Last Updated on August 15, 2026 by Karl Thompson
At half-time in a televised football match, a few million people get up and switch on a kettle.
The effect on the national electricity system is immediate and measurable. Britain’s alternating current oscillates fifty times a second, and every appliance in the country — a kettle, an MRI scanner, a railway signal — is built to run at exactly that rate. When demand jumps, frequency starts to sag. The National Energy System Operator is legally required to hold it within one per cent of fifty hertz, and in practice aims to stay within 0.2. So within seconds of the kettles going on, in a control room most people could not name, generators are being instructed to produce more.
There is no reservoir. Electricity cannot be stored economically at scale, which means supply and demand have to be matched continuously rather than reconciled at the end of the day. Whatever the country wants at any given instant, it gets in that instant or not at all.
That fact is why this article exists, and it is worth stating in its bluntest form. Electricity is not one infrastructure among several. It is what the others are made of — and because it cannot be banked, every claim on it competes with every other claim, permanently, in real time. A data centre, a housing development, a heat pump, a steelworks and an electric car are not making separate requests. They are making the same request.
Which raises a question the other systems in this series can mostly avoid. When two legitimate claims arrive on a finite resource, who decides between them?
This article is part of How Modern Society Works, a series on the systems that organise everyday life. Each article takes something ordinary and works outwards to the system that produced it.
What Is an Electricity Grid?
Strip away the jargon and a grid is three separate systems wearing one name.
Generation is where electricity is made: gas plants, wind farms, nuclear stations, solar arrays. Transmission is the high-voltage network carrying bulk power over long distances, usually from generation sites nowhere near the people who will use it. Distribution is the lower-voltage local network that takes that bulk power and threads it street by street into homes, shops and hospitals. The Energy Networks Association’s shorthand is a good one: transmission is the motorway network, distribution the local roads that get you to your door.
Almost nobody outside the industry can say where the boundaries fall, and the confusion has a particular shape. Ask who supplies your electricity and most people name the company on the bill. That company owns no cables. It buys electricity wholesale and sells it on. The wires belong to a separate, regionally licensed business you never chose and cannot switch.
The distinction matters more than it sounds, because it marks the line between a competitive market and a monopoly, hiding inside a single word. It is also the first clue that the grid is not one object you can point at. It is a layered set of institutions with different owners, different incentives, and different accountability to the public.
The National Audit Office puts the underlying constraint plainly: because electricity cannot be stored economically in large quantities, supply and demand must be matched second by second, and a whole separate layer of institutions — transmission-level balancing services — exists purely to do that matching in real time. The grid is not a static object like a bridge. It is closer to a live performance that never stops, coordinated across the whole country, every second, by people most of us will never see.
That coordination is not free, and it is not primarily technical. A 2024 report from the Royal Academy of Engineering argued that delivering a reliable, decarbonising grid at the pace Britain now needs is not chiefly an engineering problem at all, but one of governance, leadership, risk allocation and long-term investment — national coordination of the kind usually reserved for wartime mobilisation or a vaccine rollout. The wires are the easy part. Deciding who pays for them, who is liable when they fail, and whose town gets priority for new capacity is the hard problem, and it is institutional rather than technical.
The anthropologist Brian Larkin has made the general case here: large technical systems are never neutral conduits for moving something from A to B. They are material forms that carry politics with them, deciding quietly whose homes get built where, which regions are treated as worth investing in, and which are left waiting. The three-layer structure of a grid is an engineering diagram. It is also a map of decisions about whose electricity gets prioritised, and by whom.
Where Does Britain’s Electricity Come From?
Every unit reaching your home carries an invisible history — a decision, made elsewhere, about which fuel to burn or which turbine to spin. Most of us see only a bill.
Britain’s generation mix has shifted more in the past decade than in the previous half-century. The government’s Energy Trends data puts the supply between April 2024 and March 2025 at: coal 5.9 per cent, natural gas 33.3, nuclear 16.2, renewables 42.1, with a residual 2.5 from other sources. Renewables now generate more of Britain’s electricity than any other single category — a historic reversal that took about fifteen years.
It happened with remarkably little public argument. Coal stations did not close because the public demanded it the way people demand lower taxes or better trains. They closed because carbon pricing, subsidy design, falling renewable costs and international commitments made coal progressively uneconomic — decisions taken in regulatory documents and boardrooms rather than at a ballot box.
Timothy Mitchell has argued, in Carbon Democracy, that this is not incidental: the type of fuel powering a society shapes the politics it can sustain. Coal was mined and moved by concentrated groups of workers who could and repeatedly did shut down an economy through a strike, which gave coal-era labour real leverage over democratic politics. Oil and gas flow through pipelines and tankers far harder for any single group to interrupt. It is a contested claim and not every economic historian accepts it in full, but it is a useful prompt: the composition of the fuel mix is not only a technical statistic.
