Mains voltage

[reds42]

However, the main point of the phot is that the incoming ABC cable (the one coming from bottom right of pic) is not vastly bigger than the ones feeding individual installations from it, which I imagine are at least 16mm². Maybe I'm wrong, but I find it rather hard to believe that the 'main' feed cable (which looks the same throughout it's length) is bigger than 35mm² (maybe 50mm² at the very most). What do you think?

A little bit more google of ENWL documents found this document about ABC cables. On page 15 there is this table which has some maximum diameters of bundled cables (also including the possibility of 50mm2) and its interesting how little the diameter of the cable bundle increases for quite big increases in CSA. It would be very hard to judge just how big the cable is without actually measuring it, but to my eye though I would say that incoming cable does look a little bit thicker than the others in your photo - it could easily be 50mm2 but I wouldn't even rule out it possibly being 95mm.

EDIT: I'd think your incoming cable could well be 35mm2, there is no mention in this document of smaller CSA cables.

 

[JohnW2]

A little bit more google of ENWL documents found this document about ABC cables. On page 15 there is this table which has some maximum diameters of bundled cables (also including the possibility of 50mm2)

Thanks again.

and its interesting how little the diameter of the cable bundle increases for quite big increases in CSA.

Yes, it's the same with multicore cables - unless one does the sums, it's surprising how relatively little the overall cable size increases as CSA increases. However, I though I was 'subconsciously' taking that into account when I made my guesses of the cable CSA's - but I certainly accept that I may have been wrong.

It would be very hard to judge just how big the cable is without actually measuring it, but to my eye though I would say that incoming cable does look a little bit thicker than the others in your photo - it could easily be 50mm2 but I wouldn't even rule out it possibly being 95mm.

Visually (at a distance) it's certainly 'a little bit bigger' (but, I would have said only a 'very little' bit - which is why I thought that 35mm² was the most likely, but 50mm² 'possible'. I would still be surprised if it were 95mm² (or even 70mm², not mentioned in the table, if they use it - but again, I may be wrong!

EDIT: I'd think your incoming cable could well be 35mm2, there is no mention in this document of smaller CSA cables.

Possibly. Without getting a ladder out, I can't tell. The moment the ABC hits my wall, it changes to singles which travel along my wall, and they certainly look like 16mm². I have been up-close on a ladder in my time, and my recollections are that the cores of the ABC looked very similar in size to the (I presume) 16mm² singles - but, yet again I could be wrong. However, they're certainly smaller in external diameter than the modern 25mm² singles which we know and love!

In passing I'm rather interested to see that for 50mm² and 95mm², the footnote to the tables says that the number of strands in each conductor is 19±1. In general, even numbers of strands (e.g. 18 and 20) don't 'pack' all that well/efficiently. There are obviously always 'manufacturing tolerances' for anything (particularly in relation to dimensions, weights etc.) but it seems a little odd to see it being implied that a manufacturing process could get the number of strands wrong, doesn't it?

Kind Regards, John

 

[reds42]

There are obviously always 'manufacturing tolerances' for anything (particularly in relation to dimensions, weights etc.) but it seems a little odd to see it being implied that a manufacturing process could get the number of strands wrong, doesn't it?

It does seem strange given how specific all the other parameters are. I presume its to give different cable manufactures a little leeway in their designs using slightly thicker or thinner strands to meet the required CSA.

 

[JohnW2]

It does seem strange given how specific all the other parameters are. I presume its to give different cable manufactures a little leeway in their designs using slightly thicker or thinner strands to meet the required CSA.

Perhaps but, as I said, even numbers of strands generally do not pack as well. On the other hand, I find it hard to understand as a 'tolerance' of (or error in) the manufacturing process. The machine which twists strands into a conductor is presumably is fed separately with 19 strands - if one didn't "turn up", one would expect some flashing lights and audible alarms, and as to where it could get a 20th one from ....

So, the most credible explanation is probably what you suggest - that, for whatever reason, different manufacturers always (deliberately) use 18, 19 or 20 strands.

Kind Regards, John

 

[SUNRAY]

18 is not an unusual number of strands

I've amended to include the blue 21.

The sizes I struggle to see are the common 14, 16, 24

 

[JohnW2]

18 is not an unusual number of strands

It's not only a matter or being able to arrange them 'neatly'. To optimally (most efficiently) 'pack' (e.g. strands in a multi-strand conductor - or, indeed, cores within a multicore cable) they also ideally need to be symmetrical in all planes (which implies that they will fit within an outer circle, with all the outermost strands touching that circle) and to have minimal wasted space within the bundle of strands.

That is not the case with your 18-strand arrangement and nor, I think, for any of the others you illustrate, other than 3.

The first four numbers which are mathematically 'optimal' in these respects are 3, 7, 19 and 37 - which are, indeed, by far the most commonly seen numbers of strands in conductors.

Kind Regards, John

 

[reds42]

Visually (at a distance) it's certainly 'a little bit bigger' (but, I would have said only a 'very little' bit - which is why I thought that 35mm² was the most likely, but 50mm² 'possible'. I would still be surprised if it were 95mm² (or even 70mm², not mentioned in the table, if they use it - but again, I may be wrong!

