Back
Lots of Americans are electrifying their cars and homes, enticed by the prospect of lower bills, cleaner air, and less planet-warming pollution. But all that new electric equipment creates a serious challenge: It requires bigger, better infrastructure to manage the increased flow of electrons, from the electrical panels in individual buildings to the transformers and power lines that make up the grid at large.
Pacific Gas & Electric, California’s largest utility, is testing a one-two punch of technologies that could let it and customers sidestep those expensive upgrades. The first are devices from smart-electrical-panel startup Span, which plug into utility meters and control when and how a home uses power, avoiding the need for higher-capacity panels. The second are the latest digital controls from smart-meter vendor Itron, which can ensure that the collective power demands of multiple customers don’t push local grid transformers beyond their limits.
Working in concert, these technologies could help individual customers avoid thousands of dollars of upgrade costs to electrify their homes, said Quinn Nakayama, PG&E’s senior director of grid research innovation and development. And if deployed at scale, they could allow the utility to delay billions of dollars in grid upgrades, which should help reduce rates for all its customers, he said.
To be clear, PG&E isn’t promising those results right away. The pilot with Span will start by installing the company’s meter-connected devices at PG&E employees’ homes in the coming months, with a larger rollout to volunteer customers envisioned for 2027, Nakayama said. And PG&E will upgrade existing smart meters with Itron’s technology at about 1,000 homes this year; if they’re cost-effective, the utility may seek to incorporate the capability in hundreds of thousands of customers’ meters through 2030.
“Our service planners, when they interconnect new loads, always have to imagine the worst-case scenario,” Nakayama said. “This enables us to give them the tools and the assurances that those worst-case scenarios will never occur.”
PG&E isn’t the only utility looking for ways to meet growing electricity demand without blowing out its grid budget. Utility rates are on the rise across the U.S., in large part because of the increasing cost of maintaining and upgrading the poles, wires, and substations that deliver power to customers. But PG&E is under particular scrutiny from lawmakers, given its steep electricity rate hikes over the past decade.
Utilities also want to sell more power across their wires. The more they can expand capacity for EVs, heat pumps, and other power-using devices, the more money they can bring in to cover the cost of new infrastructure. This, in turn, eases upward rate pressure for customers at large.
One way utilities could sell more power over existing wires is by tapping the capacity of virtual power plants — collections of rooftop solar and battery systems, EV chargers, appliances, and thermostats that can be controlled collaboratively to reduce grid strain. In recent years, PG&E has run multiple VPP pilots with EV chargers, and it launched a project with Span, Sunrun, and other vendors in 2025 to test how smart electrical panels and solar-charged batteries that customers have already installed could relieve local grid constraints.
However, utilities are loath to rely on novel technologies to replace tried-and-true grid upgrades. If a VPP doesn’t work, for example, local transformers or neighborhood substations can overheat and break down under increased stress. That’s why PG&E’s latest experiment is covering its bases with devices that can control excess power use both at the home and on the grid.
To moderate home energy use, PG&E is using Span’s latest smart-electrical-panel device, which is designed to plug directly into utility meters. The Span device can actively monitor and control household circuits powering air conditioners, refrigerators, and clothes dryers, as well as EV chargers, heat pumps, and other more advanced energy systems.
Adding a major new power draw to a home, like an EV charger, often requires an electrical panel upgrade, which can cost thousands of dollars and add weeks to months to an installation. It can also trigger an upgrade to the local grid, which can take months to complete and cost anywhere from several thousand dollars for replacing a transformer on an overhead power line to around $50,000 for digging up and replacing underground service transformers and power lines.
“Nobody wants to pay that,” Nakayama said.
But those upgrades are predicated on the assumption that the new EV chargers will be drawing maximum power at the same time that all the other homes in the neighborhood are maxing out their electricity use, stressing their shared grid infrastructure. That’s usually during hot summer afternoons and evenings when air conditioners are running full tilt.
Span’s tech allows PG&E to offer those customers an alternative, Nakayama said: Let the smart device curb grid stress by reducing charging speeds during those peak hours. Most EVs require only several hours to recharge their batteries, giving them time to ease off on charging for a while yet still fill up overnight.
