Wooden Floor thickness UFH and heating costs

[Foldyard]

I have a wet underfloor heating system embedded in a screed.  On top of the screed I have a plywood base and oak floorboards.  The oak floorboards were kiln dried to 6% moisture to avoid movement.

 

My heating is provided via a ground source heat pump.  My heating bills have been much higher than those predicted when the gshp was installed.  The installer says that this is probably due to the thickness of the wooden floor covering, rather than any problem with the gshp system.

 

My take on it is that the thickness of the wooden flooring shouldn't have any (or more than minimal) effect on energy usage.  It just means that the floor takes longer to heat up when first turned on.  The amount of heat energy used is going to be about the same.  SInce I have the heating on all the time (in winter) I don't see how the thickness of the wooden floor should make any difference.

 

Anyone?

[Andehh]

How much insulation do you have underneath? How thick is plywood and floor boards? Wood is an insulator technically, my parents had hardie floor at 22mm thick and 14mm thick floor boards and it did take a very high flow rate to really get the rooms warming up at a decent base. Then again, they only had 75mm insulation underneath....

Edited November 18, 2023 by Andehh

[Foldyard]

I don't think insulation under the screed is an issue.  All materials on top of the screed will insulate to some degree, but that's not the point.  If it is the same material only thicker it just means that it will take longer to heat up from a cold start.  Once the temperature at the surface reaches the appropriate temperature, the energy usage should be the same whatever the thickness.

[JohnMo]

We have a mix of oak floor, tiles and carpet. The carpet is very insulating, so much so, that very little heat escapes to the room. To make it work you need to run higher temperatures. But they are bedrooms so we don't bother.

 

Anyway back to your question, to get past the insulating effect are you running a higher temperature than you expected? If so you will be getting lower CoP and have higher running costs. Having a 4 port buffer with a good amount of mixing going on, will make make matter worse.

 

So need more info really on the full system your running temp etc

[Foldyard]

Well the carpet covering is not the same material as the tile or wood floor underneath.  If it has a greater thermal resistance than the tile or wood then it might require a higher running temperature in the UFH to achieve the necessary surface temperature at the carpet/air interface, in which case there will be greater energy usage.  Alternatively it may not be possible, as a practical matter, to achieve an appropriate temperature if the thermal resistance of the carpet is sufficiently high. 

[JohnMo]

16 minutes ago, Foldyard said:

Well the carpet covering is not the same material as the tile or wood floor underneath.  If it has a greater thermal resistance than the tile or wood then it might require a higher running temperature in the UFH to achieve the necessary surface temperature at the carpet/air interface, in which case there will be greater energy usage.  Alternatively it may not be possible, as a practical matter, to achieve an appropriate temperature if the thermal resistance of the carpet is sufficiently high. 

 

My carpet isn't the issue, it was just an example.

 

1 hour ago, Foldyard said:

If it is the same material only thicker it just means that it will take longer to heat up from a cold start

No doesn't happen in practice.

 

The real answer to the transfer process is not a simple one. So will work through what happens.

 

Water flows through a pipe at temperature X. The temperature transfers to the screed or concrete in a uniform manner, up, down and sideways until it meets a thermal resistance. The resistance could be heat flow from other parts of the pipe, insulation or the floor surface.  The further the heat has travelled away from the source (the pipe wall) the lower it's temperature as the heat energy spreads outwards from a small pipe. A water pipe section will interact with the other parts of the UFH array.

 

So for well insulated modern house. 

1. For a given flow temperature, say 30 degrees, with dT of 6, mean flow temp is 27 degrees. This is around 20W/m2 output to the floor.  Floor will have a surface temperature between and 3 and 4 degrees warmer than the target room temperature. If you have insulation above the floor surface the heat transfer to the room is restricted. Move to point 2.

 

2. A heat pump will modulate it's output to keep return temperature stable, but also to manage a fixed dT and maximum preset flow temp.  Once at maximum flow temperature the heat pump will modulate output to its lowest point for a given flow rate, then dT reduces between flow and return, over a period of time (minutes or hours) below its minimum preset point, the heat pump stops adding heat (it's compressor shutdown). Move to point 3.

