PJM has picked the projects it will rush through its interconnection process to meet demand that is arriving faster than new generation. On May 2 the grid operator said it had selected 51 projects under its Reliability Resource Initiative, a one-time opportunity for shovel-ready resources with short lead times. Together they would add 9,361 MW of unforced capacity. Most, 90%, are expected online by 2030, and all by 2031.
The list says a great deal about what PJM thinks can be built quickly. It is dominated by natural gas.
What was chosen
The 51 projects consist of 39 uprates of existing plants and 12 new construction proposals. The uprates apply to existing gas, nuclear, coal and onshore wind resources and total 2,108 MW of unforced capacity, which PJM defines as the amount of energy a resource can be expected to provide during times of highest reliability risk. The new projects add 7,253 MW of unforced capacity. Of the 12 new projects, six are gas, five are batteries and one is nuclear.
PJM's table, reported in summer megawatts of energy, gives the fuel split. Combined cycle gas accounts for 7,756 MW, with 1,613 MW from 20 uprates and 6,143 MW from six new plants. Gas combustion turbine uprates add 365 MW across 13 projects. Batteries add 2,275 MW from five new projects. Nuclear contributes 1,383 MW: 496 MW of uprates at four plants and 887 MW from one new project. A single coal uprate adds 14 MW, and the onshore wind item is a roughly 20 MW increase in capacity interconnection rights only. The energy total is 11,793 MW.
Gas, in other words, provides about two thirds of the energy megawatts. Solar does not appear at all. The single biggest category is new combined cycle plants, which take years to build even on a fast track, while the uprates, at 2,488 MW of energy, can come faster because they use existing sites and connections.
How PJM chose
PJM received 94 applications. It scored them with weights that gave the most value to unforced capacity and in-service date, followed by effective load carrying capability, which measures a resource's contribution to reliability, plus extra consideration for projects in zones with the greatest need for generation. PJM had planned to cap the initiative at 50 projects, but a tie in the scores for the last two led it to select 51.
The scoring method explains the outcome. A process that rewards accredited capacity and speed favors dispatchable plants and uprates at sites that already have grid connections. Batteries scored well enough to win five slots. Solar and wind, whose accredited capacity is lower per nameplate megawatt, did not.
Why PJM needed a fast lane
The demand forecast is the reason. PJM's 2025 load forecast, published in January, projects the summer peak reaching 209,923 MW in 2035, a 10-year increase of 55,779 MW, and 228,544 MW in 2045. The winter peak is forecast at 198,175 MW in 2034/35, up 62,048 MW over 10 years. Total energy is forecast at 1,328,045 GWh in 2035, an increase of 495,264 GWh. The report attributes zonal growth largely to data centers, listing data center load as a driver in AEP, APS, ATSI, BGE, Dayton, PECO, PPL, ComEd, Dominion and PSEG, among others.
PJM's release states the problem plainly: as new generation comes online more efficiently through its reforms, the demand for electricity and the retirement of generators continue to outpace new additions.
How the fast lane fits the queue
The RRI projects will be studied in the final transition phase of PJM's interconnection reform. FERC approved that reform in 2022, moving PJM from a first-come, first-served queue to a first-ready, first-served process with milestones and progress payments designed to weed out speculative projects. PJM says the transition has so far produced 18,000 MW of projects with signed interconnection agreements through its Fast Lane for projects with lesser system impact. Those are among about 47,000 MW of projects that have completed the process and are free to move to construction.
In addition to the RRI projects, PJM will complete processing of about 62,000 MW of projects in 2025 and 2026. Its new cycle process will begin in spring 2026, with a timeline of one to two years.
PJM listed other steps: more automation and staffing that it says have reduced the backlog by 60%, a multiyear collaboration with Google announced on April 10 to use AI tools in interconnection planning, FERC-approved surplus interconnection service that lets a battery be added to an existing site using unused capacity, and a pending reform to let replacement resources use the interconnection rights of retiring plants.
The limits
The RRI is a one-time program and its scale is modest against the forecast. At 9,361 MW of unforced capacity spread over several years, it covers a fraction of the 55,779 MW of summer peak growth that PJM expects by 2035. It also relies on projects that still have to be financed, equipped and built, and PJM's own forecast keeps rising. By 2045 the load report expects total energy of 1,482,068 GWh, a 20-year increase of 649,287 GWh, for a system that serves 67 million people across 13 states and the District of Columbia.
PJM also points to the gap between clearing its studies and actual construction. It says about 47,000 MW of projects have finished the process and are free to build, and it is urging states to work with developers to get them built. Interconnection was one bottleneck; permitting, financing and equipment are others.
For data center developers and the utilities serving them, the RRI list is a guide to what the next five years of new supply in PJM will look like. It will be mainly gas, a meaningful amount of batteries, and a small but notable contribution from nuclear uprates and one new nuclear project. Whether that is enough depends on how much of the forecast data center load actually arrives.