Grid Interconnection Queue Statistics 2026
By Axis Intelligence Research
Co-author: Aidan Jad | Last updated: July 29, 2026 | License: CC BY 4.0
As of the end of 2025, approximately 2,061 GW of generation and storage capacity was actively seeking grid connection across the United States — 1.5 times the country’s entire installed generating fleet of 1,374 GW. The median project that actually reached commercial operation in 2025 spent 61 months in the queue, up from 22 months in 2008. Only 13% of capacity that entered the interconnection queue between 2000 and 2020 has ever been built. The rest — 75% — was withdrawn before it could generate a single watt.
Quick Answer: U.S. Grid Interconnection Queue in 2026
The U.S. grid interconnection queue held approximately 8,200 active projects representing 2,061 GW of combined generation and storage capacity at the end of 2025, according to Lawrence Berkeley National Laboratory’s Queued Up 2026 Edition. That figure declined 10% from 2024’s 2,290 GW active total — not primarily because more projects got built, but because withdrawal rates accelerated. According to Axis Intelligence Research’s Queue Build-vs-Withdraw Ratio (GQBWR™), for every 1 GW that reached commercial operation in 2025, approximately 14 GW were withdrawn — a ratio that captures the structural mismatch between queue ambition and grid delivery capacity.
Key Findings
- As of December 31, 2025, approximately 8,200 projects totaling 2,061 GW of generation and storage capacity were actively seeking U.S. grid interconnection, per Lawrence Berkeley National Laboratory’s Queued Up 2026 Edition — 1.5 times the country’s installed generating capacity.
- The median U.S. project that reached commercial operation in 2025 spent 61 months in the interconnection queue, up from 36 months in 2015 and 22 months in 2008, a 177% increase over 17 years, per LBNL Queued Up 2026 Edition.
- According to Axis Intelligence Research, the 2025 Grid Queue Build-vs-Withdraw Ratio (GQBWR™) was approximately 14.2×, meaning roughly 14 GW of queue capacity was withdrawn for every 1 GW that reached commercial operation — establishing this metric’s baseline reading for the U.S. transmission grid.
- Of all capacity that submitted interconnection requests from 2000 to 2020, only 13% had reached commercial operations by the end of 2025, while 75% was withdrawn, per LBNL’s 2026 Edition dataset covering 22,061 requests.
- Over 2,500 GW of renewable, large-load, and storage projects are currently stalled in grid queues worldwide, per the International Energy Agency’s Electricity 2026 report, identifying grid connection queues as “a critical bottleneck” that will require annual grid investment to rise approximately 50% by 2030 from today’s USD 400 billion.
How Large Is the U.S. Grid Interconnection Queue?
The U.S. interconnection queue peaked at roughly 2,290 GW of active capacity at the end of 2024, then fell to 2,061 GW by end of 2025 — a 10% reduction. The decline looks like progress. It is not straightforwardly good news.
To understand why, the inflow-outflow picture matters more than the headline number. During 2025, approximately 600 GW of new requests entered the queue while over 750 GW withdrew. Projects coming online — the only outflow that represents actual grid capacity additions — amounted to around 53 GW, yielding an annual throughput rate of just 2% of the then-active queue, per LBNL’s analysis. The queue shrank because attrition outran additions, not because the interconnection process became more effective at delivering capacity.
| Metric | End 2023 | End 2024 | End 2025 | Source |
|---|---|---|---|---|
| Active queue capacity (GW) | ~2,600 | ~2,290 | 2,061 | LBNL, Queued Up |
| Active projects (count) | — | ~10,300 | ~8,200 | LBNL, Queued Up |
| GW reaching COD (annual) | — | — | ~53 | LBNL, Queued Up 2026 |
| GW withdrawn (annual) | — | — | >750 | LBNL, Queued Up 2026 |
| Net change (GW) | — | — | –229 | LBNL, Queued Up 2026 |
Source: Lawrence Berkeley National Laboratory, Queued Up 2026 Edition, June 2026.
