I have a 20W continuous draw, partly the controller board, but I also suspect the MelCloud wifi module is responsible for some of that. Given that the units can draw several kW of power I think that the lower current draws in particular are not being measured accurately. And Mitsubushi are quite clear that if you want an accurate measure of the energy consumed you should fit a proper meter.

I fully appreciate that an external power meter will probably be more accurate than any internal equipment.

The point that I am making is that even an external power meter will only provide truly accurate system power consumption, if all the auxiliary equipment (pumps, valves, heaters etc.) are actually powered via such an external power meter. I assume that the external power meter is monitoring the power being fed to the outside unit. If you have the pre-piped Ecodan water cylinder, the power for the control box can be via the outside unit or from an alternative power supply. If it is from an alternative power supply then the power used by all the equipment connected to the cylinder control box will not be measured by the external power meter. This could be the reason why the external and internal readings differ.

According to my interpretation of the information contained within the manual, the controller can accommodate up to 4 water pumps, and if each of these water pumps had its own dedicated power meter, the controller could accept a power meter reading for each, and collate these values along with all the other power used by the system. In the absence of dedicated power meters for any water pumps, it is possible to set an estimated value of power consumption for each of the 4 possible water pumps that could be installed. The default setting for water pump 4 is 72W, which I assume will be the value used by the controller in its calculations whether pump 4 is installed and is running or not.

I would again suggest that you check the values that have been set within the service menu, covering water pumps and any other equipment that could be using power.

I have a 20W continuous draw, partly the controller board, but I also suspect the MelCloud wifi module is responsible for some of that. Given that the units can draw several kW of power I think that the lower current draws in particular are not being measured accurately. And Mitsubushi are quite clear that if you want an accurate measure of the energy consumed you should fit a proper meter.

My surplus power 'consumed' last year was around 630kWh, so around 1.75kWh every day. Some fairly random variation either side of this, not related to consumption, and certainly not a straightfoward percentage overestimate. A standard deduction would get you closer to the figure, but if you want accuracy then the most cost effective would be a standard electric meter (c£40ish?) that has a pulse output which can be plugged into the FTC controller and will replace the units guesstimates with accurate measurements.

glpinxit

I also spent a long period at first noting down daily figures from the controller - most tedious. But then mastered looking at the code generated by MelCloud in a browser window (in my case Firefox >Tools >Browser Tools >Developer Tools> then Network and Response options) and have managed to create an entire history of the unit's performance in a spreadsheet. This is much more useful to make comprisons with.

I have just been looking at the Ecodan manual for a different reason and remembered your post. So taking your figures of 630kWh last year, that gives 1.726kWh per day, or 71.9178W for each hour.

According to the manual, Water Pump 4 in the Energy Monitor settings has a default value of 72W.

Has anyone checked the settings within their system?

Do most UK installers provide external power consumption/ delivery measurement devices? My Portuguese contractors provided no means of measurement, but then that is only one of many of their failings!

I would suggest that you look within the 'Service' menu at the 'Energy monitor settings', since it would appear that the estimated power consumption of external water pumps has to be entered rather than being measured. Obviously even a small difference between the entered value and the actual power consumption will lead to errors over time.

It also depends upon where the power for any external pumps is derived, if it is via the heat pump equipment then it should be included in the value obtained from a heat pump dedicated power meter, but if the water pumps are powered from a separate power source, which is just being switched on and off by the controller, then it will not be measured by the the heat pump power meter.

This may be a problem with most heat pump systems, since the true overall COP should include all the electrical power consumed by the heating system, but the power used by externally powered equipment may not be included in the measurement made by an heat pump dedicated power meter.

I suppose that you could try setting any values found within the 'Energy monitor settings' to zero, and see if the controller calculated power usage reduces and is more inline with any heat pump dedicated power meter.

My surplus power 'consumed' last year was around 630kWh, so around 1.75kWh every day. Some fairly random variation either side of this, not related to consumption, and certainly not a straightfoward percentage overestimate. A standard deduction would get you closer to the figure, but if you want accuracy then the most cost effective would be a standard electric meter (c£40ish?) that has a pulse output which can be plugged into the FTC controller and will replace the units guesstimates with accurate measurements.

glpinxit

I also spent a long period at first noting down daily figures from the controller - most tedious. But then mastered looking at the code generated by MelCloud in a browser window (in my case Firefox >Tools >Browser Tools >Developer Tools> then Network and Response options) and have managed to create an entire history of the unit's performance in a spreadsheet. This is much more useful to make comprisons with.

Once I started recording I quickly noticed that the electricity consumed figure generated by the Ecodan is not consistent with the consumption measured by the meter that just measures the heat pump. (And 'no' this doesn't include any immersion heater consumption.) The difference tends to be that the Ecodan thinks it eats more electricity than the meter measures- in aggregate metered consumption in 2023 was 3138kWh and Ecodan was 3590- over 14% different! I have more confidence in the accuracy of the meter though it measures to 2 decimal places whereas the Ecodan only displays whole numbers so rounding will be a particular factor when consumption is low. In 2023 the maximum daily difference was Ecodan 53% higher and 43% lower.

This is all very frustrating. I have concluded three things:

- I assume that the Ecodan's power delivered figure is as likely to be as unreliable as its power consumed

- I hope that using the Ecodan's figures for both is more likely to give me a consistent representation for whether my tinkering is beneficial for CoP efficiency.

- looking at daily data is time consuming and is not a good basis to endlessly obsess about making frequent small adjustments (to WC curve and/or pump speeds for three different pumps). I only act once week on the basis of weekly totals and not daily. 

Good luck!

(edited for clarity)

The figures from the control panel and those from Melcloud should be the same, as yours are. I was talking about using external heat and electricity meters compared with however an Ecodan works it out internally. 

It's hard to compare COPs without knowing your LWT and OAT. When it's 10 deg outside my rads are about 36 deg and my heating COP is around 4. 

We are having a warm winter in the Algarve, so I can't do much meaningful tweaking of the Ecodan Compensation curve. I have a buffer tank and I am persuaded that the secondary pump is running too fast  and although the pump is variable speed the PWM is not available for it, so at some stage I will change the pump for a more modern version.

That aside, I calculate the January CoP as: From the FTC5 = 3.37 & from MelCloud = 3.39

How does that rate with anyone else?

House is a quite large, 4 bedroom with poor wall insulation but double glazed throughout.  Large Roca aluminium rads. No information on calculated heat loss. 

I have used both on the same system and the previous internal measurements of COP were very similar to the current externally metered ones.