Global power company AES was building large batteries a decade before the market really took off.
The company, based in Arlington, Virginia, owns $33 billion in assets, spanning regulated utilities in the U.S. and abroad, independent power production, renewables development and battery integrator Fluence. That diversified corporate structure created plenty of different arrangements with which to try this emerging grid technology.
AES first ventured into U.S. battery development a decade ago in New York state. The grant-supported 20-megawatt system, designed for ancillary services, was not a big economic success (large-scale battery development is still getting off the ground in that market). But the company gained confidence in the technology in Chile, where AES subsidiary Gener deployed batteries to take over spinning reserve duties from thermal plants.
This series focuses on U.S. asset ownership, but it's important to note AES’ global reach: The company operates 920 megawatts of storage across 22 countries, and it was the first developer of large grid batteries in many of those places, AES President and CEO Andrés Gluski said in an interview.
“We have more experience, for more time, and we have a bigger volume than anybody,” Gluski said. “When I talked about batteries three years ago, it was the lone cricket in the woods. Now we’re part of a whole chorus.”
That chorus amounts to more competition, but AES continues to grow its volume. Some of the current top-five battery owners are in the ranking due to projects that were built several years ago. AES never stopped building and is currently constructing 155 megawatts of batteries in the U.S. that it will own. Recent wins in Hawaii included a 30-megawatt/240-megawatt-hour system, pushing the boundaries of long-duration storage in today’s market.
Grid batteries have a learning curve, and AES has moved further along that curve than almost anyone.
It has years of experience running fast-paced ancillary services batteries, the typical stepping stone into the technology. But it also has longer-duration systems serving real renewables-shifting roles at a time when most developers can only talk about the potential.
“We are building a lot of renewables...we have some prime locations, and we have first-mover advantage,” Gluski said.
AES kicked off 2019 by inaugurating a record-busting solar-storage facility on Kauai, designed to let the Kauai Island Utility Cooperative absorb solar power and save it for evening peak capacity. The developer is in the midst of building the 100-megawatt/400-megawatt-hour Alamitos battery for utility Southern California Edison to shift the region’s daytime solar power for use in the evenings. That project sits next to the recently repowered AES Southland plant.
Those two flagship projects show AES is equally comfortable developing storage alongside renewable energy or existing gas plants.
Where solar is abundant, pairing it with batteries is “the killer app,” Gluski said; it’s hard to make power cheaper than that. But batteries do not generate electricity, they store it. In areas without cheap renewables, gas plants may be needed to charge the batteries.
Thermal plants also offer advantages for developing new storage plants, and AES is studying several more projects at existing power plant sites. “The distribution networks were built around big thermal plants,” Gluski said. “To the extent that those are being retired, those are ideal places to locate batteries. They were also optimally sited, taking into consideration the load.”
To take advantage of this, a developer must both own older power plants and have an interest in developing energy storage. So far, only a handful of others have tried. Vistra Energy is doing it at legacy California thermal plants in Moss Landing and Oakland. LS Power is working on a battery peaker transformation for New York City’s Ravenswood plant. That short list leaves plenty more power plants in various stages of use that could be converted into battery sites.
Utility, developer, supplier: All in one
A few business models have shown up in the top-five storage owner countdown.
Nos. 4 and 5 were regulated utilities Duke Energy and San Diego Gas & Electric; Duke built its biggest battery via its competitive renewables arm, but the battery action now is happening on the regulated side. Then there was Invenergy, the independent renewables and gas developer.
AES comprises all those business models: It owns independent power generation but also owns regulated utilities like Indianapolis Power & Light; it has a stake in renewables developer sPower; and it owns half of grid battery integrator Fluence, a joint venture with Siemens. That diversity in the corporate structure unlocks options and abilities that other companies don’t have.
“We’re open to all kinds of business, whether it be a partnership or directly owning it or selling the technology,” Gluski said.
The company’s utility subsidiaries put storage to work before there was much of a market for it elsewhere. The Chile projects helped kick things off, but AES also delivered a 20-megawatt/20-megawatt-hour battery to IPL in 2016, which it touted as the first of its kind in the Upper Midwest's MISO grid region.
The generation portfolio is undergoing a systematic pivot from fossil fuels to zero-carbon, and AES still owns plenty of coal plants. The company has promised to push its share of coal generation below 30 percent by the end of 2020; a recently announced coal plant divestment in India brings the level from 45 to 35 percent. At the same time, AES has gigawatts' worth of renewables in the pipeline, and that surge creates ample opportunities for more battery development.
Fluence's role in the storage market's development
AES turned its early experience designing battery systems into the unit that became Fluence, which recently rolled out its sixth-generation storage enclosure, moving from shipping containers to a more compact cube design.
Fluence operates independently, selling to rival developers. But it continues to supply many of AES’ most prominent projects, like Alamitos. And AES subsidiary sPower was among the first four customers announced for the cube product, as was Fluence co-owner Siemens.
“By having Fluence, we drove down costs for everybody, including ourselves,” Gluski said.
Fluence does not turn a profit currently, because it invests what it makes into research and development, Gluski said. “We’re plowing the money back into new designs and growing. It’s very different from a utility investment, where it’s all about getting the cash back. If we’re going to change the world, this thing has to continue to innovate.”
The global nature of AES' business means that innovations in one part of the corporate family can quickly be used in others. An AES utility in Brazil was required by local rules to spend 1 percent of its budget on research and development. It spent some of that requirement on drone applications for inspecting grid equipment, found some attractive uses for that technology, and now the techniques are spreading to other AES properties.
“Our people are looking around the globe a bit and seeing what happens in other areas,” Gluski said.
The challenge with running so many businesses in so many places is that sprawling global corporate hierarchies can have a way of squeezing out creativity and agility. AES takes care to maintain “entrepreneurial vigor” by keeping business units like Distributed Energy relatively independent, Gluski said.
“We approve a budget and parameters for returns and let them run free,” he said. “It’s a very different model than some of the big utilities, which [are] very centralized.”
Coming soon: "Virtual transmission"
As regulations for energy storage evolve, new business models become possible. AES has a few in mind already.
One is the idea of “virtual transmission,” in which placing batteries along a transmission corridor allows for holding electricity during constrained moments and delivering it later, when congestion clears up. It wouldn’t expand peak-hour transmission like building out new wires. But it does save excess power from being wasted, with considerably lower capital costs and permitting difficulties than a major transmission project.
AES is already working on such a proposal in Germany to assist the delivery of North Sea offshore wind to load centers in Bavaria. “We basically proposed the idea to the German grid operator, and they liked it and put out a [request for proposals] for it,” Gluski said.
The concept will also be valuable in the U.S., where “permitting [transmission lines] may not be as difficult as it is in Europe, but it's still difficult,” Gluski said.
Customer-driven solutions are another growth area. The old approach was building a power plant to meet an established capacity need; new resources allow for more creativity to meet a customer’s needs.
The deal with Kauai Island Utility Cooperative reflects this: The companies worked together to design a renewable power plant that could power the island through the night, not just during the sunny hours. “Load-following renewables” is now on the table for other customers that want it.
Last fall, AES signed a “10-year strategic alliance” with Google, to work with the tech giant on new renewables developments for data centers and use Google Cloud technology to make the grid run more efficiently.
As cities ramp up electric buses and delivery fleets, the power needed to charge them will grow. AES could make the case for battery or microgrid installations at charging depots to balance supply and demand locally, Gluski said.
History shows that ideas that work for AES customers could trickle down to the various utilities and other companies AES controls around the world.