Thank you OpenEnergyMonitor: Octopus Daikin ASHP monitoring

[John (John Gouk)]

Another very interesting thread! Like you, as we discussed on another thread you started, I had a Daikin 9kW ASHP installed by Octopus in May, and used it in earnest last weekend. My Madoka room thermostat is unfortunately in the S-facing front room (my mistake when asked by installer, it was a convenient route and installation spot :-/), any sun at all and the temp exceeds 20deg, whilst the N-facing back room was really cold. Rather than mess with thermostat offsets, which would require manual intervention any time the sun came out (!), I switched to only weather compensation i.e. flow temp determined by outside temp. This has worked really well from a comfort point of view, so as a first pass, we’re happy. Hurrah. My wife thinks the radiators are voodoo, because they never feel “hot” but the rooms are toasty! I’m inclined to agree…

I have adjusted the WD curve to be (flow/outside) 50/-7, 26/15, because the house is warm enough at 14/15 outside without heating. It was previously 50/-7 26/20 as left by the Octopus guys. However, when it got to 14/15 outside, it was still cycling on, so I discovered the register [4-02] which sets the maximum temp for the space heating to come on, and set it to 14. Bingo, HP turns off :-). I guess I could move the upper setting to say 26/20, as it was, and have the max outside temp setting turn it off, but then it might be too hot at the high setting when it’s 14 outside.

You can also set up a simple schedule to change the flow temp by ±x degrees, in the Daikin MMI, not in the app. Go to installer controls, Space Heating, and you’ll find it. I’m still experimenting with this, since the outside temp isn’t constant it’s hard to judge if the floe temp is dropping overnight - I set it to -2 22.30-6.00, but the crucial back room is hardly changing temp at all, as monitored by a simple SHT3X + Tasmota ESP8266 + MQTT + NodeRed combination.

I don’t have your sophisticated energy output monitoring, but I could… and might! I probably need to go through the Octopus survey and radiator numbers to see how they work, having read the very instructive advice you’ve been given here. My lead installer certainly seemed to know what he was doing, but there were a couple of points where I was a bit “hmm, I’m not convinced you know why you’re doing that!”. And I’ve found out lots since.

Cheers

[borpin (Brian Orpin)]

@matt-drummer

So you have convinced me that Octopus don’t really know what they are doing.

It hasn’t taken long for me to see the problems.

But it is not the installers, they just fit what they are told to and mine has been fitted well.

The problem is in the design, and maybe the choice of equipment.

(not read the whole thread TBH).

I think this is yet another demonstration, that an HP is not fit an forget by installers. It needs tweaking to get the settings correct for maximum (acceptable) efficiency.

This is what the installer hate and have zero incentive to do. Therefore, only those householders who take any notice will do anything about it.

@John

whilst the N-facing back room was really cold.

And for may that would equal - HPs are crap, when in fact it was a problem with the settings not an inhernet problem with the HP.

@John

but there were a couple of points where I was a bit “hmm, I’m not convinced you know why you’re doing that!”

Yep, but for most they don’t know to ask that question!!!

[markocoheat (Marko Cosic)]

@matt-drummer

Knowing what I now know, I think I have it set up nearly OK. It’s heating for an hour at a time, sometimes nearly two hours but less when it’s warmer, sometimes only 30 minutes.

That is absolutely fine in this mild weather!

@matt-drummer

But I think it’s nearly there, I can just up the flow temperature a bit so I get to the requested room temperature in one go more often.

What do you mean by “get to the required room temperature in one go more often?”

Are you…dialling the room temperature up and down?

So that inseatd of the radiators running cool and constantly putting heat into the property at a low temperature you’re making them run hot to get it up to temperature and then do nothing?

That will hurt your COP. If you’re aggressively changing temperature it can hurt more (through reduced COP) than it benefits (through reduced heat demand through the ventilation heat losses dropping slightly overnight)

@matt-drummer

It just says design flow temperature 50c, when I spoke to them that is 50c from the heat pump.

That isn’t a valid answer without the corresponding return temperature. It will be (should be) on the MCS docs somewhere though.

@matt-drummer

I was hoping to see COP’s between 3.50 and 4.00 during the heating season and I am really not that far from it.

I’m at 3.32 today and I have been experimenting. I have always said that I would be happy with anything 3.00 or above so it’s certainly not a disaster.