“I think most people are OK if their car charges a little bit slower, as long as it charges by 6 in the morning,” he said. That’s called managed charging, a concept that utilities across the country are exploring as they prepare to handle millions of new EVs coming online over the ensuing decades.
Span’s software also lets customers set other parameters to keep their total household electricity use below those limits, like delaying clothes dryers until later at night or easing off on air conditioning, Nakayama said. These kinds of technological solutions are going to be important for the more than 600,000 of PG&E’s roughly 5.5 million customers that the utility expects to need some kind of electrical service upgrade in the next 10 years to meet state electrification goals.
Span CEO Arch Rao said the company is working with other utilities interested in using its equipment for similar purposes. “A lot of the technical validation work has already been completed,” he said. “It’s now about customer recruitment and enrollment.”
So that takes care of individual homes. But how can PG&E ensure those controls are actually relieving local grid stress? That’s where Itron’s smart meter technology comes in, Nakayama said — or more specifically, Itron’s latest chipsets, which can be plugged into the smart meters that PG&E has already installed.
Like traditional utility meters, smart meters track a home’s electricity usage. But they use onboard computers and wireless networks to upload those readings to utilities, rather than requiring employees to come by to check the readings once a month. U.S. utilities have deployed nearly 120 million of these smart meters over the past two decades.
In utility parlance, smart meters are known as “advanced metering infrastructure,” or AMI. Older “AMI 1.0” technology can do some advanced tasks, like detect power outages and communicate via wireless networks with other meters and the utility. But it lacks the computing power and real-time capabilities to do more complex things, like actively communicate with and control devices in homes and businesses.
Enter Itron’s latest “AMI 2.0” technology. If AMI 1.0 is like a flip phone, AMI 2.0 is more like a modern smartphone, capable of uploading applications that can undertake the novel tasks that PG&E is now exploring.
In other words, “the meter is no longer just a meter — it’s a controller,” said Nick Tumilowicz, head of Itron’s distributed energy management solutions business. The company’s AMI 2.0 technology has already been controlling Level 2 EV chargers at hundreds of PG&E customers’ homes through a pilot project launched in late 2024, he said. Itron has used the same technology to manage school bus charging in New York City and Tesla Powerwall batteries for Colorado utility Xcel Energy.
Smart meters can also do something that in-home devices can’t, Nakayama said: communicate with all the other meters in the neighborhood to check how their shared electrical loads are impacting the transformers they’re connected to.
All those meters are linked in a wireless network and “speak the same language,” he said. Once an AMI 2.0 meter is connected, “it has the ability to say to its surrounding AMI 1.0 meters, ‘We’re all on the same service transformer,’” he said. “And it can do simple math, and figure out what that service transformer limit is,” as well as determine much demand the transformer faces from homes.
The tech then feeds that data back to the EV chargers and electrical panels that are linked to the AMI 2.0 meter, he said. For instance, if other nearby homes are using more power than usual and stressing the local transformer, PG&E could direct those smart panels and EV chargers to throttle power.
Finding ways for neighborhoods to electrify without crushing the grid will require a lot more solutions like these, said Ben Hertz-Shargel, global head of grid edge at research firm Wood Mackenzie.
“There is so much risk — and so much opportunity — on the distribution system. If electrification happens in an unmanaged way, it will be extremely expensive,” he said.
On the other hand, utilities have to make sure the technologies they’re deploying don’t add more costs than the benefits they deliver, Hertz-Shargel said. For example, PG&E’s new pilots are funded through state grants, and the utility will need to prove their cost-effectiveness before asking regulators to let it charge customers at large to deploy them more broadly as part of a rate case.
That evidence is particularly challenging to come up with when trying to avoid upgrades to the low-voltage network that brings power directly to houses, since most utilities don’t have solid details on that part of the grid.
“The problem is that utilities don’t have good data on these assets below the substation,” Hertz-Shargel said. “These devices need to not only solve the thermal overload problem but provide the ground-truth data to prove that they’re solving the problem — such as that the transformer stayed well below its power rating.” If that evidence is lacking, these technologies will be a harder sell to planning teams, he said.
“It’s smart for PG&E to try these different solutions,” Hertz-Shargel said. “I think the ones that survive will be the ones that are most cost-effective.”
.svg)
