The Next Power Plant Is On The Roof And In The Basement
A Department of Energy report promotes a new system that could remake the energy grid.
On any given Monday in Vermont, Josh Castonguay, the vice-president of innovation at that state’s Green Mountain Power utility, told me, he studies the forecast for the days ahead, asking questions like “What’s it looking like from a temperature standpoint, a potential-of-load standpoint? Is there an extremely hot, humid stretch of a few days coming? A really cold February night?” If there is trouble ahead, Castonguay prepares, among other things, Vermont’s single largest power plant, which isn’t exactly a power plant at all—or, at least, not as we normally think of one. It’s an online network, organized by the utility, of forty-five hundred electric storage batteries (currently, most of them are Tesla Powerwalls), spread out across more than three thousand Vermont homes. The network also includes a broad array of residential rooftop solar panels, which produce the energy stored in those batteries, and smart water heaters and E.V. chargers. The people who have these assets aren’t off the grid; they’re Green Mountain Power customers who, for a discount on their bills, agree to plug their batteries (most of which are leased to own) and appliances into the utility’s network and let the company control the devices so that they use less power at critical moments. (If customers need to override the company’s commands, they can.) This means that Castonguay (or, really, his algorithms) can program storage batteries to be charged a hundred per cent before a storm hits. Or, if it’s going to be a hot day, he can preheat water heaters in many homes in the morning, so that in the afternoon, as the temperature rises, more power will be available to run air-conditioners. He can also precool some big buildings in the morning. “Then, if you think about it, the building itself is the battery,” he said, in the sense that it stores chilled air for later in the day. “We have about fifty megawatts” of this distributed power, Castonguay told me. “At the scale of Vermont, that’s a lot.” Utilities have always been able to dispatch supply, bringing power plants, which are often in idle mode, online as demand requires. Now they’re increasingly able to call up small, individual home power plants and dispatch demand as well, turning down thermostats or delaying car charging.
Green Mountain Power is at the forefront of this push; last month, it announced plans to install storage batteries for many of its customers—two hundred and seventy thousand homes and businesses, in total—in the next decade, pending regulatory approval. (Castonguay says it is testing a new home battery system, from FranklinWH—a company named for Ben Franklin, who actually coined the term “battery”—and that this apparatus seems to work as well as the Powerwall.) But other companies are starting to follow. Green Mountain Power’s former C.E.O., Mary Powell, left three years ago and soon took over Sunrun, which supplies rooftop solar panels and storage batteries for hundreds of thousands of homes nationwide, and serves as a third-party power aggregator for several utilities. “We’re sitting on more than 1.1 gigawatt-hours of installed storage capacities just with our customers now,” she told me recently, much of it in California, where the company is based. From August through October, as a series of heat waves pushed consumption up in that state, Pacific Gas and Electric was buying up to thirty megawatts of power through Sunrun every evening to keep peak demand down in its grid system. Sunrun’s customers who provided the energy got a check for seven hundred and fifty dollars. “We went from contract to operation in six months,” Powell said. “You simply could not get a resource of that size built and operationalized any other way in that time frame.” And, she added, “it’s not just that we can make a more reliable, resilient grid” by drawing on the scattered resources; “We can also make a much more affordable grid,” because being able to use residential power means not having to build big, new power plants to meet peak demand. Taking in that money saved, she added, “We can shave ten billion dollars a year off the price of the country’s power system.”
That figure comes from a report that the Department of Energy released in September on virtual power plants (V.P.P.s), the name the industry has settled on for these systems, though not without dissent: “There’s really nothing virtual about them,” Powell said. The report’s lead author, the D.O.E.’s portfolio manager Jennifer Downing, predicted that V.P.P.s could be handling twenty per cent of peak power demand across the country by 2030, at a cost forty to sixty per cent below that of the plants that would otherwise have to provide that power. As the number of E.V.s—which are basically large batteries attached to wheels, and so can be plugged into the system to provide power or extra storage—grows, that percentage could rise. Jigar Shah, the polymathic enthusiast running the department’s efforts to deploy the hundreds of billions of dollars in federal money provided by the Inflation Reduction Act (I.R.A.), told me, “Today, our grid is used at about forty per cent of capacity. If we could fluctuate demand as we now do supply, we could get a lot more out of it.” In the process, he added, customers could save twenty per cent on their power bills.
