Chapter 5: Nuclear

A tale of two countries

France and Britain's energy policies were remarkably similar until the 1970s. Both countries had nationalised their electricity industries after World War II, both relied heavily on coal for power generation, and both were developing civilian nuclear programs. But when the 1973 oil crisis hit, their responses couldn't have been more different. France launched the ambitious Messmer Plan - a massive state-driven program to shift almost entirely to nuclear power. Meanwhile, Britain's response was characteristically muddled.

What makes Britain's failure even more striking is that it started with a clear lead. The UK connected the world's first commercial nuclear power station, Calder Hall, to the grid in 1956; by the early 1960s it had a fleet of Magnox reactors in operation and was exporting nuclear expertise to countries like Italy and Japan.

France, by contrast, lagged the UK through the 1960s and only leapt ahead after decisively committing to both improving energy self-sufficiency and technical consolidation in the 1970s. The Messmer Plan—named after Prime Minister Pierre Messmer and launched in response to the 1973 oil crisis—was a bold national strategy that standardised on pressurised water reactors (PWRs) and created a comprehensive domestic supply chain. This single-minded focus on nuclear power would see France build 56 reactors in just 15 years, while Britain stuck with a sequence of bespoke designs (Magnox → Advanced Gas‑cooled Reactor → the never‑completed Fast Breeder → belated PWRs), each change resetting learning curves, complicating procurement, and delaying delivery.

The discovery of North Sea oil and gas in the late 1960s had already complicated Britain's energy planning. Should the country double down on nuclear power like France? Or rely on this new domestic fossil fuel bounty? The result was policy paralysis. While France built 56 nuclear reactors in just 15 years, Britain's nuclear program became mired in debates about reactor designs, locations, and costs. The contrast in outcomes is stark: today nuclear provides over 70% of France's electricity at some of the lowest prices in Europe, while Britain's aging nuclear fleet supplies just 15%, is owned by the French state supplier EDF and is mostly due for retirement.

That said, the French approach isn't without long term challenges of its own. Having built so many nuclear reactors so quickly, the French grid effectively became saturated with nuclear energy. To find markets, France had to expand interconnector cables with each of its 7 continental neighbours and Britain. When the saturation occurred in the 1980s and 1990s, France also had to stop building new nuclear reactors. The massive, geographically mobile workforce and domestic supply chain which had grown up around the nationwide attempt that built 56 nuclear reactors in 15 years started to disintegrate. Talented workers went elsewhere or retired. Intermediate suppliers ramped down production facilities. In the years between 2002 when Civaux was finished and 2024 when Flamanville C was finished France didn't open a single new nuclear reactor. This important point is often ignored in British discourse comparing energy policy on either side of the channel. However, with 14 nuclear reactors representing 14 GW of capacity expected to retire in the 2030s, 22 GW in the 2040s and 20 GW in the 2050s, France now faces a cliff edge. Life extensions to nuclear plants are commonplace and substantial extensions are expected, inevitable and unlikely controversial for safety. However, the technology which was supposed to epitomise stability and self-sufficiency in energy production had instead created a multi-generational boom-bust cycle.

Financially, the French approach is also laying the foundations for longer term challenges. The investment in nuclear energy assumed a 0% discount rate. Given the vast upfront and decommissioning costs of nuclear power, there is a huge financial burden on future generations of decommissioning and containing the nuclear waste. It's quite difficult to measure the lifetime (or lifecycle) cost of nuclear in terms that you can compare to gas or other energy technologies. However, according to the French energy regulator CRE's 2022 estimate, the lifetime levelised cost of electricity (LCOE) for French nuclear power is around 61 EUR/MWh in 2022 prices (equivalent to approximately 67 EUR/MWh in current prices after 3 years of inflation), while gas-fired electricity costs 40-80 EUR/MWh depending on gas prices and plant efficiency.

