A developer in the Midwest had everything a project is supposed to need: a willing landowner, a strong wind resource, a buyer for the power, and financing lined up behind it. What the project did not have, three years into development, was a signed interconnection agreement. The turbines were not the problem. The panels were not the problem. The permission was the problem — and that story, with local variations, now describes the American energy buildout better than any technology forecast.
The scale of the logjam is public, and it is enormous. Lawrence Berkeley National Laboratory's Queued Up 2026 edition counts 2,061 gigawatts of generation and storage actively seeking grid interconnection in the United States at the end of 2025: 1,312 GW of generation plus 749 GW of storage, spread across roughly 8,200 projects. That total fell 10% from 2024, but not because projects raced to completion — more than 750 GW was withdrawn during 2025, while only about 600 GW of new requests arrived. The queue shrank because developers gave up, not because the system sped up.
The history inside that queue is the sobering part. Of the capacity in interconnection requests submitted between 2000 and 2020, only 13% had come online by the end of 2025, while 75% had withdrawn. For the projects that did make it, the median timeline from interconnection request to commercial operations exceeds five years, and the median time just to reach a signed interconnection agreement was well over three years in 2025. A developer choosing a site today is therefore not choosing a location for next year. The developer is choosing a location for the early 2030s, under rules, prices, and county boards that may all have changed by then.
None of this means building has stopped. The U.S. Energy Information Administration reported that developers planned to add 64 GW of new utility-scale capacity in 2025 — a record — led by 33.3 GW of solar, 18.3 GW of battery storage, 7.8 GW of wind, and 4.7 GW of natural gas. Those are developer plans, not completed projects, and the gap between plan and completion is exactly where permitting lives. Battery storage shows what happens when a technology clears its hurdles: EIA data show operators added 10.4 GW of utility-scale battery capacity in 2024, a 66% increase that took cumulative battery capacity past 26 GW, making storage the second-largest source of new capacity after solar. Storage moved fast because it could often be sited where the grid already had room — on or beside existing plants and substations — rather than where a new line had to be argued into existence.
The counterargument deserves a fair hearing: some of the queue was never real. Speculative requests, filed cheaply to hold a place in line, inflated the backlog for years, and the withdrawals of 2025 partly reflect reforms pushing those placeholders out. The Federal Energy Regulatory Commission's Order No. 2023, which replaced first-come, first-served serial studies with first-ready, first-served cluster studies, was designed for exactly this cleanup. But even after the cleanup, 2,061 GW remains — several times the entire installed generating capacity of the country — and the median project still waits more than five years. Clearing speculation explains part of the backlog. It does not explain the rest.
What explains the rest is that permission is a stack, not a single approval. A project needs a county zoning decision, often a state siting decision, an environmental review, a land assembly that survives holdouts, and an interconnection study whose upgrade costs can arrive late and large. Each layer has its own timeline, its own opponents, and its own appeal rights, and a failure at any layer sends the developer back to the start of another. The developer who treats permitting as paperwork to be filed after site selection has the sequence backwards: the permitting path is the site selection.
The implication for anyone building in this market is practical, not philosophical. Rank sites by permission risk before ranking them by resource quality: queue position and upgrade exposure first, county ordinance posture second, resource third. Budget five years, not two, from request to operation, and treat any timeline shorter than that as a claim to be verified, not a plan to be believed. And build the permitting file the way a litigator would — because sooner or later, someone will read it like one. The energy transition will be built project by project, county by county, hearing by hearing. The firms that master that unglamorous sequence will own the buildout; the firms that treat it as an afterthought will own the withdrawals.
Sources: Lawrence Berkeley National Laboratory, Queued Up: 2026 Edition (2,061 GW active at end of 2025; 1,312 GW generation; 749 GW storage; ~8,200 projects; 10% year-over-year decrease; over 750 GW withdrawn in 2025; about 600 GW new requests; 13% of 2000–2020 request capacity online by end of 2025; 75% withdrawn; median request-to-operation over 5 years; median request-to-agreement well over 3 years), https://emp.lbl.gov/publications/queued-2026-edition-characteristics and https://emp.lbl.gov/news/backlog-power-plants-seeking-transmission-grid-connection-eased-somewhat-2025-amidst. U.S. Energy Information Administration, 2025 planned additions (64 GW total; solar 33.3 GW; battery storage 18.3 GW; wind 7.8 GW; natural gas 4.7 GW), as reported by pv magazine, https://www.pv-magazine.com/2025/08/21/eia-projects-record-64-gw-us-power-capacity-buildout-in-2025/. U.S. Energy Information Administration, battery capacity 2024 (10.4 GW added; 66% growth; cumulative capacity over 26 GW), https://www.eia.gov/todayinenergy/detail.php?id=64705. Federal Energy Regulatory Commission Order No. 2023 (first-ready, first-served cluster study process), summary at https://www.troutman.com/insights/troutman-pepper-summary-of-ferc-order-no-2023-on-generator-interconnection-reform/.