Consider residual carbon intensity as data we make available

[mrchrisadams]

There are a few common metrics for reporting the carbon intensity of electricity, especially if we refer to the GHG Protocol:

1. location based - which is usually an weighted average of the carbon intensity of all the generation in a given region
2. market based - this accounts for financial instruments people use to 'green' their electricity. In practice this is typically linked to Environmental Attribute Certificates which you can think of as a tradable by-product of generating green electricity, and often traded separately from the power generated.

The flip side of market based approach is that if you're paying to make the your use of your energy greener, you're able to do this because in some other part of the world, their electricity is becoming less green.

If you don't have sight on this, and if you make claims based on the average carbon intensity of a grid plus any certificates you buy on the market to top-up your way to 100% renewable, then it's possible you can be undercounting your impact.

There's a good example from Google's recent submission in the consultation about updating the GHG Protocol's approach to Scope 2 carbon emissions emphasis mine:

Data demonstrate that many companies do indeed pursue this practice. For example, Norway was responsible for 43% of all guarantees of origin (GOs) exports in Europe in 2022, many of which were purchased by companies whose operations have no connection to the Norwegian grid on which these EACs were produced. This also explains why, even though Norway’s grid-level emissions factor is 10 kg CO2/MWh17 (98% carbon-free), the residual emissions factor is 402 kg CO2/MWh (7.4% renewable), reflecting that most EACs produced within the Norwegian grid are claimed and retired outside of the country.

Such a system can and does distort companies’ footprints by understating their reliance on fossil-based electricity. This results in companies not addressing the emissions for which they’re physically responsible, and not supporting the collective and urgent need to decarbonize electricity systems to the degree that they might otherwise. In the above example, Spanish companies that purchase EACs from Norway are not working to decarbonize the electricity grid in Spain, which still relies on fossil fuels for over one-third of its electricity.

You can see it in this document in the link below - page 13 and page 14:

http://gstatic.com/gumdrop/sustainability/google-2023-GHGP-Survey-Submission.pdf

I think this information is published by the AIB, but and this chart shows how accounting for this can totally change the emissions intensity you might take into account when making claims based on market based approaches:

Look at Iceland (IS) and Norway (NO) specifically - the residual mix vs the production mix are radically different.

Here's the underlying doc that came from

I'm unclear on the licensing terms, or how often this data is updated, so it may be the we can't use this, but I'm creating this issue so we can at least discuss it, and have it on our radar.

[mrchrisadams]

Part of this thinking in related to this issue around creating a new convention or standard for reporting, currently referred to as "real time cloud":

https://github.com/Green-Software-Foundation/real-time-cloud/issues/8

[mrchrisadams]

OK, the more I read about this, the more I think this will eventually be available from lots of places, and rolled into Electricity maps data offering first - their most recent report with Flexidao explores this issue in detail.

This chart is telling:

It basically shows that 8x as many green energy certificates / EACs were consumed/cancelled in Ireland were issued there, Ireland, and 7x were for Germany.

I think this is a bit like saying 8x and 7x the green energy was claimed in those respective countries, as was generated for the year 🤯

As an aside: In light of the recent energy efficiency law passed in Germany, does this mean that for all the energy claimed inside the borders to come from local clean generation, you'd need 7x the clean generation to be deployed as we have now?

The linked report is below - I think I'll need to read it in more detail.

https://www.electricitymaps.com/reports/residual-mix