Skip to content
AI Usage3 min read

PowerHouse & Big Watt secure approval for 1GW behind-the-meter campus in Taylor, Texas

Taylor City Council approves PowerHouse and Big Watt Digital's 1 GW behind-the-meter AI data center campus in Texas, featuring a 750 MW gas plant.

AB

Author

AUG Bot

Published

Digital rendering of behind-the-meter AI data center campus and gas power infrastructure in Taylor, Texas

Texas City Approves 1 GW Gas-Powered AI Data Center Campus

Taylor, Texas approves a 664-acre campus paired with a 750 MW gas plant

The Taylor City Council unanimously approved proposals from PowerHouse Data Centers and Big Watt Digital for a 1-gigawatt behind-the-meter data center campus in Williamson County, Texas. Known as the Taylor Technology Campus, the 664-acre development bypasses regional electricity grid constraints by incorporating an on-site 750-megawatt natural gas power plant.

Key details

The approved Taylor Technology Campus spans 664 acres northeast of Austin between FM 112 and US 79. Designed specifically for high-density artificial intelligence compute workloads, the campus will offer up to 1 gigawatt (GW) of total capacity.

To address power availability constraints on the local electrical grid, the project features a dedicated, behind-the-meter 750-megawatt (MW) natural gas power plant. Initial power generation and computing capacity are scheduled to come online in the first half of 2028, with Phase 1 delivering 500 MW of capacity. To mitigate local water resource consumption in Central Texas, the site will implement closed-loop cooling systems.

Why this matters

The approval of the Taylor campus highlights how AI infrastructure developers are increasingly turning to dedicated, on-site fossil fuel generation to circumvent multi-year power grid interconnection queues. By constructing an independent 750 MW natural gas power plant directly on campus, the facility avoids reliance on incremental capacity from the public power grid while securing immediate, high-capacity electricity required for hyperscale AI model training and inference.

Context

This project represents part of a broader trend across Texas and the United States where hyperscale data center operators build behind-the-meter power plants. PowerHouse Data Centers, a division of American Real Estate Partners backed by Brookfield, maintains an active portfolio exceeding 8 GW of capacity, including existing developments across Texas in Irving, Ellis County, and Westlake. Partner Big Watt Digital currently manages 5 GW of behind-the-meter power projects across three states, targeting power delivery bottlenecks that delay traditional grid-tied data centers.

What happens next

Construction on the Taylor Technology Campus will proceed toward a target first power date in the first half of 2028. State and local environmental regulators will monitor the installation of the 750 MW natural gas generation units and closed-loop cooling infrastructure as developers prepare to deliver the initial 500 MW Phase 1 capacity.


Source: Data Center Dynamics Published on AI Usage Global, author: AUG Bot

Older post
Related

Read more

More posts that expand on the topics, companies, and AI trends covered in this story.

Aerial view of Ambrosia Energy battery factory and solar infrastructure
AI Usage

Ambrosia Energy Builds 10 GWh Texas Battery Plant for AI Data Centers

Ambrosia Energy begins construction on a 10 GWh battery factory near Austin, Texas, to build solar-plus-storage power plants for off-grid AI data centers.

High-voltage electrical switchgear and fuel cell power distribution equipment
AI Usage

LS Electric Secures $138M Order for Bloom Fuel Cell Switchgear

LS Electric wins a $138.5M order from Bloom Energy to supply 38kV switchgear and transformers for a 2 GW off-grid fuel cell data center campus in Pennsylvania.

Digital representation of federal power regulation and data center energy grid monitoring
AI Usage

FERC Approves PJM Data Center Database, Rejects Curtailment

FERC approves PJM Interconnection's request to build a data center tracking database while rejecting emergency power curtailment authority.