As US grid batteries surge, risk of fires plummets

See more from Canary Media’s ​“Chart of the Week” column.

Battery fires are like shark attacks: frightening events that loom cinematically large in the American psyche and almost never actually happen.

As batteries have become America’s top choice for new on-demand grid power, community concern about their fire risk has emerged as a limiting factor on the technology’s growth. There’s no disputing that battery fires have happened, and have been nigh impossible to put out by conventional means. But a dataset compiled by the nonprofit Electric Power Research Institute shows that these fires strike in the U.S. even less frequently than sharks.

A recent EPRI report plots global failures of commercial and grid-scale batteries — including arc flash, fire, and explosion events — and shows that the annual rate of failure has fallen precipitously as installations have grown exponentially. New safety standards have taken effect since the first big batteries were installed, and the batteries themselves, as well as project layouts, have grown more sophisticated and safe over the years.

That’s a great sign for an emerging technology: Even as a lot more of it enters the world, far fewer fiery failures are happening. But there’s a lot of variance from country to country in terms of rules around energy development and safety standards. To see how the data looks for the U.S. in particular, we plotted EPRI’s failure data alongside a tally of commercial- and utility-scale battery installations from the Solar Energy Industries Association in partnership with Benchmark Mineral Intelligence.

By 2025, the annual failure rate fell by more than 97% from the 2016 rate, while cumulative deployments rose 250-fold. For three different years over the last decade, the failure rate was zero, because no known failures occurred those years. That looks like a gap in the downward trend, but it’s indicative of uneventful operations.

This dataset has its limitations, though. For one thing, it’s a small sample size. 

“There just aren’t a lot of failures, which is a good thing, but that means you don’t have a very robust dataset to analyze,” said Lakshmi Srinivasan, lead author of the EPRI report.

The raw tally of battery failures also does not capture the size and severity of the events. Some are short-lived affairs at diminutive batteries. One, outside Phoenix in 2019, sent four emergency responders to the hospital with serious injuries — the only significant bodily harm caused by a U.S. grid battery fire thus far. And then there’s California’s Moss Landing, one of the first really big battery installations to go up in a massive blaze, in January 2025; it didn’t injure anybody but largely destroyed the power plant and prompted an evacuation of the surrounding community for multiple days. 

Moss Landing pops up as an example when communities across the country weigh in on new battery development. However, EPRI calculated that grid battery units now fail at a rate comparable to oil-filled transformers, the bulky pieces of equipment that sit at substations to convert electricity between higher and lower voltage. Those transformers are ​“very well understood technology that’s been deployed on the grid for decades, and so it’s something that both the grid operators and industry is familiar with,” Srinivasan said. At this point, they’re essentially ​“invisible” to the public.

“We can say safely that battery projects are receiving a lot of public scrutiny in ways that I don’t see currently, at least, transformers or substation buildouts receive,” she said.

The data could help communities better calibrate the risks of hosting batteries, and weigh them against the benefits, like tax revenue, enhanced grid reliability, and downward pressure on electricity costs at a time when many people are worried about energy affordability.