The queue’s total capacity is still 50% larger than the installed U.S. power fleet. Grid investment, grid permitting timelines, and the interconnection study process itself have not kept pace with the volume of applications. Planning, permitting, and completing new high-voltage grid infrastructure takes anywhere from 5 to 15 years, the IEA notes in Electricity 2026 — while new data center campuses take 1 to 3 years to build. That asymmetry is the root problem, and it has no near-term solution that doesn’t require both regulatory reform and physical construction.
What Is the Average Wait Time to Connect to the Grid?
The median project that built out and reached commercial operation in 2025 spent 61 months — just over five years — from its initial interconnection request to its commercial operations date, per LBNL’s duration analysis across six ISOs and 19 non-ISO balancing areas.
That number has been climbing for two decades without interruption.
| Projects reaching COD in: | Median IR-to-COD (months) | Source |
|---|---|---|
| 2000–2008 | ~22 | LBNL, Queued Up 2026 |
| 2015 | 36 | LBNL, Queued Up 2026 |
| 2019–2024 | >4 years (~50+) | LBNL, Queued Up 2026 |
| 2025 | 61 | LBNL, Queued Up 2026 |
Source: LBNL, Queued Up 2026 Edition (emp.lbl.gov/queues). Sample includes projects with valid COD dates from 6 ISOs and 19 non-ISO balancing areas.
The figure refers to projects that actually finished the process. The projects still stuck in the queue — over 8,000 of them — have not yet started that clock. And the 75% that will eventually withdraw never complete it at all.
There are two distinct phases of delay. First, the interconnection study process itself: the median time from interconnection request to signed interconnection agreement was 45 months in 2025, per LBNL, up from shorter historical averages as cluster study transitions and reform compliance processes multiplied the workload at ISOs. Second, the post-agreement construction phase — from signed interconnection agreement to commercial operation — has also extended, to a median of 31 months in 2025, up from 21 months in 2008. Both phases are getting worse simultaneously.
Project size matters considerably. Smaller projects (under 5 MW, eligible for FERC’s Small Generator Interconnection Procedures) complete studies in a median of 12 months and reach operation in a median of 20 months, per LBNL. Projects above the 20 MW threshold — subject to the Large Generator Interconnection Procedures — cluster around a median of 30 months from request to interconnection agreement, and well over four years from request to operation. This distinction is not incidental: virtually all utility-scale renewable, storage, or gas projects — and all data center-serving generation — exceed 20 MW.
Grid Interconnection Queue by Resource Type
Aidan Jad’s read on the composition shift: the gas surge to 253 GW is the single most consequential data point in the 2025 queue. It’s not a signal that developers have given up on clean energy — it’s a signal that load is outpacing renewable buildout, and that developers believe gas is the fastest path to a signed interconnection agreement. ERCOT’s “connect and manage” model, which avoids serial cost allocation studies, makes gas competitive on timeline in Texas. Until that asymmetry closes, expect gas’s share of the queue to keep climbing even as the overall queue shrinks.
| Resource type | Active GW (end 2025) | Year-over-year change | Source |
|---|---|---|---|
| Solar | 773 | –19% | LBNL, Queued Up 2026 |
| Storage (battery, 99%) | 749 | –16% | LBNL, Queued Up 2026 |
| Natural gas | 253 | +86% | LBNL, Queued Up 2026 |
| Wind (onshore + offshore) | 220 | –19% | LBNL, Queued Up 2026 |
| Nuclear | 10.4 | +96% | LBNL, Queued Up 2026 |
| Geothermal | 4.8 | +170% | LBNL, Queued Up 2026 |
| Hydro | 2.8 | –49% | LBNL, Queued Up 2026 |
Source: Lawrence Berkeley National Laboratory, Queued Up 2026 Edition (emp.lbl.gov/queues).
Solar and storage together still dominate, at 773 GW and 749 GW respectively, though both fell sharply from 2024 peaks. The solar decline reflects the combined effect of CAISO’s cluster pause (no new requests accepted in 2024 or 2025), PJM’s ongoing transition, and expiration risk for federal tax credits that made some projects economically unviable even before connection costs were factored in. Storage’s decline follows the same market signals: fewer new requests in 2024, a year when 32% less battery capacity applied compared to 2023, has now propagated to lower active totals in 2025.
