BE 2025 - Production [EN]

Electricity generation in mainland France reached 547.5 TWh in 2025. After two years of strong growth in 2023 and 2024 of around 10% per year, due mainly to the renewed availability of the nuclear fleet and improved hydropower output resulting from more favourable weather conditions, the volume of electricity generated in mainland France grew very slightly in 2025 (+8.2 TWh, or +1.5% compared with the 2024 level).

This stability is the result of varying trends across the generation mix. In particular, 2025 was marked by a very significant drop in hydropower generation (−12.9 TWh), which had benefited from exceptional rainfall in 2024. This fall was offset by nuclear output, which returned to higher levels after the recovery that began in 2023 and continued in 2024 (+11.3 TWh higher in 2025 than in 2024), and by the increase in solar (+8.1 TWh) and wind (+2.8 TWh), driven particularly by capacity growth. Fossil-fired generation fell for the third year running (−1.3 TWh) and remains, as in 2024, at a very low level, unseen since the early 1950s.

The volume of low-carbon (nuclear and renewable) electricity generated in France reached an all-time high of 521.1 TWh in 2025. This represents 95.2% of the electricity generated in mainland France, a similar proportion to 2024 (95.0 %). On the other hand, the proportion of electricity from renewable sources fell in 2025 (27%) following the historic level reached in 2024 (27.9%).

The electricity generation fleet in mainland France is continuing its growth, with 8.7 GW of new capacity installed across the generation mix in 2025.

This expansion was driven primarily by solar capacity (+5.9 GW), which continued to grow at a high rate in 2025. It also reflects the nuclear fleet, with the Flamanville 3 nuclear reactor (1.6 GW) connected to the grid in December 2024 and commissioned gradually over the course of 2025 and early 2026. Onshore (+0.9 GW) and offshore wind capacity (+0.4 GW) also increased, though to a lesser extent. On the other hand, gas-fired electricity generation capacity fell slightly (−0.2 GW¹), mainly due to closures of small cogeneration plants.

Total French nuclear generation rose for the third year running, after reaching a low point in 2022. It amounted to 373.0 TWh, 11.3 TWh more than in 2024 (+3%). However, this increase was much smaller than the ones observed between 2022 and 2023 (+23 TWh, or +11%) and between 2023 and 2024 (+30 TWh, or +13%). The sharp rise in output in 2023 and 2024 corresponded to the recovery from the stress corrosion crisis, whose effects began to be felt from the end of 2021 and became particularly noticeable in 2022. By 2024, annual output had almost returned to pre-crisis levels. 

The trend observed in 2025 thus confirms that the availability of France's nuclear fleet has returned to a level comparable to before the crisis.

In 2022, at the height of the crisis, up to almost 65% of the fleet was shut down at the same time, due to the combined effect of stress corrosion-related maintenance and the impact of the pandemic on with stress corrosion seem to be well understood and controlled by the operator. The reactor inspection strategy now allows this phenomenon to be monitored without having to carry out the major operations initially required for reactors suspected of being at risk of this type of defect. 

The output achieved in 2025 (373.0 TWh) is slightly higher than the short-term projections set out in a number of recent publications. The Commission de régulation de l'énergie² (the French maintenance schedules. Given the high proportion of nuclear generation in the French electricity mix (between 63% and 77% over the last ten years, and 68.1% in 2025), the 2022 crisis had far-reaching consequences for the power system, including increased reliance on imports: France once again became a net importer of electricity for the first time since 1980. To a lesser extent, fossil-fired generation, mainly using gas, also increased, though at a limited rate.

The gradual completion of checks and repairs related to stress corrosion in 2023 and 2024 enabled output to gradually recover to a level in excess of 360 TWh by 2024. At the same time, France once again became a major exporter and fossil-fired generation fell to historically low levels.

The year 2025 confirms the findings of 2024: the effects of stress corrosion and the repercussions of the pandemic on production have almost disappeared. The risks and challenges associated with stress corrosion seem to be well understood and controlled by the operator. The reactor inspection strategy now allows this phenomenon to be monitored without having to carry out the major operations initially required for reactors suspected of being at risk of this type of defect. 

