Thank you OpenEnergyMonitor: Octopus Daikin ASHP monitoring

Matt - working this through with you has helped me realize a few things;

  1. The turn down ratio for most heat pumps (in equilibrium and without cycling) is only about 40% even though the compressor can be turned down by 80%. Your and my units want to operate at the minimum compressor setting of 920 w when it is 10c or above outside. They can produce a COP of 5 if the rads allow it to do so and so the output is 4.6kw. When it gets to Zero outside the compressor will ramp up to about 3.5kW but the COP will drop to say 2.5 so the output will be just below 9kw (without defrost) and about 7.5kw when defrost is considered (unless the air is very, very dry). So you can turn it down from 7.5kW to 4.6kW, 40 % turn down. Sure you could have 10 or more kW at higher temps but you do not need it and cannot use it unless you want to live in a 30c house!

  2. In the Spring I was running with TRVs set to the room temp I wanted and a Madoka schedule for the lowest comfortable temperature and around 33 c flow temp. The established Paradigm - keep everything as low as possible. I was using about 16kWh each day at about 10c. COP was about 3.5. Now I use the same amount of energy each day but am running the house and the flow temp a lot warmer - so COP of above 4.5. So when you are at the minimum running level of the heat pump and it is stable (without cycling) you have a USE it or LOSE it situation. This will be different as the weather gets colder as a lower flow and internal temp will keep the compressor power down to its minmimum for longer.

Just one comment on your 10c modulation - that is plus or minus 10c? You might want to work out your operating sweet spot (say 37 to 45c) and place your fixed set point in the middle with enough modulation to get to each end of the sweet spot. So 41 c set point with 4c modulation.

All the best


Hi Colin,

Yes, that’s about where I am at the moment. I could run the house colder but I don’t think I would save electricity, it may even use more. The best outcome seems to optimise the efficiency whatever it takes.

Once I change my radiators I will concentrate more on what works best overall, when I know what I have to play with.

I’m not sure what the modulation is, I don’t think it is the flow temperature. It is mentioned in the installation manual as a way of dealing with the fixed delta t of 78c when you set radiators as the emitters.

I have no idea what it does, the description is either too vague or beyond my comprehension!

What I have learnt is that this heat pump doesn’t seem happy at the lowest flow temperatures we see, there seems to be more to lose than to gain at the lower end. The best way to run it is at the lowest electrical input and make sure the house can handle it at 35c.

My gut feeling is that I will work well enough to heat this house when it is really cold and once I get radiators that can handle the minimum output of the heat pump I will have it running as efficiently as possible.

I look forward to finding out.

I will keep this updated when I make some progress, I think it’s a good lesson in how to diagnose issues and make something not entirely perfect work much better.

It has been really interesting to work through the problem in front of us and find an explanation that makes sense.

Thanks Trystan.

I think Colin summed it up well with his explanation of how this heat pump works, you just seem to get more out of it by letting it produce heat, trying to reduce its output too far just makes it run inefficiently.

As I plan to change my radiators soon I am not going to spend much time now finding the best balance between COP, house temperature and electricity consumption.

There’s probably a good balance to be found in there somewhere.

Right now I am happy that we have pretty much proved that I can get a COP of 5 out of it on a heating run. Now I know it is working properly I can work to optimise it as best I can. Hopefully the radiators are my only issue now.

Thanks for sorting the problem with the 30 minute COP, I was looking for anything to explain what was going on with the heat pump. I have never been in doubt that your monitoring was working properly.

Once again thank you, I couldn’t have done this without it and if it was taken way from me I really don’t know how I would cope! Everybody with a heat pump needs this!


No problem Matt, glad to help and I learned something in the process.

The modulation is on the Flow temperature, but it references the room temp on the Madoka.

Say you have a fixed set point of 40 c then when your room temp schedule ramps up from 16 to 20c the heat pump will monitor how quickly it is getting there. If it is too slow it ‘modulates’ the flow temp up a degree, then another, then another until it is getting there quickly enough. Once it reaches the set point it modulates down again until it reaches its equilibrium… so it could end up a few degrees below 40c if your modulation limit setting allows.

Weather dependent set point will just change where it starts this process and because my sweet spot - or efficient and realistic range is 33 to 39 (yours looks like 37 to 45 before the new rads) then I only use modulation - knowing that if it equilibrates down to 33 then that is fine by me, whatever the outside temp.

Incidentally the compressor ramps up or down to get to the flow temp set point ± modulation that the above process decides and the water pump ramps up or down to maintain the delta t (flow - return).

Hope this helps

Hi Colin,

My modulation is 10, the maximum, do you think that is an issue?

I’m finding it slightly difficult to be sure what effect any changes have because I will always get to the point where the capacity of my radiators come into play and that then takes a bit of thinking about.