Looknig at these two:

“Weston-super-Mare Octopus Energy Services Daikin Altherma 3 7 kW”
https://quadratic-tropicbird-6273.dataplicity.io/app/view?name=MyHeatpump&readkey=ba136cb62155e4c345349bfb0d220f86

“Ipswich Octopus Energy Services Daikin Altherma 3 9 kW”
https://emoncms.org/app/view?name=MyHeatpump&readkey=bbd1cd04728bd8c9f7acfb2ee51936d2

For very similar operating conditions *8 degC out and a positively tropical 37 degC flow temperature) the 7 kW seems to edge out the 9 kW unit at lower outputs:

This one is fun.

https://emoncms.org/app/view?name=MyHeatpump&readkey=bfd97700dd78658dedd70ff6670b453c

Big nigtht setback (so hammering the unit hard in the morning at poor COP instead of letting it tick over all day) and charging up DHW at the coldest part of the day instead of the warmest part of the day when the sun is shining etc.

Compare with a heat geek style setup:

https://emoncms.org/app/view?name=Primary&readkey=eec7334e58f2a6927a9ff7b27c8c7e63

Where is just ticks over all day with the radiators as cool as you can get away with / decent length cycles.

It’s pretty clear when folks are out / on holiday from these datasets by the way. No DHW charging since Friday 20th on that heat pump!

[markocoheat (Marko Cosic)]

I would start with the unit in an “idiot simple” mode:

• No time clock
• Pure weather compensation
• Ticks over all day with rads at a very low temperature
• DHW setpoint 50C

The introduce the load compensation:

• Target water temperature is tweaked up and down a bit if the room is too far off setpoint due to sun / cooking etc

Then time the hot water and dial back the setpoint:

• For the hottest part of the day to maximise COP for the hottest event of the day
• Setpoint 45C

And see how you get on.

Then only after you’ve got it working well like this do you experiment with setback / timing to try and reduce the amount of heat lost (by reducing air temperaturs overnight) by more than you reduce the amount of heat provided (through worse COP when reheating from the air being cold)

If the radiators really are almost 9kW “at 50C” and your heat loss is really only “about 6kW” at design condition then you should be running them barely lukewarm at the moment and getting materially better than COP 3.3

[matt-drummer (Matthew Gregory)]

The Ipswich one is mine.

Yesterday was a one off day. I turned the heating off whilst I was at work so that when I came home I would be able to change the setting form fan coils to radiators without waiting for a heating cycle to finish, I wanted to heat from a cooler starting point inside to see how it performed.

My schedule of room temperature is set at 18c from 21.00 to 05.00, 21c from 05.00 to 13.00, 22c from 14.00 to 18.00 and 23c from 18.00 to 21.00.

No big jumps and the house isn’t falling below 20c at night very often.

It just didn’t behave as I expected yesterday.

I heat the DHW at 14.00 each day, or that is my schedule. It automatically reheats. I run my disinfection cycle at 04.00 on a Friday as it uses cheaper electricity and none from my battery. My SCOP is distorted by the disinfection cycle as the immersion heater electricity use is reported but not the heat produced.

The radiator setting has a set delta t of 8c, it is fixed at this using the radiator setting.

My DHW is on an eco setting at 45c

From my data I can see that yesterday between 09.00 and 10.00 my house was absorbing 10.6 kWh of heat at a flow temperature of around 30c.

Maybe I am interpreting the data in the wrong way but it looks like my radiators are more than capable of delivering all the heat produced, even at the minimum my heat pump can produce?

[dave_walker]

@markocoheat

Big nigtht setback (so hammering the unit hard in the morning at poor COP instead of letting it tick over all day) and charging up DHW at the coldest part of the day instead of the warmest part of the day when the sun is shining etc.

Important to remember that the most important thing is cost. I run my DHW overnight as my electricity is 7.5p KWH so at a COP of 2.5 that is 3p per KW of heat. I would have to achieve a COP of over 8 in the day to get the same cost. Add to that my main hot water demand is in the morning for showers etc so would lose even more efficiency letting the hot water in the tank cool overnight, heating it and using sooner has got to be more efficient.

[matt-drummer (Matthew Gregory)]

Yes, the cost together with the comfort in the home are the most important factors.

I don’t intend to get too hung up on the COP but having the data makes it hard to resist!

I looked at quite a few other installations.

What I am pleased about is my overall setup.

I compared to one other house, a mid terrace and on the same day with similar outside temperature all day my house used the same amount of electricity to heat the house and the hot water. My house is detached and 60% larger floor area with a heated conservatory.

Whilst the COP on that day for the other house was 4+ and mine was only 3, the overall energy consumption was the same, that’s not a bad result.

But I still think my COP should and could be better.