The output achieved in 2025 (373.0 TWh) is slightly higher than the short-term projections set out in a number of recent publications. The Commission de régulation de l'énergie² (the French energy regulation commission) refers to a baseline level of 362 TWh/year between 2026 and 2028. In RTE's 2023 and 2025 Generation Adequacy Reports, the central level of generation is projected to be around 360– 365 TWh/year between now and 2030. Finally, the operator anticipates a range of 350 to 370 Twh³ for 2026 and 2027.

These projections are based on forecasting: the actual level of production that will be observed each year remains relatively uncertain, as it depends on a large number of factors associated with the power system as a whole (including weather conditions, unforeseen events affecting the availability of resources, etc.).

The current strong performance of nuclear power compared with the predicted levels offers hope of a relatively favourable situation over the next few years (barring a new crisis), supporting an acceleration in the electrification and re-industrialisation of the country, which should lead to an increase in consumption in the medium term.

The trend observed in terms of generation, i.e. a slight improvement in 2025 after significant increases between 2022 and 2024, also applies to availability, which stood at 74.0%⁴ for 2025 as a whole, a level two and a half points higher than the previous year (71.5%). Since 2024, the nuclear fleet has returned to levels of availability close to those seen in the years preceding the stress corrosion crisis.

Overall fleet availability was close to the bottom of the historical pre-crisis range⁵ (2015–2019) in the first half of 2025 and broadly within the historical range from June onwards. Availability was relatively close to the 2024 level throughout the year, with the exception of January, February and March, when it was higher. This difference is due to the fact that a large number of ten-yearly inspections had been concentrated in these months during the previous year (see below).

Analysing the monthly availability rates for each power plant shows that most of them had very high availability in January and February, which are the most important months in terms of security of supply. The operator usually schedules work on the fleet so that shutdowns are concentrated as far as possible from spring to autumn, in order to maximise availability during the winter months. 

Major maintenance, and particularly ten-yearly inspection programmes, had a relatively limited effect in 2025. Energy not produced due to ten-yearly inspections in 2025 was the lowest observed since 2016 (around 27 TWh compared with an average of 43 TWh between 2016 and 2024). This significant reduction is the result of several factors: the end of the fourth ten-yearly inspection cycle for 900 MW reactors, which was particularly busy because it corresponded to the period in the early 1980s when the most reactors were commissioned⁶; the operator's gradual implementation of a programme to improve control of reactor outages⁷ between 2020 and 2025; and finally the resolution of the availability problems arising from the stress corrosion crisis, which had led to the duration of certain ten-yearly inspection programmes being extended to carry out corrosion-specific checks that were not initially planned.

From this point of view, 2025 and 2026 are the exceptions (according to the planned timetable). The ten-yearly inspection schedule is set to intensify again thereafter. The fourth programme of ten-yearly inspections for reactors in the 1,300 MW class, designed to extend their operation to 50 years, will start in 2026. This cycle, which affects a large number of reactors and should extend their lifetime beyond the initial duration, presents major challenges from an industrial viewpoint. From 2029 onwards, the fifth cycle of ten-yearly inspections for the 900 MW reactors will also begin, bringing further major industrial challenges and a tight schedule.

Availability profiles vary considerably over time and from one plant to another.

Over the last decade, the plants that make up France's nuclear fleet have had widely differing availability rates. For a given power plant, availability varies considerably over the period depending on the life of its reactors, including major outages such as the ten-yearly inspections but also the hazards affecting their operation and maintenance. Over the period as a whole, overall availability also varies from one plant to another. For example, the two power plants where the stress corrosion phenomenon was first identified, and which were the most affected by the episode, can be clearly identified: the four reactors at the Chooz and Civaux power plants, all belonging to the 1,450 MW range, were shut down for the whole of 2022. 

