Here’s cooling/defrost energy vs heat delivered on our system for the last 3 years:
Zooming in on the peak around December 10, 2022:
4.1 kWh lost to defrosts 61.9 kWh of heat initially delivered, 57.8 kWh of heat net
Detailed view for December 10, 2022:
2579 W gross, 2408W net
Flow temperature is very low 31C, outside temp is above zero at 1C. Average room temp quite low (comfort preferences - we could have run the system hotter), peaks at 19C at end of the evening.
Looking at the Ecodan datasheet for our unit, it doesnt list a COP that matches the 31.1C flow temp and 1C outside temp exactly.
I’ve got a COP calculator that interpolates this table as part of https://openenergymonitor.org/heatlossjs/ and it suggests that the COP should be 3.74. Im actually getting 3.37. Without the defrosts I would have got 3.6…
That’s probably quite a good example of reduced performance compared to the datasheet.
I’ve never heard of this, I can’t see any reason why this would the case? Unless the evaporator or fan gets dirty, but it can easily be cleaned.
It’s useful to have a bit of capacity in reserve, to account for the occasions when you need to re-heat the house when it’s cold e.g coming back from holiday. In this case you need to input more heat than you would normally need to maintain the temperature. I’ve managed to get 85kWh of heat from my 5kW heat pump re-heating my house when the outside temperature was below 0C, this is way more than I need to maintain the temperature. But if you have a wood stove you probably don’t need to worry too much about this.
I don’t think this is always the case, in general higher rated heat pumps are noisier than lower rated heat pumps e.g my 5kW Gen6 Samsung is nice and quiet, but larger Samsung Gen6 heat pumps have a reputation for being noisy. However, the noise of the heat pump is proportional to flow temperature, there’s a good chance that people who have issues with a noisy heat pump is because it’s running at a high flow temperature. I think the quietest heatpump is the Vaillant Aerotherm+, although the new Samsung HTQ is supposedly pretty quiet. Quieter heatpumps tend to be larger physically since the larger the fan the quieter the noise. My outdoor unit is wall-mounted so having a compact unit was a big plus.
This is totally understandable, a cold customer is not a happy customer! However, I think this is the main reason that heatpumps are commissioned to run a higher than needed flow temperature. The over sizing of the heatpump is maybe due to incorrect assumptions made during the heatloss calculation, or not doing a heatloss calculation at all and sizing the heatpump based on EPC data, this is actually pretty common! 
The single thing which has the greatest effect on heat pump performance is flow temperature, if you want to maximise SCOP the focus should be on getting the flow temperature as low as possible, ideally under 40C at design temperature. Getting a low temperature will probably require radiator and possibly pipework upgrades but will be worthwhile long term.
From my experience and looking at monitoring data, flow temperature and control strategy has a much greater effect on performance than heat pump than sizing and cycling. I can think of examples where an oversized heatpump can get good performance even while cycling because it’s running at a low flow temperate with optimised control setup.
@johncantor has made a good video on the topic of heatpump cycling
Just to pick up on a few points. As of many years, compressors are scroll or rotary now, they almost never wear out or ‘degrade’. The old reciprocating (piston) type had reed valves that suffered wear & tear, so performance could drop. Refrigerant is very unlikely to ‘slow leak’ from a factory-filled system. Water condensation and defrost tends to keep heat-exchangers quite clean. Water heat-exchanger should be fine, but I guess there is a possibility of some blockage over time, and this could reduce the COP, but not really drop the heat output. It would be hard to tell performance without a heat meter, but input power and refrigerant discharge temperature might give some idea.
