Why 2,000+ GW of New Power Is Stuck in a Queue — and What the Interconnection Bottleneck Means for the Grid

What an interconnection queue actually is

To connect a large power plant — a solar farm, a wind project, a battery, or a gas turbine — to the high-voltage transmission grid, a developer must file an interconnection request. The grid operator then runs a series of impact studies to figure out what new equipment or upgrades are needed before the project can safely plug in, and assigns the cost of those upgrades. The list of projects working their way through that study process is the interconnection queue. For most of the past two decades the process was first-come, first-served and serial: one project at a time, in the order it arrived. That design worked when queues were short. It broke down as renewables, storage and now data-center-driven demand flooded in.

How big the backlog got

Berkeley Lab's annual "Queued Up" tracking, compiled with Interconnection.fyi from nearly all U.S. grid operators, puts about 10,300 projects and roughly 2,300 GW of capacity actively seeking connection at the end of 2024 — roughly 1,400 GW of generation plus about 890 GW of storage. The backlog peaked near 2,600 GW at the end of 2023. By 2025 the active total had eased to about 2,061 GW as withdrawals surged past 750 GW in a single year. The mix at the end of 2024 was dominated by solar (about 956 GW), storage (about 890 GW) and wind (about 271 GW), with natural gas capacity jumping 72% year-over-year to 136 GW — a signal that data-center and firm-power demand is pulling gas back into the queue, not just clean energy.

Why projects wait so long — and why most never connect

The wait has quietly doubled. Berkeley Lab finds the median time from interconnection request to commercial operation rose from under two years for projects built in 2000–2007 to more than four years for those built in 2018–2024. And the sobering part: of the capacity that requested interconnection between 2000 and 2019, only 13% had reached commercial operation by the end of 2024; 77% had been withdrawn and 10% was still active. Filing a queue position is cheap relative to actually building a plant, so many developers stake a claim years early, then drop out when land, permits, power-purchase agreements, equipment or financing don't line up — or when the transmission upgrade bill proves too steep.

Why it matters now: demand is finally surging

For twenty years U.S. electricity demand was nearly flat, which is why a slow queue was an annoyance rather than a crisis. That has changed. NERC's 2025 Long-Term Reliability Assessment projects summer peak demand to grow 224 GW over the next ten years — 69% above its prior forecast — with winter demand growing a further 246 GW. Data centers, artificial intelligence and the broader digital economy account for most of the increase. The assessment flags 13 of 23 assessment areas at risk of resource-adequacy challenges within five years, with five regions — MISO, PJM, ERCOT and two WECC subregions — at "high risk" by 2030 and possible shortfalls as early as 2028. NERC's John Moura is careful to frame it as "not a prediction of failure but an early warning on the trajectory of risk." The paradox is sharp: the very projects needed to meet the new demand are stuck in the queue that is slowest to process them.

What reform is doing: FERC Order 2023

The central fix is FERC Order 2023, issued in July 2023 and phased in through 2024. It attacks the bottleneck on three fronts: it replaces serial studies with first-ready, first-served cluster studies that evaluate many projects at once; it removes the old "reasonable efforts" standard and adds penalties when transmission providers miss study deadlines; and it folds in advanced transmission technologies and inverter-based-resource modeling. To flush out speculation, it raises financial commitments — bigger study deposits, site-control and commercial-readiness requirements — and penalizes withdrawn requests. FERC Chair Willie Phillips noted at the time that projects faced an average wait of up to five years. Early signals are encouraging: several grid operators set records for executed interconnection agreements in 2025, with CAISO and ERCOT each clearing more than 30 GW of signed agreements.

The bottom line: reform is bending the curve, not breaking it

The 2025 data shows the reform beginning to bite — withdrawals of speculative projects are up, and the volume of newly executed interconnection agreements is rising in most regions with data. But queue timelines remain long, and demand growth is now the wild card it wasn't a decade ago. The metrics worth watching are cluster-study throughput, how fast data-center loads actually interconnect, the pace of regional transmission buildout, and how mature Order 2023 implementation becomes in each operator's territory. The queue is no longer an obscure plumbing problem; it is one of the main gates between today's record volumes of ready-to-build projects and a grid that can keep the lights on as load accelerates.