California’s statewide power grid operator is poised to become the first in North America to deploy artificial intelligence to manage outages, MIT Technology Review has learned.
“We wanted to modernize our grid operations. This fits in perfectly with that,” says Gopakumar Gopinathan, a senior advisor on power system technologies at the California Independent System Operator—known as the CAISO and pronounced KAI-so. “AI is already transforming different industries. But we haven’t seen many examples of it being used in our industry.”
At the DTECH Midwest utility industry summit in Minneapolis on July 15, CAISO is set to announce a deal to run a pilot program using new AI software called Genie, from the energy-services giant OATI. The software uses generative AI to analyze and carry out real-time analyses for grid operators and comes with the potential to autonomously make decisions about key functions on the grid, a switch that might resemble going from uniformed traffic officers to sensor-equipped stoplights.
But while CAISO may deliver electrons to cutting-edge Silicon Valley companies and laboratories, the actual task of managing the state’s electrical system is surprisingly analog.
Today, CAISO engineers scan outage reports for keywords about maintenance that’s planned or in the works, read through the notes, and then load each item into the grid software system to run calculations on how a downed line or transformer might affect power supply.
“Even if it takes you less than a minute to scan one on average, when you amplify that over 200 or 300 outages, it adds up,” says Abhimanyu Thakur, OATI’s vice president of platforms, visualization, and analytics. “Then different departments are doing it for their own respective keywords. Now we consolidate all of that into a single dictionary of keywords and AI can do this scan and generate a report proactively.”
If CAISO finds that Genie produces reliable, more efficient data analyses for managing outages, Gopinathan says, the operator may consider automating more functions on the grid. “After a few rounds of testing, I think we’ll have an idea about what is the right time to call it successful or not,” he says.
Regardless of the outcome, the experiment marks a significant shift. Most grid operators are using the same systems that utilities have used “for decades,” says Richard Doying, who spent more than 20 years as a top executive at the Midcontinent Independent System Operator, the grid operator for an area encompassing 15 states from the upper Midwest down to Louisiana.
“These organizations are carved up for people working on very specific, specialized tasks and using their own proprietary tools that they’ve developed over time,” says Doying, now a vice president at the consultancy Grid Strategies. “To the extent that some of these new AI tools are able to draw from data across different areas of an organization and conduct more sophisticated analysis, that’s only helpful for grid operators.”
Last year, a Department of Energy report found that AI had potential to speed up studies on grid capacity and transmission, improve weather forecasting to help predict how much energy wind and solar plants would produce at a given time, and optimize planning for electric-vehicle charging networks. Another report by the energy department’s Loan Programs Office concluded that adding more “advanced” technology such as sensors to various pieces of equipment will generate data that can enable AI to do much more over time.
In April, the PJM Interconnection—the nation’s largest grid system, spanning 13 states along the densely populated mid-Atlantic and Eastern Seaboard—took a big step toward embracing AI by inking a deal with Google to use its Tapestry software to improve regional planning and speed up grid connections for new power generators.
ERCOT, the Texas grid system, is considering adopting technology similar to what CAISO is now set to use, according to a source with knowledge of the plans who requested anonymity because they were not authorized to speak publicly. ERCOT did not respond to a request for comment.
Australia offers an example of what the future may look like. In New South Wales, where grid sensors and smart technology are more widely deployed, AI software rolled out in February is now predicting the production and flow of electricity from rooftop solar units across the state and automatically adjusting how much power from those panels can enter the grid.
Until now, much of the discussion around AI and energy has focused on the electricity demands of AI data centers (check out MIT Technology Review’s Power Hungry series for more on this).
“We’ve been talking a lot about what the grid can do for AI and not nearly as much about what AI can do for the grid,” says Charles Hua, a coauthor of one of last year’s Energy Department reports who now serves executive director of PowerLines, a nonprofit that advocates for improving the affordability and reliability of US grids. “In general, there’s a huge opportunity for grid operators, regulators, and other stakeholders in the utility regulatory system to use AI effectively and harness it for a more resilient, modernized, and strengthened grid.”
For now, Gopinathan says, he’s remaining cautiously optimistic.
“I don’t want to overhype it,” he says.
Still, he adds, “it’s a first step for bigger automation.”
“Right now, this is more limited to our outage management system. Genie isn’t talking to our other parts yet,” he says. “But I see a world where AI agents are able to do a lot more.”
Read More
Alexander C. Kaufman