The electricity mix, capacity versus generation

The first distinction that matters. Installed capacity is what a fleet of power plants could theoretically produce if running flat out. Generation is what it actually produces. These diverge a lot, and conflating them is the source of a surprising share of bad takes in energy journalism.

A one gigawatt gas plant running hard can generate close to eight terawatt-hours a year. A one gigawatt offshore wind farm, running with a capacity factor of fifty percent, generates closer to four and a half. A one gigawatt solar array in the UK, at around eleven percent capacity factor, generates less than one. Same capacity, very different output.

The UK electricity mix, measured by generation over a typical recent year, is roughly a third from gas, a third from wind (onshore and offshore combined), around fifteen percent from nuclear, with the remainder split between biomass, solar, hydro, and net imports through interconnectors. The exact shares move year to year, and monthly shares move dramatically.

The capacity rebuild, by country, 2000 to 2025

Capacity is what was built. The previous section made the point that capacity is not generation, and that an installed gigawatt of nuclear, gas, solar, and offshore wind all do quite different work on the grid. Holding that distinction in mind, the next set of charts track what countries have actually built, in installed megawatts of generating capacity, across twenty-five years.

The data underneath is IRENA's Electricity Capacity Statistics, covering thirty-seven countries, nine technologies, on-grid plant only. It is the cleanest cross-country dataset on installed capacity over a long horizon. It does not cover every country, the smallest and least transparent markets are missing, and fossil fuel data is incomplete for several European countries (Germany, Denmark, France, Poland) where IRENA's coverage focuses on the renewable build. The charts below note this where it matters.

Three transitions, side by side

Three countries with three quite different transition stories. The UK has the cleanest coal-exit in the dataset. Germany's Energiewende replaced nuclear with solar at remarkable scale, while leaving fossil fuel coverage incomplete in IRENA's record. China built everything simultaneously, including more than a terawatt of new coal alongside a faster renewable build than anywhere else.

The energy mix, country by country

The stacked area is the most revealing single format for the energy transition. Bands at the bottom of each chart are fossil fuels, the dark colours. Bands above are renewables. The thickness of each band over time is the installed megawatts of that technology. For each country, pick a tab below.

Offshore wind in the renewable picture

Offshore wind is a small share of total renewable capacity in almost every market, including the markets where it is most prominent. That sounds counter-intuitive given how much attention offshore wind attracts in policy and investment circles. The explanation is partly the capacity factor advantage from the previous section, partly the fact that solar PV has scaled enormously across most of these countries, and partly the geography, only coastal markets with shallow shelves and strong wind resource have built offshore at any scale.

Four country stories worth studying closely

Four small multiples, the same stacked area treatment as the country-mix chart above, picked because each one shows a fundamentally different transition shape. Read them as four answers to one question, what does a grid look like when you rebuild it.

Key takeaways from the dataset

Capacity factors, the number that changes the argument

Offshore wind's capacity factor is the single statistic that most changes the transition conversation. At fifty percent, offshore wind produces around half of what a perfectly steady generator of the same size would produce. That is a meaningful output profile. It is why one gigawatt of offshore wind is genuinely worth more than one gigawatt of solar on the UK grid, even though both are called renewable.

Floating wind, when it matures commercially, can target higher capacity factors still, because it unlocks deeper waters further offshore where winds are stronger and more consistent. Whether it hits that potential at acceptable cost is the open question.

Turbine scaling, the driver of the cost curve

The economics of offshore wind have been rewritten repeatedly by turbine growth. Round 1 UK projects in the early 2000s used turbines rated around three megawatts. By Round 3, eight megawatts was standard. The latest generation being deployed is in the fourteen to fifteen megawatt range, with rotors exceeding two hundred and thirty metres in diameter. Announcements of twenty plus megawatt machines are already in the industrial pipeline.

Bigger turbines change the unit economics because cost does not scale linearly with power output. A single fifteen megawatt turbine replaces about five three megawatt turbines, which means one foundation instead of five, one installation operation instead of five, one set of cables instead of five, and one maintenance visit footprint instead of five. That is where much of the learning curve has come from. Whether it continues at the same pace is not certain, there are physical limits, supply chain bottlenecks on new sizes, and tower and foundation engineering challenges that grow nonlinearly with rotor diameter.

LCOE trajectory, the curve that wobbled

Levelised cost of energy for UK offshore wind fell dramatically through the 2010s. Contract for Difference strike prices went from around a hundred and fifty pounds per megawatt-hour in the early allocation rounds to the low forties by Allocation Round 3 in 2019. That decline was real, and it was unusually fast for any energy technology.

