The breaking of Britain's National Grid
A story of aging infrastructure and perverse incentives
This summer Works in Progress is hosting Invisible College, a week-long residential seminar, to give people aged 18–22 a grounding in the topics most important to us: how the world got rich, what is going wrong with science today, and the political economy of housing, urbanism, and cities. For more information and to apply, go here. Applications will close on 28th April.
The UK’s energy transmission and distribution infrastructure is broken.
There is a 15-year queue for new connections. The average transformer is now more than 60 years old. UK billpayers fund more than a billion pounds a year of curtailment payments to switch off intermittent energy sources that the grid can’t handle. The UK came close to blackouts in January 2025.
So what went wrong?
The grid has been buffeted by three interlocking forces: the aging of transmission and distribution infrastructure, the loss of power generation capacity from fossil fuels, and the challenge of rapidly incorporating renewable energy. Part of this was the inevitable product of a renewables transition that is proving difficult around the world, but it has also been driven by inattention, drift, and short-termism.
By the year 2000, the majority of the UK’s coal plants had been in service for 30 years. 2001 EU regulation demanded that these older power plants either undergo expensive retrofits or opt out and operate on restricted hours until retirement. UK nuclear output had also been in decline, as first generation reactors like Calder Hall and Chapelcross reached the end of their lives in 2003 and 2004 respectively. No new plants have come online since Sizewell B in 1995.
The UK’s capacity margin – the buffer baked into the system measured as how much the total available electricity exceeds maximum electricity demand – fell from over 26 percent at the start of the millennium to low single digits by 2015.
In 2008, the UK adopted the Net Zero target’s predecessor, the Climate Change Act. This committed the government to a legally binding target of reducing greenhouse gas emissions by 80 percent from their 1990 levels by 2050.
The UK had previously attempted to incentivise the greater production of clean energy through the Renewables Obligation, which had been introduced in 2002. The scheme required suppliers to source an increasing percentage of their electricity from renewable sources (not including nuclear). Suppliers would receive certifications for each megawatt hour of renewable energy generated, and penalties if they failed to meet a percentage target that increased each year. Certificates could then be traded so that the firms that were best at expanding renewables generation could specialise in doing so.
The scheme did not live up to its theoretical elegance. Renewables Obligations Certificates were traded in a market place and their price was subject to significant volatility. For example, in a windier year, the amount of renewable energy generated would be higher, causing the value of certificates to drop. Long-term investment in renewable energy projects requires stable and predictable revenue to justify the upfront capital expenditure. By 2010, eight years after the scheme’s introduction, renewables made up only 6.9 percent of the UK’s electricity generation, falling short of its 10 percent target.
Enter the CfD
The introduction of Contracts for Difference (aka CfDs) in 2014 was designed to fix this problem. CfDs guarantee generators a price per kilowatt hour for fifteen years, removing the risks associated with volatile wholesale prices. This proved a boon for wind and solar projects, which tend to generate power cheaply, but generate a lot of their power when prices are low, zero, or even negative. But it was a nightmare for the grid.
On one level, wind and solar were always going to pose a challenge, as they are inherently more difficult from the standpoint of grid management.
The modern grid was designed around the large rotating generators used in nuclear, coal, and gas power plants, whose physical spinning mass naturally resists sudden frequency changes (this is known as inertia), keeping the current at a consistent frequency. Electricity grids rely on alternating current, which is electricity that regularly changes direction, switching back and forth in a regular pattern. Frequency is the rate at which this current oscillates. It is typically measured in cycles per second, or hertz.
Maintaining a stable frequency is critical. The generators that power the grid and devices connected to it are designed to operate at specific frequencies. If they deviate, they can overheat, experience mechanical stress, or break. This affects everything from electric clocks to industrial motors. The nightmare scenario is a cascading failure. This is where uncorrected frequency deviation causes a small number of generator trips, worsening frequency problems, and causing more generators to cut out.
This is why the grid has a set of emergency measures. If frequency moves out of the 49.8 to 50.2 hertz band, it will begin shutting off power to parts of the system to restore balance.
The electronic inverters that convert solar panels’ or wind turbines’ power to alternating current and connect it to the grid don't provide the inertia of large rotating generators, as they have none of the kinetic energy that stabilises mechanical turbines. Something as simple as a cloud passing over a solar installation can cause rapid fluctuations in output. If capacity comes offline in one area, generators elsewhere on the grid suddenly have to work harder. The added strain causes their turbines to slow. This means the system needs to replace potentially hundreds of megawatts of power within minutes to maintain stability.
The other challenge is geography. The vast majority of the UK’s onshore wind farms are in Scotland; solar projects are disproportionately concentrated in Cornwall, Devon, and Somerset; and offshore wind in the North Sea. Moving this electricity to where it’s actually needed depends on high-voltage transmission lines. These lines need to be built not for the average production over the year – typically 31 percent of nameplate capacity for wind and 10 percent for solar in the UK – but for maximum production, otherwise they will run into capacity limitations during peak production hours.
When energy then can’t be absorbed into the local electricity grid, the government has to pay operators to switch it off through curtailment payments, or discharge it into the ground. The latter is difficult to do at scale. The grounding infrastructure that exists is designed as a safety measure to handle faults and surges, not continuous excess generation. This would risk overloading it. Meanwhile, considering the near-universal dependence of solar and wind projects on subsidies, many operators would struggle to assume the financial risk for overgeneration.
