The US interconnection queue is no longer a niche dataset for transmission planners. It is now one of the clearest leading indicators of where renewable development capital is trying to go, where grid bottlenecks are worsening, and where project competition is concentrating before a deal ever reaches the market. Load Factor tracks 38,000+ queue projects, withdrawn included, across 16 ISOs and power markets. PJM alone accounts for 10,875 of them, more than CAISO, NYISO, and SPP combined.
That scale matters because queue entries are not abstract market signals. They are named project applications with fuel type, MW, status, county, and queue date. Read correctly, the queue is the first draft of the next decade's power buildout. Read poorly, it becomes a pile of expired placeholders and withdrawn projects. The point of a serious interconnection queue tracker is to separate those two realities.
The queue is heavily concentrated in a handful of markets
Load Factor's current public queue universe spans 16 markets, but the project count is concentrated in a few large regions.
| ISO / Market | Projects |
|---|---|
| PJM | 10,875 |
| MISO | 5,545 |
| ERCOT | 3,089 |
| SPP | 2,830 |
| CAISO | 2,448 |
| NYISO | 1,419 |
| ISO-NE | 2,000+ |
| 9 others combined | 12,969 |
| Total | 41,175 |
Three points stand out immediately.
First, PJM is not just the largest queue. It is in a different category. At 10,875 projects, it is roughly double MISO and more than triple ERCOT. Any investor, developer, or advisor trying to understand queue congestion in the US has to start with PJM.
Second, the "other ISO" bucket is too large to dismiss. Nearly 13,000 projects sit outside the six or seven markets that dominate most public discussion. That means national queue analysis breaks down if it only looks at PJM, MISO, ERCOT, CAISO, and NYISO.
Third, queue scale is not the same thing as queue clarity. Some markets disclose enough detail to support project-level developer research. Others disclose less, which is one reason a cross-ISO tracker has to normalize fields, statuses, and evidence quality rather than simply stack spreadsheets together.
Solar leads by project count, but storage is now too large to treat as an add-on
Across Load Factor's current database snapshot, solar remains the largest single fuel category by project count. But battery storage is no longer a sidecar to the solar story. It is one of the central stories in the queue.
| Fuel type | Projects | Capacity (MW) |
|---|---|---|
| Solar | 14,330 | 1,381,832.9 |
| Battery Storage | 9,354 | 1,557,878.7 |
| Wind | 6,316 | 1,022,585.6 |
| Gas | 3,905 | 1,138,472.6 |
| Hybrid | 927 | 103,399.7 |
| Nuclear | 224 | 31,553.5 |
The headline is straightforward: solar leads on count, but battery storage leads on MW in the current fuel-level rollup. That is a meaningful change in how queue risk should be interpreted.
For years, queue commentary treated storage as a secondary attribute of solar growth. The project counts no longer support that framing. A queue tracker that does not isolate standalone storage, hybrid storage, and storage-heavy geographies will miss a large share of new application volume.
Gas is still substantial in MW terms, which matters for any narrative that treats the queue as exclusively renewable. But in project-count terms, the center of gravity is now clearly solar, storage, and wind.
The queue keeps absorbing new entries faster than projects exit as operating assets
The interconnection queue's structural problem is not simply that it is large. It is that new requests continue to arrive at a pace that the system has not fully absorbed.
In Load Factor's current queue-year distribution, annual project additions remained elevated throughout the recent buildout cycle:
| Queue year | Projects entered |
|---|---|
| 2019 | 2,311 |
| 2020 | 2,414 |
| 2021 | 3,051 |
| 2022 | 3,382 |
| 2023 | 2,951 |
| 2024 | 2,697 |
| 2025 | 2,754 |
Capacity tells the same story. The 2025 cohort alone represents 588,248.2 MW in the current working snapshot. That is more queued capacity in one year than most power markets can realistically absorb into construction outcomes on any normal development timeline.
At the same time, the queue carries a large volume of withdrawals. In the broader project-status rollup, 24,921 projects are marked withdrawn, while 3,981 are marked operational. That does not mean the queue is broken beyond use. It means the queue is a development funnel, not a build forecast. The right way to read it is probabilistically: where are developers still willing to spend option value, and where is attrition becoming the dominant signal?
ISO-by-ISO differences matter more than national averages
A national queue total is useful for scale. It is not enough for decision-making.
MISO's queue is large, but its fuel mix is more balanced than CAISO's storage-heavy mix or SPP's wind-heavy mix. ERCOT still shows large storage and solar volumes, but its market design and geography create a different risk profile than PJM's study backlog. CAISO's queue has become a leading indicator for storage concentration. NYISO shows meaningful storage and transmission-related activity despite its smaller project count.
That is why a cross-ISO queue tracker has to do more than count projects. It has to normalize fuel labels, statuses, dates, and geography across markets that do not publish the same way. A solar project in MISO, a battery application in CAISO, and a hybrid filing in PJM do not arrive with the same level of metadata or the same implications for viability.
What this dataset is useful for, and what it is not
Used correctly, a national interconnection queue dataset can answer four commercially relevant questions.
First, where is development competition concentrating by market, state, county, and fuel type?
Second, which fuels are crowding each queue, and how quickly is that mix changing?
Third, which project cohorts are still active versus already withdrawn, suspended, or in service?
Fourth, which queue positions can be connected to real developers rather than opaque SPV names?
What it cannot do on its own is tell you that every project in the queue is viable. Queue data is intent data. It is not outcome data. That distinction is exactly why the best queue work pairs project counts with status history, geography, and evidence-based developer identification.
Why this page should exist as a reference
Most queue commentary is either too narrow or too generic. It focuses on one ISO, one reform filing, or one cluster study cycle. That is useful for specialists, but it does not answer the broader market question: what does the national queue actually look like right now?
The answer is that it is bigger than most public summaries imply, more storage-heavy than many market narratives admit, and more uneven across ISOs than a single national number can capture. PJM is the largest problem set. MISO remains one of the richest markets for developer-level research. ERCOT and CAISO are central to the storage story. And the long tail of other ISOs is large enough that ignoring it creates blind spots.
Methodology
Headline counts in this post use Load Factor's canonical product context: 41,175 queue projects, withdrawn included, across 16 ISOs and power markets. ISO totals reflect the public market breakdown used across Load Factor's marketing pages and internal product documentation. Fuel mix, queue-year, and status distributions come from the current PostgreSQL project snapshot, using normalized fuel_type, application_status, and queue_year fields. Capacity totals use summer_mw, with winter_mw as fallback where needed. Queue data should be interpreted as a development pipeline, not a built-capacity forecast.
Conclusion
The US interconnection queue is now too large, too fuel-diverse, and too commercially important to analyze market by market in isolation. The reference view has to be cross-ISO, project-level, and current enough to show where competition is moving before projects become transactions.
If you want to work from the underlying records rather than headline summaries, search every project on Load Factor. For the evidence layer behind developer attribution, see how Load Factor identifies developers and the companion post on who is behind that SPV.