Under Floor Heating

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

RogerM

Established Member
Joined
4 Mar 2006
Messages
1,279
Reaction score
35
Location
Devon
Whilst not really woodwork, most of us do work that is aimed at improving our homes, and I thought that my venture into under floor heating may be of interest. This is NOT a definitive “how to” thread. I am not a plumber, but never-the-less the end result is one that I am pleased with.

The building work on my kitchen extension is more or less complete, so time to think about heating. We have decided on an underfloor heating wet system using our existing oil fired condensing boiler. The current system is divided into 3 separately controlled zones - downstairs, upstairs and hot water. The intention is to make the kitchen/living room extension a 4th separate zone as the levels of insulation are exceptionally high and we more or less expect to live out there in the winter, and the summer too with the bifold doors open.

Having just been charged £180 to re-route 2 pipes which had been clipped to the wall that has just been demolished, and which were intended to go under the floor, and then find after the plumber had gone that they were still going to exit the wall above floor level, I didn’t even bother to get a quote. Time to DIY.

After lots of reading around I decided to go for the Polypipe system, using their single room zone valve, and a wireless programmable thermostat.
First job is to insulate the floor with 100mm of Celotex. This is simply laid on the floor slab and the joints taped over to stop it moving about before the screed is laid.

P1040385.JPG


The walls are also insulated with 65mm of Celotex in the cavity, and a further 30mm behind dot’n dab Thermal Check K plasterboard. The floor screed will need room for thermal expansion and contraction, so a thin strip of Celotex 25mm thick is used to isolate the screed from the wall thermally. This is just a friction fit between the Thermal check plasterboard and the under floor Celotex. The loose cable on the floor will provide power to the island.

P1040502.JPG


With a 25mm strip of Celotex all round, a final layer of proprietary Polypipe expansion material is attached to the top of the underfloor Celotex with double sided tape and this bends through 90 degrees up the wall, and in front of door frames etc. It’s also important to use this where the new screed meets the old screed in the existing building so that the new screed is completely isolated to allow for expansion on all sides.

P1040506.JPG


A rigid plastic barrier is laid over the Celotex, which is castellated a bit like a giant egg box, and the 15mm underfloor pipe will be held in place by the castellations. This is attached around the edges by the double sided tape on the expansion strip.

P1040507.jpg


The cable for the island is contained within a conduit and is embedded in the Celatex under the rigid castellated sheet.

P1040505.JPG


The pipe comes in 100m rolls and we’ll be laying 240m of it, so we need 3 rolls. As there must be no joins within the floor screed, this means that all joins must be above floor level, which means laying 3 circuits and then joining them via a simple manifold once they are above the floor, and ideally near the boiler. Here’s the first circuit laid.

P1040509.JPG


..... and the second.

P1040510.JPG


.... and the third. I’ve put a removable wooden plug around the island cable so that there is an appropriately sized square hole to take a box for a floor socket. I don’t want to have to drill anywhere near the pipes! The 3 separate circuits exit the kitchen through conduits in the far wall. The 600mm space down the side and far end will be covered by kitchen units so i did’t want to heat the floor beneath them. The pipes exit the kitchen through the far wall in the 2nd photo .....

P1040520.JPG


P1040521.JPG


... come out in the utility room ...

P1040517.JPG


.. exiting the utility room under the sink. You can see the thermal expansion material crossing the floor between the old floor and the new area to be screeded, ensuring that the heating pipe is protected by conduit as it passes from one to the other. Having exited the utility room under the sink ...

P1040518.JPG


..... the pipes cross the floor of the study in conduits. Why have I lifted twice the width of old screed on the study floor as I appear to need? Well, a clear run 600mm wide was needed, and the cynical may say that having measured where the side of the excavation had to be the night before, I then excavated 600mm on the wrong side of the line. I couldn’t possibly comment! Anyway - here it is with battens screwed to the floor slab ready to be covered with 18mm floor grade chipboard. I decided against laying a concrete screed as it would take weeks to dry sufficiently to relay the carpet and I wanted it back down asap.

P1040511.jpg


.... and finally the pipes exit the wall in the garage just above the dpc, again protected in conduit. There on the right are the carefully drilled holes for the pipes in the wrong position! No harm done. They can easily be filled in. You’ll notice that I’ve taped over the ends of all pipes to ensure that no grit gets inside them whilst they are snaking about during laying. That would be the kiss of death to the pump.

