Historical Wiring.

If we were starting afresh without existing colour codes, then a sensible choice in my view would be

White for neutral, as white is a neutral sort of colour.

Brown/green striped for earth, as the earth IS brown. A striped wire is good for earth as it stands out, make the other colour green because most things growing in the earth ARE green.

Red, violet and blue. for the three phases. Cant use yellow for phase as too much like white chosen for neutral. 

Grey and black reserved for other conductors such as ELV when required to be different to line voltage.

As of this week or so, green for earth, but brown (for soil) also suits. Contrasting green and yellow is obviously distinct so I have no objection to that.

Red (for danger) for line (SP) and other primary colours for L2 and L3 (save yellow if earth is green/yellow).

Black (absence of colour) for neutral.

Hang on, aren't we back to where we were decades ago?!

I'll say it. Earth is safety critical isn't it? Perhaps that ought to be red?

Mike

Really I think folk should not be over reliant on just looking at the colour of the insulation, and to actually check what the copper is actually connected to.

mapj1 said:

I'll say it. earth is safety critical isn't it? Perhaps that ought to be red?

Only when something goes wrong.

Whilst colours are convenient, do they actually matter? It could be that they stop people actually testing what goes where and making an assumption that everything found in the socket is correct. G/Y for Earth is probably sensible, but the others? I don't care what the colours are from a practical viewpoint, and I don't even like the silly notice saying colours from a mixture of regulation dates may be present. In the HV world cables don't have colours, modern ones are all red sheathed, old ones may be lead! EHV ones are all bare aluminium. I rather liked the old colours red/black for single phase, RYB for 3 phase (and 2 way switches) etc. In larger installations black with numbered cores may well be present, and one should never believe the colours anyway, but I suppose it makes connecting up slightly quicker and allows testing to be ignored (at someones peril)! Take a trunking installation with many cables, How often does one find labels on each circuit with some ident? In good quality control panels there will be cable markers at each termination, and probably all the wires one colour, I think a better system and very helpful if there is a fault, particularly if one also has the wiring schedule and a circuit diagram, many of the wires not being live or neutral and quite possibly both at various times.

Good answer

I'll say it. Earth is safety critical isn't it? Perhaps that ought to be red?

A bit of Teutonic thinking there by any chance?

Whilst colours are convenient, do they actually matter?

One hand they are useful, but on the other having a direct association with function is a bit limiting at times - what use is a brown/blue (or red/black) pair for a lightswitch?. Yes you can buy twin brown (but then can't tell the difference between L and SL) and it's all to messy to have a similar option with 3-core - brown/brown/brown or brown/brown/blue.

The French approach (mostly in conduit, even domestics) has a practical air about it - G/Y for PE, Blue for N and anything else you like for L. That way you can use different colours for SL, stappers, or different SLs in multi-gang switches without any need for extra labels or oversleeving.

It might be simpler if the coding simply identified the core within a cable (like the numbered cores in many cored SWAs) and the association with function was looser. You could have conventions - e,g. core 1 is perm L if present, highest number if N, if present, a separate G/Y core for PE. You could even keep colour codes, but only have them identify the cores - say using an existing colour-to-digit code (as used with resistors for example), so brown = core 1, red = 2, orange = 3 etc.

   - Andy.