Forecast 2020 Q4: Back to the future

In recent weeks, wholesale electricity markets have alternated panic phases and wait-and-see phases, particularly on contracts for next winter (Q4 2020 & Q1 2021). This high volatility can be explained by the strong uncertainties surrounding the availability of nuclear power plants, but also by the macroeconomic context and the weight of a possible recession on the energy complex. Of course, these uncertainties make it very difficult to make a long-term forecast, but at COR-e we are convinced that data analysis and modelling have value to add in this situation.


We therefore propose a study focused on the impact of the weather on the forecasts for Q4 2020 contract in France.

Here we will perform different supply/demand scenarios by replaying the weather conditions of past winters; then these scenarios will be put into perspective with current market conditions to provide several possible price trajectories for the Q4 2020 contract in delivery.

We will therefore use historical temperature, solar radiations, and wind speed data from the winters 2011 to 2019 to reconstruct :

  • Hourly wind and solar production forecasts (using 2020 installed capacity),
  • Hourly power consumption forecasts

For this exercise we will also use :

  • Current gas, coal and CO2 prices quoted on the futures markets.
  • The latest nuclear availability forecasts from our proprietary model (presented in our previous Newsletter).


Before looking at the results regarding price forecasts, a few figures regarding  weather dataset we have been using :

  • Q4 2012 is the coldest quarter on average (0.4°C below seasonal)
  • Q4 2011 is the hottest quarter on average (1.4°C above seasonal)
  • There is a strong weekly variability of temperatures with deviations ranging from -4.2°C (week 44 2012) to +5.5°C from normal seasonal temperatures (week 51 2019).
  • With our simulations, consumption peaks are between 79500MW (weather of week 48 2015) and 87000MW (weather of week 50 2012).
  • With the 2020 installed capacities, the average weekly powers for renewable production (excluding hydro) would vary between 2400MW and 10000MW.

Here are now the details of our price simulations:

Beyond temperature and its impact on power demand, renewable production will be fundamental, as the saying “there is no wind when it is cold” is less and less valid in France. We can see that with current nuclear availability forecast, all scenarios see some tight weeks in November, with the risk of very high hours on week days.

There are still many uncertainties:  gas market  response to weather, hydrology, structural decline in demand. Predicting the future remains impossible, but data analysis and modeling are powerful tools for understanding the market and managing risk.

Contact us for detailed version of this analysis.

Our team is at your disposal to assist you in the construction of personalized scenarios on the tenors and countries of your choice. Our platform allows price simulations based on the inputs of your choice, whether they relate to weather, plant availability or fuel prices.

Emeric de Vigan
COR-e Founder and CEO

COR-e Logo

1 rue Hoche

83000 Toulon, France