What ho one and all,

This is a question that I am trying to get my head around, partly because a friend is building a SIP house and has sought my advice/experience on UFH.

My house is t/f with the UFH on the g/f set in screed and the f/f set into styrene insulation sheets with ali spreader plates. Laid over the pipes, is a thin UFH underlay and t&g CaberBoard with vinyl covering in the bathrooms and the bedrooms a thin UFH underlay and engineered ‘click together’ timber flooring laid directly over the spreader plates.

It works well but clicks and clacks as it heats up caused by the ali expanding faster than the pipe does. It is VERY, VERY annoying.

A neighbour with a SIP new build, has UFH on both g/f and f/f. They put a reflective material onto the fllor fist, then 200mm battens and fixed the pipes in and around the battens, effectively creating a 200mm ‘service’ void for the UFH pipework. Over this is a 22mm OSB and then solid timber floor glued down (g/) and carpet on the f/f. They have no clicking and clacking as the UFH heats up and they say it is warm.

UFH works best when there is a large thermal mass (screed) and therefore less well on t/f construction with spreader plates. My f/f heats quickly but also cools quickly, unlike the g/f screed which retains the heat evenly, but he house is never cold since the Warmcell insulation and 3g retains the heat in.

So to my question: Given that heat cannot be destroyed, and is contained within the insulated shell of the house, which is the better f/f option, the ‘service’ void or spreader plate installation methods? What mdf routed board (thermoboard?)

Thanks

Toodle pip

Rex

Tony - Why would the floor voids be draughty places in an airtight house?

Rex - Is the problem with the ali plates using it as a relatively 'high temperature' system? Run constantly at a low(er) temperature you'd not be trying to alter the temperature of the floor and causing the expansion/contraction?

Thanks for the replies. The house is more than adequately insulated and although a 'standard' t/f house, I was very anal regarding air tightness during the build. Only had to hit a target figure of 9 (whatevers) but when tested, was 1.7, so I'm pretty happy. The floor is obviously within the insulated space and it is also fully insulated with Warmcell.

My flow temp (when running) is only about 38C but because I have weather comp and a secondary pump that is controlled by the boiler (plumber advised that the boiler internal pump was not adequate to pump the volume of water through X kilometres of pipe; seems reasonable) for the weather comp to work, it seems that the boiler pump has to be running all the time. If that is happening, then the secondary pump is also running, even though the wc says that heat is not necessary. So although the boiler blurb tells me how much gas I am saving, it does not tell me that I am spending that saving on electricity to run pumps that do not need to be run. Run constantly is the answer but that is not the financially most economic option.

The way around it is to have the heating on for two periods per day. Because half the house is screed, the 'at rest' flow temp is usually around 32, but the f/f floor temp will be slightly below that. When the heating comes on, there is sufficient temp. differential to cause the ali to expand. It is annoying but not a major issue unless the start-up time is 05:00, when the clicking floor wake us.

But I was wondering if there is any real difference between insulated f/f and 'void' f/f UFH when the house is well insulated? The retailers tell me there is, but the heat has to go somewhere and it not being lost to the outside, so presumably , the retailers are wrong? Or am I missing something?

Posted By: RexThey put a reflective material onto the fllor fist, then 200mm battens and fixed the pipes in and around the battens, effectively creating a 200mm ‘service’ void for the UFH pipework.

Why such a large void? I'd have thought 50 mm or even less would be plenty. Or are the “battens” actually the joists?

Ed,

My mistake, should have read 20mm!

Tony, no, the copper flow and return are not the pipes clicking. When we are in any of the f/f rooms, and most frequently the bedroom, you can hear the floor popping all over the place. It cannot be the timber expanding because if that was the case, surely I would hear in on the g/f? And I don't.

I'm 100% certain that it is the ali expanding and as the UFH pipework has a 'sort of' rubberised feel to its outer surface, and the pipework is a press fit into the ali for good heat conduction, it is the ali expanding against the pipework.

Have met another couple locally who built a t/f house and they have the same issue on the f/f with lai spreader plates.

Simon, in all honesty, probably not, but there are two pumps, one in the boiler and the secondary external. It's more the principle; if the weather comp says that heat is not needed, and the boiler is not actually fired up, why have the pumps running to circulate water at 20C.

I have spoken at length with Viesmann about this, and their answer is, that is the way it is! Seems to me that with all the electronics and sensor monitoring, it should be possible to not run the pumps if it is not necessary.

As I understand it, the boiler uses data from the external temp sensor, flow and return temps and the target room temp, set on the boiler control panel. It does not use room or floor stats. It works well; not too much control over individual room temps, but so what, the house is cosy. But if the various parameters tells the boiler that it does not need to fire up, why should the pimps continue to run? I would have thought there could be a line in the program that runs the pumps every so often to sample the flow/return temps.

But all this is getting away from my original question. Given that f/f UFH not set into screed is less efficient, is there a heat loss in a well insulated house? OK, the floor may not feel as warm to the touch, but surely the heat is heating the environment?

Rex

Posted By: Rex: “I'm 100% certain that it is the ali expanding and as the UFH pipework has a 'sort of' rubberised feel to its outer surface, and the pipework is a press fit into the ali for good heat conduction, it is the ali expanding against the pipework.”

Not that it makes much difference but wouldn't it be the other way round?

Al coefficient of linear thermal expansion: 22.2×10⁻⁶/K

http://www.engineeringtoolbox.com/linear-expansion-coefficients-d_95.html

PEX coefficient of linear thermal expansion: 90×10⁻⁶/°F = 162×10⁻⁶/K

http://www.engineeringtoolbox.com/physical-properties-thermoplastics-d_808.html

So, for a given temperature rise the PEX will expand more than 7 times more than the aluminium.

Thanks for the comments. I do agree that the heat must be within the house so is being useful. As for whether the PEX or ali is creating the clicking, somewhat academic. I'm sure you are correct with the formulas, but in my opinion, it is still coming from the PEX / ali contact.

If PEX expands so much, what actually happens to it within a screed where there is nowhere for it to go?