Ditching the combustion engine for an electric vehicle is a good first step for cutting transportation emissions. But it’s becoming increasingly clear that owning an electric car on its own is not enough. When and how you charge the car makes an enormous difference, not only for reducing CO2 emissions, but also for helping the power grid withstand the coming electrification wave.
We know that not all charging is created equal. Location, for example, is an obvious difference-maker. In places with ample renewable energy such as hydro-dominated Washington or solar California, electric vehicles produce vastly less climate pollution over their lifetimes than gasoline cars. In places with fossil-fuel-heavy grid, the climate benefit is still there, but much smaller.
The matter of when to charge is, similarly, about aligning EV charging with the supply of renewable energy. As Heatmap has noted before, it makes sense for solar-heavy states to encourage EV owners to charge at midday when clean energy generation peaks — that would help to level out California’s duck curve rather than make it worse. That’s easier said than done, though, since not everyone’s workplace has electric vehicle chargers. Besides, the simplest form of the EV lifestyle is to plug in upon returning home from work and errands in the evening, the very moment when electricity use spikes and solar energy is dropping off for the day.
Charging’s place and time are both important for maximizing the climate good EVs can do. They are also matters of growing importance for electric utilities that must learn how to balance the coming acceleration in electricity demand without seeing their costs spiral out of control. According to new research by the group Energy and Environmental Economics, smarter ways to optimize the when and the how of EV charging could save them an enormous amount in upgrade costs.
E3’s researchers ran case studies, including one that modeled the EV-heavy territory of Southern California Edison, to find out how different approaches to widespread EV charging affected how much extra costs the utilities incurred. The researchers considered three approaches to charging. In the first, “unmanaged,” drivers plug in as soon as they get home and the vehicle charges until full. In the second, a “passive managed" scenario, the EV doesn’t necessarily charge to full immediately, but instead waits until off-peak hours when the price of electricity drops. The third, “optimized,” used Rhythmos.io’s software to imagine a system wherein a car can detect the exact moments to charge to place the least strain on the grid.
The differences were stark. E3 used California’s official Avoided Cost Calendar to measure the added costs to SCE under each scenario. Whereas unmanaged charging cost the utility $984 per EV added to the system, optimized charging dropped that figure to just $407, a 60% reduction. (The middle-ground scenario came in at $686.)
Much of these savings are attributable to avoiding the wear and tear and possible overloads that electrical transformers would suffer in a world where everyone tries to charge their EVs all at once. (The transformers that form that backbone of the power grid are rated to specific currents and voltages they cannot safely exceed, which is one of the limiting factors on how much the system can handle.) It’s a particularly pressing matter in this age of transformer shortages, when it can take years to get a replacement for a broken or outdated one.
Although the financial and resilience benefits of optimized EV charging are clear in E3’s findings, they’re far from simple to achieve in the complex moment-to-moment reality of the grid. E3 study coauthor Eric Cutter told me it starts with communication — utilities could give EV drivers a forecast a day in advance, for example, telling them when clean energy will be in good supply and prices will be low.
“They could say, ‘Tomorrow is a sunny day, so please charge during the day,’ or, ‘Tomorrow is a cloudy day, and it happens to be very hot and humid, so the air conditioners are going to be ringing, so please don't charge in the evening and charge late at night,’” he says. “And they could make that determination each day as to what's going to be the most beneficial for the system.”
But much of this work will be automatic and algorithmic. For optimized charging to work, all drivers have to do is leave their EV plugged in and be okay with whenever the system decides to send them electricity. The software will decide which cars get which levels of charge, and when, to minimize strain on grid infrastructure.
That raises another question about trust. People who don’t like the local power company — and there’s a lot of them — might not want to allow that entity to decide when their EV gets to charge. They also might not trust that they’ll have enough battery range when they need it. To combat the first issue, Cutter said, perhaps drivers will sign up for a charging management system run through their car’s manufacturer, since drivers often have a better opinion of Honda or Ford than they do of their utility. And to fight the range anxiety problem, he says, some pilot programs have given customers a button to opt out of optimized charging.
“What the programs have found out is that customers want the button, but they never use it. It's very, very rare,” he says.
The number of EVs in America, especially in markets outside California, has yet to reach a point where a smarter way to charge has become a necessity. Although their sales share is rising, EVs accounted for just 8.1% of cars sold in 2024; only California has seen the energy demand from electric vehicles exceed 1 million megawatt-hours, though the numbers are rising fast. Even with EVs and electrification facing stiff political headwinds, utilities across the nation are already at work on plans to handle the influx of EV demand.
“Ten years ago when we were talking to utilities, a lot of them would say, ‘We're not worried about EVs. Come back to me when that's 5% of adoption or 10% of load.’ But not anymore. I don't think utilities anymore are waiting until that level of adoption to start thinking about how they need to plan for them.”
