Interconnection queue reform 2026 produced three kinds of numbers: a PJM study group, an ERCOT pause and an ESIG report. How to read each one.
← All research
5 min read 4 sources DistroForge Research

Interconnection Queue Reform 2026: What Each GW Counts

Interconnection queue reform 2026 produced three kinds of numbers: a PJM study group, an ERCOT pause and an ESIG report. How to read each one.

Interconnection queue reform 2026 has put a lot of large numbers in front of utility planners, and they do not measure the same thing. PJM’s 201.5 GW is generation that qualified for a study. ERCOT’s 66.4 GW is load in a conditional category inside a paused process. ESIG’s report is a set of recommendations, not a queue of its own.

A planner who lines those up in one column ends up with a forecast built from three different kinds of evidence. Each interconnection queue reform document says what it is, if you read past the headline.

  • PJM, August 3 release: 715 projects, 201.5 GW. It counts generation qualified for Cycle 1 study, at nameplate. It does not establish signed agreements, equipment orders or completed plants.
  • ERCOT, September 14-15 board deck: 204 projects, 66.4 GW. It counts large loads conditionally included as Batch Zero base load in 2032, including load ERCOT labels energized. It does not establish permission to energize new load, a final class or a study end date.
  • ESIG, July 2026 report: recommendations, not a queue. It covers planning practices and what to change. It does not establish adoption across planners.

PJM first ready first served: a screened study group

PJM’s release says the application window drew 811 proposals, and that 715 new generation projects qualified to be studied in the first Cycle of its reformed process. The total is 201.5 GW of nameplate capacity, led by natural gas at 99.8 GW across 147 projects and storage at 60.0 GW across 314 projects. Count and capacity rank those two differently, so state which one you mean.

The first-ready, first-served screen required “meaningful up-front financial commitments and proof of site control.” That makes this a firmer population than an open queue, and it is still a study group. PJM describes a “one- to two-year process” and says it expects to add to its signed-agreement count at the completion of Cycle 1 in 2028.

The most useful lines for a buyer sit near the bottom of the release. PJM says that “only a percentage of projects that apply for interconnection will sign interconnection agreements and become operational.” It also says it has 51 GW of generation with signed agreements, and that many of those projects “are either not being built at all or are being slowed by hurdles such as state permitting and supply chain backlogs.” So in PJM’s own account, even a signed agreement is not a built plant. We covered the earlier application totals in May; the August release is the screened set.

No Cycle 1 or Batch Zero attrition rate in these documents

It is tempting to turn 811 and 715 into a survival rate and apply it to the GW. That result would describe one screening step for one cycle, not how many projects get built. The same goes for any single attrition percentage borrowed from elsewhere. PJM’s statement about “a percentage” comes with no percentage attached to Cycle 1, and nothing here supports a number for one. The nearest figures are historical and measure other things: PJM says it has processed more than 300 GW since 2020, resulting in over 100 GW with signed agreements, and ESIG reports that ERCOT discounted a 218 GW forecast of 2030 peak demand to 145 GW, using historical materialization rates for data center loads.

Large load study process: ERCOT’s dated pause

ERCOT sits on the other side of the meter. Its September board deck says it issued a market notice on August 3 pausing the Batch Zero study process, and that it paused approvals to energize data centers or crypto facilities of 75 MW or greater. It reports that 204 projects (66.4 GW) met the qualifications to be conditionally included as base load, and 158 projects met them for studied load.

The 66.4 GW is a 2032 figure for large loads of 75 MW or greater, and it is not all new load. ERCOT’s chart includes 13,176 MW labeled energized after March 25, 2022 and 9,085 MW labeled energized before that date. The deck says final classification will come in December, based on its eligibility verification and audit.

The deck prints two capacity totals for the studied group, 127.9 GW on one slide and 125.4 GW on another. Until ERCOT reconciles them, neither is a number to carry into a plan.

ERCOT says two reports go to the Commission by December 10, 2026, and it calls the effect on the process and timeline “TBD.” A reporting date is not a restart date. AEP’s Form 10-Q, filed October 28, says ERCOT “has currently paused Batch Zero classification and study activity.” Every other ERCOT statement here is dated to the deck, so check our Batch Zero coverage and ERCOT’s own notices for later changes before relying on it.

Transmission planning bottlenecks: ESIG’s recommendations

ESIG’s July 2026 report, prepared by Johannes Pfeifenberger, Warren Lasher and James Okullo, focuses on “siloed planning processes, timing mismatches between large load development and transmission expansion, and uncertainty” about future demand. Its responses include scenario-based planning and working from “shared assumptions, shared data, and shared cases.”

Those are recommendations from a task force. The report is not a survey of who has adopted them, and it is not something a municipal buyer can run alone. Its value is as a checklist of questions to put to your transmission provider.

How to read the next number

Our guide to pipeline versus contracted load covers how to question a utility’s data center figure. Before any GW figure from interconnection queue reform goes into a planning file, label four things:

  • Side of the meter. Is it generation or load? Keep them in separate columns.
  • Stage. Is it an application, a study group, a conditional class, a signed agreement or an energized project?
  • Unit. Is it nameplate, requested demand or a project count?
  • Date and status. What does the source call the next step, and is that step a commitment or a report?

Then ask the project owner for the equipment commitment directly. None of these three documents ties a queue figure to a purchase order, an equipment quantity or a delivery date, and a filing does not hold a slot at a manufacturer.

Interconnection queue reform keeps moving, and the qualifiers move with it. The Feeder is our free weekly briefing on what moved and what it means for equipment timing. Sign up here.

Frequently Asked Questions

What does PJM's 201.5 GW figure measure?

PJM's August 3, 2026 release says 715 new generation projects with 201.5 GW of nameplate capacity qualified to be studied in Cycle 1. Nameplate is what PJM calls the maximum potential output of a resource. PJM says only a percentage of projects that apply will sign agreements and become operational.

Is ERCOT's 66.4 GW an approved amount of new load?

No. ERCOT's September 2026 board deck says 204 projects totaling 66.4 GW met the qualifications to be conditionally included as base load in Batch Zero. The deck charts that as a 2032 figure, and it includes load ERCOT labels energized. The deck said final classification would come in December. AEP's October 28 Form 10-Q says ERCOT has paused Batch Zero classification and study activity.

Can PJM and ERCOT numbers be compared?

Not like for like. PJM's figure counts generation seeking study, and ERCOT's counts large loads in a conditional category. They sit on opposite sides of the grid and at different process stages.

Free Member Access

Know what changed before your next quote

Free Member tier. Pick your topics. Get a weekly digest filtered to what you actually buy.

Set my topics No credit card. Three topics minimum.