by Julian Spector
August 29, 2019

Someone in the crowd asked me what I thought was overhyped in the storage market, when Stephen Lacey and I recorded a year-in-review podcast at GTM's Storage Summit last December. In the moment, on that stage, one phrase came to mind: virtual power plants.

Part of my skepticism derived from the poorly contrived nature of the name itself. The "virtual" description connotes something that lacks physical embodiment, a disservice to the laborers who spend hours selling and installing batteries, inverters and solar panels in homes and businesses. The invocation of "power plant" equates the hopeful vision of a clean, distributed, democratized grid with the very structure that the cleantech sector is trying to replace because it's old, dirty and inefficient.

Beyond linguistic frustrations, I'd simply heard too many lofty cleantech conference panels about how "one day soon" the virtual power plant will move out of its "early days" and remake the grid as we know it. Its ratio of thought-leader mentions to operating, commercially viable projects did not support taking the concept seriously.

The times they do change, though, as we at Greentech Media will be the first to acknowledge.

Perhaps they started changing that very day, when my colleague Jeff St. John reported on Enbala winning a 1,000-home virtual power plant for Australian "gen-tailer" AGL (hey, it's hard to keep up with the news and prep for a live podcast taping). The utility was signing up customers to add batteries to their homes in the hopes of aggregating a total of 5 megawatts/12 megawatt-hours of flexibility.

Two months later, rooftop solar market leader Sunrun won its first wholesale capacity market contract in New England. This summer it followed up with distributed capacity deals for East Bay Community Energy and the City of Glendale's municipal utility.

And Vermont utility Green Mountain Power keeps adding more homes to its Powerwall army. Batteries in 610 homes sprang into action to lower the utility's consumption during the system peak last summer, saving ratepayers $600,000 in just one hour.

This march of decentralized grid assets has been in motion for a while, then, but sonnen added something new to the movement this week with the Soleil Lofts in Utah. The high-end home storage provider proved that it could outfit an entire new home development with batteries and get a real utility to pay real money for the aggregated grid services. 

"This is almost the ideal: signing a deal to deploy several hundred (and ideally in the future, several thousand) systems that can serve both customer and grid needs," said analyst Brett Simon, who tracks behind-the-meter storage at Wood Mackenzie Power & Renewables.

I covered the details for the news site, but for all you Squares I'm going to home in on what makes this project different from the ones that came before. I'll also flag a few unexpected choices that defy the usual industry logic, plus a few proof points to check back on.

Decentralized vs. centralized, VPP edition

By definition, companies in the business of selling home energy storage believe that a more distributed grid makes better sense than the centralized status quo. Sonnen's latest project reveals a point of strategic variation within this rarefied community: How distributed should the distributed energy be?

Sunrun's model, which also applies in the Green Mountain Power project and the Enbala undertaking as well, is to seek out homeowners who want solar and batteries. Once a critical mass has been assembled across a utility territory or wholesale market region, then the aggregator can roll them together into a commercially operational grid asset. If there's an identified need in a particular area, the installer can target customers more specifically, as Sunrun will do in Oakland and Glendale. But it's still a door-by-door effort.

This approach takes time and effort — the kind of effort Sunrun can afford to make with its national sales team.

Sonnen's parent company in Germany actually embraces that approach, too, finding customers across the nation and linking them up in a virtual community. In the U.S., though, sonnen's smaller satellite operation went for maximum leverage. It chose to strike individual deals that would each generate hundreds if not thousands of sales. It chose to work with homebuilders.

This choice follows from certain observations about the impacts of distributed energy. Adding rooftop solar power willy-nilly helps the customers who own it, but choosing not to coordinate those assets to serve the grid leaves value on the table; at worst, it could impose costs on the rest of customers, if an influx of solar installations forces grid upgrades to handle the new power flows.

Blake Richetta, CEO and Chairman of sonnen U.S., advocates for a managed solution, for delivering batteries to customers and crafting them into a controllable grid asset in one fell swoop.

"This is a master-planned community, managed for the benefit of community itself," he told me recently, describing the Soleil Lofts project.

This philosophy makes it easy for the utility to plan for an influx of solar power, by building in batteries onsite. Keeping it all at a single location streamlines planning relative to the same number of units spread out across a city. And putting the utility in the driver's seat — it will have full dispatch control to do whatever it wants with the Soleil Lofts batteries — allows the entity that knows the most about the local grid to use the tools however it sees fit.

A utility-friendly approach to residential battery deployment may not be fashionable among clean energy revolution advocates, but it wins the kind of friend who can make the business model work, as Rocky Mountain Power did for Soleil with its grid services payments.

