543.6.1 the protective conductor shall be incorporated in the same wiring system as live conductors (re: overcurrent protection)

Well a Wiring System is defined as "An assembly made up of cable or bus bars and parts which secure and, if necessary, enclose the cable and bus bars."

543.6.1 is concerned only with "overcurrent protective devices used for fault protection."

Z.

it is sometimes difficult for me to understand the point(s) you are trying to make 'Z'.  543.6.1 relates wiring system, protective conductor, live conductor and protective device and is not just an 'only', as you seem to imply.

I have seen other posts on 543.6.1 (and other related regs.).  My (2nd) question  still stands unanswered by you if you were intending to answer any of the three. :-)   If it is poorly worded let me know.   For example, you could just answer yes or no to it for now .

regards

(edit: made a small phrasing edit)

does it mean that a protective conductor (shared or per circuit) has to be in *each* part of the wiring system that a live conductor runs or close by ?

Only in certain contexts - so not in a TT system (or other situations where fault protection is by something other than an overcurrent device), nor where no c.p.c. is required (e.g. a system fully under section 412). [I'm presuming that in this context 'fault protection' is referring to section 411 rather than 434 - I don't think the wording is particularly clear]

But back to the point... I might also add in 521.8.1 - conductors of a circuit not to be distributed over different cables/conduits etc. So in that context what 543.6.1 seems to be saying to me is: where earth fault currents are high then route the c.p.c. in the same way as a live conductor. You might also think of 521.5.1 that requires all conductors (incl. c.p.c.) to go through the same hole in ferrous enclosures etc.

To me it all feels to be about the interaction of high fault currents - trying to keep the 'flow' and 'return' conductors close together so they cancel out as much as possible - and avoiding large separation of conductors that can cause lots of magnetic problems - from interference to increased loop impedances.

   - Andy.

"it is sometimes difficult for me to understand the point(s) you are trying to make 'Z'.  543.6.1 relates wiring system, protective conductor, live conductor and protective device and is not just an 'only', as you seem to imply."

If an R.C.D. is used the problem seems to go away. The use of a protective device that relies upon large fault currents to operate the protective device, such as a fuse or circuit breaker is the issue here with earth faults. As the R.C.D. will operate quicker than a fuse of circuit breaker in most cases of earth fault the cables do less work under fault conditions. There is less chance of a high earth fault current lingering while the protective device considers operating.

You would run a C.P.C. to all points anyway, such as to switches or socket outlets etc.

Z.

Perhaps it is the aspect of where 'best' the routing of conductors - particularly for lighting - that I am overly fussing about _*in relation to that Reg and associated others*_  when not necessarily running wiring 'together' or in 'close proximity' as detailed. The conductors are all in the same joined-up containment system as a whole, but might not be all together in the same segment at various points.

I used a pvc wiring containment example, because with steel the CPC aspect is somewhat resolved by default, via the various interconnected paths along the steel containment.

For a virtual example: some 20m away from the CU [over-current protective device] is an multi-module lighting grid switch and the most direct route from CU is along the trunking - which runs around the perimeter - along one edge of a wall to the switch.

One might presume, being direct route, to run a line conductor from the cu straight to the switch and then run the switched-lines from each switch module which then would veer off at various points, via different segments of trunking/conduit, to the each of the lighting unit banks. Keeps the number of conductors at the switch minimal.

To ensure a CPC is at the grid switch position, the CPC would run from the CU to the switch (with the line feed).

q1) would/should there then be a set (one for each switched line) of CPC's from the switch, along the same route as *each* of the switched-lines, or perhaps just run one 'running' CPC from the switch (or even from the CU ?) to each point by any efficient route ?

The neutral doesn't need to go to the switch, so could run an entirely different [efficient] route to the switched line(s) along sections of trunking/conduit from the cu to each of the lighting unit banks.

q2) would/should a CPC be also run along with the neutral in this case ?

Regards

Habs

NB: Of course one could adopt a loop-in approach if equipment etc permits, which keeps all the conductors together.

You do not want to be running any live conductor (line or neutral) on its own in a steel (ferrous) containment. It might get hot.

521.5.1

Of course a line feed down a steel conduit drop to a switch and its switch line back up again creates a cancelling effect of any magnetic fields around the two wires. But you would not want to run just a live wire singly along a steel conduit route.

Lighting cable is quite inexpensive so cost is a minor issue.

Z.

Thank you Z.

Getting hot issues aside, my scenario from the outset was not concerning steel containment,  I can't see how your response addresses specifically  q1 or q2 from my last post...how would you wire a supply feed to the grid switch and then on with the various switch-lines in the [2-plate approach perhaps] scenario I presented out of interest, if you think that approach incorrect ?

if a diagram is needed and interest not waning, I will draft one if it helps picture it.

Approaching both steel and pvc containment with a consistent 'best' approach seems sensible. 

The conductors are all in the same joined-up containment system as a whole, but might not be all together in the same segment at various points.

I'd always assumed - and now you mention it, it is just an assumption - that 'in the same multicore cable or wiring system, or immediate proximity' phrase was meant to mean that all the relevant conductors would be present at the same point. So you couldn't for instance have a complete loop of conduit or trunking running around the perimeter of a room with everything running say 'clockwise' around it  so one leg of conduit might contain just contained L (or SL) and another just N. The fact that all the bits of conduit are joined up somewhere wouldn't seem to make much difference to me (although I admit the words of the regs could be read that way).

It's like the old "spider wiring" you used to see in houses for lighting circuits from the 1960s-1980s - done in single core sheathed wiring. L would find the shortest route between switches and N the shortest route between lights and SL the shortest route between switch and light. It all worked but the large areas of "loops" can cause havoc interference-wise, especially to things like hearing-aid loops and audio equipment. I presumed that it was that sort of practice that these regs were meant to prohibit.

   -  Andy.

www.youtube.com/watch

I would be inclined not to make the ends of the wires too long as that is wasteful. Also, I would mark the switched lines to make final connections quicker and easier, especially at switches where there are three brown wires.

Also, 521.5.1 refers to ferrous enclosures, (see definition for enclosure) so where single wires enter a metal box or luminaire base through a hole,  the regulation applies, although with small currents the heating effect will, in practice be negligible.

Z.