I am doing as much as I can on my conservatory, to keep the costs down (and ensure things get done right!!, as so many tradesmen these days are bodgers).
Anyway, I have the dwarf wall done, the conservatory up, and inside is the concrete slab. It's currently 125mm below DPC, which is ALOT of screed! Figured I have an ideal oppertunity to install insulation and underfloor heating to reduce the depth of screed required to around 80mm.
I am unsure where to go now. I think I want water based heating, rather than electric, as I assume it can just plumb in, as an extension to my central heating system (my boiler is already over-spec, with regards to BTU).
Has anyone done this before, and can recommend anything...
The message from "Mark Gillespie" contains these words:
Yes, you can do it and you can do it in the way you envisage. It's not the way the pros recommend, but it works very well.
Points to watch.
You've jsut about enough room there for 50mm of insulation.
Use PEX pipe and make sure that you run your central heating boiler at a temperature that's well within the rating of the pipe.
The temperature will also have to take into account whether or not the floor is carpeted. If carpeted you'll need to run the pipe a bit hotter to allow for the inuslation effect of the carpet.
Make sure that there are no joints under the floor.
Make sure that you have valves on the flow and return legs.
Lay a snake of pipe with about 300mm between each run.
Prop the pipe up on blocks to ensure it's enclosed within the concrete but in the bottom half of the screed
Use either weldmesh or concrete reinforcement mesh on top of the blocks to tie the pipe to with zip ties to keep it in position until the concrete sets.
Brilliant system. I've been using it here for about nine years. No regrets whatsoever. Even radiant heat from the floor gives much greater comfort and no draughts such as radiators create.
This is the danger. You can go up to 60- maybe 70 for radiotors..but no more than 50C for UFH in screed.
That is not very good for a conservatory. I struggle with that spacing in a well insulated room. I would double that and go every 150mm if possible. That should net you about 100W/sq meter.
Or do what I did. Put down rebar mesh and tie wrap the pipe to it. Pipe should be laid in double spirals.
I think its better to have the mesh under: thats conventional for structures with weight on top.
A few more things. Hire a pump and pressure meter, and pressure test before the screed goes down, and keep pressure (5 bar or so) as you lay the screed.
You MUST have a separate zone valve for UFH. The time delays are such that what works fir rads will be hopelessly late for UFH. use a room stat also. Start heating up top 6 hours before you are going to use the place.
Consider a temp reducing valve..there are ones that sort of 'short circuit' the UFH if the outflow temp us too high.
My system was polyplumb. Their site is good on information, and you can read it and buy elsewhere if you like
formatting link
Let the screed dry for at least a week before firing up the system. If you have any loose bits of screed or cracks, I found a pint of PVA thinned a bit pured in, completely locked a whole loose section.
Tiles or stone is the best flooring, but laminate or wood is not too bad.
Do not under estimate how hot the floor can get under the sofa.
The message from The Natural Philosopher contains these words:
Check the spec for the specific brand of pipe you're using. As I indicated, watch that you're within spec.
The problem is the minimum radius at the ends of the room. There are various patterns you can use to get around this. Plenty of info on the web.
In this we disagree. Rebar mesh is fine. But the pipe should be IN the concrete rather than underneath it or you lose a lot of the efficiency.
Not a bad idea, but not in my view necessary.
That's where I take a different line. In my view, UFH in concrete is almost hopeless for part-time heating but perfect for situations where heat is required 24/7. The concrete floor is an enormous heat sink. The response time for effective heating is more than six hours in my experience. From absolute cold it can be DAYS. A separate zone valve is not a bad idea, but I can't really see a lot of point in a room stat unless it'll tell you what the temperature's going to be tomorrow. An outside stat might be more to the point :-). Our pattern of use is such that heat is required 24/7 and it's so simple to simply increase the flow in cold weather and decrease it in warmer weather that for us it's not worth automating the process.
It's quite common to put a mixing valve in so that the return from the UFH is fed into the flow to regulate the temperature of the water.
Give it a fornight or more, IMHO
Take all due precautions to make sure you DON'T have any iffy concrete. And remember to allow for your impregnated boards around the perimeter of each room. That hot concrete will expand.
Agreed. But carpet's fine. It's the ultimate luxury for the bathroom
I don't like the Polyplumb roomstat. It looks identical to the Speedfit roomstat, but the Speedfit's display is clearer, it has nine rather than five preset programs, and four rather than two user programs. It's also much easier to program (IMO). I don't know for certain whether you can use the Speedfit stat with the Polyplumb system, but if you can that's the way I would go.
My understanding was that it would stress and crack the screed..
Yeep. You don't need to keep to that spacing everywhere, but getting more density helps. I have a corridoor that carries all the runs to all the rooms. It looks like something out of 'born free' in the morning. Cats draped all over it. Pipes are about 3" apart there..
It will be. The rebar never lays flat, and the screed gets around it. pipes on TOP of the rebar.
Depends on how heavy and clumsy the lads doing the screed are.
Well I find that the mass is such that the temperature at the moment is modulating between around 17-18 at night, or rather early morning: If I tip the switch around lunchtime its up to the rated 19C around 7pm, and stays there on the stat till we go to bed around 1 a.m.
19 degrees is right for us..or around 20-21C if the air is dry.
The stat keeps it within a degree if run 24x7.
I half agree about the timer, but disagree about the stat. It was a nightmare without a stat. My wife kept opening the windows when it got too hot.
Ok, I did the whole separate pumb/reducing valve thing,which I think is overkill, but don't know what else is available.
