Below-average temperatures in California this summer have reduced demand and made electric grid operations uneventful so far, with the state reaching 40,000 MW of demand for the first time in July, CAISO CEO Elliot Mainzer said during a July 23 meeting of the ISO’s Board of Governors. 

Conditions have been favorable for grid operations across CAISO’s footprint: average temperatures in June were near normal, while temperatures in the first part of July were below normal, Mainzer said. 

“To [board] Chair [Severin] Borenstein, I say we are about halfway through 2025 and your attire today is indicative to a certain degree of the type of summer we are having in the Bay Area … a Mark Twain summer,” referring to the famous quote often misattributed to the author: “The coldest winter I ever spent was a summer in San Francisco.”  

In general, CAISO has avoided any major disruptions this year, but the region has August and September to go — the months that tend to offer the greatest grid challenges, Mainzer added. 

Although electricity demand has been lower than average, this year’s wildfire season is well ahead of last year’s pace in both frequency and severity, Mainzer said.  

As of July 1, 3,938 wildfires had occurred in California, compared to 3,339 over the same time last year. Acres burned is up substantially too: 182,497 acres this year so far compared to 76,152 last year. The five-year average for this time of the year is about 46,000 acres. 

California is set to meet electricity demand this summer under normal conditions, but in a worst-case scenario, the state could need to use more than 2,600 MW of contingency resources, according to a CAISO report in May. Wildfires outside the state could reduce import capacity by as much as 4,000 MW, the report says. 

AI Pilot Program Begins

Mainzer said also that CAISO is implementing an AI pilot program — called Genie — to support control room operations. The program will specifically help with facility maintenance requests and act as a “co-pilot” tool to enhance efficiency, Gopakumar Gopinathan, CAISO senior adviser of power systems technology, said in an email to RTO Insider. 

Genie can detect anomalies in maintenance requests and identify which transmission paths may be affected under certain scenarios. The AI program relies on historical data, operating procedures and related information sources to deliver clear, context-rich insights to operators, Gopinathan said. 

Genie is not meant to replace human decision-making tasks, Gopinathan said. 

“While the AI prototype can identify anomalies in maintenance requests, extract operational insights, and recommend next steps, all decisions remain under the direct oversight and authority of CAISO operators,” he said. 

Genie is being developed with Open Access Technology International (OATI) as part of CAISO’s control center modernization efforts and is not, at this time, implemented in a real-time environment, he said. 

“For now, the AI remains in testing mode, and no recommendations generated by the AI are being used in live operations,” Gopinathan told RTO Insider. “If the pilot meets performance expectations, this technology is expected to significantly enhance the support available to system operators, particularly in tasks that are repetitive and time consuming.” 

When asked if Genie will be used to help control room operators make real-time decisions about electricity flow on the grid, Gopinathan said the AI currently does not perform this task.