The following is a rough outline of the main items on our development plan in the coming months/year. Not in any particular order. This is a working document, to be added to and amended over time, please feel free to add suggestions or comments relating to these in the development discussion thread here or in separate linked threads if easier to keep discussion on topic.
HeatpumpMonitor.org
High level goals
-
Accuracy: It’s important that there is a sizeable subset of systems on HeatpumpMonitor.org that meet the highest levels of metering accuracy. This provides confidence in HeatpumpMonitor.org data and enables better data analysis and comparison with other trial data. While a majority of systems do have MID metering, it’s still important to measure, flag and understand metering errors (e.g ultrasonic air errors, temperature sensor placement related errors, sontex dirt under-reads).
Progress is being made on these, e.g default view is MID metering, air error is now flagged, measured and available on HeatpumpMonitor.org system list. We have a manual data issue flag and note that can be added to systems. Errors detection needs to be moved as much as possible from manual to automated as manual flags can easily become out of date as issues are fixed. Metering issues can also change over time, develop again e.g after further air is introduced.
Not all systems on HeatpumpMonitor.org need to meet the highest standard as there’s a lot of value in sharing data across a wider range. Default view is currently MID metering, but systems with other and heatpump integration can be easily selected for wider interest.
-
Reporting & Analysis: Previous trials have often reported disappointing low average performance results for heat pumps in the UK. Systems on HeatpumpMonitor.org are achieving much better results. How can we help share learning from these systems to lift average performance on installations across the board? While the theory of better system design is relatively straight forward, lower flow temperatures, better control, better sizing of the heat pump, pipework, heat emitters, avoiding distortion in buffers etc. What insights can we glean from the monitoring data that can feed back into design standards and processes to ensure that these results are more typical?
-
Is it possible to develop an accurate physical model that can reproduce SPF results seen on monitored systems on HeatpumpMonitor.org? E.g Heat pump SCOP calculator
-
Is it possible to develop data analysis based on supervised learning, e.g linear regression, random forest etc in order to quantify correlation between different system properties and measured performance? Inspired by John Ewbank’s work here: Heat Pump Electricity Usage Prediction Model – 10 year – John Ewbank and further discussions.
-
Specific features
- Integrate the MyHeatpump app directly in HeatpumpMonitor.org replacing the current stats overview on the system/view page.
- Implement in such a way that changes to the emoncms app are easily transferred and vice versa.
- This makes it easier to provide dashboard access without sharing the read apikey for the emoncms user.
- Potentially allows more more control of what data is shared publicly.
- Develop reporting and analysis of HeatpumpMonitor.org results including e.g oversizing analysis, flow temperature vs SCOP analysis.
- Review % carnot / simulated heat algorithm, analyse why we see such a wide variance in % carnot across systems.
- Total cost of ownership? Can we develop some example calculations and case studies on this topic.
Emoncms MyHeatpump app
- v2.9.x refinements (topic: Emoncms app module v2.9.x (database upgrade required))
- New 22nd Sep: Add button to recalculate daily stats.
- CSV export options
- More detailed stats section relating to the daily data.
- Option to show average indoor temperature and other stats on the daily view
- Use of the pre-processed daily stats for the power view when you zoom out to keep the figures on that page accurate as you zoom out.
- Option to select particular years for full year stats
- Data error identification and alerts:
- Close to zero DT between flow and return when there is significant electricity consumption, heat meter issue.
- DT between flow and return when there is flow rate and compressor is off, temperature sensor calibration/positioning issue.
Emoncms.org
- Server upgrades to larger disk capacity servers.
- Simplify options for down sampling of feed data that does not need to be recorded as high resolution. E.g kWh data could be down sampled from 10s to 30mins in many cases.
- Implement subscription billing rather than convoluted shop top up credit.
- Improve account and billing emails, feedback very welcome on this and subscription options.
- Investigate pricing based on feed interval, it’s disk space that ends up being the main cost driver and pricing could reflect this, although keeping billing simple is also important.
Emoncms
- Feed inactive notification
- Device module option to import/export devices via json, this mechanism could then be used by the sync module to make it easier to replicate accounts.
emonPi3
Please see: https://community.openenergymonitor.org/t/emonpi3-updates/26219 Thanks to @awjlogan!
emonVs mini
A lower cost single phase voltage sensor option, with separate power supply. Updated for the emonPi2/Tx5.