Heat Pump Truth or Myth#2 - Cycling uses a lot of power and ruins your COP
Heat pump cycling is one of these topics where a lot of people seem to have an opinion but no actual data to back it up. Cycling is blamed a lot for poor ASHP performance and it seems to be generally accepted that cycling or even <gasp> short-cycling is a “Really Bad Thing”. I do have data so I decided to have a look at it.
This is a graph of my 14kW Ecodan over a few hours on a recent mild winter’s day.
After the morning startup, it runs for an hour or so without cycling. As the ambient rises it starts to cycle. And because cycling is a Really Bad Thing, it's now going to use more power and my COP will be really poor, right? Let’s see.
This is a comparison of two periods after the initial warm up; not cycling (5-6am) and cycling (6am-1pm)
time |
energy consumed Wh |
energy delivered Wh |
Ave Ambient deg |
average power consumed W |
average power delivered W |
ave flow temp |
COP |
5-6am |
1082 |
4246 |
8.9 |
1082 |
4246 |
36.1 |
3.9 |
6am-5pm |
6112 |
27578 |
10.3 |
777 |
3506 |
35.9 |
4.5 |
During the cycling period, the COP is better and the power consumed is less. Surely that can’t be right? That’s not meant to happen; the power used should be much more and the COP should have plummeted.
So what happened? Well what you’d expect really. The COP is better because it’s warmer outside and the average flow temp is nearly the same. The average power consumed is less because the heat demand is less and the COP is better. The bit that might be surprising, is that cycling hasn’t made a big difference to consumption or efficiency. Why not? I think it’s because the widely touted theory that starting up an ASHP and warming up the flow a bit uses an excessive amount of energy is a load of cobblers. There are no large peaks in power. And, at the risk of stating the bleeding obvious, between peaks there are troughs, empty ones.
I am sure the cycling does make some difference and a smaller ASHP delivering 3.5kW without cycling will do it a bit more efficiently. If I look at the next size of ASHP down from mine, it can modulate down to around 3.5kW and it looks to have a 10-15% better COP at this output than I got with my cycling. Let’s say 15%. That means that I’d be saving around 120W of power, or about £1 per day over 24 hours at today’s prices.
I know there is more to it than this and there are a lot more extreme examples of cycling out there but I think cycling sometimes gets a bit of a bum rap for being the cause of all heat pumps’ ills. And I’d rather have a heat pump that cycles a bit when it’s warmer and cheaper to run than be inside shivering when it’s freezing outside, the ASHP is defrosting every hour and running flat out trying to match the number on its badge.
I’d be interested if anyone else has done any similar analyses. Or even just has an opinion. 😉
Hi Kev,
Excellent data once more.
I am not certain where the claim that heat pumps use more energy and are less efficient when they cycle originated.
Maybe someone had noticed that their heat pump was cycling (defrosting) during cold weather, and was also using lots of power, so put two and two together.
Maybe the heat pump was operating with a fixed LWT of 55C or even 60C, so hence was running inefficiently when it was operating.
As you correctly stated, as the outside air temperature increases, your home's heat loss reduces, so your heat pump does not need to produce as much heat energy. The warmer outside air also contains more energy for the heat pump to extract, which in turn should improve efficiency, though the fact that the heat pump may be using more electrical energy when it first starts running may have a negative effect.
The fact that your system was operating under correctly adjusted weather compensation may also be of importance. When your heat pump starts running it quickly achieves the calculated LWT, so the compressor is not working hard and soon stabilises and operates in an efficient manner. If the calculated LWT was at a much higher value, it would take longer to stabilise which may prove less efficient. Without carrying out very detailed testing, it is difficult to state for certain.
I believe one of the reasons why 'short' cycling may be of concern is more to do with correct lubrication of the compressor when it is running repeatedly for short periods of time. Think of the additional wear that may be experienced by a car engine if it is repeatedly stopped and started without the lubricating oil getting up to temperature. I remember reading that one of the manufacturers stated that starting and stopping the heat pump more than 6 times in a 1 hour period was to be avoided.
Posted by: @kev-mThere are no large peaks in power.
Am I misreading your graph? It looks like "consumed" hits peaks every time it starts up.
Posted by: @kev-mAnd, at the risk of stating the bleeding obvious, between peaks there are troughs, empty ones.
Of course, so for cycling to be more efficient than not, the area of the trough would need to exceed the area under the peak above the steady state. Yours does not, so cycling is using more power than if you could avoid it, or at least lengthen the troughs to be bigger than the peaks. One of the frustrations with current heat pump programming is that most will do short off cycles up to a frankly absurd level of 6 an hour and make no attempt to lengthen the "off" periods on the third or later repeat. Although you may be still losing a small bit of efficiency compared with the ideal situation (pump modulates to a low enough output), your graph is nowhere near that bad an example of cycling.
