already done that.

543.7.1.203(iii) requires two individual cpcs, each complying with the requirements of Section 543, and in Section 543, 543.2.9 requires that the cpc be in the form of a ring. So if 543.7.1.203(iii) requires two individual cpcs, each in the form of a ring, that means two individual rings.

All I can see is a requirement for two separate CPCs from a socket back to the earthing terminal at the origin of the circuit - and a single ring satisfies that.

If you think you have two separate cpcs then you must be able to completely remove one separate one and have one separate one remaining. Can you do that? If not then you cannot have had two separate ones before.

Although it might conceivably happen in some specialised environments, I have never seen or heard of a sockets circuit being wired with two CPC rings - has anyone? ... and don't forget that this discussion is all about an ordinary domestic situation - and I really cannot believe that anyone has ever seen (or installed) a ring final with two CPC rings (i.e. 4 CPCs from each socket to origin) in an ordinary domestic environment.

None of that has anything to do with what the regulations actually say. As I observed, are not alone in refusing, for some reason I cannot fathom, to simply read and think about what the regulations actually SAY