P1040515.JPG


Next job is to join the 3 flows and 3 returns into a single flow and return by constructing a simple manifold for each using elbows, spigots and T’s. Whilst not strictly necessary, I’ve put an inline stop in each circuit so that in the event of future problems I can isolate each part of the circuit independently and it will also help to balance the flow between the 3 circuits in the event that they are not even.

P1050228.JPG


Next job was to check for leaks before the screed gets laid over the top and its embedded in concrete forever. No photos of that but I simply connected a domestic hose to the flow and a spare inline stop to the return and filled each circuit in turn. Once the flow from the hose was steady I closed the inline stop in the return whilst leaving the hose switched on and checked for leaks along the entire length of the pipework. As there are no joins other than those you see in the photo of the manifolds I wasn’t expecting any, but just wanted to ensure there were no manufacturing errors.
Time to lay the screed and cover up all that work for ever.

P1040523.JPG


P1050233.JPG


Now to connect the various components to the existing system. The system was drained and a 22mm T was connected to the boiler flow and return. The UFH relies on the existing CH pump so the connection to the flow MUST be made after the pump and NOT before. This involved moving the pump slightly to make room, whilst ensuring that the pump was not at the lowest part of the system where it may be pumping sludge. The two T’s are clearly shown in this photo. I wanted to use soldered joints for each, but an annoying steady drip down the return pipe made a good joint impossible to achieve so in the end I gave up and used a compression fitting. Again i have put in a pair of inline stops so that the entire UFH system can be isolated, and it also enabled me to refill and commission the existing system as soon as the T connections had been made.

P1040525.JPG


Next job is to fit the UFH zone valve, mixer/pump unit, and connect to the underfloor pipework. A quick explanation of principles may be helpful here.

The UFH system is controlled from a programmable thermostat. When the stat calls for heat, and the inbuilt timer says that heat is required at this time, the zone valve opens and the boiler and existing CH pump fire up. The CH pump sends water through the flow from the boiler to the mixer valve and back down the return to the boiler. When the flow through the mixer valve reaches 45 degrees, the UFH pump starts and pumps water through the UFH flow and UFH return.

Whilst the temperature of the water in a radiator system may be around 65 degrees, the temperature for the under floor water must be between 45 and 50 degrees or else your floors will get too hot. The temperature for the underfloor water is set with the red knob on top of the mixer unit, and when it reaches 50 degrees it starts to mix cool water from the UFH return to the UFH flow so that the flow is restricted to 50 degrees even though the water from the boiler may be at 65 degrees. Once the return water is up to 50 degrees the boiler shuts down but if the thermostat is still calling for heat the pumps keep circulating the water until it has cooled and the boiler cuts in again.

A photo of the installation should make this clear.

P1050232%2520annotated.jpg


The existing CH wiring is nothing to do with me! However, with a combination of reading instructions and looking at the existing wiring the UFH connections were not difficult to make and it all worked first time. Just need to clip back the cables and put some insulation around the pipes and the jobs done. However it is still only 4 weeks since the screed was laid and it needs another 2 weeks minimum before the UFH is used regularly as forcing the drying with heat will weaken the screed.

Overall costs were about £1100 for all the UFH pipework and controls. I’m not sure how much I’ve saved - but it cannot be less than £1,000. Just as importantly I now have a system that I completely understand in the event of any future problems. Hope this is helpful to someone. It wasn’t difficult so give it a go!
 
Eric The Viking":2iwn5p4m said:
I'm guessing the volume is similar to equivalent radiators, and that it's not isolated from the rest of the primary water circuit and thus you don't need any sort of expansion vessel. Is that right?

I think the circulating volume will be greater than the equivalent radiators. According to my basic arithmatic, 240m of 15mm pipe will contain 42 litres of water - or about 5 buckets full. It does run off the same primary circuit though - but note my comments about the differential in temperature between the UFH flow and that for radiators. This is dealt with by the single zone pump and mixer valve. We have a pressurised system with an expansion vessel, but I have read nothing that says UFH requires it. A simple open system should be fine - as you say, just treat the floor like an additional big radiator.
 
gregmcateer":uh9kjg8d said:
Roger,

That is a great WIP - really helpful.
Can I ask the square footage/meterage? (So I can do a fag-packet calc for our proposed area)

Cheers

Greg

Greg - the kitchen/utility has an area of approx 35 sq m, but approx 12 sq m is existing solid floor which we haven't applied UFH to, and also there is an area of about 5 sq m that we've left unheated down the sides where the kitchen units will go, so in total I've applied heating to 22 sq m of floor. The single zone kits that polypipe do are fine for up to about 30 sq m.

