DOE Presses PJM to Shield Ratepayers from AI Data Center Grid Costs
Federal agency urges FERC to ensure large load customers fund 6.8 GW reliability procurement
The U.S. Department of Energy (DOE) has filed a rare intervention with federal regulators urging PJM Interconnection to reform how it assigns costs for a proposed 6.8-gigawatt reliability power procurement driven by rapid AI data center expansion. The department cautioned that PJM's existing proposal risks forcing household ratepayers to subsidize massive generation buildouts triggered by large commercial loads.
Key details
In a filing with the Federal Energy Regulatory Commission (FERC), the DOE supported federal regulatory concerns regarding PJM's cost allocation model. PJM, which operates the power grid across 13 Mid-Atlantic and Midwest states plus Washington, D.C., initially planned to procure 6.8 GW of backstop capacity after failing to acquire sufficient reserves in its last two base capacity auctions due to soaring data center load forecasts.
The FERC halted PJM's emergency procurement plan on September 30, 2026, opening a hearing process after finding that the grid operator's cost allocation rules could prove unjust and unreasonable. In response, PJM announced it will submit revised tariff proposals by October 29 following special stakeholder meetings.
The DOE emphasized that under the federal Ratepayer Protection Pledge, large energy consumers—not residential households or small businesses—must fund the electric generation and infrastructure required to serve them. Major PJM utilities including American Electric Power, Dominion Energy, Exelon, FirstEnergy, and PPL have signed the pledge. The DOE urged PJM to implement continuous project tracking so that capacity costs are dynamically assigned to specific large load customers and adjusted if planned data center projects are delayed or canceled.
Why this matters
PJM's capacity shortfalls highlight the mounting financial strain that hyperscale AI data centers impose on regional power grids. As tech giants request gigawatt-scale interconnections, grid operators are forced to procure emergency generation. Without granular tracking and direct cost assignment, utility bills for millions of households could rise significantly to fund power plants built primarily to sustain AI infrastructure.
Context
This intervention follows broader federal and state regulatory efforts to insulate consumers from AI-driven energy costs. With US data center electricity consumption projected to double by 2030, utilities across PJM and ERCOT face unprecedented demand spikes. In recent months, multiple state utility commissions and federal lawmakers have proposed targeted tariffs, exit fees, and large-load mandates to prevent cost-shifting onto residential customers.
What happens next
PJM will hold a special Members Committee meeting on October 22, 2026, to consult with stakeholders before submitting its revised reliability backstop filing to FERC on October 29. FERC will then evaluate whether PJM's revised cost allocation framework aligns with ratepayer protection requirements before allowing the 6.8 GW procurement to proceed.
Source: Utility Dive Published on AI Usage Global, author: AUG Bot
