Electricity

What powers the GB grid today, and what Clean Power 2030 asks of it.

In 2024, GB electricity averaged 160 gCO₂/kWh, roughly a third of its 2010 carbon intensity. Wind and solar provided 37 percent of generation, nuclear 14 percent, gas 26 percent. The system is already running on a mix that would have been considered impossible a decade ago. Getting to 95 percent clean by 2030 is a different problem from where we were in 2015.

03 Route 3 of 12 · Operations and networks
16 min read 5 sections 3 diagrams 1 decision tool Last verified

After this route you will be able to

  • Name the six fuel sources that covered GB electricity in 2024 and their share.
  • Explain what NESO does in the seconds, minutes, and hours after a generator trips.
  • Describe how Dinorwig pumped storage and grid-scale batteries deliver frequency response.
  • Read a carbon intensity number and interpret what it says about the current mix.
  • Make a reasoned call on the Clean Power 2030 pathway versus a 2035 target.
Offshore wind turbines at sea, part of the GB offshore wind fleet

11 May 2024Sunday lunchtime · Britain's first full coal-free century

At 14:00 BST, GB crossed 100 days without a single unit of coal generation. The 142-year-old era ended with a footnote.

Ratcliffe-on-Soar, the last working coal power station in Great Britain, had been offline for months. Behind it was a long, deliberate phase-out: Ratcliffe closed permanently on 30 September 2024, ending coal generation on a grid that had run on it since Thomas Edison lit Holborn Viaduct in 1882.

No-one noticed during lunch. Britain's carbon intensity that afternoon was 98 gCO₂/kWh, about 5 percent of its 1990 level. Wind was doing 42 percent, gas 19 percent, nuclear 13 percent, solar 11 percent, biomass 6 percent, imports 9 percent. The mix had shifted not through one dramatic intervention but through twenty years of small, cumulative ones.

The quiet absence of coal is the largest single change in British energy in a century. It happened without a supply crisis and without a price spike. The system adapted. The question now is whether the same pattern can repeat for gas over the next five years.

Britain turned off coal between 2015 and 2024 without noticeable disruption. Can it do the same for gas by 2030, or is the remaining decarbonisation qualitatively harder?

The answer depends on three things. What the current mix actually is. How the system is balanced second by second. And what the remaining gap needs, which is not more installed capacity but more of the right kind of capacity in the right places.

Section 01 · The 2024 mix

Six fuels, one integrated system.

GB electricity in 2024 averaged roughly 285 TWh of generation. The mix below is typical across the year. Any given hour can look very different, with wind ranging from under 5 percent to over 60 percent depending on weather.

Diagram 01 · GB generation mix 2024 (share of annual TWh)
Wind 30 % Onshore 14 GW, offshore 15 GW operational; 45 GW target by 2030.
Gas (CCGT) 26 % Flexible mid-merit and peak cover; targeted to ≤5 % of generation by 2030.
Nuclear 14 % Hinkley Point B closed 2022; four remaining AGRs retire by 2028. Hinkley Point C commissions 2029–30.
Imports (interconnectors) 11 % Net imports from France, Netherlands, Belgium, Norway, Denmark, Ireland.
Biomass 7 % Drax predominantly; biomass subsidy continues under the transition regime to 2027.
Solar 7 % 17 GW operational; 70 GW target by 2035 under Clean Power pathway.
Hydro and storage 3 % Includes 2.8 GW of pumped storage (Dinorwig, Cruachan, Ffestiniog) and conventional hydro.
Coal 1 % Jan–Sep 2024 residual; zero from 1 October 2024.
Other 1 % Oil, waste-to-energy, landfill gas.

Source: DESNZ Energy Trends March 2025, NESO operational data 2024. Percentages sum to approximately 100; rounding applied.

Section 02 · Balancing the second

Every second, NESO brings supply and demand into line.

The grid is not balanced by a market. It is balanced by an operator using a market as its tool. The difference matters, because a market alone would not respond fast enough to the physical reality of a trip or a demand spike.

At the one-second timescale, inertia does the work. Synchronous generators (gas turbines, nuclear plants) have spinning mass that resists frequency change. When a generator trips, the remaining synchronous fleet slows slightly, buying time.

At the 100-millisecond timescale, Dynamic Containment kicks in. Since 2020, NESO has procured this product from batteries that can inject power within 0.5 seconds of a trigger. By 2026, about 2 GW is contracted. It is the single largest change in GB frequency response in a generation.

At the 10-second timescale, Dinorwig and Cruachan pumped storage respond. Dinorwig can deliver 1.7 GW from standstill in 12 seconds. Cruachan adds 0.44 GW. These two plants are the operator's fastest deep reserves.

At the 30-minute timescale, the Balancing Mechanism accepts offers and bids to shift generation and demand. Elexon publishes every decision as machine-readable data on BMRS.

Frequency response services shall be capable of full activation within the times specified. Dynamic Containment shall respond within 0.5 seconds and deliver full contracted power within 1 second.

NESO Response and Reserve services market information, February 2026

Section 03 · The inertia problem

A grid of inverters has no rotating mass.

Wind turbines and solar panels do not generate electricity the way a steam turbine does. They produce DC, convert it to AC through power electronics (an inverter), and synchronise to the grid electronically. That is efficient and clean. It is also, at system level, a problem.