HTH
 
Hi Roger
Good post but can I ask a question? Why did you need the additional 100mm of cellotex?
To comply with building regs, your builder would have needed to insulate the new floor slab with a minimum 100mm anyway when he cast it.

or is it just extra?

Bob
 
The main floor slab has no insulation underneath it. Our architect wanted to limit the thickness of the mass of concrete that has to be heated so advised me to put the 100mm of Celotex between the slab and the screed containing the UFH pipes. It still took 4.5 tons of cement to make the screed.
 
RogerM":s73dasx1 said:
The main floor slab has no insulation underneath it. Our architect wanted to limit the thickness of the mass of concrete that has to be heated so advised me to put the 100mm of Celotex between the slab and the screed containing the UFH pipes. It still took 4.5 tons of cement to make the screed.

Ah ok, I understand his thinking.

To the best of my knowledge then, had you changed your mind about the heating, you would still have needed to install the cellotex to avoid contravening the regs.

Bob
 
Lons":3ial3n7k said:
RogerM":3ial3n7k said:
The main floor slab has no insulation underneath it. Our architect wanted to limit the thickness of the mass of concrete that has to be heated so advised me to put the 100mm of Celotex between the slab and the screed containing the UFH pipes. It still took 4.5 tons of cement to make the screed.

Ah ok, I understand his thinking.

To the best of my knowledge then, had you changed your mind about the heating, you would still have needed to install the cellotex to avoid contravening the regs.

Yes - the insulation always needed to go between the slab and the screed. Given that the Celotex needs to be laid on something level, in between the screed and the slab is the logical place. Between the oversite hardcore and floor slab would have made no sense. It remained up to us to decided whether to go for UFH or radiators - a 75mm screed would have been needed anyway.
 
.

Can I ask what are the anticipated flow / return temperatures through this array of pipes?


EDIT: Sorry - spotted the temps in the text...............

.
 
Argus":36k34y2x said:
.

Can I ask what are the anticipated flow / return temperatures through this array of pipes?

Flow from the boiler is around 65 deg, but the zone regulating unit/pump (ZRU) drops this down to a lower temperature in the 40 - 60 deg range before it enters the UFH pipework. I'm finding by trial and error that anything more than 45 degrees and the room temperature overshoots that selected due to the continued release of heat from the floor screed after the boiler turns off.

Eg with the flow at 50 deg and with a room temperature of 18 deg selected, the room temperature overshot to 19.7 which is too warm for a kitchen. With the flow temperature down to 45 deg, and the room temperature set to 17.5 deg the room temperature overshoots to 18.5 degrees. The boiler cuts out when the room temperature gets to 17.5 deg but the temperature of the room stays above the 17.5 deg selected for about 5 hours after the boiler cuts out, so it stays off for hours.

Not sure what the return temp of the UFH water will be - it feels quite cool in practice - probably no more than 30 deg - and of course this is mixed with the flow from the boiler at the ZRU to keep the flow in the UFH down to 45 deg.

HTH
 
We did a large garden room extension a couple of years ago with underfloor heating and we had to use an anhydrite screed which was pumped in and was very expensive.

Very impressed with your work Roger. Don't think i would tackle something like that myself. Me and plumbing don't get on.
 
skipdiver":2an738o5 said:
We did a large garden room extension a couple of years ago with underfloor heating and we had to use an anhydrite screed which was pumped in and was very expensive.

Anything involving pumped concrete costs a fortune. Not only do you have the cost of the materials, but also the costs of cleaning the pump and hose afterwards. For this job there was nothing special about the screed - just a simple 4:1 mix done in a simple portable mixer. This was laid by our plasterer and his mate as I know my limitations and laying 4.5 tons of screed is beyond them! I'd be happy to do a small area though. It needs to be a min of 65mm thick - in this case it's 75mm to bring the floor up to the level required to meet the existing floor level and the doors. The consistency was about the same as you would use on the beach for a sand castle - certainly drier than mortar used for blockwork.
 
