US Sector Analysis: The Power Grid Interconnection Backlog
The queue was built for generators trying to sell power. It’s now clogged with buyers trying to secure it, and that swap is what actually…
US Sector Analysis: The Power Grid Interconnection Backlog
The queue was built for generators trying to sell power. It’s now clogged with buyers trying to secure it, and that swap is what actually decides who profits.

More than 2,600 gigawatts of generation and storage are sitting in line right now, waiting for permission to connect to the American grid. That’s larger than every power plant currently operating in the country, combined.
If the whole line got built tomorrow, US electricity capacity would more than double.
Almost none of it will get built on that timeline, and the reason isn’t the one most utility investors are pricing.
The comfortable version of this story says paperwork is slow but reform is coming: cluster studies, readiness screens, new federal rules.
That version isn’t false.
It’s just answering a different question than the one that actually determines who gets paid
The Line That Outgrew the System Built to Manage It

PJM interconnection wait time: under 2 years in 2008, over 8 years by 2025
In 2008, a project applying to connect to PJM, the grid operator running thirteen states from Illinois to New Jersey, could expect to start generating within two years.
By 2025, that wait had stretched past eight.
PJM’s fix, a new process called TC1, took 544 days just to clear its first round.
More than 100 gigawatts applied.
Forty made it through the initial study.
Seventeen actually got an agreement to connect.

PJM’s TC1 process: of 100+ GW that applied, only 17 GW reached a signed connection agreement, over 544 days.
That’s real progress against what came before it.
It’s also a year and a half for one round of paperwork, inside a system where the newest class of applicant wants power delivered in twelve.
The Applicants Changed, and the Queue Didn’t Notice
For most of its history, this queue processed people trying to sell electricity into the grid: solar, wind, and battery developers adding supply.
That’s flipped.

Gas filings up 72% year-over-year. Solar and storage filings declined
Natural gas filings jumped 72% in the most recent tracked year while solar and storage applications fell, and the people filing today are increasingly hyperscale technology companies trying to secure power for AI data centers, not developers trying to generate it for sale.
A system built to check whether new supply can be absorbed safely was never built to process a wave of demand arriving all at once, wanting gigawatt-scale delivery inside eighteen months.
PJM’s own forecast has summer peak demand rising 32 gigawatts between 2024 and 2030, with data centers responsible for 94% of that growth.
Pull data centers out of the forecast and PJM’s demand is flat, possibly shrinking.
One category of customer is now generating essentially all the growth on the country’s largest grid.
Texas shows the same shift with sharper numbers.
CenterPoint Energy’s large-load interconnection requests jumped from 1 gigawatt to 8 in a single year.

ERCOT’s statewide large-load queue hit 226 gigawatts by October 2025.
Only 1.8% of that was actually operational and drawing power.
(Texas eventually added a $100,000 screening fee just to slow the filings down, which is not the kind of fee a regulator invents unless developers were reserving spots in three states at once with no intention of building in more than one.)
Position Decides the Payout, Not the Sector
Here’s where a sector-wide “own the utilities” thesis stops matching the numbers.
PJM’s capacity auction, where power plants sell their availability years ahead of time, cleared at $2.2 billion in 2024.
The next one cleared at $14.7 billion.

