Floor slab insulation. Test my logic please?

[Nickfromwales]

29 minutes ago, saveasteading said:

the screed looks like being between generally 55mm ish, but from 45mm to  85mm due to some extreme wobbles in the floor.

That’s one heck of a wobble! These are original floors? 

[G and J]

We used the u heat version (yellow rather than blue, but looks identical) the upstand on our was stiff enough to be self supporting

 

Apologies for poor image but there's nothing supporting it in front of the sliders and we didn't staple, but as you can see screed was traditional sand and cement

[saveasteading]

10 minutes ago, Nickfromwales said:

These are original floors? 

Yes there is 35mm variation on the slab, with dips occurring at various areas.  It is an existing agricultural slab which was good enough for tractors and barn use. 

There are a few local lumps too where the slab was extended or adapted but I'll have them chipped off.  Overall though, now that the rain doesn't get in, you wouldn't notice the variation, except that the sole pate is all level, packed up from zero to 35mm. The insulating slabs will follow the shapes smoothly enough. Now that the rooms are formed we can use different thicknesses of eps on the bottom, and optimise the use of insulation rather than thick screed.

But sods law combining with likelihood has it that the biggest room also has the maximum variation within it.

[Nickfromwales]

20 minutes ago, G and J said:

We used the u heat version (yellow rather than blue, but looks identical) the upstand on our was stiff enough to be self supporting

 

Apologies for poor image but there's nothing supporting it in front of the sliders and we didn't staple, but as you can see screed was traditional sand and cement

My preferred choice as my mate gets these traditional screeds perfectly flat and level every single time. Plus, he can manipulate levels for sliders etc which is a lot harder to do with wet concrete or liquid screeds.

[saveasteading]

21 minutes ago, G and J said:

the upstand on our was stiff enough to be self supporting

Ok I'll maybe buy one and fix it and give it a prod test. You appear to have a pump mix concrete that is 200mm thick or so. We will have 45mm to whatever self-levelling screed.

Is the yellow stuff stapled on?

[saveasteading]

3 minutes ago, Nickfromwales said:

My preferred choice

you mean the yellow version? JG? Is it more robust?

 

5 minutes ago, Nickfromwales said:

perfectly flat and level

Define flat and level please?  3mm under a 3m straight edge with an overall tolerance of x?

 

Your mate gets it right but a lot don't, from what I read.

[Nickfromwales]

1 minute ago, saveasteading said:

you mean the yellow version? JG? Is it more robust?

 

Sorry, I meant the dry sand and cement screed. The perimeter insulation is the same stuff from anywhere, just the colour changes. 
 

The differences are: height, does it have a polythene apron (for liquid), does it have a self adhesive strip on the back or not.

 

9 minutes ago, saveasteading said:

We will have 45mm to whatever self-levelling screed.

Ok. 
 

If it’s a poured aka liquid screed then you’ll be best off getting the foam skirting with the attached poly apron / membrane.
 

Self adhesive strip helps to just lay this out, but cannot really be relied on, so you’d end up with some kind of additional, mechanical fixing (staple / round head nail).

 

Does the Clockwork screed company operate near you? They’ve just poured cementitious self levelling screed at my Manchester project, and they were the best poured liquid screeders I have seen to date. And we’re not the most expensive quote too!

 

https://www.clockworkscreed.co.uk
 

With nobody EVER wanting to do this twice, or to have to revisit it, I’d defo get these guys in. They mix and pour from one super wagon. A very impressive setup. 

[G and J]

37 minutes ago, saveasteading said:

Ok I'll maybe buy one and fix it and give it a prod test. You appear to have a pump mix concrete that is 200mm thick or so. We will have 45mm to whatever self-levelling screed.

Is the yellow stuff stapled on?

The screed was 100mm thick, they used a pump, but it was a dryish mix, Josh shovelled sand (16ton) and cement to match like a trojan....and Jason achieved a really flat result, very happy with the job they did.

 

No we didn't staple the upstand at all.

[Nickfromwales]

Remember to get these tight into every corner as the curves and divots all need filling / foaming / cutting back etc afterwards; if you’ve not managed this by prevention, then it’s a LOT of time and effort to get these filled and made airtight with foam / passive purple / AT tapes et-al.

 

Put lots of time and effort into getting these filled foam strips put in as tight as you can. I cut and stop / start where it’s easier than forming a turn with continuous strip. Zero detriment when you stop / start, just tape over the joint. 

[Nickfromwales]

18 minutes ago, G and J said:

The screed was 100mm thick, they used a pump, but it was a dryish mix, Josh shovelled sand (16ton) and cement to match like a trojan....and Jason achieved a really flat result, very happy with the job they did.

 

No we didn't staple the upstand at all.

You don’t have to fix the upstand with a dry S&C screed, as the installers just flick a few shoots of mix against it and it pins it perfectly in place. 
 

The issue of it lifting slightly, and the screed going under it, is very real, so to remove unnecessary risk it’s always good practice to fix it so it stays down and in full contact with the insulation or sub-floor. 

