Nuclear Energy Production in France
Effects of Drought and Heatwaves
What do the results tell us?
The results show that the economy of France can be significantly affected by water coolant loss for nuclear stations as a result of drought and heatwaves, with that loss being significant even under optimistic RCPs but becoming very large in the decades surrounding 2100. Economic losses are on the scale of tens of billions of euros per decade in all three time periods. Losses are markedly higher in 2100, approximately doubling above the baseline and 2050 values. Adaptation strategies reduce these losses significantly after 2050, by between 5 billion euros (Adaptation Option 1) and 15 billion euros (Adaptation Option 2), suggesting a cost-benefit calculation would call for adaptation investments on this scale.
What can we do with the results?
The results can help decision-makers – both government and in the power sector – identify the key points of vulnerability of the economy to power loss, and to direct adaptation funds so they produce the greatest reduction in the vulnerability of the country's economy to extreme weather events. This can be done by adjusting the Sector Vulnerability Matrix so production sectors are less reliant on grid power and by allocating the residual power production in a way that minimises overall economic loss while meeting obligations under power contracts both domestically and in the export market.
How are the results obtained?
Results were obtained using an input-output model of France's economy, linked to climate projections supplied by the partners at University of East Anglia. The input-output used data (provided by EdF) on the power decline during the 2003 and 2006 drought/heatwaves were used to model the links between climate and power production. The model was then used to simulate the expected frequency of 2003/2006 conditions (or deeper droughts/heatwaves) in 2050 and 2100, and then to follow both the direct and indirect economic impacts of this loss in power during the period when the coolant water conditions were insufficient to meet demand in the power stations.
More specifically, the modelling took place in 7 steps:
- One of the Representative Concentration Pathways (RCPs) was selected from amoungst the RCP2.6, 4.5 and 8.5 pathways considered throughout the ToPDAd project.
- The projected number of days of drought and heatwaves (equal or greater in magnitude than that in the 2003/2006 events) per decade in 2050 and in 2100 were obtained from estimates provided by the University of East Anglia climate group.
- The Sector Vulnerability Matrix for each sector was determined from the 2003/2006 power loss experience.
- This information was used to calculate the loss of production capacity in each economic sector for the present day, 2050 and 2100, scaling the 2003/2006 power losses by the change in the number of days with the necessary drought and heatwave conditions.
- The indirect and direct economic losses were calculated from the input-output model for each of the three years and for each of the three RCPs.
- Adaptation strategies were then applied to repeat steps 1-6 so the effectiveness at reducing GDP loss could be produced.
Two adaptation strategies (plus one forecast with no adaptation strategy) were modelled for each RCP and year:
- Option 0: No adaptation strategy is applied.
- Option 1: Residual power during a period of reduction is allocated to maintain industrial production and meet final consumer demand in France, accompanied by a reduction in power exports.
- Option 2: Residual power is allocated to maintain industrial production, final demand and exports, with a smart grid and smart buildings introduced to allow for reduction of non-essential energy use during the ‘brown outs'.
What are the broader applications?
The same methodology can be applied to any community, nation or region for which the necessary input-output data and water-cooled power production data are available. Since the input-output data contain all economic sectors, the methodology can be applied to any sector for which the Sector Vulnerability Matrix can be supplied.
The methodology can be improved by incorporating macroeconomic projections such as are available through the GRACE and GINFORS models of ToPDAd, allowing the methodology to better reflect potential changes in the size and structure of the economy in 2050 and 2100. At present, these two factors are within the model as data entries that reflect the assumptions given above, but can be changed as the data on the macroeconomy become available.
Key Messages and Conclusions
Option 0: No adaptation strategy is applied.
Option 1: Residual power during a period of reduction is allocated to maintain industrial production and meet final consumer demand in France, accompanied by a reduction in power exports.
Option 2: Residual power is allocated to maintain industrial production, final demand and exports, with a smart grid and smart buildings introduced to allow for reduction of non-essential energy use during the ‘brown outs'.