Very interesting! We’re in a similar position, with an EDLA09 which I now realise is well over-sized. For heating we got a COP of 3.0 for the last 12 months, much less than I hoped for and less than the 3.35 estimated. We used WC mode for last year with quite low temperatures, fooling myself that this would help- when in fact the compressor would have been stuck on minimum output. Even then flow temperature rose and the compressor stopped - leaving the pump on, and still drawing 200W overall.
Now I’m trying fixed, 35C or so, and can see that at least the compressor can modulate a bit. I’m using the Madoka to run the HP morning and evening only because at 35C plus, the house heats up very quickly just like a gas boiler. Time will tell if this helps the COP, but fundamentally this HP is too large for our property.
We’re on heatpumpmonitor, the only one located in Swindon.
Hi @squarepeg77, I have some stats for Fancoil vs UFH with similar outside temps, it appears that the Daikin Algorithm is the same and has no effect on adjusting or lowering the gradient of the initial heat phase, the profiles look very similar, I think I have fettled enough, and I cannot tweak any more efficiencies but will help if I can and watch on from a far!
Hi @scanfer I checked with Daikin support just to be sure, and if the outside temp is below 7 ambient degrees the pump will flow continuously even with the compressor/fan stopped.
As we have antifreeze valves fitted, we could go with the intermittent (‘non-continuous’) setting that shunts the pump every now and then = saving power. hope this helps?
I tried DT4 and UFH yesterday evening and saw no obvious difference. This morning I tried DT6 and UFH and the system never stabilised at DT6. I don’t think my radiators can’t lose that much heat at the available flow rates from the 9kW system.
I’ve gone back to fancoils and DT5 but increased modulation to 2C.
And I’ll try and sit in my hands and not fiddle with it.
If your goal is to lower the peak energy spike, I think restricting pump performance will do that, but possibly result in longer run times in whatever phases of operation the system goes through. Did you try that? (I forget all the experiments…)
thanks for the reply, TBH I have not tried changing the pump speed limitation…I have taken note of the caveats in the MMI manual, and I have not calculated the flow rates needed, I really do not want to break the heat pump within our 1st winter season…we are at a COP (Space heating) of 4.9 at 4 degrees, I think are good…?
This should only be attempted if you have calculated the pump speeds through the pipework using a hydronics calculation tool and you are looking to balance a system delta T to the system or eliminate water flow noise. Due to higher flow rates achieving higher efficiencies it is advisable to leave this setting at “No limitation”
I limited my pump speed to 60% yesterday, the minimum.
It worked really well, established a dT of over 8c immediately at a good COP.
But it shut down when the room temperature hit 22.5c and when it restarted with a higher requested room temperature it wasn’t working so well.
I don’t know why, it had been running between 21.6c and 23.5c with 22c requested.
I don’t know how to control the range of room temperature.
I am also off the WD curve with a fixed flow of 36c and no modulation but it can’t or won’t do it.
I have experienced this all of the time, an inability to hold a flow temperature.
Some strange behaviour today and the DHW was relatively awful and the dip in flow temperature is very pronounced at times, at one point it wanted to run at 32c
This is the run with the pump speed limited to 60%, it would have been good if it had kept going but I think the Madoka shut it down, too much heat produced for the house.
I don’t think the Madoka suits my house with this heat pump, it is trying to do things the house cannot cope with.
Although my radiators can cope with the low flow temperature and deliver most of the heat available I think the Madoka is expecting the house to take longer to heat up. As the house is too small for the heat pump and the heat produced I think it messes it up.
This appears to be the same as ours ie on at -0.5, off at +1.5, against room target temp and normal behavior by design.
Madoka + modulation 8 -
Room temp set at 20
turns on at 19.5
turns off at 21.5
Our system ran continuously from 11:30pm to 6am holding room temp, I guess we could dissipate and lose the heat for the long cycle at less than 4 degrees outside temps.
If you look at the run it produced 4,500w of heat in just over 30 minutes, I guess the Madoka thought that was enough and shut down even though it had only got to 22.5c with 22.0c as the set temperature.
I have not found any way to control when the heating turns on and off with the Madoka, it seems to be hard coded at times, but then at other times not, I don’t know.
If I could control more I think I could get it to work better, I need control of the flow rate, the flow temperature and the points at which heating starts and stops.
I don’t think I have enough control though so I am stuck with the programming devised by Daikin which is I would say is for a house with different properties to mine.
I’m at a bit of a loss with it, I can get average performance generally or I can get it working really well but have to open doors and windows.
I always come back to the same issue of the minimum output, or more correctly, the minimum input being too much for a house this size.
Everybody with this heat pump seems to have the same problem, either not enough emitters and/or too much heat/too much electricity compared to a more suitably sized heat pump.
This is a good point, we always have all our internal doors open (except the stairs), could you move the Madoka to an area far enough away and in the middle of your emitters (hall way?) so you can have a more balanced gradual increase/decrease air temp?
The Madoka is in the hallway and I never shut any internal doors.
If I run continuously with no Madoka influence I have to open windows and external doors to get rid of the heat, even at the minimum electrical input the heat produced is often more than twice what we need.
My heat loss today is 2.5kw with an average of 22.6c indoors and 7.6c outside. If I just let the heat pump run continuously it will produce somewhere between 4,500w and 5,000w of heat, it’s way too much.
Now I am using the Madoka I think the house just gets hot too quickly for it.
It’s just like the gas boiler it replaced, loads of heat and the house gets hot really quickly.