It also stalled. Allocation Round 5 in 2023 received no offshore wind bids, because the reserve price was set below what developers said they needed given increased commodity, cable, and vessel costs. AR6 reset pricing upward. The post-2022 cost environment has not been kind to projects that priced their FID on 2019 assumptions. Some high-profile projects have been restructured, delayed, or partially cancelled. The long-run trend is still downward. The short-run cost curve is bumpy, and treating it as monotonic has caused serious analyst errors.

Fixed versus floating

Almost all offshore wind deployed to date is fixed bottom, meaning the turbine sits on a foundation driven or suction-installed into the seabed. Monopiles dominate in shallower waters. Jackets take over beyond roughly sixty metres of depth. Past that, floating structures become the only practical option, with the turbine sitting on a moored buoyant platform, typically a semi-submersible, spar, or tension-leg design.

Floating unlocks vast resource that fixed bottom cannot reach, including most of the Celtic Sea, much of the Mediterranean, most of Japanese and Korean waters, and the US west coast. Commercially it is still early. Installed floating capacity worldwide is measured in tens of megawatts to low hundreds of megawatts, versus tens of gigawatts fixed. Cost per megawatt-hour for floating is currently well above fixed bottom, and the supply chain, particularly dry docks capable of assembling floating platforms and heavy lift vessels for tow-out, is a near term constraint.

The grid is the bottleneck

The most common misunderstanding in public discussion of renewables is that generation is the constraint. It has not been, for several years. The constraint is the grid. Transmission lines get saturated long before the sites that want to connect are built out, and new transmission takes as long to permit and construct as the generation itself, sometimes longer.

The UK connection queue now runs into hundreds of gigawatts of projects waiting for grid connection dates, many of them pushed into the late 2030s. Scotland in particular generates surplus wind power that the existing north-to-south transmission cannot carry south to demand centres, which is the direct cause of the constraint payments that make headlines. The fix is new transmission, principally offshore HVDC links bypassing the constrained onshore network. Those are under construction, but on a timescale of years, not months.

Supply chain and critical minerals

Offshore wind at scale is a heavy industrial proposition. A single large modern project consumes several hundred thousand tonnes of steel for towers and foundations, hundreds of kilometres of export and inter-array cables, and tens of thousands of tonnes of copper. Global capacity in specific inputs is thin. Wind turbine installation vessels capable of the latest turbine sizes number in low dozens globally. HVDC cable manufacturing capacity is booked years out.

Direct-drive offshore turbines use permanent magnets that rely on neodymium and dysprosium, rare earth elements currently dominated by Chinese production. Gearboxed turbines avoid the rare earth dependency but have higher maintenance. Neither approach eliminates the copper intensity of offshore wind, which is substantial and a genuine constraint as the copper market tightens this decade.

Intermittency, and what actually solves it

Variable renewables do not generate on demand. At low penetration this is a minor issue. Above roughly forty percent of annual generation, it becomes a central planning problem. The solutions are well understood and cumulatively expensive. Batteries handle sub-daily balancing. Pumped hydro and longer-duration storage handle multi-day gaps. Flexible gas turbines remain on the system as insurance for the long dark calm weeks. Strong interconnection to neighbouring grids diversifies the wind resource geographically. Demand side response shifts some load to match supply.

The honest assessment is that the UK can reach around sixty to seventy percent variable renewables without heroic assumptions, using the tools above. Getting higher gets progressively more expensive, because the last ten percent requires over-building generation and storage that sits idle much of the year. This is not an argument against the transition, it is an argument for realistic cost expectations and for keeping a pragmatic portfolio of firm low-carbon generation, which in a UK context means nuclear, and potentially carbon-captured gas and hydrogen-fuelled turbines, alongside the wind build.

What to watch next

Three inflection points worth tracking. First, whether AR6 and AR7 pricing signals a new stable cost floor for UK offshore wind, or whether further inflation pushes prices higher again. Second, whether the first utility-scale floating projects in UK and EU waters actually deliver on budget and timeline, or repeat the early fixed-bottom experience of significant overruns. Third, whether grid reinforcement, particularly the Eastern Green Link HVDC projects and Scotland-to-England upgrades, lands on schedule. The answers to those three determine whether the UK hits its 2030 offshore wind ambitions in substance, not just in press releases.