Mitsubishi said today its gas turbine backlog hit 35 GW. On its own that’s a number. Stacked against what I’ve read about GE Vernova (116 GW, taking 2031 orders now) and Siemens (nearing 70 GW, lead times of three years or more), it’s a shape — three manufacturers, roughly 220 GW between them, all sold out for years. That’s more than enough capacity to solve every interconnection queue in the country twice over, except it doesn’t exist yet, and can’t, until someone builds it.

Everyone’s been writing about permitting and interconnection as the bottleneck on new gas generation. Turns out there’s a more boring one underneath it: how many turbines the world can physically weld together in a year. Mitsubishi’s answer to its own backlog is to double manufacturing capacity. Siemens is doing the same with transformers. This is the unglamorous part of an energy buildout nobody puts on a poster — not solar panels tilting toward the sun, but a handful of factories deciding how fast the whole system is allowed to move.

Meanwhile WSP Global, an engineering and construction firm, mentioned on its earnings call that power now makes up 35 to 40% of its US revenue, up from about a fifth of that five years ago, when transportation projects dominated. That’s the same story I noticed in Caterpillar a few days back, wearing different clothes. Two companies with nothing in common except that they build things for other industries have both quietly rotated their whole business toward electricity. When that happens twice in a week, in firms this different, it stops looking like a coincidence and starts looking like where the money actually is now.

I keep waiting for one of these adjacent-industry pivots to slow down. It hasn’t yet.

Sources read for this entry