Is nuclear modulation an oxymoron?

Before the country sank in political turmoil, a healthy debate was going on in France around the role of nuclear power in the electricity mix of the country. Nuclear power plants generated 65% of the country’s power in 2023 (RTE) and has been a cornerstone of the mix for the past 40 years. With increasing interconnections to its European neighbours and intermittent renewable energy penetration, the place and role of a gigantic baseload producer is under scrutiny.

The role and place of nuclear power in the energy mix

Nuclear power plants typically have installed capacities between 900 MW (Bugey for example) and 1450 MW for the most recent ones (Chooz and Civaux). France has a total of 56 reactors and an installed capacity of 61,4 GW. The Flamanville 3 nuclear power plant, due to be connected to the grid in 2024, will be the 57th reactor and will have a nominal capacity of 1600 MW.

Nuclear power plants have a specific economic structure. They have extremely high upfront construction and financing costs, ranging in the billion or tens of billions of euros and long construction times, often compounded by significant delays. Once operational, they have limited variable costs, mostly related to the uranium life cycle. The vast majority of their cost structure is dominated by fixed costs, and notably by the weight of amortisation and interest payments on construction debt. 

Microeconomics and basic accounting teach that businesses with important fixed costs have to sell a lot of their production before they can reach their breakeven point. They also have to produce all the time, since any downtime creates a period during which costs still accrue but revenue is no longer generated. This is typically how nuclear power plants have been designed historically. They are meant to be the power horses of the energy mix, producing vast amounts of electricity on a continuous basis, i.e. providing baseload power. 

Why modulation is increasing

This is for economic theory. In practice things have always been more complicated and continuous production is unachievable. The key variable is the availability factor of a nuclear power plant. Maintenance operations, refuelling, incidents and other exogenous factors require to stop production and are a significant source of costs and lost revenues. 

Modulation on the other hand is a situation in which production is decreased for a short period of time but the reactor is not stopped. This is achieved by a careful management of the chemical properties of the core. Boron is injected in order to dampen the neutron flux and lithium is added to balance the pH levels (full description). The reduction in output is temporary and can go as high as an 80% decrease in power generation. Nuclear power plants are however not very flexible and cannot adjust quickly to important shifts in demand. A classic adjustment program is the “12-3-6-3” during which production is at nominal capacity during 12 hours, then is progressively reduced during 3 hours, runs at the lowest level during 6 hours and then ramps up again during 3 hours. This means that for very short-term response (<3 hours), traditional peaking plants, hydro assets, batteries, flying wheels and demand response remain the best solutions.

Because of their importance in the French electricity mix, nuclear power plants have historically used modulation to match decreases in demand (also known as load following) as the supply-demand balance has to be insured at all times. It is also used to provide ancillary services to the grid, such as frequency management. These operations are however marginal and are compensated, thus offsetting the losses from reduced production. They do not fundamentally challenge the fixed cost equation.

The development of intermittent renewable energy is changing the situation however and this is idiosyncratic to France because of the weight of nuclear energy in the country’s mix. Countries such as Germany, Italy, Spain or the United States have a far greater share of traditional thermal power plants (coal, gas) whose cost structure includes more variable costs (i.e. fuel consumption) and are technically easier to modulate. Because of their intermittent nature, renewables require system-wide solutions to offset their peaks and drops in production. At roughly 14% of total generated power in 2023, solar and wind power play a significant role in the French mix and their place is bound to significantly increase in line with French and European energy transition objectives. As there is no way to pilot this production and as the energy would be lost if not consumed immediately, short-term wholesale prices experience significant fluctuations, even venturing in negative territory, when there is an excessive production. This creates a situation where nuclear power has an economic incentive to modulate its production instead of only doing baseload. 

There is often a confusion around the notion of priority of injection for renewables on the grid. In France, only small-scale installations (< 400kW) have such a priority. All significant renewable producers inject their power on the same terms as conventional and nuclear power plants, the market price is the adjustment factor. 

Technical and economic consequences

There are two questions structuring the debate around modulation. The first one relates to the technical consequences on nuclear power plants. Most of them have been designed and built in the 1970ies and 1980ies when modulation was a marginal topic. Modulation practices remain in the standard safe operation space and all parts of the plants are certified for the reduced power ranges. The question that is being raised is to know whether more frequent modulations will generate premature wear. Conditions in the primary circuitry are only marginally affected as the increased acidity due to boron injections is monitored and compensated with lithium. Temperatures remain roughly the same. 

The issue is more focused on the secondary circuitry where important shifts in temperature and pressure can be registered and where the turbines can experience accelerated wear due to regime changes. The risk is to have increased corrosion problems in the circuitry and more maintenance (and downtime) as a result. During a recent interview in front of the National Assembly, Cédric Lewandowski, EDF’s Vice-President in charge of the nuclear and thermal parcs, estimated that, should modulation requirements increase in the future, the consequences should be very closely monitored. The issue is important because it could impact the lifetime extensions that these plants may require in order to meet future energy needs before the next generation of nuclear power plants can be put online.

The second question is deeply related to the first one and regards the economic benefits of modulation. There is value for EDF to do load following for its nuclear power plants as it allows the company to capture some of the price increases due to reduced renewable production and to avoid drops in power prices when production is abundant. This short-term value has to be put in perspective with the potentially increased maintenance costs and with the potentially decreased operational lifetime of the plants. At this stage, official sources say that due to the limited use of modulation, consequences on the primary circuit are negligible and do not generate premature wear. On the secondary circuitry, things are more complex and intensive studies are under way to gain a better understanding of the long-term consequences and associated costs.

There is no definitive conclusion to the debate around modulation and this is probably why it gets so heated. It is however doubtless that modulation will play an increasing role in the French energy system in the coming years and it thus of paramount importance to gain a better understanding of its long-term consequences.

Energy Voices is a democratic space presenting the thoughts and opinions of leading Energy & Sustainability writers, their opinions do not necessarily represent those of illuminem.