The Queue Is Not the Grid

Question

What can the end-of-2025 U.S. interconnection queues responsibly tell an executive, planner, or researcher about prospective electric generation and storage, and what cannot be concluded from the headline gigawatt total?

Decision context

This demonstration can support:

• public-source market and infrastructure research;
• identification of project-maturity signals;
• comparison of queue composition by technology and region;
• tracking of study, agreement, withdrawal, and commercial-operation states;
• monitoring of FERC Order No. 2023 implementation;
• preparation of questions for utilities, developers, engineers, investors, and regulators.

It does not forecast which projects will operate, determine grid reliability, value a development pipeline, approve an interconnection, or recommend investment.

Entity model

• Project: generation, storage, or hybrid request
• Technology: solar, storage, wind, gas, nuclear, geothermal, other
• Transmission provider: ISO, RTO, or non-ISO utility
• Queue state: active, suspended, withdrawn, operational
• Process milestone: interconnection request, study, draft agreement, executed agreement, commercial operation
• Dates: request, agreement, proposed online date, withdrawal, commercial operation
• Capacity: proposed nameplate generation and storage capacity
• Related requirements: site control, deposits, study costs, network upgrades, permits, power purchaser, equipment, financing
• Policy object: FERC Order No. 2023, Order No. 2023-A, regional compliance filings and process reforms

Event chronology

1. July 28, 2023 FERC issued Order No. 2023, establishing major reforms that include first-ready, first-served cluster studies, readiness requirements, firm study deadlines, penalties for certain delays, and updated treatment of new technologies. ENG-S04ENG-S05

1. March 21, 2024 FERC issued Order No. 2023-A, clarifying and modifying parts of the final rule while preserving its overall reform direction. ENG-S06

1. End of 2025 data horizon Berkeley Lab’s compiled queue dataset contained approximately 8,200 active projects representing 2,061 GW of generation and storage. ENG-S01ENG-S02

1. May and June 2026 Berkeley Lab updated the underlying data file in May and published the 2026 Edition slide deck in June, analyzing data through the end of 2025. ENG-S01

1. July 1, 2026 Berkeley Lab published a findings article stating that active queue capacity declined 10% from 2024 to 2025, more than 750 GW withdrew in 2025, and 536 GW of active capacity had a draft or executed interconnection agreement. ENG-S02

1. November 1, 2026 evidence refresh Berkeley Lab’s current Queued Up landing page states 549 GW with a draft or executed agreement but not yet in commercial operation. The July findings page states 536 GW of active capacity with a draft or executed agreement. Both official figures are preserved because their scopes or refresh states are not identical on the face of the pages. ENG-S01ENG-S02

Source register

Claim register

Observed

ENG-O-001 At the end of 2025, approximately 8,200 active projects represented 1,312 GW of generation and approximately 749 GW of storage, totaling 2,061 GW. ENG-S01ENG-S02

ENG-O-002 Active queue volume declined approximately 10% from 2024 to 2025 amid high withdrawals and fewer new requests. ENG-S01ENG-S02

ENG-O-003 Active natural-gas capacity rose to 253 GW, an 86% increase, while active solar, storage, and wind capacity declined year over year to 773 GW, 749 GW, and 220 GW respectively. ENG-S01ENG-S02

ENG-O-004 For the historical cohort of requests submitted from 2000 through 2020, 13% of requested capacity had reached commercial operation by the end of 2025, 75% had withdrawn, and 10% remained active. ENG-S01ENG-S02

ENG-O-005 For regions with available data, the median time from interconnection request to commercial operation exceeded five years for projects built in 2025. ENG-S01ENG-S02

ENG-O-006 Berkeley Lab’s current landing page reports 549 GW with draft or executed interconnection agreements but not yet in commercial operation. Its July 1 findings article reports 536 GW of active capacity with a draft or executed agreement. ENG-S01ENG-S02

ENG-O-007 Berkeley Lab cautions that most proposed capacity will not be built and describes queue data as a general indicator for future electric capacity additions. This test does not treat that as a forecast. ENG-S01

