European power 'thought of the week'

March 15, 2017

This week, the research team takes a closer look at nuclear power's role in Germany's energy mix


« Back to All Research and Analysis


  • This week we would like to focus our attention on nuclear power generation. Its share in the energy mix has kept declining during the last 3 / 4 years as a result of the German energy policy following the Fukushima aftermath. Nevertheless, its share is still relevant (between 10 and 15% over the past year) and potential impacts on the German electricity prices may still be in place.
  • We first calculated the running correlation between the front month contract (German electricity) and the nuclear power generation for the past 191 weeks (i.e. since mid-2013 until present). We then binned such 191 values and constructed the associated probability density function.
  • It is clear that the coupling is skewed towards positive values (more specifically, 2 out of 3 weeks), which means that nuclear power production and the electricity price move in the same direction, most of the time. However, the most interesting evidence emerges when the coupling is split by season. While rolling correlation values are generally negative during the “cold months”, they are largely skewed towards a positive coupling during the “warm months”. In other words, the relationship between nuclear power generation and electricity price changes seasonally (similarly to what happens between inventory levels and gas prices). More specifically, the positive coupling is mainly observed during Q2 and Q3.
  • In order to further quantify such a relationship, we plotted the two subsets of data (i.e. electricity price as a function of power production). The associated linear fits and R² values tell us that the Nov-Feb coupling is statistically weak, while the “warm months” relationship is much stronger.
  • There surely are a number of reasons why this happens. Firstly, the European winter is dominated by demand and the large volatility in supply, which is, of course, dictated by the weather, and yet it largely impacts the production rates of conventional sources. Within this scenario, the nuclear power generation – which is in nature dispatchable – tends to have less of an impact on electricity prices.
  • On the other hand, during the warm months the weather forcing is typically smaller and the chance for nuclear power generation to impact the electricity prices increases. Therefore, even relatively small changes can lead to a visible price movement. Within such a scenario, one could argue that – under certain conditions – the levelised cost of electricity of nuclear plants is not as competitive as their CCGT counterparts. If such an inference holds, the associated coupling will be positive.