The alternative approach would be building out grid infrastructure to handle the maximum possible capacity that wind and solar might produce, only to see it underutilized much of the time. On the UK’s biggest recent day for wind production, which occurred in December 2024, 22 gigawatts of the country’s 29 gigawatts of wind capacity was used. If we saw the same proportion deployed in a world where the government had hit its 2030 goal to expand wind capacity to 90 gigawatts, it would result in the UK’s daily energy requirement being generated nearly three times over. In this universe, billpayers would be on the hook for billions in new infrastructure and curtailment payments.
Coordination failures
These challenges have been exacerbated by a failure to coordinate generation and transmission planning. New wind farms or solar projects have often been approved with the surrounding infrastructure as something of an afterthought. In the UK, several different entities are responsible for generation, transmission, distribution, and supply, following privatization in 1990. However, the UK adopted a more hands-off approach to coordinating these entities compared to other countries with privatised or partially privatised electricity systems. For example, in the Nordics, there is a common cross-border network plan shared by different regional operators, while the private transmission network operators in Germany are overseen by a federal body.
The UK’s CfD system exacerbates these challenges. CfDs offer the same strike prices regardless of project location, encouraging development in windy or sunny but grid-constrained areas – potentially areas far from where the power will actually be used.
To meet these demands, the National Grid is engaged in a £60 billion ‘Great Grid Upgrade’, a mixture of power lines, substations, and an interconnector with Denmark. Adjusted for inflation, this is costing 20 times more than the construction of the original grid and supergrid combined.
The National Energy System Operator (NESO, pronounced Neeso or Nesso), the publicly owned body that has operated the energy system since October 2024, is also reforming the grid queue. Historically, the queue operated on a ‘first come, first served’ basis. Developers could join the queue speculatively without first demonstrating that their project is viable. If NESO’s reforms succeed, applications for viable generation projects will be prioritised.
This work has gained importance as the world faces up to the growing power demands created by AI data centers. The UK is an international outlier in the length of its grid queue, but others are also struggling. Ireland, Singapore, and the Netherlands have introduced temporary data centre construction moratoria.
In Loudoun County, Virginia, the ‘data center alley’ of the US, projects still take up to four years to get a connection, as grid operators refuse to approve new connections when they do not believe the infrastructure can handle the demand. Meanwhile, energy operators have warned of the potential for rolling blackouts. There has been political pushback against a $5.2 billion plan to expand and modernise powerlines in the state, as the upgrades would be paid for by increasing energy bills.
Texas is an anomaly in the US, with a grid that is completely isolated from the rest of the country. In other parts of the country, regional grid operators will typically wait for system-wide upgrades to be completed before allowing new connections. Texas’s energy system operator, by contrast, will allow new generation capacity to be connected to the grid, provided local infrastructure is sufficiently resilient as part of a policy called ‘connect and manage’. Operators simply have to shut off production when the grid is at capacity and receive no curtailment payments. Texas is crushing every other US state in its renewables buildout. Since 2019, Texan power firms have boosted solar generation capacity by 800 percent, wind capacity by 50 percent and battery storage capacity by 5,500 percent, 80 percent more than any other state.
Texas has some advantages that the UK does not. Texas can afford not to subsidise curtailment payments, because its climate is perfect for wind and solar. This means that there is less need to overbuild generation to account for intermittency. If the UK cut the £1 billion it pays out in curtailment payments to renewables projects, their economics would deteriorate even further.
Another way to get things built without depending on the overall strength of the grid, or burdening the residential billpayer, is to allow major energy users to build their own small grids. Microsoft has signed a deal to reopen a unit of the Three Mile Island nuclear plant in Pennsylvania. Meanwhile, Google has made an advanced market commitment to buy the output of Kairos Power’s small modular reactors. But these systems still require a grid connection, because even large on-site generation systems can't economically provide the extreme reliability (99.999 percent, or ‘the five nines’) data centers demand. Large computing facilities require extremely stable power with precise voltage and frequency. The grid provides a reference standard that on-site generation must synchronize with.
Closing thoughts
As grids enter their next era, this age of improvisation will need to come to an end. Aggressively subsidizing renewables development, regardless of location or economic viability, may help with incremental progress towards the UK’s net zero goal, but has stored up significant challenges. As the UK prepares to pay large numbers of solar farms to switch off for the summer, the government will need to face up to the perverse incentives in the current system.
There are also legitimate questions to be asked about the balance between baseload power (e.g. traditional fossil fuels or nuclear energy) versus intermittent power sources. Greater use of baseload power would resolve many of our infrastructure planning interests. As John Fingleton’s Nuclear Regulatory Taskforce gets to work, they need to be given the space to make bold recommendations.
The alternative is a worsening version of the status quo: an inefficient energy transition, dependent on ever growing subsidy. The political cost of even higher energy bills and greater deindustrialisation is unlikely to be one that the government will want to pay.
Alex is an editor at Works in Progress, focused on AI and energy. He’s also the author of Chalmermagne, a Substack covering technology, policy, and finance.