The nuclear and geothermal surges are real but start from a small base. Nuclear active capacity doubled year-over-year to 10.4 GW, with MISO and PJM driving most of the increase. Geothermal tripled to 4.8 GW, concentrated entirely in the non-ISO West and CAISO. These are signal numbers — they indicate where developer attention is shifting — but at a combined 15 GW they remain far below the 1 GW-scale resource types.
Hybrid plant configurations — solar-plus-storage, gas-plus-storage — are increasingly the norm. Nearly half of all active solar capacity (376 GW, or 49% of the solar total) is now in hybrid configurations, as is 48% of active storage (358 GW). CAISO leads in hybridization: 91% of its active solar capacity is in hybrid configurations. The West (non-ISO) is similarly hybridized at 85%.
Grid Interconnection Queue by Region
Active queue capacity is highest in the non-ISO West (567 GW), followed by ERCOT (408 GW) and MISO (382 GW). The West’s dominance reflects the geographic scale of the territory and the concentration of solar and geothermal resources in the Southwest; ERCOT’s high volume reflects both the absence of study-phase barriers under its “connect and manage” approach and Texas’s extraordinary growth as a generation hub.
| Region | Active GW (end 2025) | Active projects | Notes |
|---|---|---|---|
| West (non-ISO) | 567 | 1,603 | Highest by volume; solar/geo dominant |
| ERCOT | 408 | 1,527 | Only region to grow in 2025; unique model |
| MISO | 382 | 1,646 | Queue cap implemented 2025 |
| SPP | 171 | 683 | CPP launched March 2026 |
| CAISO | 191 | 432 | No new requests 2024–2025; Cluster 16 opens Oct 2026 |
| Southeast (non-ISO) | 153 | 957 | Gas-dominant |
| PJM | 144 | 1,134 | Largest RTO; reopened queue 2026 |
| ISO-NE | 14.8 | 72 | No new requests 2025; first cluster Oct 2026 |
| NYISO | 29.2 | 190 | No new requests 2025; first cluster July 2026 |
Source: LBNL, Queued Up 2026 Edition (emp.lbl.gov/queues). Active capacity includes some estimates for missing hybrid storage capacity.
Texas accounts for 21% of all active solar, 20% of all active gas, 22% of all active storage, and 22% of all active wind capacity in the U.S. queue, per LBNL’s state-level analysis. California holds 15% of active storage and 8% of active solar.
ERCOT grew in 2025 while every other region contracted. This is not coincidental. ERCOT’s “connect and manage” interconnection service allows generators to connect without waiting for full cost allocation studies to clear — they bear congestion management obligations instead, but they can reach commercial operation significantly faster. LBNL data shows ERCOT’s median IR-to-IA duration is 24 months, versus the median 37–41 months for FERC-jurisdictional ERIS and NRIS projects at other ISOs. CAISO and ERCOT each executed interconnection agreements amounting to more than 50% of their respective peak load capacity in 2025, a record volume.
What Is the Interconnection Queue Completion Rate?
Low. Drastically lower than most industry participants assume.
Of all capacity that submitted interconnection requests between 2000 and 2020 (22,061 requests, representing approximately 3,501 GW), only 13% had reached commercial operations by the end of 2025. By project count rather than capacity, the figure is slightly higher at 19% — meaning smaller projects are proportionally more likely to complete than larger ones. By resource type, hydro has the highest historical completion rate at 29%, followed by nuclear at 18% and gas at 17%. Solar, storage, and wind — which now make up the overwhelming share of queue volume — have the lowest completion rates by capacity.
| Resource | Completion rate (cap., 2000–2020 cohort) | Note |
|---|---|---|
| Hydro | 29% | Most projects sub-20 MW |
| Nuclear | 18% | Very small historical sample |
| Gas | 17% | Higher certainty of dispatch economics |
| Wind | Lower than gas | LBNL: significant variation by region |
| Solar | Lower than gas | Large cohort, high attrition |
| Battery storage | Very low | Most requests post-2015; outcome still developing |
Source: LBNL, Queued Up 2026 Edition (emp.lbl.gov/queues). 21,927 requests from 7 ISOs and 33 non-ISO balancing areas with comprehensive status data.
