design derating for cables going through kingspan/celotex

[jonnyhifi]

Hi!

I have a problem with derating factors for cables going through phenolic thermal iinsulation:

I have been wrestling with my 17th edition: I am rewiring a house with 17century solid brick walls (no cavity). We are refurbing it, (electively), and want to hit refurb U values for heat throughout: so are drylining external walls with Kingspan (i.e .phenolic foam): and re insulating the
attic room, which had a joke of a conversion done inthe 90s: no insulation (got building regs though !!!!)

Going a little over board with the insulation as the marginal cost to over egg it is slight: Ill never come back and do it again. so: we are putting 50mm celotex + 77 mm kingspan (i.e. 12 mm plasterboard bonded to 65mm phenloic foam if i remember rightly), "dot and dabbed" into alcoves... and behind this is an ideal place to run cables, clipped tothe external brick wall (i.e. insulation one side, solid 9 inch brick the other). Reading my regs carefully : if the lighting is done in 1.5mm: all is fine: it could be derated 50%; and I'm fine for a 6A MCB. 2.5 mm ring mains is the difficult case. nonethelss; I can apply installation method 102, p265.. and table 4D5 p282 says I'm fine...

so ... next case to be mentally ticking off is... if I go through say 100mm of this stuff anywhere; how do I derate the cable ... ? (e.g. when coming from behind the board to a plasterboard double 13 socket/ patress)

p104 reg.523.7 & table 52.2 looks good for this : "cable surrouned by thermal insulation".. BUT: it has the caveat "for insulation having thermal conductivity greater than .04 W/m/K...
BUT Kingspan/ celotex is .022 or so. Holy Crap !!does this mean that to go through as little as 1 cm I have to derate "50%" ...? this would be mad: if i do that: I can't use a 32A MCB: (2.5mm -->13.5A..., 4mm is still not enough...... so 6mm ring .. ?!?!?? crazy !)

It seems bizare the 17th edition as far as i can see almost pretends that ceotex/ kingspan deosn;t exist, yet it is now very very common as the only practical way to hit the modern thermal efficiency targets.

Any suggestions as to what regs allow me to sanely justify what I should be doing ?


As i read it; as long as the cable is "behind" the insulationl; withinsulation on one side, the standard rule of thumb of "2.5mm ring / 32 A mcb " for sockets holds. as soon as i go through a slab of this stuff I'm screwed.... ?!?!?

what would you all do ?

 

[NotHimAgain]

what would you all do ?

Probably nothing - but I am like that .

Table 52A is based on research by the ERA and was introduced in the first release (Red Cover) of the 16th Edition.

It had always been assumed that cables covered by thermal insulation over short distances would not be subject to stress as the heat would dissipate along the cable to the area that was not covered by insulation. The ERA research showed that this was not the case. However, the research was limited and has not, AFAIK, been extended. You will note that it only applies to cables up to 10mm2 - this is not helpful when assessing large cables passing through fire barriers in switch room walls.

When 52A was first introduced the thermal conductivity was 0.0625 W/(Km) and it stayed at that value until the 17th Edition.

So the figure has been up rated but not to the level you require. Now this table goes back to a time when the ERA would do bits of research like this for the fun of it . They are now a commercial organisation and they have to make money. This means that someone has to pay them and JPEL/64 are unlikely to do that.

The up shot is you have to decide - BS 7671 does not offer assistance.

You could mitigate the problem by ensuring that the insulation is not in direct contact with the cables where they pass through it. IMO, it is unlikely to cause any problems - but you must decide.

 

[chapeau]

It had always been assumed that cables covered by thermal insulation over short distances would not be subject to stress as the heat would dissipate along the cable to the area that was not covered by insulation. The ERA research showed that this was not the case. However, the research was limited....

That's interesting. There are very few materials out there with a higher thermal conductivity than copper. Is there a link to the actual paper?

 

[chapeau]

Actually, having given this some more thought, the effect of this insulation on the cable can be worked out from theory and you don't need to do any practical experiments to find it out. The way that an electric current heats copper up is known, as are the thermal conductivities of copper, rockwool/kingspan and pvc are known, as are the dimensions in 2.5mm2 cable. The cable cannot be at > 90C along it's length.

So all we need is somebody to work it out. This would have been a simple A-level physics question in my day, though my day was in the 1970s, and I have since forgotten how to do it. So I turn it over to somebody else.

 

[NotHimAgain]

It had always been assumed that cables covered by thermal insulation over short distances would not be subject to stress as the heat would dissipate along the cable to the area that was not covered by insulation. The ERA research showed that this was not the case. However, the research was limited....

That's interesting. There are very few materials out there with a higher thermal conductivity than copper. Is there a link to the actual paper?

Why don't you ask them - try here first http://openlibrary.org/a/OL1337684A/Electrical_Research_Association.

 

[chapeau]

Well, i suppose as it's customary for the person referring to research to cite the source I thought I'd ask you first

Your link isn't very helpful. However a google search threw up http://www.scribd.com/doc/21665568/Current-Carrying-Capacity which appears to have all the formula in for working it out from scratch.

Not smart enough to do it myself anymore

 

[NotHimAgain]

Well a quick read of that should convince anyone that the best solution is to avoid the problem. I would just make sure the insulation is not in direct contact with the cable where it passes through .

 

[chapeau]

Indeed

What I would say is:

The tables give a derating factor for insulation > 0.04 Wm-1K-1. The OP wants to use insulation with conductivity of 0.02 Wm-1K-1.

These are pretty much identical, especially so when compared to the conductivity of the copper (400 Wm-1K-1) and the difference is negligible when compared with the copper.

The vast vast majority of the heat generated at the centre of the insulation is going to be lost via the copper in both cases. I would use the standard tables.

If LABC came in and picked up on it I would make the effort and calculate it, but I have far better things to do.

 

[jonnyhifi]

Well: It seems to me the pragmatic solution here, is to "over bore" the hole you drag the cable through; --- since this is slab insulation, it won't slump after the event. As a rule of thumb probably the same cross sectional area againleft as "free air", as the cable, and ensuring separate crcuits are drawn some distance apart.

All a bit handwavy: but clearly this common situatin is beyond the specific scope of the regs, and doesn't sound like it will change any time soon!

It also implies that where mounting plaster board patresses into this stuff, the insulation should be relieved for the wires being poked through from behind, more than the simply "being poked through" clearance would (not) give you.


Thanks guys.. all very illuminating.


The difficulty with these kind of pragmatic fitting details, is that nothing is really clearly visible later as an attempt to "comply": so much as what will look like sloppy hole making but I guess to have worked that out is important. The devil is always in the details of course... !

My guess is 99.99% of electrician s have never worried themselves about this one,and never will... the things that never even get though about for periodics eh ?

Jonathan

 

[ColJack]

I wouldn't "come through" the foam anywhere..
just fix a 47mm backbox to the wall and use an extension box or two to bring it flush ( or as near to.. ) the front of the board..

yes I know that this means you need a 47mm backbox and 2 x 35mm extensions for each outlet, but at least you won't have clearance problems with any flat plate sockets you decide to use..

 

[jonnyhifi]

So... have 90mm or so of box present, being made of 2 separate ones "stacked" ?


I see the logic: equally i could just cut straight throught the insulation so there is no insulation right the way through the slab, fo the full width and height of the patress box..... but then.... I start getting a big cold bridging problem... I was reading elsewhere earlier in skeilings in attic conversions; the problem of flush mounting GU10 fittings, if cut in with a ventilated cold roof, corroding badly internally over as little as 18 months: due to condensation inside...

Mmm life is all compromises...