Grid-scale batteries are a hard business to break into -- and a hard business to regain market share lost to the competition. But for A123 Systems, it appears that Wanxiang Group, the Chinese company that bought it out of bankruptcy two years ago, has been serious about keeping its grid energy storage business alive.

That’s one impression to take from a slew of contracts announced by A123 Energy Storage Solutions in the past month. First came news from China, where the non-automotive arm of A123 announced its 2-megawatt battery was in commercial operation delivering frequency regulation services for Beijing-based Ray Power Systems Co. Ltd.. It’s A123’s first such project in China, although the firm's batteries are being used for similar fast-reacting, power-centric applications by AES Energy Storage in the United States.

Then, A123 announced two more contracts, using its new Long Duration Grid Battery System, developed to emphasize multi-hour energy storage, rather than quick-burst power density, as a core attribute. The first is a 1-megawatt, 3-megawatt-hour system for Spanish Transmission System Operator (TSO) Red Eléctrica de España (REE), delivered as a turnkey solution to integrate renewable generation to the grid while providing reactive power, frequency regulation, and voltage regulation.

The second 1-megawatt, 2.8-megawatt-hour system, announced last week, is capable of similar services, but for a corporate customer. That’s Japanese manufacturer IHI Corporation, which wants to manage its peak power demand and integrate on-site large-scale solar PV at an IHI factory in the Tohoku region of Japan, the northern section of Honshu Island that’s still struggling to meet electricity needs in the aftermath of the 2011 Fukushima disaster.

These aren’t huge new contracts, compared to some of the projects being announced from lithium-ion battery contenders like BYD, Mitsubishi, Panasonic, Samsung, LG Chem, Johnson Controls and Saft. Indeed, they're small compared to A123's 110 megawatts of batteries already deployed, in projects like the 32-megawatt array backing AES’s Laurel Mountain wind farm, in Chile's Atacama Desert, or on the Hawaiian island of Maui.

But they do represent important proving points for a business that lost some significant contracts during its bankruptcy, including a project in Chile with AES Energy Storage and an 8-megawatt, stimulus-grant-funded project with Southern California Edison. What’s more, A123’s launch of its long-duration storage system indicates that Wanxiang is serious about investing in products that take on a core challenge for lithium-ion batteries in grid-scale applications.

“The long-duration system is actually a completely different energy storage cell” than the company’s previous “high-rate” systems, although it uses the company’s same lithium-iron-phosphate chemistry, Roger Lin, director of product marketing for A123 Energy Solutions, told me in a January interview. Specifically, while its high-rate systems are built with its 26650 cylindrical cells, the long-duration batteries are built with its AMP20 prismatic pouch cells, the same that are being used for electric vehicle customers like General Motors for its Chevy Spark, he said.

These prismatic cells, stacked together into storage racks and integrated into formations suitable for supporting grid-scale tasks, can add up to 4 megawatt-hours of storage within a 53-foot storage container, he said. Lithium-ion has so far been seen primarily as a “power” application, unsuited for long-term energy storage or shifting, as compared to competing technologies like sodium-sulfur batteries, flow batteries or emerging aqueous electrolyte-based metal batteries.

A123 is supplying more than batteries, he added. “We have a stable of inverter providers that we use,” he said. “We do all of our own controls hardware and software internally. That includes battery management systems for the cells, and at the highest level, overall control systems for the battery solution.” That’s an important distinction in the emerging grid-scale battery industry, where makers of battery chemistries now in active grid deployment (lithium-ion, sodium-sulfur, advanced lead-acid) are both equipment suppliers and key partners in integrating their products into grid-ready configurations.

So far, A123 has focused on containerized systems, but it’s also looking at launching smaller-scale, behind-the-meter applications in the next year or so, he said. GTM Research predicts that customer-sited energy storage could be one of the first applications to break the economic barriers -- namely, the high price of batteries -- that have held them back from serving more roles on the grid.

Today’s grid battery applications include a few niches, such as frequency regulation, or using it to reduce peak stress on substations or transmission lines that would otherwise require expensive renovations. But future applications, like smoothing out renewable energy’s grid disruptions, shifting energy loads from off-peak to peak demand hours, will require both cheaper batteries and more sophisticated control and integration to put them to multiple uses.

Whether or not grid-scale energy storage can actually meet these needs in the short to mid-term is much-debated question in the industry. Moves by grid regulators in California and other important regions to mandate grid storage or create energy market mechanisms for it could help. At the same time, the recent bankruptcy filing of Xtreme Power highlights the challenges still facing grid-scale battery vendors as they seek to capture projects in a slow-growing, untested market.

On the cost side, there’s no doubt that lithium-ion batteries are quickly falling in price, based on the rapid growth in manufacturing supporting their rollout in consumer electronics and electric vehicle applications. Lin estimated that today’s multi-megawatt lithium-ion systems are being priced at about $800 to $1200 a kilowatt for an hour of storage for AC systems -- quite a bit cheaper than some of the project costs cited in industry reports from a few years ago.

At the same time, “Cell costs for long-duration systems can be up to 50 percent to 60 percent for some of the four-hour systems out there, but it’s not the whole picture. Close to 30 percent of the cost comes from balance-of-system” components, he noted. “While cell costs have come down, that balance-of-system cost cannot be ignored -- and it’s the systems integration that can start to really squeeze those costs down.”