By region, only ERCOT achieves more than 20% capacity-weighted completion from the 2000–2020 cohort. ISO-NE ranks second by project count (over 25%), with CAISO (8%) and the non-ISO West (10%) lowest. These completion rate differences reflect market structure differences as much as study process differences.
Signing an interconnection agreement does not guarantee a project gets built. Of all capacity that signed interconnection agreements between 2000 and 2022, 41% had subsequently withdrawn by the end of 2025, according to LBNL. The economics, permitting, and financing conditions that made a project viable when it entered the queue may no longer hold after four or five years of study process.
Axis Intelligence Research: Grid Queue Build-vs-Withdraw Ratio (GQBWR™)
The Problem with Reporting Queue Size Alone
A queue number without a throughput number is a waiting room headcount. It tells you how many people are sitting in chairs; it tells you nothing about whether the door to the treatment room ever opens.
Axis Intelligence Research created the Grid Queue Build-vs-Withdraw Ratio (GQBWR™) to quantify the throughput dysfunction in a single, citable metric. It answers the question: for every gigawatt that makes it through the process and gets built, how many gigawatts give up and leave?
Formula
GQBWR (year) = GW withdrawn in year ÷ GW reaching commercial operation in year
Both inputs are taken from LBNL’s Queued Up annual dataset, which covers 7 ISOs/RTOs and 50 non-ISO balancing areas — approximately 98% of U.S. installed generating capacity.
2025 Baseline Reading
Inputs (end-2025, LBNL Queued Up 2026 Edition, retrieved July 29, 2026):
- GW withdrawn in 2025: >750 GW (LBNL lower bound; exact total estimated at approximately 750 GW using LBNL’s inflow-outflow methodology)
- GW reaching commercial operation in 2025: ~53 GW
GQBWR (2025) = 750 ÷ 53 = 14.2×
This is the baseline reading. Axis Intelligence Research will update this figure annually as LBNL publishes successive Queued Up editions.
Interpretation
A GQBWR above 10× indicates that the interconnection process is functioning primarily as an attrition mechanism — most capacity that enters exits via withdrawal, not via delivery to the grid. A ratio trending toward 1× would indicate a process delivering approximately equal output to its withdrawal rate.
The 14.2× baseline establishes that in 2025, the U.S. transmission grid was consuming roughly 14 gigawatts of developer effort, financing, and study costs for every gigawatt that successfully added capacity. That is the underlying economics of the interconnection backlog.
License and Citation
GQBWR™ is published under Creative Commons Attribution 4.0 International (CC BY 4.0). To cite: “Axis Intelligence Research Grid Queue Build-vs-Withdraw Ratio (GQBWR™), 2025 baseline, axis-intelligence.com/grid-interconnection-queue-statistics/.”
How Are ISO/RTOs Reforming the Interconnection Process?
FERC Order No. 2023, issued July 28, 2023 and described by the Commission as among the largest rules in its history, mandated that all transmission providers transition from a serial “first-come, first-served” study process to a “first-ready, first-served” cluster study approach. The compliance deadline ran through May 2024. As of mid-2026, all major ISOs are either implementing the cluster approach or are actively in transition — but LBNL explicitly notes that “most data presented in this report pre-date these reforms” and that it is “too early to measure and assess their full impact.”