On the other hand, some plants have high availability rates of over 80% in certain years: these are years in which no major maintenance was planned on any of the reactors at these plants. Such years were few and far between over the past decade due to the sustained programme of long-term maintenance, and particularly the Grand Carénage major refit programme (see figure 2.8). Apart from years that are exceptional for one reason or another, the causes underlying a power plant’s availability level are very varied, and often specific (see below).

The overall availability of the majority of power stations over the decade from 2015 to 2025 was in the 60–75% range. Apart from generic maintenance massive development of the nuclear fleet from the late 1970s onwards, the country could call on abundant electricity generation, making it possible different and depends on specific events and hazards. Nevertheless, the following observations can be made: 

• the four reactors in the 1,450 MW range, the most recent apart from the EPR, have a significantly lower rate than the average: this is because this was the range most affected by the stress corrosion phenomenon;
• the 1,300 MW plants all performed better than average, with the exception of the Paluel and Flamanville plants, which were faced with special circumstances⁸.

The nuclear share of France's electricity mix is an important factor in the operation of its reactors and the power system as a whole. 

In most countries with nuclear power, the plants were built to operate as baseload generation. In France, given the importance of nuclear as a share of the electricity mix, the reactors were designed from the outset to modulate their electricity generation when in operation and thus adapt to variations in electricity consumption, both between seasons and on a daily or weekly basis. With the to plan for a reduction in dependence on fossil fuels: as a result, electric heating was encouraged more than in other countries, leading to significant variations in consumption between winter and summer. Furthermore, the nuclear fleet has historically adapted to the fact that consumption is generally lower at night and at weekends. 

In addition, given the large number of reactors in service, it is extremely unlikely that the entire nuclear fleet will be available at the same time, due to the industrial constraints on maintenance schedules, which mean that maintenance shutdowns are spread out over the year. As a result, French nuclear generation almost never reaches its maximum installed capacity. In other countries, with much smaller numbers of reactors, the fleet produces at maximum capacity almost every year, for varying lengths of time

The reasons that may cause the nuclear fleet to be unavailable differ widely. As an outline, starting from the maximum annual energy that can theoretically be generated, i.e. the energy that would be generated if the whole fleet operated without interruption at its rated power level and under reference weather conditions⁹, a number of effects of very different kinds limit the actual generation level of the fleet

1. Lower production due to the seasonal variability of the reactors' maximum effective power, which depends partly on the temperature of the cooling source. As this can vary from the reference conditions at which the reactors' rated power is calculated, these deviations tend to slightly reduce the maximum energy that can be generated over the year.
2. Planned or unplanned outages. These include ten-yearly inspections (“VD”, usually lasting around six months), partial or periodic inspections (“VP”), outages for refuelling (about every 12 to 18 months, usually lasting one month) and outages for routine maintenance or testing. Outages may also occur due to technical faults in a reactor, industrial action or environmental constraints such as the temperature of the cooling source being too high, particularly in the case of power stations located along rivers. For a given reactor, the operator generally alternates between partial outages and shutdowns for refuelling only. Ten-yearly inspection programmes are explicitly required by law, while the frequency of refuelling outages is dictated by the fuel cycle. The exact schedule for partial inspections is determined on a caseby-case basis by the operator, working with the safety authorities. These periods of unavailability correspond to operations that may be complex and are often combined; in addition, although these are scheduled operations, unplanned schedule overruns can occur. As a result, the actual duration of this category of shutdowns may differ from the standard durations. Given the repercussions for the performance of the fleet, the operator takes particular care to control these longer shutdowns¹⁰. 
3. Energy not generated due to modulation: this may involve reactors contributing to balancing the system, modulation due to a lack of economic outlets or modulation for fleet management purposes. These modulations can take the form of shutdowns varying in length (over a weekend, for example), or power reductions with the reactor still running. An example of modulation for fleet management purposes is modulation to save fuel, which depends on the schedule of refuelling outages for the reactor concerned and overall fleet scheduling. As the refuelling interval is limited (between 12 and 18 months), if a reactor is subject to heavy demand and/or if the next outage is scheduled a little later than usual (typically to wait for the end of winter), it may at some point become necessary to save a certain amount of fuel to “hold out” until the next refuelling¹¹. France's nuclear fleet was designed from the outset to incorporate a degree of modulation into its operation.