Sadly, I've lost my data from the spring to a software/hardware failure (SD cards are not good backup devices... it now backs up to a USB disk) and didn't suffer any particularly bad cycling this autumn for various reasons (including the long mild spell and none of the control systems I tried were particularly vulnerable to cycling, reacting to the pump shutting down). In winter, the pump is working hard enough that short-cycling is rare. In summer, it's working so little that it's also rare.
Posted by: @mjrPosted by: @kev-mThere are no large peaks in power.
Am I misreading your graph? It looks like "consumed" hits peaks every time it starts up.
Posted by: @kev-mAnd, at the risk of stating the bleeding obvious, between peaks there are troughs, empty ones.
Of course, so for cycling to be more efficient than not, the area of the trough would need to exceed the area under the peak above the steady state. Yours does not, so cycling is using more power than if you could avoid it, or at least lengthen the troughs to be bigger than the peaks. One of the frustrations with current heat pump programming is that most will do short off cycles up to a frankly absurd level of 6 an hour and make no attempt to lengthen the "off" periods on the third or later repeat. Although you may be still losing a small bit of efficiency compared with the ideal situation (pump modulates to a low enough output), your graph is nowhere near that bad an example of cycling.
Sadly, I've lost my data from the spring to a software/hardware failure (SD cards are not good backup devices... it now backs up to a USB disk) and didn't suffer any particularly bad cycling this autumn for various reasons (including the long mild spell and none of the control systems I tried were particularly vulnerable to cycling, reacting to the pump shutting down). In winter, the pump is working hard enough that short-cycling is rare. In summer, it's working so little that it's also rare.
I remember reading in one of the manuals (don't ask me which) that the heat pump was prevented from restarting for a period of 10 minutes, which would reduce frequent cycling.
During the warmer periods when cycling is more prevalent, setting the system to operate under thermostat control should extend the running and none running periods, which should help reduce the cycling frequency.
Posted by: @mjrPosted by: @kev-mThere are no large peaks in power.
Am I misreading your graph? It looks like "consumed" hits peaks every time it starts up.
Posted by: @kev-mAnd, at the risk of stating the bleeding obvious, between peaks there are troughs, empty ones.
Of course, so for cycling to be more efficient than not, the area of the trough would need to exceed the area under the peak above the steady state. Yours does not, so cycling is using more power than if you could avoid it, or at least lengthen the troughs to be bigger than the peaks. One of the frustrations with current heat pump programming is that most will do short off cycles up to a frankly absurd level of 6 an hour and make no attempt to lengthen the "off" periods on the third or later repeat. Although you may be still losing a small bit of efficiency compared with the ideal situation (pump modulates to a low enough output), your graph is nowhere near that bad an example of cycling.
Sadly, I've lost my data from the spring to a software/hardware failure (SD cards are not good backup devices... it now backs up to a USB disk) and didn't suffer any particularly bad cycling this autumn for various reasons (including the long mild spell and none of the control systems I tried were particularly vulnerable to cycling, reacting to the pump shutting down). In winter, the pump is working hard enough that short-cycling is rare. In summer, it's working so little that it's also rare.
There are peaks but they are small and narrow. They are not using a lot of energy. I didn't say cycling wasn't less efficient; my point is that while cycling is less efficient than not cycling , the degree of inefficiency is (in this case anyway) small.
Posted by: @kev-mI didn't say cycling wasn't less efficient; my point is that while cycling is less efficient than not cycling , the degree of inefficiency is (in this case anyway) small.
And I also say your system is barely cycling in the example, so proves/disproves nothing.
Posted by: @mjrPosted by: @kev-mI didn't say cycling wasn't less efficient; my point is that while cycling is less efficient than not cycling , the degree of inefficiency is (in this case anyway) small.
And I also say your system is barely cycling in the example, so proves/disproves nothing.
Of course I haven’t proved or disproved anything with such an unscientific experiment. I think what I have done is show that when my heat pump has 15 starts in 7 hours, it doesn’t significantly worsen efficiency.
Another way of looking at it is to compare these two hours. They are delivering very similar amounts of heat in similar conditions in the same house, with the same settings. The only difference is that in the second hour the one the ASHP is cycling 3 times per hour (barely?) and in the second it isn’t. That's the most frequent cycling I can find on my ASHP btw.
time |
energy consumed Wh |
energy delivered Wh |
Ave Ambient deg |
average power consumed W |
average power delivered W |
ave flow temp |
COP |
5-6am |
1100 |
4318 |
8.9 |
1082 |
4247 |
8.9 |
3.9 |
6-7am |
1006 |
4429 |
9.0 |
929 |
4088 |
9.0 |
4.4 |
There is an interesting video on youtube from John Cantor (OpenEnergyMonitor and heatpumps.co.uk) driven off data in OpenEnergyMonitor (and presumably some investigations into performance). Effectively shot-cycling in itself is not necessarily too bad for performance (a few percent drop in performance, especially for more modern inverter driven compressors), although in the case of mis-use of a heat pump (timer, turning it off overnight) or incorrect sizing will still cause performance problems. The mention of buffer tanks, as well as a plumbing diagram with a plate heat exchanger, were interesting.

