PJM’s Backstop Procurement and Connect-and-Manage Plan Turn AI Load Growth Into a Bilateral Capacity Market

PJM’s July 1 stakeholder update clears the publish bar because it is not just another warning that data-center demand is rising too fast. PJM said it is preparing two July filings at FERC: one for a Reliability Backstop Procurement and one for a Connect-and-Manage framework. The stronger Grid Report angle is that PJM is moving toward an explicit market structure for large loads that do not fit neatly inside the normal sequence of queue studies, utility service, and capacity-market procurement.

The primary-source facts are unusually specific. PJM said the Reliability Backstop Procurement would create a one-time process to purchase new supply resources to serve new data centers and other large loads. It also said that procurement will be combined with a bilateral contracting process in which PJM matches large-load customers with developers of new supply resources for long-term contracts. Separately, the Connect-and-Manage proposal would allow large loads to connect while recognizing that the system may not be able to serve them on peak days.

PJM is no longer treating AI-scale load growth as an ordinary queue problem. It is building a parallel service model with backstop supply procurement, bilateral matching, and explicit curtailment rules.

That combination matters more than the headline phrasing suggests. In practice, PJM is sketching a three-part operating model for AI-era demand growth: procure incremental supply outside the normal auction calendar, pair large customers directly with new generation or flexibility resources, and make curtailment risk explicit when load arrives ahead of firm system capacity. That is a much more concrete answer than simply telling developers to wait for the queue to clear.

This is why the story belongs in energy-grid rather than a generic policy bucket. The useful change is operational. PJM is treating large-load service as a reliability design problem that now requires its own procurement lane and its own service logic. The grid operator is no longer acting as though ordinary load-growth assumptions can absorb tens of gigawatts of new data-center demand without a separate mechanism.

The timing is also important. PJM said the Board will develop final Section 205 filings in July and expects to finalize commitments from the Reliability Backstop Procurement before the Dec. 9 capacity auction for the 2029/2030 Delivery Year. In an earlier June 25 update, PJM also said the backstop process could run in September and October to address an anticipated capacity shortfall for the 2028/2029 Delivery Year. That makes this a near-term reliability tool, not a distant planning concept.

One detail that deserves more attention is the proposed Large Load Registry. PJM said several Connect-and-Manage proposals centered on a public registry identifying large loads and their reliability impact by zone and by state. States could use that data to shape curtailment, flexibility, and cost-allocation policy. That pushes large-load governance out of abstract debate and into a more visible framework where regulators can see which projects are driving local system stress.

For operators, the practical implication is that speed to power in PJM may increasingly depend on whether a project can bring contracted supply, accept curtailment conditions, or fit inside a bilateral match process rather than simply requesting traditional service and waiting. For developers and investors, the implication is similar: the value of flexible load design, colocated resources, and contract-backed generation just went up again.

This is also what makes the story distinct from the site’s earlier June 22 PJM interconnection and June 24 Manual 13 coverage. Those stories were about process reform and emergency treatment of backup generation. This one is about PJM building a specific commercial and operational lane for AI-scale load growth, with procurement, matching, registry, and curtailment all on the table at once.

That is enough to publish. Searchers looking for PJM’s July 1 update do not need another generic recap of data-center demand. The more useful answer is that PJM is constructing a parallel service model for large loads: if enough new power is not available through the ordinary system in time, the grid operator is prepared to procure supply, broker long-term pairings, and formalize who gets curtailed when the system is tight.

PJM Inside Lines, “Stakeholder Process on Backstop Procurement and Connect and Manage Moves Forward,” published July 1, 2026: https://insidelines.pjm.com/stakeholder-process-to-procure-generation-and-serve-large-loads-moves-forward/

PJM Inside Lines, “The Capacity Auction Is Coming. Here’s What We’re Doing Now,” published June 25, 2026: https://insidelines.pjm.com/the-capacity-auction-is-coming-heres-what-were-doing-now/

PJM Inside Lines, “FERC OKs Temporary Process To Fast-Track Large Capacity Projects,” published June 11, 2026: https://insidelines.pjm.com/ferc-oks-temporary-process-to-fast-track-large-capacity-projects/

Federal Energy Regulatory Commission, “Fact Sheet | FERC Takes Action to Supercharge America’s Grid for Efficiency, Reliability, and a Bold Energy Future,” updated June 18, 2026: https://www.ferc.gov/news-events/news/fact-sheet-ferc-takes-action-supercharge-americas-grid-efficiency-reliability-and