Viability of retrofit UFH not all at once?

[danlightbulb]

Hi all, so this is a speculation question at the moment.

I have a 3 bed semi currently on combi boiler and radiators. Solid wall 1930s. Loft well insulated and newish double glazed windows.

Downstairs is suspended timber floors, uninsulated, drafty with some gaps. Rooms are hallway, kitchen, living room, separate dining room. So 4 independent rooms all of which currently have radiators.

I have a broken joist in one room so will soon be pulling up the floor to repair it. At the same time I was planning to insulate that floor. Eventually (over a period of years) I was planning to insulate under the whole ground floor, room by room.

Is it viable to swap to underfloor heating, on a room by room basis over a period of time, perhaps using the existing radiator pipework in each room? If so how would this be carried out in terms of ordering of the retrofit?

 

[Madrab]

Is it viable to swap to underfloor heating, on a room by room basis over a period of time, perhaps using the existing radiator pipework in each room? If so how would this be carried out in terms of ordering of the retrofit?

Yes it is viable. Could you use the current pipework, not really. Multi zone UFH uses a different type of pipe (Pert/Pert-al-Pert/PEX, etc) running from a centralised manifold and is laid in a serpentine/concentric type of arrangement.

How it would be approached would really be down to what system is used but yes it could be done room by room. Suspended floors can really only use a couple of methods, inter laid between the joists or overlay.

 

[danlightbulb]

Yes it is viable. Could you use the current pipework, not really. Multi zone UFH uses a different type of pipe (Pert/Pert-al-Pert/PEX, etc) running from a centralised manifold and is laid in a serpentine/concentric type of arrangement.

How it would be approached would really be down to what system is used but yes it could be done room by room. Suspended floors can really only use a couple of methods, inter laid between the joists or overlay.

So I'd be comfortable DIYing the floor reconstruction including the insulation, floor boarding system and integrating the UFH pipework in the grooves. However then Im left with a UFH pipework system not connected to anything.

What would be the step of installing the appropriate manifold when all Im doing is one room now, and the remaining rooms could be years down the line? I can't pull up the remaining floors to connect the new room UFH back to the boiler.

Hence what I was asking was can I use the existing radiator supply heating pipework in that first room as the location for the manifold, but then once that floor is down then how would I ever connect the other rooms to it in the future? So could I utilise the existing radiator supply and return pipework and have a manifold per room, rather than one central one? But then would I be left with something difficult to hide and it looks unsightly.

 

[Madrab]

An UFH manifold would usually always be located central to all of the loops and always above the floors to allow for control and maintenance, usually in a downstairs cupboard in the hall, that allows the min length of pipe run before getting the supply gets to any given loop in any given room. When a room is to be converted then the loop is laid and then run to the manifold location usually through the wall to the central location

You would size the manifold with the number of loops in mind and run a new feed and return to it from the CH's 22mm backbone and then install a 2 port valve on it to control the supply.

 

[dilalio]

You could fit the manifold (which really needs primaries to it, from the boiler in 22mm flow and return).

Install the loop for the first room.

Run lengths of UFH as tails from the manifold for all the other rooms to positions where they can be picked up later, when doing each additional room, so when first room floor is up, you use this access to get tails into the other rooms as far as possible, so they can be joined onto when doing the next room and so on.

It's not best practice though, as you would likely end up with inaccessible joints! However, these couplings are what you'd have to use as repair if a loop ever got damaged.

Best if you could do each room and still get back to the manifold with a complete loop.

But only you know the logistics and layout of your house.

 

[danlightbulb]

Is it worth the hassle do you think? If I'm insulating under floors and keeping a radiator obviously that would be much simpler and cheaper. It would have been a good opportunity to do UFH if I'm getting floors up, but maybe it's not worth the faff.

 

[Madrab]

UFH, if done correctly and coupled with good insulation is far superior in terms of user experience, efficiency and overall comfort than radiators, without a doubt.

 

[dilalio]

Is it worth the hassle do you think? If I'm insulating under floors and keeping a radiator obviously that would be much simpler and cheaper. It would have been a good opportunity to do UFH if I'm getting floors up, but maybe it's not worth the faff.

If you're doing a lot of work on the floors it's well worth it. UFH needs much lower flow temp so you'd be future proofing for when we are forced to use low temp heating systems.

 

[ericmark]

I would say no to UFH, because of how long it takes to heat up and cool down. I can see when using heat pumps, the whole idea changes, due to them not getting the water as hot, but in my house at least, I heat rooms as and when required, I do maintain a background heat in the rooms, but the time from deciding I want to use the room, to the room getting to a comfortable temperature is important, and the one room we with mother house put underfloor heating in, it took ½ hour to feel any heat in the floor, an hour before the floor was really warm, and 3 hours before the whole room was warm.

The other problem is the reverse, for the floor to stop heating the room, again mother house large bay windows in living room caught the sun, so morning sun could overheat the room, changing the TRV head to an electronic type, speeded up the time it took to react, and setting the lock shield valves, resulted it the radiators staying warm, rather than alternating between hot and cold, this helped, and where the sun was taking one end of room to 32ºC, when set to 22ºC, on getting settings correct, that was reduced to 25ºC when the sun hit the room.

I would think with an old people's home, where the heating is static, day or night, temperature is static, UFH may work well, but I do not like wasting money, so bedroom not heated in the day, and living room not heated overnight, and office, craft room, dinning room, garden room etc, only heated when we want to use them, so reaction time is important.

I will admit the system is not perfect, for some reason, when the house was built, radiators were placed under the windows, this means the TRV head is also under the windows, on an outside wall, since next to an outside wall the TRV tends when heating is not running to be cooler than the room in general, so they open and when linked, start the boiler when the room is really warm enough, so Autumn and Spring I have to set them to a lower temperature to what I really want, in winter they are fine, the radiators cause an air flow in the room, so room is spot on, just Autumn and Spring when the TRV head can activate the heating when not really required. Using wall thermostats would likely cure the problem, but it is such a minor problem, hardly worth the effort.

Again, if a room does get too hot, it does not take long to cool with radiators, but with UFH the reaction time is simply too long, there is a reason why we stopped using the Hypocaust

it looks good, but likely it did not work that well, just the Italians were not use to our climate? What did the Romans do for us?

 

[Madrab]

With traditional thermal mass UFH (TM UFH), then yes, the way that has to be run certainly follows on with those requirements. other types of UFH do work differently though.

Overlay UFH reacts reacts just as quickly as rads do (heat up and cool down) and doesn't need to be run constantly (which TW UFH really should be). Each loop is zoned and each loop is typically a single room with it's own thermostatic control, therefore there is individual room control. Rads can also be retained as an augmentation and controlled the same way and/or by individual TRV's, giving even more overall control.

When it comes to space heating UFH is far more efficient. Radiators, using convection, have to heat a whole lot of unused space, all the space above shoulder height up to the ceiling. As UFH heats from the floor up it therefore the room heats more evenly and eficiently and can be run at a lower temp, typically a couple of deg cooler.

 

[dilalio]

there is a reason why we stopped using the Hypocaust

I think control logic is a little better these days!