Process
Grid interconnection queues
The administrative line to be connected to the electricity grid. In several markets it is the slowest step in building anything.
Searching is free and needs no AI. Summaries use your own AI model.
01Key numbers
What the text on this page claims, and the number behind it
a multi-year administrative queue between a signed site and a live megawatt
MeasuredLBNL: the median US project built in 2025 took 61 months from interconnection request to operation.
Typical/median time from interconnection request to commercial operation (projects built that year): Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition)Open the source → months, 2025Active generation + storage capacity in US interconnection queues: Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition)Open the source → GW, 31 Dec 2025
Lead time · USTypical/median time from interconnection request to commercial operation (projects built that year)Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (2025)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 2025Open the source → months202555 (2024) → 61 (2025).How this is computedThe latest value in a series minus the value before it, divided by the earlier value, same series and unit.55 (2024) → 61 (2025).Method →Code → vs 2024 · tightens
Typical/median time from interconnection request to commercial operation (projects built that year), months, 2008 to 2025. Every value is listed in the table below. 2040608022 months2008Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗36 months2015Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗55 months2024Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) ↗61 months2025Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗Period Value Source 2025 Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (2026-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 2026-06Open the source → months Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗ primary The median project built in 2025 took 61 months from the interconnection request to commercial operations1, compared to 36 months in 2015 and 22 months in 2008
2024 Source: Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) (2025-12-15)SourcedLawrence Berkeley National Laboratory (Queued Up, 2025 edition) · 2025-12-15Open the source → months Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) ↗ primary The typical project built in 2024 took 55 months from the interconnection request to commercial operations1, compared to 36 months in 2015 and 22 months in 2008
2015 Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (2026-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 2026-06Open the source → months Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗ primary The median project built in 2025 took 61 months from the interconnection request to commercial operations1, compared to 36 months in 2015 and 22 months in 2008
2008 Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (2026-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 2026-06Open the source → months Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗ primary The median project built in 2025 took 61 months from the interconnection request to commercial operations1, compared to 36 months in 2015 and 22 months in 2008
Read from the sources Read from the source by an agent; quote matched mechanically against the fetched page; not yet reviewed by a person, 2026-09-24. Each value links the page it came from and quotes the sentence that carries it.
Queue · USActive generation + storage capacity in US interconnection queuesSource: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (31 Dec 2025)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 31 Dec 2025Open the source → GW31 Dec 20252,290 (31 Dec 2024) → 2,061 (31 Dec 2025).How this is computedThe latest value in a series minus the value before it, divided by the earlier value, same series and unit.2,290 (31 Dec 2024) → 2,061 (31 Dec 2025).Method →Code → vs 31 Dec 2024 · loosens
Active generation + storage capacity in US interconnection queues, GW, 31 Dec 2022 to 31 Dec 2025. Every value is listed in the table below. 2,0002,2002,4002,6002,040 GW31 Dec 2022Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) ↗2,600 GW31 Dec 2023Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) ↗2,290 GW31 Dec 2024Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) ↗2,061 GW31 Dec 2025Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗Period Value Source 31 Dec 2025 Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (2026-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 2026-06Open the source → GW Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗ primary Total (cumulative) active capacity in queues decreased 10% in 2025 to 2,061 GW
31 Dec 2024 Source: Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) (2025-12-15)SourcedLawrence Berkeley National Laboratory (Queued Up, 2025 edition) · 2025-12-15Open the source → GW Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) ↗ primary Total (cumulative) active capacity in queues decreased 12% in 2024, to 2,290 GW
31 Dec 2023 Source: Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) (2024-04)SourcedLawrence Berkeley National Laboratory (Queued Up, 2024 edition) · 2024-04Open the source → GW Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) ↗ primary Total (cumulative) active capacity in queues is now nearly 2,600 GW (2.6 TW)
"nearly 2,600 GW".31 Dec 2022 Source: Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) (2023-04-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2023 edition) · 2023-04-06Open the source → GW Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) ↗ primary Active capacity in queues (~2,040 GW) exceeds installed capacity of entire U.S. power plant fleet (~1,250 GW)
Approximate ("~"); headline elsewhere says "Over 2,000 GW".Read from the sources Read from the source by an agent; quote matched mechanically against the fetched page; not yet reviewed by a person, 2026-09-24. Each value links the page it came from and quotes the sentence that carries it.
