Somewhere in America right now, several terawatts of proposed wind, solar, battery, and gas projects are sitting in line, waiting for permission to connect to the grid. The total capacity in interconnection queues routinely exceeds the entire existing U.S. generating fleet. Only a fraction will ever get built. The interconnection queue — the administrative process by which new generators get studied, assigned upgrade costs, and approved to connect — has become one of the most important and least understood chokepoints in the energy transition. Understanding it explains why announcements of new projects so often dissolve into years of silence.
What the queue is
When a developer wants to build a power plant, connecting it to the grid is not as simple as plugging in. The new generator changes power flows across the network: it may overload existing lines, require substation upgrades, or trigger stability issues hundreds of miles away. Before it can interconnect, the grid operator must study these impacts and identify the network upgrades needed to accommodate the project safely. The developer typically pays for those upgrades — or a share of them — plus the direct cost of the connection facilities.
The queue is the ordered list of projects awaiting these studies. Each regional grid operator (and each utility balancing authority outside organized markets) runs its own queue with its own rules, but the traditional structure has been serial: projects are studied in the order they applied, one after another, with each study building on the results of the ones before it.
How the process works, step by step
A typical interconnection process runs through several phases:
- Application and feasibility: the developer submits site control evidence, technical specifications, and a study deposit. An initial screening identifies obvious problems.
- System impact study: engineers model the project against the existing and queued system, identifying thermal overloads, voltage issues, and stability concerns. This produces the list of required network upgrades and their estimated cost.
- Facilities study: detailed engineering of the upgrades themselves — what equipment, what construction, what timeline.
- Interconnection agreement: the binding contract allocating costs, milestones, and deadlines. The developer posts security; the transmission owner builds the upgrades.
Each phase takes months, and the full cycle has historically run 3 to 5 years — longer in congested regions. For illustration, a project entering a queue in 2024 might reasonably expect a commercial operation date in 2028 or later, with the interconnection process consuming much of the gap between announcement and construction.
Why the queue is so backlogged
Three dynamics compound. First, volume: the number of proposed projects has exploded — driven by state clean-energy mandates, corporate procurement, tax incentives, and load growth — overwhelming study teams sized for a fraction of the applications. Queues that once held tens of gigawatts now hold hundreds.
Second, serial restudies: because studies build on each other, when a project ahead of you withdraws or changes, everything behind it must be restudied. Withdrawals are common — historically, the large majority of queued projects never reach commercial operation — so restudy cascades ripple through the queue repeatedly, invalidating months of work.
Third, upgrade cost allocation: under the traditional model, the project that triggers a needed upgrade pays for it, even when the upgrade benefits the whole system. A developer can receive a study assigning it hundreds of millions of dollars in network upgrades for a project worth less than that — and rationally withdraw. This “first mover pays for shared infrastructure” problem means the upgrades the system needs most are the ones least likely to be funded by any single project.
The cluster-study reform
FERC’s interconnection reform (Order 2023) attacked the serial problem directly by requiring cluster studies: instead of studying projects one by one in application order, operators study groups of projects together in periodic windows. Projects in a cluster share upgrade costs by formula, withdrawals trigger restudies of the cluster rather than cascades through a serial line, and readiness requirements (site control, deposits, commercial milestones) weed out speculative projects before they consume study resources.
The reform also imposed study deadlines with penalties, standardized the modeling assumptions, and required operators to consider grid-enhancing technologies as alternatives to traditional upgrades where appropriate. Implementation is phased across regions and still working through backlogs accumulated under the old rules — reform changes the process going forward, but the existing mountain of queued projects must still be worked down.
What the queue data really tells you
Queue totals are among the most misread numbers in energy. Headlines citing “two terawatts in the queue” imply a tidal wave of imminent construction. The reality: queues are options portfolios, not construction pipelines. Developers submit multiple speculative projects for each one they intend to build, because queue positions are cheap relative to the value of optionality. Historical completion rates — the share of queued capacity that reaches commercial operation — have typically been in the low teens to around twenty percent.
The useful signal is not the total but the composition and the cost curve: what technologies are queuing, where, and what upgrade costs the studies assign. Rising assigned upgrade costs in a region signal genuine transmission scarcity — the point at which building shared network infrastructure becomes cheaper than making each generator pay individually. Falling completion rates signal a queue clogged with speculation rather than real projects.