In January 2026, FERC’s statement highlighted accelerating automation and AI in interconnection study processes. MISO piloted automated study software (SUGAR, by Pearl Street Technologies) that reduced study timelines by 98% in testing. SPP launched a Consolidated Planning Process in March 2026. CAISO’s Cluster 16 opens in October 2026, as does ISO-NE’s first cluster. NYISO’s first cluster application window opens July 31, 2026.
| ISO/RTO | Status (as of Queued Up 2026 Edition) | Key Reform |
|---|---|---|
| CAISO | Cluster 15 closed; Cluster 16 opens Oct 2026 | AI-assisted study automation (GridUnity) |
| ERCOT | Ongoing (not FERC-jurisdictional) | “Connect and manage” model; record IA volumes in 2025 |
| ISO-NE | Transition to cluster; first cluster Oct 2026 | Order 2023 compliance; GridUnity partnership |
| MISO | Cluster in effect; queue cap active | SUGAR study automation (98% time reduction in pilot); ERAS fast-track |
| NYISO | First cluster July 31, 2026 window | Order 2023 compliance |
| PJM | Cycle-based cluster; working through legacy backlog | Google Tapestry automation; Reliability Resource Initiative fast-track |
| SPP | CPP launched March 2026 | Hitachi AI solution; GridUnity automation; ERAS fast-track |
| Non-ISO | Transition ongoing | Order 2023 compliance |
Source: LBNL, Queued Up 2026 Edition (emp.lbl.gov/queues); FERC, “Energized for 2026,” January 14, 2026 (ferc.gov).
PJM deserves the longest parenthetical. It is the largest U.S. electricity grid, serving 13 states and Washington, D.C. and more than 65 million people. After pausing new interconnection requests for several years to work through its backlog, PJM reopened its queue in 2026 and expects to begin processing new requests under its reformed cluster cycle with projected wait times of one to two years once fully operational. That timeline, if achieved, would represent a structural improvement from the multi-year medians of the serial process. PJM’s backlog exceeded 300 GW during the reform transition.
Whether the reforms actually close the gap between study speed and queue volume is the critical question for 2027 and beyond. FERC’s Order 2023 also introduced stricter financial readiness requirements, higher at-risk deposits, and withdrawal penalties — changes designed to reduce speculative entries that inflate queue volume without delivering real capacity. LBNL attributes part of 2025’s withdrawal surge to these deterrents flushing out projects that had entered under the old rules with limited commercial readiness.
What Percentage of the Queue Has Reached Commercial Operation by ISO?
ERCOT is the outlier — in a useful direction. Its completion rate from the 2000–2020 cohort exceeds 20% by capacity, the only ISO to cross that threshold. ISO-NE ranks second by project count (25%+). CAISO (8%) and the non-ISO West (10%) are the lowest. These differences reflect market structure: ERCOT’s “connect and manage” approach assigns congestion costs to generators rather than requiring cost certainty before connection, enabling faster throughput.
The PJM 2026–2027 capacity auction hit the FERC-imposed price cap of $329 per MW-day across its entire footprint, according to analysis commissioned by GridLab and executed by Aurora Energy Research. That price cap being binding across PJM reflects acute supply-side pressure — the direct consequence of slow interconnection. The same Aurora analysis estimated that if just 10% of the 107 GW of land-based renewables in PJM’s pre-2024 queue had come online in time for the 2026–2027 auction, consumers would have saved approximately $3.5 billion.
Grid Interconnection Queue: The Data Center and AI Dimension
The interconnection queue statistics above describe generation-side interconnection — power plants seeking to supply electricity. Data centers represent the load side: large power consumers seeking to connect directly to the transmission grid rather than through distribution systems. These are tracked separately.
The Department of Energy directed FERC in late 2025 to initiate rulemaking on federal load interconnection standards, covering facilities above 20 MW — effectively all utility-scale data center campuses, per the Rocky Mountain Institute. Load interconnection has historically not been federally standardized, unlike generator interconnection, which has operated under FERC’s pro forma procedures since the early 2000s. DOE’s directive signals a recognition that the two queues — generation and load — cannot be solved independently when data centers are consuming power at a rate that outpaces new generation delivery.
A single AI hyperscale campus drawing 500 MW or more requires a dedicated high-voltage transmission interconnection, which inherits much of the same study process and timeline as a generator interconnection request. The constraint is the same transmission infrastructure. The IEA notes in Electricity 2026 that planning, permitting, and completing new grid infrastructure can take 5 to 15 years, while data centers take 1 to 3 years to build — making the grid, not the building or the hardware, the binding timeline constraint for AI infrastructure deployment.
