On the morning of July 22, 2026, more than 3 GW of power demand — roughly 3% of system load at that moment — disappeared from the grid in Northern Virginia as data centers switched rapidly to backup power during a major transmission disturbance, according to a report published by Brandon Owens, founder of AIxEnergy.io, in Utility Dive. While PJM Interconnection said the event caused no reliability impact and Dominion Energy stabilized the system within minutes, the report describes the incident as "more of a harbinger of things to come than an isolated event" and a warning about mounting risks from the wave of large loads being added to power systems worldwide.
The Northern Virginia event wasn't the first time large loads have suddenly vanished from the grid. Between November 2020 and March 2023, the Electric Reliability Council of Texas identified eight incidents in which faults near a large Texas Gulf Coast industrial facility produced repeated demand reductions of approximately 400 MW to 700 MW, with system frequencies climbing as high as roughly 60.11 Hz. On Dec. 7, 2022, multiple faults and delayed fault clearing after a breaker failure caused an approximately 1,560 MW load reduction in West Texas — with ten large power-electronic loads accounting for about 162 MW of the drop, oil-and-gas facilities for around 420 MW, and thermal generation for 112 MW. Overseas, Irish grid operators EirGrid and SONI documented four major data center demand reductions tied to 220-kV transmission events: 74 MW on Jan. 7, 2022; 204 MW on Dec. 13, 2022; 321 MW on Jan. 26, 2025; and 387 MW on May 8, 2025.
According to the report, most debates about large loads focus on whether enough generation and transmission can be built, project timelines, financial security, and who pays for infrastructure upgrades — questions that address only "whether the system can connect and serve the load — not how that load behaves once it is operating." A data center might procure generation, fund network upgrades, and meet capacity and interconnection requirements while still presenting a poorly modeled common-mode transfer risk during a disturbance, the report states. The Irish system operators proposed new grid code modifications requiring facilities to restore at least 90% of pre-fault demand within 500 milliseconds after fault clearance and voltage recovery, though the proposal remains under regulatory consideration.
The report argues that grid operators need a more comprehensive operating architecture for large loads, one that identifies the largest plausible simultaneous demand reduction at both facility and electrical-cluster levels, establishes performance-based voltage and frequency ride-through requirements, and requires verified as-built models of IT loads, cooling systems, uninterruptible power supplies, protection settings, backup generation, transfer logic, and reconnection timing. The framework should also provide grid operators with timely telemetry on real and reactive power, voltage, transfer status, and expected restoration behavior, while preserving a shared forensic record after significant events. In May 2026, NERC issued a Level 3 Alert calling for improved computational-load modeling, studies, instrumentation, commissioning, operational coordination, protection, and control, though it's not an enforceable reliability standard. In July, FERC directed NERC to develop new or modified computational-load reliability standards and registration criteria, with filings due by Dec. 31, 2026.
For most of the power industry's history, reliability planning has focused on the sudden loss of supply — but the computational era is creating its mirror image: the sudden loss and return of demand. The next major contingency may not begin with a power plant or transmission line failure, the report warns, but behind the meter, through thousands of power-electronic devices responding simultaneously to the same disturbance, each protecting its own facility exactly as designed.