It also produces a good return on energy invested. A small team working in executive meetings with builders and utility personnel can broker a deal that moves hundreds of battery systems at once. As long as the deal comes together, it's a savvy way for a small team to extend its influence in the market.


That's not to say that everything about the Soleil Lofts deal makes sense on the surface. Several key tenets appear counterintuitive if not puzzling, in particular the reasons for the batteries being there in the first place.

This is a solar-powered community, but the customers don't receive any of the solar power. They get to come home and see the panels on the roof, while a separate legal entity related to the development sells the power.

Similarly, as described to me by the three parties to the deal, the customer has no control or functional interaction with the battery system. And yet, Wasatch Group thought it very important to put the sonnen ecoLinx units in each individual apartment.

The company even considered a "large battery bunker," since the batteries will be operated en masse by the utility, but opted for the more interactive model of putting them in the home.

"It’s more about the apartment being a living organism," said Wasatch COO Jarom Johnson. "We’ve never lived in an age where technology was more interactive and more a part of your family."

American installers typically relegate batteries to the garage, out of sight, but "in Germany, a lot of our clients are so proud of their clean energy battery in their living room," Richetta said.

In Soleil Lofts, the customers can look up from their dinner table, see their sonnen unit glowing faintly and derive pleasure from the knowledge that someone in a control room miles away could be signaling for that device to provide frequency regulation at that very second. Or, lacking detailed knowledge of grid services, the tenant may get by on the "using solar power to help the grid" explanation.

The one case where the batteries could have a practical impact on residents' lives is in the event of a blackout, when the batteries will keep individual apartments and common spaces powered up through the dark times.

Backup power was the initial engine behind sonnen's U.S. sales; back in 2016, 90 percent of all installed sonnen units were set in backup mode. As an employee told me at the time with refreshing clarity, “This is an emotional sell, totally an emotional sell.”

That's still the case here. Soleil is built in a relatively new part of the grid with minimal risk of outages, Rocky Mountain Power Managing Director Bill Comeau told me. This isn't California with its wildfires, Arizona with its sandstorms and thunderstorms, Florida with its hurricanes, or New England with its snow and ice. Instead, Comeau said, "It’s more about the vision."

"Utah’s got a unique culture — we like to be independent," he said. "If there was a natural disaster, this is a community that chose, 'We’re going to be resilient.'"

What does the tenant actually get from this virtual power plant, then? The feeling of living in a renewable, carbon-neutral community, without having to sign any special contracts to achieve it. Proximity to innovative grid solutions. A cooler perk than the usual pools and grills and fitness centers.

What we do know is the structure generates a large purchase order for sonnen ecoLinx. It offers Rocky Mountain Power an appealingly forward-thinking way to spend grid modernization dollars earmarked by a state law from 2016. And it gives Wasatch Group a differentiated marketing platform largely paid for by tax credits and the local utility.

Clever financial engineering seems to be filling the gap left by the usual reasons for customers to buy battery storage. That's not a knock on the project; clever financial engineering is the only way solar and batteries ever made it to the grid.

Much remains to be seen

Thanks to the work of sonnen and its peers, industry thought leaders can finally discuss virtual power plants in the present tense. The next question is whether the early instantiations of the concept reflect well on it in practice.

"It's entirely possible the whole thing could fail," cautioned Simon, the WoodMac storage analyst. "The technology might not work properly, the value may be lower than expected, regulations may change, customer uptake may be slow, etc. But the opposite is also true, and if it succeeds, it's a great example for what aggregated, residential solar-plus-storage can do."

There are two major proof points here.

First, do customers actually want this? That question takes on different meaning depending on the type of VPP. For Sunrun, where the grid services layer on top of a customer value proposition that has to make sense, the uncertainty revolves around what happens to a customer's battery when it cycles for grid services, and whether customers feel they're getting a fair cut of the revenue.

Sonnen took some of that uncertainty off the table. The customers don't get any revenue, so they have nothing to complain about. Nor are they likely to use the batteries for any purpose, so even if the grid cycling wears them down over the years, the tenant would have no way of knowing.

The broader question, as far as grid decarbonization is concerned, is whether the utility gets its money's worth in grid value. In theory, splitting the costs among multiple parties should give the utility capacity at lower prices than single purpose assets. That calculus relies on sonnen aggregating and delivering power as promised.

With any luck, Rocky Mountain Power will divulge performance data when it has some runtime to analyze. Unless cause for concern emerges, we can expect the steady march of virtual power plant announcements to continue.