In 2025, outages for refuelling and routine maintenance (whether combined or not) remained by far the leading cause of energy not generated (compared with the theoretical maximum volume), with a total impact of around 80 TWh (approximately 50 TWh for partial inspections and 30 TWh for outages for refuelling). These operations occur frequently and in large numbers, which allows for a degree of expansion and explains the relative stability over time – apart from exceptional situations such as the pandemic or stress corrosion. As far as the tenyearly inspection programmes are concerned, 2025 was characterised by a relatively light schedule. Eleven reactors had undergone a ten-yearly inspection at one time or another in 2024. At the beginning of February 2024, eight reactors were shut down simultaneously for ten-yearly inspections, which is quite exceptional: this was made possible partly by a comfortable situation in terms of security of supply. In 2025, only six reactors were shut down for this reason. In total, it is estimated that the energy that could not be generated due to ten-yearly inspections in 2025 was much lower than in 2024 (27 TWh compared with 46 TWh), which contributed to the high availability levels recorded. 

Modulation (across all types) represented around 30 TWh in 2025, a relatively stable volume compared with the previous year. This is not a new phenomenon. For example, comparable or even higher levels were reached in the 1990s and early 2000s, in a context where generation capacity exceeded consumption. Over the last few years, while the volumes have been comparable, the modulation profile has changed, and the amount due to a lack of economic outlets has increased. In the early 2000s, modulation mainly took place at night during the summer. Recently, the phenomenon has been more prevalent in the afternoons, particularly at weekends between April and October (when prices are lower). The changes associated with this new modulation landscape in the nuclear fleet were documented in detail in the 2025 Generation Adequacy Report.

Finally, environmental constraints in 2025 led to a reduction in production of less than one terawatt-hour. The effects associated with this factor thus remained limited this year, as did the effect of strikes¹² (also less than one terawatt-hour) and network constraints (negligible volumes).

After reaching a level in 2024 that had not been seen since 2013, hydropower generation fell sharply in 2025 (−12.9 TWh/−17.2%), reaching 62.4 TWh, a level close to the average annual output over the last decade (61.4 TWh on average between 2015 and 2024). 

The fall in hydropower generation between 2024 and 2025 essentially reflects lower rainfall in 2025 than in 2024: whereas 2024 was one of the ten wettest years since 1959, cumulative rainfall in 2025 was close to the norm¹³.

After falling by more than 10% in 2024, onshore wind generation recovered slightly in 2025, reaching 43.9 TWh (+1.1 TWh/+2.5% compared with 2024). This shift was due to growth in onshore wind capacity over the course of the year, and to wind conditions in 2025 that were broadly similar on average to those in 2024 across mainland France (with an improvement in the western and central regions and a slight decline in the north and east of France), although they remained unfavourable compared to the average over the previous decade (see below). 

On average over 2025, output from onshore wind covered around 10.0% of national consumption. This coverage rate was slightly up on the previous year (9.6% for 2024) due to the relative stability of gross consumption.

The load factor for French onshore wind power was 21.4% in 2025. This is an exceptionally low level, following a decline that had already been seen in 2024.

The low load factor for onshore wind power is primarily due to wind conditions. The average wind speeds recorded in 2025 were below their average levels over the last decade for the whole of mainland France, except in the north of the Nouvelle-Aquitaine region. This wind deficit was more pronounced in the northern half of France, where most of the onshore wind capacity installed on the mainland is concentrated.

Over the course of 2025, French onshore wind farms occasionally modulated their output downwards, either in response to economic incentives during episodes of negative spot prices (the majority by volume), or when RTE called on them to guarantee the balance of the power system (for more limited volumes; see the “Modulation” section). Over 2025, these modulations created a downward effect on onshore wind generation of around 1.3 TWh, more than a third higher than the estimated level for 2024. Overall, modulation of onshore wind generation – whether in response to economic market conditions or at RTE's initiative – was responsible for a drop of around 0.7 points in the load factor compared with a theoretical situation with no modulation.

