I had very helpful advice from several forum members back in January, however still experience heat pump short cycling, the background is as follows.
2024 new build house in central Scotland. 260 sqm floorspace, well insulated and triple glazed with 36 sqm of south facing glass, so lots of solar gain to contend with. Heat loss of 8.6kW at -4 degrees from the SAP design calculations, which I now believe may be overstated by 1.5 to 2 kW after I adjust for the very large ventilation losses that were assumed.
Grant Aerona 17R32 heat pump (nominal 17kW, but I believe a bit lower in practice), Grant LLH, UFH to ground floor with seven zones, 4 rads. to first floor, electric UFH to bathrooms. The heating system was installed by my builder who subcontracted the heat pump installation, connection and commissioning to a local Grant registered engineer. In my opinion the commissioning was poor and left setup as for a conventional boiler. (45 deg. C flow temp to rads, 30 deg. C to UFH, whole system on/off three times per day)
Our first experience of cold weather operation was in December 2024 when high electricity consumption started my self-education journey into heat pump setup and optimisation. I found that the return to the HP had no flow regulator fitted and with advice from the forum, decided this needed correcting as a key first step to improving control of the system. This was also confirmed by Grant technical support.
The Grant installation engineer said the flow regulator "would not make any difference as they all cycle on and off anyway” and after waiting three months and him still not doing the work my builder used an alternative (Daikin) engineer to fit it in May 2025. By then the weather was not ideal for testing, but with the flow regulated to around 18 L/min the HP modulated down to 600W at a fixed flow temp. of 31 degrees and ran for a two-hour test without cycling at OAT 10 degrees. After several similar results I was pleased and waited for winter to start setting the weather compensation.
In spite of this good result, by this time I had learnt that my HP is probably significantly oversized, and that the design of my heating system is not really to best practice. But without significant changes I will probably need to accept some compromises in its operation.
My (compromise) thinking and initial target is to run at 29-33 deg. C flow temp. controlled using w/comp. for the radiators as that seems to be sufficient to maintain first floor temperature, and blended down to say 27 deg. C for the UFH. If possible, with room stats set at high level, just to protect against overheating when we have winter solar gain.
With consistently low temperatures in the last week, I have run more tests with the following observations. (I disabled W/Comp to avoid any variable flow temp. targets during the tests).
At OAT of 6 deg. C and above the HP modulates down to 600W at my 31 deg. C flow setpoint and will run continuously without cycling, (as in May 2025 tests) and even with only half of the zones calling for heat.
At below 5 deg. C and all zones calling for heat the HP short cycles, best cases after 12-15 minutes, worst cases after only 6-7 minutes. I have the same result at 0 deg. OAT
On one test at 13:00 hrs, 4 deg. OAT in the shade, the HP in sunshine with its thermometer reading 7 deg. OAT it modulated down to 600W and ran without cycling.
I am surprised (and confused!) by these results as they appear to go against what I have understood about short cycling in warmer v colder temperatures.
I understand that at lower OAT a HP will consume more energy to achieve a given flow temperature, and that I am probably trying to run the heat pump at the bottom end of its operating window. But if my HP will modulate down to 600W at 10 deg. OAT, why will it not modulate below 1500W at 5 deg. C or lower? It continues to run at 1500-1800W, may finally modulate to 1200W for around 30 seconds but the flow temp. rises 2-3 degrees above the set point within a few minutes and it then stops.
It’s as though the control system is saying “it’s really cold now, so work extra hard and never drop below 1500W”. I can see no parameters in the Chofu manual that would cause this behaviour.
My thoughts are to take this low OAT short cycling issue to Grant technical support, but I would value some guidance from the forum before starting that discussion.
My questions are:
Please could anyone explain what may be causing short cycling when OAT drops and if there might be a solution?
Please could anyone with a Grant HP please share if they have experienced similar? (Particularly if the HP is oversized).
