Only had the heatpump since March and you'll have seen I've posted a few queries already as I try to get it fine tuned. The other day with outside temp of 3° I was really pleased to see it running in steady state with a good COP even if took a while returning from setback which I was fine with.
Last night tho with Melcloud screenshot attached, it seemed a bit haywire. It shot up to 35ish flow then cycled off and did this each hour. It led to brilliant COPs of well over 4 but obviously didn't heat the house too well as it was cycled off most the time. It was obviously a fair bit warmer last night so I expected cycling but this seemed extreme especially with the long wait between cycles. I thought if it was because of too warmer outside temps, then the return temp would have got close to the flow temp and so that would have caused the cycling, but there looks a healthy dT of 5 throughout.
I've suspected it's how I set the curve, that it is set to say 30° at that temp and so when it gets to 30° + 5 it switches off as it's exceeding the autoadapt parameters. I had something like 45 at -2 and 20-odd at 10°. So I've changed to 30 at 10°. I've also changed the time.interval to 30 mins, I had at one hour after reading some of the debates on here about autoadapt. It does seem though like the heatpumps incapable of steady state in the low/mid-thirties flow temp which seems disappointing considering that's where I should get some great COP.
By way of background - I have a radiator system with 6kw Ecodan and 5.2kw heatloss at -2. 22mm primary/10mm to rads, no buffer, open system.
Only had the heatpump since March and you'll have seen I've posted a few queries already as I try to get it fine tuned. The other day with outside temp of 3° I was really pleased to see it running in steady state with a good COP even if took a while returning from setback which I was fine with.
Last night tho with Melcloud screenshot attached, it seemed a bit haywire. It shot up to 35ish flow then cycled off and did this each hour. It led to brilliant COPs of well over 4 but obviously didn't heat the house too well as it was cycled off most the time. It was obviously a fair bit warmer last night so I expected cycling but this seemed extreme especially with the long wait between cycles. I thought if it was because of too warmer outside temps, then the return temp would have got close to the flow temp and so that would have caused the cycling, but there looks a healthy dT of 5 throughout.
I've suspected it's how I set the curve, that it is set to say 30° at that temp and so when it gets to 30° + 5 it switches off as it's exceeding the autoadapt parameters. I had something like 45 at -2 and 20-odd at 10°. So I've changed to 30 at 10°. I've also changed the time.interval to 30 mins, I had at one hour after reading some of the debates on here about autoadapt. It does seem though like the heatpumps incapable of steady state in the low/mid-thirties flow temp which seems disappointing considering that's where I should get some great COP.
By way of background - I have a radiator system with 6kw Ecodan and 5.2kw heatloss at -2. 22mm primary/10mm to rads, no buffer, open system.
I think that some time ago when Ecodan owners started using Auto Adapt, they found that when the heat pump was first started up after a shutdown, it raised the LWT higher than required and took some time to settle down.
Can your system control the speed of the water pump and if so what is its setting?
@derek-m no doesn't look like it. Just checked and we're at 14l/m. It was on setting 5 on the FTC, I reduced that down to 3 from setting 5 but it didn't change the pump speed so the pump must operate independently.
Nominal for the 6kw is 17.2//m and range can be between 8.6 and 17.2. I did wonder whether reducing pump speed a little would help?
As a side note - is there an easy way to work out pipe velocity? As I know you're supposed to keep below 1.5m/s for copper
@derek-m no doesn't look like it. Just checked and we're at 14l/m. It was on setting 5 on the FTC, I reduced that down to 3 from setting 5 but it didn't change the pump speed so the pump must operate independently.
Nominal for the 6kw is 17.2//m and range can be between 8.6 and 17.2. I did wonder whether reducing pump speed a little would help?
As a side note - is there an easy way to work out pipe velocity? As I know you're supposed to keep below 1.5m/s for copper
You could try manually changing the speed on the water pump itself and see what effect that has.
