Take from Inside Climate News written by Dan Gearino on August 14, 2025

If what happened on July 29 in California is any indication of the future of the power grid, then there is reason for optimism.

Let’s look at it from the perspective of Tom Weldon, who owns a house on the south end of San Jose and has rooftop solar and a battery system.

Weldon didn’t do anything out of the ordinary on that day. But his batteries, hooked up to software run by the solar company Sunrun, were part of a two-hour test to see the capabilities of home-based batteries when they work together to send power to the grid.

He could see the test was happening by looking at the Sunrun app on his phone. Otherwise, there were no changes in his home.

“When the actual event occurs, I don’t really see or hear anything different,” he said.

Sunrun said the July 29 event was likely the largest of its type ever held in the world, with more than 100,000 households sending small amounts of power to the grid which collectively generated 535 megawatts. For perspective, that’s enough power to meet about half the peak demand on a normal day for San Francisco, where Sunrun is based.

The effect was large enough that it showed up in a graph of demand for the California Independent System Operator, or CAISO, the grid manager for most of the state and part of Nevada. During the test period, which lasted from 7 p.m. to 9 p.m., demand for centralized power plants dropped by a noticeable amount, which reduced the need to burn natural gas or use other resources.

The test was a glimpse of what’s possible. For years, researchers have discussed the potential of virtual power plants, which are networks of home-based batteries that are connected by software, working together to function like a large power plant when needed. A Department of Energy report released in January, during the final days of the Biden administration, highlighted the potential.

“Virtual power plants are solutions that can be deployed at scale in a short timeframe to maximize the use and value of existing grid infrastructure, minimize costs to ratepayers, and ensure a resilient, reliable, and secure grid for all Americans,” the report said. 

The cost considerations are important because virtual power plants are reducing the need for natural gas peaker plants, which are expensive to build.

The July 29 test was the culmination of years of work for Sunrun, said Chris Rauscher, the company’s head of grid services.

“This is really a watershed moment when it comes to distributed power plants,” he said. “Folks saw in real time the impact that we had on the CAISO, on the CAISO demand, and saw that this is a reliable, dispatchable resource at a scale that is significantly larger than literally dozens and dozens and dozens and dozens of natural gas power plants in the state of California.”

Notice that he said “distributed power plants” instead of “virtual power plants.” Sunrun and others in the energy industry are trying to retire the word “virtual.” The idea is that the word “virtual” doesn’t do enough to convey that these systems function like real power plants.

I don’t think the word “distributed” is a great alternative, maybe because I’m used to talking about virtual power plants. But let’s not get waylaid by a quibble over nomenclature.

Rauscher noted that networks of batteries have been most effective for Sunrun in two very different markets, California and Puerto Rico. He views this as proof of the systems’ flexibility.

The main impediment to approving virtual power plants in other places is regulatory, with some states and grid regions not having a process in place to accommodate them. Utilities and regulators need to be on board.

In California, utilities were part of the test and have helped to facilitate Sunrun customers’ participation.

“This wasn’t just a tech demo,” said a news release from the utility Pacific Gas & Electric. “It was a real-world rehearsal for when [the batteries will] really be needed.”

Back in San Jose, I asked Weldon how he explains a virtual power plant to neighbors. His two-story house has solar panels on the roof, two Tesla Powerwall batteries inside and two EVs in the garage.

“Basically, the energy that I produce and store on the battery can be distributed to other houses,” he said. “It relieves strain on the grid. You don’t have to transmit as much electricity as far. It smooths out the bumps in the demand.”

Weldon, 71, works in technical sales for a company that makes memory for computer systems.

He said he is pleased to do his part to contribute to a cleaner grid. He also receives compensation for participating in the virtual power plant.

Sunrun gives each participating customer an annual payment of $100 or $150 for each battery. Since Sunrun uses a lease-based business model, it owns the batteries. Other companies have other business models for managing and compensating customers who are part of virtual power plants.

Sunrun’s contract with customers allows them to opt out of the virtual power plant, and it says that the local battery will retain enough power to serve the customer’s needs in case of a power outage later that same day.

The idea is that the software and the battery do the work, while people like Weldon won’t notice a difference unless they are monitoring the Sunrun app.

What’s the downside? Some consumers have concerns about the utility or some other company being able to remotely control a battery. I think the best way to deal with this is to have easy-to-navigate opt-out rules, with just a click or some other simple action needed.

The concerns, in my view, are minor compared to the opportunities. If every state had several gigawatts of virtual power plant capacity, it could be transformative for the grid’s ability to handle periods of high demand and do so in a much cleaner way than burning gas.