Thanks James and Derek. The LLH is inside the house, in the downstairs bathroom with all of the other indoor parts.  When you mention the poor COP, remember that when the system is allowed to run normally (no cycling), the COP is between 4 and 5, which seems fine to me.  The only time it drops to below 3 is when it's cycling. Fundamentally, when allowed to run freely without cycling, it behaves according to spec, which is why I'm not sure a complete system review will actually make any difference. If it continued to run poorly in those circumstances, I'd agree that we have a fundamental problem, but I have to think that basically there's nothing majorly wrong with it when it's running.

Switching to 35C flow temperature with a 5C differential target has brought the COP up to just over 3 today in temperatures of around 12C. Maybe tomorrow I'll open the windows (if it's not raining like today!), set the TRVs to 25C, and ensure that the system is running continuously all day. If that gives us a COP of 3.5 upwards, then we know the system is just fine when it's not cycling.

@rgledhill

What are the on - off periods when it is cycling?

It's tricky to tell when it's using the difference between out and return flow temperatures rather than the Honeywell thermostat.  The Honeywell system is set to cycle no more than 3 times in an hour, so when it was using this (via the External Thermostat setting), the COP was a bit worse but not much different.

Posted by: @rgledhill

↑

Thanks James and Derek. The LLH is inside the house, in the downstairs bathroom with all of the other indoor parts.  When you mention the poor COP, remember that when the system is allowed to run normally (no cycling), the COP is between 4 and 5, which seems fine to me.  The only time it drops to below 3 is when it's cycling. Fundamentally, when allowed to run freely without cycling, it behaves according to spec, which is why I'm not sure a complete system review will actually make any difference. If it continued to run poorly in those circumstances, I'd agree that we have a fundamental problem, but I have to think that basically there's nothing majorly wrong with it when it's running.

Switching to 35C flow temperature with a 5C differential target has brought the COP up to just over 3 today in temperatures of around 12C. Maybe tomorrow I'll open the windows (if it's not raining like today!), set the TRVs to 25C, and ensure that the system is running continuously all day. If that gives us a COP of 3.5 upwards, then we know the system is just fine when it's not cycling.

Pleased to hear all of that, you seem to have decided what to do and why, so my best suggestion is to do it and see what happens.

If the only material problem in your system is cycling, and that is accounting for a significant reduction in COP, then that is very interesting data indeed. If you could post any data you collect that may help others to optimise their systems.  As things currently stand I am not aware of any quantification of the downsides of cycling.  Its said to be bad, there is a very sound argument in thermodynamics for it resulting in some degradation, but nothing like the factor of 2 or more which I think you are suggesting.  Of course there may be model-specific factors at work here which I certainly don't know about, so please post any measurements as it may well help others!

Posted by: @jamespa

↑

If the only material problem in your system is cycling, and that is accounting for a significant reduction in COP, then that is very interesting data indeed

Hi James, here is some limited data from my Samsung 12kW Gen 6 ASHP. For reasons best known to itself, it decided to cycle between 1am and 4am and then ran continuously till 6am. When my Samsung cycles it does so every 8 minutes, so 7-8 times an hour. The hourly consumption kWh were broadly similar throughout, but the COP was worse during the hours of cycling. Here are the COP values for the hour that follows the time given:

01:00 3.51

02:00 3,50

03:00 3.42

04:00 3.68

05:00 3.74

The first 3 values are when it was cycling and the last two when it was running continuously. Not a huge difference and certainly nowhere near a factor of 2, but significant nevertheless. The OAT dropped by around 0.5 degC between 01:00 and 06:00

Posted by: @mike-h

↑

Posted by: @jamespa

↑

If the only material problem in your system is cycling, and that is accounting for a significant reduction in COP, then that is very interesting data indeed

Hi James, here is some limited data from my Samsung 12kW Gen 6 ASHP. For reasons best known to itself, it decided to cycle between 1am and 4am and then ran continuously till 6am. When my Samsung cycles it does so every 8 minutes, so 7-8 times an hour. The hourly consumption kWh were broadly similar throughout, but the COP was worse during the hours of cycling. Here are the COP values for the hour that follows the time given:

01:00 3.51

02:00 3,50

03:00 3.42

04:00 3.68

05:00 3.74

The first 3 values are when it was cycling and the last two when it was running continuously. Not a huge difference and certainly nowhere near a factor of 2, but significant nevertheless. The OAT dropped by around 0.5 degC between 01:00 and 06:00

Was your system actually cycling or defrosting?

Posted by: @mike-h

↑

Posted by: @jamespa

↑

If the only material problem in your system is cycling, and that is accounting for a significant reduction in COP, then that is very interesting data indeed

Hi James, here is some limited data from my Samsung 12kW Gen 6 ASHP. For reasons best known to itself, it decided to cycle between 1am and 4am and then ran continuously till 6am. When my Samsung cycles it does so every 8 minutes, so 7-8 times an hour. The hourly consumption kWh were broadly similar throughout, but the COP was worse during the hours of cycling. Here are the COP values for the hour that follows the time given:

01:00 3.51

02:00 3,50

03:00 3.42

04:00 3.68

05:00 3.74

The first 3 values are when it was cycling and the last two when it was running continuously. Not a huge difference and certainly nowhere near a factor of 2, but significant nevertheless. The OAT dropped by around 0.5 degC between 01:00 and 06:00

Thanks, interesting.

If I have done the maths correctly that's about a 6% increase in energy consumption during cycling (1/3.5 vs 1/3.7), ie small but worth eliminating if you can without otherwise compromising performance ... which is what I would hope to be the case.  As you say nothing like a factor of 2.

