The points you raise are all very valid and ones we’ve been becoming familiar with.
We are all too familiar with understanding Solar gain, draught controls and insulation. You may recall some of our exploits in this regard from some of our earlier discussions on these forums. So I’m guessing some of your points are rhetorical rather than actual.
Nonetheless, my comments above go to show I was surprised how much stored energy there was in a heated radiator system. So your comments are a timely reminder to understanding the context of a heating system.
Rest assured I am not planning on rushing to an early conclusion on any changes to our system. My comments above were more a reflection of the ground we have covered. But also to, at some point, draw some conclusions and take positive steps to improve things.
I hope others are finding this thread as useful as I have.
Yes, I do tend to repeat myself, one of the pleasures of growing older.
The point that I was trying to make is not to jump to conclusions that what is happening is actually being caused by the heating system.
Changes is ambient temperature will have virtually no effect on an industrial system controlling at 1000C, but can have quite a dramatic effect on a home heating system controlling at 21C.
One of the other problems is the long time delays and their inconsistency. At lower OAT heating will take longer and cooling will be much faster, so a controller optimised during milder weather conditions may not perform as well when it is colder.
Take WC for example, it is more than probable that the OAT will reduce by some degree before any effect is noticed indoors, but once the IAT starts to reduce the heating system could be playing catch up to prevent the IAT from falling too low. WC was therefore included as a 'feed forward' signal to try to maintain the desired IAT. In some circumstances the action performed by WC may be too slow, whilst in other circumstances it could be too quick, but from a chart is it possible to differentiate that effect from someone using an oven to cook a meal and in the process raising the IAT.
One of the other problems is the long time delays and their inconsistency. At lower OAT heating will take longer and cooling will be much faster, so a controller optimised during milder weather conditions may not perform as well when it is colder.
Take WC for example, it is more than probable that the OAT will reduce by some degree before any effect is noticed indoors, but once the IAT starts to reduce the heating system could be playing catch up to prevent the IAT from falling too low. WC was therefore included as a 'feed forward' signal to try to maintain the desired IAT. In some circumstances the action performed by WC may be too slow, whilst in other circumstances it could be too quick, but from a chart is it possible to differentiate that effect from someone using an oven to cook a meal and in the process raising the IAT.
indeed this is part of the conundrum But I think Auto Adapt, and it’s tendency to overshoot then cool back is also part of the same problem. You may recall a comment from @alec-morrow back in July 2022- on the ‘Ecodan room temp mode’ forum
“Its accepted that stable room temperatures are more comfortable than temperature swings, and indeed a lower stable temperature can be enjoyed if there are no temperature swings.”
This chimes with my earlier comment- I was possibly going to look at a shorter interval period to reduce the 2 hour “off” period after an overshoot of the set temp. This isn’t about jumping to conclusions... all that means is we are just working through the permutations.
Even a 22c temperature falling to 21c can be seen as uncomfortable, more-so than a constant 20c.
At the moment I can control a more consistent stable room temperature using an accurate weather compensation curve with much lower flow temperatures (32 to 35) and the room gently coasts to 20. If there is solar gain for a few hours then my Mitsubishi wireless acts as an over-temp off switch and cuts back in 30 minutes later. This really does work.
In the end the smoothest heat build at the lowest flow temperature is the simple yet elusive objective.
I suppose the starting point would be to try to ascertain if your desired control philosophy is actually achievable.
If you expect (as some may) to be able to lower the temperature setting from 22C to 20C and the IAT will immediately change from 22C to 20C, then you are going to be disappointed. Even in heating mode it could take quite a number of hours to raise the IAT back from 20C to 22C.
Using a hypothetical home with a 14kW ASHP supplying a calculated heat loss of 12kW @ -3C, with heat emitters with a total heating capacity of 24kW.
At an IAT of 21C and an OAT of 9C the probable heat loss will be 6kW, which is the quantity of thermal energy the heat pump would need to supply to maintain the IAT at 21C. To do this the heat pump would need to operate at a LWT of 39.7C.
To increase the IAT from 21C to 22C would require raising the LWT to 41.9C, to supply the necessary 6.5kW of thermal energy, but even then it would take 24 hours for the IAT to achieve 22C. To achieve 22C in say 5 hours would need a LWT of 43.1C, initially supplying 7.4kW of thermal energy, and if this rate was maintained for 24 hours, the IAT would increase to 22.55C.
Lowering the setpoint from 22C to 20C, would probably cause the heat pump to stop, in which case a home with low thermal mass would cool from 22C to 20C in approximately 2 hours.
If the same experiment was carried out on a home with a thermal mass 2.5 times greater, to raise IAT from 21C to 22C in 24 hours would need a LWT of 42.35C. Raising the temperature in 5 hours would require a LWT of 46.5C, which if maintained would increase IAT to 23.6C after 24 hours.