Are my thoughts on flow temps. for the radiators and UFH reasonable or might there be a better way? For example, would it be feasible / better to run all the UFH zones closer to an “open loop” at the same slightly higher temp. as the first-floor radiators (29-33 deg.) if I can get the individual room circuits better balanced/restricted to avoid overheating?
I have a Grant 10kW (nominal 9kW for the UK), which is, I believe correctly sized, but this is a retrofit on a 70s converted bungalow, with radiators, so my heat demands are greater than yours, so not comparable set-ups, so I'm not sure I can help you, but here you go...
This is what I see at various, recent, outdoor conditions (explanation below each graph):
A mild day (OAT around 10C) with "slow and steady" demand, as HPs are happy running. No cycling at all, typical using 700W.
A slightly colder day, around 5C, but again, no cycling, just slightly more energy consumed to hit the higher LWT required.
A colder day, just above freezing and the (four) 'dips' in energy usage are caused by defrost cycles. As the day warms, heat demand decreases, and the HP turns off as temperatures are reached. There are then short periods of heating, what I would call typical "boiler cycling" as the temp drops below target, and then goes back above after a 15 minutes from the HP. Not HP "short cycling" just, bursts of heat to keep temps up.
This morning - emerging from a sub-zero night, and there are defrost cycles every 40 odd minutes, as HP struggles to hit target LWT. This is typical for my HP in very cold nights, and tends to recover once the daylight arrives.
(I hope the graph labels are clear - the two 'Outdoor temperature' readings are the HP one; and another external thermostat on the opposite side of the house, facing north. The HP is on the south side, so is affected by sun glare off the rear wall; the grey line is 'Target LWT')
------
As I mentioned earlier, with radiators and older house, my required LWT will be higher than yours, hence higher energy usage for 'comfort', but I do not experience short-cycling. My observations have been that the Grant is very good at modulating down to hit and stick to the target LWT, but then I feel my HP finds that "easier" as it's sized correctly for the heat demand of the house.
My only observations would be:
- how do you know you're short-cycling and not just performing defrosts in moist, cold, UK (I assume) climate?
- the only parameter that might help is "21 41" (Hysteresis of water set point) which by default is 8C and I've reduced to 4C, and might mean the HP turns back on again sooner (i.e. after a drop of only 4C, rather than 8C), but prizes for anyone following the logic in Section 8.1 of the Installation Guide?
@mikefl, thanks for sharing your operating data, very interesting.
For the power consumption, are you using the compressor frequency values that I view on function 01 of the user monitoring section of the "remote" controller? I am interested to compare the your consumtion across the temperature range with mine.
Yesterday morning we had -2 to 0 deg. OAT for a while and the HP was consuming c 3500W to achieve 32 deg. flow temp which seemed disproportionate to the 1800W I was seeing at 3-4 deg. OAT. I have had the flow rate reduced down to c 18L/min as I understood this was needed to try to reduce short cycling, considering my likely oversized unit. Thinking the HP was possibly working harder than necessary to reach flow temperature due to this restriction I opened up the flow rate to 30 L/min and it soon modulated back to 1200W and later to 700-800W for around 45 minutes, which is closer to the performance I have been expecting. Hopefully this change will be a turning point!
OAT is 12 deg now, so I will monitor further when it drops again.
To your valid question, I am not confusing the defrost cycles with my short cycling problem. Mine defrosts hourly once in that temperature zone and I am aware of it when it occurs.
My water hysteresis has been at 3 deg, I think increasing by a few degrees will only extend the "off" periods by a couple of minutes as it tends to fall quite quickly when the compressor stops, and has already fallen below the lower limit for restarting before the unit reaches it's 3 minutes of minimum off time, which is not adjustable to my understanding. The drop in radiator temperature is noticable in the 3 minute off periods when the flow temp. drops from say 32 to 26 degrees and there will also be a slug of cooler water working through the UFH that can't be helpful. I don't want to increase that cooling effect and clearly if I can stop the short cycling the issue is removed.
