UFH - insulation vs buffer size trade-off

[Nick Thomas]

Just trying to get a head-start on planning the UFH here. Heat loss survey done at the time of heat pump insulation puts the annual space heating demand at 6250kWh, and suggests a flow temperature of 30°C would save a good 500kWh of electricity/yr compared to 55°C, so if I can get it in relatively cheaply, it seems like it'd be a win:

 

 

(Weather compensation is on and the radiators never feel like they're at 55°C; I don't have a good idea of what temperature it's actually running at, so saving 500kWh/yr is probably the best case scenario).

 

It's a relatively-new-build, with a solid concrete floor downstairs. The original EPC gives the floor a U value of 0.28W/M²K, so presumably 75mm PIR or so underneath the concrete - feels insufficient? Although on par with the walls 🤷‍♂️.

 

 

Digging it all out is unappealing, especially after reading threads here from people who've done it! So I've been poking at low-profile insulation like https://www.kevothermal.co.uk/products-solutions/kevothermal-flooring-insulation/ . My issue is that if I put that down, it stops the existing concrete slab from acting as a heat store for the UFH, right? It'd just be using the new screed that goes on top of the insulation as a store, which is a completely different dimension of unappealing ^^.

 

Given a trade-off of (relatively) poor insulation with a large thermal store, or good insulation with a small thermal store, I'm not quite there with working out which one is better. Presumably it's all about required flow temperature - the larger store would let the HP run colder, which would reduce heating loss and increase CoP, to be offset against the losses made from the slab - but I'm not sure how to go about plugging that into a formula. Or does one of those terms dominate to the extent that it's not worth a calculation? Thoughts?

 

Keeping the depth of the new floor down is fairly important - the existing carpets are, I dunno, 12mm or so, but as a new build box the ceilings already feel a bit oppressive.

 

I came across these cool customers: https://www.jk-gb.com/jk-in-ground-ufh.php - which seems like it'd keep depth down to the absolute minimum, but is obviously incompatible with extra insulation. Anyone done anything like this?

 

I'm also wondering about screeding options that could be treated as the final finished floor, rather than needing to pop tiles or LVT on top of that, which would save space for insulation if needed. E.g. http://www.thermotech-screed.co.uk/products.html talks about "Decor Screed", which is just the regular screed with colour added, then polished to a finish. Would that work with any screed? Decent, or naff?

Edited January 17, 2023 by Nick Thomas

[Nick Thomas]

FWIW, while saving money on heating costs is *a* motivator, the biggest motivator is getting rid of the carpets, which I hate with a passion. That gets blocked by the UFH as I really don't want to be putting down a hard floor, just to take it up again later if/when it's UFH time ^^.

[Nick Thomas]

On 17/01/2023 at 09:29, Nick Thomas said:

I came across these cool customers: https://www.jk-gb.com/jk-in-ground-ufh.php - which seems like it'd keep depth down to the absolute minimum, but is obviously incompatible with extra insulation. Anyone done anything like this?

 

Just following this up - I got a quote off them today. They do the grinding and install the pipes quite happily, but you have to get someone else to pour a 3-10mm screed on top for most finished flooring types, and of course, that has to dry before the UFH can be used. To avoid that:

 

> NB: Tiles, Natural Stone and bonded wooden floors can be applied directly on top of the JK® system without any levelling or smoothing compounds. A suitable flexible, UFH suitable, adhesive need to be used in all cases.

 

but I was never going to do any of those.
 

Pricing detail:

 

 

So nearly £80/sqm (inc VAT) for everything except the covering screed and a bit of electrical work. Grinding the grooves in a sand/cement base is ~£33/sqm (+VAT) at my scale; doing it in concrete or tile adds ~£12.50/sqm to that. The kitchen is (badly) tiled so the idea that they could put the grooves straight into what's there, rather than me having to take it all up, is attractive.

 

All in, probably too much for me, and more disruptive than I was hoping ^^ - but maybe useful information for someone else.

 

I was wondering how difficult it would be to DIY with a wall (floor!) chaser; suspect the answer is "too difficult".

Edited January 20, 2023 by Nick Thomas

[Roger440]

Hire a groove cutter and do that bit yourself? https://www.flooriq.co.uk/why-milling/#

 

Ill be doing this later in the year. On a much bigger scale.

 

Im confused. You say you dont want carpet. So thats a hard floor. Which means no screed necessary. You then rule it out?