Sensitivity of European power systems to energy scenarios and climate change projections

(climate) uncertainty and power systems in Europe

The impacts of climate change on energy systems may be complex. And what about its impact on the energy systems we will have in Europe in 2050? This is the topic of the paper led by Hannah Bloomfield from the University of Reading, UK, recently published on Renewable Energy and that I co-authored.

In the paper we see that in the next decades Europe will be generally hotter, leading then to less use of electricity for heating in winter and a more for cooling in summer (especially in the southern countries).


This shows the impact of climate change on European annual and seasonal electricity demand (difference between 2045-2065 mean and 1980-2000 mean) using a present-day (2015) power system scenario (in other words, we show only the change due to climate). Coloured bars show the multi-model mean for each emission scenario, and individual models are given by black points with the black bars.

Climate models (we used data from the EURO-CORDEX initiative) show little agreement on how and where will wind & solar change. If you look at the seasonal changes, the uncertainty shown by climate change projections is rather high. This is consistent with other studies (for example, here, here and here): it means that the information given by climate change models must be used with caution (but in general uncertainty is not a bad thing, it reflects how complex is our world).


This figure is similar to the previous one, but it shows changes in wind power (again assuming the current technology and capacity). It is evident how the various climate models disagree and the extent of the uncertainty they show.

The paper digresses also a bit on a very important topic: the storyline-based approach when communicating uncertain climate information.

[…] In the absence of any reason to discount one or more of these climate models, each of these scenarios should be considered equally plausible estimates of future climate. Moreover, as all climate models frequently share many elements of code, they cannot be considered as unbiased estimators. This means that […] it is still possible to identify plausible scenarios of future changes […]

The work is based on the energy scenarios developed by e-Highway2050 project and on the climate-based reconstructions of electricity demand, wind and solar generation based on atmospheric reanalysis. The data, which is free (and currently available here), has been created in the first proof-of-concept of the Copernicus Climate Change Service (C3S) on the energy sector.

To conclude, we see that there is a lot of climate uncertainty in fundamental aspects of power systems (demand, renewables' generation) and this uncertainty will become more and more important in low-carbon power systems. This paper then tries to assess the uncertainty (both considering non-climatic sources of uncertainty, e.g. scenario paths) using the best available data for climate (multiple models from EURO-CORDEX) and for energy (from C3S). Enjoy the open access to the paper!

Senior Data Scientist