Data discrepancies on Daikin units

Hardware Heatpump
1

Good evening all,

I am seeing that the Daikin system in Wokingham seems to have some interesting readings and COP - Emoncms - app view

When my unit Farnborough, Hampshire (same EDLA08E2V3) is idle I see 16w of power usage but for the Wokingham unit I see only 9w.

When the pump is running other units show 75/100w of usage @ 16-20lpm but this unit shows 40w@16lpm

There also seems to be a lack of compressor pre heating that is usually seen with other daikin units.

Just wondering if the owner is on the forum anywhere and if there are any configuration changes I can make to optimise my unit further, if there is a Daikin fw update available or if some of the energy usage is not being reported and the SCOP is not fully accurate.

Having played with my unit and the 9kW beast before I am keen to hear any configuration guidance to optimise this unit.

Thanks in advance.

1 Like
2

Hello Zak,

That’s my system in Wokingham. Power is measured with a Shelly 3EM clamp meter flashed with esphome using the shelly calibration data extracted prior to flashing:

Idle power can be as low as 9W but is usually in the 12-13W range. I’ve reduced the brightness of my thermostat to minimum as well as the blue LED, perhaps that has helped? The clamp could well be under reporting though as my COP is a lot better than I expected it to be. I have another uncalibrated clamp meter that I can calibrate and attach to the mains feed for comparison.

I’ve read through your journey and used a lot of the settings you’ve spoken about, it’s unlikely that I have done anything drastically different.

3

Hello Ali,

Thanks for replying to Zak.

I have had a similar journey to Zak, a bit of a nightmare!

Can I ask a few questions?

Your data shows only 40w at a pump speed of 16lpm, mine always uses 75w and Zak’s uses 100w at 20lpm.

Any explanation?

Your heat pump doesn’t appear to be preheating the compressor, all of the others do every time, even when short cycling.

Again, any explanation?

How do you manage to get your heat pump to run at close to 20c outside?

I can’t do that, it’s just not possible for me or anybody else I can see. Not that I want to do it, that would be mad for me as the house would be like an oven, I can’t get mine to run anywhere near that.

Do you have overshoot adjustment in all three heating modes?

What heating mode do you use and what dT if in fan coils or ufh?

How big are your radiators, that is total output at dT50c?

It says Octopus designed to run at 50c at -3c outside, did they really?

You must have a missive radiator capacity? But when it is colder the data suggests otherwise, it’s a bit confusing.

With a heat loss of only 5kW at -3c how does your house accept so much heat at 15c and above without becoming a sauna?

You also say you have a buffer tank, do you?

Or just a volumiser.

Sorry for lots of questions, we’re just interested nerds!

Regards,

Matthew.

4

No problem, I’ll answer what I can.

I’m not 100% confident that my energy monitor is accurate. It is supposed to be pre calibrated, but it could be under reporting. I’ll get my alternate clamp meter up and running to compare tomorrow hopefully.

I wasn’t aware of the preheating thing, I can’t explain it. Can you show me an example of a system preheating?

I haven’t done anything very much different to you two I don’t think. I set underfloor heating to get DT3 and the ability to set overshoot to 2. I started out with a leaving water temp of 26 to dial in the curve and use home assistant to adjust the offset up and down. Currently the daytime offset is +5 and night time +4.

Whilst Octopus changed ten of thirteen radiators from type 11 to 22, I was able to swap the remaining three and bumped two up from their original size (type 21 to 22 and type 22 to 33). I’ve also balanced them properly. Total output before the last three were swapped was 7556W. edit I think that’s at DT26.5 or whatever Octopus use, it’s what’s printed on the radiator schedule

My radiator schedule shows a design temp of 50 degrees at -2.2 but I’m sure their air change calculation is way off. Other than the cat flap the house is draught free and decently insulated. Cavity insulation was increased on April 17th from a partial fill to a full fill.

My flow and return pipework is quite lengthy, I suspect when it gets colder I’m losing heat there. Perhaps that explains why when it gets colder performance drops quite a bit.

My heat loss could be wrong, one of the plumbers told me and it was either 5.6 or 6.5 I forget exactly. I’ve adjusted to 6 for now to avoid confusion and checked the data issue flag until I confirmed power usage.

I have a volumiser in the loft, the standard 20L Octopus one. edit I’ve removed buffer from my profile and left volumiser checked.

BTW Matt I read through your giant thread when you had issues with the 9kW unit. I was also supposed to get the volcano unit originally and managed to convince Octopus to fit the 8kW unit without any pushback. Many thanks!

2 Likes
5

Hi Ali,

Thanks for the reply.

I’m glad you avoided the 9kW, good work!

All of these 4/6/8kW heat pumps preheat the compressor, even when short cycling.

Look at any of them, mine or Zaks as they are MID metered just to be sure.

You will see electricity consumption every time before heat is produced, yours goes straight to heating.

Your heat pump will be preheating, they all do, something is wrong.

So you had 7,556w of radiators, I guess from your radiator schedule from Octopus?

If so that will be at dT30c, that’s about 15kw at dT50c and maybe with your upgrades, maximum 20kw.