There are various water velocity calculators that can be found online, or if you calculate the internal volume of a 1 metre length of pipe, then divide the number of litres by this value, it will give the flow velocity. Don't forget that flow rate is normally measured in litres per minute whilst velocity is normally given as meters per second.
and so when it gets to 30° + 5 it switches off as it's exceeding the autoadapt parameters. I had something like 45 at -2 and 20-odd at 10°. So I've changed to 30 at 10°. I've also changed the time.interval to 30 mins, I had at one hour after reading some of the debates on here about autoadapt. It does seem though like the heatpumps incapable of steady state in the low/mid-thirties flow temp which seems disappointing considering that's where I should get some great COP.
By way of background - I have a radiator system with 6kw Ecodan and 5.2kw heatloss at -2. 22mm primary/10mm to rads, no buffer, open system.
Whats the dt50 value of all your radiators? Also, when you observed the cycling how much of your rad system where closed eg. Thermostatic controlled? Was the above graph created while operating in AA Or was it WC?
apart from your set parameters there is a minimum operating load which might be below the output target 🎯 you may be surprised how low the heat loss will be through the rads at 30c flow temperature esp if the room temp is almost reached.
@derek-m no doesn't look like it. Just checked and we're at 14l/m. It was on setting 5 on the FTC, I reduced that down to 3 from setting 5 but it didn't change the pump speed so the pump must operate independently.
Nominal for the 6kw is 17.2//m and range can be between 8.6 and 17.2. I did wonder whether reducing pump speed a little would help?
As a side note - is there an easy way to work out pipe velocity? As I know you're supposed to keep below 1.5m/s for copper
One thing that I would suggest is that you switch off AA mode for a while and optimise the WC curve, since my understanding is that in AA mode the controller still uses the WC curve, and fine tunes the LWT with reference to the room temperature. If the WC curve is not correct, then the controller may struggle when in AA mode.
and so when it gets to 30° + 5 it switches off as it's exceeding the autoadapt parameters. I had something like 45 at -2 and 20-odd at 10°. So I've changed to 30 at 10°. I've also changed the time.interval to 30 mins, I had at one hour after reading some of the debates on here about autoadapt. It does seem though like the heatpumps incapable of steady state in the low/mid-thirties flow temp which seems disappointing considering that's where I should get some great COP.
By way of background - I have a radiator system with 6kw Ecodan and 5.2kw heatloss at -2. 22mm primary/10mm to rads, no buffer, open system.
Whats the dt50 value of all your radiators? Also, when you observed the cycling how much of your rad system where closed eg. Thermostatic controlled? Was the above graph created while operating in AA Or was it WC?
apart from your set parameters there is a minimum operating load which might be below the output target 🎯 you may be surprised how low the heat loss will be through the rads at 30c flow temperature esp if the room temp is almost reached.
Open system, no TRVs (no volumiser or buffer either). It was in AA mode. I spoke to an engineer at Ecodan on the helpline today, didn't think it helped much - just stressed was about the ambient temps but the flow temp peak has been pretty similar lately whether at 7° or 11° outside. What I really wanted to get an idea of was whether this was cycling behaviour (can't modulate low enough) or AA behaviour (algorithms and settings). He seemed to indicate it could be AA behaviour to send it to a peak flow temp before immediately cycling down as a way of retaining low flow temps I guess. I'd have expected it to get to steady state.
It's been taking quite a while to get up to temperature from setback, I reckon because of the cycles. However I guess looking at today for example, even with the cycles it's putting in 2-3kwh of heat per hour which I guess would hint at hitting the minimum for the 6kw heatpump (modulating down to 1/3).
@derek-m no doesn't look like it. Just checked and we're at 14l/m. It was on setting 5 on the FTC, I reduced that down to 3 from setting 5 but it didn't change the pump speed so the pump must operate independently.
Nominal for the 6kw is 17.2//m and range can be between 8.6 and 17.2. I did wonder whether reducing pump speed a little would help?
As a side note - is there an easy way to work out pipe velocity? As I know you're supposed to keep below 1.5m/s for copper
One thing that I would suggest is that you switch off AA mode for a while and optimise the WC curve, since my understanding is that in AA mode the controller still uses the WC curve, and fine tunes the LWT with reference to the room temperature. If the WC curve is not correct, then the controller may struggle when in AA mode.