Queue · USActive natural gas capacity in US interconnection queuesSource: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (31 Dec 2025)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 31 Dec 2025Open the source → GW31 Dec 2025136 (31 Dec 2024) → 253 (31 Dec 2025).How this is computedThe latest value in a series minus the value before it, divided by the earlier value, same series and unit.136 (31 Dec 2024) → 253 (31 Dec 2025).Method →Code → vs 31 Dec 2024 · tightens
Active natural gas capacity in US interconnection queues, GW, 31 Dec 2022 to 31 Dec 2025. Every value is listed in the table below. 010020030082 GW31 Dec 2022Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) ↗79 GW31 Dec 2023Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) ↗136 GW31 Dec 2024Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) ↗253 GW31 Dec 2025Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗Period Value Source 31 Dec 2025 Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (2026-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 2026-06Open the source → GW Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗ primary Active natural gas capacity (253 GW, +86%) increased in 2025, while solar (773 GW, -19%), storage (749 GW, -16%) and wind (220 GW, -19%) capacity decreased.
31 Dec 2024 Source: Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) (2025-12-15)SourcedLawrence Berkeley National Laboratory (Queued Up, 2025 edition) · 2025-12-15Open the source → GW Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) ↗ primary Gas (136 GW) capacity increased by 72% year-over-year, whereas Solar (956 GW, -12%), Storage (890 GW, -13%), and Wind (271 GW, -26%) all decreased in 2024
31 Dec 2023 Source: Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) (2024-04)SourcedLawrence Berkeley National Laboratory (Queued Up, 2024 edition) · 2024-04Open the source → GW Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) ↗ primary Solar (1,086 GW) , Storage (1,028 GW), and Wind (366 GW) make up 95% of active capacity in queues, with 3% (79 GW) from Gas.
31 Dec 2022 Source: Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) (2023-04-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2023 edition) · 2023-04-06Open the source → GW Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) ↗ primary Only 82 GW of gas capacity active in the queues, less than 10% of active solar capacity
Read from the sources Read from the source by an agent; quote matched mechanically against the fetched page; not yet reviewed by a person, 2026-09-24. Each value links the page it came from and quotes the sentence that carries it.
Queue · USShare of queued capacity that reached commercial operation (request cohort)Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (31 Dec 2025)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 31 Dec 2025Open the source → %31 Dec 202513 (31 Dec 2024) → 13 (31 Dec 2025).How this is computedThe latest value in a series minus the value before it, divided by the earlier value, same series and unit.13 (31 Dec 2024) → 13 (31 Dec 2025).Method →Code → vs 31 Dec 2024 · flat
Share of queued capacity that reached commercial operation (request cohort), %, 31 Dec 2022 to 31 Dec 2025. Every value is listed in the table below. 1313.51414 %31 Dec 2022Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) ↗14 %31 Dec 2023Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) ↗13 %31 Dec 2024Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) ↗13 %31 Dec 2025Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗Period Value Source 31 Dec 2025 Source: Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) (2026-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2026 edition) · 2026-06Open the source → % Lawrence Berkeley National Laboratory (Queued Up, 2026 edition) ↗ primary About 19% of projects (13% of capacity) requesting interconnection from 2000-2020 reached commercial operations by the end of 2025
Cohort: requests 2000-2020; 19% of projects.31 Dec 2024 Source: Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) (2025-12-15)SourcedLawrence Berkeley National Laboratory (Queued Up, 2025 edition) · 2025-12-15Open the source → % Lawrence Berkeley National Laboratory (Queued Up, 2025 edition) ↗ primary About 19% of projects (13% of capacity) requesting interconnection from 2000-2019 reached commercial operations by the end of 2024
Cohort: requests 2000-2019; 19% of projects.31 Dec 2023 Source: Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) (2024-04)SourcedLawrence Berkeley National Laboratory (Queued Up, 2024 edition) · 2024-04Open the source → % Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) ↗ primary Only ~20% of projects (14% of capacity) requesting interconnection from 2000-2018 reached commercial operations by the end of 2023
Cohort: requests 2000-2018; ~20% of projects.31 Dec 2022 Source: Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) (2023-04-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2023 edition) · 2023-04-06Open the source → % Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) ↗ primary Only ~21% of projects (14% of capacity) requesting interconnection from 2000-2017 reached commercial operations by the end of 2022
Cohort: requests 2000-2017; ~21% of projects.Read from the sources Read from the source by an agent; quote matched mechanically against the fetched page; not yet reviewed by a person, 2026-09-24. Each value links the page it came from and quotes the sentence that carries it.
