Energy Vault Signs 1.25 GW Texas AI Data Center Power Deal Using Battery Storage and Off-Grid Generation
Energy Vault signed an agreement to supply battery storage, grid-forming power systems and control software for an initial 1.25 GW of integrated power infrastructure serving hyperscale AI data centers in Texas.
Published · Updated · 5 min read
The Energy Vault Texas AI data center power deal announced August 7 will deploy an initial 1.25 GW of integrated power infrastructure for hyperscale artificial intelligence data centers in Texas, combining battery energy storage, grid-forming power conversion, dispatchable generation and control software. Energy Vault said the systems are intended to operate independently of traditional utility interconnection schedules, making the agreement a significant example of data center developers turning to behind-the-meter power as grid access becomes a constraint on AI expansion.
The initial 1.25 GW deployment is backed by a hyperscaler customer contract and is expected to begin rolling out over the next four to 12 months, according to Energy Vault. The company expects the agreement to contribute approximately $500 million to $600 million of revenue during the second half of 2026 and 2027, although those financial expectations remain forward-looking.
The platform combines batteries, generation and grid-forming technology
Under the August 7 commercial agreement announced by Energy Vault, the company will supply battery energy storage systems, grid-forming power conversion systems and AI infrastructure control software.
Those components will be combined with turnkey power generation, Caterpillar generator sets and engineering, procurement and construction capabilities supplied through Energy Vault’s strategic infrastructure partner.
The result is effectively a modular power plant designed around the operating requirements of a hyperscale computing campus rather than a conventional utility-connected industrial load.
Energy Vault said the architecture will coordinate generation, batteries, power conversion equipment, redundancy systems and electrical infrastructure through a common control platform. The system is intended to balance power flows and maintain voltage and frequency while responding to rapid changes in AI computing loads.
Off-grid power is intended to bypass long interconnection timelines
The most consequential aspect of the agreement is not simply its scale but its intended relationship with the utility grid.
Energy Vault says the companies will jointly deploy fully integrated off-grid systems that can bring computing capacity online without waiting for conventional utility interconnections. The architecture can later incorporate utility electricity, renewable generation and additional distributed energy resources as individual campuses evolve.
That approach reflects a growing speed-to-power challenge facing the data center industry. Large AI campuses can require hundreds of megawatts or even gigawatts of electricity, while new transmission infrastructure and utility generation can take years to plan, permit and construct.
The Energy Vault Texas AI data center power deal instead places substantial generation and storage directly with the computing load, allowing developers to treat power infrastructure as part of the campus itself.
Battery storage performs a different role than the generators
Although dispatchable generators provide the underlying electricity supply, Energy Vault’s batteries and grid-forming inverters are intended to manage some of the characteristics that make AI loads difficult to serve.
Large GPU clusters can change their electricity consumption rapidly as computing workloads increase or decrease. Those swings create power-quality, frequency and voltage-management requirements that cannot always be addressed efficiently by generation equipment operating alone.
Energy Vault says its control software will dynamically coordinate power flows, reduce generator cycling and maintain electrical stability while batteries provide rapid response to changing loads.
This creates an architecture resembling an advanced industrial microgrid, although Energy Vault describes the offering more broadly as integrated AI power infrastructure.
The agreement builds on Energy Vault’s push into powered AI campuses
The August 7 agreement is Energy Vault’s second recently announced strategic framework focused on AI infrastructure.
On July 27, the company announced the start of work on a powered AI infrastructure campus in Snyder, Texas, where it plans to combine power infrastructure with modular data center capacity.
Energy Vault’s broader business includes utility-scale battery storage, gravity storage, hydrogen storage and energy-management software. The company’s strategy has increasingly expanded from supplying individual storage systems toward developing and operating integrated energy infrastructure.
The new agreement takes that model further by positioning batteries, generation and electrical controls as a packaged solution for large computing customers.
The customer and Texas project locations remain undisclosed
Important details about the 1.25 GW deployment have not yet been made public.
Energy Vault identified the end customer only as a hyperscaler and did not disclose the specific Texas data center locations. Its national power infrastructure partner also was not named in the announcement.
The 1.25 GW figure describes integrated power infrastructure capacity and should not be interpreted as 1.25 GW of battery capacity alone. Energy Vault did not disclose the battery megawatt or megawatt-hour capacity within the initial deployment.
The company also described deployment over the next four to 12 months as an expectation rather than a completed construction milestone.
Behind-the-meter power is becoming part of the data center development model
The agreement illustrates how the collision between AI development schedules and electric-grid development schedules is changing data center power architecture.
Instead of treating utility service as the only primary source of electricity, developers are increasingly evaluating combinations of on-site generation, batteries, renewable resources and microgrid controls that can provide firm power before or alongside a permanent grid connection.
Energy Vault’s architecture is specifically designed to preserve that flexibility. The company says utility electricity and additional distributed energy resources can be integrated later, meaning an initially off-grid campus does not necessarily have to remain permanently isolated.
For utilities and grid planners, that distinction will become increasingly important. Large behind-the-meter power systems may reduce immediate interconnection requirements, but gigawatt-scale computing campuses still influence regional fuel infrastructure, emissions, land use and long-term electricity planning.
For data center developers, the commercial proposition is more immediate: the ability to secure reliable electricity may now determine how quickly new AI computing capacity can actually be deployed.
Jonas Muthoni is Editor-in-Chief of Microgrid Media, where he oversees reporting and analysis on microgrids, energy storage, distributed energy resources, data center power demand, grid modernization, resilience, renewable energy, and electricity policy and markets. His work focuses on the infrastructure, technologies, and regulatory developments reshaping the power system.
As Editor-in-Chief, Jonas leads Microgrid Media’s editorial strategy and standards, including story selection, source verification, technical accuracy, and the development of original reporting and analysis. His coverage draws on regulatory filings, government records, utility and company disclosures, technical documentation, and independent industry sources to explain significant developments across the evolving energy system.
