The Interconnection Queue

The constraint moved and almost nobody repriced

Grid operators require that a new generator or battery be studied before it is built, so that connecting it does not destabilize the system. The studies are reasonable and the backlog they produced is not. Roughly 2,600 gigawatts of generation and storage were actively seeking interconnection as of a June 2026 Berkeley Lab count, more than double the 1,279 gigawatts the American grid actually has installed. A project that went from initial request to commercial operation in under two years in 2008 now takes nearly five.¹

The queue does not discriminate. It blocks batteries, solar farms, and natural gas plants alike. So the cheapness of manufactured energy, which is real and compounding, piles up at the gate rather than on the grid.

An engineering problem became a permission problem

This is the part that changes who wins. When supply was limited by the cost of building generation, the advantage went to whoever could build most cheaply. When supply is limited by the wait for a study, the advantage goes to whoever has the shortest legal path to a live meter. Capability stops being the scarce input and standing does.

Which is why anything already interconnected quietly became an asset class. Zach Dell's Base Power is the pure expression: chop the utility-scale battery farm into thousands of pieces, install them on houses that were connected decades ago, and the queue simply does not apply. The homeowner supplies the interconnection for nothing, along with the land and the site work.

The a16z essay makes the arithmetic vivid. During the July 2026 Texas demand records, Base discharged roughly 150 megawatts, about the size of a full utility-scale battery site. Had someone begun building that site in 2024, it would still have been sitting in the queue.¹

What the pattern does not survive

The moat is procedural, and procedures can be reformed. Faster studies or cluster processing would compress the advantage, and the sources that name the queue as an opportunity are all held by investors who benefit from it persisting. A reader should treat the queue as a condition with a plausible expiry rather than a permanent feature of the landscape, and ask of any company built on it what remains when the line moves.

The reader is directed to the file on Wright's Law, which explains why the equipment waiting in this line keeps getting cheaper while the line itself gets longer. The two facts together are the whole opportunity.