If possible, I would suggest that you start by improving the level of insulation, since better energy efficiency will reap increased benefits long into the future. 

@derek-m There may be some improvements that can be done in future in the bedrooms, but the main space already has triple-glazing, wall insulation etc.

Posted by: @wintergreen

↑

@derek-m There may be some improvements that can be done in future in the bedrooms, but the main space already has triple-glazing, wall insulation etc.

I would suggest that you carry out room by room heat loss calculations, since so many other factors will be dependent upon the results.

I do have heat loss calculations for all the rooms, but while they may inform the required heat pump capacity, the point remains that I will need a device that can output 55 degrees at a market-leading COP. Hence the first question. 

@wintergreen, at a flow temperature of 55C, you're edging towards the realm of HT heat pumps, and COPs tend to suffer as a consequence. You may want to look at some of Daikin's models, as they have a reputable line of HT heat pumps. Having said that, most reputable heat engineers and heat pump installers, will still be adamant that a 'conventional' ASHP, designed and installed correctly, will get the job done. Why does the "emitter value of the UFH drops significantly?" – what are you basing this on?

Based on your description of your property and your daily hot water requirements, it seems like using a dedicated internal heat pump cylinder such as the Vaillant Isostor could be a more economical option.

Posted by: @wintergreen

↑

I am about to put in 10MW of solar PV (expected annual yield 7MW).

Hmmm. Do you have the figure for the max output of the solar panels, as opposed to the annual accumulation?

By my quick reckoning, this sounds like your roof is going to have 27kW of PV panels on it.
That's very high

The maximum permitted export to the Distribution Grid is 16A (per phase), which equates to 3.68kW.

So if this PV installation is to use the common grid-tied 'string' inverters, then you'd expect to require a 3-phase connection.

Alternatively, a better option is to not export anything, but store it instead.
In that case we need to be discussing storage batteries... which aren't (yet) on your shopping list above.

I've posted a brief overview of the two main approaches to battery storage from PV panels over on the Hello topic.

Posted by: @wintergreen

↑

fitted with UFH under an insulated floating engineered wood floor. The emitter value of the floor is on the low side, and it was specified at 55 degrees

Ouch! I've used three different types of engineered wooden floor with UFH in my house.

To avoid the boards warping and the tongue & groove joints separating, the surface temperature shouldn't exceed 25°C.
That means the aluminium heat-spreader plates shouldn't be receiving water higher than 45°C.

This photo shows construction detail of those plates in position and the PEX pipe being clipped into place.
Is this how your floor was installed?

As you'll notice, 45°C is a very good match for the output from a heat-pump.
You shouldn't require 55°C  😲 

Neither of the above points fits within the title of this topic.
So if we need to discuss these further, we'll move these elements to categories where others will find them.

Right, on the PV front, we anticipate 10.1kW of panels, zero shading, largely south-facing. I have confirmed the yield calculations from a separate source, so they look ok to me. I intend to install around 30kWh of batteries alongside that.  We're on a 3 phase supply.

As for the design of the UFH, the only construction detail I have is shown below.  Apologies for the quality.

Short of digging up the floor, we're stuck with the 55 degrees.  It may well be unavoidable anyway, because it's a massive space:

It's an amazing room, but it's probably an example of architects not paying enough attention to heating design.  In any event, it seems clear that the performance of an ASHP at 55 degrees is a critical constraint for getting rid of the gas boiler (massively overspecced at 46kW!).

Posted by: @wintergreen

↑

First, the 55 degree specification is quite important.

😰

Firstly you've gone with NuHeat for the UFH components.
That's good. They're a reputable company, based in Axminster, Devon.

Ignore the concrete floor for the moment.

The detail of the suspended floor states that there's a 9mm plywood layer overlaying the joists.

Have you got an engineered floor on top of that?

Posted by: @transparent

↑

The detail of the suspended floor states that there's a 9mm plywood layer overlaying the joists.

Have you got an engineered floor on top of that?

Yes. That was the design Nu-heat came up with to go under an engineered wood floor.

Oh dear! 😥 

Firstly, I'm not 100% certain that your suspended floor is compliant with Part A of the Building Regulations (structural elements).

If your floor joists are the usual 400-450mm spacing, then a wooden floor requires a thickness of 20-22mm.
You're achieving that only by adding together the thickness of the 9mm plywood and the (14mm?) engineered floor boards.

Secondly, the 14mm diameter FastFlo tubing is effectively surrounded by insulation on all sides.
How the heck is that meant to work? 🤔 

Let's look at these elements individually:

A. the pipe is called FastFlo because it's a smaller diameter than the more commonly-used 16mm PEX UFH pipe.
Given the same pump pressure, the water in FastFlo pipe willhave a higher velocity.

14mm with a 2mm wall thickness gives a bore of 10mm and a circumference of 44mm through which the heat is radiated.
A 1m length contains 0.31 litres of water, and has an external surface area of 0.044m²

16mm with a 2mm wall thickness has a 12mm bore and 50mm circumference.
A 1m length contains 0.45 litres, and has a 0.05m² area.

So you've got 30% less fluid trying to radiate through 12% less surface area.

That's not a great start.

B. The pipe is resting in 25mm thick polystyrene board, Nu-heat Tri-Panel.

Polystyrene is an insulator.

C. The diagram you posted suggests that the 'heat transfer plates' are below the polystyrene layer.
I doubt that.

It's more likely that the arrows have been displaced, and that the FastFlo pipe is actually clipped into 900mm lengths of radiating surface like this:

Compare the flat surface area of that with the 400mm-wide aluminium heat diffusers which I've used:

D. Above the narrow heat-transfer plates, you've then got 9mm plywood before the heat even reaches the floor itself.

What conclusion shall we arrive at regarding the efficiency of that heat-transfer design?

You can ignore the maths; what does common sense tell us?

Look at the larger heat-transfer surface which Nu-Heat themselves offer by using their ClippaPlate system when installing from below.