The new mix has not escaped its own version of the problem. Gas still supplies a third of the country’s electricity and remains important for a specific reason — unlike wind or solar, it can be ramped up within minutes, which makes it the most reliable way of covering a still, cloudy day. Renewables generate more in aggregate than anything else, but they generate when the weather allows rather than when people need it. That mismatch is one of the defining problems of the modern grid, and it returns below.
Transmission: Moving Bulk Power
If distribution is the local roads, transmission is the motorway — except this motorway is planned decades ahead, costs billions, and increasingly has to run through places that do not want a pylon.
Britain’s renewable power is disproportionately generated somewhere remote: offshore wind in the North Sea, turbines on Scottish hillsides, solar on flat rural land far from the cities that will use it. That geographic mismatch is the single biggest reason transmission has become one of the most contested infrastructure questions in the country.
A 2025 report from the Institution of Engineering and Technology laid out the choice in unusually concrete terms. Overhead pylons remain by far the cheapest way to move large amounts of power over long distances, and also the most visible and most resented. Burying the same cables costs on average around four and a half times more per megawatt per kilometre. Running them under the sea costs roughly five times more for a simple point-to-point connection, and up to eleven times more for a fully networked offshore system linking multiple sites. There is no cost-free version of this decision. Every mile is a trade between price, visibility, disruption and delay, and someone — a regulator, a planning inspector, ultimately a billpayer — has to choose.
That someone is working against the clock. The International Energy Agency’s review of transmission investment found grid expansion worldwide now constrained as much by supply chains as by planning permission: specialist transformers, high-voltage cable and skilled engineering capacity are all short, with lead times for some critical components past four years. A wind farm can be built in a couple of years. The line to carry its power to market can take a decade — which has left swathes of new generation sitting in connection queues, able to produce electricity the grid cannot yet move.
It is tempting to read that as a purely physical bottleneck: not enough steel, not enough transformers. But Stephen Graham and Simon Marvin’s argument about networked infrastructure applies well here. Large technical networks do not expand evenly. They get built toward the places judged most valuable to connect first, and they leave others waiting — sometimes for years, sometimes indefinitely. Every queue has an order, and that order is a decision rather than a law of physics. The question worth carrying forward is not only whether Britain can build enough transmission fast enough, but whose power gets carried first.
What Does a Substation Actually Do?
Substations are the least glamorous and most consequential buildings in the system, and their job sounds trivial: change the voltage of electricity as it moves through the network.
In practice that job is doing an enormous amount of work. Electricity leaves a power station at one voltage, is boosted far higher for efficient long-distance travel, and then has to be stepped back down repeatedly before it is safe to plug a kettle into. UK Power Networks operates thousands of these facilities across London, the South East and the East of England alone, from unremarkable green boxes behind housing estates to sites the size of small industrial parks. Most carry nothing but a yellow-and-black warning sign. Hospitals and schools often have their own dedicated substation; the rest of us share.
Inside, the work is done by transformers, which Hitachi Energy describes with a useful metaphor: a volume control for electricity. Step-up transformers near power stations raise voltage as high as it can safely go, because higher voltage wastes less energy as heat over distance. Step-down transformers reverse that at the far end, repeatedly, until what reaches your street is low enough to be safe.
Substations are not only switches. They are the interconnecting nodes that let the network reroute power when one path fails — a built-in redundancy that matters enormously when something goes wrong. In March 2025 a fire at the North Hyde substation in west London cut power to Heathrow and forced the temporary closure of one of the world’s busiest airports. Flights were grounded worldwide and tens of thousands of passengers stranded. What is striking about the government’s response to the subsequent review is how ordinary the underlying failure was: not sabotage, not an unprecedented technical frontier, just one piece of equipment in one building that global aviation turned out to be depending on. The surrounding network’s automatic protection stopped the damage spreading much beyond the airport, which is the system working as designed even as it produced a spectacular outcome.
By the time electricity reaches a socket it has been reduced to a standard written into regulation: a nominal 230 volts, held within a tolerance of roughly -6 to +10 per cent, monitored continuously by engineers working around the clock.
Distribution: Who Owns the Wires Under Your Street?
Once bulk power arrives from the transmission network it passes to one of fourteen regionally licensed Distribution Network Operators, each with a fixed territory set by Ofgem. You cannot choose yours any more than you can choose your local water pipes.