@steam-powered Hi - for 'power consumption' I use the '01 06' value (from the controller), which is in units of 100W, but most of the figures presented on the controller aren't very fine (i.e. all temps are giving in multiples of 0.5C, which is too coarse to be that useful). Although as you probably notice 'power' is directly related to 'compressor frequency' in most instances.
@damonc I was refering to the 'standard' energy sticker that comes with the HP which I think indicates (for my 10kW Grant) that output rating varies by location, and I think I'm in a '9kW' location whereas someone in St Ives or Lisbon would be getting 11kW (for CH):
I assume similar stickers come with all HPs (like they do with all white goods in the UK) but it's the sort of thing an installer usually throws away or leaves tucked in pages of manuals, so I guess I was lucky that my installer stuck this on my LLH.
@steam-powered I'd suggest dropping the cold weather end of your curve down, what are your curve setting now? It sounds like your ASHP is oversized and so it's just hitting the curve temp quickly because it's producing a lot of heat then cycling off until it hits the hysteresis temp. Try dropping the cold weather flow temp until it's more in balance between your heat loss and the output of the ashp it should cycle much less as it will take a lot longer to get to your target flow temp albeit will still do a defrost cycle from time to time.
@damonc, my current weather comp. settings are 32 @ -4 , 26 @ 20. At 5 - 7 deg OAT for the last day the house is generally at 20-21 deg and only calling for heat intermittantly, so I agree it should be set lower.
The other constraint I have is that the first floor rads. are sized for c 40 degrees flow and they are just keeping the colder north facing bedroom to around 18-19 with 30 degrees flow temperature, although with the HP cycling the rads. are cycling between 26 and 30 degrees.
I have followed your advice and dropped to 30 @ -4 and will monitor.
Do you think at lower flow temps. reducing the flow rate at the HP return regulator would then help? I had it down to 18-20 L/min (Chofu say the minimum is 15L) but currently have it fully open at 32 L/min. I seem to have less cycling at the higher flow rate, which does not make sense to me. When it was down at 18L the compressor would sometimes work really hard for a short time (3kW consumption) and then overshoot and cycle off. 32L/min appears to avoid that issue.
Does anyone sucessfully run a Grant HP at <30 deg. flow temp. without any cycling? There does not appear to be any published Grant or Chofu performance data that advises the minimum stable operating conditions that are possible.
@steam-powered I'm currently running 30@-7 23@15 so currently it's target temp is 26 at OAT of 5.5 and it's running steadily at 700watts. That's a really lean curve for me as I'm experimenting a bit and so I may need to nudge it up a bit. Counter intuitively (for me) I'm now trying a lower curve target at the cold end and think it may help reduce cold weather cycling as when I had it at a higher setting (35@-7) the pump was powering it up to the target temp much faster than the house was losing heat and then turning off (not defrosting as far as I could see) for until the temp dropped back, so cycling about 1.5 times per hour. I'll see if having the heat creation building up more slowly in the house and being closer aligned to the heat loss gives me a longer run time. I also upped my hysteresis to 6 from 4 so I've changed a few things and am waiting for it to settle out. The hysteresis change and mild temps yesterday meant the flow took a long time to cool down to meet the new curve setting overnight and so the house had cooled but it kicked back in at about 0330 and the inside temp has stabilised at about 19.5 and I think it will now start building back up, I'll see where it settles this evening but the pump is probably x2 oversized and can only modulate down to 600w so if the heat it produces at that input level comes up too far it'll cycle in warm weather but hopefully less.Will have to see. A higher flow rate may help reduce cycling by robbing the primary circuit of heat so it's not bumping up against the target all the time, my compressor only ever reaches 3kw and above when it's doing hot water
@damonc, thanks for sharing your settings. It's great to know that you are achieving what I aim to do with the same Grant HP.
I ran trials over the last couple of days with the following observations.
1. At OAT of 6 deg and above my HP will run at 700W and flow temps. of 28 to 31 deg very consistantly without any short cycling, even with only a few of my zones calling for heat.
2. As soon as the OAT moves to 5 deg or lower the modulation is quite erratic and in the 28 to 32 deg flow range it will ramp up to 1500-1800W, overshoot the target temp and short cycle every 8-10 minutes.
