Firstly I am well aware that the best thing to do is kango up the floor slab and insulate below but we decided 15 years ago not to do this and bought external doors to allow for 50-80mm floor-height raise. So whilst I remain interested in sums analysing this option I'd like to skip a load of advice suggesting that in this thread if possible, and concentrate of what is least-bad above-slab. We can direct people to http://www.greenbuildingforum.co.uk/newforum/comments.php?DiscussionID=13875 for that.

This job (floor insulation) has now reached the top of the pile so its design time.

I am trying to quantify the pros/cons of various flooring retrofits and whether to put in UFH despite limited height and thus insulation. Sums are much harder for floors than for other building elements because simple area calcs are just wrong. Has anyone come across any tools other than PHPP to help with this?
Someone must have worked out a way to do some FEM analysis with a FreeCAD model by now?

So what I have is a rectangular 1960s detached house which is all close to enerPHit grade now apart from my original concrete slab floor. I also have perimeter insulation (100mm XPS) on 3 sides down about 700mm below ground to the top of the foundation. The 4th side has a modern (2014) PH-grade extension with 170mm XPS below 150mm slab with UFH-in 50mm screed on top. Some more details at http://wookware.org/house/retrofit/

The combination of very low heat-load due to good airtightness and 3G plus U=0.15 walls all round, and perimeter insulation is one reason why I think it might still make sense to put wet UFH in the floor despite the fact that there will only be 30-50mm insulation under the pipe which various rules of thumb say is 'not enough'.
What I need is some quantification of how this all works in practice.

It seems to me that if the flooring is designed with high conductivity so most of the heat does up not down, and low flow temps so losses are minimised, this might be reasonably efficient. But maybe, in fact, it will still be worse/no better than than radiators? How to do the sums is my question.

Current flooring is 20mm parquet on 2mm bitumen. And there is ~20mm screed which comes up fairly easily in the one place I have checked so far (but one can presume that that may be quite variable across the whole ground floor. I have 60mm above the parquet to the back door threshold and 70mm above to the front door. So means we have 80mm to play with comfortably - maybe 90-100 if the screed removal works out well.

The current plan is something like:
5mm LVT (0.4 W/mK)
18mm cellecta hideck gypsum cement board (0.45 W/mK)
ali spreaders
50mm XPS or PIR containing 16mm pipe. (0.033 W/mK)

I can't find anyone selling pre-routed PUR or PIR sheets. They are all EPS or XPS. Seems to me they would be useful for the 'limited height' retrofit case.

There are plenty of suppliers that will sell you a 20mm EPS/XPS sheet along with 16 or 12mm UFH pipe to retrofit a concrete floor, but that's obviously pretty crummy (4-8mm of actual insulation below pipe - i.e. bugger-all, and thinner pipe adds to pumping loses). We can clearly do much better than that.

There are thinner floor-coverings that can work. I used 12mm multipro (calcium silicate, 0.26W/mK)) in my workshop (over XPS) and it's fine, but it's not T&G so you have add joint-pieces to stop differential movement. T&G floors are a lot less faff.

PHPP says my current overall heating demand is 52kWh/m2.a, and with the floor done as above it reduces to 35kWh/m2a, so 33% improvement. Digging up the whole floor to put 170mm XPS underneath would get it down to 28 - a 46% improvement.

But PHPP isn't allowing for the large thermal bridge of the central chimney or other internal walls on the slab which won't get insulated (maybe it has an approximation for that)? Nor that the perimeter insulation is different on one side, so those numbers are somewhat approximate.

It does show that this is the difference between enerPHit or not. (in previous iterations I did get the heat loss down to 28kWh/m2.a without the floor-digging, but I think there were mistakes. All my refinements since have made the headline number slightly worse so I'm not as close to enerPHit as I thought I was).

Hmm, I'm sidetracking myself here! What I was hoping to get feedback on was best floor-buildup options given 80mm on top of slab, how to evaluate this UFH vs radiators only (vs both), and any other useful analysis tools for how much heat goes up and how much down, and how sensitive this is to detailed material choice/thickness.

"FEM analysis with a FreeCAD model" A quick google of that threw up lots of stuff. Adding "thermal" in front of it still produced some interesting-looking hits. CAD is not my thing, so that's about as far as I will go.

What's the central chimney used for? i.e. does it still need to be there? I remember that you put a course of insulation in your external walls; presumably you could do the same in your internal ones if you were keen?

What's the screed made from? Presumably it can be made to come up everywhere by a suitable application of violence. Worst case you might need to make good the slab underneath in some places.

I make the sums add up a little differently.
Available height: 60 mm free door height + 20 mm parquet + 2mm bitumen + 20 mm screed = 102 mm
So you could add a little more insulation, I think.

As regards UFH, I think you're pretty much committed to the idea that the earth under the house will stabilise at a somewhat warmer temperature and so limit heat loss a bit. It sounds like your heat load is well within the capabilities of a UFH system so I wouldn't bother with radiators (you could plumb in some spare pipes just in case maybe). The system should work regardless of how much analysis you do or don't do

Sorry I haven't answered most of your questions.

Posted By: wookeyfloor-buildup options

Using conductive and high-heat-capacity materials (screedboard) immediately under the surface is good for conducting UFH heat during the winter hours while UFH is running. But feels cold underfoot during hours when the UFH is turned off, if you like to walk round without shoes. For your house, the UFH might not be on for many hours of the year? Even a few mm of timber (eg ply or fibreboard) makes vinyl warm underfoot, though can be disastrously absorbent if there is a slow water leak (we found!). Just a balancing act...

Posted By: wookeyhow to evaluate UFH vs radiators

The table on pg9 of "MCS-021 Heat Emitter Guide" is useful for this
https://mcscertified.com/wp-content/uploads/2021/10/MCS-021-.pdf

Posted By: wookeyhow much heat goes up and how much down

This is harder than meets the eye. Are you hoping to heat the block of soil inside your floor perimeter insulation to a) room temperature or b) UFH temperature?

If a), the UFH heat will flow down into the cooler soil and up into the equally-cooler room, in proportion to the R values upward from the pipe (surface+LVT+screed) vs downward from the pipe (insulation), about 10:1.

If b) the soil, once heated to UFH temperature, will not absorb any more UFH heat, which will all rise into the room. Once the UFH is turned off, some of the heat stored in the soil will slowly return back to the room, the rest will eventually find its way to the outside air or the subsoil, in proportion to the U values of the floor vs the soil (about 1:1).

If the UFH only runs relatively few hours per year, the soil will end up closer to room temperature (a) than to UFH temperature (b), and v.v. if the UFH runs continuously as it might with a heatpump.

Yes indeed - thanks. Link edited.

So. I've made mess and dug up a lot of parquet and bitumen and screed. I now have a somewhat irregular floor which is going to need some self-levelling goo before I put down insulation and a floor.

Do people recommend anything in particular? I reckon there is probably 15mm variation in slab top c.f. 'level' over the whole area, which feels quite thick for self-levelling? I can spend time cutting down the high spots to get it more consistent but it's tedious. It's important that the base is solid, flat and well-stuck down so as to avoid noisy and flappy floors.

(I found the original iron internal gas pipe in the process (it's in a slight channel in the slab and stuck up somewhat into the screed. I'll be getting rid of that before insulating as it no longer connects to anything and we will be disconnecting the supply in the next year or so.)