Impact of levelling compound (epoxy?) on UFH

[Dan F]

Hi,

 

We are currently working out UFH zones/spacing etc. and I'm keen to ensure that we get things right to enable us to run at low temps and still cover our heating demand.  Given I like to understand/play with the details rather than just take someones word for it, I'm using LoopCAD to model the circuits and their required/actual floor surface temperatures to meet the heat load at different spacings and flow temperatures.  LoopCAD takes into account the depth of concrete, height of UFH pipes in concrete, u-value of insulation under concrete and the heatingd demand of each zone and even the type/size of UFH pipe.

 

This is an MBC passive slab so I'm using:

Concrete depth:150mm

Depth of UFH pipe: 50mm (?)

Pipe: 16mm Pex-al-pex

Insulation under concrete 0.1 u-value.

Heating demand: M&E guy has calculated 20W/m2 on average using standard heat-loss calcs down ground floor rooms only.  PHPP gives 15W/(ground floor m2) as heating load for the whole house.

 

What I'm not sure is what to use for r-value of flooring.  75% of the ground floor will almost certainly be targe format porclain tiles , and 25% engineered wooden flooring but our architect has allowed for 10mm self-levelling compound on top of the floated slab.  I can't find any data anywhere that tells be the r/rsi/tog value of products like this though, and concern if might be a bad (or worse) than a thick carpet depending on the product.  Maybe the 10mm levelling compound isn't justified, but I'm pretty sure we'll use some and might need to use the full 10mm if we want to ensure level-thresholds still work (currently our FFL is 35mm above slab).  Any idea?

 

I have a few other questions on UFH, but will save those for a sperate thread once I've got things modelled.  Be interesting to see how closely LoopCAD numbers match @Jeremy Harris table here.  From initial experimentation it looks spot on!  I still like playing with the paremters myself though...  Also interested to see if spacing can be increased to 300mm or more and still supply the heat load required with ASHP temp <=35C. (althought not sure what we'd loose by using 200mm instead of 300mm if it's the same cost ...I assume 200mm will mean more circuits to cover floor area and a larger manifold though).

 

.

 

Edited April 19, 2020 by Dan Feist

[Jeremy Harris]

Self levelling compound almost certainly has about the same thermal conductivity as concrete or similar materials, so around 1.2 W/m.K.  FWIW, we notice zero diference between the side of the house that has bamboo flooring, with some rugs over it, relative to the side of the house that has travertine tiles.  In theory the travertine should be a lot more thermally conductive than the bamboo, but in practice the floor surface temperature is always within about 0.1°C, no matter where it's measured.

 

I would personally not like to have pipes at 300mm centres, as I think the response time will really start to suffer.  We have ours at 200mm centres, which is OK, but I think the response time might have been better had they been at 150mm centres.

[Dan F]

Isn't epoxy a plastic and therefore signifcinalty more insulating?  I assume there are various non-epoxy self-levelling compounds also though.  Anyway i've used RSI of 0.1 for now..

 

On the spacing LoopCAD gives me the the following:

- Using 200mm, the required flow temp is 29.3C (DT of 4C) to achieve floor temperate of 23C and output of 22 W/m2.  (seems to match your experience)

- Using 150mm, flow temperature required drops to 29C (DT of 4C) for same output.

- Using 300mm flow temperature required goes up to 30C (DT of 4C) for same output.

- Using 450mm flow temperature required goes up to 32.6C (DT of 7C) for same output.
- Using 600mm flow temperature required goes up to 34.2C (DT of 7C) for same output.  

 

And peak output with 45C flow temperature:

150mm Peak = 56.2W/m2 (27C floor temp)

200mm Peak = 54.1W/m2 (26C floor temp)

300mm Peak = 51.5W/m2 

450mm Peak = 45.7W/m2 

600mm Peak = 40.4W/m2 (25C floor temp)

 

I'll almost certainly use the seemingly standard 200mm, but it's interesting that with such low heat load, in theory, even 600mm space with <35C flow temp would suffice.  This doesn't take into account response time though of course, also it's clear that if you went away on holiday for a month then the larger the spacing the longer it's going to take to heat the slab to 21C I assume...

