25 August 2022 |

Virtual power plants power up


Last week, a ‘virtual power plant’ system that aggregates energy storage from distributed energy resources, like the Tesla Powerwall, successfully flowed electricity back to the grid in California during a peak period where the grid was constrained. A group of more than 2,000 Tesla Powerwall owners supplied as much as 16 MW during a period of peak load for the grid

To get up to speed, the virtual power plant concept refers to distributed energy resources working together to provide electricity back to the grid. Tesla’s Powerwall is a lithium-ion battery energy storage system for homeowners to stockpile solar energy for backup power (e.g., for use in the event of a blackout). They can also help their owners save money when electricity prices are high and can provide electricity back to the grid, a la a virtual power plant.

This ability of energy storage systems to flow electricity back to the grid is an important decarbonization tool, not just a way to prevent blackouts or reduce prices. Natural gas plants are the default choice for extra electricity generation during peak load because of how quickly they can be ‘turned on’ to generate more electricity. Plants dedicated to meeting peak load are called ‘peaker plants.’ By using stored energy produced by renewables when demand is high, virtual power plants can displace natural gas, or in some cases even coal, usage, and their resulting greenhouse gas emissions.

Virtual power plants power up 

The ‘virtual power plant’ name makes what happened in California sound harder to understand than it is. It isn’t that complicated: Energy storage systems store electricity produced via renewables and deploy it back to the grid when a) their buildings don’t need additional electricity and b) there’s demand for it elsewhere.

Considering that ~2,000 homes participated in the virtual power plant event in California last week, generating 16 MWs at peak, it’s enticing to imagine what this looks like at scale. There are more than (roughly) 10M homes in California

Before you do the simple math of extending the virtual power plant example to its logical extreme, it’s important to note that coordinating these distributed resources isn’t easy, and the technology required isn’t cheap. One Tesla Powerwall, for instance, costs more than $10,000. While it may well pay for itself over a lifetime of use, that’s still a high upfront capital expenditure.

Similarly, as we explored last week, inputs for battery energy storage aren’t limitless and come with environmental costs. Further, tech and infrastructure upgrades are often needed to allow a home to flow energy bi-directionally.

A home equipped with bi-direction energy systems (from dcbel)

These constraints don’t outweigh the benefit of building out infrastructure to support ‘larger’ virtual power plants networks with more capacity. What could more distributed energy systems look like in the future?

  • Most homes and buildings have an energy storage system + solar panels 
  • Many homes are largely self-sufficient in terms of their own electricity needs (and can charge their EVs accordingly, too)
  • Those that aren’t can get electricity from energy storage systems in their community
  • Larger electricity users, like big buildings and industry, can tap virtual power plants
  • Excess electricity can also be used in all kinds of other applications, like producing green hydrogen for use in industry or for more transportable energy storage 

Pipe dream? Perhaps – it would require a massive amount of investment, metals, data, and software applications to coordinate. But it’s a future worth working towards.

One important lever to scale adoption includes ensuring consumers can make money for participating in virtual power plants. At present this is only the case for people who are part of a net energy metering program. And even when systems to provide incentives, whether for sending electricity back to the grid or for demand response are in place, they’re not always equitable, nor are they necessarily used to their full potential. Room for improvement!

How novel is this really?

It’s important to recognize that the ideas at play here aren’t all that novel. The concept of reducing demand in response to periods of high demand has been around for a long time – it’s called demand response. In many markets, consumers can get credits and / or rebates for participating in demand response; by helping reduce load during forecast ‘busy’ periods, they make some money back on future electricity consumption. 

Nor is the concept of using energy storage to reduce load novel. Homeowners and businesses with energy storage systems can use them to reduce their electricity consumption during peak events, reducing total load.

But flowing electricity back to the grid, especially at a not-insignificant scale, is newer. The 16 MW provided by the consortium of Tesla Powerwall owners last week in California is comparable to the output of eight typical onshore wind turbines (or one ‘colossal’ Chinese offshore wind turbine).

Tesla isn’t the inventor of the ‘virtual power plant’ concept. In Vermont, the Green Mountain Power utility has helped customers get home battery systems and uses them during peak demand. Further, other groups in California are working to aggregate groups of homes and other buildings with energy storage to form virtual power plants, too.

While not the first-mover to conceive of and form VPPs themselves, Tesla’s Powerwalls are quickly becoming the staple energy storage system of choice for utilities, companies, and consumers nationwide who want to participate in virtual power plant networks. In Vermont, there’s actually a larger network of Powerwall owners than the one that went to work in California last week. That said, there are other players, e.g. companies like Swell in New York

The net-net

When people talk about a ‘smart grid,’ virtual power plants are a good example. As renewable energy resources grow and energy storage systems get more sophisticated, there’s an essential layer of data, devices, and software that can help maximize resources and monitor the complex interplay between systems. 

While one 16 MW event might not seem like a massive impact, at scale, virtual power plants, coupled with optimized and scaled forms of demand response, will play a significant role in the next generation of the grid.

Folks have been excited about the evolution of the grid and beckoning in a smart grid revolution for a long time. Last week’s events were a sign it’s starting to happen. Everything that goes into these smart grids, from energy storage systems to new electrical panels and metering infrastructure, represents areas for significant investment, innovation, and business opportunities across hardware and software. 

Virtual power plants won’t solve the energy crisis gripping much of the world right now alone. A smarter grid is a way to enhance and optimize energy use; firm power sources that are significant in output, reliable, 24/7, and ideally, low-carbon, are still the critical, foundational layer of the grid. On that front, companies like Fervo Energy, which raised a massive fundraising round announced on Monday, are innovators to watch.