Lead time · USTypical/median time from interconnection request to commercial operation (projects built that year)Source: Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) (2023)SourcedLawrence Berkeley National Laboratory (Queued Up, 2024 edition) · 2023Open the source → years20235 (2022) → 5 (2023).How this is computedThe latest value in a series minus the value before it, divided by the earlier value, same series and unit.5 (2022) → 5 (2023).Method →Code → vs 2022 · flat
Typical/median time from interconnection request to commercial operation (projects built that year), years, 2022 to 2023. Every value is listed in the table below. Period Value Source 2023 Source: Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) (2024-04)SourcedLawrence Berkeley National Laboratory (Queued Up, 2024 edition) · 2024-04Open the source → years Lawrence Berkeley National Laboratory (Queued Up, 2024 edition) ↗ primary The typical project built in 2023 took nearly 5 years from the interconnection request to commercial operations1, compared to 3 years in 2015 and <2 years in 2008.
"nearly 5 years".2022 Source: Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) (2023-04-06)SourcedLawrence Berkeley National Laboratory (Queued Up, 2023 edition) · 2023-04-06Open the source → years Lawrence Berkeley National Laboratory (Queued Up, 2023 edition) ↗ primary The typical project built in 2022 took 5 years from the interconnection request to commercial operations1, compared to 3 years in 2015 and <2 years in 2008.
Stated in years; later editions state months.Read from the sources Read from the source by an agent; quote matched mechanically against the fetched page; not yet reviewed by a person, 2026-09-24. Each value links the page it came from and quotes the sentence that carries it.
Every series on every bottleneck — filter, compare and download.
02Why it holds things up
Not a material, not a machine, and frequently the binding constraint: a multi-year administrative queue between a signed site and a live megawatt.
Part of the process · Permission, which takes years to a decade to loosen once somebody decides to.
03Severity
How this is computedSum of four judged scores, each 0–3: concentration + substitution + lead time + inelasticity.Method →Code → / 12
An administrative queue measured in years between a signed site and a live megawatt. Relief is a rule change, not a factory.
04What it rests on, and what rests on it
Depends on
- High-voltage cable and switchgear
“May require modest upgrades (breakers) or reconstruction of several high-voltage transmission lines.”
eta-publications.lbl.gov · primary source · LBNL (Feb 2026), on network upgrades a new generator triggers. A cost item in the queue, not a stated supply constraint.
Depended on by
- Microsoft
“Limitations in power availability, delays in obtaining power connections, outages, shortages, increased energy costs, or requirements imposed by utilities, regulators, or other market participants could restrict our ability to develop or expand datacenter capacity.”
sec.gov · primary source · Fiscal 2026 10-K risk factor. Names delays in power connections, not the interconnection queue by name.
- Google
“We have experienced and may in the future experience supply shortages, price increases, quality issues, or longer lead times that could harm our operations, driven by raw material or component availability, manufacturing capacity, labor shortages, industry allocations, logistics capacity, inflation, foreign currency exchange rates, tariffs, sanctions and export controls, trade disputes and barriers, forced labor concerns, sourcing requirements, geopolitical tensions, armed conflicts, natural disasters or pandemics, the effects of climate change, power and transmission availability, and significant changes in the financial or business condition of our suppliers.”
sec.gov · primary source · Alphabet’s 2025 10-K. "Power and transmission availability" is one cause in a long list.
- CoreWeave
“Limitations on generation, transmission, and distribution may limit our ability to obtain sufficient power capacity for our potential expansion sites in new or existing markets.”
sec.gov · primary source · Power capacity for new sites; does not say "queue".
- Anthropic
“The U.S. can enable the buildout of large-scale infrastructure for AI training in strategic locations, including by accelerating permitting, supporting targeted transmission projects, and collaborating with utilities to speed up grid interconnection processes.”
anthropic.com · primary source · A policy recommendation for US AI infrastructure, not a statement about Anthropic’s own sites.
A recorded dependency is a sentence, not a share, a volume or a contract. Walk it on a portfolio →
05Rules that govern it
Order No. 2023: improvements to generator interconnection procedures and agreements
Rewrites how new generation joins the US grid: first-ready first-served clusters, deadlines and penalties, aimed squarely at the queue.
06What would remove it
- Small modular reactorsEstimate by Substrata, open to challenge, 2026-09-15: Designs are licensed or in licensing and first units are under construction; none is yet delivering power at scale in a Western market.Estimate, not a measurementDesigns are licensed or in licensing and first units are under construction; none is yet delivering power at scale in a Western market.By Substrata, open to challenge · 2026-09-15Supporting source → · Prototype at meaningful scale
Sited next to the load, it can skip the transmission queue that a distant plant cannot.