Its a common notion that a small heat pump can ‘run hard’ as a small car might. However, they are quite different to cars. My old 5kW Ecodan was fitted (in 2008) alongside a 30kW boiler. It ran for months mid-winter without stop (apart from defrosts) at around 35C. No ‘straining’, just a lot of steady-state running
just wanted to comment on this. as I have an evohome system and a self installed Samsung16kw. When we were on Gas, I made heavy use of the evohome system to run the house with “the least possible gas energy input I can get away with” consistent with “just about acceptable comfort in the rooms we were using”. Letting unused rooms run cool. This gave us a gas consumption significantly below the EPC (14kkwh vs 23kkwh, 210m2 house) at the cost of always wearing two jumpers (apart from our 1 cosy, the evening TV room)
Now that I have an ASHP (but only since March), I removed the TRV heads from all of downstairs as part of the commissioning. With the TRV heads in place there were a number of issues (which I was expecting, to be fair): 1. when rads partially shut down, not a large enough effective total emitter area available to the heat pump thus cycling , can’t sustain needed output; 2. not enough flow rate to meet the rate need to sustain the heat output required at the DT thus again cycling, can’t sustain needed output; 3. Loud Harmonic noise (like your pipework becomes a church organ) due to reed effect of nearly-closed TRV with water being pushed through it at at least 4x higher flow rate and several Meters higher head than previously with the gas systems circulating pump.
solution is all TRVs removed downstairs, balancing with lockshields only. With tweaking of flow temp (tweaking WeatherComp curve until you get a stable comfort level), and doors left mostly open (they were mostly closed before) rooms don’t overheat. Effective total comfort level of the house became much higher because all rooms are heated to 20-21C. Once the ASHP in, we didn’t wear jumpers in march or april inside, even in the cold snaps. A side effect is that bedrooms end up warmer (19C ish) than we were used to (we ran them at 15C usually) purely by the heat coming up from all rooms constantly heated downstairs. I had trouble sleeping for first month due to the warmth.
Total annual heat input into the house with the ASHP will I have no doubt be higher than the best I did on gas with the evohome system fully turning down unused rooms. Total energy input will I expect course be lower due to COP, but I don’t have the data for that yet.
Dear all, some really really valuable experience being shared here, such a fantastic community of helpful, interested and well informed people! Thanks for sharing that video by @johncantor @glyn.hudson, and thank you for making it John!
Trying to summarise:
In terms of tracking performance then these 2 charts are a good way to track current performance. In my case I am planning some minor upgrades to my radiators. So looking at these graphs once the next heating season starts should reveal whether there has been any noticeable change in the heat output and COP etc from our ASHP resulting from those changes.
Sorry, a bit slow responding. So, a 10kW unit running at 50% should be considerably quieter than a 5kW at 100%. Re size. Ideally, the unit will operate between say 100% and say 30%, with quite a lot of operation around 60-70% where COP should be best (though only marginally best). An oversize unit will spend more time at lower speeds, and more time off/cycling. It may tend to ‘rev up’ when you dont really want it too. Again, efficiency differences not great
Hi All
This is my first post here so…HELLO.ooooo
I have collected some data from the lovely open energy files to compare power usage on the coldest day of the year, December 14 2022.
Taking the coldest 24 hours energy use, dividing by 24 and comparing that figure with the capacity of the heat pump on that day. I have learnt that Heat Pump capacity changes depending on ambient temp and also flow temp. So the capacity figures have been adjusted.
Hope it is of interest
Yours
Andrew Cunningham @suburbanpirate
Larger units with large evaporator coils with have advantages in our moist winters.
I’m currently running an old Samsung 16kw R410a Unit, which is way over sized for my property.
But I runs and modulates smoother than my Newer Grant 10kw Unit R32, this is a really small unit and I found it suffered with frequent defrost cycles over the winter.
Our system is fully open loop, rads and UFH, no buffer/hydronic separation.
Hello Andrew, Ken, great to see you both on here! and thanks for posting this up here Andrew.