If you live in London, the South East or the East of England, your electricity runs through cables owned by UK Power Networks, which serves around eight million supply points. Its ownership is not something a customer would find anywhere they might look. From November 2010 the company was owned by a Hong Kong consortium — CK Infrastructure Holdings and Hongkong Electric with forty per cent each, and the Li Ka Shing Foundation with the remaining twenty — which bought the business from the French utility EDF for £5.775 billion. In 2026 it was reported to be changing hands again, in a sale to the French group Engie valued at around £10.5 billion.
Roughly a fifth of the country’s electricity distribution was owned from Hong Kong for a decade and a half and is now reportedly being sold to Paris, and at no point in that sequence was any customer consulted, because there is no mechanism by which they could be. That is not a scandal. It is the ordinary operation of a regulated monopoly, a form of ownership that deliberately separates the question of who profits from the question of who is served. Suppliers compete for your custom and can go bust, merge or rebrand overnight, as many did during the 2021–22 energy crisis. The network underneath them keeps running the wires and collecting a regulated fee, insulated from that competition entirely.
For most of the grid’s history the local network had one job: carry power one way, from substation to socket. That is changing fast. Rooftop solar, home batteries and vehicle chargers sit at the very edge of the distribution network and increasingly push power back into it rather than only drawing it out. A 2018 technical paper from the IET found that roughly ninety per cent of Britain’s solar generation connects directly to the distribution network rather than the transmission system it was originally built around — meaning a system engineered for one-directional flow now manages millions of small, unpredictable, two-way ones.
UK Power Networks responded by formally reinventing part of what it does. In April 2023 it became the first operator in the country to launch an independent Distribution System Operator function: a shift from distributing electricity to actively managing it, coordinating distributed energy resources it says amount to enough capacity to power more than six million homes. The local monopoly that used to deliver power is increasingly also the institution deciding, second by second, whose rooftop solar gets to export first.
Which is where Graham and Marvin’s splintering reappears at a scale small enough to be uncomfortable. Networks splinter between cities and hinterlands, and they splinter between neighbours. Two houses on the same street, both with panels, can face different export limits depending on how much headroom is left in the cable they happen to share — a decision made by an algorithm inside a company neither household chose, answerable to a regulator neither has heard of.
How Is Supply Matched to Demand?
Two threads meet here. Electricity cannot be stored at scale, so supply and demand must be matched continuously. And renewables now produce more of Britain’s power than anything else, but they produce it when the weather allows.
Put those together and you get the defining operational problem of the modern grid, and the reason the country needs a control room staffed around the clock doing a job with no real equivalent elsewhere in the economy: matching supply to demand live, for sixty-eight million people, without being allowed to get it wrong for long.
The mechanism is frequency, as described at the top of this article. When demand rises, frequency dips and generators are instructed to ramp up within seconds. When demand falls, the reverse.
That job has become substantially harder over the past decade, for reasons of physics rather than competence. A widely cited 2018 study by Iain Staffell and Stefan Pfenninger modelled how weather affects both sides of the equation at once and found the British system moving into territory it has never operated in. Peak demand is projected past seventy gigawatts as heating and transport electrify, while intermittent renewable output can already exceed total demand on the windiest, sunniest days — a scenario that would have sounded like science fiction to the engineers who designed the original coal-and-gas grid. The same study found hour-to-hour swings in the supply-demand balance widening by roughly fifty per cent, and year-to-year variability by around eighty, as the system leans harder on weather it cannot control.
None of this is abstract in the control room. It shows up as cost, decided in real time. Every additional piece of flexibility the system needs — reserve generation on standby, fast-response batteries, extra headroom in engineering tolerances — has to be paid for, and those costs land on billpayers. A regulatory decision taken in December 2025 reduced the minimum inertia the system is required to hold in reserve, a change Ofgem estimated saved consumers over £200 million in a single year. Balancing the grid is not only an engineering exercise. It is a continuous negotiation about how much risk the system carries, and who pays when that risk is priced.
The result is a system operating closer to its margins than at almost any point in its history, doing so precisely because it is decarbonising successfully — and successful decarbonisation is what makes second-by-second balancing harder, not easier.
What Do Interconnectors Actually Buy?
Some of the electricity in a British socket has, in the last few hours, been Norwegian or Danish. Not metaphorically. Britain is an island whose electricity system stopped being one decades ago.
The newest and longest of these links is Viking Link, a subsea and underground high-voltage cable running 475 miles between Lincolnshire and southern Denmark. National Grid described it on entering full commercial operation as the longest land-and-subsea interconnector on Earth: a £1.7 billion joint venture with the Danish system operator Energinet, capable of carrying enough electricity for up to 2.5 million UK homes and projected to save consumers over £500 million across its first decade, largely because Danish wind is often cheaper than running a British gas plant to cover the same demand. The North Sea Link does the same job with Norway — 450 miles of subsea cable connecting Northumberland to a village near Stavanger, built to let Britain trade with a country whose hydropower reservoirs function as enormous natural batteries.
