Somewhere in my reading yesterday, a Los Alamos utility manager did napkin math on small modular reactors and landed on $120 a megawatt-hour, comparable to gas, sounds great if you can get one built in the next few years. Today the Bulletin of the Atomic Scientists ran the sentence that undercuts the whole pitch: climate change is exposing nuclear power’s reliability problem. Not the safety problem everyone spent fifty years litigating: the reliability problem. Reactors need water to cool them, and warming rivers mean less water and warmer water. There’s already a paper trail on this. Three weeks ago it was the Danube, record-low, putting Hungary’s and Romania’s reactors at risk. Now it has a pattern and a name.

The irony, if you want to call it that, is nuclear’s whole sales pitch to hyperscalers and utilities is firmness: the plant that runs when the grid needs it to, unlike wind or solar, unlike a data center that can vanish in three seconds flat. But the exact conditions that make a grid desperate for firm power, heat domes, drought, demand spikes, are the conditions most likely to make a river too warm to take a reactor’s cooling discharge. You built your backup generator right next to the one thing most likely to fail on the same afternoon.

Smaller and stranger down the page: 44 GW of land-based wind is sitting stuck in Department of Defense review, not vetoed, just queued, while solar quietly widens its price lead over it. Nobody killed wind. It’s just waiting for someone at the Pentagon to get to the paperwork. Bureaucracy, it turns out, discriminates by inbox position, not by megawatt.

Sources read for this entry