One question I have in my mind at the moment is what is the actual capacity of each heat pump during regular defrost conditions? I think it might be lower than some? manufacturers are stating on the datasheets, mostly going on the Daikin I mentioned, I may be wrong about this, It would be interesting to try and excercise next winter during near zero conditions to try and max out our heat pumps over a 4 hour period to see how close to datasheet output they can achieve? This is clearly not an issue for the Vaillants on there which have datasheet capacities above their badge.
Building on the earlier table above, I thought I would try finding the maximum output of each heat pump on the list over a 4 hour period where outside temperatures were below 1C. The script that I’ve put together to generate the following can easily be modified to change the period to e.g 2 hours, or filter for specific outside temperatures, so happy to re-run this with different parameters, here are the results:
| location | heatpump output | heatpump model | heat loss survey | heat demand | div 2.9 | Max output below 1C | Outside temp | % of heatpump output | % of heat loss |
|---|---|---|---|---|---|---|---|---|---|
| Oxford | 3.5 | Vaillant Arotherm+ | 3 | 7500 | 2.6 | 5.3 | -7.5 | 151% | 177% |
| Heat Geek Towers, Camberley | 5 | Vaillant Arotherm+ | 5 | 6000 | 2.1 | 5.2 | -2 | 104% | 104% |
| Sheffield | 5 | Vaillant Arotherm+ | 5.4 | 9000 | 3.1 | 5.0 | 0.5 | 100% | 93% |
| North Hertfordshire | 5 | Vaillant Arotherm+ | 6 | 8000 | 2.8 | 4.8 | -9 | 96% | 80% |
| Machynlleth, Wales | 6 | Mitsubishi Ecodan | 1.5 | 7500 | 2.6 | 5.7 | 0.5 | 95% | 380% |
| Cambridge | 8 | Samsung HTQ | 7.8 | 10897 | 3.8 | 7.0 | -2.5 | 88% | 90% |
| Fife, Scotland | 7 | Vaillant Arotherm+ | 0 | 23500 | 8.1 | 6.1 | -7.5 | 87% | |
| Stratford-upon-Avon | 7 | Vaillant Arotherm+ | 7.57 | 7300 | 2.5 | 6.0 | -2 | 86% | 79% |
| Banbury | 7 | Vaillant Arotherm+ | 6.5 | 0 | 0 | 6.0 | -2 | 86% | 92% |
| Llanberis, Gwynedd | 5 | Samsung Gen 6 | 3.5 | 8000 | 2.8 | 4.1 | -1 | 82% | 117% |
| St Albans | 5 | Vaillant Arotherm+ | 4.1 | 9143 | 3.2 | 4.1 | 0.5 | 82% | 100% |
| Nottingham | 12 | Midea | 8 | 11332 | 3.9 | 9.8 | -0.5 | 82% | 123% |
| Basingstoke | 10.6 | Daikin 11kw monobloc | 9.8 | 21000 | 7.2 | 8.1 | 0.5 | 76% | 82% |
| West Yorkshire Peak District | 8.5 | Mitsubishi Ecodan | 8 | 15452 | 5.3 | 6.3 | -1.5 | 74% | 79% |
| Malton, North Yorkshire | 14 | Mitsubishi Ecodan | 14 | 41699 | 14.4 | 10.3 | -4 | 74% | 74% |
| Gloucestershire | 5 | Mitsubishi Ecodan | 3.5 | 7200 | 2.5 | 3.6 | 0.5 | 72% | 103% |
| Weston-super-Mare | 7 | Daikin EDLQ07CV3 | 0 | 9524 | 3.3 | 5.0 | -1 | 71% | |
| Llanberis, Gwynedd | 5 | Mitsubishi Ecodan | 4.5 | 8000 | 2.8 | 3.3 | -3.5 | 66% | 73% |
| Manchester | 6 | NIBE F2040-6 | 3.5 | 7000 | 2.4 | 3.8 | -4 | 63% | 109% |
| Weymouth | 14 | Mitsubishi Zubadan | 0 | 27328 | 9.4 | 8.8 | 0 | 63% | |
| Manchester | 5 | Mitsubishi Ecodan | 3.4 | 9500 | 3.3 | 3.1 | -0.5 | 62% | 91% |
| North Yorkshire | 6 | Grant Aerona3 | 6.25 | 15500 | 5.3 | 3.7 | -5 | 62% | 59% |