Storage and hybrid projects: the new complexity
Batteries and hybrid solar-plus-storage projects have added new wrinkles. A battery both draws from and injects into the grid, so its impact depends on when it charges and discharges — behavior the developer controls. Studying a battery under worst-case assumptions (charging at system peak) can assign it upgrade costs for a scenario it would never create. Operators are developing new study methodologies — “operating assumption” studies that model realistic dispatch — but the modeling is genuinely harder than for a one-directional generator.
The cost of the queue: who bears the delay
Queue delays are not free — they reallocate costs and risks across the system. For developers, a multi-year interconnection process means years of carrying costs on land options, deposits, and development teams, plus the risk that assigned upgrade costs render the project uneconomic after millions have been spent. This development risk premium flows into power prices: projects that survive must earn enough to cover the ones that did not.
For the system, the cost is delayed capacity. Utilities planning for load growth or retirements need new resources online by specific dates; when interconnection pushes commercial operation back by years, the options narrow to delaying retirements, paying scarcity prices, or procuring emergency capacity. Inadequate interconnection throughput is thus a resource adequacy problem with a time lag — today’s queue delays become tomorrow’s capacity shortfalls.
For consumers, the cost is diffuse but real: slower entry of cheap generation means higher wholesale prices for longer, and the upgrade costs assigned to generators ultimately flow into the project economics that set contract prices. An inefficient queue is a tax on the energy transition, paid by everyone.
Affected systems: when your project impacts the neighbor’s grid
A complication that surprises newcomers: a generator interconnecting in one system can trigger required upgrades on a neighboring system. Power flows do not respect balancing-authority boundaries, and a large project near a seam can overload the neighbor’s lines. The neighbor — the “affected system” — then studies the impact and assigns upgrade costs, adding another study track, another timeline, and another cost allocation to the project.
Affected-system coordination is among the slowest parts of interconnection because it involves two (or more) operators with different rules, models, and priorities. FERC’s reforms imposed coordination requirements and timelines, but seams remain friction points. For developers, the lesson is locational: projects near system boundaries carry affected-system risk that inland projects do not, and experienced developers price it in.
How developers navigate the queue
Sophisticated developers treat interconnection as a core competency, not an administrative afterthought. The playbook: site for the grid, not just the resource — a slightly worse wind site near strong transmission beats a great site behind a constrained interface. Enter with readiness — site control, permits in progress, and commercial backing — to survive cluster-study readiness screens that cull speculative projects. Model early — run independent power-flow analysis before applying, to estimate upgrade costs rather than discovering them in the system impact study. And maintain optionality — hold multiple queue positions across regions, expecting most to fail.
The queue has also created a secondary market: late-stage queue positions with completed studies and assigned upgrade costs are valuable assets, sometimes worth more than the development work itself. Acquiring a de-risked interconnection position can be faster than starting a new application — which tells you everything about where the scarcity lies. It is not in turbines or panels. It is in studied, approved grid access.
The large-load wrinkle: data centers in the queue
Interconnection queues were designed for generators, but the fastest-growing queue entries in some regions are now large loads — data centers requesting hundreds of megawatts or gigawatts of new service. Large-load interconnection raises mirror-image issues: instead of studying whether the grid can absorb new supply, operators study whether it can serve new demand, and the required upgrades (new substations, feeder buildouts, transmission reinforcements) can be just as large.
The emerging questions are cost allocation and timing. Should a single data center pay for substation and transmission upgrades that will serve future customers too? How should speculative load requests — developers shopping multiple sites — be screened so they do not clog study resources the way speculative generation did? And how do large, flat, 24/7 loads interact with resource adequacy planning built around weather-driven peaks? Regions are developing large-load interconnection procedures, minimum-load-factor requirements, and in some cases, requirements that large loads bring their own generation or curtail during emergencies. The queue, it turns out, was only the first interconnection challenge; the load side is the second.
Why it matters beyond developers
The queue is where the physical limits of the grid meet the financial plans of everyone else. Utilities counting on new generation for resource adequacy need to know which queued projects are real. Policymakers counting on queued clean energy to meet mandates need to know the completion rate. Investors need to distinguish developers with late-stage, de-risked queue positions from those holding early-stage lottery tickets.
The deepest lesson of the queue is the one from the transmission bottleneck: generation is easy to propose and hard to connect. The United States does not have a shortage of proposed power plants. It has a shortage of studied, cost-allocated, buildable grid capacity to absorb them. Until the queue moves faster — through cluster reforms, proactive transmission planning, and realistic speculation filters — the gap between announced megawatts and operating megawatts will remain the defining feature of the buildout.