UFH Plan

[marshalp]

Hello,

 

Currenly doing an extension and renovation of a 1930s semi detached house in South Manchester. I’m not a plumber or tradesman and have come up with the attached plan by research, dangerous I know!

 

5 Zones UFH 

• Living/dining room
• Kitchen
• Utility
• Sitting room
• Hall

 

The majority of the house is suspended wooden floors with a void approx 500mm, the extension which includes kitchen, utility, part of the hall and dining area is concrete.

 

The concrete areas have 100mm of kingspan embedded within, I was planning on fitting 100mm kingspan between the joists below the floorboards, the 250mm kingspan over the top of everything, 16mm UFH pipe then wet screed approx 35mm.

 

The majority of pipe is laid at 200mm centres, all zones will be controlled using Heatmiser NeoStats. In the living and sitting room we will have gas fires. Upstairs will be radiators in most rooms, I was planning on using a manifold to create separate zones for these rooms and installling electric UFH in the bathrooms, don’t have the height to run wet UFH.

 

I’ve attached my pipe layout plan, your thoughts would be appreciated.

[PeterW]

Download loopcad (free for 30 days) and model the rooms and the loops. You won’t get them tight enough to do some of those loops. 

 

Also, 200mm is very wide and 35mm is very thin for a standard build house. You would be better with 150 or 100 and run at much lower temperature. 

[epsilonGreedy]

1 hour ago, PeterW said:

Also, 200mm is very wide and 35mm is very thin for a standard build house. You would be better with 150 or 100 and run at much lower temperature. 

 

 

Is this advice specific to the OP's thin screed layer or UFH design in general?

[PeterW]

Near passive with low heat input requirements can cope with 200mm centres - houses such as @JSHarris and @TerryE for example. Lower insulation values and higher losses mean more heat required and as the input is w/m2 then basically you have to up the wattage - this is either done by making there be more pipe per square metre or changing the delta between the flow temp and room temperature. The downside of this is once you have set the pipe centres the flow temperature is the only thing you can control so to get more heat out you have to increase the temperature. 

 

35mm screed is very thin, especially on a joist based floor and is likely to crack so I would want to go thicker and lose the 25mm of Kingspan above the floor. 

[marshalp]

Based on the plans above the main living space loop is 108m, if I reduce the pipe spacing I would need to split the loop into two.

 

I appreciate the screed depth is thin, what screed depth would be recommended? I plan to use castellated panels to run the pipe.

[Nickfromwales]

Are the areas that are suspended going to be concreted? Or spreader plates?

[marshalp]

22 minutes ago, Nickfromwales said:

Are the areas that are suspended going to be concreted? Or spreader plates?

 

Neither, I was planning on laying the UFH and screed above.

[Nickfromwales]

Ah, so little / no insulation either  ?

[PeterW]

C’mon @Nickfromwales read the whole thread ...?

 

100mm between joists, 25mm over and then 35mm screed

 

lots of insulation but very little strength in the screed

[epsilonGreedy]

4 hours ago, PeterW said:

Near passive with low heat input requirements can cope with 200mm centres - houses such as @JSHarris and @TerryE for example. Lower insulation values and higher losses mean more heat required and as the input is w/m2 then basically you have to up the wattage - this is either done by making there be more pipe per square metre or changing the delta between the flow temp and room temperature.

 

 

Oh interesting, I had been thinking the main challenge of UFH was persuading the concrete slab to shed enough watts into the air above through whatever floor finishings were fitted. Now I appreciate that charging up the slab temperature through the available pipe surface area requires careful thinking.

[Jeremy Harris]

Concrete is a reasonably good conductor of heat, close to that of water for a typical medium density concrete.  Here's a quick example of the thermal conductivity for a few materials:

 

Water - thermal conductivity ~ 0.6 W/m.k

 

Concrete ~ 0.4 to 1.2 W/m.k (denser concrete has the higher thermal conductivity)

 

Softwood ~ 0.12 W/m.k

 

Hardwood ~ 0.16 W/m.k

 

[marshalp]

What about anhydrite screed, at 45mm? 

[Nickfromwales]

2 hours ago, PeterW said:

C’mon @Nickfromwales read the whole thread ...?

 

100mm between joists, 25mm over and then 35mm screed

 

lots of insulation but very little strength in the screed

 

Read the thread !? I'm in the office today ?

 

 

Just skimmed the op, so my bad. 

Ill add something of value later 

[CC45]

Neat bend on that pipe @Nickfromwales - how did you do that?  Bit of heat?

[Nickfromwales]

2 hours ago, CC45 said:

Neat bend on that pipe @Nickfromwales - how did you do that?  Bit of heat?

Bingo bango

[CC45]

Neat. Thanks @Nickfromwales

[jack]

Bear in mind that those of us with 200mm centres have our pipes buried under at least 40mm of concrete. That will tend to provide more even heat than the same spacing under ~20mm, which is all you'll have above yours with 16mm pipe at the bottom of 35mm screed.

[TerryE]

In my, Jack's and Jeremy's slabs we have 100mm of concrete directly poured in a single pour slab with no covering screed (as the pour itself is power floated flat).  The UFH piping is embedded in it.  The slab also has 200 mm box cross-bracing ribs and 300ml (in my case) ring beams and centre beams. The concrete itself acts as a thermal store and is a key part of ensuring the thermal stability of the house.  My peak input into 70m² of slab is 3kW, and I rely on the mass of the slab to smooth the transfer into the environment.  I discuss this in more detail in my blog.