Observations on my ...
Clear all

Observations on my Ecodan's estimates of energy consumed and delivered

1 Posts
1 Users
Trusted Member Member
399 kWhs
Joined: 1 year ago
Posts: 24
Topic starter  

Comments I made in the thread 'Electricity use for heat pumps - comparing homes and consumption' seemed to merit a thread of their own. So here it is, after a short break while I stripped out the bathroom (and then waited for March data).

The short version is that my Ecodan invariably overestimates its electricity consumption. But it does this inconsistently which leads me to assume it's estimates of energy delivered may also be unreliable. So unless it's estimates of energy consumed and delivered are inaccurate but consistent with each other, any Coefficient of Performance derived from either or both of them is probably also inaccurate.

I've read plenty of people observing this inaccuracy. Here I'm quantifying it (in terms of energy consumed but not delivered). I'd be interested in the experience of other Ecodan owners- especially if they have third party measurement of energy consumed and delivered.

The long version is that my 11.2kW Ecodan was installed in the Summer of 2022 with an FTC6 controller. At the planning stage I specified that I wanted a separate meter installed so I could monitor energy consumed. In February 2023 I started keeping comprehensive daily records to help me judge whether changes I made led to any improvement in performance efficiency. Up to that point I had only kept a note of the figures from the consumption meter (which is accessible- unlike the FTC6 which entails burrowing at the back of the airing cupboard to read). My daily records since then cover energy consumed and delivered as 'measured' by the FTC6 controller, consumption measured by the meter and daily maximum and minimum outside temperatures (from my home weather station which is broadly consistent with the Ecodan/Melcloud OAT reading).

The attached table only gives the % difference between the energy consumed month-by-month as measured by the meter and by the Ecodan. In every month the Ecodan overestimates consumption. I'm not sharing energy delivered in this thread as I don't want it to become a 'how to improve efficiency' discussion- we've already got plenty of those! And I lack confidence in the accuracy of the energy delivered figures reported by the controller. Given the widely varying difference between my two sources of energy consumption data any CoP figures derived are unlikely to be reliable or accurate. Note that energy consumed by the immersion heater isn't included in these figures.

Observations and thoughts:

  • consistent over-reading: Viewed at the monthly level, the Ecodan figures always overstate energy consumed. At the daily level the picture isn't so absolutely consistent but there were only 35 occasions in time covered by this data-set when the metered energy consumed exceeded the Ecodan estimate.
  • data factors: The consumption meter gives figures to two decimal places while the Ecodan forecasts are all whole numbers. It naturally follows that 'if' the Ecodan rounds its forecasts (as opposed to only ever counting in whole numbers) then there will be a degree of inconsistency. But this would only ever be less than 0.5 kWh and even at the daily levels the inconsistency between the two data sources is usually greater than this.
  • I turned off the heating manually for the summer. Despite this the Ecodan still 'allocated' some of its energy consumption to heating. I haven't reworked the data to explore what happens if I treat the heating consumption as zero.
  • based on my temperature recording there is good correlation between energy consumed and OAT at the daily level.

Conclusion ?

The over-estimating of energy consumption is worst when consumption is low. There seem to be two main contributory factors- rounding and Ecodan estimates of consumption by the heating side when it is actually zero in the summer.

The uncertainty of reported energy consumed and delivered figures make using daily CoP figures a very unreliable way to measure the effectiveness of small changes to things like a weather compensation curve. Weekly figures are a better basis.

Surely things should be better than this!

IMG 1920

SKD and Mars reacted


Join Us!

Latest Posts

Members Online

x  Powerful Protection for WordPress, from Shield Security
This Site Is Protected By
Shield Security