That is less than Zak has, I think he has about 30kW and I have about 50kW.

Is a lot of your lengthy pipework outside?

It shouldn’t affect your dT and therefore your heat production data but it might explain where the heat is going?

You have your dT set to 3c on ufh?

From your current data your dT between flow and return is about 5c

Octopus normally set the emitters to fan coils, you changed it?

I think I am starting to understand.

At times the data looks quite reasonable, in line with mine, a heat loss of around 5kW and with what I would have assumed were large radiators.

But it didn’t make sense when it was colder or warmer, then it gets very different to me.

Something isn’t right with the data if you have it set at a dT of 3c an you are showing 5c.

And it may be that a lot of your heat is not reaching the house?

That would explain how you can run continuously when it’s warmer and the rest of us can’t and why your performance drops when it is cooler.

I think you need to do some checking.

Not to satisfy me or anybody else but I suspect your efficiency is nothing like you are showing.

If your radiators are about what I said there is no way your data is correct.

I have never seen Octopus fit enough radiators, they always design to 50c at -3c.

I only have what I have because I changed them all to make my 9kW work.

I had about 18kW of radiators post install, about the same as you.

Zak was the same and convinced them to go bigger when his heat pump was changed.

I did mine myself as I really wanted to make the 9kW work, and I really did try!

Your house is bigger than mine and Zaks so your heat loss may well be a bt higher.

I am heating an all glass conservatory though which is a bit of a problem at times, but does effectively make my house a bit bigger and closer to yours because of the heat loss.

I think you need to check our electricity and your dT.

If you can it would be useful to check the dT where the flow and return enter the house.

6

Here is the start of one of my heating cycles

image
image1137Ă—729 44.8 KB

And this is yours

image
image1171Ă—728 60.5 KB

There is no preheat in yours, it produces heat as soon as electricity is applied, even before.

All the others in this range of heat pumps are like mine.

And here is your pump running at 16lpm consuming 38w of electricity

image
image1123Ă—735 31.1 KB

And mine doing the same uses 75w

image
image1113Ă—745 36 KB

7

So if I assume you have a maximum of 20kW of radiators you are currently running at 31c flow and 26c return with a room temperature of 21.5c

That is a mean flow temperature of 28.5c and a dT between radiators and room of 7c.

But it will be less because of your pipe run, the flow at the house will be lower than at the heat pump.

At a dT of around 5c your radiators will put out 1,000w of heat.

You are producing about 2,500w of heat according to your data.

I suspect less than half of it, if it is even true, is going into your house.

That might explain why the heat pump can run when it is warm outside, why your house doesn’t overheat and how the return temperature remains under control.

I’m guessing you are heating your garden!

8

Looks to me like there’s some lag in the heat measurements of your system, as electrical power starts before heat and also stops earlier. Perhaps due to the distance between the compressor and the flow temp sensor, and the flow rate.

Screenshot_20240502_235131
Screenshot_20240502_235131720Ă—625 93.7 KB

I.e. the heat power graph is shifted to the right.

This probably won’t affect overall COP.

9

Hey Ali,

Thanks for reaching out and good job on avoiding the 9kW beast.

My system specs are available in the post here - Struggling to get good COP from Daikin 9kW - #18 by KnightPhoenix

A few minor changes from the new rads page is the DT of 3 and the design flow temperature of 38 (meaning the bathroom is slightly undersided).

I have just over 20kW of radiators following the changes but would like to increase this in the future (by adding UFH downstairs and keeping a few smaller radiators to get to a 30°c flow temperature), the future being when the good lady allows me to start doing more renovations again following having the house in bits for weeks at a time for various projects.

As you may have guessed Matt and I have been at this for a while and are trying to figure out his these Daikin heat pumps should be configured to actually work at a reasonable efficiency level (although Matt’s radiator solution is somewhat more impressive than my system).

I fully get that different meters may be a bit out and the data from ESPAltherma is not always 100% (I use P1P2SERIAL for my DHW cycles and internal temperature reference and it has it’s moments).

I did have a play with the madoka room stat led but it got a bit too dim tonread. I will take another crack at that next week or over the weekend and see if that can save a few watts (shame it doesn’t have an auto off after x seconds settings).

Just out of interest where is your clamp meter connected, is this inside one of the consumer units or inside the ASHP as it could just be there is power being taken to another part of the system before it gets to the shelly.

Having spent soo long taming the 9kW beast this becomes somewhat addictive and any potential to squeeze a bit more performance out of this unit is something I am keen to investigate and review.

Thanks for all of the info on your setup and if you have any questions or want to throw a few ideas about don’t hesitate to shout. The more everyone collaborates the more we all win.

10

I can see the delay tim and as you have pointed out with a pipe run between the ASHP and the heat meter this delayed start is pretty much expected.

The oddity here I believe is the hump at the start of the cycle (the pre heating hump) in most of the Daikin units you see a short spike in both the electric usage and the heat output before the system settles down (this has been the bain of many discussions and seems to have lead to a lot of cycling at lower temperatures).