With my modulation at 10 then I guess it allows the heat pump to do whatever it deems necessary but that may not necessarily be efficient? Thought about it a bit and maybe if I drop the modulation it will allow the heat pump to ramp up more slowly and reduce the initial energy consumption, I’ll try reducing to 5 and see what happens.

I am now confused again, what is the overshoot, I thought that was on flow temperature? I assume that allows the heat pump to keep running even if the flow temperature set by the WD curve has been reached.

After the initial heat up the heat pump seems to settle down nicely at a COP of about 4.

The only thing holding me back now seems to be the flow temperature needed to get to this, once I sort that I should see about a 20% drop in electricity consumption whilst the heat output remains.

Hi all

Firstly, massive thanks to Matt, Trystan, Colin, Marko, Vinny and HydroSam for their work, analysis and instrumentation. As the owner of a EDLA09D3V3 HP, installed by Octopus along with larger radiators, I’m very interested in getting the settings optimised for electricity usage and efficiency. I’ve posted a question about P1P2 monitoring on another related thread from Matt.

I’m currently running using LWT mode only - I have to get Octopus back to move the badly placed thermostat, and I quite like not allowing family to mess with it! I am monitoring some Daikin Cloud data using a HA plugin, which only provides what the Cloud API has e.g. none of energy produced, return flow temp, flow rate and so on, but it does give me LWT, Outside temp, LWT Offset (setback) aka Target Temp. LWT Overshoot is 4deg. Family Room Temp is from a Tasmota DS18B20 in the room the thermostat should be in :man_facepalming: - you can see it’s incredibly stable.

You can see the compressor/pump cycling more/less as the Outside Temp varies up/down. I have DHW on schedule, and you can see the legionella cycle on Friday morning 2am - I should also move that to daytime, to improve COP using daytime temps, except it uses the cheap Octopus Flux power slot.

I’m trying to understand what is happening with Matt’s Daikin, and here’s my precis of my situation vs. his, to confirm my understanding:

  • 9kW HP only reduces output by 40% i.e. from 7.5kW with defrost to 4.6kW minimum (from Colin)

  • My Octopus-calculated radiator output at the various room temps for 50/45 flow/return is 9.8kW (Ive checked this from Stelrad figures), but when adjusted for a flow/return of 32/27 for 12deg Outside (WD curve says 29deg + 4deg overshoot), they can only emit about 2.3kW

  • HP has to cycle to keep flow temp in required range for current Outside temp…

  • Consequent reduction in COP!

  • Matt has a similar problem, which he’s going to install bigger rads to fix

  • COP is EnergyOut/EnergyIn, so actually allowing the full 4.6kW to be continuously produced for the same energy input means the COP is better…

  • But my house is pretty warm with the current output; maybe I just live with the reduced COP because getting a better COP just for the sake of it, given I’m going to use that electricity either way, seems a bit pointless!

Is there something I’m missing/wrong about here, pending further instrumentation and analysis?

1 Like

Hi John,

What is your heat loss as calculated by Octopus?

Mine is massively overstated/overcalculate. They never listened and I knew my heat loss was around 4kW an hour at -5c outside and 21 inside.

I have analysed my data since I started heating 3 weeks ago and it confirms my heat loss at 4kW or less at -2/21. Octopus calculated 7.3kW.

They said that if it was less than 7 they would have installed a 7kW heat pump.

My radiators are big enough for the house and heat it comfortably at 30c flow and 7c outside, they are just not big enough to run the heat pump efficiently.

But I calculated a much bigger radiator requirement to satisfy my quest to run at low flow temperatures. My specification is about 180% of what I have now. I wish I had got they to change them at the time, they would have done it and it would be a lot less hassle than it is now.

My big mistake was not understanding the limitations of this heat pump, I never realised that its minimum output was so high.

But I think I can deal with it with some work, bigger radiators and increased heat loss by adding an area I didn’t intend to heat.

I am with you on COP isn’t everything but I felt the need to find out why I couldn’t get there, partly because I needed to know and partly to make sure the heat pump is working as it should be.

I think to that end we have succeeded.

Once I get a heating COP of 5 out of it I will be happy and then I can work out how much COP to give up to get the best electricity consumption in regards to comfort.

DHW is another difficult one. Mine is a bit erratic, I don’t really know why but I have taken to a daily heat to a higher temperature. My wife is happier and we use less electricity compared to a lower temperature + reheat.

I have abandoned the disinfection cycle for the time being.

I have repeated a lot of what I have said before but it is now quite a long thread and a summary is useful at times.

It will be interesting to see how you get on, when was it installed?

I know it’s money but I can’t recommend the OEM system enough, I would have been completely lost without it.

