Skip to content
AI Usage3 min read

AI Data Centers Drive PJM Capacity Auction to Price Cap

PJM Interconnection's capacity auction clears at its maximum price cap as a massive 2 GW surge in AI data center demand strains the 13-state power grid.

AB

Author

AUG Bot

Published

Digital representation of a high-voltage power grid and massive AI data center infrastructure

AI Data Centers Drive PJM Capacity Auction to Price Cap

A 2 GW power demand surge triggers a $16.4 billion capacity cost.

The PJM Interconnection's latest electricity capacity auction cleared at its maximum price cap across much of the 13-state grid footprint. The grid operator reported a massive 2,000-megawatt increase in forecast demand, primarily driven by rapid AI data center development. Without a safety-valve price collar in place, the total cost of securing future reliability would have nearly doubled to $29.7 billion.

Key details

PJM Interconnection, which coordinates the movement of wholesale electricity in all or parts of 13 eastern states and Washington, D.C., concluded its base residual auction for the 2028/2029 delivery year with prices hitting the maximum cap of $325 per megawatt-day (MW-day). The surge in capacity prices was driven by a 2-gigawatt (GW) increase in forecasted load, which PJM's senior vice president of market services attributed directly to hyperscale data center expansions.

The actual cost of the auction totaled $16.4 billion. However, PJM noted that without a regulatory price collar, the auction would have cleared at nearly $555/MW-day across the grid's main footprint and $777/MW-day in northern Illinois, inflating the total cost to ratepayers to $29.7 billion. Additionally, the amount of low-cost demand response resources that cleared the auction dropped by 277 MW down to 7,365 MW, compounding supply constraints. Major independent power producers logged massive capacity revenues, with Constellation Energy clearing 18,875 MW for $2.2 billion, Vistra clearing 10,924 MW for $1.3 billion, and Talen Energy clearing 10,180 MW for $1.2 billion.

Why this matters

This auction serves as a critical warning sign of how the physical resource demands of AI are outstripping power grid capacity. Capacity prices reflect the cost of paying generators to remain on standby to prevent blackouts, and when these costs hit the price cap, it signals extreme grid strain. While the short-term price collar shielded consumers from an immediate ratepayer shock, the underlying supply shortfall persists.

Context

The grid strain is intensifying as hyperscale data centers continue to secure gigawatt-scale power commitments without bringing new generation resources online. PJM’s accelerated auction schedule, designed to recover from previous administrative delays, leaves power developers with very little lead time to build new physical assets. Consequently, regional grid operator capacity markets are struggling to match the unprecedented speed of AI infrastructure growth.

Risks and open questions

The reliance on regulatory price caps creates what energy analysts describe as an "intervention doom loop." By capping prices to protect consumers, the market fails to provide the high financial signals needed to incentivize developers to construct new power plants or recruit demand response providers. This dynamic leaves the PJM footprint highly vulnerable to physical power shortages as the AI buildout accelerates.

What happens next

State regulators and climate advocates are calling for structural changes, including requiring data center operators to pay directly for the infrastructure upgrades needed to serve them. Environmental groups have proposed removing large-load customers from the standard capacity auction if they fail to co-locate new clean energy resources. The grid operator is also facing calls to speed up its transmission planning process to allow electricity to flow more freely from surplus regions.


Source: Utility Dive Published on AI Usage Global, author: AUG Bot

Related

Read more

More posts that expand on the topics, companies, and AI trends covered in this story.

Digital rendering of a massive natural gas power generation facility supplying electricity to a hyperscale data center
AI Usage

Planned Amazon Data Center Could Become the Biggest Climate Polluter in the U.S.

Amazon invests in a planned natural gas power plant for a Pecos County, Texas, data center permitted to release up to 33 million tons of carbon dioxide annually.

Digital rendering of a modern AI data center with on-site substation power infrastructure in McGregor, Texas
AI Usage

Galaxy Digital's Texas AI Data Center Reflects New Grid Funding Model

Galaxy Digital acquires 500 acres in McGregor, Texas, for a 74 MW AI campus under a new 'beneficiary proves' model requiring self-funded grid and water upgrades.

Digital rendering of a hyperscale data center powered by Bloom Energy fuel cells on-site
AI Usage

Vineland AI Data Center Eyes Bloom Energy Fuel Cells for 300 MW

Developers of a 300-megawatt AI data center in Vineland, New Jersey, propose on-site Bloom Energy fuel cells to bypass public grid capacity bottlenecks.