For analysis of data center electricity demand growth driving this load surge, see Axis Intelligence Research’s AI Data Center Statistics 2026 and Data Center Grid Impact Statistics.
Global Grid Connection Queues: How the U.S. Compares
The U.S. queue is the largest and most precisely measured in the world, but it is not unique. The International Energy Agency’s Electricity 2026 report identifies grid connection queues as having “reached record levels worldwide,” with over 2,500 GW of renewable, large-load, and storage projects stalled globally.
Europe’s queue is the second-most documented. The IEA’s 2024 data cited in LBNL’s Joule paper (Gorman et al., 2024) counted 596 GW of wind and solar capacity in queues across the UK, Italy, Spain, France, and Germany — equal to roughly twice their combined installed capacity in those resource categories at the time. That queue has grown materially since.
The IEA’s modeling in Electricity 2026 concludes that grid-enhancing technologies (dynamic line rating, advanced power flow control, topology optimization, storage as a transmission asset) and non-firm connection agreements could collectively unlock sufficient capacity to connect between 1,200 and 1,600 GW of advanced-stage projects currently stuck in global queues — without waiting for the 7+ years required to build new high-voltage lines. Dynamic line rating alone could increase existing line capacity by 20–30% at costs that are a fraction of new construction.
Those unlocking pathways exist in theory. Their adoption requires regulatory frameworks that permit non-firm connections and that incentivize system operators to deploy grid-enhancing technologies in operational planning — a change that is happening, but slowly.
What Happens to Projects That Withdraw From the Queue?
Three-quarters of all capacity that entered the U.S. interconnection queue between 2000 and 2020 has been withdrawn, per LBNL. The mechanics of withdrawal vary by stage. Most withdrawals in 2025 occurred during the system impact study phase, before the project reached the later and more expensive interconnection agreement stage, per LBNL’s phase-of-withdrawal analysis.
Late-stage withdrawals — after an interconnection agreement is signed — are more costly and more disruptive. A project withdrawing after signing takes sunk deposits with it, and can trigger mandatory re-studies for other projects in the cluster whose cost allocations depended on the withdrawing project’s network upgrade cost share. LBNL found that 41% of all capacity that signed interconnection agreements between 2000 and 2022 had subsequently withdrawn by end of 2025.
The average duration from interconnection request to withdrawal has itself been increasing — the queue is taking longer to reject projects as well as longer to approve them. Median withdrawal duration reached new highs in 2024, though it declined slightly in 2025 as reforms motivated older requests to exit faster rather than continue cycling through re-studies.
Interconnection Queue Costs: The Invisible Driver
LBNL maintains a separate dataset on interconnection upgrade costs. The Queued Up report itself notes that “limited available transmission capacity has resulted in high interconnection upgrade costs, which contributes to high withdrawals and may deter developers from submitting new requests.”
Interconnection costs — the network upgrade expenses assigned to project developers when their project requires transmission infrastructure improvements — are a function of where on the grid a project seeks to connect. High-congestion points generate high cost allocations. When those allocations exceed roughly 10% of a project’s total capital expenditure, economics typically require withdrawal.
This is the structural mechanism behind the 75% withdrawal rate. It is not that developers cannot build projects in principle. It is that the grid’s existing transmission capacity, built for a different generation mix and a different load profile, cannot accommodate the volume of applications without requiring network upgrades whose costs, assigned to individual project developers, regularly destroy project economics.
FERC Order 2023 requires grid operators to evaluate grid-enhancing technologies as alternatives to traditional upgrades before assigning full upgrade costs to projects. That provision, if implemented consistently, should reduce cost allocation for some projects and lower the withdrawal rate at the margin. Whether it does so at scale will determine whether the 13% completion rate has a structural floor or a structural ceiling.