[G and J]

12 minutes ago, Nickfromwales said:

Remember to get these tight into every corner as the curves

Now he tells me......

[Nickfromwales]

Just now, G and J said:

Now he tells me......

You never asked for your 1hr free consultancy 

[saveasteading]

1 hour ago, Nickfromwales said:

The issue of it lifting slightly, and the screed going under it, is very real

For other readers, it is indeed a real and a very bad possibility. 

I have been there when it happened. Our  subcontractor didn't tape the joints properly. Our plumber had insisted on doing everything from insulation up. and it was the only time ever that we relinquished control of the slab concrete. It must have been 20 years ago and he insisted he was experienced and that it was fundamental that he had the package. I became wiser on that day,

 The pump-mix concrete got under the PIR which floated up to the surface.  

Suddenly I became the only person with a clue what to do, and I had them

1. stop pumping.

2. bring in concrete blocks, tread the pir down and lay  the blocks on them.

after about an hour the blocks came out and levelling continued.

 

There will still be a skin of concrete under that pir, but it all turned out ok.

 

 

In hindsight I think they taped the joints in the dpm but either not properly or the concrete was spouted at an overlap and forced it off. Plus there wasn't an edge detail (the edging under discussion here, did not then exist).

 

it was potentially very horrible, and the need for sealed laps and careful pouring was demonstrated. Thus I will supervise and inspect. Maybe I will do the first room myself to test for any difficulties.

 

Any recommendations for a good sticky tape? 

[Nick Laslett]

On 03/01/2026 at 14:14, dpmiller said:

graphite EPS takes it to 0.031 or so

I believe that the graphite EPS not only has slightly better U-value, but also reflects more heat back at the source. Which I don’t think u-value measures but I might be mistaken. BASF Neopor is the stuff used in my ICF blocks. 

[JohnMo]

11 minutes ago, Nick Laslett said:

but also reflects more heat back at the source

Doubt in an UFH system. There is no space for radiation to occur.

[ADLIan]

The effect of the graphite is accounted for in the lower thermal conductivity compared to standard EPS so not dependant on application.

[SteamyTea]

Graphite has a k-value of around 200 W.m^-1.K^-1

 

 

[saveasteading]

I wonder if it is hyped the same way as 

Multifoil, ie tested in a particular context then sold in another. A reflective surface only works if there is somewhere to reflect it to.

 

"Graphite is a good electrical and thermal conductor".

[G and J]

7 minutes ago, saveasteading said:

ie tested in a particular context then sold in another.

Isn’t that the root of the cladding problem with Grenfell?

 

I can easily see me falling into that particular pitfall.  Trouble is, when one is scribbling away, it’s so seductive thinking one has come up with a really neat solution (“why had nobody else thought of this? Gosh I’m clever!”).

 

Sigh. 

[saveasteading]

3 minutes ago, G and J said:

problem with Grenfell?

I'd have to check the facts but I recall that there was more than one factor, one being that a fire test sample was 'adapted' to perform better.

[JohnMo]

15 minutes ago, saveasteading said:

I wonder if it is hyped the same way as 

Multifoil, ie tested in a particular context then sold in another. A reflective surface only works if there is somewhere to reflect it to.

 

"Graphite is a good electrical and thermal conductor".

I asked AI got this answer when asking about graphite embedded in EPS insulation 

 

Value is added because the "air space" required for graphite to reflect radiant heat is actually located inside the material itself, not as a separate external gap between the insulation and the concrete.

1. Internal Reflection Within Micro-Cells

Graphite insulation, specifically Graphite EPS (GPS), is composed of individual closed-cell beads. Even when tightly embedded in or under concrete: 

The "Air Space" is Internal: Each bead contains millions of microscopic trapped air pockets.

Infrared Mirroring: The graphite particles are infused into the walls of these tiny internal cells. As radiant heat attempts to travel through the board, it hits these graphite "mirrors" and is reflected back and forth hundreds of times within the cellular structure.

No External Gap Needed: Because this reflection happens at a microscopic level inside the insulation board, the board can be in direct, "intimate" contact with concrete without losing its reflective benefits. 

2. Physical and Efficiency Gains

Even without an external air gap, graphite adds measurable value over standard white insulation in concrete applications:

 

[saveasteading]

It's BASF so not likely to be untrue.

I've found a thing saying the eps nodules work identically, being bubbles of air, but the conduction route over the bubbles' convoluted path is less conductive because infra red gets mirrored back whence it came. Hmmm.

 

Anyway its quoted insulation is 10 % better than white eps, but I think the publushed cost is double, and I wouldn't expect so much discount.

Edited 8 hours ago by saveasteading

[dpmiller]

when we were ordering ours, white EPS was £11.57 for a 50mm, vs £13.70 for graphite

[saveasteading]

11 minutes ago, dpmiller said:

11.57

Per board or m2?

At present my best price for 100mm white eps is just over £15/ board of 2.4 x 1.2.

 

[dpmiller]

2 minutes ago, saveasteading said:

Per board or m2?

At present my best price for 100mm white eps is just over £15/ board of 2.4 x 1.2.

 

/board. A few years ago mind...