For more than a century, we have thought of power plants in much the same way. They are massive industrial complexes built behind fences, connected to distant communities by transmission lines, and fueled by coal, natural gas, uranium, falling water, or vast fields of renewable energy equipment. They require years of planning, billions of dollars in investment, and frequently become flashpoints for communities worried about pollution, water consumption, habitat destruction, safety risks, or rising utility bills. Whenever electricity demand begins to climb, the conversation usually starts with the same assumption: we need to build another power plant.
That assumption deserves another look.
In June, Sunrun, Tesla, and Renew Home announced a framework to coordinate home batteries, rooftop solar systems, smart thermostats, and other connected devices into what the companies describe as the largest distributed power plant in the United States. Together, their portfolios represent as much as 16.8 gigawatts of flexible capacity spread across the country.
The companies involved are notable, but they are not the most important part of the story. What matters is the idea their announcement demonstrates. Instead of responding to every increase in electricity demand by pouring more concrete, laying more pipelines, and constructing another industrial facility, a significant part of the solution may already exist inside American homes and neighborhoods.
A virtual power plant is not one facility operating at one location. It is a network of smaller energy resources coordinated through software so they can respond to the needs of the electric grid. A home battery may release stored electricity during a hot evening when air conditioners are pushing demand toward its peak. A smart thermostat may briefly adjust cooling by a few degrees. An electric water heater can shift when it reheats its tank, while an electric vehicle can charge later in the night rather than during the busiest hours.
One household may contribute only a small amount. When hundreds of thousands or millions of devices respond together, however, their combined effect can begin to resemble a conventional power plant. Some resources supply electricity, while others reduce demand. Both can help close the gap between how much power the grid has available and how much customers are using at a particular moment.
The announced 16.8 GW should not be interpreted as a single generator capable of producing its maximum output continuously. It is a geographically dispersed portfolio of flexible resources, including hundreds of thousands of batteries operated by Sunrun and Tesla and more than 8 million thermostats and connected devices managed by Renew Home. Those devices would need to be organized into regional programs, enrolled with customers’ permission, and dispatched where local utilities and grid operators can use them.
Even with those qualifications, the scale is remarkable. The Hoover Dam power plant has a nameplate capacity of about 2,080 megawatts. Comparing a hydroelectric facility directly with a virtual power plant is imperfect because the technologies operate differently, but the contrast helps illustrate how large a nationwide portfolio of 16.8 GW could become. Distributed energy is no longer confined to a handful of experimental pilots. It is entering the same conversation as major power infrastructure.
Traditional energy planning tends to treat homes and businesses as passive consumers. Electricity flows from a central power station through transmission and distribution lines until it reaches the customer. Virtual power plants challenge that one-way model. A building can consume electricity, generate it, store it, and adjust when it uses it. Coordinated across a community, those capabilities can provide reliability services that utilities once expected only from conventional generators.
This matters because the United States does not simply face a shortage of generating equipment. It also faces a problem of timing and utilization. Electricity systems must be designed to handle the highest periods of demand, even when those peaks occur for only a limited number of hours each year. Utilities may propose expensive plants, substations, and grid upgrades to serve those brief periods, with ratepayers covering costs for infrastructure that may remain underused much of the time.
Virtual power plants can reduce those peaks by using resources that are already connected to local distribution systems. They do not eliminate the need for all new infrastructure, but they can help defer or avoid some investments while making better use of equipment customers have already purchased. The U.S. Department of Energy has estimated that expanding virtual power plants to 80 to 160 GW by 2030 could address 10 to 20 percent of national peak electricity demand and save roughly $10 billion annually in grid costs.
Those savings could come from avoiding unnecessary generating capacity, postponing certain transmission and distribution upgrades, and reducing dependence on expensive peaker plants that operate during the most stressful hours. The Department of Energy has also reported that some utilities have launched basic virtual power plant programs in less than six months with relatively modest upfront costs, showing how quickly distributed resources can begin contributing when regulators and utilities make them a priority.
The environmental advantages are equally significant. A virtual power plant does not require a new smokestack or a fuel pipeline. Coordinating an existing battery or thermostat consumes no additional land and requires no cooling water. It can strengthen the electric grid without placing another industrial facility in a neighborhood that already carries a disproportionate pollution burden.