Methodology
Data collection. All statistics in this article were retrieved from primary sources during the production session of July 29, 2026. The primary dataset is Lawrence Berkeley National Laboratory’s Queued Up 2026 Edition (PDF and Excel data file), published June 2026 and covering interconnection queue data through December 31, 2025. Data were compiled by LBNL and Interconnection.fyi from 7 ISOs/RTOs and 50 non-ISO balancing areas representing approximately 98% of U.S. installed generating capacity. Secondary primary sources include FERC’s “Energized for 2026” statement (January 14, 2026) and the IEA’s Electricity 2026 Grids chapter.
GQBWR™ calculation. Grid Queue Build-vs-Withdraw Ratio = GW withdrawn in calendar year ÷ GW reaching commercial operation in calendar year. Inputs for 2025: withdrawn capacity >750 GW (LBNL lower bound; the exact figure is estimated at approximately 750 GW using LBNL’s stated inflow-outflow method from Slide 12 of the 2026 edition, as exact withdrawn-year data was available for only 63% of withdrawn requests); capacity reaching commercial operation approximately 53 GW (LBNL Slide 12, stating an annual throughput rate of 2% from the 2024 active base of 2,290 GW). Formula verified by Python on July 29, 2026: 750 ÷ 53 = 14.2×. This is the baseline reading; historical GQBWR values will be calculated in a future edition.
Scope. LBNL’s dataset includes only transmission-connected generation and storage interconnection requests — it excludes load interconnection (data centers and industrial facilities seeking direct grid connection), distribution-connected generation, and behind-the-meter projects.
Limitations. Interconnection agreement dates are unknown for approximately 45% of records in the LBNL dataset; withdrawn year is unknown for approximately 37% of withdrawn requests; commercial operations dates are valid for approximately 73% of operational projects. These gaps are structural to the data and are documented by LBNL. Wait time statistics cited here represent projects for which valid dates were available and should be treated as indicative, not exhaustive. Completion rates for more recent cohorts (post-2020) are understated because many projects take more than five years to complete.
About This Dataset
Coverage: U.S. grid interconnection queue data, 2000–2025, with point-in-time statistics as of December 31, 2025. Includes GQBWR™ baseline for 2025.
Primary underlying source: Lawrence Berkeley National Laboratory, Queued Up 2026 Edition, June 2026.
CSV download: CC BY 4.0. Cite as: “Axis Intelligence Research, Grid Interconnection Queue Statistics 2026, axis-intelligence.com/grid-interconnection-queue-statistics/.” Hugging Face / Kaggle / GitHub: Upload recommended following publication.
Frequently Asked Questions
What is an interconnection queue?
An interconnection queue is a list of proposed power plants — and increasingly, large electricity consumers like data centers — that have applied to connect to the transmission grid and are awaiting the completion of engineering studies that determine what grid upgrades are required and what they will cost. Transmission operators (ISOs, RTOs, and utilities) require these studies before any new connection can be finalized. The queue is public data; LBNL compiles it annually across more than 57 transmission providers.
Why is the U.S. interconnection queue so large?
The U.S. queue grew from under 200 GW in 2010 to a peak of roughly 2,600 GW in 2023, driven by a surge in solar, storage, and wind applications following federal tax credit expansions and the Inflation Reduction Act’s passage in 2022. The grid’s transmission infrastructure was built for a smaller, less distributed generation fleet, so applications now routinely exceed available connection capacity at any given point of interconnection — triggering cost allocation studies that can make projects economically unviable and contributing to the high withdrawal rates.
How long does it take to get grid interconnection approved?
The median project that completed the interconnection process and reached commercial operation in 2025 spent 61 months — just over five years — from its initial request to operation, per LBNL’s 2026 Queued Up report. The study phase alone (from request to signed interconnection agreement) took a median of 45 months. Smaller projects (under 5 MW) average around 12 months to agreement; large projects (over 200 MW) average nearly five years from request to commercial operation.
What percentage of interconnection applications are approved and built?
Only 13% of capacity that submitted interconnection requests from 2000 to 2020 had reached commercial operations by the end of 2025, per LBNL. By project count the figure is 19%, reflecting that smaller projects complete at slightly higher rates than large ones. The remaining 75% of capacity (by GW) was withdrawn — meaning it never generated power.