 

3 hours ago, Jeremy Harris said:

I would personally not like to have pipes at 300mm centres, as I think the response time will really start to suffer.  We have ours at 200mm centres, which is OK, but I think the response time might have been better had they been at 150mm centres.

 

Given you have a passive standard MBC build, just wondering in what scenarios the response time of the slab is important to you?

 

 

[Jeremy Harris]

It's the fillers in levelling compounds that dominate in terms of thermal conductivity, I think.  Even the epoxy compounds are mainly filler.

 

The response time is key at this time of the year, where we can have sudden changes in outside temperature.  Recently we had a long period of warm weather, followed by a few cool days with cold nights (including frost).  It's handy to be able to inject a bit of heat after 48 hours or so of cool weather, as the house temperature starts to dip, but only for a short time.  We've played with using the warm air system to do this, and it certainly works relatively quickly, but it also heats the whole house (the MVHR isn't zoned) and so the bedroom tends to get warmer than we'd like.

[PeterW]

1 hour ago, Dan Feist said:

Isn't epoxy a plastic and therefore signifcinalty more insulating?  I assume there are various non-epoxy self-levelling compounds also though.  Anyway i've used RSI of 0.1 for now..

 

On the spacing LoopCAD gives me the the following:

- Using 200mm, the required flow temp is 29.3C (DT of 4C) to achieve floor temperate of 23C and output of 22 W/m2.  (seems to match your experience)

- Using 150mm, flow temperature required drops to 29C (DT of 4C) for same output.

- Using 300mm flow temperature required goes up to 30C (DT of 4C) for same output.

- Using 450mm flow temperature required goes up to 32.6C (DT of 7C) for same output.
- Using 600mm flow temperature required goes up to 34.2C (DT of 7C) for same output.  

 

And peak output with 45C flow temperature:

150mm Peak = 56.2W/m2 (27C floor temp)

200mm Peak = 54.1W/m2 (26C floor temp)

300mm Peak = 51.5W/m2 

450mm Peak = 45.7W/m2 

600mm Peak = 40.4W/m2 (25C floor temp)

 

I'll almost certainly use the seemingly standard 200mm, but it's interesting that with such low heat load, in theory, even 600mm space with <35C flow temp would suffice.  This doesn't take into account response time though of course, also it's clear that if you went away on holiday for a month then the larger the spacing the longer it's going to take to heat the slab to 21C I assume...

 

 

Given you have a passive standard MBC build, just wondering in what scenarios the response time of the slab is important to you?

 

 

Unless you’ve played with all the settings, Loopcad is very pessimistic when it comes to calculating heat loads. I would go with 200mm as you can always turn down the flow temp, turning it up you will stand the risk of overshooting your target temperature and end up with a very warm house. 

[Dan F]

28 minutes ago, PeterW said:

Unless you’ve played with all the settings, Loopcad is very pessimistic when it comes to calculating heat loads. I would go with 200mm as you can always turn down the flow temp, turning it up you will stand the risk of overshooting your target temperature and end up with a very warm house. 

 

I didn't let it calculate the heat loads, I put them in manually using 22W/m2.  This is the same value @Jeremy Harris mentioned in the other thread as a maximum from his experience, and also happens to be around 150% of what PHPP calculates as the heat load for the whole house.  I started out wanting to ensure 200mm was close enough spacing, but then explored the larger spacing just out of interest more than anything else.

 

What I don't understand about UFH (yet) though, is if there are any minimum flow rate constraints.  For example:

- Our smallest ground-floor room (a study) is approx 5.5m2.

- With 200mm spacing, flow temp of 33C and delta T of 7C LoopCAD estimates a the flow rate required is just 0.38l/min.

 

Is this kind of flow rate a Salus actuator could comfortably deal with, or does it have a minumum?

 

[Jeremy Harris]

The Salus actuators don't really care too much about flow rate, as they seem to be able to modulate over a pretty wide range.  I've found that our UFH works a great deal better with a lower flow temperature than 33°C though.  If I ran ours anywhere near that high it would be pretty uncontrollable.  We run ours at a flow temperature of about 25°C to 26°C, and have found that this gives an overshoot of maybe 0.5°C.  When we ran it at about 28°C the overshoot went up to about 1°C.