> solar at something like 25%

The graph at https://ourworldindata.org/grapher/solar-pv-energy-consumpti... seems to indicate the real world outcome is something more like 12.9%. That is, pick a dot on the graph and look at the capacity (watts) versus how much was generated in 2024 (watt-hours), and the number ends up vaguely looking like 1000 watt-hours generated for every watt of capacity. Given that there's 8760 hours in a year, that's vaguely in the 12% range.

The number for "World" is 2,110,000 GWh consumed for 1,866 GW of capacity, which means 2110000÷(1866×8760) = 12.9% of "capacity". Running the numbers for every country (there's a csv!) shows expected cloudy/northerly countries down near 8-9% (UK, germany, norway) and the sunnier ones near 20%... The USA is 19.8% which tracks given how popular solar is in the sunnier regions in particular.

Nobody in their right mind should be surprised by this, since the sun doesn't always shine, it gets dark at night, etc... it's unrealistic to assume this number will ever meaningfully change for solar. It's just the baseline expectation.

So yeah, "capacity" is misleading indeed. It means that for solar, "50% of global capacity" would mean something more like "6% of energy consumed".

But it's still super exciting to see the clear exponential growth here. (Speaking as someone who installed a 14KW array on his roof last year, solar makes me super excited.)