Not a bad call. I think my issue sometimes is not leaving it alone to do the learning! The urge to tweak is too great! In reference to my messages earlier in the year and your responses at the time - I think I also sometimes expect too much when returning from setback - I should be accepting slightly higher flow temps for that than I have been or accept slower warm up times if I tame it. Frustration often though has been it not going into steady state, it seemed to need to be about 4° outside for that to happen.
I can't remember how many industrial control systems I have fine tuned over my 50+ years career, but when possible the first thing that I would do is observe how it responds to a slight change, often in manual control rather than automatic.
Other than the actual designers, I don't think that there will be many people who actual know the inner working of a heat pump controller, so it is often necessary to make an educated guess based upon standard control philosophy.
By now I feel that most forum members have an understanding of WC control, but AA control is a little more involved, particularly since it is not very well explained in the manual.
In its most basic form, automatic room temperature control is quite simple, a sensor measures the room temperature, this is compared against the desired setpoint, and the controller raises or lowers the required LWT to correct any deviation. Simples. 🙄
But then response times raise their ugly head, which is where the Adaptation part of AA, will probably be required.
If the room temperature is too low, and the controller starts to increase the LWT too rapidly, the net result will be that the room temperature will increase to the desired temperature, but then continue to rise until the thermal energy supply and thermal energy demand balance. By now the room temperature will be above the setpoint so the controller will lower the LWT, which may lead to temperature cycling or even heat pump cycling.
In an industrial control system there are controls which can be adjusted to match the response of the controller to that of the process, and hence prevent cycling and provide stable control. Similar controls do not appear to be present in a heat pump controller, other a timing setting and some operating limits.
My understanding is that AA was developed to supplement WC and provide more accurate room temperature control, but still requires correctly adjusted WC for the system to function in the most optimum manner. The start should therefore be to optimise the WC curve over a period of time and then try to incorporate AA into the equation.
I would suggest running the system in AA mode at a constant setting initially, to prove that it controls the room temperature in a stable manner. Only then should any form of setback be tried.
and so when it gets to 30° + 5 it switches off as it's exceeding the autoadapt parameters. I had something like 45 at -2 and 20-odd at 10°. So I've changed to 30 at 10°. I've also changed the time.interval to 30 mins, I had at one hour after reading some of the debates on here about autoadapt. It does seem though like the heatpumps incapable of steady state in the low/mid-thirties flow temp which seems disappointing considering that's where I should get some great COP.
By way of background - I have a radiator system with 6kw Ecodan and 5.2kw heatloss at -2. 22mm primary/10mm to rads, no buffer, open system.
Whats the dt50 value of all your radiators? Also, when you observed the cycling how much of your rad system where closed eg. Thermostatic controlled? Was the above graph created while operating in AA Or was it WC?
apart from your set parameters there is a minimum operating load which might be below the output target 🎯 you may be surprised how low the heat loss will be through the rads at 30c flow temperature esp if the room temp is almost reached.
Open system, no TRVs (no volumiser or buffer either). It was in AA mode. I spoke to an engineer at Ecodan on the helpline today, didn't think it helped much - just stressed was about the ambient temps but the flow temp peak has been pretty similar lately whether at 7° or 11° outside. What I really wanted to get an idea of was whether this was cycling behaviour (can't modulate low enough) or AA behaviour (algorithms and settings). He seemed to indicate it could be AA behaviour to send it to a peak flow temp before immediately cycling down as a way of retaining low flow temps I guess. I'd have expected it to get to steady state.
It's been taking quite a while to get up to temperature from setback, I reckon because of the cycles. However I guess looking at today for example, even with the cycles it's putting in 2-3kwh of heat per hour which I guess would hint at hitting the minimum for the 6kw heatpump (modulating down to 1/3).