The main advantage of the French approach is stability. When wholesale energy prices spiked in 2022, France should at least in theory have been well-placed to ride out the period of high prices. The cost of nuclear power generation in the short-run is relatively low, nuclear fuel costs only about 5-7 EUR/MWh generated compared with a peak of 500-600EUR/MWh in 2022. The value of nuclear exports should have massively boosted France's balance of trade and the trading results of its state-owned electricity generator, EDF. And even if energy prices in France were allowed to rise, to respect the country's commitments to the EU single energy market, there would be considerable tax revenue from France's energy companies, most significantly EDF to offset the cost of any state support afforded to households and businesses.

However, EDF's financial position has deteriorated significantly in recent years. The company's debt ballooned to over €60 billion by 2022, driven by the costs of maintaining France's aging nuclear fleet and extensive repairs over the years. The French government was forced to fully renationalise EDF in 2022, buying out minority shareholders at a cost of €9.7 billion, reversing the privatisation of 15% of the company in 2005. The liability for decommissioning France's nuclear plants rests with EDF, though the French state has created dedicated funds to cover these costs. However, these provisions are widely regarded as insufficient, with the French Court of Auditors estimating that EDF's nuclear decommissioning and waste management liabilities could reach €74 billion - exceeding the company's total market value and highlighting the long-term financial burden that nuclear power places on the state.

Compounding these financial problems, France didn't see the expected windfall from nuclear exports in 2022 for a number of reasons:

Nuclear Fault - In 2022, France's nuclear fleet was operating at just 60% capacity due to widespread corrosion issues discovered in key reactor components. This forced the shutdown of 32 reactors for safety inspections and repairs, dramatically reducing France's nuclear output and export potential.

Hydro Drought - France's second-largest electricity source, hydroelectric power, was severely affected by drought conditions across Europe. Reservoir levels fell to historic lows, cutting hydro generation by 40% compared to normal years and forcing France to import electricity rather than export it. Heatwave conditions also restricted cooling water intake on some river sited nuclear plants for a while too.

The combination of these two crises meant that France, normally Europe's largest electricity exporter, became a net importer for the first time in decades. France imported 16 TWh of electricity in 2022 - equivalent to the annual consumption of 4 million households. This reversal from France's typical 50 TWh annual exports contributed significantly to Europe's energy shortage during the Ukraine crisis, pushing wholesale electricity prices to €400/MWh in August 2022 - ten times the pre-crisis average. For context, France's annual electricity consumption is around 450 TWh, in Britain it is 300 TWh. France's sudden need for gas-fired power from neighbouring countries like Britain, Italy and the Netherlands had a particularly marked effect on prices. This demand brought less efficient, peaking gas power stations into operation, which due to marginal pricing forced up the cost of power more generally in Europe.

Current British nuclear programme

There are just 5 nuclear plants left online in Britain, each of which is expected to close shortly:

1. Hartlepool - Advanced Gas-cooled Reactor (AGR), 1,190 MW capacity, expected closure 2026
2. Heysham 1 - AGR, 1,155 MW capacity, expected closure 2026
3. Heysham 2 - AGR, 1,320 MW capacity, expected closure 2028
4. Torness - AGR, 1,360 MW capacity, expected closure 2028
5. Sizewell B - Pressurised Water Reactor (PWR), 1,198 MW capacity, expected closure 2035

All except Sizewell B are Advanced Gas-cooled Reactors built in the 1980s and early 1990s. Sizewell B is Britain's only Pressurised Water Reactor, built in the 1990s as a prototype for a wider PWR programme that never materialised. Together, these plants currently generate about 15% of Britain's electricity, down from a peak of around 25% in the 1990s.

Hinkley Point

The vast new nuclear plant at Hinkley Point C (3,200 MW) has been under construction since the year 2017. When completed it will be the largest nuclear plant ever constructed in Britain. Hinkley Point C is an EPR reactor.

The construction has been a disaster. It was originally supposed to open in 2023, but is now not expected until 2030 at the earliest. Every year that its completion is delayed, the National Grid is without 3GW of low carbon electricity, which is invariant to weather conditions and reduces the need for gas generation and imports.

