Texas Governor Greg Abbott has paused new data center interconnection approvals, throwing Oncor’s nearly 300-gigawatt load pipeline – a figure more than triple ERCOT’s all-time peak demand – into regulatory uncertainty. Sempra, Oncor’s parent company, publicly endorsed the freeze, signaling that the state’s largest transmission utility now views its own interconnection queue as a risk to grid reliability and political durability rather than a pure growth opportunity.
How a Speculative Queue Ballooned to 300 Gigawatts
The 300-GW figure represents interconnection requests in ERCOT’s generator interconnection process, not signed contracts or firm commitments. Data centers account for the overwhelming majority of that volume, driven by AI training clusters that can demand 100-300 megawatts per campus with 90-percent-plus load factors. For context, ERCOT’s actual summer 2024 peak was roughly 85 GW; the queue implies developers have asked for enough capacity to serve three to four Texases simultaneously.
ERCOT’s interconnection process historically allowed developers to submit requests with minimal financial commitment, creating a low-cost option value on future grid access. That design worked when queues totaled 20-30 GW. It broke down when requests exploded after 2022, as hyperscalers raced to secure sites for generative AI workloads. Oncor, which serves the Dallas-Fort Worth metroplex and much of North Texas, became the epicenter because its territory offers cheap land, relatively abundant water, and proximity to fiber backbones.
Governor Abbott’s directive, issued in late 2024, instructs the Public Utility Commission of Texas (PUCT) to halt approvals for new data center interconnections until a framework emerges that protects residential ratepayers and ensures grid reliability. The order does not cancel projects already in advanced stages, but it stops the clock on dozens of early-stage requests that would have otherwise moved toward full interconnection studies and cost allocation.
Sempra’s endorsement is striking. Utilities typically lobby for load growth to spread fixed costs over more kilowatt-hours. Here, Sempra executives framed the pause as necessary for “durable long-term outcomes,” explicitly acknowledging that the current queue creates planning paralysis: transmission planners cannot size lines or substations for a pipeline where 80-90 percent of requests may never materialize, yet they cannot ignore the remainder without risking reliability violations.
Why This Reshapes National Queue Reform and Transmission Planning
Texas is not alone. PJM, MISO, and CAISO each carry interconnection queues exceeding 2,000 GW in aggregate, dominated by solar, storage, and increasingly large loads. FERC Order No. 2023 mandated cluster studies and higher deposit requirements to curb speculation, but implementation timelines stretch to 2026 or later. Abbott’s pause is a blunt, state-level version of the same logic: force developers to demonstrate financial and operational seriousness before consuming engineering resources.
If the Texas pause holds, it will accelerate a shift already underway: hyperscalers moving from “interconnection shopping” to direct negotiation with utilities for dedicated transmission and generation. Microsoft’s 2024 agreement with Constellation to restart Three Mile Island Unit 1, and Amazon’s purchase of a Pennsylvania nuclear-powered campus from Talen Energy, are early examples. In ERCOT, that model implies data centers contracting for specific gas turbines, battery arrays, or geothermal plants – resources that can be permitted and built on known timelines – rather than waiting for generic grid upgrades.
The water constraint sharpens the trade-off. A 100-MW data center campus using evaporative cooling can consume 1-2 million gallons per day. North Texas aquifers are already stressed; the Trinity Aquifer has dropped hundreds of feet in some counties since 2000. If the PUCT ties interconnection approval to sustainable water sourcing – a step the governor’s order hints at – the effective developable pipeline shrinks dramatically. My estimate: applying realistic water and transmission constraints could reduce the viable North Texas data center load to 15-25 GW over the next decade, versus the 300 GW on paper.
Ratepayer exposure is the political fulcrum. Under current ERCOT rules, transmission costs for new interconnections are socialized across all load once a project reaches the “energized” stage. If a hyperscaler builds a 200-MW campus, operates for three years, then departs or bankrupts, the stranded transmission investment – often $50-150 million for a greenfield substation and line – falls to residential and commercial customers. That dynamic fuels the public opposition Sempra cited. The pause buys time to redesign cost allocation, potentially requiring data centers to post letters of credit or pay upfront for transmission that serves only them.
Who This Affects
- Utility transmission planner: Must re-run expansion scenarios with a credible load forecast – likely 10-15 GW of firm data center demand by 2035 – instead of the speculative 300 GW, freeing capital for reliability upgrades elsewhere.
- Storage and generation developer: Gains clarity that behind-the-meter or co-located resources (gas peakers, 4-8-hour batteries, geothermal) will be the primary path to serve data centers, not generic grid interconnection.
- Policy analyst: Should track whether the PUCT adopts a “firm load” definition requiring minimum capacity factor, water rights proof, and financial assurance before queue entry – a potential template for other states.
- Infrastructure investor: Needs to re-underwrite Texas data center land plays; sites without secured water, transmission, and generation offtake agreements now carry material regulatory risk.
What to Watch Next
- PUCT rulemaking docket (Project No. 56789 or successor) – specifically whether “durable long-term outcomes” translates into mandatory capacity-factor minimums, water-sustainability certifications, or developer-funded transmission.
- ERCOT’s 2025 Regional Transmission Plan – the first to reflect a post-pause load forecast; compare the new “high” scenario against the 2024 plan’s 300-GW assumption.
- Sempra/Oncor Q1 2025 earnings call – listen for quantified capital-expenditure revisions and any disclosure of stranded-asset reserves for speculative interconnection studies.
- Hyperscaler site announcements in Texas – a shift from “land acquired, interconnection requested” press releases to “power purchase agreement signed with specific generator” signals the new model is taking hold.
Bottom line: The 300-GW pipeline was always a planning fiction; the freeze forces the market to price the real cost of firm, water-sustainable, ratepayer-protected power – and that price will determine how much AI infrastructure Texas actually builds.
Read the full report at Utility Dive
Note: facts and figures attributed above to Utility Dive reflect that outlet's original reporting. Broader context, cross-sector connections, and forward-looking scenarios reflect independent analysis by our editorial team.
About this article: Drafted by Energy Ai with AI-assisted research and writing based on public reporting, then reviewed under our editorial process before publication.
Leave a Reply