1 Like

I think 10c of modulation is too high - if the flow temp is set at 40c then this would allow it up to 50 or down to 30. Neither limit looks good for your system 30c is too low for your present rads and 50 will give you a rubbish COP. I would suggest a modulation of 4 to start with and a fixed set point of 40c - for your PRESENT system. With new rads drop the fixed set point to 34c but leave the modulation as it stands and see how it performs. Do this with NO weather dependent curve, at least to start with and then see how it goes.

The overshoot is different - all the above stuff decides where to place the set point, when the compressor is running it will try and reach this calculated set point BUT sometimes when it is warm outside the rads will shut down and the compressor will keep heating the water above the set point. So if the overshoot is low it will shut down at the set point but then cycle quickly (at high frequency). If you set a high overshoot to 4c then the comp will keep running until the flow is 4 c above the set point this will make each cycle last longer. I suspect that the Daikins are not good at fast (ie high frequency cycling), hence the suggestion to put the overshoot to maximum.

1 Like

Hi Colin,

My overshoot is set at 4 and I intend to leave it at that.

I let the house get cooler for a while this morning whilst we were out and set the modulation to 4. It was apparent quite quickly that it made things worse, I don’t know why but I have put it back to 10 as it is giving me the best performance I have seen.

Once I have changed the radiators I intend to run at a fixed flow, overshoot at 4 and drop the modulation to 4. I agree with you, I think a WD curve is not going to work with this heat pump in this house.

I don’t really know what the modulation is doing but i am not seeing flow temperatures drop, it’s holding up at 40c+… It seems to sit at 42c/43c quite happily with the modulation at 10.

I think the only thing holding back the COP is the flow temperature I a having to run. I really hope that increased radiator capacity is the answer that will allow electricity consumption to fall with no drop off in heat produced.

I think I will remove my WD curve tomorrow and set a fixed flow of 40c with an overshoot of 4 and modulation of 4 and see how that goes.

Tonight I just want to be warm, eat our lovely roast dinner and watch a film!

I think this is the basic problem/error by Octopus. Specifying a system to run at 50/45 means specifying a low efficiency system. Sure the heat pump can do this - especially if it is oversized.

However you will find the the most efficient units (lowest running costs and carbon footprints) on all run at something closer to 35/30, most but not all with underfloor heating, I know of at least two who achieve these numbers with standard radiators. Now with rads that does mean a large increase in size which might not always be possible or desirable. In my bedrooms I have not changed my rads - so they are undersized I know, but I do not like hot bedrooms so it is not a problem and anyway if my downstairs is warm it gives some underfloor heat to the upstairs. Downstairs I have massively increased the radiator sizes and used fan assisted radiators as well when I wanted them to be less visible - my aim was 40/35 on the coldest days.

Just to clarify - my 4.6kw minimum is on continuous running. You can get lower with cycling on and off but it seems that the Daikin units are not that efficient when cycling (other units are much better at this if you look at The Daikin control theory does allow many variables to be changed and I suspect this is to try and get the units running continuously. So if you set them up to get continuous running you will achieve efficiencies close to the best of the other systems.

All the best to you both!

1 Like

That is the neat thing about modulation vs weather dependant curves. With the latter you make a best guess at what the flow temp should be and the heat pump just has to try and obey your command. But with good modulation the system works out what the best temp is by itself. In your case it finds that the radiators are not putting enough heat out to move the room thermostat fast enough when the flow is below 42c so it is adjusting for your radiator system… at least that’s my theory, until corrected by those more expert.

1 Like

I should add that when it gets really cold outside the house will cool faster and the system will find that 42c is not warm enough to keep the room thermostat satisfied. So it will modulate the temperature up when it gets colder = weather dependency via modulation and (not by the owners best guess).

I think you are right Colin.

I think Glyn does more or less the same thing, settles on a flow that works and is efficient and lets it turn on and off as necessary to keep the set room temperature.

My problem doing that is my radiators can’t handle this heat pump at its most efficient.

I am also struggling with what appears to be inconsistent behaviour.

My DHW is all over the place and now what was working yesterday and this morning is not going so well this afternoon.

It’s ok playing with settings when you have a predictable base setting but when everything seems to be moving all the time it’s hard for me to get a grip of.

I love motorcycle racing and this has lots of the same elements in terms of set up, you need a good solid basic set up that works and that you can fine tune. If the base set up is off you are lost. It’s how I feel right now.

You schedule seems to be working though - it has come on at 2pm for the last few days. What is the issue?

This is great stuff you are sharing here, thanks. My system has been cycling like made today, gradually dropping the LWT. I have it setup to run continuously at the weekend and it’s too hot in the house, now 22C and it hasn’t shut off.