The logic is straightforward: it is rarely windy, sunny, still and cold everywhere in northern Europe at once, so a cable that lets Britain import when its own renewables are becalmed and export when they are not makes the whole regional system cheaper and more secure than any one country balancing alone. National Grid operates six such interconnectors, linking Britain to France, Belgium, the Netherlands, Norway and Denmark, and expects that by 2030 ninety per cent of the electricity flowing in through them will come from zero-carbon sources.
What that tidy logic conceals is how contingent the arrangement is. After the Brexit referendum, researchers at Imperial College London’s Grantham Institute warned that continued UK access to the EU’s Internal Electricity Market — the framework that lets interconnector capacity be used efficiently rather than sitting idle — was not something Britain could assume it would retain, and estimated that losing barrier-free access risked adding hundreds of millions a year to energy system costs on top of the capital already sunk into the cables.
The physical infrastructure turned out to be the easy part. What determines whether a cable delivers cheap Norwegian hydropower or sits underused is a set of treaties, regulatory agreements and diplomatic relationships that no customer will ever read. Geography decides where the wires can go. Politics decides how much of them counts.
What Actually Depends on the Grid?
The obvious answer is the lights, and it is the least interesting one.
Water arrives at a tap because electric pumps pressurise the network, not because Victorian gravity is still doing the work. Hospitals hold backup generators sized to bridge short gaps, not to sustain full operations indefinitely. Mobile masts carry battery reserves measured in hours. Payment terminals, traffic signals, railway signalling and the refrigeration in every supermarket and distribution warehouse all treat continuous electricity as a background condition so basic it barely registers as an assumption.
The government treats this seriously. The Cabinet Office’s National Risk Register includes dedicated scenarios for national and regional electricity failure, explicitly modelling how an outage would cascade into water supply, telecommunications, transport, healthcare and food distribution rather than staying contained as a single-sector problem. Analysis of the register noted that electricity failure now sits among the risks government rates as combining relatively high likelihood with genuinely catastrophic potential impact.
What is worth dwelling on is how thoroughly this dependency was built without anyone choosing it. Nobody voted for hospital operations to depend on continuous mains power, or for supermarket supply chains to assume electrified refrigeration at every link. Those dependencies accumulated gradually, technology by technology, until modern life became something that cannot function through an extended blackout — not because anyone designed it that way, but because each individual choice to electrify or automate made sense on its own terms, and nobody was responsible for totalling up the cumulative fragility.
There is at least one natural experiment showing how fast the pattern can shift. During Britain’s COVID-19 lockdowns, researchers at Imperial College London’s Energy Futures Lab found household electricity consumption rose by more than ten per cent as people spent their days at home — but that rise was more than offset by a sharp fall in national demand as commercial and industrial activity collapsed. The result was a grid running on more renewable power than usual and a measurable fall in household carbon emissions, purely as a side effect of where people spent their time.
And the dependency is now growing again, faster, in a form that makes the competition explicit. Electrifying heat and transport moves two enormous energy demands onto the grid that were previously met by burning things at the point of use. Data centres are being built at a rate that has already exhausted network capacity in parts of west London, where developers were told connections for new housing might not be available until 2030 and in some cases 2037. Every one of those is a policy the government supports. They will compete for the same wires.
Who Owns the System?
Before 1990 Britain’s electricity industry was state-owned outright. The House of Commons Library’s account of what replaced it is direct: since the late 1980s, gas and electricity supply has been privately owned and run, regulated rather than operated by government, through Ofgem.
Generation and supply became genuinely competitive markets, with companies able to enter and exit — which is part of why so many smaller suppliers collapsed during the 2021–22 energy crisis, a level of churn that would be structurally impossible under a single nationalised utility. But the wires were treated differently, because a monopoly was never going anywhere: nobody was ever going to build a second competing set of power lines down the same street. Instead of competition, those networks got regulation — private ownership with Ofgem controlling what they are allowed to earn.
That arrangement has a documented track record and it is not spotless. A National Audit Office review found that around twenty per cent of a typical household’s annual electricity bill — roughly £130 a year — goes directly toward running, maintaining and upgrading these privately owned networks, and that Ofgem’s price controls, in one significant period, allowed network companies to earn real shareholder returns of around nine per cent against a typical UK company average of five to six. The NAO calculated that better use of available evidence on financing costs alone could have saved consumers at least £800 million. Ofgem accepted the substance of the finding and has tightened subsequent price-control rounds.
