Looking for the live map? It now lives at app.electricitymap.org
electricityMap logo

How to trace back the origin of electricity

April 19, 2021 · 7 min read

Electricity can be generated in many different ways, each with their own environmental footprint. As a concerned citizen (or as a responsible company), I might be interested in knowing to which extent I was powered by renewable energy. Furthermore, I might be interested in knowing the amount of greenhouse gas emissions I’m responsible for when I consumed that electricity.

Tracing back the origin of electricity consumed might seem like a daunting task, especially since part of that electricity might not be locally generated: it could also be be imported from neighboring areas. Furthermore, these neighboring areas might import electricity from other areas, which in turn could also import from even more distant areas…

Does this mean electricity is untracable? Absolutely not! This blog post explains how we trace the origin of electricity as used in electricityMap.

How electricity behaves

At its core, electricity is always produced by power plants. Coal power plants produce coal-based electricity, and wind turbines produce wind-based electricity. When the electricity from each power plant is combined, we end up having mixed electricity: partly coal-based, party wind-based.

The share of each that is present in the resulting mix is determined by the share of power supplied by each power plant. For example, if a coal power plant produces twice the power of a wind turbine, then we’ll end up with a mix which is twice as coal-intensive as it is wind-intensive. This leads to a mix consisting of 67% coal and 33% wind, as illustrated below.

mixing-electrons.png

Rule #1 (proportional mixing): electricity sources mix proportionally to their amount of power.

A simple way to think about this is to think about what happens when a smoothie is prepared. The raw ingredients are blended, and even though the individual constituents are not present anymore in the mix, we can still say that the smoothie was made with a certain proportion of each ingredient.

Note that similar to the smoothie preparation process, the electricity mixing process is irreversible. When drinking from the straw, it becomes impossible to drink exlusively the strawberry part of the smoothie, and one inevitably ends up drinking the mix.

In the same way, it is not possible to specifically consume wind-based electricity once it has been mixed: it is only possible to consume electricity in proportions given by the share of sources.

mixing-smoothie.png

Rule #2 (irreversibility): Just in the same way that one can't unmix a smoothie to get the raw ingredients back, one can only consume electricity with proportions given by the share of sources.

The electricity network

The electricity grid is a network which contains power lines and power plants spanning across multiple areas. A simplified representation consists in dividing up the electricity network in areas within which the electricity is assumed to flow freely without any restrictions. These so-called zones can represent countries, states or even islands, depending on the data available.

Electricity can flow between these zones using so-called interconnectors, which represent the imports and the exports of electricity between neighbouring zones. The limited capacity of the interconnector imposes restrictions on how much electricity can flow to and from a neighboring zone.

topo.png

The (simplified) European electricity grid used by electricityMap.

Two things happen on this network given the behavior of electricity previously described:

  1. The electricity exported from a given zone is the same as the electricity available in that zone, as one can’t unmix the electricity to selectively export some elements of it.
  2. The electricity available in a given zone is the result of mixing electricity generated from each local power plant with electricity imported from each neighboring zone.

Note that both rules lead to a peculiar consequence: as the mix of electricity imported might come from a zone that also imports electricity, the mix of electricity available in a given zone is influenced by the entire chain of imports. That chain starts with neighbouring zones that provide imported electricity, extends to their respective neighbours, the neighbours’ neighbour etc… and ends when no neighbors providing imported electricity can be reached.

It’s as if smoothies were repeatedly mixed with new ingredients, where each mixing operation represents a situation where imports (the smoothies) are mixed with local generation (the new ingredients).

In order to make matters worse, the chain of imports might be infinite, as so-called loop-flows might arise in situations where a zone ends up indirectly importing from itself (see figure below).

loop-flow.png

Presence of a loop flow where electricity in East Denmark is imported from Germany, which imports it from Sweden, which imports it from East Denmark, which imports it from Germany, which...

Fear not, as mathematical modelling is quite able to cope with these problems!

The flow-tracing methodology

The solution to this problem is a methodology called flow-tracing, and is a concept introduced in this peer-reviewed paper and applied on the European electricity grid here. It can be used to trace back the mix of electricity available in a given area, even in the presence of loop-flows.

The resulting mixes in each zone depict where the electricity available in a given zone originates from. Furthermore, flow-tracing shows the propagation of electricity locally generated (see figure below).

eu-mix-paper.png

Representation of the electricity flows during the first hour of January 1, 2017. The width of the arrows is proportional to amount of electricity transmitted. The cascade of electricity flows from German wind and Polish coal are highlighted with blue and brown arrows, respectively. The cascade stops when the share of German wind or Polish coal becomes too small to be drawn. This happens e.g. in France as the local generation is very large compared to the imported German wind, which causes the exported electricity to contain almost no German wind. Note that some countries, e.g. Switzerland, are excluded due to a lack of data. Taken from our joint paper with Bo Tranberg.

Once the origin of electricity is identified and broken down by power plant type, we can use IPCC 2014 greenhouse gas emission factors to assess the carbon impact of one’s electricity. The breakdown by power plant type also allows us to determine the share of renewable and the share of low-carbon electricity (which includes nuclear power).

header-emap.png

The electricityMap App showcases the result of solving the flow-tracing algorithm in real-time. Even though East Denmark has no nuclear generation, some nuclear electricity is still available through the imports, thus creating a small difference between the low-carbon and renewable percentages shown here (nuclear is low-carbon but not renewable).

The importance of taking into account not only neighbouring imports, but also the full chain of imports (neighbours of neghbouring imports..) is not to be underestimated, as even the slightest changes can have large consequences.

In the figure below, even though the import from East Denmark only represents 2% of Swedish electricity, it stil ends up representing 15% of Swedish emissions. Nearly half (41%) of these imported emissions are actually not domestically produced in Denmark, but rather imported from Germany, being Sweden’s neighbor’s neighbor!

situation.png

Selected snapshot from the electricityMap App showcasing the importance of taking into account imports beyond the first neighbour, as even the smallest power imports might be responsible for a large share of emissions as it propagates through the network.

Conclusion

Flow-tracing can be used to trace the origin of electricity in real-time, even in the presence of loop flows.

As more accurate grid data becomes available, we foresee electricityMap to be able to increase in resolution, ideally down to distribution networks or even transformers. Furthermore, we will be looking at improving the methodology in the future, by for example taking into account transmission losses or by having a more accurate model of storage systems.


Olivier Corradi

Written by Olivier Corradi
Founder @ electricityMap, CEO
follow them on