At the end of 2025, onshore wind capacity in operation in mainland France represented 23.9 GW. With a further 0.9 GW coming on stream in 2025, France's onshore wind fleet continued to expand. However, the rate of growth slowed in 2025 for the third year running, reaching its lowest level since 2013.

French offshore wind generation in 2025 stood at 5.7 TWh, an increase of almost 43% (+1.7 TWh) compared with the 2024 level. 

This increase largely reflects growth in offshore wind capacity. The Saint-Brieuc and Fécamp offshore wind farms were gradually commissioned over the course of 2024, and their entire installed capacity (just under one gigawatt for the two farms combined) has been operational since May 2024. In 2025, a fourth commercial wind farm whose turbines are installed offshore off the islands of Yeu and Noirmoutier was gradually brought on stream (411 MW out of a total of 488 MW were operational by the end of 2025). In addition, the 25 MW of France's first floating offshore wind farm pilot¹⁴ were commissioned in June 2025 in the so-called “Faraman” zone, 17 km off the coast of Port-Saint-Louis-du-Rhône. As of 31 December 2025, the total installed capacity of the offshore wind fleet off French mainland coasts was 1.9 GW.

The load factor for commercial offshore wind farms¹⁵ commissioned before 1 January 2025 was 38.8% over 2025 as a whole¹⁶. In 2024, the load factor for the Saint-Nazaire wind farm, the only one operating at full capacity for that year, was 32.9% (compared with 36.3% in 2025). The increase in the load factor for French wind power is due both to the commissioning of new wind farms in more favourable locations and to the higher availability of the SaintNazaire wind farm in 2025 compared with 2024¹⁷.

Offshore wind capacity off the French coast is set Courseulles-sur-Mer and 60 MW of floating wind to continue growing in 2026 with the commis- turbines at two sites off Leucate and Gruissan. sioning of around 450 MW of wind turbines off 

French solar generation stood at 32.9 TWh in 2025, up 32.7% on the 2024 level.

Electricity generation from the solar sector covered an average of 7.5% of consumption in mainland France in 2025. This rate has gradually increased over the last decade, in line with the momentum of solar generation.

The increase in French solar output results from growth in installed solar capacity (see below) and the significantly better sunshine conditions in 2025 than those experienced in mainland France in the previous year: whereas 2024 was one of the least sunny years that France has seen over the last thirty years, sunshine over the year in 2025 was slightly higher than the average for the previous ten years (2015–2024)¹⁸. The surplus sunshine over the past year was more pronounced in the northern half of mainland France than in the south, where most of the country's solar electricity generating capacity is concentrated.

The improvement in sunshine conditions resulted in a sharp rise in the average annual load factor for the French solar fleet, which stood at 13.6% in 2025, compared with 12.8% in 2024. However, the load factor for solar power in 2025 was still below its tenyear average (14.4%), despite slightly more sunshine than normal. This relative weakness is partly due to the fact that French solar producers modulated their output downwards more in 2025 – in response to economic incentives during periods of negative spot prices or at the request of RTE to guarantee the balance of the power system – than they have done historically (see the “Modulation” section). Modulation of solar power represented around 1.6 TWh in 2025 (RTE estimate), i.e. almost two and a half times more than in 2024. This modulation was responsible for a drop of around 0.8 points in solar power's load factor in 2025 compared with a theoretical situation with no modulation. 

the end of 2025, the cumulative capacity of solar power generation facilities in mainland France was 30.4 GW. Over 2025, 5.9 GW of new capacity was installed, continuing the acceleration in the growth of French solar capacity observed since 2021.