thanks for this, very interesting and (I think) broadly confirms what I said, but with a bit more rigour and credibility!
Old thread but it does address part of my question hence not creating a new one:
Does this mean that oversizing an ASHP isn't necessarily a terrible thing? My installer for some reason ignored my request for an 8.3kW Air-to-Air ASHP ('Aircon') and put in a 12kW Mitsu MXZ-F outdoor unit for 4 fan coil units (2x 2.5, 2x3.5). Of course the math seems to work out (2.5+2.5+3.5+3.5 = 12) but this is a passivhaus so it is super unlikely that all these FCUs would run at full tilt, let alone at the same time. So far I really have only been running one 2.5 fairly regularly, and the ASHP can only go down to 3.6kW....
The unit I was expecting (8.3) has other advantages, such as being meaningfully quieter and a bit more efficient (10% better) but on the downside, it has no option to add another FCU and I might just want to do so in the near future (one un-ACed room gets pretty hot..)
Thoughts? Spec sheet of ASHP attached.
I understand that A2A heat pumps have better turndown than A2W thus sizing is less critical. I haven't seen the proof that this is true however.
4kW peak of solar PV since 2011; EV and a 1930s house which has been partially renovated to improve its efficiency. 7kW Vaillant heat pump.
- 26 Forums
- 2,114 Topics
- 46.4 K Posts
- 85 Online
- 3,410 Members
Latest Posts
-
RE: Replacing Refrigerant on Mitsubishi Heat Pump
@jamespa The pipework was badly installed and even wors...
By DavidAlgarve , 17 minutes ago
-
RE: Compressor and Evaporator Stratification
Thank you. That's an excellent find. As you can probab...
By trebor12345 , 2 hours ago
-
Indeed. With a battery its presumably the battery con...
By JamesPa , 2 hours ago
-
RE: No-code, plug-and-play monitoring for your heat pump
To update an ESP device using ESPHome in Home Assistant...
By Grantmethestrength , 20 hours ago
-
RE: Is this just haze or something more?
I am not a scientist nor a meteorologist and I know the...
By Toodles , 3 days ago
-
RE: Benefits of an extra horizontal loop for GSHP
It will improve performance, but by how much depends on...
By Brendon Uys , 3 days ago
-
RE: The good, the bad and the not that great – my heat pump installation
Said I would pop back with a update. Only Mr Jesus h...
By Burtis , 4 days ago
-
RE: Buffers, hot water and cooling
Thats presumably heat pump DT, what about radiator DT a...
By JamesPa , 4 days ago
-
RE: Who's your electricity provider and what's your tariff?
@chandykris Exactly the same here. We now have 13.5kWh ...
By Old_Scientist , 4 days ago
-
RE: Wood burner with ASHP - direct air?
This depends on the size of your wood burner and size/t...
By Lucia , 5 days ago
-
RE: Difficult positioning of air source heat pump
@rikiarn It's a not a buffer it's a volumiser which is ...
By Lucia , 5 days ago
-
RE: Is your heat pump insured?
To make things clear, lots of insurance companies cover...
By Lucia , 5 days ago
-
RE: Boxt Heat Pump Installation Experience
@adrian any updates since installation? How is it perfo...
By Ashfp , 5 days ago
-
RE: Say hello and introduce yourself
Hi @seatonian - Yes, I know Seaton, and I have the Nat...
By Transparent , 6 days ago
-
@seatonian got your email. Will connect you with a reco...
By Mars , 7 days ago
-
@downfield and @toddles, this isn’t in the settings cur...
By Mars , 1 week ago
-
RE: Homely launched for Grant Aerona ASHP control
I haven't bought the Aerona Smart Controller (£700!) bu...
By Prunus , 1 week ago
-
RE: Antifreeze top up for my heat pump - is this a rip off?
For the record, Grant supply ethylene glycol for their ...
By Prunus , 1 week ago
-
RE: Compressor attempts start but fails - Nibe Fighter 2005 8kW
@skyefarmer amazing. Well done!
By Mars , 1 week ago
-
RE: Air source heat pump roll call – what heat pump brand and model do you have?
Ebac 5kw - Monobloc R32. Developed and made in UK - Mit...
By marcinwloch69@gmail.com , 2 weeks ago