-- Attachment is not available --
Thanks for your reply... it’s a good sign you’re getting a COP of 4 and operating an open loop system. So my first sentence asked what is the rated output of all your radiators added together. This is normally listed in the manufacturers data sheet as DT50 output. This is a standardised scale of heat transfer from a radiator heated to a mid point temperature of 70c less a notional room temperature of 20c (70-20=50 =DT50)
Once you know that you can convert the DT to approximate the total output at any flow temperature eg a 35c Flow temperature will only deliver 16.5% of the rated radiator output. This is one contributing factor which would inevitably cause rapid cycling if it drops below the minimum operating output of your heat pump.
So that is why I was suggesting it would help if you know the combined rated output of your radiators?
The Mitsi engineer will probably be inundated with cycling calls at this time of year. And was giving you a valid yet benign reason for possible cycling but not committing to giving you a cause.
I'm watching this thread with interest as I'm in an identical situation to @walkers-heatpump. 6kw Ecodan, with rads, open system, 5.7kw heat loss calc. System been in for nearly a year now, very happy with it, just feel some 'disappointment' regarding lack of steady state running unless ambient temps are near to zero.
@sunandair I've just totaled up the output of all my radiators, they come to 12000 watts. I'm not familiar with what conversion factor needs to be applied for lower flow temps (I'm currently at 35c) - after a quick 'Google' I've come up with a total output of 2950 watts at 35C, have I got that correct(ish)?
@walkers-heatpump - have you looked at Quiet Mode? There's little information available, but from what I've found it reduces the compressor speed (and therefore output) - I've been running in Quiet Mode since last winter in an attempt to 'calm' the unit down (it does appear to have the desired effect to a degree).
Also worth noting that I ran a fixed flow temp of 38C last winter - 38C provided sufficient to get to target room temps at -5 ambient temps, and much less cycling at milder ambient temps. Basically I would have to run a very 'shallow' curve of say 38C at -5, 35C at +15 to provide sufficient heat at cold temps and have an acceptable amount of cycling at milder temps - in the end I saw no benefit from running the curve so settled on fixed. I've reduced to a fixed 35C in the last week or so to see how that behaves (I think it will be going up a degree or to (37-38C seems to be the sweet spot) - it would be nice if the rad output calcs back that up (I think(?) they do).
I'm watching this thread with interest as I'm in an identical situation to @walkers-heatpump. 6kw Ecodan, with rads, open system, 5.7kw heat loss calc. System been in for nearly a year now, very happy with it, just feel some 'disappointment' regarding lack of steady state running unless ambient temps are near to zero.
@sunandair I've just totaled up the output of all my radiators, they come to 12000 watts. I'm not familiar with what conversion factor needs to be applied for lower flow temps (I'm currently at 35c) - after a quick 'Google' I've come up with a total output of 2950 watts at 35C, have I got that correct(ish)?
@walkers-heatpump - have you looked at Quiet Mode? There's little information available, but from what I've found it reduces the compressor speed (and therefore output) - I've been running in Quiet Mode since last winter in an attempt to 'calm' the unit down (it does appear to have the desired effect to a degree).
Also worth noting that I ran a fixed flow temp of 38C last winter - 38C provided sufficient to get to target room temps at -5 ambient temps, and much less cycling at milder ambient temps. Basically I would have to run a very 'shallow' curve of say 38C at -5, 35C at +15 to provide sufficient heat at cold temps and have an acceptable amount of cycling at milder temps - in the end I saw no benefit from running the curve so settled on fixed. I've reduced to a fixed 35C in the last week or so to see how that behaves (I think it will be going up a degree or to (37-38C seems to be the sweet spot) - it would be nice if the rad output calcs back that up (I think(?) they do).
I have attached a radiator output calculator in a spreadsheet. The yellow highlighted cells are the ones that can be varied.
If you know the flow and return temperatures then put those in the LWT and RWT cells, set the indoor temperature, and the answer will appear in the left-hand side table.
If you know the required thermal energy output, put that in the Heat Output cell in the right-hand table, the cell next to it gives the required average water temperature at the heat emitters. Half the Delta T needs to be added to this value to obtain the required LWT from the heat pump.
Thinking about installing a heat pump but unsure where to start? Already have one but it’s not performing as expected? Or are you locked in a frustrating dispute with an installer or manufacturer? We’re here to help.