That is what a regulated monopoly means in practice: not the absence of profit-seeking, but profit-seeking channelled through a negotiation between a private company and a public regulator, with the household billpayer footing whatever the negotiation lands on.
The one part of the system that has moved back toward public ownership is, tellingly, the part that holds everything together in real time. In October 2024 the operator responsible for balancing the grid second by second — previously a subsidiary of the private company National Grid — was reconstituted as the National Energy System Operator, a fully government-owned public corporation, under powers created by the Energy Act 2023. Generation competitive, supply competitive, networks privately owned but regulated, and the institution coordinating all of it brought back under direct state ownership: as if, having privatised almost everything else, Britain concluded that the conductor of the orchestra needed to answer to the public.
Who Pays for the Wires?
Everyone. Not everyone in the same proportion.
The network charge described above — roughly a fifth of a typical bill, about £130 a year — is levied per customer. It does not vary with income, and it does not vary with whether the building can be heated.
Both of those things vary enormously. The government’s most recent figures put an estimated 9.4 per cent of English households — 2.36 million homes — in fuel poverty in 2025, meaning low-income households living in properties too inefficient to heat affordably. That is an improvement on the previous year. A second, broader measure in the same report tells a starker story: 30.4 per cent of English households, over 7.6 million homes, spending more than a tenth of their income after housing costs simply to meet basic energy needs. Nearly a third of the country. And the rate is sharply regional — around eleven per cent of households in England, but thirty-four per cent in Scotland and twenty-four per cent in Northern Ireland, driven by colder climates, older and less efficient housing stock, and heavier reliance on expensive off-grid heating fuels.
The geographer Stefan Bouzarovski, who has spent much of his career studying energy vulnerability across Europe, argues that fuel poverty is not simply a matter of income. It is produced by the intersection of low income, inefficient housing and high energy costs, layered onto a household’s particular circumstances: disability, age, whether they rent or own, whether they can afford to move or renovate. Two households on identical incomes a few streets apart can face entirely different exposure depending on the age and condition of their specific building.
A flat per-customer charge has no way to register any of that. And the routes out are gated by exactly the characteristics the charge ignores. Every subsidy for insulation, every incentive for a heat pump, every network upgrade to accommodate vehicle charging is easier to access if you own your home, have savings for the upfront cost, and live somewhere with existing infrastructure to build on. A tenant in a poorly insulated property cannot install a heat pump, and the landlord who could has limited reason to.
That is the transfer, stated plainly. The cost of the wires is recovered identically from a household in a well-built flat and from one in a property that cannot be heated affordably; the margin on that recovery, for a period, ran several points above the ordinary corporate return; and the mechanisms for escaping the underlying cost were available in practice to owners with capital. None of it was a deliberate design choice by any single institution named in this article — not Ofgem, not the network companies, not government. It is the cumulative effect of a system built around efficiency and average cost, which was never designed to ask who specifically ends up paying most for the electricity everyone equally depends on.
Two Blackouts, Two Systems
A working grid tells you little about how it is governed. A failing one tells you a great deal.
Britain, 9 August 2019. At 4:52pm, lightning struck a section of high-voltage transmission line in Cambridgeshire. On its own that should not have mattered much; lightning strikes transmission infrastructure fairly often and the system is built to absorb it. But within two minutes a large offshore gas-fired power station and a big offshore wind farm both dropped off the system, for unrelated technical reasons, at nearly the same moment. Ofgem’s investigation found that the near-simultaneous loss of that much generation caused frequency to fall sharply — and when frequency falls too far the risk is not inconvenience but total system collapse, with transformers and generators tripping out across the country in a cascade that can take days to recover from.
To prevent that, the system did what it is designed to do: it sacrificed part of itself to save the whole. Automatic Low Frequency Demand Disconnection cut power to roughly five per cent of Britain’s electricity demand within seconds, disconnecting just over a million customers deliberately and automatically, with no human in the loop making that call in real time. Trains stopped. Hospitals switched to backup. Within about forty-five minutes the vast majority of customers were reconnected. It was a controlled failure: painful for those affected, engineered specifically to stop a local problem becoming a national catastrophe.
Texas, February 2021. The physics was comparable — extreme conditions overwhelming a system faster than it could adapt — and the institutional response was not. When a severe winter storm hit, power plants across the state, most of which had never been required or financially incentivised to winterise, began failing simultaneously: frozen sensors, frozen fuel lines, frozen turbines. The federal regulator’s investigation found the resulting shortfall forced the state grid operator, ERCOT, into the largest manually controlled load-shedding event in US history — not a brief automated few minutes, but rolling blackouts leaving over 4.5 million customers without power for periods stretching into days, in freezing conditions.