| Banbury | 8.2 | Vaillant Arotherm+ | 7.6 | 0 | 0 | 4.8 | 0.5 | 59% | 63% |
| North Yorkshire | 11 | Mitsubishi Ecodan | 9.4 | 22000 | 7.6 | 6.3 | -2 | 57% | 67% |
| Kidderminster | 7 | Acond PRO-N | 37 | 6570 | 2.3 | 3.9 | -1.5 | 56% | 11% |
| Bangor, Wales | 11 | Mitsubishi Ecodan | 5.5 | 11000 | 3.8 | 6.0 | 0.5 | 55% | 109% |
| Newbury | 14 | Mitsubishi Ecodan | 11.5 | 12000 | 4.1 | 7.5 | -2.5 | 54% | 65% |
| Sheffield | 11.2 | Mitsubishi Ecodan | 9.989 | 23594 | 8.1 | 5.8 | -1 | 52% | 58% |
| Derby | 8 | NIBE F1145 | 0 | 4542 | 1.6 | 4.1 | -1.5 | 51% | |
| Sofia, Bulgaria | 8 | Atlantic Alfea Extensa | 5 | 12800 | 4.4 | 4.1 | -8 | 51% | 82% |
| Swansea | 11.2 | Mitsubishi Ecodan | 0 | 30000 | 10.3 | 5.4 | 0.5 | 48% | |
| Fleet, Hampshire | 12.12 | Daikin Altherma | 0 | 12700 | 4.4 | 4.5 | 0 | 37% |
Plot of max heat output over 4 hour period as a proportion of badge capacity (outside temp below 1C)
I should overlay this with the max heat output over a 24 hour period to see how it changes…
I also extracted the maximum, attempted so far? heat output over a temperature range for a number of the heat pumps on the list.
My Ecodan, 5kW:
Here’s @glyn.hudson’s samsung 5kW:
Yorkshire, Ecodan, 11kW:
Manchester, Nibe, 6kW:
Heat Geek, vaillant, 5kW:
Daikin 11kW (This is the one I know is struggling to do much above 7.5-8 kW below zero), heat loss of the house is 9.8 kW based on previous years 24h gas use.
Brilliant stuff Trystan.
Interesting to see how some are clearly not producing as much heat as you might expect at low temps.
The Daikin 11KW clearly maxes out at about 0C.
I wonder how defrosts are affecting temps? Also maybe a WC curve too low? As you say it’s hard to know if the heat pumps could have done more.
Room temps is something I would have added which would be a measure of ‘customer satisfaction’ but I found limited data on that. Maybe you could overlay room temp data?
I would say the value of these comparisons would be to see at a glance how likely it is for a heat pump install to keep you warm in winter but at the same time the chances of being hugely oversized.
Andrew
Thanks Andrew,
Yes agreed, I will definitely try a number of max output tests next winter 
Some folk are monitoring room temp, so including that for those that have it should be possible.
Perhaps also a drop down selector in the form? to gauge perception of comfort and capacity of heat pump to reach higher temps? Given that people have such different comfort levels?
Agreed that customer satisfaction is probably more important than exact room temps achieved.
People will have such different needs.
Some will really want and need at least 21 C all day and night.
While others might relish putting on 2 jerseys at 17C and toughing it out.
Also some properties might be packed with people and cooking and computers.
Hello
Hi Ken
Im interested in your take on this.
Andrew
As long as the refrigerant remains and the compressor has had a good life, ( ie not short cycling, or High pressure issues due to poor heating design ) then should be good for over 20 years.
My current Samsung is already 12 years old and second hand.
