Great Britain’s Monthly Energy Stats | National Energy System Operator

What the numbers show

Wind set a new instantaneous generation record of 23,880 MW at 13:30 on 25 March, supplying 60% of GB electricity at peak — enough for roughly 23 million homes on NESO's own estimate. Across the full month, wind provided 34.8% of generation, making it the single largest source. Gas came second at 22.2%.

Peak demand hit 38,160 MW on 3 March at 18:00, down 5,405 MW from February's peak. This is the seasonal pattern: longer daylight, milder temperatures, falling heating load.

On gas supply: UK and Norwegian fields provided 74%, LNG imports 24%, storage withdrawal 2%. Distribution networks carried 65% of gas to end consumers; power stations took 16%.

Trends

The wind record in context. The 23,880 MW peak is a system milestone but it is a single half-hour. Wind's monthly share at 34.8% is strong but not unprecedented — January 2025 hit similar levels. The gap between the 60% instantaneous penetration and the 34.8% monthly average is the intermittency problem compressed into one data point. The system can physically handle very high wind penetration for short periods. The question is what happens in the other 700-odd half-hours of the month when wind is producing less.

Gas remains structurally necessary. At 22.2% of monthly generation, gas is doing exactly what the system needs it to do: filling the gaps that wind cannot. That 22.2% is not a failure of decarbonisation. It is the cost of intermittency. Every percentage point of wind's monthly share that falls below its instantaneous peak must be met by something dispatchable, and in GB that means gas. Nuclear provides baseload but cannot flex. Interconnectors help at the margin but are price-takers. Storage is growing but nowhere near the scale needed to shift energy across days, let alone weeks.

Demand is falling seasonally but the structural question is rising. The 38,160 MW peak is unremarkable for early March. What matters is the trend in underlying demand over the next 12-24 months as data centres, electrification of heat, and EV charging begin to show up in the numbers. NESO's own Future Energy Scenarios project significant demand growth from the late 2020s. Monthly peaks in winter 2026-27 will be the first real test of whether that growth is materialising or whether efficiency gains and behind-the-meter generation are offsetting it.

Gas supply mix is stable. The 74% domestic-plus-Norway share reflects GB's post-Russian-gas supply structure. LNG at 24% is the swing supplier. Storage at 2% in March is normal — storage is drawn down through winter and refilled in summer. The key number to watch is the LNG share: if it rises significantly, it signals either domestic production decline or increased competition for global cargoes, both of which feed through to wholesale gas prices and therefore to electricity prices via the gas marginal pricing mechanism.

What to watch

Curtailment costs behind the record. NESO's press release celebrates the 23,880 MW wind record. It does not mention what that record cost in constraint payments. When wind generation exceeds what the transmission network can physically transport from where it is generated (predominantly Scotland) to where it is consumed (predominantly England), NESO pays wind farms to curtail and pays gas plants further south to generate instead. The constraint cost of a high-wind day can run into tens of millions of pounds. The March data release does not include balancing costs — those appear in Elexon's settlement data — but anyone using the 60% figure to argue the system is ready for higher wind penetration should be asked: at what constraint cost?

Implications for the TNUoS charging review. Ofgem's ongoing Transmission Network Use of System charging review will determine how the costs of transporting power from northern wind farms to southern demand centres are allocated. The 25 March record is a data point in that review: the system can accommodate the power, but the network behind it is paying the price. Stronger locational signals in TNUoS would change the economics of where new wind gets built. The current methodology socialises constraint costs across all consumers.

Gas plant economics. At 22.2% of generation, gas is running enough to remain commercially viable but not enough to be comfortable. Every month that wind's share rises, gas plant load factors fall, and the Capacity Market clearing price — the mechanism that keeps gas plants open for when wind drops — comes under pressure to rise. The March numbers are another data point in the slow squeeze on gas plant economics that will eventually force a reckoning on how much the system is willing to pay for dispatchable backup.