Inspired by discussion on the Sharing heat pump data, An open heat pump dataset? thread and @frh’s heat pump profiles example:
I’ve put together a new Emoncms app to generate a similar profile output.
The app is currently available in the master branch of emoncms for testing and is also live on emoncms.org. Here’s a live demo: https://emoncms.org/midterrace/app/view/Profile and a couple of screenshot examples of it in action:
Heat pump electricity consumption
Resolution is set to 10 minutes here, we can see fairly steady but slowly rising heat pump electric demand through the average day in January. Shoulder months start to see heat pump demand fall off overnight and summer months show no space heating demand. The larger spikes show the heat pump heating hot water.
Lighting demand only
Something quite satisfying about this one, showing lighting demand growing as the hours of day light decreases. It looks like there’s a data error perhaps in December as none of the other months show overnight consumption… or did we leave the attic light on for a whole month!?
Lighting, appliances and cooking
Cooking and overnight timed washing machine cycles showing up quite clearly here:
EV charging showing a very strong over night off-peak pattern. We dont yet have solar unfortunately
Timing here influenced by 00:30 to 04:30 Octopus Go off peak timings.
It’s interesting to see that the EV overnight charging completely dominates the overall profile. I guess that will have to change in the future when adoption gets into the high 80-90% ranges, otherwise we may invert the traditional demand curve… in the short to medium term it’s beneficial of course as it’s more likely to smooth out the overall demand profile.
Ah, that looks good.
Could it be done by year as well rather than just month?
It certainly could and would be good to have selectable years too, just a start for now.
Nice work. I like that I can chart any cumulative feed. Here’s my average solar generation:
Some minor feedback:
- it would be nice if it kept the month selection when switching feed or time period.
- it would also be nice if months with zero kWh were hidden (I only have this year)
- I’d also like to be able to chart virtual feeds, e.g. house consumption without ASHP
Good ideas thanks!
Zero months now hidden:
I’ve also added support for Virtual feeds and a greater variety of feed types such as power feeds and temperature feeds. Here’s an example with outside temperature:
And inside temperatures:
We started to make a bit more effort to save energy in March, April and May, hence lower temperatures than January and February…
Another request: show the name of the feed somewhere; maybe on or under the title?
Looks great @TrystanLea
Here’s my total house usage… Filling the Mixergy overnight via immersion really ramped up overnight usage after its installation in February.
Hoping ASHP will reduce that considerably.
Plus more overnight battery filling when no PV to rely on during the darker months.
Feature Request: Would it be possible to sort the table at the bottom (high to low)?
The profile chart makes a mess of my “House Consumption Total” feed, which is a virtual feed that subtracts heat pump and immersion from cumulative consumption. The source feeds come from two different devices, with different time base and frequency.
Thanks for this @TrystanLea! I think these sort of plots are a useful midpoint between the high time resolution detailed monitoring and higher level stats like SCOP. Plotting the temperature is also really interesting combined with the heat pump demand.
Interesting that the EV dominates the combined profile - can I ask how often you’re generally charging the car vs the hot water tank? Looks like more driving in winter (on top of lower EV efficiency)?
Tariffs like Octopus Go may need to evolve to offer different low cost times to different people if EVs are generally dominating as seen here. But someone with a bigger/peakier heat pump demand, and less driving or slower charging, could easily overwhelm your nighttime peak.
Go already has spin-off variants called Go-Faster.
GO-22-07-05 (standard Go, 4H from 0030)
Go Faster variants:
Hi Trystan, could you post link to the relevant part of the github repo plse?
It’s part of the Emoncms ‘app’ repo:
@TrystanLea Trystan - the new profiler app is great. I have used the export function to create monthly heat pump electric in, heat out and COP plots. Example attached for my heat pump in December at 10 minute intervals.
Great to see! Those are some nice and high COP’s for December!
@TrystanLea Yes - quite happy with the heat pump efficiency. Nominal flow temperature of 35degC at design external of -2degC - although the heating curve does shift up and down a small bit as I use adaptive controls.