Texas approved 765 kV ultra-high-voltage transmission line expansion to support anticipated multi-gigawatt AI data center loads.
Texas regulators' April 2025 approval of a $13.8 billion, 765-kilovolt transmission strategy represents a fundamental shift in grid planning — the state is now building massive transmission capacity for data centers and large electrical loads before developers even break ground. The Public Utility Commission of Texas (PUCT) backed the extra-high-voltage backbone, the first of its kind in ERCOT history, after concluding it would better accommodate decades of growth in data centers, cryptocurrency mining, hydrogen production and similar large loads than continued expansion of the existing 345 kV grid.
The decision capped a five-year planning effort that initially focused on electrifying the Permian Basin's oil fields. ERCOT's 2019 Delaware Basin study and 2021 Permian Basin planning work aimed to replace onsite generation with utility power as drilling operations electrified. By 2024, however, transmission providers were forecasting an "exponential increase" in non-oil-and-gas demand — primarily from cryptocurrency mining, data centers and hydrogen electrolysis. Those projects represented 11.6 GW of projected demand by 2030, nearly matching the oil-and-gas load that originally justified the initiative. Combined demand reached 23 GW by 2030 and 26.4 GW by 2038.
The strategy echoes ERCOT's past success with Competitive Renewable Energy Zones (CREZ), a transmission buildout originally designed to move West Texas wind generation that inadvertently created high-capacity corridors and substations later attracting hyperscale AI development. The 765 kV decision reflects a more deliberate effort to build transmission before planners know precisely where the next generation of major power users will emerge.
ERCOT chose the 765 kV option after analyzing alternatives: the comparable 345 kV plan cost $12.95 billion, while a 500 kV alternative was projected at $15.32 billion. The full 2038 reliability plan under the 765 kV option included common local upgrades, import paths and incremental local transmission improvements. PUCT staff concluded the modest premium for extra-high-voltage architecture was justified by greater transfer capability — 600 MW to 3,000 MW more regional capacity than the 345 kV alternative — a 5% reduction in annual transmission losses, fewer transmission corridors and greater flexibility to accommodate future generation and large electrical loads.
Historically, transmission planners designed grids around known generation additions, regional reliability needs and identifiable industrial customers. AI is forcing a different approach. Neil Osnato, founder of Persistence Analytics Group, explained that AI demand is "larger, faster-moving, less geographically fixed, and more sensitive to power availability than many traditional forms of economic development." For utilities, higher-capacity transmission becomes essential when the timeline for new load additions can be as short as 6 to 12 months — ERCOT noted the system is experiencing "unprecedented load growth" and actual demand in the Permian could exceed 2038 forecasts if interconnection timelines accelerate.
PUCT staff argued in their recommendation that a 765 kV network would better accommodate that uncertainty while providing the foundation for "a looped, networked statewide plan." ERCOT cannot know precisely where future load and generation will ultimately site in Texas; higher-capacity transmission allows utilities to absorb that uncertainty. "In that sense, flexibility becomes infrastructure," Osnato said.
The commission unanimously adopted the recommendation. Commissioner Kathleen Jackson said the Permian Basin had "outgrown its infrastructure" and, while the 765 kV option carries a higher initial cost, it provides "more transfer capability" and longer-term benefits. Building the system "for the future" while using today's dollars, she argued, represented sound policy.
For hyperscale developers, the practical benefit isn't simply bigger transmission lines — it's more places where campuses requiring hundreds of megawatts can connect to the grid. Higher-voltage transmission increases the number of substations capable of supporting AI infrastructure, easing congestion and expanding developers' siting options. Osnato noted that transmission capacity is becoming a competitive advantage for regions pursuing AI investment, as power deliverability joins land, fiber, tax incentives and permitting as deciding factors in site selection. "The winning regions will not be the ones that simply announce land, tax incentives, or data-center interest. They will be the ones that can prove power can be delivered at scale, on schedule, and without shifting unpriced risk onto utilities, ratepayers, or other customers."
Though the reliability plan focuses on West Texas, ERCOT's broader planning documents point toward a larger vision. The 2024 Regional Transmission Plan concluded that a statewide 765 kV network would provide greater transfer capability, reduce congestion and transmission losses, and expand siting options for both generation resources and large electrical loads. The plan would remain largely necessary even if forecast load growth falls roughly 20 GW short of current expectations.
Osnato believes the planning philosophy will spread beyond Texas. "Texas may simply be showing the rest of the country what the next planning problem looks like earlier than others." The state's wager is straightforward: build the transmission first and let the next generation of major power users follow.