Novel idea for UFH.

[The Natural Philosopher]

Or, the trouble when two engineers start talking.;-)

Basically a mate is redoing a room with a suspended floor, and we got to talking about radiators, and since he was doing underfloor insulation I said 'what about UFH under the boards' and we talked..but the boiler temp was probably too high..and so it was either temp reducers or radiators. Now I couldn't' think of a passive - non pumped - temp reducer, so thoughts turned to how to reduce the pipe temperature by other means..

Insulation. So the first thought was to slip insulating sleeves over the hottest parts of the pipes..but that just moved the 'hot spot' further down the pipe..and so I said 'what we want is insulation that gets thinner as it goes along the pipe ' and he said 'Nope. Do it the Radar way: slots in the insulation getting longer as you get further away from the pipe source'.

So there it is folks! the phased array underfloor heating system for suspended wooden floors that eliminates temp reducers.

Just lay copper - or plastic - pipe, and then get some slip over insulation and cut it into short lengths and starting at the hot side, put it on the pipe with a 3mm gap, then a 4mm gap, then a 5mm gap till at the far (cool) end of the pipe it's *all* gap!

Add a bit of balancer and a TRV and you have a UFH system you can run off a standard 85C boiler circuit with no mods.

I reckoned a pair of insulated feeders running *across* the joists at one side of the room full of 'T' pieces taking a loop of copper or plastic between each joist pair would do the trick. So hot up and cold down.

Whaddya think?

[Andrew Gabriel]

Over-engineered!

It's a suspended floor, so you use air as the insulator, simply by hanging the pipes an inch or so under the floor. For your balancing along the length, just run a loop up and back between each floor joist. As you go along the floor, the average temp of the two sides of the pipe run will be about constant (near as makes no odds anyway).

I've done this with 10mm copper, and it seems to be working well. Mindful of copper's low emissivity, I got a spray can of black paint and sprayed a thin film of it along the top of the pipe, to help the convection heating with a bit of radiated infrared too. I have a TRV poking out too.

[robgraham]

But how do you stop all the heat being blown away by the underfloor ventilation ? Rob

[Andrew Gabriel]

You'll need to insulate against that whatever you do.

In my case, it was on the top floor, so it was a ceiling underneath and I didn't bother (heat still goes into the house).

[The Natural Philosopher]

How hot did you run it?

[The Natural Philosopher]

Oh, he's insulating that..basic idea is slats under the joists, breathable membrane over that for draughtproofing, then rockwool, then boards. I was just suggesting that having done all that why not scrap the rads and use UFH.

His argument against was 'cost and commplexity' so looking for a cheap way..

[The Natural Philosopher]

Id like to hear a lot more about this Andrew..

IN partcualar what boiler temps you use and what materials the floor an ceilings are.

Ive got an upstairs room this might work in - 19mm chip over 6x3 timber with herringbone.

Currently carpeted, but laminate might be a better bet - its a work room

[Andrew Gabriel]

,

It's a bathroom at a friends. It was to be completely stripped and redone.

SHMBO decided they wanted rid of the radiator, and just wanted a chrome towel rail. I warned that on a north facing bathroom, this would leave it cold, but there was no way a radiator was going in this room. There had always been a bit of bathroom floor just in front of the basin which was always nice and warm, which turned out to be due to a radiator pipe being there. Having got the floor up, I decided to get a roll of 10mm copper tube, and I snaked a run up and down the length of the room between each floor joist, supported on periodic lengths of 1x1 I screwed between the joists. Stapled carpet to the 1x1's to prevent any creeking as the pipework expands. This was a small room and I split the 10mm runs into 2 circuits. A larger room would need more parallel runs. One end is brought up above the floor with a TRV head poking out the end of the bath panel.

Floor was creaky floor boards which I chucked out, and replaced with 18mm waterproof ply to take ceramic floor tiles. The resulting floor surface doesn't feel noticably warm, but it's not cold either, and it seems to keep the bathroom warm. The other important thing I did was to put in a new lower ceiling (original was rather high for a small room). In doing this, I put 50mm cellotex between the ceiling timbers (that's in addition to the original ceiling and insulation which also remains in place). This probably also helps with retaining the heat.

