By Austyn Gaffney
Over 4 million individuals lost power Friday morning after the massive storm system known as Hurricane Helene struck Florida and progressed northward.
This storm exemplifies the growing susceptibility of utility companies to extreme weather phenomena, which are becoming more frequent and severe due to climate change.
“There are numerous indicators of climate-associated weather hazards to infrastructure,” remarked Catie Hausman, a public policy professor at the University of Michigan. These hazards encompass hurricanes and floods, wildfires, heatwaves, and heightened risks of tornadoes or unusual cold spells in regions that are less accustomed to such events.
Intense weather is increasingly taxing the electrical grid, making it the leading cause of significant power outages in the United States. In certain locations across the country, the likelihood of power outages caused by hurricanes may rise by 50% in the upcoming decades as these storms intensify.
Wind and precipitation are key factors that can burden power grids, according to Andrea Staid, a researcher focused on energy systems and climate analysis at the nonprofit Electric Power Research Institute. The institute’s projections indicate that as hurricanes increasingly impact the Gulf and Atlantic coastlines, the incidence of power outages could rise unless grid improvements are made.
“Hurricanes aren’t merely an issue for coastal areas,” Staid noted. “Our models indicate that these storms can extend well inland with formidable winds, so we must ensure that individuals and communities possess the information and resources needed for preparedness.”
Currently, to enhance predictions regarding incoming storms and to shift away from a fossil-fuel-dependent grid, utilities are exploring a new wave of technologies powered by artificial intelligence.
Hausman mentioned that although she is unsure if AI is the solution for reinforcing the grid, contemporary data and computational power are essential for grasping the challenges at hand.
“Whether AI provides us with innovative insights or becomes an unintelligible black box will depend on the provider and the tools they utilize,” Hausman stated.
What is evident? Utilities must invest substantially to modernize their power systems to effectively handle both current and impending storms.
“We are still conceptualizing the grid as we have for the last century, and it’s increasingly apparent that this approach must evolve,” expressed Mark Dyson, managing director of the carbon-free electricity program at the global research organization Rocky Mountain Institute.
“The convergence of extreme weather, decaying infrastructure, and emerging technology presents an opportunity to utilize improved technology, including AI-enhanced software, to help maintain an operational grid and manage costs,” Dyson remarked.
Among the emerging entities in this developing field is Rhizome, established in 2022, which employs AI technologies to assist electric utilities in pinpointing and preparing for vulnerabilities that may lead to power disruptions.
“Rhizome’s AI platform aims to deeply comprehend the connections between hurricane conditions and grid effects,” stated Mishal Thadani, co-founder of Rhizome. The firm’s data, derived from thousands of hurricane-related asset failures and their causes, can provide utilities insights into their long-term risks from hurricanes, facilitating decisions on where to reinforce poles, bury power lines, or trim vegetation.
“Ultimately, we’re capable of estimating how many future potential hurricane-induced outages can be reduced for each dollar invested by utilities,” Thadani explained.
Determining how to strengthen the grid and expand the national transmission network could decrease consumer costs, increase the availability of renewables, and lessen power outages, according to Hausman.
Many utilities are already implementing some form of machine learning or artificial intelligence technology, noted Booga Gilbertson, a former utility executive and current investor in Rhizome. Presently, the more eye-catching AI applications are language models such as ChatGPT from Microsoft or Gemini from Google. Machine learning has been analyzing an extraordinary volume of data for over a decade, according to Gilbertson.
“It feels like there’s a new player entering this domain almost every month,” she observed. “Products like these are quite new to the marketplace, and the rise of AI and computing capabilities has made them more accessible. This gives utilities a new tool in their arsenal.”
Artificial intelligence is also recognized as a significant energy user and could increase the nation’s electricity consumption by as much as 20% by the end of the decade. However, the United States, like other major nations, aims to be a leader in AI technologies. In September, the White House gathered a coalition of utilities, AI firms, and data center operators to strategize future developments.
While the national transmission framework faces challenges from rising electrification and an evolving generation mix, the expanding array of AI-enhanced products may potentially change AI’s unfavorable reputation regarding climate effects.
“I believe that the energy expended in this area could outweigh the costs in maintaining a reliable and affordable grid,” Dyson posited. “However, the evidence to back this hypothesis is currently nonexistent. The industry is still in its infancy, and we lack projections for which sectors could benefit most from AI technologies.”
