Brookhaven National Lab research could help data centers expand nationwide
Brookhaven National Lab, in partnership with Amazon Web Services, is developing an AI-powered tool called GridSearch to identify optimal grid connection points for data centers, aiming to reduce development time and costs nationwide. This initiative occurs amidst significant local and state opposition to data center expansion, with New York Governor Kathy Hochul issuing an executive order to block large hyperscale data centers and the Brookhaven Town Board enacting an 18-month moratorium on AI data centers.
Brookhaven National Laboratory, located in Upton, New York, is spearheading the development of an AI-powered tool named GridSearch, designed to significantly expedite and reduce the cost of connecting new data centers and other large energy users to the electric grid. Hendrik Hamann, the lab's chief AI scientist, stated that current grid connection processes can take three to eight years and cost hundreds of millions of dollars, a major obstacle the tool aims to overcome. GridSearch, developed in partnership with Amazon Web Services, will analyze billions of grid scenarios to identify optimal connection points, thereby bypassing the need for costly infrastructure upgrades and accelerating data center expansion nationwide.
This technological advancement emerges amidst growing local and state opposition to data center development, particularly on Long Island. Concerns range from noise pollution and environmental impact to potential increases in utility bills for ratepayers. New York Governor Kathy Hochul recently issued an executive order prohibiting the permitting of hyperscale data centers exceeding 50 megawatts, reflecting a statewide effort to regulate their growth. Concurrently, the Brookhaven Town Board unanimously enacted an 18-month moratorium on AI data centers, with Islip town officials also considering a similar pause.
While proponents like Hamann emphasize GridSearch's potential to create a more efficient electrical grid and save developers and ratepayers money, critics, such as Philip Odonkor from Stevens Institute of Technology, caution that such tools could accelerate development in areas with less informed or empowered residents. Eric Hittinger of Rochester Institute of Technology also noted that Long Island's high electricity costs and dense population make it less ideal for data centers, suggesting they might gravitate towards less populated areas with less organized opposition. The lab, however, stresses that AI data centers do not need to be physically located on Long Island to serve its residents, highlighting the national scope of the GridSearch tool.