Hey guys, A small site that checks, every day, whether making electricity cheaper when the grid is clean would actually cut CO₂.

It reads yesterday's generation mix from ENTSO-E and the EIA, works out the carbon intensity hour by hour, and compares a normal tariff against two carbon-aware ones.

It started for Switzerland, which turned out to be a good place to start for an odd reason. Swiss electricity is already very clean — about 34 gCO₂/kWh — and yet it's one of the best grids in the set at 2.4%, because it imports from dirtier neighbours and its carbon intensity swings through the day. Louisville, at 741 g/kWh, gets 0.01%: it burns coal at the same rate around the clock, so there's no cleaner hour to move into.

Across 38 grids, the correlation between the saving and how dirty a grid is comes out slightly negative. With how much it varies, it's 0.91. Being dirty doesn't help at all — being uneven is the whole thing.

Fair warning: the demand response is a model rather than measured behaviour, and it uses average carbon intensity, not marginal. Happy to hear your thoughts.

Proportional decrease is not as useful in a climate change sense than actual decrease in co2 emissions. A grid that is highly sustainable with lots of wind and solar, but fossil peaker plants will have a lot more potential co2 savings in this model. It essentially makes dirty grids look cleaner.

Is it assumed that the total consumption remains the same (i.e there is an increase in consumption in clean energy when there is a decrease in less clean ones)?

Otherwise, there would be an obvious solution which is to increase the price all the time to reduce demand and thus reduce CO2 consumption, but it is impractical politically.

How are you modelling how demand changes in response to price changes?