Due to the hourly profile of solar generation, with a peak in the late morning lasting until early afternoon, the growth of solar capacity in France and Europe has distorted the hourly profile of electricity prices, resulting in low or sometimes negative prices during these times, particularly in the spring (see the “Prices” chapter). This reveals a need to develop the flexibility of the power system to guarantee that supply and demand are balanced, in terms of both generation flexibility (the downward modulation of renewable production, for example, which must be controlled to avoid too-sudden variations) and consumption flexibility. In particular, the peak/off-peak signal system introduced in the 1960s has been adapted to include the early afternoon among the off-peak hours.

Negative spot prices result from anticipation of surplus generation compared with consumption and trading forecasts. These negative prices then constitute an economic signal incentivising all producers who have the technical capacity to do so to reduce their output or stop generating. This incentive applies in particular to nuclear power units (which modulate their output based on market prices), dispatchable hydropower capacity, storage facilities and a proportion of solar and wind generation.

In 2025, episodes of negative prices led to around 3 TWh¹⁹ of solar and wind output not being generated. This is almost double the 2024 volume.

The increase in the modulation of wind and solar generation during negative price episodes between 2024 and 2025 is linked to negative prices occurring more frequently (see the “Prices” chapter), the strong increase in solar generation (in France and the rest of Europe) and the growth of capacity that can be curtailed during negative price episodes.

The majority of onshore wind capacity and some photovoltaic installations are now being developed under the premium scheme (“complément de rémunération”). This framework includes incentives for the installations involved to stop production when the spot price is negative. Capacity developed under the premium scheme, like non-supported installations²⁰, is therefore likely to stop generating during periods of negative spot prices. Between 2024 and 2025, this capacity increased by 3.3 GW.

The remainder of the wind and solar fleets are covered by purchase obligation contracts²¹. Historically, the arrangements for remunerating generation under these contracts have not encouraged producers to modulate their output in line with market prices. However, since the end of December 2025, solar and wind installations with an installed capacity of more than 10 MW²² have had a financial incentive to reduce or stop their production at the request of buyers with purchase obligations²³. This provision could therefore lead to some wind and solar farms with purchase obligation contracts curtailing their output during periods of negative spot prices from 2026 onwards. 

The modulation of renewable output contributes to the balance of the power system, but this must now be controlled in order to ensure the balance between supply and demand is managed safely in as near-real time as possible. Stopping or restarting very large amounts of generation at the same time can lead to frequency disruptions, risking the operation of the system, or to the use of costly adjustment resources. The 2025 Finance Act and its implementing decree of December 2025 introduced a number of measures aiming to smooth renewable generation shutdowns and restarts, whether they are caused by negative price signals or a buyer with a purchase obligation.

The legislator has given RTE responsibility for balancing electricity supply and demand²⁴. As a result, RTE can request generation adjustments in as close to real time as possible to guarantee the safety of the power system. In particular, when the decisions taken by market players result in surplus power generation, RTE can order cuts in output. Solar and wind capacity can now be called on to adjust their output downwards in order to guarantee the balance of the power system. 

Downward adjustments requested by RTE in 2025 led to the curtailment of 0.1 TWh of solar and wind generation. Although the contribution of wind and solar to the total downward adjustments requested by RTE over the year as a whole remained moderate compared to other power sources²⁵, the volumes of solar and wind generation adjusted rose sharply. Their level in 2025 was almost five times higher than in 2024, while the downward adjustments made by RTE across the generation mix remained stable over the same period (+0.3%). 

The greater reliance on wind and solar to guarantee the balance of the power system is partly due to the fact that the volumes generated by hydropower and thermal plants fell significantly between 2024 and 2025, reducing the scope for these facilities being called on to adjust their output downwards. It also coincides with increasing participation by these installations in the balancing mechanism. The wind and solar capacity that can be mobilised under this mechanism has increased tenfold over the past year, reaching almost 5.6 GW by the end of 2025. This growth results from successive changes to the legal framework over the course of 2025. 

The 2025 Finance Act of 14 February 2025 enabled all plants supported by a purchase obligation or the premium scheme to take part in the balancing mechanism. This option has been available since 1 October 2025 and applies across the generation mix regardless of power thresholds. Participation in the balancing mechanism was even made compulsory by the “DDADUE” law passed at the end of April 2025²⁶, which introduces an obligation for all installations with a capacity greater than 10 MW²⁷ – including solar and wind farms – to offer as much capacity as is technically available to the balancing mechanism. This obligation came into force on 1 January 2026, and some players were able to comply earlier.

