Thermal break required

[eandg]

We have a precast insulated slab which we had intended laying UFH onto then pouring 50mm liquid screed. Had overlooked what appears to be an obvious issue of requiring some form of thermal break to prevent the UFH simply conducting heat below. And ideally one that's as minimal depth and as cheap as possible given we have little wiggle room on the appertures (kit may need to be re-engineered) and budget. U-value of the floor is 0.11 so not necessarily needing any additional thermal insulation just something that'll provide a break and ensure UFH is efficient. Any thoughts?

[Barryscotland]

What’s the make up of the existing slab?

[eandg]

It's 325mm EPS with 50mm structural concrete topping, with concrete deeper at the joins. 

[IanR]

Have you poured the Slab yet are or you at design phase?

Most of us with insulated slabs/rafts, put the UFH in the concrete without a screed. But, typically, minimum 100mm thick, rather than 50mm.

Who spec'd the slab?

 

If you've poured already, 60mm screed on 50mm slab, is pretty close to what most of us have, ie. 100mm concrete. UFH works fine. Some have 200mm concrete an still report no issues.

[eandg]

Slab is precast, specced by us (self designing) and signed off by engineer and for warrant, but not yet installed. Wunda advise we'll need a thermal break for UFH but that may not be the case then?

[IanR]

I'm not understanding your build up. Is the slab ground bearing?
ie. you are placing EPS on a prepared sub-floor that is sitting on earth, and then you plan to place precast concrete sections on to that EPS?

What are your load bearing walls sitting on? A Wall-Floor Section would help.

Wunda probably isn't expecting you to have 325mm insulation under your slab, hence the thermal break to the screed.

[eandg]

No, it's a precast insulated concrete slab suspended on blockwork creating a void to a groundbeam which sits on piles - we're not falling down. Have advised wunda of build up but still advising they need a thermal break which I hadn't planned for. 

Edited December 6, 2021 by eandg

[IanR]

Is there a cold bridge through slab -> block work -> ground beam -> piles?

 

I can't imagine there's EPS sandwiched between Slab and Block work? There wouldn't be the surface area to take the loads...would there?

 

[JohnMo]

Do you have a drawing?

[eandg]

23 minutes ago, JohnMo said:

Do you have a drawing?

I do but not on my phone unfortunately. It's just basically a spantherm type slab. 

 

But it is basically the same detail as the screenshot (taken from spantherm website).

[SteamyTea]

So is it basically block and beam, with insulation under it?

[eandg]

4 minutes ago, SteamyTea said:

So is it basically block and beam, with insulation under it?

Aye, could have saved myself some bother if I'd just said that!

[SteamyTea]

4 minutes ago, eandg said:

Aye, could have saved myself some bother if I'd just said that!

Would have been better saying it with numbers.

Talking of which.

If you can find out the thermal conductivity of all the connected elements, then you may be able to set the UFH pipework inboard from the cold parts and still not loose too much power "out the sides".  Just a matter of doing some calculations.

Edited December 6, 2021 by SteamyTea

[eandg]

9 minutes ago, SteamyTea said:

Would have been better saying it with numbers.

Talking of which.

If you can find out the thermal conductivity of all the connected elements, then you may be able to set the UFH pipework inboard from the cold parts and still not loose too much power "out the sides".  Just a matter of doing some calculations.

I'll need to go find and then work out what it all means! The perimeter will all be insulated (60mm insulation) with exception of doors which I imagine will help out the sides, it's just the potential to lose it down the way. 

[SteamyTea]

4 minutes ago, eandg said:

The perimeter will all be insulated (60mm insulation) with exception of doors which I imagine will help out the sides, it's just the potential to lose it down the way. 

It is all going to be a compromise, but if you can add extra in one place, that helps, especially if it is a side that does not see the sun.

[IanR]

I don't believe it sits on 60mm for the entire perimeter, EPS couldn't take the loads on such a small area, so the there's concrete going all the way down to the blocks.

 

 

The brochure gives a 2D Psi calc for the cold bridge, that looks reasonable, but it does cut the section where there is EPS between concrete and block plinth. Would be good to see the same calc done for the section where the concrete goes all the way down to the block plinth.

 

Having said that the overall U value they calculate (0.11W/m².K) includes the perimeter cold bridges. As long as they are not using smoke and mirrors then I feel the UFH could go in a screed straight on the precast sections. I probably wouldn't run the pipes too close to an external wall though, ie. at least 200mm away as there is still a bit of a cold bridge.

[eandg]

Thanks Ian - from the drawings it goes further but doesn't go all the way to the block plinth with the join to be grouted on site. There's various requirements (needs to be 215mm block for example) and calculations but I'm paying other people to work that out. This is the drawing provided by the manufacturer and from correspondence the thermal conductivity of the EPS is 0.031W/mK.

 

[IanR]

I don't believe that section shows he whole picture, as I mentioned before the EPS can't take the loads on to a 100mm block plinth.

 

This image shows it better

 

 

While the section you showed is not wrong, If you the moved the section along the span a little it would show pre-cast concrete going all the way down to the block plinth.

 

But, this image also shows that along the length of the pre-cast section, there's only 1 localised bearing point every 1m or so, which isn't bad. Although along the width of the pre-cast section it looks like 100mm of localised bearing point (cold bridge) for every 100mm on EPS.

Edited December 7, 2021 by IanR

[Olf]

With only 50mm to play with it will be hard to find something that efficiently provides thermal break. Bonding screed layer with the slab would be imho beneficial, as you get more thermal capacity (unless you are sold on quicker reaction times). With the cold bridges due to the specifics of the construction method I think you need to adjust the UFH circuit layout: pipes running along the areas where concrete forms bearing points would mean heat is transferred this path anyway (not contributing to space heating), so by avoiding laying near the perimeters you don't loose too much on house heating, yet reduce losses.

[eandg]

15 minutes ago, Olf said:

With only 50mm to play with it will be hard to find something that efficiently provides thermal break. Bonding screed layer with the slab would be imho beneficial, as you get more thermal capacity (unless you are sold on quicker reaction times). With the cold bridges due to the specifics of the construction method I think you need to adjust the UFH circuit layout: pipes running along the areas where concrete forms bearing points would mean heat is transferred this path anyway (not contributing to space heating), so by avoiding laying near the perimeters you don't loose too much on house heating, yet reduce losses.

Thanks; adjusting layouts would appear to make sense. 

[eandg]

Turns out I have a wee bit more room to play with and could just about get away with 25mm foil-faced PIR instead but itwould leave only 13mm tolerance for the plasterboard (i.e. between the bottom of the floor joist and top of the screed) which I'd rather not be paying to have trimmed. Anyone have any thoughts?

[Mr Punter]

Can you just double up the sole plate on the timber frame and increase the building height by 38mm?

[eandg]

5 minutes ago, Mr Punter said:

Can you just double up the sole plate on the timber frame and increase the building height by 38mm?

That's a (potentially very interesting) thought I hadn't had... will look into it, thanks.