Texas PUC forces data centers to hold through grid events as ERCOT queue hits 438 GW
Unanimous approval of fault ride-through requirements puts grid stability before connection speed as Texas's interconnection queue tops 438 GW.
The Texas Public Utility Commission on Thursday (2026-07-09) unanimously approved rules requiring large computational loads — data centers and crypto-mining facilities — to remain stable and connected to the ERCOT grid through voltage and frequency disruptions rather than tripping offline when conditions deteriorate. The decision arrives as developers have filed study requests for more than 438 GW of large load projects within ERCOT's footprint, a pipeline that dwarfs the state's current installed generation capacity.6
The significance for Texas power markets lies in what large loads do when a grid event strikes. A facility that disconnects abruptly during a frequency excursion can accelerate a manageable disturbance into something the grid cannot absorb. NERC's 2026 Summer Reliability Assessment found that large data centers and industrial facilities "pose risks of sudden load loss, which can trigger cascading outages," a hazard that scales with each new gigawatt added to the queue.4,2
The case for regulatory authority was contested before the vote. Industry opponents argued ERCOT lacked the explicit statutory basis to mandate ride-through capability. ERCOT staff rejected that framing: "Staff respectfully submits that delegated authority is sufficient," the regulator wrote, and the commission accepted that argument without dissent. Walker, citing the staff submission, was direct: "There is no debate that voltage and frequency excursions on the transmission network create reliability concerns, which increase with the interconnection of each new large computational load."6
The practical stakes are considerable. Some individual data center projects in the ERCOT queue could require more than 4,000 MW apiece, enough to serve roughly one million homes, according to earlier regulatory filings. Even if only a fraction of the 438 GW eventually connects, the cumulative demand would far exceed recent load growth. NERC's summer assessment noted Texas demand rose roughly 1.9% year-on-year, compressing reserve margins from 34% to 29%. Accelerated data center additions would tighten that buffer faster still.1,2,6
ERCOT had been moving toward this posture incrementally. The grid operator approved two landmark rule packages on Tuesday (2026-06-02), one establishing batch review for large load interconnection requests and the other setting technical requirements for large users. The Public Utility Commission followed with formal approval of the interconnection framework on Thursday (2026-06-18), creating what it called "Batch Zero," the first cohort to navigate the new process.3,5
Opposition from the Texas Blockchain Council and Texas Industrial Energy Consumers centered on cost and timeline. Trade groups argued that compliance could require billions of dollars in facility redesign and years of implementation, potentially making grid connection unfeasible for some projects. That argument did not move the commission.1
The grid's vulnerability at the margin has a specific shape. In Texas, the highest-risk hour has been shifting toward 9 p.m., when solar output drops off but cooling demand and data center load remain elevated. A large facility tripping at that moment, rather than riding through a voltage dip, can worsen a frequency event the grid is already straining to manage.4
What remains unresolved is the gap between rulemaking and physical compliance. Ride-through requirements that necessitate facility redesign face lead times that the interconnection queue does not wait for. With 438 GW pending and the commission having endorsed delegated rather than explicit statutory authority, legal challenges from industry groups remain possible. How courts treat that authority question will determine how much of the queue actually advances to construction and on what timetable.6,5