When lowering the setting from 22C to 20C it would take slightly over 4 hours for the IAT to reduce.
Repeating the same tests at an OAT of 12C gives the following results:-
Heat loss of 4.5kW at IAT of 21C with LWT at 35.9C
Low thermal mass requires LWT of 38.2C to supply 5kW for IAT of 22C in 24 hours. A LWT of 39.55C would be required for IAT of 22C in 5 hours, with an IAT of 22.58C after 24 hours. Cooling from 22C to 20C takes slightly over 2 hours.
Higher thermal mass requires LWT of 38.67C to supply 5kW for IAT of 22C in 24 hours. A LWT of 43.1C would be required for IAT of 22C in 5 hours, with an IAT of 23.65C after 24 hours. Cooling from 22C to 20C takes slightly over 5 hours.
The results indicate that only a 3C change in OAT can have a measurable effect upon how the system reacts to setting changes.
Looking at the Room Temp. control details in the manual, I would suggest setting the 'Mode' to normal and try setting the 'Interval' to 60 minutes, which I suspect will provide the slowest response. Then set the desired room temperature and allow the system to operate in AA mode without any setpoint changes. Provided the OAT does not rise too high to cause cycling, the control should hopefully stabilise over a period of time, which could be a day or several days.
One of the other problems is the long time delays and their inconsistency. At lower OAT heating will take longer and cooling will be much faster, so a controller optimised during milder weather conditions may not perform as well when it is colder.
Take WC for example, it is more than probable that the OAT will reduce by some degree before any effect is noticed indoors, but once the IAT starts to reduce the heating system could be playing catch up to prevent the IAT from falling too low. WC was therefore included as a 'feed forward' signal to try to maintain the desired IAT. In some circumstances the action performed by WC may be too slow, whilst in other circumstances it could be too quick, but from a chart is it possible to differentiate that effect from someone using an oven to cook a meal and in the process raising the IAT.
indeed this is part of the conundrum But I think Auto Adapt, and it’s tendency to overshoot then cool back is also part of the same problem. You may recall a comment from @alec-morrow back in July 2022- on the ‘Ecodan room temp mode’ forum
“Its accepted that stable room temperatures are more comfortable than temperature swings, and indeed a lower stable temperature can be enjoyed if there are no temperature swings.”
This chimes with my earlier comment- I was possibly going to look at a shorter interval period to reduce the 2 hour “off” period after an overshoot of the set temp. This isn’t about jumping to conclusions... all that means is we are just working through the permutations.
Even a 22c temperature falling to 21c can be seen as uncomfortable, more-so than a constant 20c.
-- Attachment is not available --
At the moment I can control a more consistent stable room temperature using an accurate weather compensation curve with much lower flow temperatures (32 to 35) and the room gently coasts to 20. If there is solar gain for a few hours then my Mitsubishi wireless acts as an over-temp off switch and cuts back in 30 minutes later. This really does work.
In the end the smoothest heat build at the lowest flow temperature is the simple yet elusive objective.
So much of that resonates with me (and I think I have you to thank for re-visiting AA mode again).
I've attached a couple of charts from the past 24 hours, it's about as good as it gets imo. I'm getting very long run times (heating on) at the lowest of average flow temps with AA mode which equates to that 'lower for longer' sentiment. This does mean the heat pump cycles every 45mins approx at current OAT's but I can't(?) see that being problematic. I'd expect(?) those cycle 'on' times to get longer as OAT decreases (at the same flow temp) then as it gets colder still the flow temp will need to increase (time will tell what it actually does).
Worth noting that as I run quite a 'skinny' room target temp (set to 18C) I'm not looking at night time setback. Again - the lower stable temperature sentiment wins for me (24/7). Our lounge is well insulated so will get to 20-21C, the rest of the house 18.5C.
I think also worth pointing out a couple of other things that have worked well for me:-
Positioning of the remote stat: I've placed it in the most neutral position I can find. It's in a room that has no solar gain, little or no drafts from open doors etc, not occupied for most of the day etc. All radiator balancing / target temps are all using this as a stable reference point.
Quiet Mode: I turned this on last year, not to make the unit quieter but to restrict it's output (there were a couple of references on this forum that stated that in order to make the unit quieter the compressor output was reduced). I saw this as a way to limit the ASHP output capacity (which I believe to be slightly oversized) - it immediately made it less 'excitable'. It may be worth a try if you've not looked at this before - it may(?) help with the smoother heat build at the lowest flow temp objective.
thanks for the information Derek…. There’s a lot of data in your post to digest.
By way of interim feedback, we have been running our sYstem in WC mode to check its alignment with heat loss today. As luck would have it, we’ve had a very sharp drop in OAT tonight. This has been welcome to see how our WC curve adjusts to the rapid drop in temperature.
