Retrofit in-screed UFH

Joined
21 Sep 2021
Messages
7
Reaction score
0
Country
United Kingdom
I'm renovating a small 3 bed 70's semi-detached property, with ~50m2 downstairs floor area. modern double glazing, cavity wall insulation. No central heating currently, and i've opted for a Viessmann 35kW Combi (ASHP cost prohibitive, which is crazy!)

I've opted to fit oversized rads upstairs and wet UFH downstairs. Already dug up the screed, and limited to floor buildup height by external door sills.

So i'm working with 90mm, not including floor finish. Which leaves me with 40mm PIR (0.022W/m.K) and 50mm of liquid screed.

Has anyone got any practical experience if this is 'enough' insulation, will my house be warm in winter? Equivalent to rads etc

I realise this is below current regs, but this is the best i can do with the depth i have to play with. 200mm pipe spacing. Could really do with a word of experience from someone who knows.

Cheers
 
Doubt you'll find anyone with practical experience, as its too thin really.

The room will get warm, but will loose heat more quickly. Which in-turn means more gas used.
Presume you are down to the concrete slab, which is why you can't go deeper?

If you are worried about heat, you could bring the pipe centres closer.

Have the UFH and the rads on separate zone valves, you can then either on/off independently.

That said, these retro-fit spreader plates sit in a sheet that is thin, they must work.
 
Thanks for your reply Mr Chibs, yeah i'm down to the slab. Rads and UFH are set up on different circuits, with their own flow and returns and valves.

In my mind i know 40mm isn't enough, but (as far as i know) the only other options are
* more insulation (floating floor) and routed board...£££ and potential for creaky floors
* or radiators!

Can I really not do any better than radiators here?

cheers
 
What had you intended to have as the floor covering for down stairs?

I ask, as this has an effect on the performance for UFH.

I was going to have wood in my kitchen, but after research, deemed it not that suitable, so went limestone instead, this works great with UFH.

One thing worth noting, is there is a diminishing return with insulation, once you've reached a level twice the thickness of insulation is not twice as effective.

Think of a freezer, this only has something like 30mm insulation around it...
 
I think you might regret 40mm insulation.

what you are doing falls under building regs, as you are replacing a thermal element, so insulation should be to b/regs

personally I would gun up the slab and lower the floor, then put in 100mm insulation.
 
What had you intended to have as the floor covering for down stairs?

Wood effect porcelain tiles is the plan, to hopefully give the heat an easy route up

I think you might regret 40mm insulation.

what you are doing falls under building regs, as you are replacing a thermal element, so insulation should be to b/regs

personally I would gun up the slab and lower the floor, then put in 100mm insulation.

Yeah i understand your point, Notch. It is a retrofit and not a new build, so it was always going to be a compromise. Didn't have the appetite to dig up the slab as its our first renovation, and now its been plastered so its definitely off of the cards.

I understand its not going to be as good as it should be, but I'm still not sure if it will be good enough. I've done the heat loss calcs based on Jeremy Harris' calcs spreadsheet, and at the worst case heating requirement there is 24% heat loss. 100mm insulation hits the 10% loss limit on the nose. Doesn't look like too much of a compromise...although they are just numbers!

1) Insulate floor, change tack and fit rads downstairs?
or
2) Insulate floor, send it with UFH and see what happens?
 
I'd go 2, it's worth a risk, even less efficient UFH with tiles will give a more comfortable warmth than rads.

Don't forget decoupling membrane for your tiles:eek:
 
Ever considered a low profile overlay system?
I haven't to be honest, what are the benefits? I thought the floor will still need screed, and if the pipework doesn't go in it, the floor height will increase won't it?
 
I haven't to be honest, what are the benefits? I thought the floor will still need screed, and if the pipework doesn't go in it, the floor height will increase won't it?
Nope ... A low profile overlay system uses an formed EPS400 foil covered board (22mm) to lay the pipe into, you then lay whatever on the top. Some advantages are its low profile, doesn't need a screed layer, much faster warm up times and doesn't require a 'mass' to be kept warm.

Can be overlaid with engineered wood, tiles, cementitious boards etc. I've laid it in a few situations and now used it in my own property and have found it to be excellent without the need to dig out slabs and pour/lay screed.

In my own house I have a concrete slab, 40mm PIR, 22mm P5 flooring then the 22mm EPS400 boards and in different rooms - engineered wood - tile - and 6mm cementitious board with an underlay and carpet on top. It all works supremely well once the system was balanced properly.
 
Nope ... A low profile overlay system uses an formed EPS400 foil covered board (22mm) to lay the pipe into, you then lay whatever on the top. Some advantages are its low profile, doesn't need a screed layer, much faster warm up times and doesn't require a 'mass' to be kept warm
How flat does the subfloor need to be? i've got just less than 20mm variation across the ground floor at the moment. Does it bounce at all? The cost of the routed boards and cement boards/OSB overlay bumps the price up a fair bit.
 
You wouldn't over lay with OSB then another floor covering as wood insulates, you put the final floor covering directly on top unless it's carpet, that's when a cemintitious board would be used. Not really noticed any bounce and the EPS400 board is very dense and should be fixed/glued to the floor and itself. I actually screwed it to the flooring as I have P5 boards below it. That being said there is a little bounce from the engineered wood but it has a 3mm foam underlay and it is a floating floor so movement is expected, I only have tiles in the bathroom and not really noticed much but it is well laid on a membrane.

If there is a variance I would self level the floor first.

I guess the extra cost is down to it being a hybrid solution, the best approach of course is screed but where height doesn't allow then it is a middle ground solution.
 
If building regs is your main problem, there is a caveat that could help you:
upload_2022-5-6_19-44-48.png
 
Back
Top