[markocoheat (Marko Cosic)]

Ok. We can watch what’s happening.

@matt-drummer

My schedule of room temperature is set at 18c from 21.00 to 05.00, 21c from 05.00 to 13.00, 22c from 14.00 to 18.00 and 23c from 18.00 to 21.00.

No big jumps…

On the contrary.

Your heat pump and readiators are spending most of their time doing sweet f**k all; then it’s tipping in all the heat at quite a high target water temperature:

Try setting to 22C 24/7 with a lower curve.

@matt-drummer

I heat the DHW at 14.00 each day, or that is my schedule. It automatically reheats. I run my disinfection cycle at 04.00 on a Friday as it uses cheaper electricity and none from my battery. My SCOP is distorted by the disinfection cycle as the immersion heater electricity use is reported but not the heat produced.

Kill the disinfection cycle. Not required.

@matt-drummer

The radiator setting has a set delta t of 8c, it is fixed at this using the radiator setting.

Evidently that’s the target dT that the unit has decided represents the best balance between improved compressor performance and pumping cost.

@matt-drummer

My DHW is on an eco setting at 45c

Good choice. Now kill the disinfection cyce. Not required on domestic with high turnover of water. Legionella can’t grow fast enough to be a concern before you use that water.

[quote=“matt-drummer, post:46, topic:24589”]
From my data I can see that yesterday between 09.00 and 10.00 my house was absorbing 10.6 kWh of heat at a flow temperature of around 30c.[/quote]

And a DHW reheat?

Yes to 30C:

Now heat for longer at a lower flow temperature and you’ll get an idea of ust what the Daikin unit is capable of. For example the beginnign of the reheat period you’re up at a COP of 5. Spread that throughout the day and you caan afford to waste 20% more heat from the home (due to increased ventilation losses) before it costs you a penny extra in elextricity vs a COP of 4…

@matt-drummer

Maybe I am interpreting the data in the wrong way but it looks like my radiators are more than capable of delivering all the heat produced, even at the minimum my heat pump can produce?

I would agree with that. Looks like 3.8 kW @ 35/30 is the minimum ouput of that unit:

Perhaps your main issue holding back efficienchy in use is in fact the aggressive setback and the unneccessary immersion heat?

[markocoheat (Marko Cosic)]

@dave_walker

Important to remember that the most important thing is cost. I run my DHW overnight as my electricity is 7.5p KWH so at a COP of 2.5 that is 3p per KW of heat. I would have to achieve a COP of over 8 in the day to get the same cost

That’s a fictional cost - the suppliers is hoping that by offering artifical / fake price variation they win more business through “OMG look how hceap heat pumps are and how amazing batteries are” than they lose through some peopel successfully gaming the game.

Whilst it lasts you are doing the correct thing - economically / personally - by sacrificing efficiency to play the supplier at their pricing game.

Environmentally it is less ethical. As unethical as battery storage. (work out the emissions math on charge/discharge vs exporting and displacing gas fired generation elsewhere, and you come to the conclusion that Putin would endorse peopel buying batteries)

[markocoheat (Marko Cosic)]

@matt-drummer

But I still think my COP should and could be better.

It could and should.

Start with the dumbest possible control.

Introduce complexity only if it actually helps.

Al these smarts and mucking about with setbacks, for the most part, waste energy when it comes to heat pumps.

The only one that’s worth doing environmentally, or economically as UK plc, is avoiding running at say 4-7 pm (or whenever the notification is) during winter, so that you avoid adding to grid peak and causing less efficient peaker plants to run.

Otherwise set and forget is hard to beat unless there are OHER reasons for wanting it.

2-3 kWh/day here on a £1500 air2air unit set to 20C during waking hours in an averagely crap 75 m2 end terrace.

The setback is done for comfort rather than performance/efficiency. Too much heat leaks upstairs to bedrooms when it is mild out. When it gets cold enough it runs 24/7 and enough heat is lost from roof to compensate for the heat coming up the staircase / from the ground floor ceiling through the floor to ensure that the bedrooms remain cool enough to be comfortable at night.

Edit:

Handover pack received privately via email. Excerpts returned via DM.

Apart from the property address, a radiator schedule, and a note that the design flow temperature is 50 degC it contains nothing much of any value.

It doesn’t give system design parameters (flow/return temps to the radiators, flowrates for each radiator), there’s no evidence that the radiators were balanced etc, and the commissionign sheets for the heat pump whilst included aren’t fileld in.

