Underfloor heating with less of the "under floor"

[Tim S]

Interesting...

As some may remember, I'm fixing a 1950's bungalow, with a solid concrete uninsulated ground floor.[1]

I'd ruled out insulation, believing it had to go *under* the screed (or slab). Turns out I was wrong, when I came across Marmox board - a polystyrene insulating board faced with fibre glass/cement which can

*allegedly* be tiled straight onto (compressive load 30 tonnes/m2 according to manufacturer).

I then ruled out UFH, believing electric was the only option in this case.

Ho hum: While I was looking to see if there was anything cheaper than Marmox, came across this:

"Easy panel" is the product of interest - 25mm polystyrene with grooves to take 12mm or 16mm PEX pipe for wet underfloor heating. 30t/m2 compressive load, 15t/m2 prolonged.

Looks like the company is importing a Swedish, or at least Scandanavian product. Claims you can tile straight onto it. Had a response to an email earlier to them, just waiting to see the cost of the insulated 25mm board (non insulated products are priced on the website).

*If* it compares favourably to Marmox + cost of rads+vertical rads + fan convectors (which I won't then need) I shall have to give it serious consideration.

I'm not asking advice, because I'd be amazed if anyone here has actually installed this stuff (please speak up if you have or know someone!) - thought it might be of interest though...

If the quote looks reasonable, I'll ask for an offcut to look at and do a test bond to the floor to see if it looks like it will actually handle the job. Really don't want to be putting tiles down and taking them up again before I'm dead, if I can help it ;->

Cheers

Tim

[1] Repeated my boss's measurements on Sunday: 1 thermometer 0.7m above floor, another 0.05m above floor. 4C difference - and my feet were freezing off through socks and shoes. Granted the building is only minimally heated for frost protection (I brought the air upto 12C with half a central heating system and a gas fire over 6 hours) - so the slab is extra cold, but all the same... It's probably going to be a cold floor if I just tile over it.

[John Stumbles]

you priced up getting the existing floor dug out and replaced with proper insulated slab with embedded UFH? No system using thin insulation on top of an existing uninsulated slab is going to give similar efficiency. Plus is the existing slab properly DPCed?

On a job I've been working on the builders did just this - to an area of about 7m x 3m total. Took 2 labourers about a day to dig it out, maybe half a day to level the area with concrete, maybe another half day to lay polythene DPC + 100m polystyrene (+ UFH piping), another day or so to concrete. They were mixing up the concrete in a barrowmix: on a bigger job with readymix one would obviously save some time over that.

[Tim S]

Hi,

John Stumbles coughed up some electrons that declared:

No - partly because the thought of the amount of work, mess and expected cost scared me (even if I didn't do the job!). Perhaps it might be less bad than I thought if I did price it, but it's too late now - I've made the decision not to do that and I'm too single minded to change my mind at this stage as work's started (which is probably a good thing IMHO with a large project). The idea of adding any insulation to the surface was an opportunistic one as it doesn't add much to the work and doesn't break the basic plan. The UFH is even more opportunistic - but I'm not commited until I prove it to not be an utterly stupid thing to do. At least the extra work for UFL plumbing is offest by no radiator plumbing.

If I think for one minute that sticking stuff under floor tiles is going to make them less than long lived, I'll scrap this and go back to the original plan as robustness and longevity is a higher priority for us.

Agreed. The whole thing started off due to discussion at work about how effective 25mm of Jablite was compared to nothing. Of course, there'll be more losses with UFL as it runs at 27C as opposed to the floor tiles getting to maybe 18C with insulation but no UFH. There is an element of a previous discussion suggesting losses away from the external walls may be less due to the ground retaining some heat - I'll never calculate that though. But a worse case calculation based on the board's k-value assuming a certain (constant) ground temperature will be easy enough.

No. None whatsoever. The email I got back from teh company said to use cementous tile adhesive to stick the panels down as what they'd normally recommend isn't very good in the damp. But I'm looking at a couple of coats of Aquaseal (Aqua Stop IIRC) onto the concrete / screed anyway for good measure. The floor has survived 50 years with a variety of floor coverings including vinyl, clay tiles and wood tiles with no obvious problems so I'm not particularly concerned. No mushrooms on the walls, though that may be due to what looks like a chemical DPC that's been injected some time ago (found the injection holes behind some skirting I pulled off last week).

That's interesting. The cost break point for me for 95m2 is about 2500 quid. That is the combined cost of Marmox board (stronger jablite in effect) and all the rads, soem vertical rads (not cheap but a necessity due to room shapes) and a couple of Myson fan convectors (also not cheap, but required).

I wonder why no-one's done a marmox type board based of PIR foam? Would seem to be an obvious product, unless there's a technical difficulty... Did scan the Celotex and Kingspan sites, but there's nothing identical in application.

I'll come back when I have some more info from them.

Cheers

Tim

[Colin Wilson]

One claim that strikes me as taking the proverbial was...

"Whereas an electrical system can easily be damaged irreparably by a lightning strike or power-surge, The Box system utilises a water pipe to heat the floor, meaning it's safe from any such damage."

I haven't heard of any internal underfloor electric heating suffering from a direct lightning strike personally, but i'm sure it's no more likely than a muppet drilling fixings into a wet floor system.

Oh, and then there are the (minor) discrepancies, like...

"It has been established through careful testing that the PEX water pipe used in the system could last, in use, for over 400 years."

"In special tests carried out by the Royal Institute of Technology, Stockholm, it is estimated that our water pipe could last, in use, for up to 500 years!"

