The issue is that comparing "capacity" as a percentage is misleading. A baseload generation source can have average generation above 90% of its rated capacity, solar at something like 25%, wind something like 25-40%. Which means that saying "nearly 50%" of capacity can imply something closer to 15% of generation, and potentially even less if the amount of local capacity is high, because then you get periods when renewable generation exceeds demand and the additional generation has nowhere to go, which effectively reduces the capacity factor even more.
And on the other side, natural gas peaker plants can have a capacity factor even lower than solar and wind because their explicit purpose is to only be used when demand exceeds supply from other sources, so that "nearly 50%" in a grid which is entirely renewables and peaker plants could actually imply more than 50% of total generation. This is much less common in existing grids but it makes looking at the nameplate capacity even more worthless because you can't just multiply it by a fixed factor to get the real number.
Whereas if they would just publish the percentage of actual generation, that's what people actually want to know. But then you'd have to say "13%" or "24%" or whatever the real number is, instead of "nearly 50%".
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.)
https://pv-magazine-usa.com/2026/02/25/solar-and-storage-to-...
Humanity does far more wasteful things than build some extra solar panels.
But in much of the world, setting up PV is economically sound simply because it displaces a certain amount of kWh generated over the course of a year from other sources that are more polluting and more expensive.
In this regime, the dynamics of production over time don't matter yet.
At some point, when renewable generation has very high penetration, you'll reach a point where building more is uneconomical, and to then displace the remaining other power sources you'll need to overpay (ignoring externalities).
However, that's assuming no technological change on the way there, which is a whole separate topic.
I'm in favour of having it but the reason why you need to over provision is because of the intermittency. This can also push out proper base load (e.g. nuclear) although it's not simple.
You have to think about the portfolio.
In Britain at least there is also a bit of a sleight of hand where the marginal costs are reported but not the CFD strike prices used to incentivise the buildout.