What is an interconnection queue?

You cannot simply plug a power plant — or a gigawatt-scale data center — into the grid. The connection has to be studied and approved by the grid operator to make sure it won’t destabilize the system or overload transmission lines. The interconnection queue is the ordered line of projects waiting for that study and approval — in plain terms, the meaning of an interconnection queue is a waiting list for permission to connect to the electric grid.

The term shows up in two closely related forms. Grid interconnection is the process — the studies, agreements, and physical upgrades that let a new generator or large load join the transmission or distribution system. The interconnection queue is the backlog of requests moving through that process, ordered by when they were filed. So a project "in the queue" has asked for grid interconnection and is waiting its turn to be studied.

Each request triggers a series of engineering studies that model how the new project affects power flows, what upgrades the grid needs to accommodate it, and who pays for them. Only after clearing that process can a project sign an interconnection agreement and energize. Because every study depends on the assumptions of the projects ahead of it, the queue behaves less like a ticket counter and more like a chain — which is a large part of why it moves so slowly.

A request does not clear in one step. It moves through a staged sequence of studies, each one more detailed and more expensive than the last, and a project can be knocked out at any stage.

• Feasibility / initial study. A first-pass model of whether the grid can absorb the project at its requested point of interconnection, and a rough estimate of the network upgrades it would trigger.
• System impact study. The core engineering analysis — power-flow, short-circuit, and stability modeling that determines exactly what transmission upgrades the project requires and how it interacts with everything else in the area.
• Facilities / interconnection study. The detailed design and cost of the specific equipment (substations, lines, protection) needed to physically tie the project in.
• Interconnection agreement. The contract that assigns the upgrade costs and lets construction and energization proceed.

Modern reform bundles these into cluster studies: instead of studying every request one at a time in arrival order, the operator groups projects in a geographic area and studies them together, sharing the cost of common upgrades across the group. This is meant to break the old pattern where a single withdrawal near the front forced everything behind it to be re-studied. The staging is also why timelines run in years, not months — each phase takes quarters, restudies reset the clock, and a network upgrade identified late can add its own multi-year construction schedule.

People searching for "grid connection" and "grid interconnection" often mean the same real-world goal — getting a project tied into the power grid — but the words carry a useful distinction.

Grid connection is the general, everyday phrase for hooking something up to the grid: a rooftop solar array, a home, a factory, a new substation. It says nothing about scale or process.

Grid interconnection is the formal, regulated version of that for anything large enough to affect the bulk power system — utility-scale generation, storage, or a very large load. It is the term of art used by grid operators and regulators precisely because those connections require the study-and-approval process described above. When a data center or a power plant seeks grid interconnection, it is not just running a wire to the nearest line; it is entering a queue, commissioning impact studies, and negotiating who pays for the transmission upgrades its connection demands.

So every grid interconnection is a grid connection, but not every grid connection is a "grid interconnection" in the regulatory sense. The distinction matters for anyone sizing a project: a small connection may be a permitting formality, while a large one lands in a multi-year interconnection queue.

The queues have ballooned. Berkeley Lab’s Queued Up analysis found more than 2,000 GW of generation and storage capacity waiting nationally — more than the entire existing US power fleet — with the median project taking about four to five years from request to commercial operation (LBNL, 2025).

Several forces drive the jam:

• Volume. A flood of solar, wind, storage, and now large-load requests far exceeds the capacity to study them.
• Serial studies and withdrawals. When an early project drops out, downstream studies often have to be redone, cascading delays.
• Transmission scarcity. Many projects need network upgrades that are expensive and slow, and disputes over who pays stall them further.

The queues are public. Every regional grid operator publishes its active interconnection queue — CAISO (California), PJM (Mid-Atlantic), ERCOT (Texas), MISO (Midcontinent), SPP (Southwest Power Pool), NYISO (New York), and ISO-NE (New England) all maintain searchable databases listing each pending request with its size, fuel type, requested location, and study status. If you want to know what is waiting to connect in a specific region — or where a competitor’s project sits in line — the ISO’s own queue database is the primary source.

For the national picture, the standard reference is Berkeley Lab’s Queued Up series, which compiles the individual ISO and utility queues into one dataset and tracks capacity, wait times, and completion rates over time. The ISO databases tell you what is in line today; Queued Up tells you how the line is trending.

Because the raw queues are large, messy, and formatted differently by each operator, a small market of interconnection queue software and analytics tools has grown up around them — products that normalize the ISO/RTO spreadsheets, track a project’s status over time, and flag where transmission constraints cluster. The underlying data is still the public ISO queue; the software mostly makes it searchable and comparable across regions rather than adding new information.

For an AI operator, the queue is the difference between a plan and a working facility. A campus that needs power in 2027 cannot rely on a grid connection that clears in 2030. Even when capacity exists on paper, the study-and-upgrade timeline can blow past the build schedule.

This is the single biggest reason data centers are turning to behind-the-meter generation — building power on-site to bypass the queue entirely. The trade is straightforward: instead of waiting in line for the grid, you take on the cost and responsibility of generating your own power, and you control the timeline.