Attributing solar energy to individual devices to help measure how changed behaviours can improve carbon footprint

I am trying to assign my solar production to appliances that i am tracking the energy of. I saw another post that said its not possible, but i feel that with all appliances using power at different parts of the day, and solar having different production over time, there would be quite big differences. For example, lighting would use very little solar, as its on at night. Conversely, our washing machine is run during the day. So the washing machine is greener. And this could help the family change behaviours by using the house differently (or even automate)

I have created a template helper to track the “green ratio” over the day (looking at the proportion of consumption that comes from solar.

{% set pc = states('sensor.solaredge_power_consumption') | float(0) %}
{% set gp = states('sensor.solaredge_grid_power') | float(0) %}
{% set sp = states('sensor.solaredge_solar_power') | float(0) %}
{% if gp < 0 %}
  {{ 100 }}
{%- else -%}
  {{ min (sp / pc * 100, 100) }}
{%- endif %}

This shows that for many hours, the house is 100% green (ie no grid imports).

The idea is to line this up with different appliances, and apply that ratio to their consumption. This needs to be done at regular time intervals (as you can’t just apply a daily figure). I was thinking that 15min intervals would be OK (1 hour may be too course, and 1min may be too much data?)

Here is a chart of the green ratio and a couple of appliances with different time profiles.

From this we can see the laundry is a lot “greener” than the hot water, as it was used when solar prodution was much higher. In times where there is both solar and grid (ie green ratio <100%), i plan on just pro-rating across everything consuming at that time.

So i am at the point where i have energy for each appliance - i think i need to chunk this up into time blocks. Thinking a 15min or 1hour utility meter would create the blocked data. (the apex charts has group by, and does this automatically - wish i could access this data). Then i would apply the ratio. Then i could chart it, and also would re-sum across the day to see the best and worst uses. All of this is a bit like the Energy dashboard individual device view, but with a solar overlay.

This feels pretty inefficient. I would be creating multiple template helpers and energy meters for every appliance. Has anyone done this before, or is there a better way to apply this flexibly (maybe repurposing a tariff?)

Or use the tariff feature of the utility meter to measure the green/non-green energy consumption of each device.

1 Like

No you don’t. Because you have batteries on the system each devices ‘solar percentage’ will be the same as the battery.

Because the battery stored. All unused energy (assuming yours is not topping off and sitting idle full) all goes into the battery - all comes out of the battery. That’s where percent is measured and therfore all devices drawing that pool have the same ratio.

You don’t use more ‘solar’ just because of when you happened to run the device.

I get what you’re trying to encourage but it’s not technically accurate.

I don’t have batteries in my system so I think the flow concepts do make sense. I agree that if there is energy stored in the battery, the situation is quite different and any attribution is probably arbitrary. Not sure what your thoughts are about a system without batteries, and how this might be achieved?