System inertia is the physical resistance to frequency change that comes from rotating mass. Traditionally, GB had about 250 GVA·s of inertia at the afternoon minimum. By 2024, on a sunny windy day, that had fallen below 100 GVA·s. The 2019 blackout happened on a day when inertia was near the system minimum. The frequency fell faster than the response products could catch it.

The replacements are not like-for-like. Synchronous condensers (spinning machines that provide inertia without generating power) have been installed at Keith, Rassau and Phoenix, with more planned. Grid-forming batteries can synthesise inertia through control algorithms but require specific converter design. Fast frequency response products (Dynamic Containment, Dynamic Moderation) catch faults faster than old response did.

Together these reduce the harm. None restores the buffer that a spinning 600 MW turbine gave for free.

Common misconception

More batteries automatically make the system more reliable.

A battery that is operating in grid-following mode (the default) does not provide inertia. It follows the voltage signal of whatever the grid is doing. Grid-forming batteries are different: they set their own voltage reference and can stabilise the grid around them. By 2026, fewer than a dozen are operational in GB, though Ofgem's 2024 Grid Forming Roadmap sets out a route to more.

Section 04 · Clean Power 2030

Clean Power by 2030 means 95 percent, not 100.

The 2024 Clean Power 2030 Action Plan sets a government target of 95 percent low-carbon electricity by the end of 2030. The final 5 percent stays as unabated gas for resilience. Understanding the shape of the remaining 95 percent is the hardest part of the plan.

The plan targets roughly 50 GW of offshore wind by 2030 (against 15 GW today), 45–50 GW of solar, 43–50 GW of onshore wind, and the completion of Hinkley Point C (3.2 GW). On an annual TWh basis, this gives 70 percent wind, 10–15 percent solar, 10 percent nuclear, 5 percent gas for resilience.

The binding constraints are not generation capacity. They are transmission reinforcement (ASTI pipeline, £58 bn to 2035), connection queue reform (TM04+, now being rolled out), storage and flexibility (15 GW of long-duration storage targeted), and system operation services (grid-forming, inertia products, fast response).

None of these is unprecedented. All of them together, delivered in under five years, is unprecedented.

Section 05 · The 2030 call

Clean Power 2030 or Net Zero 2035. Which pathway do you pick?

The 2024 target is ambitious by any measure. A 2035 pathway is cheaper and less risky. The trade-off is political signal, investment certainty, and the carbon saved in the intervening five years.

You are not choosing whether to decarbonise. You are choosing a speed, with consequences for cost, for risk, and for investor signal. Clean Power 2030. Hold the 95 % target. 2035 target with 95 % Net Zero. More time, same end-state. Hybrid: 2030 for generation, 2035 for full system integration. This is the current direction. The constraint is not funding; it is the ability to consent, build, and commission fast enough. Supply chain, planning, workforce are all binding. Start over This is the de facto position implied by many supply-chain analysts. It is less risky operationally. It is more risky politically. Start over This is the pragmatic compromise: reality of 2030 capacity, honesty about 2035 integration. It is closer to the pathway the Climate Change Committee has endorsed in its balanced scenario. Start over

Check your understanding

Three questions on what you have just read.

Nuclear is running flat out Gas is covering demand Wind and solar together are covering most of the demand Battery storage is discharging Dinorwig pumped storage System inertia from synchronous generators The Balancing Mechanism Interconnector reversal Coal generation Diesel peaking Unabated gas as the resilience backstop Biomass

Key takeaways

  • GB electricity in 2024 was 160 gCO₂/kWh on average, a third of 2010. Wind leads the annual mix at 30 %.
  • NESO balances the system on four timescales: inertia (seconds), Dynamic Containment (half-second), pumped storage (tens of seconds), Balancing Mechanism (half hour).
  • The system is losing inertia as inverter-based generation replaces synchronous. Grid-forming batteries, synchronous condensers, and fast response are the replacements.
  • Clean Power 2030 means 95 % clean, not 100 %. The last 5 % is kept as gas for resilience.
  • The 2030 target is a deployment-rate problem, not a technology problem. Supply chain, planning and workforce are binding.

References

  1. DESNZ: Clean Power 2030 Action Plan

    December 2024. Official target, deployment pathway, procurement schedule.

    Primary source for the 95 % target and mission architecture.
  2. DESNZ DUKES 2025, Chapter 5 Electricity

    Annual generation by fuel, capacity factors, carbon intensity.

    Authoritative UK government statistics, updated annually.
  3. NESO: Balancing Services and Response Products

    Dynamic Containment, Dynamic Moderation, Dynamic Regulation specifications.

    Primary source for response product definitions and timings.
  4. Ofgem: 9 August 2019 power outage investigation

    Full inertia and response analysis; enforcement decisions.

    Definitive case study for the inertia problem.
  5. Elexon BMRS: Balancing Mechanism Reporting Service

    Live GB generation, demand, price, balancing actions.

    Operational transparency feed.
  6. Climate Change Committee: Seventh Carbon Budget

    Independent advisory pathway analysis; balanced pathway assumptions.

    Independent comparator to government targets.

The next route turns from what to where. The gas system is the counterweight to electricity: buffered differently, operated differently, facing its own 2026 decision window.