'Thin' Underfloor Heating

[Richard Downing]

My bungalow has a big sitting room with one wall made of double glazed panel - the view is fantastic and worth the energy losses, it's good double glazing though. It also has a 9 foot ceiling, so there is a lot of air to heat. (House has 300mm loft insulation and cavities filled)

Currently there is an enormous double radiator on the back wall providing the heating and although we can easily get the room warm the air flow is the wrong way round. Cold air forms on the windows and runs down onto the solid concrete (Cork Tiles) floor and then rises from the radiator to the ceiling. I'd like it to run the other way so that warm air came over my feet!

It seems to me that the best route to this is to replace the big rad with a big rad on the floor - I.E. underfloor heating, and a 'conservatory pack' would be about right cost and output-wise to do this.

The trouble is that I don't have much floor 'depth' to play with. The three brick walls are fine, but there is only about 85mm under the window. If I put down say, 50mm of insulation there is only room for

30mm of slab. Using just 20mm of insulation doesn't sound very 'building regulation' friendly, in fact neither does 50mm. Any ideas?

R.

[Owain]

I wouldn't use underfloor heating - as you say you haven't really got the depth.

You don't need to heat the whole room from the glazed end - just provide enough heat at the bottom of the window to counteract the down-draught.

You might find (surface) floor-mounted electrical tubular heaters sufficient - the put out about 40/60W per linear ft and at about 90mm high would not interfere with the view. Ideally, it would only need to run when the radiator (CH pump or zone valve) was on.

It would be a lot cheaper to experiment with a couple of them (screwed to moveable planks before doing a permanent install) than laying a new floor. Cheaper to run too if you were going to use electric underfloor heating. If you have small children check the touch temp of the tubulars first, for safety.

You might also consider a ceiling fan to reduce stratification of all the warm air above your head.

Owain

[MikeH]

So presumably you're proposing to raise the existing finished floor level to allow the insulation and UFH system to be installed on top of the existing concrete floor? Or have I misunderstood something?

I've recently had a concrete floor removed back to bare earth in order to install a new slab, 100mm of insulation and an 85mm screed containing the UFH pipes. From what I've read, the secret of a successful UFH installation is the insulation that prevents you heating up the earth under your house. Maybe you could get away with 50mm, but you'll need a lot more than 30mm for a screed unless you're considering an electric system. Even then I suspect the screed will need to be a lot thicker (or otherwise strengthened) because it will be laid on top of what is basically a fairly flexible material and might otherwise be prone to cracking. For wet systems I believe the minimum thickness of screed is closer to 65mm.

Can't comment on what the building regs might require.

Mike

[Ian Stirling]

You've got several options.

Firstly, dig up the existing slab, put in insulation underneath it, and then re-pour, over UFH pipes. This is possibly the best way to do UFH - won't move, and a large thermal mass to buffer the temperature in the room.

You say 'big'. I'm going to take a wild stab in the dark, and assume

6m*6m. This is about 40m^2.

25mm of kingspan will do .8W/m^2/K. If the ground is at 20C below the room temp, you're looking at around

1000W over the floor area. In practice, the ground will also insulate somewhat, so maybe 800W or so.

Doubling the thickness of kingspan helps quite a lot.

If I Was You.

The easy way.

75-100mm of kingspan, 50mm or so of screed, in which the pipes are in, 10mm of flooring. At the edge of the room, a 15cm or so strip of 25mm of kingspan, and flooring over it, with a step down.

Into this 'step' next to the window go concealed uplighters, and maybe a plant or two, depending on your decor.

'It's not a bug, it's a feature'.

[Richard Downing]

Thanks for the replies, they confirm what I thought. I'm now thinking maybe I could refit the windows 'two bricks' higher (I just looked at the way they meet the ceiling and this is feasible). This would give me the depth for a proper insulation layer and thick hot-slab.

SWMBO doesn't like the idea of tubulars or a ceiling fan.

I'll get a quote for someone to dig out the slab and re-lay though, at my age I don't fancy swinging a pick :-)

As with all my projects, I'll think about it for a few months.... You might hear about it again.

Thanks once more,

R.

[Owain]

Could you build an internal wide/low "window sill" with gratings and hide the tubulars or whatever inside it?

Owain

[john.sabine]

Interestingly, I was chatting to a bloke who specifies UFH the other day, and he reckons this conventional wisdom is cobblers.

He reckons that the heat movement through the earth is such that you're wasting your time insulating: OK, this means you heat the top 30cm or so of the soil, but once it's up to a reasonable temperature it simply becomes part of the thermal mass of your home and provides extra temp stability.

Accordingly, he specifies some edge insulation (so you don't lose heat to the ground outside your walls, and recommends nothing under the main body of the screed.

It all felt a bit wrong to me, but on the other hand the numbers he showed me (none of which I can remember) seemed to hang together, and my own laziness means I want to believe him.

John

[Ian Stirling]

Well... There is an unfortunate problem with this. Namely that it's bollocks. The argument that you heat the top 30cm is rubbish.

If you heat a 10m circle (for simplicity) on the ground to 30C, then what happens? Assuming that there is uniform sand underneath, with no heat flow. (the best case).

So, 30cm down, in the middle of the heated patch, the heat just stops? Of course it doesn't - you end up - over a long period - heating to more or less the temperature of the slab a sort of squashed hemisphere of sand, maybe 2-3 meters deep. This conducts the heat away to the rest of the earth.

Google perimeter/area ratio sand u-value.

[john.sabine]

I didn't say I believed him, just that I wanted to ...

Well, no, of course not - I was guilty of over-simplifying.

His argument was that the temperature gradient in the soil would be such that you would get a significant temp rise at the top (I think in about the top 30 cm or so), tapering to a minimal rise by the time you were a metre or two down.

Aye, this is the bit I had difficulty with. I guess his notion is that there is no significant heat transmission to the bits of the rest of the earth that aren't under your house.

John

[Ian Stirling]

Yeah - the problem is - if this is true, you've got 30cm of earth doing the same job effectively as the insulation in the walls. Unfortunately - perfectly dry sand - as an example - insulates as well as 1/10th its thickness of rockwool. And moist - well, you don't want to go there.

I wish it could be so simple. I'm in the annoying position of actually having calculated all of this for my house, in the approved manner, and for the numbers to actually match my energy bills pretty much.

Annoying - as it means that the 'long pole' in the tent in my case is the walls, (by a factor of 4) which are really, really annoying to insulate.

[The Natural Philosopher]

My 2000 system was 50mm poly and 75mm screed. To regs.

My be more these days. Should have gone 100mm insulation.

[The Natural Philosopher]

Hire a kango.

[The Natural Philosopher]

This is to an extent true. The building regs on floor insulation take it into account.

Sadly in my case it was a suspended concrete floor with a vented airgap under..and boy does that lose heat on cold windy days.

However just earth alone will not give you the all U values you need, and it will make for a very long time constant although it does make for a nice micro climate just outside the walls for tender plants.