The decision to shut down eight nuclear reactors in the year following the Fukushima disaster contributed to higher electricity prices in Germany. It also led to the partial replacement of nuclear generation by electricity generated from coal, lignite and gas, and thus to an increase in CO2 emissions and local pollution, the health effects of which have not been considered.
Diffuse and concentrated risks
The dispersal of risks and accidents is a factor that facilitates their social acceptance. Thus, although much less deadly than road accidents (which cause more than 3,000 deaths a year in France), plane and train accidents cause more emotion. Even if we add to this the 20,000 or so injured people hospitalized each year as a result of road accidents, the car is perceived as not very dangerous, and public opinion erupts when the authorities seek to impose speed limits. In the evaluation of public policies, this difference in perception is generally not considered. At best, economists calculate the balance of aggregate benefits and costs without taking into account how these benefits and costs are felt by the public. Politicians, on the contrary, if only for electoral reasons, are more sensitive to the concentration or dispersion in time and space of the effects of their decisions.
Phasing out nuclear power
Germany's nuclear power plant policy is an example of this sprinkling effect. In 2011, after the meltdown of the Fukushima Daiichi reactors, the German government, which had just relaunched the industry, decided to gradually close all its nuclear power plants. The last ones will have to stop in 2022. No cost-benefit assessment was carried out to support this decision, but the federal authorities considered that it would be well accepted by the population, which has often shown its hostility to this technology because of the risks of accidents and the problem of waste reprocessing. Even today, 81% of Germans are still in favor of this radical change of policy. For environmental organizations, the move away from nuclear power is a win-win situation, as plant closures are compensated for by the production of electricity from renewable sources, which is operated at close to zero cost and without emitting greenhouse gas or particulates. It is undeniable that green electricity production has sharply increased over the decade. But in an industry where the product is not storable on a large scale, it is necessary to ensure that production is available at the times when consumers want to use it. Because wind and sun are intermittent, unless considerable capital is invested in energy storage facilities, green energy cannot be an identical replacement for dispatchable plants that operate at baseload levels and, if necessary, increase their output to meet peak demands. The problem arises in particular at the end of the day when demand increases while the photovoltaic panels are not producing. At times when green energy is not available, conventional German thermal power plants, or imports, have had to be used to replace closed nuclear power plants. The shutdown of the reactors has therefore been accompanied by an increase in production from green energy at times when it is available and, at times when it is not, by an increase in production from power stations burning coal, lignite and gas, as well as by an increase in imports, especially from Denmark, France and the Czech Republic.
Generalized cost
In total, output from thermal power stations has fallen, but to a lesser extent than if the nuclear power stations had not been shut down. A study conducted at Berkeley measures the difference between the observed trajectory and that of the alternative scenario in which the nuclear reactors would not have been shut down early (the counterfactual). It assesses the costs associated with the more intensive use of conventional thermal power plants and the increased imports resulting from the nuclear phase-out. The authors analyze the shutdown of 10 reactors (out of the 17 initially operating) between 2011 and 2017. The originality of the study is that it is not limited to production costs and aggregate carbon emissions. It also integrates the spatial dispersion of effects in terms of local air pollution. By combining data on hourly and plant by plant operating volumes and costs between 2010 and 2017 with information on electricity demand and prices, local weather conditions and primary energy prices, the authors calculate what the hourly output per plant would have been if the government had not ordered the closure of the ten reactors. They deduce the additional costs incurred in terms of primary energy consumption and environmental damage. They end up with some estimates of the effect of the closure decision on the average production of thermal power plants (+15%), on imports (+37%) and on wholesale prices (+4%). As the four major industrial groups owning nuclear power plants also owned coal-fired power plants, they were little penalized by this energy substitution. However, the higher prices were indeed a cost for German consumers, estimated by the authors at $1.6 billion per year. But this is not the main cost. The impact on health and the climate must also be considered.
Local pollution and excess mortality
The combustion of fossil fuels causes emissions of CO2 and pollutants with local health effects (SO2, NOx and fine particles). According to the study cited, the replacement of nuclear power plant production resulted in an additional 12% for each of these emissions. The additional carbon dioxide emissions amount to 36 million tons per year, which, based on a social cost of carbon of $50/tCO2, represents an additional cost of $1.8 billion per year.
But the most important effect comes from additional local pollution: it would have caused an excess mortality of 1,100 individuals per year. This number is calculated from a geographical analysis of the populations living around the thermal power plants, the variation in local production induced by the government decision and estimates of the harmfulness of local pollution (ESCAPE project). It only considers residents in Germany, without integrating the effects in neighboring countries. Taking a value of $7.9 million as the price of a German's life, the authors of the study estimate that the additional cost of the closure decision due to local pollution amounts to $8.7 billion per year.
Benefits of closure
The costs are therefore substantial. What about the benefits? By closing the country's nuclear power plants, the German government wants to eliminate the risk of accidents and solve the problem of waste treatment and storage. The evaluation of the induced gain depends on the accident probabilities retained and on the way in which the populations in the vicinity of the damaged plants are taken care of (evacuation, re-housing, care, etc.). Current estimates are around $3 per MWh. As the exit policy has resulted in a reduction of 53 TWh/year in nuclear generation, the estimated gain is less than $200 million per year, far from the annual cost of $12 billion resulting from the addition of the additional cost of production ($1.6M), the additional CO2 emissions ($1.8M) and the additional mortality linked to local emissions ($8.7M).
*
* *
Assessing the economic and environmental impact of a nuclear power plant is not easy. It requires approximate assumptions that are open to criticism. But even if one can think that the costs are overestimated and the gains underestimated by the study cited above, the gap between the two is so great that it is difficult to see how the decision to close could be justified on a purely techno-economic level. However, in hindsight, even in possession of these data, it is likely that the anxiety-producing potential of the images of the earthquake/tsunami/reactors-meltdown sequence looping through the 24-hour news channels would have led the German government to take the same decision. The scattering of illnesses and deaths caused by local pollution considerably reduces their weight in the face of an event such as Fukushima, where we find the ingredients of the classic tragedy: unity of place, unity of time and unity of action. It is a psychological fact that explains why political rationality often frees itself from economic and environmental analysis.