ENG-O-008 FERC Order No. 2023 replaced serial first-come, first-served study architecture with a first-ready, first-served cluster approach and added readiness, deadline, penalty, and technology reforms. Order No. 2023-A later clarified parts of that framework. ENG-S04ENG-S05ENG-S06

Corroborated

ENG-C-001 The active queue is a pipeline of requests, not a committed build schedule. Both the current dataset and the historical outcome record show that most requested capacity does not reach operation. ENG-S01ENG-S02

ENG-C-002 A draft or executed interconnection agreement is a meaningful maturity signal, but it does not establish commercial operation. Berkeley Lab reports substantial post-agreement withdrawals and identifies other dependencies such as permits, land, buyers, equipment, financing, and upgrades. ENG-S01ENG-S02

ENG-C-003 The 549 GW and 536 GW values are both present in official Berkeley Lab materials. A responsible record must preserve the source, publication date, wording, and apparent scope rather than silently choosing one. ENG-S01ENG-S02ENG-S03

Derived

ENG-D-001 The minimum useful project record is not name plus megawatts. It requires technology, region, queue state, study phase, agreement state, dates, proposed online date, network-upgrade context, and known non-interconnection dependencies.

ENG-D-002 Headline capacity should be segmented into at least active, suspended, withdrawn, and operational states, with agreement maturity shown separately.

ENG-D-003 The historical 13% completion rate is a cohort outcome, not a conversion factor that can be mechanically applied to every current project, region, or technology.

ENG-D-004 The effect of Order No. 2023 cannot be established solely from the 2025 decrease in queue volume because reform implementation, regional process changes, developer behavior, and market conditions moved together.

Inferred

ENG-I-001 Executives evaluating prospective power supply should ask what stage and evidence sit behind a megawatt claim before using it in a demand, reliability, construction, or financing narrative.

ENG-I-002 A credible monitoring system should separate policy adoption, regional compliance, operational implementation, and measurable outcome. Those are different events.

These are Delta Arc systems inferences. They are not engineering, investment, procurement, or regulatory recommendations.

Contradictions, limits, and unknowns

• Berkeley Lab’s landing page gives 549 GW, while the July 1 findings page gives 536 GW, for closely related agreement-stage concepts. The package does not invent a reconciliation. It preserves both source-specific claims and requires a dataset-level refresh before public quantitative comparison.
• Agreement-status information is unknown for a material share of records. The Queued Up deck identifies missing agreement or study-phase status for approximately 39% of records.
• Interconnection-agreement dates are incomplete for many records, limiting duration analysis.
• Queue nameplate capacity is not necessarily deliverable capacity during peak or critical hours.
• Storage and hybrid capacity can involve imputation or component overlap, so totals require the source methodology.
• A proposed online date originates with a developer and can change.
• Regions differ in market design, queue rules, data quality, withdrawal meaning, upgrade allocation, and fast-track programs.
• A queue record does not establish land rights, permits, offtake, equipment, financing, fuel, construction, or commercial-operation readiness.
• The full effect of FERC Order No. 2023 remains too early to measure definitively in Berkeley Lab’s 2026 assessment.

Safe Delta Arc outputs

• Versioned queue evidence map
• Project-maturity and agreement-state dashboard
• Technology and regional composition brief
• Withdrawal and duration chronology
• Order No. 2023 implementation tracker
• Source-discrepancy register
• Executive question set
• Qualified-review briefing package

Prohibited outputs

• Forecast that queued megawatts will be built
• Grid reliability or resource-adequacy determination
• Interconnection approval or engineering study
• Investment, acquisition, lending, or securities advice
• Project or developer valuation
• Procurement or siting decision
• Network-upgrade cost estimate
• Claim that an agreement guarantees operation
• Mechanical application of the 13% historical outcome rate to a current portfolio
• Silent substitution of 536 GW for 549 GW, or the reverse

Changes made at release

• A browser check on November 2, 2026 found that Berkeley Lab does not use the words "not a forecast". The claim now follows its Queued Up page, and that framing is identified as this test's own.