C&I storage company Stem has come back for more money after closing an oversubscribed Series D in January.

The company has added $26 million, led by BNP Paribas and Magnesium Capital, bringing the Series D to $106 million.

“We looked at it as a board, and there were some great additional partners that we could include in the round,” said CEO John Carrington. “It was just prudent to add the capital.”

Additionally, the Ontario Teachers’ Pension Plan followed up its January investment with a CAD $200 million financing facility (USD $153 million) to support Stem’s efforts in the province. The company intends to use that $200 million over the next 12 months to finance projects in Ontario, Carrington noted.

Stem started out in California using battery storage to reduce electricity bills for commercial entities while aggregating across storage installations to serve utility contracts or bid into the wholesale market.

Ontario currently lacks opportunities to use distributed storage for grid services, Carrington said, but the company has developed a different business case. It will help larger industrial customers mitigate the Global Adjustment Charge, which is set by their 5-hour-long consumption peaks in a year.

So far, Stem has announced one customer there: Inoac, a large injection molding plastic company involved in the automotive space.

Provided Stem’s analytics can successfully anticipate when some or all of those five peaks occur, it can save customers money without any disruption to business activities.

“Customers don’t want to change their behavior,” Carrington said. “Our system, in conjunction with the software platform, lets them do what they do best.”

Company materials now tout that software first and foremost, pitching Stem as a purveyor of “artificial-intelligence powered energy storage.” 

All grid-connected batteries require some layer of software to govern operations. When Carrington talks about AI, he frames it as the ability to learn from all the hours of run time that Stem has accumulated over the years to train the algorithm and improve it.

“We’re still doing the same business model that we had before,” he said. “We realized that all the data [and] the run-time hours that we were generating were extremely valuable.”

When Stem installs a new system, he added, it will take a month or so to “learn the building,” assessing when the loads rise and fall, before the official bill-reduction service starts. The accuracy of the algorithm typically increases 15 percent over the first 45 days.

“It improves, then it’s off to the races,” Carrington said.

The raise demonstrates that investors have at least some appetite for commercial storage.

At some point, future investment may hinge on demonstrating success outside of the nurturing embrace of California’s Self-Generation Incentive Program. California accounted for 88 percent of all U.S. commercial storage megawatts deployments in 2017, according to GTM Research's Energy Storage Monitor.

Stem has dipped into other markets, like Hawaii, Massachusetts and New York. Those remain minor plays compared to California, but the financing in Ontario could jump-start a substantial business where energy storage had been almost nonexistent.

By continuing to raise capital, Stem avoided the buyout route taken by C&I developers Green Charge (now Engie Storage) and Demand Energy (bought by Enel). Competitor Advanced Microgrid Solutions maintained its independence by raising a $34 million Series B last year, with a similar emphasis on its software.

Stem has choices for its path ahead. It’s one of the few current cleantech companies where talk of an IPO doesn’t sound implausible.

“The board is by no means saying, ‘You’ve got to sell this company tomorrow,’” Carrington said. “Let’s continue to execute, and with execution good things will happen.”