IN Brief:
- GEM and CREA estimate Chinese wind and solar curtailment reached 360TWh during the first half of 2026.
- Australia, Japan, and India are also recording increased renewable curtailment as generation growth tests network capacity.
- Transmission expansion, storage, flexible demand, and market reform are becoming critical alongside additional renewable generation.
Global Energy Monitor and the Centre for Research on Energy and Clean Air estimate that China curtailed 360TWh of wind and solar generation during the first half of 2026, exposing an increasingly large mismatch between new generating capacity and the infrastructure available to absorb its output.
The estimate is 49% higher than for the same period of 2025 and exceeds the entire increase in Chinese electricity demand during the six months. In physical terms, enough renewable electricity was available to cover that additional consumption, but transmission constraints, operating arrangements, and competing generation prevented a substantial proportion from reaching users.
The research groups’ estimate is considerably higher than official Chinese figures. The National Energy Administration has reported first-half curtailment of 8.6% for solar and 9.1% for wind, while GEM and CREA estimate that 26.1% of potential combined wind and solar output was rejected after accounting for unreported curtailment through weather-adjusted analysis.
The methodological gap makes the precise scale contested, but both datasets point towards the same system problem. Renewable generation is being installed faster than the networks, storage capacity, operating rules, and flexible electricity demand needed to integrate it consistently.
Generation capacity is outrunning the grid
Curtailment occurs when an otherwise available generator is instructed or forced to reduce output because the wider electricity system cannot accept it. The immediate cause may be a congested transmission line, insufficient local demand, system-security limits, inflexible conventional generation, negative pricing, or a lack of storage capable of shifting electricity into another period.
Some curtailment can be economically rational. Building enough transmission and storage to accommodate every possible renewable peak would create infrastructure that may be lightly used for much of the year, so power systems normally tolerate a degree of rejected output where reinforcement costs would exceed the value of the electricity recovered.
The calculation changes when curtailment becomes frequent enough to undermine project economics or materially alter national generation planning. Developers may connect new wind or solar capacity but sell substantially less electricity than its resource and nameplate rating suggest, while networks face pressure to justify why new generation is being built faster than reinforcement around it.
China’s position is complicated further by continued additions of coal-fired capacity. GEM and CREA report that 30GW of new coal generation entered service during the first half of 2026, the highest first-half addition for a decade, while only 2.7GW was retired.
Contracting arrangements can reinforce that physical imbalance. Where coal plants retain guaranteed operating commitments, the system has less room to reduce conventional generation during periods of high wind and solar output. Renewable curtailment can therefore reflect both limited wires and rules governing which generating assets are allowed to remain online.
The problem is spreading beyond China. Australia’s National Electricity Market curtailed approximately 2.93TWh of wind and solar output during the first six months of 2026, around 7% of generation and 37% more than a year earlier. Japan rejected approximately 2.35TWh, equal to about 4% of renewable output and an increase of 34%.
India curtailed 8.13TWh of solar electricity in the quarter ending June, equivalent to approximately 14% of output during the period. Its separate difficulties allocating transmission connectivity to delayed renewable projects show how generation and network development are now colliding at both ends of the project cycle — before construction, through scarce connections, and after commissioning, through constrained output.
The figures are not directly interchangeable. Each system uses different definitions, market structures, generation mixes, and reporting methods, but the direction is sufficiently consistent to make renewable integration a network-engineering question rather than simply a generation target.
New transmission remains the most direct response where renewable resources are located far from major demand. High-voltage lines and substations can transfer electricity between regions, but their planning, consenting, procurement, and construction periods are generally longer than the development cycle for a solar farm or many onshore wind projects.
Battery storage can be deployed more quickly in some locations and absorb renewable electricity during periods of excess supply. It does not substitute for a transmission corridor where energy fundamentally needs to move between regions, but it can reduce peak export requirements, provide frequency and reserve services, and shift part of the available generation into higher-demand periods.
Flexible electricity demand offers another route. Industrial processes, electrolysers, data centres, electric heating, and vehicle charging can be operated more heavily during periods of abundant renewable production where commercial arrangements and operating requirements allow, reducing pressure to either curtail electricity or store all of it for later use.
Market arrangements determine how those physical assets are used. Connection agreements, locational signals, dispatch rules, ancillary-service markets, and long-term generation contracts can either expose or obscure the value of locating flexible assets where the grid actually needs them.
China’s 360TWh estimate is therefore less a measure of failed renewable technology than a measure of an increasingly unbalanced power system. Wind turbines and photovoltaic projects can now be deployed at a pace that transmission corridors, substations, storage fleets, system controls, and commercial structures struggle to match.
The investment signal will change as curtailment rises. Developers will increasingly have to assess usable output rather than theoretical generation, while network operators will face sharper scrutiny over where reinforcement, storage, or flexible connections produce the greatest reduction in wasted electricity.
Adding renewable gigawatts remains necessary, but the more difficult engineering metric is becoming how much of their output can be delivered when it is available. In several of the world’s fastest-growing power systems, that gap is now becoming too large to treat as an occasional operating constraint.