It DOES show that the radiators have an output of 8.8 kW at the 50C design tempeature. If this is the case, and the heat losses are more lik 6 kW at the 21C living room / 18C bedroom design temperature that they have deisgned to, then you should be able to set a curve qute a bit lower than that secified and enjoy a higher sCOP.

If you like the rooms at 23C (probably because the walls/floors/ceilings etc are all cold due to the night setback; which will make you feel colder for a given air temperature due to convenction currents and radiaton) then you will need a higher output though.

The performance estimates are also only valid for pur weather compensation; not any of this setback nonsense that the unit is currently configured for. The handover pack makes no attempt to explain any of this.

[Timbones (Tim)]

@markocoheat

Now kill the disinfection cyce. Not required on domestic with high turnover of water. Legionella can’t grow fast enough to be a concern before you use that water.

Seconded. Here’s the science…

Legionella and Water Temperature: What You Need To Know

Legionella cycle is doubly bad for performance, as the next heat pump DHW cycle runs against a hotter than normal tank which drops the COP right down. HP is most efficient heating up a cold tank, so reheating every 1 or 2 days is better, depending on domestic usage.

Most systems measure the temperature at the bottom of the tank, so might appear to be “cold” when there’s still plenty of hot water coming out of them top.

[matt-drummer (Matthew Gregory)]

So I have turned off the disinfection cycle, switched to pure weather compensation and the pump to sample mode.

Well see how it goes.

[John (John Gouk)]

Your analysis of the Handover pack matches that which I’ve just done on mine. I’ve looked at the Stelrad output spec for my rads, the design temps and the effective output at various dTs (room vs rad dT) and I reckon that for the -2.2 deg outside temp, the HP flow/return is intended to be 50/45. I imagine that mark-drummer’s is the same. My radiators have an output of 9.8kW in total, and the calculated heat loss is 8.4kW, presumably at the -2.2 deg design temp. But this isn’t a very airtight house, so it’s probably higher :-/.

[matt-drummer (Matthew Gregory)]

So now I am on pure weather compensation with no other influence.

The heat pump is using 900w of electricity to produce 3kWh of heat.

That looks like it is as good as it gets.

Outside temperature is around 10c and the internal temperature is steady, so working OK. Flow temperature is about 30c.

Looking at other heat pumps they are only using 600w to produce around the same amount of heat.

I can only conclude that this Daikin heat pump is inefficient compared to something like a Vaillant.

Am I wrong?

[markocoheat (Marko Cosic)]

Hmmm!

The COP is indeed nothing to write home about even though it’s warm out and the flow tempreature is very modest.

If I interpret the datasheet that @Vinny posted correctly you ought ot be seeing a COP of 5 or so in these conditions:

Either:

• Something is drawing more electricity than it ought to

• Something is putting less heat out than it ought to

• You’re not measuring electricity accurately

• Youre not measuring heat accurately

• The performance of the heat pump really does fall off a cliff at partial load (3 and change kW, rather than the 5ish kW that the datasheet is rated for)

• I’m completely wrong in my expectations on how that unit should perform

Interesting.

I don’t have any thoughts as to which of the those is most likely.

Ask Octopus?

[matt-drummer (Matthew Gregory)]

Thanks Marko.

In reference to my topic title this is why the monitoring is so useful.

I have at least three ways of confirming the electricity consumption to verify it’s accuracy, I have these meters, another meter on the feed to the heat pump and data from the Daikin MMI controller.

The Daikin MMI controller also reports heat produced but I need to be at home to see it and I’m at work now.

Nothing I have done in the last week has made very much difference to the COP, it is more or less the same if I run it as fan coils, radiators, WC with Madoka influence or pure WC.

It seems to be set up really well in terms of the weather curve, the flow temperature is reasonably low and the house is a steady 22.5c according to the Madoka unit. The temperature in the house is also constant according the OpenEnergy thermostat I am using here.

I don’t think it’s me, I think I know how to set up a curve and make it work.

So, as you say, I am either reporting incorrect data somewhere or the heat pump isn’t performing as we might expect for some reason.

A quick check of my three sources of measuring electricity consumption confirm that the consumption is as reported.

I will check what the Daikin MMI says about heat produced.

I think my heat meter is installed properly, I see no flow problems and I don’t hear any noises that sound like trapped air.

Maybe after partially draining the water to fit the heat meter I have failed in some way to fill it up properly and that is causing the poor performance?

[markocoheat (Marko Cosic)]

@matt-drummer

In reference to my topic title this is why the monitoring is so useful.

Absolutely. OEM is fab!