 

3. For a given heat flow into the floor (product of flow rate and dT). The more restrictive the surface finish is, the more difficult the heat transfer mechanism. So instead of transferring the heat from the pipe, to the floor, to the room (heat flow is restricted by floor covering), the return flow temperature stays slightly warmer, decreasing dT quicker than normal and shutting down the heat compressor sooner. Less heat is actually transferred to room.

 

To compensate you either increase flow temperature or you accept a cooler room, as in the case for our bedrooms.

 

Add a 4 port buffer, you could easily be flowing 5 or more degrees hotter than you thought you needed.

 

Also if you are running zones and on/off timings, these all add to a rubbish CoP.

[Andehh]

It's heat sandwiched between two layers of insulators. Wood on top, insulation before....the more you 'insulate' the bottom the more heat is retained for it to force its way through the wood... If you have (for example) 10mm insulation below the screed and 25mm of wood above more heat will be dissipated below the 10mm then above the wood.

 

On the flip side if you have 200mm insulation below screed then your post is correct, wood will slow it down but temperature will eventually find its way through (though return temps will remain higher then if the heat could escape quicker....)

[JohnMo]

1 minute ago, Andehh said:

On the flip side if you have 200mm insulation below screed then your post is correct, wood will slow it down but temperature will eventually find its way through (though return temps will remain higher then if the heat could escape quicker....

No it won't, it doesn't work like that. I thought did until I tried it, with my summer house - the heat pump works against you, not for you. The only way it works like that is by turning up the flow temperature or running a way lower dT (moving the mean flow temperature up)

[Andehh]

Oh really? Because you end up with a really small delta T and ASHP struggles to compensate?

[Foldyard]

I appreciate everyone's comments.  I understand how various factors influence the efficiency of the system, but for the moment I am focussed on floor covering thickness.  I am also thinking about it in abstract terms, which may not accord with experience.

 

JohnMo said:

 

"3. For a given heat flow into the floor (product of flow rate and dT). The more restrictive the surface finish is, the more difficult the heat transfer mechanism. So instead of transferring the heat from the pipe, to the floor, to the room (heat flow is restricted by floor covering), the return flow temperature stays slightly warmer, decreasing dT quicker than normal and shutting down the heat compressor sooner. Less heat is actually transferred to room."

 

I get the bit aout "the more restrictive the surface finish it etc..."  ie if the thermal resistance of the material is greater it may require more energy (and a higher temperature of the water in the pipes) to get the surface temperature to the 2 or 3 degrees above room temperature than a material with a lower thermal resistance.  What I am struggling with is where you have a material of a given thermal resistance and them make it thicker.  The material has the same thermal resistance and so heat is going to flow though it the same, whatever the thickness.  You have an extra amount of material to heat up initially so it will take longer, but once it reaches an equilibrium temperature, I would have thought that the temperature of the water in the pipes would be more or less the same.  I am probably wrong but I would like to understand why.

[Conor]

How much insualtion do you have under the screed and what flow temps using? Everything else is incidental.

[JohnMo]

43 minutes ago, Foldyard said:

material of a given thermal resistance and them make it thicker.  The material has the same thermal resistance and so heat is going to flow though it the same, whatever the thickness

Again no, the same material that is thicker has an increased thermal resistance.

 

A material has thermal conductivity defined by W/mK,

 

The thermal resistance or R-value is calculated by using the formula

 

R-Value = l/λ

 

Where:

 

l is the thickness of the material in metres and

 

λ is the thermal conductivity in W/mK.

takes into account the actual material thickness, the

 

'U' value is just the inverse of the 'R' value 

 

So thickness has a big impact. Double the thickness half the heat transfer. Otherwise you just insulate your house with 1mm insulation and all would be good.

[Foldyard]

Yes but it only slows down the heat transfer.  If the heating stays on constantly, then over time the extra layer of wood will heat up to the same temperature as the underlying layer, more or less, given that the two layers have the same conductivity and heat capacity.  Or am I wrong?  With your insulation example it takes longer for the heat to get through the extra layers so heat is being lost at a slower rate than the heat being added to the interior of the house.

