Fair points, but note it is not "necessary" or a "technical requirement" to replace radiators to install an ASHP. Ten years ago we had a Daikin Altherma HT which ran our existing rads happily at 80⁰C, being designed as a drop in replacement for oil boiler with two compressors in series. Most of the current crop of R32-based ASHPs are happy to run at 65⁰C, which is a direct replacement for an existing condensing gas boiler without touching the radiators.

So if it were a "technical" or "physics" criteria, then replacement rads would not be "necessary" and would not be zero VAT. I doubt we would be happy with tax accountants deciding about physics questions though!

However, larger rads are "better" as they increase the SCoP, so depending on the existing sizes, it might be worthwhile to replace them, or some of them. In our case, it is financially worthwhile to replace about half our rads, the others turned out to be diminishing returns to replace, as they were previously oversized for various mundane reasons. So it is good that the rads are zero VAT rated based on this value judgement, as the passage I quoted makes clear.

It is also not physically necessary to replace pipework, but is "better" for us on certain runs, because then we can run a lower DeltaT with less noise, so that is zero VAT as the guidance explained.

In our previous house we converted the loft by adding rigid insulation boards, with an "ancillary" need to plasterboard the ceilings, build dwarf walls, add trims, decorate etc. Was all at reduced VAT, the installer's accountant checked that HMRC were happy. Sounds like you were unlucky not to get that on your roof insulation, did you push it?

Posted By: WillInAberdeen

It is also not physically necessary to replace pipework, but is "better" for us on certain runs, because then we can run a lower DeltaT with less noise, so that is zero VAT as the guidance explained.

What if you only have microbore pipework in a manifold system for your C/H?

Hi Phil, yes I understand so, anything that makes the ASHP "better" is zero VAT, even decorative things like trunking and designer radiators.

Hi Jeff, with microbore it depends how much heat you need to move through it, and how long each run is, and whether it is 10mm 8mm 12mm etc. Personally I dislike it and would take the chance to replace it all, but we do have a few towel rads etc with minor heat loads where we will leave it. In principle if you can get enough heat through it to keep the radiators warm at the moment, then it is big enough. In practice the installers were keen to replace any component that might possibly eventually contribute to them getting called back under warranty, including the existing microbore.

Simon, I think you may be thinking of the previous generation of R410-based heatpumps which required low flow temperatures. The MCS-021 Heat Emitter Design Guide was issued in 2013, based on that older generation of heatpumps.

The newer heatpumps use better refrigerants (R-32 et al) for good efficiency at warm flow temperatures, to avoid the necessity to replace radiators.

For example the Arotherm Plus (R-290) has a SCoP of 3.63 at a flow temperature of 55⁰C, and has weather compensation so that it would only flow at that temperature on the very coldest days. That's the same flow temperature as our existing condensing boiler.

If the radiators were made bigger, it could flow at lower temperatures, eg at SCoP=3.9. It is therefore a value judgement whether to pay to replace the radiators in order to gain 10% on the SCoP on certain days. Many people would not bother as it is not essential, but we already wanted to move a few around.

To support these decisions, the MCS-031 spreadsheet allows you to try out different flow temperature to find the most economical radiator sizes for each home's requirements, balancing capital vs running costs. The MCS installers we dealt with, were also accredited by Samsung and Vaillant to use their specific software, so they were not using the MCS publications.



Regarding pressure loss in pipes, you have perhaps overlooked that the ASHP feeds a buffer vessel. A separate circ pump feeds from there to the emitters, so the DeltaT across the ASHP can be chosen to be lower than across the rads if desired. If we choose a 10⁰ DT across the rads, we need less circulation flow, so possibly smaller pipes and a less powerful circ pump than if we choose 8⁰C. But we'd need larger rads as their average surface temperature is about a degree cooler, so this is a factor in the radiator sizing economics, above.

This degree isn't an issue with the warm flow temperatures with R-32 heatpumps, but your logic of using lower DT and bigger pipes was correct for the old style R-410 heatpumps, where every spare degree was important.


If like me you are interested in flow physics, I used to teach my students to start from the Darcy equation arranged in terms of flow:
DP = L.(8.f.rho./pi). Q^2/d^5

From this we see that, all other things equal,
1m of 10mm pipe is equivalent DP to (1.5^5)= 7.6m of 15mm pipe.
And a 15mm pipe has (1.5^5)^0.5 = 2.8 times the flow capacity of a 10mm pipe for the same DP.

In my house I have some rads that are much more than 7.6m from the manifold, and one within 1m. Their heat loads differ by more than 2.8x. Therefore they have widely different pressure losses in their pipework, which we balance by adjusting the lockshield valves rather than having different size pipes for each radiator.

I calculated the pressure effect of choosing different DeltaTs for the ASHP and rads, and whether this just meant adjusting the lockshield, or whether the pipe would constrain the radiator capacity. Generally it was not necessary to replace pipe, but it was "better" for the longest runs to the biggest rads. The installers took a simpler CYA approach - see above!

(Edit to add: your pressure calc may have gone awry because you were trying to move 5.5kW in a single run, but no radiator can take that much heat. Try splitting the heat flow into a 0.5-1kW to each room.)



An observation is people should not assume the existing radiators were correctly sized in the first place. Some of ours were already 3-4x oversized, the previous installer had probably just picked standard sizes that looked right, and we had also added insulation. We are not changing the size of those ones!

Finally, the radiators for fossil boilers are often oversized so the house warms up quickly after being left unheated. This is less common with an ASHP which tends to run continuously. Again this contributes to the radiators not needing to be changed out.

Our ASHP was £1,200 for an 8kW Samsung plus £300 for controls. Installers seem to inflate prices hugely.