[Andrew Gabriel]

It just runs directly from the heating circuit. That house does have a condensing boiler now, but the rest of the system was designed for 80/70 and it often seems to be set higher than ideal for condensing operation.

Pipe is suspended a couple of inches under the floor.

[Richard Skeen]

Just a thought, but...

Isn't the reason for the normal thermostatic mixing valve etc. to limit the maximum temperature any part of the floor could reach?

I would have thought that partially insulating the pipe would reduce the amount of heat which was transferred to the floor, but not guarantee that the temperature of the floor at the least insulated part of the pipe wouldn't go higher than you wanted.

Limiting the supply temperature instead of insulating the pipe could mean that you could actually transfer more heat a cold the floor if the flow rate was sufficient, without risking overheating the floor.

Maybe bit like trying to supply a variable size electrical load from a higher voltage supply using only a series resistor to limit the voltage, as opposed to a voltage regulator.

Rich.

[The Natural Philosopher]

Yes, of course. Thats especially relevant in screed.

It probably is also relevant if you have a huge output and go walking in bare feet.

What I was wondering though,was when you have an airspace being heated rather than screed, how important that is.

Its a much more complex equation, because you are balancing heatloss from the pipe to teh underfloor airspace with heatloss from the floor top the room.

I already know from my screed floor, that if you cover the room with a rug and a sofa, under the sofa is very warm indeed. Heated seating ;-)

So this system offers no guarantees that things wont get uber hot under the floor, juts a fairly reasonble guess that the underfollr temps will be roughly constant. Andrews idea of running an up an down loop in the same joist space is pretty damned good, as that should balance the heat between the flow and return parts.,

Wel thats why its done of course. I wouldnt risk this in a screed floor,. but under a wood one..I dunno. It has distinct potential I think.

Exactly.

I mean everybody offers dire warnings as to what will happen if a wood floor gets too hot, and yet I have hot water pipes in studwork walls and under chipboard floors. I DO know that one particular wall that is so equipped shrank a bit faster when the house was built, yes. But its stable enough now.

[Cicero]

========================================= Since you're heating a closed air space you might save a lot of effort by fitting a few of these connected by microbore piping:

[The Natural Philosopher]

I love those, but they are too expensive and too high a heat output.

[Andrew Gabriel]

I found I took a couple of photos part-way through. Looks like I had just threaded through, soldered up, and was pressure testing all the new copper pipework at this point, before it was connected up to the heating system.

Also, looks like I did 3 runs of the tubing between each joist, except where joist spacing was less than normal.

I didn't take a picture when it was all finished, before the floor went down.

[Tim S]

Andrew Gabriel coughed up some electrons that declared:

That looks very cool Andrew. Given me an idea for my upstairs with the added advantage I could put the header tank lower.

Two questions:

a) did you make any special allowance for bleeding it, or do you find that the pump can push the air out OK?

b) What's the ambient temperature like in the void underneath - I'm thinking about the effect it will have on electric cables?

Cheers

Tim

[Doctor Drivel]

You will probably find the return temperature is too high reducing condensing efficiency.

[Doctor Drivel]

I know someone who did something similar. He used 8mm copper coiled pipe. He clipped it to the side of the joists, about 1/4 way down, all along the length of each joist in one continuous loop. He use two loops to even out the heat - hottest in the centre of the room. The pipe did not touch the joists, running though joists through 3/4" holes.

The control was a simple room stat and zone valve. The boiler was set to

65-70C. He did not lay the floor covering until one month after in case of floor shrinkage. It works OKish. The on-off control does not make matters better with temperature overshoots. The floor is warmish, but lots of insulation beneath the pipes in the floorspace.

His boiler is old and he plans to fit a condensing boiler I suggested a Broag dual temperature boiler using a 3-way valve DHW priority system, This, runs the boiler temp to maximum when heating DHW. He plans to put UFH in a similar way under all his suspended ground floor. I told him to put larger rads upstairs and run the lot on the weather compensator, with 55c max temp on the compensator slope.