At the same time as these legal and regulatory changes, amendments to the purchase obligation contracts for the Saint-Nazaire, Fécamp and SaintBrieuc offshore wind farms meant that these farms have participated in the balancing mechanism since spring 2025. As a result, adjustments to offshore (7.2 GWh), the first year in which RTE resorted to wind generation at RTE's initiative (via the balancing downward adjustments to offshore wind generation mechanism or emergency orders) amounted to 63.8 to ensure the balance of the power system.

French fossil-fired electricity generation totalled 18.7 TWh²⁸ in 2025. For the second year running, this was the lowest level since the early 1950s; in 2024, fossil-fired generation stood at 19.9 TWh, which was already the lowest total recorded during this period. 

Gas, coal and oil-fired power stations, which are by far the most emissions-intensive, generated less than 4% of the total electricity generated in France in 2025. While the French electricity mix as a whole has long been low-carbon thanks to the major contribution from hydroelectric and nuclear power, fossil-fired generation still accounted for around 10% of production at the turn of the 2010s, with around a third from coal. 

The momentum seen since then leads to two conclusions: firstly, fossil-fired generation as a whole is now almost residual in the French mix; and secondly, of the three main conventional fossil fuels, gas, by far the least emissions-intensive, now accounts for the bulk (87%) of the remaining fossil-fired generation in France. Electricity generated using oil and coal has almost entirely disappeared: in 2025, these two generating sources accounted for 1.7 and 0.7 TWh respectively, less than the electricity produced from biogas.

Gas-fired generation is usually divided into three main categories in terms of their place in the power system and their operation:

• combined-cycle gas turbines (CCGTs): these are power stations that combine a combustion turbine with a steam turbine (which uses the heat from the first turbine's exhaust gases); this is the most modern technology and offers the best efficiency (between 50 and 60%) and thus the lowest emission intensity of all fossil fuels. This category accounts for most of the so-called “dispatchable” gas-fired generation in France. The marginal cost of CCGTs is competitive compared with other fossil-fired generation methods, which means that they can be called on during periods of high consumption, such as in winter or, to a lesser extent, during heatwaves. The Landivisiau plant, the most recent to be commissioned in France in the summer of 2024, falls into this category;
• conventional gas-fired power plants, including combustion turbines: these are less optimised because, unlike CCGTs, they are based on an open cycle, i.e. with no recovery of the heat produced during combustion. This makes them less efficient (between 30 and 40%) and their output is more emissions-intensive. France generates very little electricity at this type of plant: they only operate at peak times or, in some cases, to contribute to the short-term balancing of the power system or to resolve network pressures;
• cogeneration²⁹: this refers to facilities that produce both heat (e.g. for industrial needs) and electricity. For this type of installation, electricity generation is generally secondary to heat production. They are usually attached to a main installation, often a district heating network, or sometimes an industrial facility. Output from cogeneration is characterised by the fact that it is driven by the heat requirements of the underlying process, and depends to a lesser extent on the economic conditions that govern electricity generation in the short term. From the viewpoint of the power system, therefore, this is partly inflexible generation.

Based on this categorisation, gas-fired generation can be broken down broadly into two parts:

• a “base” level corresponding to the continuous output of the different types of cogeneration. As cogeneration linked to heating networks accounts for a large proportion of cogeneration in France, this base level varies according to the season: for 2025³⁰, it is estimated at around 1,600 MW during the heating period (between November and April), and 250 MW in summer (a level that corresponds more to industrial cogeneration of various kinds); in addition to this base level, a dispatchable or market-sensitive portion, essentially made up of CCGTs and, to a lesser extent, conventional power plants, including combustion turbines

There are also hybrid systems, which operate in specific ways, with part of their output as the base level but with the possibility of increasing their production beyond this base if the economic conditions justify it. This is the case, for example, with the Dunkirk combined-cycle gas power plant (see below).