What is FERC Order 2023 and what does it change?
FERC Order No. 2023, issued July 28, 2023, is the largest overhaul of the generator interconnection process since FERC established its pro forma interconnection procedures in the early 2000s. It requires all transmission providers to transition from a serial (first-come, first-served) study approach to a cluster-based (“first-ready, first-served”) process, where groups of projects are studied simultaneously. It also introduces stricter financial readiness requirements, higher deposits, and withdrawal penalties designed to remove speculative applications. FERC is also fast-tracking permitting for 50+ GW of shovel-ready projects. Full compliance is underway across ISOs; material impact on wait times is not yet measurable in the 2025 data.
How does the U.S. interconnection queue compare to other countries?
The U.S. holds the largest and most well-documented national interconnection queue at 2,061 GW as of end-2025. Globally, the IEA counts over 2,500 GW of renewable, large-load, and storage projects stalled in connection queues worldwide. Europe’s queues — particularly across the UK, Italy, Spain, France, and Germany — collectively held approximately 596 GW of wind and solar capacity in 2024, roughly double those countries’ installed capacity in those resource types, per LBNL’s Joule paper (Gorman et al., 2024, doi:10.1016/j.joule.2024.11.008).
Why is natural gas capacity in the queue growing while renewables are declining?
Natural gas active queue capacity rose 86% year-over-year to 253 GW in 2025. Gas projects benefit from faster permitting in some regions, more predictable revenue under utility contracts, and — critically — ERCOT’s “connect and manage” interconnection approach, which allows generators to reach commercial operation faster than the cost-allocation-intensive FERC LGIP process. Rising load forecasts from AI data centers and electrification are also creating market signals that favor dispatchable generation, which gas can provide. Solar and storage declined due to queue pauses at CAISO, PJM’s transition backlog, and tax credit uncertainty affecting project economics.
What is the GQBWR and how should I interpret it?
The Grid Queue Build-vs-Withdraw Ratio (GQBWR™) is an Axis Intelligence Research proprietary metric that divides the gigawatts withdrawn from the U.S. interconnection queue in a given year by the gigawatts that reached commercial operation. A ratio of 14.2× in 2025 means that for every 1 GW built, approximately 14 GW gave up and withdrew. The metric captures queue dysfunction more concisely than either queue size or wait time alone, because it reflects the balance between successful delivery and failed attempts. A ratio converging toward 1× would indicate a functioning process; above 10× indicates structural attrition dominates delivery. See the Methodology section for formula details and source citations.
Citation Block
APA: Axis Intelligence Research & Jad, A. (2026, July 29). Grid Interconnection Queue Statistics 2026: 2,061 GW Stuck, 14× More Withdrawn Than Built. Axis Intelligence Research. https://axis-intelligence.com/grid-interconnection-queue-statistics/
MLA: Axis Intelligence Research and Aidan Jad. “Grid Interconnection Queue Statistics 2026: 2,061 GW Stuck, 14× More Withdrawn Than Built.” Axis Intelligence Research, 29 July 2026, axis-intelligence.com/grid-interconnection-queue-statistics/.
Chicago: Axis Intelligence Research and Aidan Jad. “Grid Interconnection Queue Statistics 2026: 2,061 GW Stuck, 14× More Withdrawn Than Built.” Axis Intelligence Research, July 29, 2026. https://axis-intelligence.com/grid-interconnection-queue-statistics/.
Related Research
- AI Data Center Carbon Footprint Statistics 2026 — The HCDI™ index, hyperscaler emissions disclosures, and Scope 1/2/3 analysis
- AI Data Center Energy Consumption Statistics 2026 — Full TWh breakdowns, grid mix, and the AEII™ index
- AI Data Center Water Usage Statistics 2026 — Cooling technology, water consumption, and regional water stress
- Data Center Battery Storage Statistics 2026 — BESS market, chemistry breakdown, and the IEA 20-25 GW battery projection
- Data Center Electricity Cost by State 2026 — DCECI™ index, all 50 states ranked
- Nuclear Energy for Data Centers
- Big Tech Clean Energy