Not surprisingly, the cost of construction has also ballooned, from an original estimate of £16 billion to current projections of £32-35 billion. Thankfully for the British taxpayer and billpayer, because of the way the CfD contract is structured, EDF will be remunerated only when the power station generates electricity. However, the negotiation with EDF that took place before construction began guaranteed EDF a very high price for Hinkley's output; £128/MWh in 2025 prices, which will rise in line with inflation. This is slightly more expensive than the latest expected costs for offshore wind (about £116-117/MWh)

As noted above, this price is about double the average cost of EDF generating nuclear power in France and is higher than the current UK wholesale price, meaning it will be subsidised by billpayers. One could even argue that the delay is saving customers money; at least until power prices rise.

More nuclear plants

A follow-on nuclear project at Sizewell C has been announced by the UK Government, replicating about 85% of the EPR design from Hinkley Point. As with Hinkley Point, by using an existing nuclear site, it is hoped that the local population will be more amenable to the project, recruiting staff with the right skills will be easier, and there will be an existing grid connection to use.

The financing model for Sizewell C is called RAB (Regulated Asset Base) and is the opposite of Hinkley Point; the taxpayer/billpayer will finance and take on the risk of construction. EDF was reluctant, especially after Hinkley Point to repeat the CfD arrangement. RAB (Regulated Asset Base) is the same as the water industry and some other privatised monopoly utilities.

If lessons are learnt from Hinkley Point, it's feasible that Sizewell C could end up being cheaper or at least that not much more expensive than Hinkley Point. Hinkley Point is the first in a generation nuclear project in Britain, and the first time an EPR reactor had been built in the UK. If some savings passed through to Sizewell C, the billpayers would benefit from the savings, unlike the CfD arrangement where savings would accrue to EDF.

However, underspend on nuclear plants is not the norm! And substantial increases in interest rates since 2012 (when they were close to zero), mean that with the best engineering advances, it's much more likely that, should things go wrong at Sizewell, billpayers will end up paying for any increases in the budget and suffer the consequences of any delays to the generation of power from it.

Bloomberg New Energy Finances has recently estimated the cost of Sizewell C's power in levelised cost terms (a sort of average NPV or net present value) at £286/MWh; which is extremely high. It's no surprise perhaps that EDF is choosing to ditch the EPR and build a lower price simpler reactor for the 6 new projects it intends to build in France by 2038. However, this begs the question whether the UK is wise to continue using the combination of EDF and the existing EPR design.

Is nuclear reliable?

Traditionally, the strongest argument for nuclear power is its supposed reliability. Compared to the main low carbon alternatives (wind and solar), nuclear is "dispatchable", you can control if a nuclear reaction is to take place and the amount of power that's produced. Unless a drought prevents a nuclear plant getting cooling water, you're generally not affected by weather, and a nuclear plant is designed to run 80% of the time, with planned outages just for maintenance.

More recently, this reliability has come into serious doubt for two reasons:

Construction Delays

Every nuclear plant built in Europe in the last 25 years has been severely delayed and over budget. France's Flamanville 3 (1,650 MW) was under construction from 2007 to 2024 - 12 years later than originally planned, with costs escalating from €3.3 billion to €19.1 billion (€11,600/MW). After finally opening in 2024, it was immediately shut down due to technical issues and operational problems. Finland's Olkiluoto 3 (1,600 MW) took 18 years to build (2005-2023) and cost €11 billion compared to the original €3 billion estimate (€6,900/MW). Britain's Hinkley Point C (3,200 MW) is currently projected at €37-41 billion versus an original €19 billion estimate (€11,600-13,000/MW). Hungary's Paks 2 project (2,400 MW) was originally planned for €12.5 billion but has been suspended due to escalating costs. The pattern is clear: nuclear construction in Europe averages 10-15 year delays and 3-6x cost overruns, with final costs of €7,000-13,000/MW. Note: In "How Big Things Get Done", author Bent Flyvbjerg highlights nuclear power plants as notoriously underperforming megaprojects due to cost overruns, findings which are mostly data driven.

