Talk to almost anyone about the future of British energy and you will, sooner or later, be shown the same small table. Offshore wind, it says, can now be built for £44 per megawatt-hour. New gas, by contrast, is said to cost £85–£110. The conclusion appears obvious: wind has won, gas is obsolete, and the market has spoken. The table is drawn from the Department for Business, Energy and Industrial Strategy’s annual “Electricity Generation Costs” report — a sober looking document that carries the authority of Whitehall. Yet the conclusion is false, because the table is not describing a market price at all. It is describing something called the Levelized Cost of Electricity, and LCOE is a mirage.
The idea behind LCOE is perfectly reasonable. Add up every pound you will ever spend on a power station — concrete, turbines, fuel, insurance, interest, eventual demolition — discount the future flows, and divide by the number of units the station is expected to produce. Out pops a single figure that seems to allow fair comparison between a nuclear plant that lasts sixty years and an offshore wind farm that lasts twenty five. The trouble is that the calculation rests on three assumptions that hold only for the lost world of the dispatchable baseload plant. When those assumptions are applied to weather dependent renewables, the result is not insight but propaganda.
First, LCOE assumes that every megawatt hour is created equal. In the world for which it was designed, the world of coal, gas, nuclear, that was broadly true. A CCGT running at 50 per cent load factor can choose to run harder on a winter evening when demand is high and it can choose not to run at all at 3 a.m. when demand is low. The electricity it produces is therefore worth roughly what the market will pay at the moment of delivery.
An offshore wind farm enjoys no such freedom. On a still December evening it is precisely when its output is zero that the system needs it most; on a gusty May night it may be producing at full tilt when the market is already oversupplied. The Future Energy Scenarios of the National Grid show that even after 100 GW of wind capacity is built the residual peak demand met by dispatchable plants barely falls. Those plants must still exist, must still be financed, and must still recover their fixed costs across ever fewer running hours. The true marginal value of wind power therefore falls as penetration rises, but LCOE, frozen in its spreadsheet cell, never notices.
Second, LCOE assumes the plant is built close to demand. In the original studies the power station sat beside the steelworks or the city it served. Britain’s best wind and solar resources are nowhere near its load centres. The Dogger Bank offshore zones lie more than a hundred miles from Yorkshire, the sunniest fields of East Anglia are separated from London by congested 400 kV corridors. Famously, much of our wind is in Scotland — the windiest part of the UK — behind the infamous B6 corridor. Being paid to turn off every time the hosepipe of wind tries to squeeze through the pinhole of grid capacity. National Grid’s “Pathway to 2030” transmission plan identifies £54 billion of reinforcements, new subsea cables, onshore pylons, reactive compensation, needed to accommodate 50 GW of offshore wind. Spread across the additional wind output this is roughly £20 per MWh, yet it appears nowhere in the £44 headline figure.
Third, LCOE assumes the electricity can be sold at the calculated price. Half-hourly settlement data show that British wholesale prices already fall close to zero whenever wind output exceeds 15 GW. Curtailment payments to wind farms — money paid for electricity that is never generated — have risen from £53 million in 2016 to £1.5 billion in 2024. The cost of balancing actions — paying gas plants to start quickly, paying batteries to soak up excess — adds another £2–£3 billion per year and is rising. None of these costs is debited to the wind farm that caused them. They add approximately another £32 per MWh.
We should not be making policy on the basis of a single, flawed metric
Once these hidden costs are tallied, the picture changes. One of the few attempts to calculate a “system LCOE” was undertaken by Robert Idel in 2023 and found that system costs increase exponentially the more penetration there is of variable renewables on a grid. This aligns with an OECD study that shows a similar trend of system costs per MWh increasing the more wind and solar comes online.
When Ed Miliband steps to the despatch box, the figure in his mind is £51 per megawatt hour — the price, he insists, that offshore wind will reach in 2025. That number did not emerge from open bidding or complicated engineering. It was lifted, verbatim, from the Climate Change Committee’s latest pathway, a document produced by advisers who have spent a decade polishing the same LCOE spreadsheet. Four months before the CCC published its pathway, the Government’s own Contracts for Difference auction had to pay £85 per megawatt hour for projects that will not be commissioned until 2028. This week the Secretary of State raised the strike price cap for the next auction by 30 per cent. The message is plain: new British offshore wind is already priced well above the CCC’s fantasy, and the gap — once transmission, storage and balancing are counted — will be wider still.
We should not be making policy on the basis of a single, flawed metric while ignoring reality. Labour has committed to decarbonise its power system by 2030. Achieving that goal will require not just gigawatts of wind and solar but also new transmission, firm low carbon capacity, storage, demand flexibility and market reforms. Pretending that the headline LCOE figure already contains these elements is like pricing a car by quoting only the cost of the engine and then claiming surprise when the wheels, brakes and steering wheel have to be bought separately. But worse, we’re buying the engine and then lying to ourselves that it cost half the price we actually paid.
The next time someone tells you that offshore wind is cheaper than gas, ask them: cheaper for whom, under which grid, and at what hour? Until the answer accounts for the whole system, the headline remains what it has always been: a convenient fiction.