[Tim S]

Colin Wilson coughed up some electrons that declared:

I noticed that. But PEX is PEX, so I'm happy with that bit - other people use it for UFH.

It's the ability of the panels not to deform and to remain stuck that I'm most concerned about.

It's very hard finding people who've tried this stuff - Marmox is nigh on impossible to verify and that's a more common product AFAICS...

Cheers

Tim

[fred]

In article , Tim S writes

I'm not convinced that composite boards are such an advantage in flooring applications, I think I'd prefer a layer of easily laid PIR foam sheets carefully sealed at joints, followed by your Marmox with staggered joints and bonded down with contact adhesive.

PIR foam sheets rated for use in floors are available.

[meow2222]

Maybe best bet is get a piece, put a tile on it, stand on it and have someone measure it accurately. Any movement and forget it.

NT

[Tim S]

[e-mail] coughed up some electrons that declared:

Sound advice - then I can make sure the DPM/adhesive/board/adhesive/tile bonds are all happy.

Waiting for a price by email - my ISP crashed for the entire day so email for me is backed up on other people's servers :(

Cheers

Tim

[Tim S]

Tim S coughed up some electrons that declared:

OK. The ISP lost their ATM uplink to BT's 20C network. Ow.

But it's back, I have the email and it looks like:

Marmox 20mm and EasyPanel 25mm [1] are *very* similar in price. PEX pipe and manifolds are comparable to the rads/convectors I would need.

So - insulation vs insulation+UFH is comparable. Either is 2k more expensive than doing neither (what I originally intended).

I think it would be worth getting a sample of the EasyPanel for examination - after I run a crude heat loss calculation...

[1] The PEX pipe is 16mm, so even in 25mm, some of the pipe will have a lot less insulation between it and the floor slab. 20 vs 2

Cheers

Tim

[Tim S]

Ah - seem to have found the actual manufacturer at last:

pictures and technical info in this PDF:

to rely on 0.1mm ali foil to strengthen the top layer, nothing apparent on the bottom face. Need Mapei Kerabond to stick tiles on.

[The Natural Philosopher]

Tim: the heatloss down to heat loss up ratio is entirely dependent on the ratio of insulation underneath to isnualation above `(flooring, carpets dogs/cats sofas etc)

50mm is probably the minimum for a screed floor but this should work well enough.

[Andy Champ]

Seems improbable.

If the floor's at 25, and the room at 20, but the ground underneath at 5 I'd expect a *lot* more loss downwards that if the ground underneath was at 15.

And I suspect the "U" value of several metres of dry soil could be quite high - whereas wet soil, with water trickling through it, could steal enormous amounts of heat.

Andy

[The Natural Philosopher]

You run UFH at around 45C water.

At that temp the differential between heat lost down to up, is much less dependent on outside temp: inside temp.

3 meters of wet soil is about the same as 50mm of polystyrene.

Bur the thermal 'mass' is vast.

takes days of heating to get it up to temp..

[Tim S]

The Natural Philosopher coughed up some electrons that declared:

And wet (well quite damp) soil is what I have.

And it's going to arrive at a mean away from the walls at this time of year, assuming room heated 50% of the day, of perhaps avg(5,27) = 16.

Repeating crappy calculation from earlier:

Area=19m2, deltaT=11C, U=1.63 - so total loss down is 340W

With 20mm Marmox and no UFH, lets say the mean earth mass is avg(5,20) = 12C ish U=1.3, deltaT=8C so total loss down is 200W ish.

With that scenario, we could say the "luxury" of UFH costs us an extra 140W for 19m2 room at 50% duty cycle heating. That's more like 2.50 quid/month for the colder months.

Most of the other rooms downstairs are bedrooms, so should be on a lower duty cycle, but as 2 are kids bedrooms, that may not always be true...

Hmm.

I'm off to have a look at doing better calculations...

Cheers

Tim

[Tim S]

Tim S coughed up some electrons that declared:

Marmox looks robust - both faces are surfaced with fibreglass bonded with a cementous polymer. Sounds solid stuff...

seems Ok on the top (0.1mm foil). But the bottom looks weak. How well is that stuff going to stick down, and stay stuck down?...

I'd have it under a floating floor, where minor compression isn't problem. I'm not yet convinced it's going to provide solid lasting support for tiles...

Sample time I think...

Cheers

Tim

[terry]

This is 'presumably?' electric cable underfloor heating? An acquaintance has underfloor water heating cables, tied in with part of an older installation of baseboard water radiators in the basement area, from a conversion from oil furnace to electric hot water furnace by a previous owner.

[Tim S]

terry coughed up some electrons that declared:

No - it's wet. Electric would be too expensive.

Cheers

Tim

[<me9]

You should be able to run the roms at a lower temperature with UFH, as the heat is where you want it. This will negate some of the extra losses thrpugh the floor.

[The Natural Philosopher]

yup.

I have a roomstat set to 18C and SWMBO says its 'fine with a jumper on'

By running 24x7, that 18C is utterly even: with timing, I used to feel cold until the stat showed 19-20C.

[Tim S]

The Natural Philosopher coughed up some electrons that declared:

That's interesting. Do you have a feel for whether it was much more expensive to run at 18C all the time.

BTW - found an approximation formula for U-values of uninsulated floors:

good reading material there)

U = 0.05 + 1.65(P/A) - 0.6(P/A)^2

where P=length of exposed perimeter wall, A=floor area.

I've got IIRC a P/A of 0.5, so base U-value of 0.73.

I'll try some numbers with that after I call the gas/leccy boards...

Cheers

Tim