Just wondering if Ali had something switched off or configured differently that may allow us to get a nice steady startup that his unit seems to be experiencing as it looks and seems to perform much nicer than way.

11

The only issue here is the numbers we are dealing with are soo small that even a virtually un noticeable dip in house temperature of 1/2°c taking the system from dt7 to dt8/9 can make quite a difference to the power output from a radiator. At dt5 on a 20kW system the max output is 1kW at dt10 you are at 2.5kW and that also depends on the heat being radiated from radiator circuit pipework and any other heat being unintentionally exported into the house or loft (fixed a big of that myself too).

If you are up for investigating further Ali I would love to know what the MMI export shows as this should allow you to see the whole system configuration (and compare it to mine and others on the forum). If you do have a cheat code / field setting enabled to turn off the preheating then that is something a few of us would be very keen to review further :+1:.

12

Thanks for the input chaps, lots to work out here it seems!

@matt-drummer the pipe run is quite long, but is well insulated. Where the team left a few gaps I was able to cover up and stop any potential water ingress. I have 15mm plastic upstairs and 10mm copper downstairs, I suspect that restricts my radiators ability to output what its supposed to. That’s mostly why I set DT3 to try and get the pump speed up. Though with overshoot it seems that I’m getting DT5.

I did wonder whether the pump temperature sensors might be reading high/low, but when the pump is running without any heat input both flow and return read exactly the same so I assume they’re not terribly out. I would like to test flow and return temps at the point of entry into the house, not sure how to do that accurately though.

@KnightPhoenix the clamp meter is in the consumer unit next to the pump outside. I’m planning to add the alternate clamp at the meter where the pump feed splits off from the incoming supply.

The missing preheating is rather puzzling, I’ll dump my MMI settings for you guys to check out, just need to find a usb stick.

13

Looking at last nights heating run, it appears to have pre-heated:

image
image1963Ă—871 94.7 KB

1 Like
14

Hi Ali,

If your dT is set at 3c on the MMI then there is only one way the heat pump could be running at a dT of 5c and that would be if the circulation pump was running flat out and the radiators were still losing more than 3c.

I can’t think of any other way it could happen other than data issues.

How long is the pipe run and where does it go?

In simple terms this is what I think might be happening.

The water is leaving the heat pump at 31c and arrives at your house at 30c. It goes around your radiators at 30c and leaves the house at 27c but by the time it gets back to the heat pump the return has lost another 1c so arrives back at the heat pump at 26c.

You see a dT of 5c at the heat pump but in the house it is only actually 3c.

If you are producing 2,500w of heat at the time then something like 1,500w is going to the house and 1,000w is lost to the air outside.

It doesn’t take to much of a temperature loss on your pipe run on both the flow and return to make a massive difference.

Apart from any other issues, this is one of my problems with EspAltherma, it is not guaranteed to be measuring heat actually delivered to the home because the flow and return are not measured at the point of entry to the home.

Your actual temperature sensors look fine as they converge together with the pump running, the absolute values may not be 100% accurate but they are matched nicely it seems.

I also don’t see any issue with electricity most of the time. On the occasions when I have looked and seen similar conditions between your heat pump and mine, the electricity consumption for the heat produced at the same flow and outside temperature agree.

I think there is some issue with the electricity meter as the consumption with just the pump running is too low compared to all of the others.

But I think your big issue is losing temperature in the external pipe runs, the outside is acting like a big radiator and you are not delivering the heat to your house the data suggests.

It explains pretty much everything and fits with you radiator sizes.

15

It does now seem to be preheating at times.

It looks like the electricity data is erratic and inconsistent, possibly, I don’t know.

16

Here’s the export:
db-export.txt (7.2 KB)

17

New clamp meter has been calibrated and is logging to emoncms since ~11.40am. It does appear to be reporting a higher wattage and the frequency of reporting to home assistant is a lot greater. Hopefully this will give more accurate stats.

1 Like
18

Getting a few cycles now, pre-heating is being captured. Thanks for the info guys! I’ll have to 3d print a case for my alternate clamp meter which is why I bought the Shelly in the first place. Oh well.

Not sure what to do with the last months data, do you think I should dump it and start again?

19

Hi Ali,

I would keep your data, I can’t speak for everybody but I suspect most of us don’t care.

I am still interested in what your dT is set at.

There is a problem if it is set at 3c but your data is showing 5c or more.

If I were you, I would want to know how much heat I was losing outside, if any?

How long is your external pip run and how well insulated is it?

From what you have said and the data I have seen, you could find yourself with a problem when it is really cold.

When was your heat pump installed? The beginning of April or before?

20

I was mistaken, dT is set to five.

I’m not sure there’s much more I can do with the external pipe run. The plumbers left some “glue” and tape and I was able to fix up a few areas where the insulation wasn’t quite fitted right around the pipe clips but that was it.

The external run is about five meters, probably a bit more as it wraps around a chimney breast. I can certainly improve the insulation to the volumiser in the loft as that’s just plastic pipe clips and grey insulation.

Install date was the week beginning the 18th of April.