@matt-drummer

It seems to be set up really well in terms of the weather curve, the flow temperature is reasonably low and the house is a steady 22.5c according to the Madoka unit. The temperature in the house is also constant according the OpenEnergy thermostat I am using here.

Agree.

@matt-drummer

A quick check of my three sources of measuring electricity consumption confirm that the consumption is as reported.

Can you sanity check what is in scope? That you’re not picking up something daft alongside the heat pump? (clutching at straws)

Can you move any of the meters easily out of interest? For example to measure JUST the heat pump controls / compressor rather than heat pump and circulator pump?

There does seem to be a fair old circulator pump power draw vs what one would expect. (assuming that the 125W whic sometimes appears after compressor-stop is the circulator)

This, if unresolvable, might favour running the unit hotter for a shorter period even if that’s worse for the compressor. Doesn’t feel right though.

@matt-drummer

I think my heat meter is installed properly, I see no flow problems and I don’t hear any noises that sound like trapped air.

Photos as a sanity check? (for location as much as anything else)

I would not expect an ultrasonic unit with air in to register any flow at all (they ought to trip out and register 0 if they’re ultrasonic units)

Yours is a vortex-shedding unit. It’s oblivious to air, oblivious to dirt, and oblivious to being installed backwards even. (it can’t tell the difference between vortices being shed from accurate geometry vs inaccurate geometry caused by air/dirt/backwards install)

The heat output is rather higgledy-piggledy compared with others with the Sontex meter; which I can’t make sense of unless the unit it modulating the circulator in a higgledy-piggledy manner:

[matt-drummer (Matthew Gregory)]

I just did a rough calculation of the heat produced in a stable 30 minute period, not particularly scientific but the best I could come up with.

The flow was steady at 24lpm and so I could calculate the change in temperature over time of the water mass and came up with exactly the same as reported.

I will load a picture of the installation when I get home later.

The circulation pump was using even more until I turned it down to it’s minimum of 60%. The standby consumption of the heat pump is 23w and when the pump was running with no heat produced the consumption was 175w or so.

I can’t see why the pump is running so hard and producing no heat, seems like a waste?

Running like it is now is not ideal for me, the pump/water circulating is quite noisy, don’t know if that is normal. Running all the time is noticeable.

I have three electricity meters, one fitted to each of the supplies to the heat pump. The first one is the heat pump power including the circulation pump, the second is the immersion heater and the third is a booster heater inside the heat pump. The electricity use reported is only from the first meter apart from during the disinfection cycle when use on the second meter is recorded. There is no consumption on the booster heater so it is not inadvertently using that to provide some of the heat.

I still can’t get away from the heat being produced by the heat pump at the electricity consumption reported. Surely it should be higher and then shut off sooner at those flow temperatures if my radiators cannot handle the heat produced at that low a flow?

It just looks like the heat pump isn’t producing much heat compared to the electricity consumption.

If I increase the flow temperature now by moving the offset on my WD the house will just get warmer I assume?

[Timbones (Tim)]

@markocoheat

There does seem to be a fair old circulator pump power draw vs what one would expect. (assuming that the 125W whic sometimes appears after compressor-stop is the circulator)

For a comparative data point, my Mitsubishi heat pump consumes 120 W when running its 2 circulation pumps and they run continuously between cycles. While this is undesirable for COP, it doesn’t have a massive impact. My morning heating ran with COP of 3.9, vs. 4.1 if I ignore consumption below 200. 5% difference.

[matt-drummer (Matthew Gregory)]

So I have been looking at another 9kW Daikin Altherma 3’s data

This morning, at the same time and same outside temperature I see the following

My heat pump is consuming 926w and producing 3,520w of heat at a flow temperature of 30c and a return of 27.8c

The other heat pump is consuming 988w and producing 4,556w of heat at a flow temperature of 38c and a return temperature of 32c

The electricity consumption is almost the same with the difference being the flow and return temperatures.

It is interesting that there is a bigger difference between the two flows and returns.

I can’t run my heat pump on pure weather compensation at 38c, it will be like a sauna in here, it’s already 22c inside running at 30c flow.

As Marko has noticed, my heat output is up and down compared to others.

I am going to increase my flow by the WD curve offset and see what happens.

If that improves things then I will have to reintroduce Madoka room temperature control.

A couple of hours later and it hasn’t improved the COP.

I have just had my Octopus quality of installation audit.

We went through the performance of the heat pump and he said the best SCOP he had seen on one of their installs was 3.60.

They are going to send somebody to look at it and go through the setup.