RogerM":2hlffnwz said:
skipdiver":2hlffnwz said:
We did a large garden room extension a couple of years ago with underfloor heating and we had to use an anhydrite screed which was pumped in and was very expensive.

Anything involving pumped concrete costs a fortune. Not only do you have the cost of the materials, but also the costs of cleaning the pump and hose afterwards. For this job there was nothing special about the screed - just a simple 4:1 mix done in a simple portable mixer. This was laid by our plasterer and his mate as I know my limitations and laying 4.5 tons of screed is beyond them! I'd be happy to do a small area though. It needs to be a min of 65mm thick - in this case it's 75mm to bring the floor up to the level required to meet the existing floor level and the doors. The consistency was about the same as you would use on the beach for a sand castle - certainly drier than mortar used for blockwork.


Yeah, mixed and laid plenty of normal screed in my time but the BCO on this job insisted on this anhydrous stuff which was sourced from Lafarge and taken out of our hands. We still had to do all the prepping for them though. It is very much stronger and lighter than normal screeds and allows heat to travel through it well. Can't remember the reason for him making us do it but think it was something to do with the thickness. We didn't have 75mm depth and BCO said normal screed wouldn't be strong enough with all the pipes in it. Had to pass the extra cost onto the customer and they were not best pleased.
 
Hi Roger,

I have finished my underfloor heating - inspired by your excellent WIP - so thank you!

I hope a quick question - do you run yours at a constant temperature, or do you have it timed on and off - and if the latter, what sort of time do you find you need it on and off?

I hope that's not too difficult or complex to answer!

TIA

Greg
 
Normally with screeded underfloor heating you would use what's called a night setback thermostats so that the temperature drops back to a preset level overnight based on a time signal - to keep running costs down. The response time on screeded or pipes inset into concrete is fairly long, if you turn it off altogether it can take a while to get back to the desired temperature.
 
Greg - it's pretty much as mentioned by jimmy-s. We have a Polypipe setback programmer which is set for 17.5 deg during the day and 15 deg at night. The timer calls for the daytime temperature at 06.30 and is on for about 2 1/2 hours, or until the temperature exceeds 17.5 deg. We find that in practice, the room temperature overshoots the set temperature by about 1 - 1.5 deg, particularly if we are working in the kitchen. The timer comes on again at 15.00 for 1.5 hours but only fires up the boiler if the room temperature is below 17.5 deg, and again, in practice, we find that this only happens on the very coldest of days. It's not happened so far this autumn.

During the rest of the day and night the boiler will only fire up if the room temperature drops below 15 deg, which it never does due to the very high levels of insulation.

Depending on how well you're insulated, and on how much the room temperature overshoots the set temperature, you may want to reduce the temperature of the flow. Whilst the instructions suggested 40 - 50 deg, we found that far too hot and at the moment it's set to 35 deg which is plenty to keep the room at around 18 - 20 all day after the morning firing and without any topping up during the day, and we only increase it to about 38 deg even in the coldest weather.

HTH.
 
Hi Roger I had underfloor heating installed by a so called mate a few years ago .As soon as we turned on the boiler blew up .This happened because no bypass valve was fitted .My so called mate refused to came back to have a look at it and i had to pay a plumber to solve the problem.I wouldnt want you to have this problem and i dont know if you have fitted a bypass valve but here is the diagram from polypipe.

UNDERFLOORHEATING_zps1418bfe3.jpg


Cheers Bern.
 
A good point Bern. We do have a bypass in the system, but would only expect it to be used in the event that there is a failure elsewhere. All zones, including the radiators, hot water and the under floor heating are controlled by a programmer which opens/closes a zone valve so that in the event that none of the zones are calling for heat, the boiler is switched off, so the boiler will never be on with no where for the primary flow to go. It only fires when one or more of the zones is calling for heat and the zone valve for at least one zone is open to allow the primary to flow.

I would also expect the thermostat on the boiler to cut out before it was allowed to overheat. Sounds like either you had a failure somewhere in the system, or (more likely) there was a mistake in the wiring that meant the boiler didn't switch off despite there being nowhere for the pumped primary to go?
 
Back
Top