PJM’s capacity auction: $2.2B (2024) to $14.7B (2025). 63% of the increase, $9.3B, is attributed to data center demand.
PJM’s own Independent Market Monitor attributed 63% of that jump, roughly $9.3 billion, to data center demand pushing against a supply base that couldn’t grow fast enough.
A utility only collects from that number if it has capacity actually cleared into the auction: built, connected, bid.
A utility still sitting in the queue earns nothing from this cycle.
It’s carrying the cost side of the equation without any of the revenue, and the cost side has gotten brutal on its own.
Network upgrade costs in PJM went from $29 per kilowatt between 2017 and 2019 to $240 per kilowatt between 2020 and 2022, an eightfold jump in three years.
Nearly 80% of projects that enter the queue eventually withdraw before finishing, and for those that do withdraw, interconnection costs alone can consume 30 to 37% of the total project budget, against 6 to 8% for the projects that make it through.
Grid Strategies, an independent analysis firm, calculated that if PJM had simply moved a portion of its stalled queue faster, ratepayers could have saved roughly $7 billion in that single auction cycle.
Not from new construction. From unsticking projects that had already been studied and were just sitting.
A Price the Regulators Are Already Working to Undo
The same scarcity generating this windfall has already triggered enough political pushback to put a ceiling on it.
Five governors, from Illinois, Maryland, Delaware, New Jersey, and Pennsylvania, have publicly called for changes to how PJM’s market operates.
Pennsylvania sued over the auction’s design and won a $21 billion settlement, now flowing back to ratepayers to offset the exact price spike data center demand helped cause.
American Electric Power floated leaving PJM entirely, unable to move fast enough inside the current structure, which is a strange thing for an incumbent to consider if the current price were actually stable.
Maryland’s Office of People’s Counsel put the scale plainly in an October 2025 letter: PJM’s 2026 load growth requests alone represent a 37% jump over 2025’s entire peak demand, and PJM covers only about a fifth of the country’s population.
Regulators aren’t reading that as a growth story.
They’re reading it as a solvency question, and a capacity price that draws lawsuits and nine-figure settlements within a year of clearing isn’t a floor. It’s a number under active compression.
What This Actually Requires Checking
The AI power demand thesis isn’t wrong.
It’s aimed at the wrong unit.
“Utility” isn’t one exposure here.
The real exposure sits in three narrower places: whether a company holds cleared capacity in a specific auction, where it sits in line for dispatchable generation hardware like gas turbines (whose order backlogs already stretch years out), and which regulator oversees it, because operating in a state actively suing over auction design is a different risk than operating somewhere reform hasn’t landed yet.
None of that shows up in a sector-level “AI tailwind” headline.
It shows up in interconnection queue disclosures, turbine order books, and state regulatory filings, none of which draw the attention a $14.7 billion number does on its own.
I won’t treat a utility’s AI-demand narrative as an earnings input until I’ve checked two things: whether it holds cleared capacity in the relevant auction, and whether its home regulator is one of the five currently working to bring that price back down.
The queue number is real. What it means for any single company depends entirely on where that company actually stands inside it.
Sources & References
- Lawrence Berkeley National Laboratory. Queued Up: Characteristics of Power Plants Seeking Transmission Interconnection in the United States (2024 Edition). Primary source for U.S. interconnection queue size, installed-capacity comparisons, project withdrawals, and technology mix. https://emp.lbl.gov/news/grid-connection-barriers-new-build-power-plants-united-states
- Lawrence Berkeley National Laboratory. Queued Up: 2025 Edition. Updated national interconnection queue statistics based on projects submitted through the end of 2024. https://www.osti.gov/biblio/3008763
- PJM Interconnection. Critical Issue Fast Path: Large Load Additions. Source for PJM’s projected peak-demand growth and the contribution of AI data centers. https://www.pjm.com/-/media/DotCom/committees-groups/cifp-lla/2025/20251014/20251014-informational-only---critical-issue-fast-path---large-load-additions---problem-statement---updated---clean.pdf
- RMI. PJM’s Speed to Power Problem and How to Fix It. Source for Transition Cycle 1 (TC1) statistics, including application volumes, study timelines, and draft interconnection agreements. https://rmi.org/resources/pjms-speed-to-power-problem-and-how-to-fix-it/
- Lawrence Berkeley National Laboratory. PJM Data Show Substantial Increases in Interconnection Costs. Source for network-upgrade cost escalation. https://emp.lbl.gov/news/pjm-data-show-substantial-increases
- Grid Strategies. Analysis of the PJM interconnection backlog and consumer-cost impacts. https://pv-magazine-usa.com/2025/02/26/consumers-will-pay-billions-due-to-very-slow-interconnection-in-the-pjm-grid-study-says/
- Utility Dive. Coverage of PJM capacity auction clearing prices and market impacts. https://www.utilitydive.com/news/pjm-interconnection-capacity-auction-prices/753798/
- Congressional Research Service. PJM Capacity Market. Background on PJM market design, reforms, and state policy concerns. https://www.congress.gov/crs_external_products/R/HTML/R48553.html
- Commonwealth of Pennsylvania. FERC Approves Governor Shapiro’s PJM Settlement. Official source for the projected consumer savings from the settlement. https://www.pa.gov/governor/newsroom/2025-press-releases/ferc-approves-gov-shapiro-settlement-pjm-prevent-unnecessary-pri
- Utility Dive. Coverage of governors’ calls for PJM capacity-market reform. https://www.utilitydive.com/news/governors-pjm-capacity-market-auction-opsi-p3-ls-power/731186/
- ERCOT. System Planning and Weatherization Update. Source for Texas large-load interconnection requests. https://www.ercot.com/files/docs/2025/12/02/16.2-System-Planning-and-Weatherization-Update_Revised.pdf
- Renewable Energy World. Coverage of CenterPoint Energy’s large-load interconnection request growth. https://www.renewableenergyworld.com/energy-business/new-project-development/a-fundamental-shift-centerpoint-sees-700-increase-in-data-center-interconnection-request-queue/
- Interconnection.fyi. Public database for project-level interconnection queue data across U.S. grid operators. https://www.interconnection.fyi/
- U.S. Department of Energy. DOE Distributed Energy Resource Interconnection Roadmap. July 31, 2026. https://www.energy.gov/cmei/i2x/doe-distributed-energy-resource-interconnection-roadmap
- E&E News by POLITICO. PJM Proposes Data Center Registry with Teeth. July 31, 2026. https://www.eenews.net/articles/pjm-proposes-data-center-registry-with-teeth/
- POWER Magazine. Texas Audit Could Delay 49.8 GW of Data Center Load, Cost Projects Up to $15 Billion, BNEF Warns. August 2026. https://www.powermag.com/texas-audit-could-delay-49-8-gw-of-data-center-load-cost-projects-up-to-15-billion-bnef-warns/
- PJM Inside Lines. PJM Board Directs Action on Resource Adequacy, Affordability and Large Loads. July 31, 2026. https://insidelines.pjm.com/pjm-board-directs-action-on-resource-adequacy-affordability-and-large-loads/
메타데이터
- post_id
- 0bbc1469e332
- slug
- us-sector-analysis-the-power-grid-interconnection-backlog-0bbc1469e332
- url
- https://medium.com/write-a-catalyst/us-sector-analysis-the-power-grid-interconnection-backlog-0bbc1469e332
- canonical_url
- https://medium.com/write-a-catalyst/us-sector-analysis-the-power-grid-interconnection-backlog-0bbc1469e332
- author_url
- https://medium.com/@MrChandravanshi
- status
- ok
- fetched_at
- 2026-09-17 19:58:22