Carbon intensity (gCO2-eq/kWh) statistics


Dear community,

I am looking for data related to carbon intensity measured in gCO2-eq/kWh (and not gCO2/GDP) for the EU28 plus Norway and Switzerland, ideally over the timerange 2006-2016.

The EEA proposes such statistics (for EU28) but only based on the electricity generated domestically. What I would need is data derived from electricity consumed, that is including the carbon from imports (and excluding the carbon from export electricity).

Would anyone be aware of such data ?

Many thanks in advance,



Hello @Michaelweber If you need hourly marginal data, then try electricityMap. Their machine-learning algorithm is described by Carradi (2018). ElectricityMap must have historical data to train their algorithm but I don’t know how far back that goes.

Notwithstanding, I recently asked electricityMap about the problem of unaccounted cross-border carbon (shorthand for GHG) flows, anti‑parallel to the energy flows if you like, but have yet to receive an answer. I think the underlying allocation problem is probably quite tough and interactions at the margins do not capture the information being sought.

If you need annualized average country‑specific data, then perhaps using national economic accounts, including input/output tables might prove a better approach? Perhaps CGE modelers can help out here? They could also project forward. I am not familiar with the literature here though. But you will probably need to apply decomposition techniques. Patterson et al (2006) might offer clues? When I coded up their tutorial problem in Matlab, I got a different solution. I told Murray that but he wasn’t the slightest bit interested.

This question is quite topical because Germany claims its carbon trajectory is lousy because it now exports significant quantities of hard coal generation to the Netherlands, Great Britain, and elsewhere. HTH, R.


Corradi, Olivier (3 July 2018). Estimating marginal carbon intensity with machine learning. Tomorrow. Copenhagen, Denmark. Blog.

Patterson, Murray G, Graeme C Wake, Robert McKibbin, and Anthony O Cole (15 March 2006). “Ecological pricing and transformity: a solution method for systems rarely at general equilibrium”. Ecological Economics. 56 (3): 412–423. ISSN 0921-8009. doi:10.1016/j.ecolecon.2005.09.018.


Hi @robbie.morrison !

Thank you for this pertinent answer. I also tried to look at the resources that electricityMap could offer, but unfortunately I couldn’t find what I needed.

I am looking at annual average data; my work covers many different energy indicators and I can’t allow myself to spend too much time gathering data just for (consumed) carbon intensity. Worst case scenario will be to use EEA’s data. However, as you’ve mentioned it, carbon reporting is significantly different whether we take into account generated or consumed kWhs.

Thank you nonetheless.


Hi @Michaelweber Just to note that nodal pricing (also known as bid‑based, security‑constrained, economic dispatch with nodal prices) provides directly the marginal information that electricityMap infers. Maybe the entire European electricity system will under nodal pricing one day … who knows? R.


Hi again @Michaelweber The United Kingdom Carbon Intensity API website is worth looking at. It is restricted to Great Britain so does not address your original question. But like electricityMap, it uses “state‑of‑the‑art machine learning”. The site offers two‑day forecasts. Backers include Oxford University Department of Computer Science, National Grid TSO, and WWF. The documentation reads (emphasis added):

Summary: National Grid’s Carbon Intensity API provides an indicative trend of regional carbon intensity of the electricity system in Great Britain (GB) up to 2 days ahead of real‑time. It provides programmatic and timely access to both forecast and estimated carbon intensity data. The Carbon Intensity forecast includes CO2 emissions related to electricity generation only. [And] includes emissions from all large metered power stations, interconnector imports, transmission and distribution losses, and accounts for national electricity demand, embedded wind and solar generation.

I would guess that the model does not extend to western Europe and the import intensities are based on typical values for France (IFA, 2 GW), the Netherlands (BritNed, 1 GW), and Northern Ireland (2 × 0.5 GW). I couldn’t see any academic publications listed but doubtless the methodology will be written up at some point. HTH. R.


Great resource, thank you! I wish it would be extended to EU28.