Of course there maybe models which perform differently as this is probably more about engineering than the raw thermodynamics.

Thanks all. I'm as surprised as you that cycling is having such a big effect, but at the moment this is where logic is pointing. Tomorrow I'll force the system to target 25C in all the rooms, but leave it set to 35C out, targeted 30C return, so that the valves should all be fully open and I'll control the temperatures in the rooms with opening or closing windows! I'll also set it to be controlled by the external thermostat as this should ensure it runs continuously too.  I know this is a bit wasteful of energy, but basically I want to have the system running at 35C, continuously on but at a low setting, for several hours and see what happens to the COP. The nearest I've had to this so far was when the house cooled by about 5C while we were away for 24 hours and then the system brought everything back up to temperature the next day (~4 COP).

If nothing else, it's interesting to speculate and then evaluate the results!

Posted by: @rgledhill

↑

Thanks all. I'm as surprised as you that cycling is having such a big effect, but at the moment this is where logic is pointing. Tomorrow I'll force the system to target 25C in all the rooms, but leave it set to 35C out, targeted 30C return, so that the valves should all be fully open and I'll control the temperatures in the rooms with opening or closing windows! I'll also set it to be controlled by the external thermostat as this should ensure it runs continuously too.  I know this is a bit wasteful of energy, but basically I want to have the system running at 35C, continuously on but at a low setting, for several hours and see what happens to the COP. The nearest I've had to this so far was when the house cooled by about 5C while we were away for 24 hours and then the system brought everything back up to temperature the next day (~4 COP).

If nothing else, it's interesting to speculate and then evaluate the results!

Noted - but remember the old saying: correlation does not equal causality

Having said that if cycling is causing such a large degradation (and its genuinely a heat pump degradation, not a function of some other part of your system), there is something very odd about the Daikin unit.  Given that they appear to firmware-limit the output, its potentially an even bigger problem than one might at first imagine, because more cases will be affected.  They have a good reputation, albeit that they are, at least in my limited experience, more than a little reluctant to provide detailed technical information to mere end customers (which is always suspicious)

Good luck.

@rgledhill I had an evohome system, I removed it from everywhere except bedrooms to optimise the ASHP efficiency. Its still there in the bedrooms to allow them to run cooler overnight.

I take your point that you have struggled due to your solar gain to live completely without evohome but there are things you can do to try to "tune it out". Firstly have you set the "fully open" parameter on the TRV's - I can't remember what its called. so that when there is any demand, the valve opens fully. you don't want the "open a little bit" situation to ever occur, that is designed for modulating the flow rate through the radiator right down, in a boiler setup.

Secondly set your TRV setpoints to 1.5 - 2C higher than the target temp that you want for that room. Tune the lockshield / flow rate as best you can to get the room temp right in normal circumstances.  Use the TRV to only shut that room down if its becoming uncomfortably overheated when the solar gain kicks in. Not to control the room temp in normal running.

the objective of the above is that as many radiators are as open to flow as possible as much of the time as possible. which WILL reduce cycling effects. and these WILL have a material positive effect on COP. if you can run at 35C , you should be able to aim for COPs around 4 in mild temps (7C +). 

from discussion on other forums you should be glad you've got an 8kw daikin.  the consensus seems to be this is a genuine 8kw unit that modulates down reasonably well. whereas their 9kw is actually a firmware de-rated 16kw and an absolutely ***t to get to modulate down, floor is about 4.5kw.

Posted by: @jamespa

↑

If the only material problem in your system is cycling, and that is accounting for a significant reduction in COP, then that is very interesting data indeed. If you could post any data you collect that may help others to optimise their systems.  As things currently stand I am not aware of any quantification of the downsides of cycling.  

some data from my system today that I think proves and quantifies this. It always cycles first thing when mild. Pretty sure, I know why, because three large rads in the bedrooms are off (by virtue of the evohome programmable TRV's). I believe downstairs isn't quite enough emitter area for stability. What I haven't yet figured out is the best  way around it!

Mean LWT is 34-35C in both. Room temp 20-21C in both. ambient 6-8C in both.

first: with cycles, 0430-0620.  COP 3.72

second: without cycles, 0630-1000, long clean run. COP 4.05

that is about an 8% hit. Plus some compressor startup wear/tear that would be nice to tune out.

Thanks everyone.

Today I basically did what everyone's (correctly) proposing - I set all rooms to 25C on the TRVs, full valve opening on lock nuts and TRV setting, 35C flow temperature and set to maintain 5C differential between incoming and outgoing water.  The system ran all day continuously at a steady, low flow, without cycling (as far as I can tell).  Rather than faff with locknuts (I'm working from home, not retired!) I simply opened windows a crack when it got to 23C, to keep rooms safely below the 25C point where the EvoHome might intervene and partially shut valves.

The end result? An identical COP of about 3.2, same as yesterday, and I barely had to open any windows except for half an hour when it was sunny.

So I'm puzzled as to why it didn't get to around 4 as it did last weekend when it was heating the house from about 15C, though I guess the temperature difference in the rooms  compared to the flow temperature would be higher so heat emission would be greater.

Today I've still been using the Grundfos pump in its auto mode based on the pressure on the system (i.e. the more open the radiator valves are, the lower the back-pressure, so the higher the pump speed needs to be to get decent flow rates). I have noticed that this is still a significantly slower speed than when on it's lowest fixed setting, which makes for a blissfully peaceful house but could also result in lower heat emission rates.

Tomorrow's test: back to normal target temperatures in the house, but with the Grundfos CHW pump set to speed I (lowest fixed speed but higher than it's been running previously).  I shall report back...