Neel Somani Examines Grid Fairness as Data Centers Scale Across U.S. Power Markets

Electricity demand in U.S. power markets is undergoing a significant transformation due to the proliferation of data centers and AI infrastructure, according to researcher Neel Somani. This shift from traditional industry to digital technology is remaking the electricity distribution system and prompting critical discussions about how grid costs and benefits are allocated among consumers. Somani's research emphasizes the need for updated thinking in power markets regarding infrastructure cost allocation, particularly given the rapid scale of this expansion. Data centers represent one of the fastest-growing electricity loads, making them a central topic for energy executives and the public alike. Regional grid operators, utilities, and regulators are tasked with integrating these large loads into wholesale markets, as consumers and governments face the challenge of rising energy prices and environmental considerations. The demand for new data centers frequently necessitates costly infrastructure upgrades. Somani's analysis delves into how technological advancements, from faster processors and better storage to the rise of cloud computing and artificial intelligence, are directly influencing the evolution of power systems. AI applications, such as speech recognition and language translation, rely on massive computing clusters that place unprecedented demands on regional electricity grids. A key debate centers on whether hyperscale data centers, which consume enormous energy, should bear a larger portion of grid expansion costs. Somani also explores the potential for flexible loads, like AI training clusters, to help stabilize regional power markets by balancing supply and demand. Globally, data centers accounted for about 1.5 percent of electricity consumption in 2024, growing 12 percent annually over the past five years. Somani's ongoing research contributes to discussions on the environmental impacts of AI, including energy and water usage, carbon footprint, and hardware waste, by examining how grid policy and infrastructure planning can adapt to a rapidly expanding digital economy while ensuring fairness for all stakeholders.