I'll not argue with that. It was more like 6 weeks before I got my wiring sorted and fired it up..Left it running with no stat and on minimum boiler temp all night. The builders 'UFH is only background heat mate, it won't heat a house' came in in the morning to about 23 degrees of house. They had the grace to apologise..
Dear Mark I wholly endorse all the alternative changes and suggestions of the Natural Philosopher I have 150 mm gap on 20 mm Pex My pipes are on the bottom of the screed and connected to the insulation with plastic clips and I have kept all ferrous material out of the screed - its worth paying a bit more for plastic clips (I used the stuff from Kestrin and can recommend their specilist James Dale if you are willing to pay for design (but for this job I would probably design it myself) Make sure you read the pump scale right 5 bar is only just a bit up from the bottom on most of these hire pumps! I put mine up to 6 bar initially (and as the pipe expands the pressure goes down) and then left it at 4 bar for for an hour then down to 2 during the screeding I would certainly consider a mixing / temp reducing valve to get UFH temp down as NP suggests If you are prepared to pay - I recommend Kestrin but if price is the main factor (as opposed to getting the design right) it is much cheaper to DIY. Risk is getting it wrong! chris
It is screed not concrete so no NEED for mesh but a very good idea is to buy the Fibromesh (basically like glass fibre (but not) tensile strenthening) that one can put in the drymix of the screed. Works a treat. Side insulation not greater than 25mm and largely covered by skirtings Chris
I think one should be careful to distinguish between what you MUST do, what is not half a bad idea, and what is optional in all this.
Things you MUST do are
- avoid underfloor joints
- incorporate insulation (and at ground level, DPM) under and SOME at the sides. The more the better.
- avoid running the screed too hot. 60C is recommended as a maximum. - locate the pipes *in* the screed somehow.
- use at least 3" of screed.
Those that are very GOOD to do are
- cram as much pipe is as you can for faster warm up times,
- add a temp reducing system to protect the screed,
- pressurize the system to test and keep it hard while laying the screed
- Use zones and stats to control them. Running untimed is one thing, but without a stat? not for me, no.
- put pipe in conduit tube at slab edges - when going from room to room. Otherwise slab expansion may stress the pipes.
What is IMHO optional, but nice.
- separate pump
- any fancy and expensive pipe location system
- screed reinforcement (but it gives you something to tie the pipes to).
- fancy predictive timers and suchlike.
HOW you achieve all these things and where you draw the line, is down to the individual situation.
I have two inches of side insulation. For example. The tiles and laminate covers it easily. But if laying carpet its not a good idea.
As with all these things, you need to separate out from the plethora of info, what is vital, what is good practice, and what is something the manufacturer wants to sell you, or what is presented as correct just because it 'worked for me'.
If I were doing mine again, I would
- put in even more insulation
- put in even more pipes
- put in even more zone stats.
I would not put in a predictive stat.
Nor have I had any reliability problems associated with the UFH apart from an external motorised valve that failed and stuck open. GOD it was hot when that happened...
I have to say that I have seen installations that consist of nothing more than a few loops of copper pipe in an upstairs bathroom floor.Those work pretty well too. Sure some heat goes downwards..so what? sure you OUGHT to not have 70C water running in that space, but beacause its all air surrounded, the heat loss from the pipe is less, and the floor doesn't overheat.
Slightly confused about this. I already have my concrete slab down, so the pipe was going to live ontop of the insulation, and under the screed. Should be able to do 130mm (50mm insulation, 80mm screed).
Ok. What Mark is saying is that you want to maximise heat transer into the screed, rather than to a pocket of hot air. This keeps the screed hot and the pipe cool.
His solution is to raise the pipes a little so the screed gets underneath, and then lace to an overlay of rebar mesh.
Mine is to put the mesh down first, and lace the pipes on top, and then screed all over.
I prefer to keep pipes as low as possible, so that there is less chance of screws going in when mounting stuff onto the floor.
Don't leave us in the dark - whose roomstat did you use? I haven't found that many UFH-compatible roomstats (i.e. with a setback capability to stop the temperature falling too much during the 'off' periods).
The message from The Natural Philosopher contains these words:
There's certainly a danger of stressing and cracking a cement/sand screed. Best to lay the floors "in a oner" -- a good 4" or more thick if there's piping embedded in them
There was a case for insulating at least some of the pipes in that instance :-)
Agreed pipes on top of rebar. The rebar isn't there to reinforce and there's something to be said for the lighter weldmesh. But definitely the pipes to be in the concrete.
Definitely a case for a do-it-yourself job ! And preferably an all-in-one-go concrete mix.
There's more than one way to skin a cat. For us, the timing + seaonal control of the llow and return have been more than adequate control. But remember that we've got a lot of concrete down there as a heat sink and we've got it well covered by carpet which controls the heat release.
The message from The Natural Philosopher contains these words:
I'd put down the mesh first and fasten the pipes on top.
I like to get the pipes totally enclosed -- but I agree that one has to bear in mind the danger of having people screw things to the floor, especially at doorways. There's a case for a steel plate there! :-). I favour laying the pipe in concrete rather than in a sand/cement screed. Less likely to disintegrate. However, there are arguments both ways.
Well its actually only a small area, and the landing above it has no heating whatsoever, so it all works out.
100% agreement there then. The architect did nlot specificy rebar in the screed - but said it would help, and if it gave a pipe laying matrix, so much the better.
Lay extra pipe near walls and near outside doors where cold bridging and other extra heat loss is likely.
I cant plaster and I can't screed.
The builders I used made a mess of it too.
Had to glue the cracks together and use self levelling under teh wooden floors..even now its not perfect..the slatees had 'variable cememt' from
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.