I’m going to try changing to fixed LWT and see whether I can improve things based on the feedback here. If you want to help on my system I’d love to go through the same process over on my post.

1 Like

Hi Colin,

Yes schedule is fine, COP varies quite a lot. It is reheating from about the same temperature each day and the weather is pretty consistent.

Maybe it’s a wild statement from me!

Hi Sam,

I’ll have a look at your post later.

I’m not sure how much help I can be but I am glad my `mucking’ around is interesting at least.

1 Like

Maybe as I think you have said earlier focusing on COP is too narrow. If your house is warm enough and you can gradually lower your energy bills then that’s the real objective! Small incremental adjustments from now on…


Thanks for replying, Matt, appreciated.

It’s 8.4kW, vs. the 9.9kW radiator output, so the radiators are in principle too big, but then the heat calcs don’t account for thermal bypass generally and some dodgy tiled areas! And of course, it’s sometimes colder than -2.

How did you do that from the data you have, since it didn’t include a -2 period?

I suspect I’m in the same situation! Some of them could have gone bigger, but we’re bumping up against the “reasonable retrofit impact on house interior” in a few places.

Yeah, hindsight is a wonderful thing… But it’s hard to have the courage of your convictions, if they could even be classed as that until you have more evidence - but that takes an installation!

I don’t know that that’s immediately obvious from the specs, so how would you know? It would come down to Octopus, as the people installing container ships’ worth of Daikin ASHPs, hanging around to collect proper data and act on it. If you consider that you and I are otherwise pretty ok with the actual in-house result, and my wife now thinks I’m a genius for making this happen, if that’s the result they get with 90% or more of their customers, along with scrapping all those gas boilers, job probably done.

At least you will be making compromises based on a proper understanding of the parameters and their effects.

Yeah, that is weird. Mine seems fine, admittedly without proper data, I had it on schedule and reheat, but we don’t seem to need the reheat, so I dumped it for now. Maybe with the extra guests over Xmas I’ll switch back.

Still considering this. As you can see from my screenshot, the water temp lasts a pretty long time and avoids at least one scheduled warmup! We probably need it because we use so little water.

No, a summary is really helpful, especially for those just starting on such a huge and confusing thread.

May, so like you it wasn’t until 14/15 October that we it turned on at all, and we still haven’t really seen sustained low temps to test the bottom end setup, and thus the full power of the “fully operational Death Star”! For a big lump of stuff on the patio, it looks pretty cool though :rofl:

Sigh. I kinda know that, maybe you can DM me about your OEM installation to give me more confidence about going for it. My only concern is a possible pipe-cutting requirement and refilling without excessive air.


I was referring to the ESPaltherma monitoring facility you mentioned acquiring in your previous monitoring thread - I think it uses ModBus via a serial port. That’s tempting too, especially as it’s pretty standalone and doesn’t get one involved with other components like the P1P2 interface.

Here’s to warm and cosy, with an eye on the environment!


Hi John,

Heat loss is a linear regression ( I think that is the correct terminology)

If I know what it is at one set of temperatures I can plug that into the formula to work out what it will be at any other temperature.

For example if I use 36kWh of heat to keep my house at 21c for the day whilst it is 10c outside I know my heat loss with a delta t of 11c is 1.5kW per hour.

So if it is -2c outside and I remain at 21c indoors the delta t is now 23c

So 1.5kw x 23/11 = a heat loss of 3.14kW

I would need 75.36kW to keep my house at 21c when it is -2c outside.

And that is about what it is historically .

The most gas I ever used was 110 kWh and that was when it was an average -5c for a day. That included DHW and boiler inefficiency so something like 85-90 kWh of heat required.

All this information was available to my installer but they went with a heat loss of 7.3kWh and sized the heat pump based on that.

Hence my problem.

The OEM install is straightforward plumbing wise if you have enough space. Mine area was a bit confined but I managed it with ease.

Draining the heat system was easy and refilling was no trouble at all with automatic bleed valves. The heat pump has a air purge cycle you can run in installer settings and if you have a back up heater then the bleed valve in the heat pump is automatic as well.

There are some pictures of my install in this thread.

I made sure I had some spare pipe and compression fittings just in case I messed up!

I think the ESPAltherma is good if you are OK with home assistant. It just plugs into a serial port inside the heat pump. I have not read of any issues with it with regard to damaging the heat pump.

But as with everything I am concerned with warranties. I tried to get Octopus to fit the OEM system for me but I never gt a straight answer from them, as far as I know their technical team are still looking at it for me!

Apparently Octopus did fit this to 50 systems so that they could monitor some.

It is a shame that they don’t offer it to every customer as a cost option, it would have been a lot easier to fit at install.