An incomplete nuclear plant is useless. Not only do they not generate any power, but given the scale of investment involved, grid planners need to know with reasonable certainty when a nuclear-sized plant will be available. If it isn't, they must make provision for alternative firm capacity, as Britain is doing for Hinkley Point, which was supposed to be the largest power plant on our grid by 2023. The knock-on effects of nuclear delays ripple through the entire energy system - when a 3.2 GW nuclear plant is delayed, it affects grid planning, energy security, carbon targets, and consumer bills across the country. The absence of Hinkley Point C is particularly problematic because it would have generated over 3 GW of power in the South West, a region that already lacks adequate transmission capacity to import power from the North and East of England. A construction delay at a medium-sized onshore wind farm might cost a few hundred megawatts and be relatively easy to compensate for, but nuclear delays create system-wide planning challenges. In reality, we're fortunate that Britain has existing gas plants that can be kept in service rather than needing to build entirely new capacity to replace the delayed nuclear plant. However, it's worth noting that construction delays for wind (even offshore), battery storage, solar, and gas plants are much lower and more predictable than nuclear. This reliability in planning and delivery makes these technologies far more valuable to grid operators than nuclear power.

Unscheduled Outages

The French nuclear outages in 2022 occurred when a common fault was identified that affected a large number of reactors. These were the Pressurised Water Reactor (PWR) type, specifically the 1300 MW and 1450 MW models, and 32 reactors were built between 1980 and 2000 with the same design flaw - stress corrosion cracking in the emergency cooling system piping.

French nuclear plants also suffer frequent strike-related outages, with EDF workers regularly walking out over pay disputes, pension reforms, and working conditions, further undermining the reliability that nuclear power is supposed to provide. While UK nuclear workers are unionised and have the legal right to strike, their track record of industrial action is relatively limited compared to France - though wildcat strikes at Sellafield and Heysham in 2009 showed the potential for disruption.

Were Britain to replicate the French model and build a large number of identical nuclear plants, it might achieve economies of scale and reduce nuclear costs closer to French levels of 60EUR/MWh, which are about half the cost of Hinkley Point. However, this approach would expose Britain to the same systemic risk that France experienced in 2022, where a single design flaw could simultaneously take out a significant portion of the entire nuclear fleet. The critical difference with nuclear power is that faults cannot be ignored or tolerated - unlike gas turbines or wind turbines, nuclear reactors must be immediately shut down when safety issues are discovered, regardless of the economic consequences.

Small Modular - a potential way out

Small Modular Reactors (SMRs) have been touted as a potential solution to Britain's nuclear construction problems. Unlike traditional nuclear plants built on-site with bespoke designs, SMRs are designed to be manufactured in factories and assembled on-site, with proponents claiming this approach could reduce costs and construction times through standardisation. The government has already committed £210 million to the SMR design competition and a further £157 million to support Rolls-Royce's SMR programme, despite no SMRs yet being built commercially anywhere in the world.

Proponents suggest factory-built reactors could achieve levelised costs of electricity (LCOE) of around £60-80/MWh - significantly below Hinkley Point's £128/MWh - though these are theoretical estimates based on untested technology. The modular approach could potentially open up distributed deployment options that traditional nuclear cannot match, such as powering individual industrial complexes or data centres. Rolls-Royce brings submarine reactor experience to this market, and while they sold off their civilian nuclear business to Westinghouse in the early 2000s, they do have extensive civilian power generation expertise through their aero-derivative gas turbines, which are widely used in power plants worldwide.

While this is definitely one to watch, and if it works out could be a technology that Britain could export globally, the track record of nuclear construction in Europe suggests that theoretical advantages often fail to materialise in practice. However, given the downsides of centralised nuclear at Hinkley Point and elsewhere, modular reactors merit further attention and investment.