Underfloor heating tips

inkydolphin

Hi. I just moved into a bungalow with underfloor heating in the kitchen-diner powered by gas combi-boiler. The combi-boiler also runs the radiators (eight including two heated towel rails), one of which (albeit quite small) is in the kitchen-diner.

I was after some advice on general principles to use for a setup like this in terms of settings for the thermostat governing the underfloor heating (it works on air temperature not floor temperature) - i.e. should this be set to a lower temperature than the thermostat in the hall that controls the radiators, what flow-rate temperature to set for the underfloor heating water (presuming it's adjustable) and how long it should be on for and at what time of day to make it worth having.

I realise there will be a certain amount of trial and error with these things but any thoughts would be greatly appreciated.

QrizB

inkydolphin said:

Hi. I just moved into a bungalow with underfloor heating in the kitchen-diner powered by gas combi-boiler. The combi-boiler also runs the radiators (eight including two heated towel rails), one of which (albeit quite small) is in the kitchen-diner.

Congratulations on your new bungaow with gas-fired wet underfloor heating! We often get posts from people with electric underfloor heating and huge energy bills. You shouldn't have that problem.

inkydolphin said:

I was after some advice on general principles to use for a setup like this in terms of settings for the thermostat governing the underfloor heating (it works on air temperature not floor temperature) - i.e. should this be set to a lower temperature than the thermostat in the hall that controls the radiators, what flow-rate temperature to set for the underfloor heating water (presuming it's adjustable) and how long it should be on for and at what time of day to make it worth having.

I'd suggest you set the UFH thermostat to suit your needs in the kitchen-diner, and ignore the radiator in that room (or turn the valve off). Run the rest of the house like a "normal" system.
The UFH water temp is probably set by a blend valve, and might not be user-adjustable.

JSHarris

Worth noting that radiators are usually a lot more efficient than underfloor heating.  UFH always has higher heat losses from the underside, even with lots of insulation (and very often it's installed with woefully inadequate insulation).

We have wet ground floor underfloor heating, run from a heat pump.  I went OTT with insulation, so there is a 300mm thick layer of rigid foam underneath it.  This is a great deal more than most installers use, but even so our UFH loses about 10% of its heat input through the insulation to the colder ground beneath.

Radiators will rarely lose anything like as much heat to the outside world, especially if not positioned on outside walls (where they will lose a small amount of heat through the wall behind).

I was really shocked at how much heat was lost downwards through our thick layer of insulation, so I thoroughly checked the heat loss calculations, but they were spot on.  The problem is that the ground under a house will always be cool, typically between 8°C and 10°C, so the temperature differential between the UFH pipes and the ground is much greater than that from the room temperature to the outside air temperature.  It's this that drives the high losses.

BellaBlondykeTheThird

JSHarris said:

Worth noting that radiators are usually a lot more efficient than underfloor heating.  UFH always has higher heat losses from the underside, even with lots of insulation (and very often it's installed with woefully inadequate insulation).

We have wet ground floor underfloor heating, run from a heat pump.  I went OTT with insulation, so there is a 300mm thick layer of rigid foam underneath it.  This is a great deal more than most installers use, but even so our UFH loses about 10% of its heat input through the insulation to the colder ground beneath.

Radiators will rarely lose anything like as much heat to the outside world, especially if not positioned on outside walls (where they will lose a small amount of heat through the wall behind).

I was really shocked at how much heat was lost downwards through our thick layer of insulation, so I thoroughly checked the heat loss calculations, but they were spot on.  The problem is that the ground under a house will always be cool, typically between 8°C and 10°C, so the temperature differential between the UFH pipes and the ground is much greater than that from the room temperature to the outside air temperature.  It's this that drives the high losses.

Can you quote some studies on this as we have UFH and the bills went down significantly compared to radiators. In no way us it less efficient that radiators.

The selling points were one big radiator. Lower requested temperatures form the heat source, even heat and 25% more efficient which seems to bear out with our bills.

The concrete slab above and below becomes one big radiator, low and slow.

JSHarris

BellaBlondykeTheThird said:

JSHarris said:

Worth noting that radiators are usually a lot more efficient than underfloor heating.  UFH always has higher heat losses from the underside, even with lots of insulation (and very often it's installed with woefully inadequate insulation).