3. With OAT in the 2 to 3 deg. range I did tests with constant fully open system at various flow temps up to 38 deg (to see if my lower flow temp. expectation may be the issue) and had erratic behaviour, examples being:
a. 2400W @ 38 deg for 20 mins,
b. 1600W @ 37 deg for 20 mins,
c. 1600W @ 36 deg for 20 mins,
d. 2800W @ 35 deg for 20 mins,
e. 3000W @ 34 deg for 15 mins (then I stopped that test as it showed no attempt to modulate!)
f. 2400W @ 33 deg for 25 mins,
I realise these are relatively short time periods, but my short cycling occurs within 8-10 mins, so I am happy that a 20 min run indicates more stability than I normally see.
4. I had a period where the OAT was moving up and down between 5 and 6 deg. At 30 deg flow it was stable at 700W with 6 deg OAT but as soon as OAT dropped to 5 deg it ramped up to 1300W, the flow temp increased to 32 deg.
5. At 6 deg OAT, 31 deg flow & 700W, I increased flow to 33 deg and it slowly modulated up to 1400W but did not overshoot the set point, I increased to 35 deg and it slowly modulated up to 2000W, did not overshoot and modulated back to 1600W. I consider this to be expected performance.
My conclusion is that at 6 deg OAT and above it modulates perfectly well for flow temps 28 to 35 deg. At 5 deg OAT and below the modulation is generally poor, running at high power consumption, but with two examples (b & c ) of more reasonable running.
Can you please tell me if you notice any change in your HP's modulation when OAT drops to 5 deg and below, (or any other temperature change point), or do you still manage to run at 700W between say 6 and 3 deg OAT with only gradual ramp up as OAT drops?
Additionally also I find the correlation between compressor frequency and power consumption to be erratic. For example at 30 Hz I can have power ranging from 1200W to 2000W. Do you notice anything similar with your HP please?
I am wondering if this factor may indicate some control issue in my HP?
My next step is to share this data with Grant technical support as I believe there may be a fault in the control system of my HP. If you can advise if you see anything similar with your HP that would be helpful background for me.
@steam-powered 17kw heat pump for a 8.4kw heat loss is bonkers! I am in a 1970s bungalow with an over calculated heat loss of 11kw (I think it should be 8kw when you move from 2 ACH to a more realistic 0.9). They installed a 13kw Aerona3 and it is definitely oversized. Mine modulates every twenty five minutes or so, but my power usage isn’t too bad usually around 700-1200w but yeah it would be nice to have it just pootle along at an even pace. I am currently experimenting with night mode i.e. restricting the compressor frequency but it rarely goes much over 50hz. I might look at extending my cold end of the WC but one set of tinkering at a time!
For background I use home assistant connected to the ASHP via modbus so I can see everything in real time. I am also completely on rads so no UFH.
@steam-powered 17kw heat pump for a 8.4kw heat loss is bonkers! I am in a 1970s bungalow with an over calculated heat loss of 11kw (I think it should be 8kw when you move from 2 ACH to a more realistic 0.9). They installed a 13kw Aerona3 and it is definitely oversized. Mine modulates every twenty five minutes or so, but my power usage isn’t too bad usually around 700-1200w but yeah it would be nice to have it just pootle along at an even pace. I am currently experimenting with night mode i.e. restricting the compressor frequency but it rarely goes much over 50hz. I might look at extending my cold end of the WC but one set of tinkering at a time!
For background I use home assistant connected to the ASHP via modbus so I can see everything in real time. I am also completely on rads so no UFH.
I am so envious, your are so fortunate to have all of the data at your finger tips!!
2024 build bungalow
Southern england 179 m2 High level of insulation Underfloor heating All 12 circuits are fully open all the time 1 thermostat in family room 7KW heat pump 50 litre buffer tank (4 port)
3.6KW solar panels Energy used by heating 2527 KWh - 7527 KWh (SCOP 3.5 approx)
