Looking at the data on Oversizing vs SPF/SCOP for 22 Daikin heat pumps on HeatpumpMonitor.org the correlation is very weak but Im sure our experience here on this forum in terms of all the discussion around Daikin performance has suggested that right sizing can have a significant beneficial impact on performance.
How do we make sense of this?
@matt-drummer @ColinS @meatballs @mortstar @MikeEvans @1nc @KnightPhoenix any thoughts?
Maybe a useful comparison between two units:
8 kW, Daikin Altherma 3, Bournemouth, SPF 4.0, unit max output is ~2.2x measured and projected demand, average flow temperature when running 34.6C:
https://heatpumpmonitor.org/system/view?id=113
6 kW, Daikin Altherma 3, West Oxfordshire, SPF 3.5, unit max output is ~1.2x measured and projected demand, average flow temperature when running 34.3C:
https://heatpumpmonitor.org/system/view?id=162
9kW, Daikin Altherma 3, Reading UK, SPF 3.3 (1.2x oversize factor), average flow temp when running 33.9C
https://heatpumpmonitor.org/system/view?id=111
Hey @TrystanLea,
Great to meet you in Farnborough the other week.
One thing to keep in mind is that Matt and I had our units swapped out at the start of this year so your data is mostly from the previous heating season with the old units.
From personal experience my old unit would give me a cop in the 3.2-3.4 region at best and the new one is showing 4.33 for the last month with 4.97 for space heating.
In addition the lower flow temperature leaves the house feeling more comfortable and a consistent temperature (not a easily measured statistic but keeping the other occupants happy is very important).
It was great to meet you too @KnightPhoenix!
Thanks! that’s a night and day difference, which suggests the above chart is missing something substantial…
What does it look like if you round up the oversizing factor to the actual physical units without their software derating?
E.g. a 4, 6 are all 8s, 9 to 14 are all 16s (not 100% sure if this applies to older models - I know my 11kW has a slightly lower output power than the 11kW mono).
Daikins are probably the biggest cohort with heatpump integration reporting (and not via official HP API but third party/individual implementation) so something to be wary of.
Would be interesting to see what the difference is between the units of Matt and I for September/ October last year to this year.
The units were swapped in January/ Feb time so that may give you some ideas as to the difference the correct size unit makes on two separate houses (not a great sample size but a good start as far as direct comparisons go).
You may need to factor in for the degree days but would still be interesting.
There’s a problem with this I think.
Oversizing is based on the headline output of the heat pump.
On the face of it my heat pump is oversized, and it is technically.
But if I had a 4kW or 6kW version of my heat pump I wouldn’t notice any difference as they are all the same.
So, my 8kW is too much for this house but somehow it works even though I never need more than 4kW.
It’s pot luck really. If you have to have a Daikin you need a heat loss maybe between 4 and 6kW or 12 and 16kW
In the middle, no current Daikin heat pump will work very well.
It’s either the 8kW or the 16kW, that is the biggest of the two ranges and a heat loss to match. Otherwise, I say, look elsewhere.
This is a great before and after for your system @KnightPhoenix:
Before 8.6C outside, Room temp 20.9C (EDLA09DA3v3)
After 8.4C outside, Room temp 21.1C (EDLA08E2V3)
Monitored using class 1 electric and class 2 heat meter (kamstrup), so not a difference in internal sensors in the unit either… and pattern of operation is obviously so much better.
@matt-drummer do you have some good before and after comparisons under similar circumstances for yours? I think we cleared your old data in the app didn’t we?
Hi Trystan,
I don’t really know if the data is cleared.
What is shown now is just the new heat pump from 20 February 2024.
You helped me with this, so maybe if you help me again you could reset it to 7 October 2023 when I installed the heat meter and maybe the data is there? Or you have access to it?
There are a few previous graphs available on the forums and some here - I am lost and confused with my Daikin heat pump - #21 by matt-drummer
Another good example by @Rob-G :
Sorry yes the old data is there! great! here’s the change over:
COP jumps up from 3.98 over this first period with outside temp at ~7C:
up to 4.96 over this second period with the 8 kW unit installed:
Lower mean flow temperature delivers the results! and % carnot is a bit higher.
That’s 3 examples of just downsizing the unit resulting in pretty decent improvements in COP…
Here’s a very quick attempt at setting the oversizing factor based on compressor size of the Daikin’s , I’ve removed the daikin splits and HT models as they are another model range.
Maybe it’s just the reality that performance is dependent on so many other factors causing a wide spread of values that makes it appear like there is no trend when plotting oversizing factor vs SPF on the chart.
Would be interesting to see on that table which units are from which batch.
There is some corolation in the form of two backwards slashed and a few scattered in the middle but another factor on scop is going to be the radiator and system design.
While we know the standard approach of our tentacled friends is to aim for 50°c they also have a tendency to leave oversized radiators in place and not replace or design accordingly.
This could be leading to either overheating or trv’s choking off the system volume and impacting performance.
I think it is important to recognise one important factor when looking at mine and Zaks data.
In order to accommodate the 9kW Daikin heat pump we were both going to extreme lengths to get the best efficiency we could and enduring massively more heat than we wanted.
In my case I was getting over 4kW of heat to get a SCOP around 4 when I only neededhalf that or less.
Most of the heat went out of the windows and doors.
If we had actually tried to get less heat our SCOPs with the 9kW would have been way lower because we were both operating at the minimum electrical input of that heat pump.
I changed every radiator in the house after the Octopus installation increasing the overall size by 250%. Before this I had to run at a minimum flow temperature of 42c (with increased electricity consumption and even more heat) to get anything close to a COP of 4 and I think if I had stuck with it as installed I doubt I would have seen a SCOP over 3. And then there was the defrosting!
The improvement for us between the two heat pumps, that on paper seem quite close, was much bigger than the numbers suggest.
Not only has the SCOP improved considerably but we have much more comfortable lives and use less than half the electricity doing it.
Great points @matt-drummer
I think the overall conclusion is that plotting over-sizing factor vs SCOP gives a misleading / meaningless result. Suggesting that I should be more careful when trying to interpret aggregated high level data on HeatpumpMonitor.org.
Made another mistake above in the compressor size corrected comparison, I plotted oversizing factor vs the wrong SPF/COP column (DHW not combined!). Here’s the corrected plot, that does actually show somewhat of a trend, though again probably one to be very careful with as the visual trend is influenced strongly by your system @matt-drummer with it’s extra large radiators giving the SPF/COP near 5 and the Shirehampton 6kW (8kW compressor) with 1.5 kW heat demand. A system with 1.5x oversizing factor could easily get a much lower COP if the emitter system dictated a much higher flow temperature of course.
This is my rough estimate of how both of your systems have moved @matt-drummer and @KnightPhoenix since swapping the units:
@matt-drummer what was the date of your radiators changes? just want to make sure I dont mix up effects in my analysis here? @KnightPhoenix did you change rads as well?
Hi Trystan,
My radiator changes were completed on 02 December 2023.
I did it over three days and it should be quite clear which days they were as the heat pump was off for hours during the day. The upstairs were done on a Saturday, maybe 18 November 2023 and the first half of downstairs was done on a Tuesday, maybe21 or 28 November 2023.