It’s worth thinking for a moment about when America’s electric system first came online. It began with Thomas Edison’s urban experiments in the late nineteenth century and expanded rapidly across the country in the early to mid-twentieth century. That’s roughly the same period that America’s broadcast communications system grew from a few local trials in radio to continent-spanning broadcast networks. Timing isn’t the only similarity: they both also ran one way. A few giant utilities (such as Consolidated Edison) produced electricity at mammoth power stations, and ran the power down lines to homes and businesses, just as a few broadcast giants (CBS, NBC, and ABC) produced content at a handful of centralized locations, and transmitted it over the airwaves into America’s living rooms.
Then, with the advent of the Internet, every person became a potential content producer and was connected laterally to everyone else. Now the electric grid is belatedly starting to follow that model, with millions of homes and businesses becoming energy suppliers and storage nodes. All this is coming at a crucial moment, as the demand for electricity is expected to soar. Fighting climate change means electrifying almost everything and providing the electricity required as cleanly as possible; making household batteries and appliances part of the system should help ease what’s likely to be one of the largest industrial transitions in U.S. history. Electric utilities are often regarded as the stodgiest of businesses, because their profit is guaranteed and also capped (public-utility commissions in each state set rates based on the costs to utilities, plus some percentage). And the bottom-line imperative from the beginning of the industry has been reliability. That has led to a culture of conservatism. “Largely, you still have a cultural system in grid operators where anything new is looked at with skepticism,” Powell said, “with ‘Well, take ten years to prove it to me.’ ” It’s also clear that networked clean energy that’s produced in private homes and businesses could cut into the companies’ profits. A consultant for Arizona Public Service, that state’s main utility, confirmed to regulators at a state hearing in September that the increasing growth of rooftop solar panels “may supplant investments that the utility may otherwise have been able to make, and earn a return on, for the benefit of their shareholders.”
But recent developments are starting to change that equation, most of them stemming in some way from the convulsive effects of climate change. Since power supply produces about a third of the country’s greenhouse gases, most jurisdictions have enacted laws that call on utilities to generate less of their power from fossil fuels. Meanwhile, mandates from states and the federal government designed to cut emissions from the transportation and industrial sectors are pressuring utilities to provide more power for, say, electric Ford F-150 Lightning pickups and the electric-arc steel furnaces that are starting to replace fossil-powered foundries. The electrical grid needs to be modernized and expanded, and the cost of that means that electricity rates are going up fast—so fast that twenty million American households, about one in six, are currently behind on their utility bills, the highest number on record. Last year, in Maine, the state’s two private utilities cut off ten per cent of their customers for nonpayment, which helped spark a referendum campaign to take those companies public. (The Maine chapter of Third Act, which I helped organize, worked to pass the initiative on the ballot earlier this month; the utilities outspent the advocates thirty-five to one, and defeated that effort.)
Unprecedented heat waves are also straining systems—Texas was on the edge of rolling blackouts much of last summer and, in August, during one peak usage period, more than eleven gigawatts of gas and nuclear power were unexpectedly knocked offline. “Luckily for Texans,” Powell wrote, “renewable energy resources such as solar and storage delivered energy when it was needed most. This story, in which solar and storage prevent grid failure, will become more common in the coming years.” In Vermont, ever-stronger storms are pushing the utilities: Green Mountain Power wants all its customers to have a battery because it spent fifty-five million dollars on storm recovery this year, up from an average of less than ten million from 2015 to 2022. “Our three worst storms were this year,” Castonguay said.
“Based on what I’ve seen with utility contracts this year, I think we’ve arrived” at a place where “utilities are coming to us,” Suleman Khan, the C.E.O. of Swell Energy, told me. Like Sunrun, his company is also based in California, and it also aggregates household power and storage, and sells the energy to utilities as dispatchable power. It built a V.P.P. in Oahu—Khan said that Hawaiian Electric asks Swell to dispatch power from household batteries when the grid needs it, which is about ten times a month. As a result, those customers now get a payment from Swell. “We will lease you a battery and solar panels, for, say, two hundred bucks a month.” Khan said. “And now we can pay you maybe fifty dollars a month for the ability to dispatch your power. It’s bringing down the cost of ownership, which is good for well-off people, but decisive for people who really care about the difference between a hundred and fifty and two hundred dollars a month.”
Utilities are now building their own big battery farms out near their big solar or wind installations, which is helpful—studies show that there aren’t enough rooftops in the U.S. to hold the solar panels needed to run a fossil-free power system. But power generators and batteries situated closer to the point of consumption are uniquely useful, Khan said. “One thing is transmission losses. If you can generate and store where you consume, you avoid the five- to ten-per-cent generation losses that come with running power down a long line. And, if you’re trying to cope with load growth on a particular circuit, where, say, a lot of new housing is going on, a compelling way to do that is to have batteries around on that circuit.”
