One of the key things I want to understand better relating to this high vs low temperature heat pump question, is what design flow temperatures can existing radiator systems support with an *accurate* heat loss calculation? There are various bits of research that have been published on this but I’m not confident in their conclusions given all the issues with accurate heat loss calculation on which the calculated design temperature ultimately depends. If the heat loss calculation is over-inflated the design temperature will also be over-inflated as a result.
I changed all the radiators and pipework in my house as I believed at the time that my heat loss was much greater and that the low temp R410a heat pump would have been unable to supply the temperatures required (~67C flow temp). In reality at my measured heat loss that same single panel radiator system would have only actually needed a design flow temp of 47C which technically is within the operating range of a low temperature heat pump ..
I actually thought my upgraded radiator system had a design flow temperature of 47C based on the original over-estimated heat loss (before I understood the issues with over-estimated heat loss calcs), in reality it only needs ~36C and hence the SPF of 4.2, which would be even higher with a newer heat pump.
It would be great to have a large number of examples of original radiator system capacities and then real world measured heat loss.
If anyone with measured heat loss data on HeatpumpMonitor.org can share their original rated radiator capacity before radiator upgrades it would be great to have more examples. This radiator schedule tool may be useful in calculating this rated output figure Radiator Schedule.
With my gas boiler, the radator capacity will increase the next time I am in the loft, as I believe it will behave a little better with the pump on a higher setting. Hence I don’t believe it possible to measure what flow temperature a heatpump will need based on most existing none optimized gas boilers. (We also can’t set a different flow temperature for CH then DHW reheat.)
I posted this under a rad subject previously but I think it is relevant here, modified the text a bit..
Heat loss from our survey for the only room with a rad in was 872 Watts, with the existing 600x1400 K2 rad, output was quoted as 1138W, 50C at -3C.
21C room target.
However our LWT during winter is 30 to 33C.
That rad at LWT with delta T of around 10 to 13C gives out around 612 to 650W.
It actually performed a bit better than we expected. Short fall of ~250W from the survey.
The room got to around 17C.
Various room heat loss websites calculators I tried gave 940W and 1100W as loss!
I found a website with Rad Delta Ts from 10C upwards - most useful.
That showed that a K3 600x1400 rad as a direct replacement would output around 1000 to 1100W depending on our exact LWT input.
This made up the shortfall and we get 21C nicely.
That room is now as warm as the rest of the UFH house.
With the UFH we wouldn’t want a higher LWT as it simply over heats the floor / house and has to be switched off / cool / on cycle endlessly.
Added:
The estimated output from rad calc in your link seems rather low compared to the data I found and based my change in.
A 2021 study from BEIS estimated that 90 per cent of UK dwellings would require new radiators to meet peak heat load for a common maximum heat pump flow temperature of 55°C, rising to 99 per cent at 45°C. However, data from the new study showed this to be an overestimate. While few dwellings could operate at the 45°C level without upgrades, 31.5 per cent of homes could operate at 55°C or less, with 66.5 per cent ready to operate at 65°C or less.
Although the study showed that more UK homes were suitable for heat pumps than previously thought, the researchers noted that high temperature heat pumps capable of operating at 65°C are significantly less efficient than their low temperature cousins. Nonetheless, they could still play a role in decarbonising home heating.
I’m a bit skeptical of these conclusions but perhaps I am being overly so? I believe the BEIS study used RdSAP estimated heat demand calculations. It would be great to see a study based on measured heat demand.
We’ve been running our gas boiler at 45° flow temp for quite some time now, and it works just fine. We only heat to 18.5° (occasionally 20°) however. The boiler struggles to modulate that low sometimes - often it achieves 47°. When it does nail it, there is a dT of 5°. This is with all TRVs and lock shields fully open.
Octopus calculated our heatloss at 4.1kW and are installing a 4kW Daikin in 2 weeks, with 7 rad swaps (albeit small differences). Design temp of 50°. Knowing that we can run at around 47° on existing rads, I’m hopeful we can go down to around 40° with the upgraded rads.
We’re a 2018 4-bed detached 115m2.
