Today, we’re launching a new improvement to Electricity Maps which will improve the precision of our carbon emission data. We call this improvement regional emission factors.
An emission factor corresponds to the amount of greenhouse gasses released in the atmosphere from the production of a single unit of electricity by a specific power plant. Renewable technologies tend to have a low emission factor (typically lower than 50 gCO₂eq/kWh), whereas fossil fuel power plants tend to have a high emission factor (typically higher than 400 gCO₂eq/kWh). Multiplying an emission factor by the amount of electricity consumed results in the total amount of greenhouse gasses emitted.
Emission factors used on the Electricity Maps App take into account the whole life-cycle of the power plants. Thus, emissions resulting from the extraction of resources required to build the power plant, emissions from its operations, and its end-of-life, all need to be accounted for. Our commercial API includes both life-cycle and direct emission factors.
We previously used emission factors from the 2014 IPCC Fifth Assessment Report, which aggregates emission factors estimated by a multitude of peer-reviewed studies. These emission factors depend on the type of power plant, but stay the same irrespective of the location.
However, the emission factor of a power plant depends on multiple factors. The quality of the fuel used will impact the factor. For example, lignite power plants have higher emission factors than hard coal power plants as lignite has a lower carbon content. The efficiency and the technology of the power plant also need to be considered. In the case of a combined heat and power (CHP) plant, heat is recovered from electricity production and used for heat production, significantly increasing the efficiency of the plant compared to separate electricity and heat installations. Less fuel is burned and consequently, total emissions are reduced. As a result, emissions from electricity generation in a CHP plant are lower than a conventional thermal power plant.
All of this means that there’s an opportunity to increase the precision of emission factors, especially as newer and more granular data sources appeared in the meantime.
The US Environment Protection Agency publishes the Emissions & Generation Resource Integrated Database (eGRID) every year. The eGRID data includes generation, emissions among other attributes for almost all power plants on the US territory. The dataset is available here.
To compute regional emission factors for the US, more than 5000 power plants in the eGRID dataset were matched to the corresponding Electricity Maps zone. Emission factors for each plant in a given zone are then aggregated to create the emissions factors per power plant type and per zone.
In the example below, we can see that resulting biomass emission factors vary greatly depending on the location. In the US, reported emissions from biomass depend on the type of combustible burned. Some combustible types are considered as carbon-free, while others are not. Consequently, the emissions reported vary greatly between power plants (see eGRID2020 Technical guide section 184.108.40.206).
The European Commission publishes verified emissions and allocations for all power plants that are part of the EU emissions trading scheme (EU-ETS). This dataset is updated on an annual basis and is available here. Furthermore, ENTSO-E publishes generation information per power plant for the majority of European countries. Generation per unit is available on the ENTSO-E Transparency Platform here.
Crossing these two datasets enables us to determine the emission factors of each power plant.
The emission factor of a type of power plant in a given Electricity Maps zone is obtained by aggregating the emission factors of all power plants located in that zone. As an example, when considering gas power plants, we see quite a diverse set of emission factors depending on the country. As explained above, emission factors depend on the technology of the power plant. This explains the difference in distribution between Romania and Italy. In Italy, the average gas emission factor is lower and the distribution of factors is smaller as there are more CHP plants. On the other hand, the large variance in Romania suggests that gas installations only produce electricity and are less efficient.
The methodology to compute these figures is completely transparent and will be continuously updated on our wiki. Anyone is encouraged to contribute to it in order to help us publish the most accurate and up to date information. The resulting emission factors are open source, and can be accessed here. They are also accessible directly on our App.
This new methodology will impact carbon intensities in the US and Europe in the following ways.
Finally, the following table provides an overview of the new emission factors, averaged across the whole US/EU.
We continuously improve our methodology to ensure the highest level of data accuracy. Adding the regional emission factors helps us reach that goal. This will also allow us to have a better understanding of a zone’s electricity sector as well as its actual contribution to global electricity emissions.
We would like to thank Mirko Schäfer for his insights on the EU-ETS data and for his help on the methodology.
We’d also like to thank Thomas Gibon for helping Electricity Maps identify some of these key data sources as well as Dave Jones and Ember for reviewing the methodology.
This work was supported by a grant from the Google.org foundation.