In 2025, gas-fired power plants generated a total of 16.4 TWh. The volume of generation that can be considered as operating independently of the power system is around 7 TWh, or just under half. This corresponds to the output of cogeneration, and other thermal sources whose behaviour can be likened to cogeneration. The rest of gas-fired generation comes essentially from CCGTs. In other words, the generation actually called on in order of economic precedence represented only 9 TWh, or less than 2% of all the electricity generated in France.

Some CCGTs linked to industrial sites can also choose to increase their production beyond the base level due to cogeneration from industrial processes, depending on a trade-off between market prices and the constraints of the industrial processes to which they are linked. This is the case with the Dunkirk combined-cycle power plant (also known as DK6), for example. This plant, with a total capacity of around 800 MW, consists of two CCGTs connected to the ArcelorMittal steelworks, from which they recover some of the steelmaking gases. When operating solely to burn steelmaking gases, the two units produce around 80 MW each continuously; when market conditions³¹ justify it, the plant can increase its output to its maximum capacity (almost 800 MW). These two levels, as well as the periods of highest output and periods of complete or partial shutdown of the site, are clearly visible in the annual production profile (Figure 2.34). Analysis shows that the DK6 plant very rarely operated beyond the base level linked to the combustion of steelmaking gases in 2025, indicating that market conditions causing the plant to start up at its rated output were infrequent.

The majority of gas-fired generation, mostly combined-cycle plants, takes place during clearly defined periods, corresponding to periods of high consumption, and particularly high residual consumption³². This is true in winter, when these situations are most frequent, but also – more occasionally and to a lesser extent – in summer, particularly during hot spells³³. For example, a third of annual gas-fired generation is concentrated on the coldest 10% of days, representing around 35 days

Finally, there is a slight inverse relationship between the level of gas-fired generation in France and exports: the highest levels of exports are recorded when gas-fired generation is at its lowest. 

On 31 December 2025, total battery storage in service in France amounted to 1.6 GW. Around a third of this capacity is connected to the transmission network, and the rest to the distribution network. In principle, the higher the voltage level, the higher the capacity of the installation. The majority of batteries in service in France today are thus of modest size, in the megawatt range. The most powerful batteries, connected to the transmission grid, are few in number (just over 20 at the end of 2025).

Diffuse batteries, i.e. small-capacity, low-voltage batteries installed in homes or small businesses, are underdeveloped in France, with a total of just 140 MW³⁴. In other countries, this type of installation represents a significant proportion of ambitions for the growth of storage (see the Europe chapter).

For more powerful batteries, the waiting list, i.e. projects for which developers have already secured access to the grid, had reached around 14 GW by the end of 2025³⁵. Most of these projects are connected to the electricity transmission network (high and very high voltage).

It now seems clear that not all of these projects will come to fruition³⁶, or at least not in the next few years: as with all types of grid connection projects, most of the projects on the waiting list have not yet reached the point of a final investment decision. Analysing the battery connection waiting list in detail suggests that this sector is likely to develop in a very different way from what has been seen so far. The general trend is towards bigger installations: almost 90% of the total capacity on the waiting list involves transmission system connection projects. This corresponds to around 180 projects. Even within the transmission system, the size is tending to increase, with an average capacity of almost 75 MW for projects on the waiting list compared with an average of 25 MW for the batteries already connected to the transmission network.

Generally speaking, battery storage installations can take part in a wide range of mechanisms and markets:

• providing system services, such as frequency regulation; 
• taking part in wholesale markets, and particularly intraday markets: this involves storing surplus generation at certain times of the day and releasing it during daily consumption peaks. The typical example of this type of use is to take advantage of solar power, which is often in surplus in the middle of the day;
• contributing to the management of network pressures;
• taking part in the capacity mechanism, and thus in ensuring security of supply.

Today, however, the main activity of the batteries already in service in France is still the provision of system services, and particularly system services linked to frequency regulation.