China is throwing away enough clean power to run Mexico. The grid is the problem.

China is generating more clean electricity than ever, and turning a record share of it away. The bottleneck in the renewable transition is no longer the panel factory. It is the wire, and the rules that decide which electrons get to use it.

In the first half of 2026, China curtailed 360 terawatt-hours of solar and wind power its grid could not absorb, according to a Reuters wire dispatch by Colleen Howe and Sudarshan Varadhan, citing a joint report from Global Energy Monitor and the Center for Research on Energy and Clean Air. That is enough electricity to run Mexico for a year, and it is up 49% from a year earlier. Curtailment is the industry term for turning solar and wind output down because the wires, the dispatch rules, or the local demand cannot take it. China is the world's largest installer of solar panels. It is also the world's largest discarder of their output.

The waste is not a one-off. The same Reuters dispatch flagged Australia and Japan as failing to keep pace with their own renewables buildout. Each is hitting the same wall at a different scale: more panels and turbines, but transmission, market rules, or both, lagging behind.

Two numbers illustrate how badly the official picture undersells the problem. China's National Energy Administration reported that 8.6% of solar output and 9.1% of wind output were curtailed in the first half of 2026. Global Energy Monitor and CREA, using weather-adjusted data that captures generation operators did not even bother to report, put the combined wind-and-solar curtailment rate at 26.1%. The methodology gap is itself the story. NEA stopped publishing monthly curtailment-by-province data in March 2026, and the gap between the government figure and the independent figure has widened as the buildout has accelerated.

Why is so much clean power being turned away? Two structural reasons. First, the transmission grid has not been built fast enough to move solar from the desert provinces in the west and north to the demand centers on the coast. Second, supply contracts that guarantee coal-fired plants a minimum number of operating hours effectively force the dispatcher to keep coal running first and to throttle solar and wind second. The result is a system that has spent the last five years building cleaner power plants faster than the wires and rules to use them.

The market is starting to price that in. New solar installations in China dropped 66% this year, Reuters reported, linked to both the curtailment pressure and a new policy that removed the guaranteed fixed price for renewables. Yuan Ren, an analyst at Wood Mackenzie, described the problem as "structural, not a temporary bottleneck." A temporary bottleneck gets fixed by a new line and a few quarters of paperwork. A structural one changes which projects get built.

The signal is already moving capital. New clean-power investment in China is shifting away from standalone solar farms and toward solar-plus-storage projects, which can store midday solar output and dispatch it when the grid has room. Storage does not solve the transmission problem. It solves the timing problem. The other response, slower but more durable, is dispatch reform: rewriting the rules so that coal plants are paid for capacity and ramp, not for running flat-out, and so that clean power is dispatched first when it is available.

China's coal-fired generation is set to rise again in 2026, ending a decade of decline. Read narrowly, that looks like a verdict against the transition. Read alongside the curtailment numbers, it is a verdict on how fast the wires, storage, and dispatch rules are catching up with the panels.

The next data point to watch is whether NEA resumes monthly curtailment data and whether the methodology gap with the independent trackers narrows. The wire will move on to the next solar-installation figure. The underlying constraint, the wire and the rule that decides which electron gets used, will not.