What derating table do I use for multicore armoured thermosetting cable in insulation.

Hi Alan, why are you using SWA ?

I have to cut a hole next week to double check, but I think the cable is running less than 50mm from the surface of a cupboard built in to the roof space. As far as I can see the cable is sitting in or very close to the eaves of the roof,  the cupboard is built so that it also goes as close as possible to the eaves. There is a good chance that there's just 9mm of plasterboard between the surface of the cupboard and the cable.

Life would be much easier and cheaper for the customer if I could just leave the existing 16mm twin and earth in place.

Not too sure what you are describing. It looks like you want a 60A distribution circuit and have an existing 16mm2 T/E in place but installed as reference method 100 thereby giving a table 4D5 current carrying capacity of 57A. 
I can’t imagine a constant 60A demand in domestic circumstances but if you want to be squeaky compliant then you are correct to seek an alternative, which might include a more detailed consideration of the maths.

Personally, all other things being ok, it wouldn’t cause me a second thought to leave the existing arrangement in place. I am not even sure if I would write it up as a deviation on the EIC!

My main concern is a non armoured cable, without 30mA rcd protection running through the fabric of a building. If I put a 30mA RCD on the submain most of the house circuits are protected by a single RCD. If I wasn't making changes I would probably just code with a C3; but I am changing the CU on the supply and load side, therefore believe the circuit needs to comply with current regulations.

More questions I'm afraid..

Method 100 - that's only good for insulation of up to 100mm (fluffy stuff)- while 300mm is generally needed to meet current building regs in lofts - how much is actually there? (or could reasonably be added during the lifetime of the installation?)

If you're in a small space just below the roof surface, with little opportunity for heat (from solar gain on the roof) to escape into the living space, is a 30 degree ambient plausible?

What's the CU on the end of this cable going to be supplying? Traditional CUs rarely had to carry much more than 10A for most of their lives, with relatively short durations at higher currents (maybe 20A or 30A for an hour or so when cooking, or 50A when showering but only for a few minutes at a time) - so large cables barely had the chance to heat up before the loads dropped down again. Only ones supplying off peak heating were pushed closer to their capacity. While renovating my house I had everything, including an electric cooker and a fan heater to keep warm, running off an old 30A fuse without any problems at all. Some modern loads - EVs or heatpumps, or other long hour loads (immersion heater perhaps) might be challenging, but if it's just a few sockets and lights, maybe the circuit could be rated a lot lower?

Using 90 degree rated cable and allowing to run over 70 degrees in the loft - i.e. well away from terminations - might be one approach.

   - Andy.

The main load on the cable will be all the normal stuff you will find in a kitchen and utility, hob, oven, dishwasher, washing machine, tumble dryer. Plus sockets in lounge and a coupe of bedrooms. 

Wouldn't be surprised with a double oven in the future and fairly big hob. If there was an EV I would expect it to be connected via a different DB. It's a large house from the 1800's, heat pump technology needs to get much better for it to be a practical solution.

If heat pump or EV charge point is to be installed, they have a bigger issue that they are on a looped supply, new supply would probably enter at rear of property on overhead lines and current supply head is at front of property. It's actually tempting to ask for permission to connect an EV now to force unlooping, then board with all the heavy loads would probably be near where the supply enters the house and what would then be the sub board would just be powering a few bedrooms, office and hallways. But that would put the project back a couple of months, which is far from ideal.

Neighbour probably won't be impressed because the DNO will discover that their supply cable is running from back to front of his property, concealed in the fabric of the building.

Before you ask I don't know what size the main supply fuse is at this point, now that I know it's looped 60A would be my guess.

Other thought, is rather than use 16mm² 3-core, use 25mm² 2-core (it's about the same diameter) and if you don't trust the armour as a PE for downstream, add a 16mm² G/Y along side (all permitted these days). The 90 degree version with a method C rating of 146A should be good for 63A under almost any conditions.

   - Andy.

I had considered 25mm two core. How does the oval shape work with standard swa glands, do I need a special gland. The widest diameter of the cable is the same as 3 core, do I gain a lot in flexibility. It's  a TT earth, so as long as I have some sort of earth path it will be enough for the 100mA time delayed RCD I will have up front anyway.

Alan B said:

I had considered 25mm two core. How does the oval shape work with standard swa glands,

A 2-core SWA is round, not oval.

Alan B said:

It's  a TT earth, so as long as I have some sort of earth path it will be enough for the 100mA time delayed RCD I will have up front anyway.

RCDs can't provide protection against overcurrent .. for protection against overcurrent, you'd need to take into account the maximum prospective fault current, which is not the same as the minimum fault current which is used for ADS.

Just worth a check there are no parallel paths that might increase prospective fault current.