The death toll is itself disputed and worth stating carefully rather than rounding. The Texas Department of State Health Services’ final official report attributed 246 deaths to the storm, most from hypothermia among people aged sixty and over. Independent researchers have argued that is a significant undercount: an excess-mortality analysis reported by BuzzFeed News estimated the true toll between 426 and 978, with a best estimate around 702, the gap explained by indirect deaths — delayed medical care, exacerbated chronic illness — that never reached a storm-related death certificate. Both figures are worth holding: the state’s conservative count, and the independent estimate roughly three times larger.
A timeline compiled by the University of Texas at Austin traces the deeper cause to a decision made decades earlier. Texas deliberately built its grid to be electrically isolated from the rest of the United States, specifically to avoid falling under federal regulatory oversight. That isolation meant ERCOT could not import large amounts of emergency power from neighbouring states — the same interconnector logic that lets Britain lean on Norwegian hydropower on a still day simply was not available, by design rather than by accident.
Put the two events side by side and the difference in outcome tracks the difference in governance almost exactly. Britain’s system is a regulated monopoly, legally required to plan for extreme events and answerable to a national regulator; when it failed, an automated mechanism sacrificed a defined slice of demand for minutes and a public investigation followed within months. Texas’s grid was built to escape that oversight and, not coincidentally, to escape the cost of winterisation that regulation would likely have required; when it failed, it failed for days and killed people.
Texas also exposed a narrower group to a different kind of harm. Customers of Griddy, a retailer passing real-time wholesale prices directly to around 29,000 subscribers with no cap, were billed a combined $29 million during the storm as prices spiked to $9,000 per megawatt-hour, leaving individual households with bills reported as high as $17,000 for a few days’ usage. Most Texans, on fixed-rate contracts, were shielded. The minority who were not found out in the most direct way possible what it means to bear a system’s pricing risk with no buffer at all. That contrast — between those who had bought protection from a system’s failure and those who had not — is the subject of a later article in this series, and it recurs well beyond electricity.
How Would Britain Ration Power?
Suppose the worst happens: not a substation fire or a frequency dip, but total collapse of the transmission system. It sounds like the premise of a disaster film. It is also something the government has a legally binding plan to recover from, on a clock measured in days.
The National Emergency Plan for Downstream Gas and Electricity sets out what is formally called the Electricity System Restoration Standard. From 31 December 2026, NESO will be legally required to have the capability to restore one hundred per cent of Great Britain’s electricity demand within five days of a total blackout — though the same document is candid that on current capability a full restoration could realistically take up to seven. Getting there relies on Black Start: a small number of specially contracted generators, able to start with no external power at all, gradually energising islands of the network and connecting them together until the system is rebuilt from nothing.
If demand has to be rationed before full restoration, there is a formal mechanism, updated as recently as April 2026. The Electricity Supply Emergency Code divides ordinary customers into eighteen roughly equal groups, or load blocks, and rotates power cuts between them in three-hour windows across each day. Hospitals, key national infrastructure and other designated protected sites are shielded; everyone else takes their turn.
It is worth pausing on how unusual that document is in the context of everything above. Almost every allocation described in this article happens through a mechanism that refuses to make a judgement — a queue ordered by filing date, a charge levied per customer, an export limit set by cable headroom. The emergency code is the exception. It is the one place where somebody sat down in advance, decided explicitly who would be protected and who would take their turn, wrote it down, and published it. Rationing is the only part of the system where the allocation is admitted to be an allocation.
Set the five-to-seven-day figure against the dependencies described earlier and it stops reading like a technical specification. Water and wastewater are among the most electricity-hungry sectors in the country. The restoration standard is not really a promise about lighting. It is a judgment, written into government policy, about how long sanitation and public health can be expected to hold while the network is rebuilt from nothing.
How Was Capacity Allocated, and What Changed?
Twice already this article has run into the same unanswered question — first between regions, then between neighbours — and left it hanging, because until recently Britain had no principled answer. It had a queue instead.
For most of the past decade a project wanting to connect to the transmission network joined that queue, and its position was determined by the date its paperwork was filed. Not by whether it was funded. Not by whether it had planning permission. Not by whether the system needed it. By arrival order.
At its peak the queue held roughly 800 gigawatts of proposed capacity — around four times what the country needs to meet its 2030 clean power target — and the House of Lords Industry and Regulators Committee found projects that had waited up to a decade. Ready schemes sat behind speculative ones that might never be built, because a speculative application filed in 2019 outranks a shovel-ready one filed in 2022, and the rule contains nothing capable of distinguishing between them.
