Lambda Oklahoma Data Center Plans Flexible Load and Full Energy-Cost Coverage
Lambda plans a new AI data center at MidAmerica Industrial Park in Oklahoma with commitments to fund its power infrastructure, pay its energy costs and make some computing load flexible during grid peaks.
Published · 4 min read
Lambda announced plans on September 27 to build a new AI data center at MidAmerica Industrial Park in Mayes County, Oklahoma, pairing the project with commitments to pay its own electricity-related costs and operate as a more flexible grid customer. The Lambda Oklahoma data center is still in development, and the company has not yet disclosed the facility’s planned electrical load, capital cost or construction schedule.
The announcement is notable because power infrastructure and cost allocation have become central constraints on large data-center development. Lambda said it will fund the resources and infrastructure needed to serve the facility, including 100% of its energy costs, while Grand River Dam Authority said Lambda will fund the upfront infrastructure costs required to serve the project. The company also said the site is being designed so certain computing loads can be shifted or reduced during periods of peak grid demand.
Lambda Oklahoma data center will be built at MidAmerica Industrial Park
Lambda selected a site in the southern section of MidAmerica Industrial Park, within the Chouteau-Mazie Public School District tax boundary. MidAmerica and Grand River Dam Authority used what the industrial park described as a reverse request-for-proposals process to identify a developer for the site, while Cloverleaf Infrastructure was selected to prepare the property for Lambda.
Lambda estimates the project could generate about $500 million in taxes over 10 years at current rates, create up to 1,000 construction jobs and support up to 100 permanent positions. Those figures are company and project-partner projections rather than independently realized outcomes, and Lambda said more information about the project timeline and expected economic impact will be released as planning advances.
The project adds another large computing development to a market where access to electricity is increasingly shaping site selection and project design. Microgrid Media has previously examined that shift in its coverage of large AI data-center campuses that are being planned alongside major power infrastructure rather than treating electricity supply as a later-stage utility connection.
Oklahoma’s ratepayer law is central to the power plan
Lambda said the project will comply with Oklahoma’s Data Center Customer Ratepayer Protection Act of 2026. The state law establishes protections intended to allocate the costs of serving qualifying large-load customers to the customers responsible for those costs. Oklahoma’s official legislative record shows that HB 2992 was approved by the governor, and the Oklahoma Corporation Commission’s Public Utility Division lists administration of the act among its responsibilities.
The law defines a large-load customer to include certain new data centers, cryptocurrency mining operations and artificial-intelligence computing facilities that contract for 75 MW or more of new load after July 1, 2026. Lambda has not disclosed the planned megawatt demand for the Mayes County facility, so the project’s exact treatment under each threshold and tariff provision cannot yet be independently determined from the information released Sunday.
The power-cost commitment also echoes a broader policy debate over whether households and smaller businesses should bear grid-upgrade costs created by very large computing loads. Microgrid Media recently covered the federal Data Center Ratepayer Protection Act, which addresses a separate federal proposal for large data centers. Oklahoma’s statute is already state law and should not be confused with that federal measure.
Flexible computing load could give the grid another operating lever
Lambda said it is working to make the facility a more flexible grid customer, with the ability to shift or reduce certain computing loads during periods of peak demand. The announcement does not quantify how much load could be curtailed, how quickly it could respond, or whether flexibility will be governed by a tariff, contract or grid-services program.
Those details will matter. Large data centers can create concentrated additions to electricity demand, but some computing workloads may also be more schedulable than traditional industrial loads. Grid operators are increasingly examining whether large-load flexibility can reduce reliability pressure during constrained periods. Microgrid Media’s coverage of PJM’s data-center load proposal shows how that question is moving from theory into formal grid-planning and emergency-response discussions.
Lambda also said the Oklahoma facility will use closed-loop cooling intended to reduce water consumption compared with traditional data-center cooling approaches. The company has not yet published a project-specific water-use estimate, energy-efficiency target, power-supply mix or on-site generation plan.
Major project details are still undisclosed
The September 27 announcement establishes the location, development partners and several operating commitments, but it leaves major engineering and commercial details open. Lambda has not disclosed the site’s planned megawatt load, electrical interconnection configuration, generation sources, backup-power strategy, construction start date or expected commercial-operation date.
Those disclosures will determine the ultimate grid significance of the Lambda Oklahoma data center. For now, the clearest power-sector takeaway is the project’s explicit combination of customer-funded infrastructure, full energy-cost responsibility and planned load flexibility, three issues that are becoming increasingly important as utilities respond to rapid growth in AI-related electricity demand.