I am guessing but I would expect the Madoka is expecting a slower increase in room temperature than it sees, I don’t think it is programmed to work like this.
I declined mainly because they weren’t nice about it, they weren’t prepared to accept they got it wrong, and I really want to make what I have work.
I have a Daikin engineer coming next week and one of the things they say they will do is assess the suitability of the heat pump to my house and report to Octopus if they find it is unsuitable.
To be fair I was expecting a lot more grief too. I mentioned it to the installer while calling up about a different issue and I had a call back that afternoon to offer a swap.
The main reason it has taken this long is that I wanted a heat geek (Szymon from urban plumbers) to come out and complete a full heat loss survey (as I wanted to run at 40 and I wanted to know the actual heat loss per room).
Octopus have agreed that they messed up and that they will be replacing the 9 for the 8 on Tuesday 9th. My actual heat loss by Szymon is 5.45kw, according to HTC myself is around 5kw but there is no point going for the 6 as same issue is just software derated and has no lower power benefit. The original heat loss calculation has left the house radiators unbalanced to such a degree (needs 43°c downstairs and 28°c upstairs) they will be changing out some of the radiators to allow us to achieve a 40°c flow rate.
This is all FOC and they have not asked for anything (I did offer as to get to 40°c we will need to swap out a new K2 for a 500*2200 k3 (and I know they are not cheap and that is above and beyond the original specification).
I get they should have not messed up the heat loss in the first place but everyone is human and can make mistakes (I always think of Hanlon’s razor here due to the amount of stuff I have seen in my professional life). The fact that I said it was wrong and they came back with options to fix it without much grief is a good sign.
When you are literally installing hundreds of systems a week there will be mistakes but it’s how you handle them that counts in my book.
Without my open energy monitor results and the fact the operations manager was involved in my install from the off (due to having them wire up the smart grid functions via the Eddi relay board) I cannot say if I would have had it so easy (or if I would have necessary noticed) but hopefully soon we can see what my house can actually achieve with the correct sized system.
I think the problem is deeper than errors in the heat loss calculations.
They don’t help when you are on the cusp of an 8kW or 9kW but the real problem is Daikin.
If their 9kW heat pump had a sensible range of operation none of us would be talking here and contemplating/actually changing heat pumps.
If the 9kW modulated down to around a third of its rated output then I think we could all live with it.
The error that Daikin have made is we have a 16kW heat pump in all but the highest output, it’s the same compressor and the same circulation pump designed for a 500m2 home.
If it went down to 500w to 600w of input power and 3,000w of heat we would all be happy, it’s more than we need some days but perfectly manageable and a sensible step up from the 7kw or 8kW models.
That sort of lowest power from a 9kW heat pump is not an unreasonable expectation.
900w of electricity and 4,500w of heat is not realistic for any 100m2 house with any half decent insulation.
It’s Daikin’s issue and not really the fault of Octopus, a relatively small overstatement of the heat loss and a bit of a margin for error and comfort and we end up with a monster of a heat pump on the average heating day.
I would say though that Octopus have to accept some blame, they don’t understand this heat pump, how it was designed, its limitations and what it is like to live with. It may meet their modest expectations in terms of performance but there was so much better available on the market that they could have chosen to partner up with and install.
I still can’t believe Daikin made this heat pump and thought it was a viable solution for heating a home in the UK.
Matt, hopefully once Daikin have been and visited you too may be able to swap out for the 8kw unit (or check my stats from Wednesday next week and see how it performs).
With all the improvements and work you have done now surely it must be a SCOP of 5 or bust.
@tiger_cook If you are uncomfortable limiting the pump speed, how about enabling one of the quiet mode settings? That will probably cut the peak power spike. You might need to disable it on the coldest days if you don’t get the temperatures you require.
Thanks, but I’m not sure that would help.
When the Madoka thermostat has tripped, the circulation pump stops and only runs intermittently for freeze protection below 6C. Between compressor cycles it still runs though, which at low flow temperatures can be a long time.
thanks @squarepeg77 like you, I think I need to sit on my hands and accepts a few undesirables during shoulder (higher) outside temps, you can never achieve perfection especially with UK weather that can fluctuate 10 degrees from day to day, I am learning how to live with the system, it does not really learn how the house heats up and adapts that quickly, one thing I noticed at ~4.5 degrees outside the HP did not stop and circulate for 20-30 mins compared to a 7 degree outside
I think the peak is ok for me as it does not exceed our battery output and we would not draw from the grid, I do not think Daikin expose ‘quiet mode’ with their API, so I could not use Home Assistant to automate this dependant on the outside temps (maybe >7 degrees), and I think doing it manually is flawed as I am the SPOF and human.
There is a new Repo (Fork) that is cloud dependant that exposes more control (on/off function and set temp), I use this as well as the ESPAlterma (local monitor only)
caveat to the below: I have zero knowledge of the business relationship between Octopus and Daikin.
I have seen this kind of thing within the telecoms service provider and enterprise data networking industry (what I do). I’m sure others have similar business experience in whatever it is we all do. buy-side engineers (analogy being the Octopus engineers in this case) will do all manner of testing of equipment from vendors A,B,C, to see which best meets a particular requirement. They may well be able to prove with all manner of technical reports that A is the best fit. But if C has the best commercials, and/or best business relationship with the company doing the buying, unless they can say that C is completely unfit for purpose, C will get the deal, and the engineers have to do their best to work around whatever the limitations of C are.