The first document I found describes the use of M20 bolts through 22mm holes in the posts. Now assuming that your post is using the same standard looking fittings and M20 bolts I present this terrible piece of image editing consisting of a coloured in bolt cut, rotated, and pasted over the incoming cable. I'd say there's a good chance it is a 95mm2 supply as two M20 bolt widths would pretty much match the 39mm diameter of a 95mm2 cable.

 

[eveares]

My voltage (under low/normal loads) tends to hover around 242v within a volt or so.

 

[JohnW2]

The first document I found describes the use of M20 bolts through 22mm holes in the posts. Now assuming that your post is using the same standard looking fittings and M20 bolts I present this terrible piece of image editing consisting of a coloured in bolt cut, rotated, and pasted over the incoming cable. I'd say there's a good chance it is a 95mm2 supply as two M20 bolt widths would pretty much match the 39mm diameter of a 95mm2 cable.

Judging by your investigative activities, it looks as if you are a 'man after my own heart'

I really need to take some higher resolution photos (the one's I'm presenting are pretty old, from my archives), but see below.

It looks to me as if the cable bundle is probably 1.3 - 1.5 times the size of the bolt. IF the bolt is 20mm, that would make the cable bundles some 26mm -30mm. The individual cores look to be 0.5 - 0.6 times the diameter of the bolt, which would (with 20mm bolt) translate to 10mm - 12 mm.

Per the table you previously posted, my estimates would seem to suggest either 35mm² or 50mm², perhaps slightly more likely 35mm². What do you think? We are obviously both having to assume that the bolt is 20mm.

Kind Regards, John

 

[reds42]

It looks to me as if the cable bundle is probably 1.3 - 1.5 times the size of the bolt. IF the bolt is 20mm, that would make the cable bundles some 26mm -30mm. The individual cores look to be 0.5 - 0.6 times the diameter of the bolt, which would (with 20mm bolt) translate to 10mm - 12 mm.

Per the table you previously posted, my estimates would seem to suggest either 35mm² or 50mm², perhaps slightly more likely 35mm². What do you think? We are obviously both having to assume that the bolt is 20mm.

Looking at the right-hand edge of the right hand overlay I would be thinking 1.5 times the bolt putting it at possibly 50mm2, but then the left end does look more 1.3 times and 35mm2.

I think the key thing to bear in mind is that the bundle is 4 cores and not perfectly round in cross section, and the description in the table is of a circle diameter which would circumscribe the outer bounds of the cable - so a circle that would encompass all the corners of the bundle. Its looks like the cables have a twist to them so really we need to look at the visibly thicker areas of the cable where we are seeing the maximum distance across the bundle to gauge the diameter of circle that could circumscribe it.

In this case I would say the right-hand edge of the right-hand overlay would be the most appropriate to use as a gauge so I would be thinking stronger odds on being 50mm2.

 

[JohnW2]

Looking at the right-hand edge of the right hand overlay I would be thinking 1.5 times the bolt putting it at possibly 50mm2, but then the left end does look more 1.3 times and 35mm2.

Indeed so.

I think the key thing to bear in mind is that the bundle is 4 cores and not perfectly round in cross section, and the description in the table is of a circle diameter which would circumscribe the outer bounds of the cable ....In this case I would say the right-hand edge of the right-hand overlay would be the most appropriate to use as a gauge so I would be thinking stronger odds on being 50mm2.

I'm inclined to agree. I certainly think we're probably both right in saying that it's either 35mm² or 50mm², definitely not 95mm² (or even 70mm², which doesn't appear in your table).

What perhaps may have made me think that it was 'slightly more likely' to be 35mm² was the core diameter (for which the 'bundling considerations/ complications' don't apply). Given the complication of shadows etc. (and 'fuzz'!) I would doubt that the cores are more than 12mm at most, quite possibly nearer 10mm - which would, per the posted table, make 35mm² more likely.

Kind Regards, John

 

[Nige F]

105 yards, it is in the church yard, at the end of the street.

has it got a dead short somewhere ?

 

[ericmark]

Does this all matter, has no one brought up the skin effect, gone are the days of 14/0.012 cables, or 28/0.012, and as far as bunching cables, just as likely to cool as to heat cables up, rarely are all cables in a wiring harness loaded to the max, I have opened up so many harnesses where one wire has had a massive overload, only to find no damage in the harness, only where it enters and leaves are there signs of damage.

 

[JohnW2]

Does this all matter, has no one brought up the skin effect ....

'Skin effect' is very small with 50 Hz, and therefore will not be a significant factor in relation to short lenths of cable, as in LV distribution networks.

The reason for the discussion about cable dimensions and bundling etc. is to enable us to determine (at least get a best guess of) the CSA of the distribution cable supplying my house (and, I presume, many others) - which, in turn is relevant to my original question (and subsequent discussion) about the voltage variation ('voltage drops') I see (and don't see) in my supply (e.g. when people downstream of me turn on electric showers).

Kind Regards, John

 

[Harry Bloomfield]

Does this all matter, has no one brought up the skin effect,

The effect at mains frequencies is minimal, it only becomes important at RF frequencies.