It is worth being precise about why a rule like that exists, because the obvious reading is that it was lazy and the obvious reading is wrong. First-come-first-served is what you adopt when you have decided that the body operating the queue must not pick winners. A regulated monopoly free to choose which customers to favour would be a private company making national industrial policy with no mandate and no appeal. Arrival order is unfair in the ordinary sense and scrupulously fair in the only sense it was designed for: it cannot be lobbied, and it cannot prefer.
What it also cannot do is decide. And that changed in December 2025.
The system operator assessed around three thousand applications and cut the queue to a prioritised pipeline of 283 gigawatts, discarding over 300 gigawatts of projects judged not ready or not aligned with national targets. The principle it now runs on is stated as first ready, first needed, first connected — a direct rejection of arrival order, replacing filing date with a judgement about which projects will deliver the electricity the country needs, and where. NESO estimates the reform could unlock up to £40 billion a year in investment that had been stuck behind a paperwork bottleneck. Ofgem frames the wider programme as central to reaching a clean power system, and frames it too as a genuinely difficult balance between speed, fairness to existing applicants, and the physical limits of how fast transmission can be built.
ADD IMAGE
The significance is not the efficiency gain. It is that somebody now decides, on stated criteria, and can be argued with about them.
Nothing about that is a solution, and the sequel proves it. Within months the reform produced a consequence nobody forecast: batteries, which clear planning faster than other technologies, were disproportionately favoured by the new criteria, leaving prioritised battery capacity well above the government’s 2030 range and a projected surplus beyond what the system needs even by 2035. Ministers and Ofgem went back to reconsider protections they had just finished designing. Clearing a paperwork backlog also builds no pylons; the construction still takes years, still costs billions, and still runs into the same trade-offs between overhead lines, buried cables and local opposition.
What the Grid Decides
Go back to the kettle at half-time. Nothing about switching it on required you to know where the electricity came from, who owned the wires it travelled through, or what would happen if it stopped. That is the system working as intended, and it is a genuine achievement — a national machine run to tolerances of a fifth of a hertz, reliable enough that the country has organised its water, health, food and work around the assumption it will not stop.
The achievement has a shadow, and it is not that the system is secret. Connection rules are published. Price controls are consulted on. Fuel poverty statistics come out annually and the emergency rationing code is a public document. Anyone could read all of it. The shadow is that almost none of it presents itself as a decision anyone could argue with. A filing date is not a policy. A standing charge is not a judgement about who should bear what. An export limit is not a view about whose panels matter. Each is a rule that produces an outcome while declining to take responsibility for it, and the outcomes are substantial: which regions build first, which households spend a tenth of their income staying warm, which company earns nine per cent on the wires.
That is what makes December 2025 more interesting than a procedural reform. For one part of the system, the refusal was withdrawn. A rule that could not choose was replaced by a body that does choose, publicly, on stated criteria, and can be held to them — and immediately got something wrong, and was argued with, and went back to reconsider. That is not a tidier outcome. It is a worse-looking process attached to a better kind of accountability, which is usually what accountability looks like from the inside.
Every other allocation in this article still works the old way. The flat network charge is not a decision anyone takes; it is a recovery method. The export limit on a shared cable is set by capacity, not by a view. There is no forum in which a household in a cold house can put its claim against the ordinary corporate return of a network company, because no process treats the two as competing. Rationing aside, the only part of this system where somebody sat down and decided who should be protected is the plan for what to do when it has already collapsed.
More in this series
Previously: How GPS Works: The Clock We All Share — why every fix your phone calculates is a timing measurement, and what happens to the services that cannot afford their own clock.
Next: Undersea Cables — the infrastructure that lies outside any state’s authority, and what that means when one is cut. (Placeholder — this article is drafted but not yet published. Until it is, point “Next” at the hub rather than leaving a dead link.)
All nine articles on hidden infrastructure: How Modern Society Works.
References
References
Government, regulators and parliamentary committees
National Energy System Operator, What is frequency? — the fifty-hertz standard, the one per cent legal tolerance and the tighter operational target.
National Audit Office, Electricity Balancing Services — why electricity cannot be stored economically at scale, and the institutional layer that exists to match supply and demand in real time.
National Audit Office, Electricity Networks — the share of a household bill that funds the networks, the returns allowed under Ofgem’s price controls, and the sum the NAO calculated could have been saved.
Ofgem, Decision notice establishing the National Energy System Operator — the October 2024 move of the system operator into public ownership under the Energy Act 2023.
