14 hours ago, Oz07 said:

crack just anywhere

That can be controlled by putting in crack inducers, esp at doorways.

Our own floor over screed is 12mm thick tongue and groove, including mm oak veneer. Glued down onto screed that has had 8 weeks to stabilise.

1 hour ago, Oz07 said:

I thought you had to have floating on top of ufh? I like the look of parquet flooring in the smaller traditional size but all the modern floating stuff seems to be massive planks

Mine said clearly to glue down. Floating you have lots of air gaps and they hinder heat transfer. 5 years of heating and cooling, look as good as when installed

@Alan Ambrose the frying pan intuition is reasonable but the numbers say the screed doesn't spread heat sideways much. Screed is about 1.2 W/mK. Take a 65 mm screed, 5 K difference between two rooms, heat travelling 1 m sideways: 1.2 x 0.065 x 5 / 1 is about 0.4 W per metre of wall. Basically nothing. So the floor surface temperature follows the pipe density and flow underneath it, room by room, which is why balancing at the manifold works even in a well insulated house.

At PH-level losses (say 10-15 W/m²) the floor only needs to be 1-1.5 K above the room, so it self-regulates a lot, and wide spacing is fine. That's the real argument for keeping it simple: similar loop lengths under 100 m as Spinny and torre said, one loop per room or group of similar rooms, all open, and set the flows once. +1 to stevenlee on pressure testing each loop and keeping a sketch.

Thanks for a very informative post - I hadn’t done the calc or modelling, I was just using the ‘concrete is a good thermal conductor’ shorthand. There’s nothing like an actual calc or some actual data :). The concrete underneath the screed might be having an effect - or maybe not much.

1 hour ago, Alan Ambrose said:

The concrete underneath the screed might be having an effect

Why both concrete then screed, why not just concrete to the design height?

What thickness of concrete/screed and UFH depth?

1 hour ago, JohnMo said:

Why both concrete then screed,

Good question. It is a lot of work and complication to fit pipes then pour a slab around them, which is a gutty job with welly boots clambering about. Also there is precision required in positioning pipes.

I prefer to think structure first, then Internals.

But then again, getting a slab flat is a rare skill.

@Alan Ambrose can we go back a stage and consider the overall construction or us that fixed? ( you have prob told us already, but my memory confuses other discussions.

Where is the insulation?

I pleased to report that that ship has sailed. We have a raft with 150mm slab on 300mm XPS with a door threshold ~95mm above that and a floorplan. Plus walls, roof etc, no windows yet. Expect to be laying the loops / screed in January.

So, the only remaining decisions I think are whether to make that all 95mm screed or, say, 25mm XPS plus 70mm screed …and… the loops.

I’ve been buying into the ‘whole house at 20/21 degrees 100% of the time / use the slab as a thermal battery with cheap rate overnight’ thinking - hence treating the whole slab as one uniform thing rather than trying to heat each room separately like the good old days of radiators. Hence my original question.

Ahh.

So now you have a concrete slab, without the ufh pipes.

You already have the insulation beneath the slab so that the heat stays in the room.

so, the last thing you should have is more insulation on top of the slab, or even a screed.

Why do you want a screed? How flat and smooth is the concrete?

Look at having no more insulation and the minimum of screed that will enclose the pipes. 45mm.

The price of screed won't double for 95mm so that is up to you.

You will have 245mm of slab which is unnecessary. It will act as a good heat sink but with a slow response.

But no more insulation! It would trap heat in the thick slab.

@Alan Ambrose +1 to saveasteading on skipping the extra XPS. With 300mm already under the raft, another layer on top of the slab just cuts the 150mm of concrete off from the pipes and you store heat in whatever screed is above it. For the overnight battery idea, thicker screed toward your full 95mm gives more mass to charge on cheap rate and slower response, which is kind of the point. Minimum cover around 45mm as he said if the slab is flat enough.

On loops, uniform layout still works for that approach, but it'll be several circuits not one. 16mm pipe tops out around 100m per loop before pressure drop gets silly. Keep them roughly equal length at the same spacing, maybe a bit tighter near the big glazing once you know where the windows go.

If you go with some insulation the response time for the heating will be quicker but the heat won't go into the slab much.

The layer of insulation (between concrete and screed) will allow downwards heat travel and then slowly warm up the concrete below, but once warmed it will be effectively trapped heat until you start cooling. Then the cooling will make the heat migrate upwards until temps stabilise. It will make the start of heating slow and same for transition to cooling. But once heat soaked it not going anywhere and will be stable.

But the big BUT, your goal of batch charging with 70mm of screed engaged is flawed you're going need 100mm and more for that to be effective on a cold day. Not sure if you are single storey or two, as that will also make a difference to how effectively you store heat.

I would do UFH on top of your concrete, but choose your screed well.

Slow and batched charged floors is good. But ignore everything you have read about weather compensation, because it will not work and give almost zero advantage and plenty of disadvantages. Fixed flow temp at design day flow temps will work well and allow a high rate of heat transfer so your heat pump doesn't start modulating down to much, which you don't want when batch charging. Because of this you can also get a great CoP as dT stays pretty wide and you have quite low flow temps. This morning (4-7am) running at 33 degs for 3 hours of cosy window, I got a CoP 5.95 at 9 degs outside.

But if you are taking this approach to heating consider how you will heat DHW, as using the heat pump kills the batch charge approach very well. Just use the immersion.