This does not mean the equipment itself has no environmental footprint. Solar panels, batteries, vehicles, electronics, and communications systems require raw materials and manufacturing. New devices also cost money, and their benefits will remain unevenly distributed if programs are available only to affluent homeowners. A credible virtual power plant strategy must include renters, multifamily buildings, lower-income households, community institutions, and residents who cannot afford to purchase batteries on their own.
That equity question should shape how the technology grows. Public incentives should not merely subsidize private equipment and then allow corporations to capture most of its grid value. Households that contribute their batteries, appliances, or ability to reduce demand should receive fair compensation. Programs should protect customers’ backup power needs, privacy, comfort, and right to leave. Communities should share in the savings when distributed resources allow utilities to avoid costly infrastructure.
Handled correctly, virtual power plants can shift investment away from another round of centralized projects and toward the people who actually provide the resource. Instead of asking ratepayers to finance a gas plant that generates returns for utility shareholders, part of that money can support batteries, efficient appliances, weatherization, demand flexibility, and backup power in homes and community facilities.
False Solutions has repeatedly challenged the claim that growing electricity demand automatically requires more fossil fuel infrastructure. In The Cleanest Power Plant Is the One We Never Build, we argued that utilities too often reach for construction before fully considering efficiency, conservation, demand management, and distributed energy. The virtual power plant announcement gives that argument a new scale. It shows that equipment spread across millions of buildings can become infrastructure when institutions are willing to treat it that way.
The issue also intersects with the rapid growth of energy-intensive computing. As discussed in The Cheap AI Myth Is Starting to Crack, the costs of new electricity generation, transmission, water systems, and grid upgrades are increasingly being shifted from large technology companies to ordinary customers. Communities are then told that new gas plants or other major projects are unavoidable because power demand is rising too quickly.
Virtual power plants cannot erase every impact associated with data centers or other large industrial loads, and they should not become an excuse for approving unlimited electricity consumption. They do, however, weaken the argument that combustion infrastructure must always be the first response. Before communities are asked to accept another gas plant, regulators should require utilities and large customers to demonstrate that they have seriously pursued efficiency, flexible demand, distributed storage, and locally available clean energy.
That is especially relevant in California, where communities are already pushing back against the land, water, and energy demands of large developments. As documented in The Data Center Backlash Has Arrived in California, residents are beginning to question why they should carry the environmental and financial consequences of projects whose benefits largely flow elsewhere. Distributed energy cannot answer every concern, but it offers a far less destructive starting point than building another generation of centralized infrastructure.
Virtual power plants are not a magic replacement for every generator or transmission line. Home batteries usually provide power for a limited number of hours. Their usefulness depends on location, customer enrollment, communications, utility rules, compensation, and the ability of grid operators to count on them when needed. Regional resources must also match regional needs. A battery in California cannot solve a local reliability constraint in Virginia.
These limitations are real, but they are not an argument for sidelining the technology. They are reasons to design better programs. Berkeley Lab has found that regulators, utilities, and virtual power plant providers already have a growing body of experience with program design, customer recruitment, market participation, and the integration of distributed resources. Its research identifies practical strategies for increasing enrollment and scaling these systems beyond small pilot programs.
The most useful question is not whether virtual power plants can do everything. No single resource can. The question is whether they can meet part of our electricity needs faster, more cleanly, and with less disruption than the conventional projects being proposed in their place. In many cases, the answer appears to be yes.
For decades, utilities have treated electricity demand as a reason to build something larger. The emergence of virtual power plants offers a different philosophy. Before constructing another industrial facility, use the batteries already installed in garages. Before expanding a substation solely to serve a short period of peak demand, reward customers for shifting when they use electricity. Before placing another polluting project in a frontline community, invest in the distributed resources that can make homes and neighborhoods more resilient.
The largest power plant in America may never be visible from the highway. It may have no cooling towers, no smokestacks, and no single address. It may simply be millions of homes working together, proving that a power system does not have to burn something, consume another piece of land, or sacrifice another community to keep the lights on.