@gfletcher 

I have a similar setup to you – an 8.5kW Ecodan alongside PV – and have noticed a similar pattern over the summer. Melcloud has reported 50-70kWh per month used in heating mode against 0kWh output, whereas Mitsubishi claim a daily use of c0.6kWh per day to keep the compressor warm, and then there will be some to run the FT6 unit though not much I would think. I haven't worried about this as a lot of use will be met by the PV and not imported from the grid. I have also noticed that the reported consumption by the HP and metered grid imports are strikingly similar amounts, meaning our normal import of around 100kWh per month seems to have melted away! Magic! It makes more numerical sense if I regard the HP as overestimating the power it consumes by 1-2kWh per day.

In the technical literature it explicity states that if you want accurate measurements then you should add extra electricity and heat meters –  these meters get wired in to the controller board and Melcloud uses those readings in its reporting. So no extra meters means relying on the Ecodan's self-confessed inaccurate estimates of energy consumption and generation – though I think generation is likely to be more accurate as long as the thermistors on your system are not dangling in mid-air as mine were after installation and are closely insulated to the pipes.

@harriup +1 - my thermistors were hanging in mid air too! Eventually I worked out which was flow and which return and then found the pockets on the pipes to insert them into.

Posted by: @hjd

@harriup +1 - my thermistors were hanging in mid air too! Eventually I worked out which was flow and which return and then found the pockets on the pipes to insert them into.

I think you both had the same installers as I did. 😆

I spoke to Mitsubishi Support in the UK re the metering. Noticed last spring COP was getting worse rather than better as things warmed up and heat pump wasn't on so often.

They said there was an issue with the consumed power metering when system was idle - it reports higher consumption than actual - 2-3kWh per day instead of around 600w per day actual in idle. They said they had done tests and standby consumption was around 25W. They did not know if anything would be done about the issue.

I don't have any external metering to verify.

Decided not to fret about it - in winter (Nov - March) heat pump consumes enough that idle errors really don't amount to much! System reports COP in winter as 3.5 - 4.5 and running costs are OK.

May investigate further by fitting a CT to heat pump supply - wondering about turning heatpump off entirely for summer and using immersion from solar for hot water. One point I've noticed is that turning off the heat pump at the controller doesn't turn it off entirely - all the sensors are still active in MELCloud - wondering if turning it off at the controller makes any difference to actual consumption?

hmmm. This is both annoying and reassuring in equal measure. Annoying in that the reported cop data is frustratingly inaccurate. And reassuring in that the energy savings of a regular idle phase probably are there but not being reported in the data. The attached 32c temperature operation with its own idle periods was still reporting 12kw per day consumption when that was what I was getting on cooler days at 40c constantly. 

Posted by: @steevjo

↑

They said there was an issue with the consumed power metering when system was idle - it reports higher consumption than actual - 2-3kWh per day instead of around 600w per day actual in idle. They said they had done tests and standby consumption was around 25W. They did not know if anything would be done about the issue.

The figures you’ve mentioned seem to correlate with what we’ve found this weekend. We have been running our Ecodan 8.5 in “stand by” while we were away over Easter and have seen that even though the heat pump never called for heat the MELCloud still reported around 3kw energy consumed each day. 

The attached report shows data of consumed energy for 7th, 8th and 9th April.  Also there is a mode of operation pie chart for the same period. Clearly you would expect some energy usage for the occasional freeze stat (antifreeze) pump circulation but we noticed that on the 9th April it was mild enough overnight not to activate the freeze stat function.

In other words no energy was used to either heat or defrost or circulate on 9th April 2023.

So to look a little closer at just the data for 9th April when no heat pump activity was recorded by MELCloud it can be seen there is still a registering of 2.8kw useage in the 24hour period. 

This is for the full 24 hour period of operation (9th April) when the auto adaptation mode was selected with a room temperature was set back to 10c so that the room stat did not call for heat (because the room temperature never dropped below 14.5c.) 

it is unclear how much a typical active heating day when there may be cycling occasionally would affect the true energy use. 

Posted by: @sunandair

↑

So to look a little closer at just the data for 9th April when no heat pump activity was recorded by MELCloud it can be seen there is still a registering of 2.8kw useage in the 24hour period. 

I have MMSP metering so am not using the internal Ecodan power meter.  My 14kW Ecodan always uses 24W when in standby; that's no freeze stat, no water pumps, just switched on. That's (obviously) a lot less.

Posted by: @kev-m

↑

I have MMSP metering so am not using the internal Ecodan power meter.  My 14kW Ecodan always uses 24W when in standby; that's no freeze stat, no water pumps, just switched on. That's (obviously) a lot less.

Thanks @kev-m that sounds like the best way to keep up to date with performance. 👍🏼

It's good that this thread is still going and from a selfish perspective I'm glad that it's not just me! Also, thanks @steevjo for posting the response you had from Mitsubishi. I have done a clumsy work around to try to get a balanced idea of COP and energy use - basically just taking previous years electricity data and removing an average daily use from the whole house electricity data and applying that to the info that comes from the unit. It is giving me COP's of between 3.5 and 4.5 instead of 2.5. 

I also now have a really good idea of winter running usage - it's working really well. Our whole house electricity use has been just under 13kWh a day Nov-April, the delivered heat and direct heatpump usage is approx 1/3 of that predicted by the heatloss survey. The house is average 18.5 degrees. All things considered, apart from the dodgy readings it has been a thoroughly pleasant experience. My neighbours are so impressed that they're looking at installing one!   

I also have a Mitubushi ecodan.

At the moment the heating is turned down to 10c with the hot water on a timer to come on for only 2 hrs a day on a low/night rate between 5 and 7am.

Below are images that shot the heat pumps usage. 
As an above post, has to be reading incorrectly??

@westkent 

I think you are suffering from the same problem described above; Melcloud over estimating consumed energy when the heating is off. 

@kev-m 

one issue I’m having is that my hot water loses very close to 1 degree an hr throughout the day.

I do have an annual service in a few weeks so I’ll mention it.