07Science pipeline
judged, with a citation claimed, no citationNumber: papers, preprints and grants the feeds collected — unreviewed
- Integration of Distributed Solar Energy into Saudi Smart Distribution Grids: Voltage Regulation, Hosting Capacity and Protection Challenges
The deployment of distributed solar photovoltaic generation is leading Saudi Arabia's distribution networks to move from a primarily one-way arrangement to a more actively managed system in which power, voltage and fault current can change direction at various stages of…
Fundamental research · No laboratory or pilot wording.From OpenAlex · not yet reviewedDOI
- Using cross-regional grid interconnection to relieve growing energy storage demand in renewable energy integration: A case study from China
With the rapid growth of wind and solar generation, power systems face increasing balancing and energy storage needs. Cross-regional interconnection provides spatial flexibility through tie-line power exchange, yet the implications of tie-line scheduling flexibility for capacity…
Zhejiang UniversityNARI Group (China)
Fundamental research · No laboratory or pilot wording.From OpenAlex · not yet reviewedDOI
- Energy infrastructure for AI data centers: A critical review of grid integration, techno-economic challenges, and future pathways
Artificial intelligence (AI) workloads are rapidly changing the energy landscape of today's data centers, both in terms of total electricity usage and power density at the facility level. The ongoing deployment of large-scale AI training and inference infrastructure is further…
Fundamental research · No laboratory or pilot wording.From OpenAlex · not yet reviewedDOI
- Coordinated integration of renewables, STATCOMs, and V2G for hosting capacity improvement under uncertainty
Abstract Although incorporating renewable energy sources (RESs) into electric distribution networks (EDNs) offers significant sustainability and environmental benefits, improper management of RESs may threaten the technical performance of the network. To assess the safe…
Fundamental research · No laboratory or pilot wording.From OpenAlex · not yet reviewedDOIPDF
- DYNAMIC AND SECURITY-CONSTRAINED ASSESSMENT OF PHOTOVOLTAIC HOSTING CAPACITY IN A LOW-INERTIA POWER SYSTEM: IEEE 39-BUS VALIDATION AND EMPIRICAL APPLICATION TO SENEGAL
As phоtоvоltaiс (PV) penеtratiоn rises aсrоss interсоnnеcted grids, hоsting cаpасity (HC) саn nо lоnger be inferred sоlely frоm vоltagе mаrgins аnd еquipment-lоаding соnstraints. The соmbinеd effeсts оf rеducеd synchrоnоus inеrtia аnd sоlar variability requirе a unified treatment…
Fundamental research · No laboratory or pilot wording.From OpenAlex · not yet reviewedDOIalso at doi.org
08What we have predicted
- 2026-09-15open until 2029-03-31
Median US interconnection queue duration will not fall below three years by the end of 2028, even though FERC Order No. 2023 took effect in November 2023.
Settled by · The Lawrence Berkeley National Laboratory queued-projects series, or an equivalent published dataset, reporting median duration from request to agreement for projects completing in 2028.
09Who could fund relief
Is funding the constraint? · Money is not the constraint
An administrative queue. There is nothing to buy. Money can fund the transmission that shortens it eventually, but the queue itself is a process, and process is relieved by rules.
- Department of Energy Loan Programs OfficeDevelopment bank
Lends at scale to energy projects that commercial lenders will not finance alone. It now operates as the Office of Energy Dominance Financing, under the authorities the Energy Policy Act of 2005 gave it.
Development bank will not fund · Anything outside its mandate, however commercially sound. It is patient and cheap and narrow, in that order.
A mandate covering the kind of asset that would relieve this row. Not a claim that anyone has funded one, and not advice.
10What has happened
Nothing recorded about this one yet.
Events are added when a source is read and accepted by hand.
11What this holds up
- Frontier training runone turn: Weeks to months
Worst judgement on this loop: How this is computedThe highest severity score among the bottlenecks recorded as gates on the loop.Method →Code →, judged 2026-09-15
- Energyone turn: Months to years
Worst judgement on this loop: How this is computedThe highest severity score among the bottlenecks recorded as gates on the loop.Method →Code →, judged 2026-09-15
A loop is one design, build and measure turn. Severity is a dated judgement on a 0–12 scale, not a measurement, and the judgements are not added together — the figure shown is the worst single one.
Discussion
Comments use threadkit (permission is participation). Sign in with OrangeCat to write. Fact-check is the same assistant as Ask, grounded in the corpus.
- No comments yet.