But when utilities really engage they can dramatically speed up the pace at which households can supply more power. For companies like Swell and Sunrun, one of the biggest costs is customer acquisition. They have to buy ads, or hire people to cold-call potential clients, and it’s hard for consumers to separate a good offer from a bad—or even fraudulent—one. Some utilities are now actively helping the aggregators acquire customers. “Even in cases where regulations make that hard, at a minimum their endorsement really matters,” Khan told me.
And, as these partnerships expand, so do the opportunities for financing. Swell recently landed a hundred-and-twenty-million-dollar round of venture capital, in part, Khan said, because a lot of investors are more interested in financing V.P.P.s, which in essence are more like utilities—an asset class with a much longer history, and a more predictable payback—than they are like pure solar installers, which are more like a startup. The fact that the Department of Energy, with money from the I.R.A., offers loan guarantees to these companies adds another layer of comfort for investors, Khan said. “It’s a strong signal to the marketplace,” he added. Strong enough that “I think over the next thirty years every home will have a battery. I don’t know if every home will have solar—some have too much shade, or the wrong roof. But batteries are going to become an appliance, just like a dishwasher or a television.”
He predicts particularly rapid growth in the southeastern U.S., where traditionally low electric rates and obdurate utilities have slowed the home-solar industry. In recent months, Duke Energy, which once helped convince regulators that an environmental group that had installed some solar panels on the roof of a Black church and was selling it power below what the utility charged should not be allowed to do so, announced plans to enroll thirty-five hundred customers in a trial that would let the utility call on their batteries thirty-six times a year, in exchange for discounted rates. “Utilities have an opportunity to be the heroes of this story,” Shah, at the Department of Energy, told Solar Power World last month. “By supporting increased electrification, they’re enabling consumers’ home efficiency upgrades, enabling economic growth, reducing emissions from enterprises and more.”
Advocates say that a good example of this trend is Portland General Electric (P.G.E.), in Oregon, whose C.E.O., Maria Pope, was part of the D.O.E.’s advisory board as it was preparing the September report. In mid-August, the Oregonian reported, a three-day heat wave broke “daily, monthly, and decades-long records.” But it didn’t break the P.G.E. grid, in part because the company has a long-standing commitment to its own virtual power plant, with about two hundred megawatts’ worth of solar panels, batteries, and thermostats in customer’s houses that it can call on. P.G.E.’s software went to work precooling houses in the morning so that it could turn thermostats three degrees lower during the afternoon peak, and introducing the staggered charging of electric cars so that they could wait till the wee hours to refill their batteries, once the strain was off the system. All this is voluntary—customers sign up to participate, and if they need to override the thermostat they can. “But about twenty per cent of our customers are bought in,” John Farmer, the utility’s spokesperson, said. “We saw at least ninety-megawatt reductions in peak usage three days in a row.”
In fact, if utilities don’t move fast enough, they may find themselves not with new opportunities but with new competition. In October, in Santa Barbara, a team of local environmentalists announced plans for a thirty-megawatt V.P.P. that one of them, Leah C. Stokes, a key architect of many of the climate provisions in the I.R.A. and a professor of environmental politics at the University of California campus, called “a step towards having the community run one hundred percent on clean power 24/7.” She added, “Right now, Santa Barbara is at the end of a transmission line—they call it the ‘Goleta Load Pocket.’ And, hence, we’re highly vulnerable to natural disasters disrupting electricity access.” In 2017, an enormous wildfire almost cut the region’s connection, which depends on three transmission lines coming through a single hundred-and-forty-foot-wide corridor that traverses mountain and forest. So the goal is to create a system that runs mostly on locally generated solar power, and has the capability to isolate buildings “from the broader grid in the event of an emergency. Individual homes can already do this with solar and batteries,” Stokes told me. “But our vision is much bigger. We want to build a large project that can protect the community and supply power locally, especially during emergencies.”
Working with yet another aggregation startup—Scale Microgrids—the group has mapped warehouses and parking lots with fifty thousand square feet of potential solar exposure, and are aiming to invest a hundred million dollars in venture capital to build out a V.P.P. that will “serve as a model for cities around the world.” Stokes says that many of her neighbors want to “do bold things to act on climate” and that they also know they’re “on the front lines with drought, fires, and extreme rainfalls leading to mudslides.” That combination of utopian hopes and dystopian fears may boil the climate crisis down to its essentials—and the sum of those hopes and fears may look very much like a battery in the basement and a panel on the roof, magically connected to thousands of others.