IN Brief:
- Nuclear reactors generated 2,702TWh in 2025, 35TWh above the previous record set during 2024.
- Capacity under construction increased to 82GWe after eleven reactors started construction during the year.
- World Nuclear Association's 1,457GWe 2050 scenario assumes existing reactors remain operating, identified projects are delivered, and national capacity targets are met.
World Nuclear Association says reactors generated a record 2,702TWh of electricity during 2025, extending nuclear power’s previous annual high while highlighting the much larger gap between current construction and national capacity ambitions for 2050.
Generation increased by 35TWh from the 2,667TWh recorded in 2024, which had itself set a global record. Nuclear supplied around 9% of world electricity during the year, with the operating fleet maintaining an average capacity factor of 83.7%.
Capacity factor measures actual generation against the theoretical output available if equipment operates continuously at maximum capacity. The 83.7% fleet average therefore reflects more than installed megawatts: outage duration, maintenance, refuelling schedules, plant reliability, uprates, and operating performance all influence how much electricity existing reactors deliver.
That makes continued operation of the installed fleet central to near-term nuclear output. New reactors normally require long development and construction programmes, whereas maintenance improvements, justified uprates, plant restarts, and lifetime extensions can preserve or increase generation from assets whose sites, grid connections, workforces, and supply chains already exist.
World Nuclear Association reports 82GWe of new capacity under construction, up from 76GWe in the previous edition of its outlook. Eleven reactors entered construction during 2025, nine in China and two in Russia, while three reactors connected to the grid and seven were permanently shut down.
The organisation expects 54GWe of today’s construction pipeline to enter service by 2030 and another 28GWe by 2035. Those projects provide a comparatively visible route to additional nuclear output because physical construction has begun, although individual schedules can still be affected by financing, licensing, component delivery, quality issues, and commissioning performance.
The longer-term numbers require considerably more caution. World Nuclear Association’s 2026 outlook gives a potential 2050 global capacity of 1,457GWe, but that figure is conditional rather than a forecast that every identified project and national target will be delivered.
The scenario assumes currently operating reactors continue for up to 80 years unless a shorter operating period is identified, reactors classed as under construction, planned, or proposed enter service by 2050, and additional capacity required to meet national government targets is also built. Changing any of those assumptions changes the resulting total.
The gap between stated ambition and the physical project pipeline is substantial. The report says more than 550GWe required to meet national goals is not yet associated with projects classified as under construction, planned, or proposed. Its detailed projection puts additional capacity needed solely to satisfy government targets at 559GWe.
That distinction is the industrial challenge behind the headline 1,457GWe figure. A national capacity target can be announced without a selected reactor design, licensed site, financing structure, construction contractor, component orders, fuel contract, trained operating organisation, or completed grid connection. A construction programme cannot.
Even projects classed as planned or proposed retain delivery risk. World Nuclear Association explicitly notes that not every planned or proposed reactor can be assumed to result in an operating unit and that some existing reactors may not remain in service to 2050. The report therefore assesses countries separately for policy, institutional, infrastructure, project, and programme readiness.
The scale of the supply chain required becomes increasingly important as the pipeline expands. Reactor pressure vessels, steam generators, large forgings, turbines, pumps, electrical equipment, instrumentation and control, nuclear-grade valves, specialist cables, civil structures, and fuel all require qualified manufacturing capacity that cannot be expanded instantly.
Workforce is another constraint. New-build programmes compete with operating stations and lifetime-extension projects for nuclear engineers, welders, quality specialists, commissioning staff, regulators, operators, and project-management capability. Countries attempting several builds at once need repeatable programmes rather than assembling an entirely new delivery organisation for every site.
The report projects 1,303GWe of its 2050 capacity in countries that already operate nuclear reactors, while new entrant countries contribute another 154GWe. Existing nuclear markets have an advantage in regulation, operating experience, grid infrastructure, and established suppliers, although many must simultaneously manage ageing reactors and new construction.
World Nuclear Association’s own recommendations consequently emphasise keeping existing reactors operating where they continue to meet safety requirements, completing plants already under construction, scaling proven designs, developing newer technologies, strengthening financing, and expanding industrial and workforce capability.
The 2,702TWh generation record is a measured result from operating equipment. The 1,457GWe figure is a conditional description of what could be operating in 2050 if a much larger collection of policy intentions becomes physical projects. Between those figures sits the harder part of nuclear expansion — licensing sites, manufacturing equipment, financing construction, training workforces, and commissioning reactors repeatedly rather than announcing them.