We have wet ground floor underfloor heating, run from a heat pump.  I went OTT with insulation, so there is a 300mm thick layer of rigid foam underneath it.  This is a great deal more than most installers use, but even so our UFH loses about 10% of its heat input through the insulation to the colder ground beneath.

Radiators will rarely lose anything like as much heat to the outside world, especially if not positioned on outside walls (where they will lose a small amount of heat through the wall behind).

I was really shocked at how much heat was lost downwards through our thick layer of insulation, so I thoroughly checked the heat loss calculations, but they were spot on.  The problem is that the ground under a house will always be cool, typically between 8°C and 10°C, so the temperature differential between the UFH pipes and the ground is much greater than that from the room temperature to the outside air temperature.  It's this that drives the high losses.

Can you quote some studies on this as we have UFH and the bills went down significantly compared to radiators. In no way us it less efficient that radiators.

The selling points were one big radiator. Lower requested temperatures form the heat source, even heat and 25% more efficient which seems to bear out with our bills.

The concrete slab above and below becomes one big radiator, low and slow.

It came up when I was putting together the energy assessment that formed part of the EPC process for the house when I was building it.  Surprised me so much that I spent ages re-checking the downward heat loss calaculation, but there was no error at all.  The calculation is exactly the same as that use to determine the heat lost through the walls, roof, windows, doors etc.  

The input variables are the same, the thermal conductivity of the composite structure (in my case a concrete floor with pipes embedded, sat on top of a membrane, then 300mm of EPS foam, then compacted stone that is on top of hard clay soil).  The killer is really the temperature differential.  For the walls, roof, windows etc the temperature differential is the room temperature minus the outside air temperature, for the heated floor it's the UFH pipe temperature minus the ground temperature.

The formula is straightforward, and is the one universally used to calculate heat loss, and is:

Heat loss (W) = Floor area (m²) * (UFH pipe temperature(K) - ground temperature(K)) * Floor composite U value (through section from pipe lower surface to underlying ground)

The U value can be worked out using the industry standard method, using the thermal conductivities of the various materials that make up the floor "sandwich" together with the thickness of each layer and the boundary effects (i.e. air on the inside surface, soil on the underside surface).  In my case the floor U value is pretty low, at 0.0975W/m².K  This is a passive house level of floor insulation, most UK homes will be a great deal worse than this, as it's rare to have more than around 100mm of underfloor insulation and exceedingly rare for homes to have anything close to the 300mm of foam insulation we have.

The actual heat loss figures through our floor are low in absolute terms, but high in relative terms, and it's the relative bit that impacts efficiency.  For example, our house typically needs about 550W of heat input to the air in the house (24/7) when it's 5°C outside and 21°C inside, so about 13.2kWh of heat input per day (this is pretty low because it's a passive house).  The heat loss through the floor is a pretty constant 99W, as the UFH temperature tends to be fairly constant.  The efficiency calculation for these conditions is easy, if I'm putting 550W into the house to keep it warm and losing 99W through the floor then the loss is 15.25% for those conditions.  When it's colder outside and the house needs more heating the proportion of the total lost through the floor is smaller (because the ground temperature doesn't really change), hence the 10% quoted as an example.

QrizB

JSHarris said:

The actual heat loss figures through our floor are low in absolute terms, but high in relative terms, and it's the relative bit that impacts efficiency.  For example, our house typically needs about 550W of heat input to the air in the house (24/7) when it's 5°C outside and 21°C inside, so about 13.2kWh of heat input per day (this is pretty low because it's a passive house).

If you had a more "normal" house, though, it might need 2kW of input under those conditions. Your 99W loss is then only 5%.

Also, getting back to the comparison with radiators, you'll still be losing heat through the floor (admittedly from a 21C room temp not a <whatever>C water flow temp) so the increase in losses due to UFH is going to be less than the total loss.

If your flow temp is 35C, then the delta-T to 10C soil is 25C. From a 21C room it's 11C. You'd expect heat losses from a radiator-heated room to be 11/25ths as much, so about 44W, all other things being equal. The increased loss due to UFH is 55W, 10% of the total heat demand in your case, 2.5% for my hypothetical normal house.

JSHarris

QrizB said:

JSHarris said:

The actual heat loss figures through our floor are low in absolute terms, but high in relative terms, and it's the relative bit that impacts efficiency.  For example, our house typically needs about 550W of heat input to the air in the house (24/7) when it's 5°C outside and 21°C inside, so about 13.2kWh of heat input per day (this is pretty low because it's a passive house).

