Sounds perfect - what's the problem/question?

Posted By: hitormissPolished concrete supposedly needs at least 100mm over heating pipes.

That doesn't agree with what I've seen elsewhere. Its common to tie the underfloor heating pipes to the steel mesh which normally sits centrally in the concrete slab. So 20mm pipes in a 100mm slab would typically have around 30mm of cover.

Where are your underfloor heating pipes? In the concrete slab or in the flagstone bedding layer?

Where's the insulation? Is it below the slab?

David

A neighbour has some self-catering cottages with wet UFH under random (i.e. crazy paving like) slabs of Caithness slate. From what I recall of seeing the slate on the pallets it was probably 25-50mm (I'm sure there was variation in thickness)? The pipes were cast into a concrete slab before the slate was bedded in on top. I don't know the depth of the pipes in the concrete, but I could ask if you were interested (the last was built a few years ago now, but as they built 4 in total they may remember!). The system works very well.

Tom - S'obvius innit?
Actually I was thinking of resistance - for a specific rate of heat transfer to the air above, the thicker the layer the hotter the pipe, and so - if you're using a heat pump - the lower the COP... Or for a condensing boiler, the less condensing. But you know this already. I suppose this dwelling could have such a small heat demand that the heat supplied by the floor is so low - almost acting as active insulation - that these differences would not amount to much, but we don't know enough to say that.

Whether it is a good idea to have so much thermal mass above the heating layer seems harder to say, other than that with the same thickness of stone but in two layers, one above and one below the hot bit, you'd have about the same thermal inertia but a quicker response. But you know that too...

Posted By: fostertomMore inertia the better

Explain that more please.

I would like to see some data to back your assertions up Tom. There is enough of us interested in this that we should be able to come up with a simple way to do it. The only person I know who had some data on a low mass house was, like me, amazed at just how stable his low mass place was.
Anyone got temp data for their houses over a few months or better still years?

Tom, - whilst a lot of what you say is the considered wisdom as it were, it is highly dependant on the occupation and what tolerances can be allowed.

For very similar occupation patterns and temperatures, there is very little difference between high mass and low mass building - all other parameters like airtightness being equal.

Either you put a lot of heat in fast, in a low mass building, then throttle back to tick over to maintain condition or you put a longer lower heat input to a higher mass building and maintain slightly more stable internal conditions (which may well extend into periods where you are heating the building when not required.


Basically, though, the total energy burden will be pretty equal in either type.


I agree with your approach to very low delta T between the room condition and the floor (adjusting for a much higher radiant fraction), but basically, the more resistance (and thus delta T) between the heating fluid and the heat emitter surface then the higher the fluid temperature needs to be to achieve heat transfer.

The last school I did had 300mm of insulation above a ground slab, and had the UFH in approx 250mm of concrete on the insulation - it was actually sitting on and tied to the lower mesh which sat on 50mm chairs.

Floor temperature is designed to be circa 22C.

In terms of energy input, it's now clear to us, that we could have moved the UFH significantly shallower in the slab, tightened up on the density and used a lower water temp to max out the heat pump CoP's.

The modelling would start to show that putting the UFH on spreader plates and just covering with a resonably conductive flooring would allow even lower temps.

The counterpoint to this is stability of temperature - something you may well need more of in a continuously occupied building than one with cyclic or intermittent occupancy.

In my experience, just having mass doesn't actually equate to lower energy consumption.

Regards

Barney

Posted By: SteamyTeaExplain that more please.

Is it really fair to derail “every” thread with this discussion?

Posted By: fostertomBut the conductivity of the flags + concrete is so high that it'll make little difference to delta-t.

Even with this?

Posted By: owlman…"bedding in" slabs of such varying thicknesses to create a decent, level FFL. Maybe some sort of semi-dry mix?

Not sure but can't help feeling that such a mix will have a lot lower conductivity than stone or concrete.

Would sorting the flagstones into thick and thin and having two levels of underlying concrete to suite be a feasible way to minimize the amount of bedding?

In terms of thermal response, you just need a bedding material of similar conductivity to flag - the "heat" doesn't give a monkeys what it passes through - but if it passes through at similar rates by matching bedding to flag then it doesn't matter - you'll end up with a reasonably equal floor temperature.

So perhaps stone dust rather than sand in the mix ?

Regards

Barney

Posted By: Ed Davies

Posted By: fostertomBut the conductivity of the flags + concrete is so high that it'll make little difference to delta-t.

Even with this?

Perhaps increases delta-t across its thickness, relative to normal screed + tiles, from 1C to 3C? So mean water temp might be 42C instead of 40C. Not v significant, compared to the benefit of greater thermal mass, as described, and nothing compared to wood or carpet as finish.

I think you are worrying far too much. Our slabs vary from less than 2 inchs to about 4. Some of them I had to cut down in thickness from 7-8 inches (basically sandstone blocks). The thick ones were original, the thinner ones were the closest match I could find in the reclamation yard.
Crushed recycled glass and lime screed. Lime pointing.
Underfloor runs off the woodstove so cannot comment on COP.
No problems, but then the walls are about 2 foot thick so the floor layer is probably thin in comparison.

Sorry, that was as meant to be as close to the top flags not slab, as possible.