Edited November 18, 2023 by Foldyard

[Mike]

3 hours ago, Foldyard said:

What I am struggling with is where you have a material of a given thermal resistance and them make it thicker.  The material has the same thermal resistance and so heat is going to flow though it the same, whatever the thickness.  You have an extra amount of material to heat up initially so it will take longer, but once it reaches an equilibrium temperature, I would have thought that the temperature of the water in the pipes would be more or less the same.  I am probably wrong but I would like to understand why.

OK, here's my attempt at a simple explanation:

 

Temperature is a measure of the average speed of vibration (kinetic energy) of the atoms within a material. Heat is transferred by atoms / molecules bouncing off each other. However there is no guarantee that any one atom / molecule will encounter another - if it doesn't no heat will be transmitted between them.

 

The amount of heat transmitted will depend on the density of the material (the denser it is, the more the atoms / molecules are packed together, so the higher the probability is that one atom / molecule will encounter another to bounce off) and the thickness (as thickness increases, the heat has to be transferred through an increasingly long chain of atom / molecule bounces).

 

To turn that into numbers, if the probability of one atom bouncing off another is 90%, the probability of that second atom bouncing off a third is 90% of 90% = 81%, and the probability of the third bouncing off a fourth is 81% x 90% = 72.9%; as the number of bumps increases the heat transferred goes down.

 

[JohnMo]

2 hours ago, Foldyard said:

If the heating stays on constantly,

There is a difference between heating being on continuously and the heat pump giving a heat rate at a constant rate. As detailed in point 2. above. 

 

But still no detail on your system specifics or how you operate it? I will await details before responding again.

[Foldyard]

I found this bit of info which probably answers my question as regards whether a thicker wooden floor should cause a large increase in fuel bills:

 

"The flow temperature depends on the required output and the chosen flooring. Normally, the flow temperature is 30 to 45 °C, and it should not exceed 50 °C. The flow temperature has a limited impact on the heat energy consumption. If the temperature is raised from 30 to 45 °C, heat energy consumption only increases by 6%, as energy consumption is determined by the difference between the flow and return temperature. It is only marginally more expensive to heat a thick plank floor than a thin engineered wood floor, but it takes a higher flow temperature to ensure the required surface temperature."

[JohnMo]

25 minutes ago, Foldyard said:

If the temperature is raised from 30 to 45 °C, heat energy consumption only increases by 6%

That may be true for a gas boiler, which has an almost direct relationship between consumption and power output, but not a heat pump. For my ASHP (similar would be true of a GSHP) - for the same weather conditions electric consumption increases by 25%, not 6% if you increase flow temp from 30 to 45.

[SteamyTea]

On 18/11/2023 at 11:55, Foldyard said:

  If it is the same material only thicker it just means that it will take longer to heat up from a cold start

Almost right.

Depending on the thermal inertia. The room air temperature may be in equalibrium, but that could be because energy is leaking out via thermal paths that are varying in temperature more rapidly than you think. The heating system may be trying to compensate for this by increasing flow, or temperature, or both.

This is more pronounced on a heat pump than a gas or oil boiler.

They work differently. A gas or oil boiler is a fixed temperature device, a heat pump, by the nature that the 'fuel' i.e. the ground or air, varies in temperature, causes the power to drop if you try to keep the same flow rate or temperature.

 

[SteamyTea]

20 hours ago, Foldyard said:

flow temperature is 30 to 45 °C, and it should not exceed 50 °C. The flow temperature has a limited impact on the heat energy consumption

I am in a cafe at the moment but that does not seem right.

You can take a stab at the sums and see what f it makes sense.

 

If you assume that the heat transfer medium is water (4.2 kJ/kg.K) and take a small packet of water, say 1 kg, to keep the sums easy.

Also assume that the return temperature is the same, let's say 27⁰C.

 

Then.

 

@Flow 30⁰

4.2 x (30 - 27) = 12.6 kJ

 

@Flow 40⁰

4.2 x (40 - 27) = 54.6 kJ

 

@Flow 50⁰

4.2 x (50 - 27) = 96.6 kJ

 

Now that is a very basic, and almost certainly a very extreme calculation as the return temperature is fixed, but say @40⁰ the return is at 33⁰ then:

 

4.2 x (40 - 33) = 29.4 kJ

An energy difference of 16.8 kJ. A 43% increase.

 

Without some real numbers the above is speculation.

So it is well worth getting some real temperatures and times the system is actually generating energy.