If you had a more "normal" house, though, it might need 2kW of input under those conditions. Your 99W loss is then only 5%.

Also, getting back to the comparison with radiators, you'll still be losing heat through the floor (admittedly from a 21C room temp not a <whatever>C water flow temp) so the increase in losses due to UFH is going to be less than the total loss.

If your flow temp is 35C, then the delta-T to 10C soil is 25C. From a 21C room it's 11C. You'd expect heat losses from a radiator-heated room to be 11/25ths as much, so about 44W, all other things being equal. The increased loss due to UFH is 55W, 10% of the total heat demand in your case, 2.5% for my hypothetical normal house.

Exactly right, it's the relative numbers that matter in terms of efficiency, not the absolute numbers.  Nevertheless, all other things being equal, UFH will always increase the heat loss rate through the floor when compared to radiators, just because of that increase in temperature differential.  Might only be a small increase, but the principle holds true, simply because the floor temperature is being raised and that's in thermal contact (albeit through some insulation) with the ground.  Radiators are only in poor thermal contact with outside walls usually, because the the air gap behind them and the effect of convective air movement moving heat upwards through that gap.

Reed_Richards

There are actually two types of underfloor heating:

1. The UFH heats a large slab of concrete (on top of some insulation).  This acts a bit like a giant night storage heater and it is very slow to both heat up and cool down.  I imagine this is what you would have in a new build.
2. The UFH is embedded in a shallow layer of screed on top of some insulation.  This type would respond faster  but is more likely to be found in a retrofit where the insulation underneath might not be so thick, so as to avoid reducing the ceiling height too much.

I think the right answer to the question depends on which type it is.

inkydolphin

Reed_Richards said:

There are actually two types of underfloor heating:

1. The UFH heats a large slab of concrete (on top of some insulation).  This acts a bit like a giant night storage heater and it is very slow to both heat up and cool down.  I imagine this is what you would have in a new build.
2. The UFH is embedded in a shallow layer of screed on top of some insulation.  This type would respond faster  but is more likely to be found in a retrofit where the insulation underneath might not be so thick, so as to avoid reducing the ceiling height too much.

I think the right answer to the question depends on which type it is.

@Reed_Richards Hmmm not sure on that one - sort of half and half - it's in a ten year old extension...

DeeQS

Reed_Richards said:

There are actually two types of underfloor heating:

1. The UFH heats a large slab of concrete (on top of some insulation).  This acts a bit like a giant night storage heater and it is very slow to both heat up and cool down.  I imagine this is what you would have in a new build.
2. The UFH is embedded in a shallow layer of screed on top of some insulation.  This type would respond faster  but is more likely to be found in a retrofit where the insulation underneath might not be so thick, so as to avoid reducing the ceiling height too much.

I think the right answer to the question depends on which type it is.

We have a new build that’s EPC B, and I think from the building spec it’s number 1 above. It took some getting used to, no instant heat for example. Now we’ve gotten used to it, it does act as you say like a giant storage heater.

The heating runs for a few hours in the morning to get the rooms to 18/19c (each room is individually controlled) and then it stays this temp for 12+ hours.

JSHarris

Our UFH has the pipes embedded in reinforced concrete, with a mass of about 15 tonnes.  Takes forever and a day to heat up when we first turn the heating on (usually around the beginning of November), but stays warm for a very long time.  We use it as a storage heater, heating it up with the heat pump during the Economy 7 off-peak period, with the floor then releasing heat to the house over a long period.

This works pretty well, but can get caught out if there's a sudden increase in outside temperature, as happened last night.  At 10 o'clock last night the outside temperature was down at 2.5°C and stayed low until the early hours (so the heat pump upped the flow temperature to compensate).  Right now the outside temperature has risen to 9.8°C, so the floor has a bit too much stored heat and the room temperature is about 0.7°C above the thermostat setting of 21.5°C.

We have a similar slight issue with the house getting a bit warmer than it should if we get an unexpectedly sunny day in winter, because there's no way to stop the warm floor from giving out heat (although the heat output does significantly reduce as the room temperature increases - Newton's law of cooling at work).  Not a major issue, we aim for a room temperature of 21.5° and it sometimes gets as high as 22.5°C when something like this happens, so not too high.