The PJM Capacity Crisis: How Data Centers Are Redefining Power Market Reliability

The electricity landscape across the PJM Interconnection—the largest regional transmission organization (RTO) in the United States—is undergoing a seismic shift. Driven by an unprecedented surge in demand from hyperscale data centers and artificial intelligence infrastructure, PJM has reached a critical inflection point. The December 2025 capacity auction for the 2027/2028 delivery year exposed systemic vulnerabilities, clearing at US$333.44 per megawatt-day (MW-day). This price, hitting the Federal Energy Regulatory Commission (FERC)-approved cap for the third consecutive year, represents a staggering 1,053% increase from the 2024/2025 auction, which settled at a mere US$28.92 per MW-day.

As the grid operator scrambles to balance reliability with skyrocketing demand, it has unveiled a radical two-part policy proposal that aims to shift the financial burden of new capacity away from residential ratepayers and onto the massive corporate entities driving the consumption spike.

A Chronology of the Capacity Crunch

The path to the current crisis was paved by a rapid, unforeseen acceleration in electricity demand. In recent years, PJM—which serves all or parts of 13 states and the District of Columbia—found itself in a position where supply could no longer comfortably outpace the hunger of the digital economy.

• 2023/2024–2024/2025 Auctions: The baseline for capacity pricing remained relatively stable, with prices hovering in the sub-US$30 per MW-day range. During this period, the integration of renewables and the retirement of older thermal plants were managed through standard market signals.
• December 2025 Auction: The market hit a wall. Total procurement costs surged to US$16.4 billion. Analysis revealed that data centers were responsible for US$6.5 billion of that total—approximately 40%. Even more alarming was the revelation that US$6.2 billion of that data center cost was tied to facilities that have not yet even broken ground.
• September 2026 (Upcoming): PJM has proposed a one-time "Reliability Backstop Procurement" (RBP) auction to run from September 30 to October 21, 2026, intended to bridge the 6GW shortfall identified in the recent cycle.

Supporting Data: The Impact on Ratepayers

The economic implications for the average citizen are significant. According to projections from the Natural Resources Defense Council (NRDC), households within the PJM footprint could face bill increases of roughly US$70 per month by 2028 compared to pre-surge levels.

Evidence of this impact is already emerging. In Washington, D.C., customers of the utility Pepco saw a US$10 monthly increase on their bills stemming from the 2025/2026 delivery year alone. This "cost-spreading" model, where the burden of building new, expensive power plants is socialized among all ratepayers, is the primary target of PJM’s new regulatory framework.

The Two-Part Strategic Overhaul

To mitigate the risk of grid failure and protect residential budgets, PJM is pivoting to a dual-pronged strategy that fundamentally changes how large-load customers interact with the wholesale market.

Part 1: Reliability Backstop Procurement (RBP)

PJM’s proposed RBP is designed as a targeted intervention to address the 6GW capacity gap. By acting as a facilitator, PJM will conduct a one-time auction to procure new capacity. Crucially, the costs associated with these 15-year contracts will be passed directly to the utilities serving the specific zones where the large loads reside.

The mechanism is designed to incentivize speed. The auction will prioritize projects that can come online rapidly, including natural gas generators, nuclear facility uprates, and—most notably—Battery Energy Storage Systems (BESS). Because BESS can be deployed with greater agility than traditional base-load power plants, they hold a distinct competitive advantage in this procurement cycle. The success of this model hinges on state regulators ensuring that utilities force data centers to foot the bill, preventing these costs from being leaked into residential rate bases.

Part 2: Interim Resource Adequacy Service (IRAS)

Perhaps the most contentious and transformative element of the plan is the Interim Resource Adequacy Service (IRAS). Under this framework, PJM will terminate the practice of procuring capacity supply for new, large-load customers through its standard capacity auctions.

Instead, the onus shifts to the data centers themselves: they must secure their own power supply. If a large load fails to bring new capacity to the grid, PJM will no longer act as their safety net. These entities will be permitted to connect to the system, but they will be subject to strict curtailment orders during emergency periods. By prioritizing the protection of residential and commercial ratepayers from rolling blackouts, PJM is effectively telling the tech giants that their growth cannot come at the expense of public grid stability.

Official Perspectives: The NRDC and the Regulatory Horizon

Claire Lang-Ree, a clean energy advocate at the NRDC, characterizes the PJM proposal as a necessary, if imperfect, evolution of the energy market. "It’s a clear signal that large loads are no longer going to be allowed to ride on the backs of other ratepayers," she states.

However, the NRDC remains concerned about the "locational" element of the RBP. Currently, a power plant built in Ohio to serve a data center in Virginia could satisfy the contractual requirement while failing to account for the massive, multi-state transmission infrastructure needed to move that power. These "baseline transmission projects" represent billions in potentially hidden costs that could still be socialized. The NRDC is actively lobbying FERC to mandate that new generation resources be sited in closer proximity to the large loads they serve. Such a requirement would not only reduce transmission costs but would also favor energy storage, which can be sited closer to urban and industrial centers than traditional, large-scale generation.

Regarding the implementation of IRAS, the NRDC emphasizes that the heavy lifting must now move to the state level. Each state within the PJM footprint must draft new "large load tariffs" and establish precise protocols for curtailment. The goal is to ensure that when the grid is stressed, the lights stay on for families and small businesses, while data centers—who have failed to provide their own supply—are the first to be throttled.

Implications for the Future of the Grid

The uncertainty surrounding FERC’s upcoming decision adds another layer of complexity. Because FERC must approve or reject the entire filing, they cannot "cherry-pick" the positive elements of the RBP while refining the locational requirements. This all-or-nothing approach forces a high-stakes standoff between grid reliability, corporate growth, and consumer protection.

If approved, this policy will mark the end of the "business as usual" era for the North American power grid. Data centers, which have historically treated electricity as a commoditized, limitless resource, will now be forced to become active participants in energy procurement and infrastructure development.

The implications for the energy storage sector are particularly profound. As the grid moves toward a model that favors dispatchable, fast-responding capacity, BESS technology is poised to transition from a supporting actor to a lead player. By providing the flexibility to balance intermittency and respond to local load demands, batteries may prove to be the most efficient solution for the "large-load" era.

As the industry looks toward the 2026 Battery Asset Management Summit, the conversation will undoubtedly be dominated by these systemic changes. From the roles of artificial intelligence in load forecasting to the necessity of second-life battery applications, the industry is preparing for a reality where power is no longer just a utility—it is a finite, highly contested asset. Whether PJM’s two-part plan succeeds will depend on the political will of state regulators and the ability of the market to adapt to a reality where the digital economy must finally pay the true cost of its own insatiable demand.