Be aware that the CO2e emissions shown in this app give an indication but have pretty big caveats:

• The CO2 emissions of nuclear that are shown are most likely too low. Electicity Maps just used the absolute lowest of a number of estimates.

• The emissions of reservoir-based hydro vary depending on climate and biomass in the reservoir but Electricity Maps uses a single standard value. Some hydro power in warmer climates is connected to methane emissions from the reservoir, because of anaerobic processes at the bottom of the reservoir. Some reservoirs emit as CO2e as a coal plant with a similar capacity. [edited slightly, see below↓]

• They are showing solar/wind emitting emissions when they're producing electricity. However, solar and wind don't work that way because they're not using consumables. Their emissions mainly come from manufacturing and transport. In essence, solar and wind are emitting just as much whether they are producing electricity or not.

That UI is beautiful.

Australia, what you doin'!? So much coal, what the fuck. In a country where you'd expect they have so much space and sun, they could just go solar.

A related tool for Germany: https://peakpick.de/webapp/index.php

The idea behind this:

• Electricity generation from wind and solar is not constant. PeakPick visualizes and forecasts the share of renewables.
• By choosing when to run your home appliances, you can use more of the renewable capacity, and less of fossil fuel backups

This made me aware that the best time to use electricity is around noon (+/- a few hours). By shifting your electricity consumption to this time, you can cause less emissions, regardless of your contract.

Open question: A complete picture would require plotting the consumption along the production. For example, while renewables might peak at 1 PM, maybe all that capacity is already eaten up by consumption at that time. If hypothetically consumption was below capacity at 3 PM, it might be better to consume then, even if capacity is lower then.

No data available... Yeah I doubt that

South Africa's data likely doesn't account for all the diesel burning to avoid worse Loadshedding stages.

Shouldn't there be some normalization based on population? Like some of the low population states / siberia have bad emissions but that should have much less effect than a smaller difference in a more populated location.

Kind of surprised South Carolina is so bad. Just drove by so many solar farms you'd think the state ran it alone. Apparently my random sample was incredibly biased. Also they're fairly recent installations, so maybe it hasn't filtered into the data.