PJM's July Peak Broke Records, but the Bill Reveals Who Paid for Running the Grid to Its Ceiling

A landmark piece in Power Magazine reported that PJM Interconnection, serving 67 million people across the Mid-Atlantic, Midwest, and South, set an all-time peak demand record of more than 168,000 MW on July 2, breaking a mark that had stood since 2006.^[1] The grid did not fail. It also did not hold up in any meaningful sense. Holding it up required two federal emergency orders under Section 202(c) of the Federal Power Act, forced curtailment of data centers onto backup generation, activation of emergency demand response, a recall of generators from scheduled maintenance, and operating reserves slashed to roughly 5,100 MW, about half the prior level.^[1] Day-ahead wholesale prices topped $2,000 per megawatt-hour in parts of the footprint, roughly triple comparable peaks a year earlier.^[1] When federal emergency powers become routine, the system is not holding up. It is operating at its ceiling, and the customers who kept their lights on paid triple for the privilege.

This was the third federal emergency intervention for PJM in 2026 alone.^[1] The pattern is not a surprise; it is a choice baked into the structure of monopoly regulation and the incentives it creates. The investor-owned utilities that dominate PJM's footprint, particularly in the Mid-Atlantic where Dominion Energy and Exelon operate, have for decades collected depreciation and maintenance allowances in rates for distribution and transmission upkeep, underspent on the actual assets, and distributed the collected cash as dividends to shareholders. The mechanism is discoverable in FERC Form 1 filings: accumulated depreciation grows faster than actual spending on vegetation management and pole inspection; dividend payouts climb in parallel; and the grid becomes fragile, shedding reliability margin under stress. When stress arrives, as it has this summer with back-to-back heat waves and explosive electricity growth from data centers,^[2] the system does not fail outright because federal and regional operators buy time through price signals so extreme that demand responds or prices ration out marginal users. That is not reliability. It is an accidental tax on ratepayers who cannot shift consumption or supply.

The accountability chart is missing from every emergency order and every news cycle celebrating that the lights stayed on. PJM faces near-term reliability challenges that demand additional resources, according to PJM's own planning statements,^[3] and the regional operator has worked throughout 2024 to adapt electricity markets to meet growing demands amid generation fleet changes.^[6] But adapted markets are not adapted infrastructure. The 1.7 percent annual summer peak demand growth forecast by PJM^[5] collides with a generation fleet where major coal and nuclear plants are retiring, such as Talen Energy's planned retirement of the 1,282-MW Brandon Shores power plant outside Baltimore,^[7] and the new supply entering the queue is mostly solar and wind, which are cheap to build but require grid storage, transmission capacity, and distribution hardening that the utilities owning those assets have deferred for years. The capacity auction for the 2025-2026 delivery year cleared enough capacity to meet demand plus required reserves,^[6] a statement that means the market cleared at the price needed to attract generation at the margin. That price reflects scarcity, not abundance. And scarcity pricing is a mechanism that transfers wealth from ratepayers to generator owners and backup-fuel suppliers, not one that builds the grid.

The same utilities that let reliability margins erode also own the transmission and distribution lines where the actual failures occur in storms and heat waves. Under cost-of-service regulation, they are paid a return on capital they deploy, which rewarded them for building the grid once and for decades not maintaining it fully, because maintenance is expensed and does not earn a return. The fix is performance-based regulation: binding reliability targets (SAIDI and SAIFI metrics) backed by symmetric penalties and rewards so that a utility that lets reliability slide loses revenue, not profits. Britain's RIIO framework and Hawaii's 2020 adaptation of it show the model: multi-year revenue caps with totex (total-expenditure) allowances that remove the capex bias and lock reliability outcomes to rates in advance. No utility has asked for a summer surcharge yet, but when one does, the remedy is a prudence review that asks not whether the spending is necessary now, but whether the neglect that made it necessary was prudent then. If it was not, the cost belongs to shareholders, not ratepayers buying the grid twice.

The alternative is to stop waiting for the next emergency order. Ratepayers in PJM's footprint, particularly in Virginia, Maryland, Pennsylvania, and New Jersey, can demand that their state public utility commissions adopt binding reliability performance standards with penalty mechanisms and require utilities to file multi-year capex plans tied to SAIDI and SAIFI targets, not to dividend policy. Those states can also open their interconnection queues to distributed energy resources and storage faster, reducing the need for marginal generation and buying time for transmission to be built. And they can accelerate public power alternatives: municipalization campaigns in targeted franchise areas, particularly around data centers and high-demand districts, reduce dependence on a single owner's maintenance choices and give ratepayers a control group against which to measure the IOU's performance-per-dollar.