Ofgem, Frequency Risk and Control Report 2025: Authority Decision — the December 2025 reduction in required system inertia and the estimated saving to consumers.
Ofgem, Investigation into the 9 August 2019 power outage, and the accompanying Energy Emergencies Executive Committee technical report — the sequence of generation losses, the frequency fall, and the automatic disconnection of around five per cent of demand.
Department for Energy Security and Net Zero, Energy Trends and the Fuel Mix Disclosure Data Table 2025 — the generation mix for April 2024 to March 2025.
Department for Energy Security and Net Zero, Annual Fuel Poverty Statistics in England, 2026 [URL needed] — the fuel poverty rate and the broader measure of households spending more than a tenth of income on energy.
Cabinet Office, National Risk Register 2025 [URL needed] — the scenarios for national and regional electricity failure and their modelled cascade into water, telecommunications, transport, healthcare and food.
Department for Energy Security and Net Zero, Government response to NESO’s North Hyde report — the March 2025 substation fire and the closure of Heathrow.
National Emergency Plan for Downstream Gas and Electricity, July 2023 — the Electricity System Restoration Standard, the five-day requirement from 31 December 2026, and the candid assessment that current capability may take seven.
Electricity Supply Emergency Code guidance, revised April 2026 — the eighteen load blocks, the three-hour rotation, and the designation of protected sites.
House of Commons Library, Public ownership of industries and services, CBP-8325 — the privatisation settlement and what was left under regulation rather than competition.
House of Commons Library, Fuel poverty, CBP-8730 — the regional breakdown across England, Scotland and Northern Ireland.
House of Lords Industry and Regulators Committee, report on grid infrastructure delivery — the scale of the connections queue and the length of the waits.
National Energy System Operator, connections reform announcement, December 2025 — the assessment of applications, the prioritised pipeline, and the shift to first ready, first needed, first connected.
Federal Energy Regulatory Commission, final report on the February 2021 Texas freeze — the largest manually controlled load-shedding event in US history.
Texas Department of State Health Services, final report on winter storm deaths [URL needed] — the official toll of 246.
Peer-reviewed and academic
Iain Staffell and Stefan Pfenninger, ‘The increasing impact of weather on electricity supply and demand’, Energy 145 (2018), 65–78 — projected peak demand and the widening of hour-to-hour and year-to-year variability.
IET/CIRED technical paper on distributed energy resource integration, 2018 — the share of British solar generation connecting to distribution rather than transmission.
Brian Larkin, ‘The Politics and Poetics of Infrastructure’, Annual Review of Anthropology 42 (2013), 327–343.
Stephen Graham and Simon Marvin, Splintering Urbanism (Routledge, 2001).
Stefan Bouzarovski, ‘Energy poverty in the European Union: landscapes of vulnerability’, WIREs Energy and Environment 3:3 (2014), 276–289.
Timothy Mitchell, Carbon Democracy: Political Power in the Age of Oil (Verso, 2011).
Engineering and research institutions
Royal Academy of Engineering and the National Engineering Policy Centre, report on delivering clean power by 2030, 2024 — the argument that the constraint is governance rather than engineering.
Institution of Engineering and Technology, electricity transmission technologies report, 2025 — the cost multiples for undergrounding and subsea routing against overhead lines.
International Energy Agency, Building the Future Transmission Grid — global supply-chain constraints and component lead times.
Grantham Institute, Imperial College London, interconnectors, the EU Internal Electricity Market and Brexit, 2017 — the estimated cost of losing barrier-free market access. Predates the eventual UK–EU trade agreement and is cited as contemporary analysis rather than current fact.
Energy Futures Lab, Imperial College London, briefing on lockdown effects on electricity demand.
University of Texas at Austin, technical timeline of the February 2021 blackout [URL needed] — the origins of ERCOT’s isolation from federal oversight.
Industry and company sources
The following are published by companies and industry bodies with an interest in the subject, and are used here for technical description and for their own figures.
Energy Networks Association, Energy networks explained and the network operator lookup.
UK Power Networks, what is an electrical substation, voltage in your home or business, and the Distribution System Operator function.
Hitachi Energy, High-Voltage Basics.
National Grid, announcements on Viking Link and the North Sea Link.
CK Infrastructure Holdings and Hongkong Electric, completion of the acquisition of EDF Energy’s UK network business, 1 November 2010, and Électricité de France, half-year results, 30 July 2010 — the ownership of UK Power Networks and the terms of its sale.
International Energy Agency, Electricity Grids and Secure Energy Transitions.
Reporting
S&P Global Market Intelligence, on the UK battery storage surplus and the review of queue protections, April 2026.