The coronavirus response reduced electricity demand in every country across Europe last weekItaly, France and Spain have been most affected. UK electricity demand has been least impacted so far.
23rd March 2020
Every country in Europe has seen electricity demand fall 2-7% week-on-week. These are very significant falls in the context of electricity demand, where temperature-adjusted changes are normally small. Italy, Spain and probably France show twice the impact of any other country. UK electricity demand has been least impacted so far.
In Italy, although the week-on-week fall was 12%, there was already an 8% impact from the previous week, implying a total impact of 20% over the last 2 weeks. The graph below shows the Italy impact started early, with falls in electricity demand already in the week of 2nd to 8th March; these did not happen elsewhere in Europe. In addition, Italy imposed further lockdown yesterday, meaning even more industry and services will shut down and the impact on electricity demand could even exceed 20%. Last Friday, the Italian utility A2A said that coronavirus was responsible for a 15% fall in electricity demand across the whole of Italy and that the fall was as much as 25% in northern regions of Italy.
In Spain, the weekly impact was -10% on electricity demand, but Thursday saw the biggest fall at 16% (see below). In France, ENTSO-E data stopped on Wednesday, so we can only see falls of 10-13% from Monday to Wednesday.
Germany’s electricity demand fall accelerated to 6% by Friday as more industries closed down throughout the week (no data was reported yet for the weekend by ENTSO-E). The UK has seen one of the smallest impacts so far, with little impact observed until Saturday. Every country has seen some impact – for example, Poland had daily falls of around 5% from Monday to Friday.
Context: Although the first case of coronavirus in Europe was confirmed almost 2 months ago, it was only in the last 10 ten days that the crisis has gradually paralysed Europe. Therefore, we thought a one-off deep analysis of week-on-week changes would give a good idea of how this has started to already impact electricity demand across Europe. We went through the additional step of weather-correcting the demand data to make the comparison more robust.
Obviously this is really early into the disruption, so constant monitoring will be needed to track how this changes, and we shouldn’t read too much from these early results. China provides some clues that the falls will be deep and long-lasting: in Jan/Feb electricity demand was down 8% year-on-year. Since the Wuhan lockdown only started at the end of January, this is more likely to be a national demand impact of around 16% in February. Hubei still has electricity demand 30% below the level compared to last year, two months after the lockdown began.
Methodology: We took electricity demand data from ENTSO-E (they call it “total load”). This is data aggregated from local system operators across Europe. We performed a weather correction for temperature (the data already should strip out all distributed generation including wind and solar generation, so reflect pure electricity demand). We manually copied the historic capital temperatures from TimeandDate. The colder the temperature, the higher electricity demand should be, so we adjusted electricity demand by 0.8% for every degree from the same hour from the week before. This regression is roughly in line with historic analysis for the UK and Italy (especially in context of the 5-15 degree range of mean temperatures for the last fortnight). France shows a bigger temperature sensitivity there due to high use of electric heaters; so we adjusted demand by 2.5%/degree.
Note: The week-on-week weather temperature change was small across Europe, except for Spain which saw 5 degree cooler temperatures last week, so this is not a critical step.
Data plays a vital role in how we optimise our response to the coronavirus. Something as seemingly banal as electricity demand provides some insights into our response and its impacts.
More than ever, we need this data and analysis in the open public domain, not behind paywalls. We would urge all organisations to put data and analysis into the public domain so that we can all respond more effectively to the pandemic.
France: ENTSO-E has not yet reported data beyond Wednesday lunchtime.
Germany: ENTSO-E did not report data for the weekend yet. See the pick-up in falls from Thursday, which coincided with more announcements of industries – particularly the car industry – closing down.
Italy: Note there were big falls prior to the Thursday from the week before – there is a graph in the summary above which included the previous week of 2-8th March.
Netherlands ENTSO-E load data may not be reliable.
Spain: note the biggest falls seem to be in the morning, maybe a shift of routine? There were big temperatures changes in Spain last week, with every day in Madrid being about 5 degrees cooler than the week before, therefore the weather-adjustment we have made is quite critical here.
Switzerland: saw small falls, but with a big increase in Saturday early morning confusing the overall weekly averages.