I’m very encouraged that room temperature has continued its slow climb to achieve comfort without noticing any lag or overheating. So far, that is.
The late afternoon graph shows higher flow temp, reflecting the colder OAT.
While the earlier daytime graphs show flow temps around 32 to 35c
Here is an interesting observation. The two graphs are in similar conditions. one using AA and one WC. It's interesting to compare the flow temps and energy used. It's also interesting to see how AA and WC compare when starting up after the night time setback. WC immediately goes full throttle, creating a big peak, whereas AA ramps up slowly. This is to be expected. But the surprising thing is that the same thing happens with DHW, with an initial peak in WC mode but not in AA. I wonder why that is?
Also, my 48 deg peak when I changed the settings hasn't happened again...
So much of that resonates with me (and I think I have you to thank for re-visiting AA mode again).
I've attached a couple of charts from the past 24 hours, it's about as good as it gets imo. I'm getting very long run times (heating on) at the lowest of average flow temps with AA mode which equates to that 'lower for longer' sentiment. This does mean the heat pump cycles every 45mins approx at current OAT's but I can't(?) see that being problematic. I'd expect(?) those cycle 'on' times to get longer as OAT decreases (at the same flow temp) then as it gets colder still the flow temp will need to increase (time will tell what it actually does).
Worth noting that as I run quite a 'skinny' room target temp (set to 18C) I'm not looking at night time setback. Again - the lower stable temperature sentiment wins for me (24/7). Our lounge is well insulated so will get to 20-21C, the rest of the house 18.5C.
Thanks for your kind words.
Operating at 18c should give you a good COP and, as you’re charts suggest, an even temp. I have found a useful temp setting since we are using a 17c setback with a gradual climb to evening temp of 21. The attached chart shows how the Auto Adapt has learned to overrun the target temp by half a degree then modulate to only slowly rise by another half a degree. It then shuts down for between half an hour and an hour then reheats, etc, etc. (Blue line is target room temp, black is actual room temp.)
This seems to be giving us a respectable 4.1COP Which we’ve never had before. The long modulated runs followed by a pause in operation seem to be helping.
We’ve noticed if the target room temp rise is greater than 1c at a time then Auto Adapt will try a higher gradient flow temperature going over 40c.
It's interesting to compare the flow temps and energy used. It's also interesting to see how AA and WC compare when starting up after the night time setback. WC immediately goes full throttle, creating a big peak, whereas AA ramps up slowly. This is to be expected. But the surprising thing is that the same thing happens with DHW, with an initial peak in WC mode but not in AA. I wonder why that is?
Hi Kev. Have you had any results from new settings of temp differentials?
i altered my wc curve and interval plus shortened temp differential.
we have achieved our lowest flow temp for a given room temp.
basically dropped the warm end of WC CURVE to 28c at 12c and shortened the interval to 20 minutes and set the differential to +3 and -3. And then ran it for a week.
Even though I expected it to cycle at 10 c it appears AA operated it slightly above the cycling temperature of 32c. We’ve had lower and slower flow temps as a result.
There was one period (before we made the adjustments) where the target room temp was passed by half a degree and AA appeared to pulse heat to try to reach an extra half a degree. Then go into a steady rise until the room temp rose by half a degree.
The system then closed for two and a half hours. This is what prompted the adjustments.
It's interesting to compare the flow temps and energy used. It's also interesting to see how AA and WC compare when starting up after the night time setback. WC immediately goes full throttle, creating a big peak, whereas AA ramps up slowly. This is to be expected. But the surprising thing is that the same thing happens with DHW, with an initial peak in WC mode but not in AA. I wonder why that is?
Hi Kev. Have you had any results from new settings of temp differentials?
i altered my wc curve and interval plus shortened temp differential.
we have achieved our lowest flow temp for a given room temp.
basically dropped the warm end of WC CURVE to 28c at 12c and shortened the interval to 20 minutes and set the differential to +3 and -3. And then ran it for a week.
Even though I expected it to cycle at 10 c it appears AA operated it slightly above the cycling temperature of 32c. We’ve had lower and slower flow temps as a result.
There was one period (before we made the adjustments) where the target room temp was passed by half a degree and AA appeared to pulse heat to try to reach an extra half a degree. Then go into a steady rise until the room temp rose by half a degree.
The system then closed for two and a half hours. This is what prompted the adjustments.
that's very interesting. I've been a bit tied up with family stuff so not done a lot. The only change I made was changing the differential to +-3 from +-5. After a day or two of quite erratic behaviour it settled down. It's behaving much the same as before except the LWT range is now 38-43 instead of 40-43. That is, once it's warmed up the house after setback, LWT is lower than it was. It is cycling a bit at higher OATs however so I'll keep an eye on that.
My next move will be similar to yours to try and get the LWT down a bit more.
