by Julian Spector
September 08, 2017

Hello readers, next week I'm heading to Vegas for Solar Power International. My mission: to scour the swankiest bars, poolsides and imitation European landmarks for all the latest insights into where the storage and solar industries are headed. If you're going to be there or have any tips to share, let me know at [email protected].

The latest Energy Storage Monitor just flew off the digital presses, and there's been a notable shift in the market since last quarter.

In terms of megawatts deployed, Q2 was unremarkable, down 11 percent year-over-year and down 47 percent from a strong Q1. Total megawatt-hours nudged up slightly year-over-year.

It turns out the largest-ever storage procurement is a tough act to follow. With Aliso Canyon wrapped up in January, the next quarter was doomed to disappoint in comparison. That's unusual in the history of storage deployments, which tend to increase from Q1 to Q2. 

A new source of strength arose this quarter, though, in the previously humble behind-the-meter segment.

Residential and commercial storage produced 32 megawatt-hours compared to utility-scale's 18.5 megawatt-hours, accounting for 63 percent overall. That resulted from a surge in behind-the-meter activity coinciding with a down quarter for grid-scale, where the few projects that were built had shorter durations.

The rise of distributed storage is something we should get used to seeing, said Brett Simon, who tracks that sector for GTM Research.

"This quarter was a bit of an outlier, rather than the start of a trend that will continue every quarter," he said. "Still, as we get into the 2020s, it will become more common until it's the norm."

The sheer number of deployments doesn't say much about how big they are, but it's worth noting that the behind-the-meter segment set a record with 443 projects installed in Q2. That expanding number of projects bodes well for the industry, which has been tightly clustered historically.

Much of the growth came from the residential sector, which very nearly doubled the number of projects it delivered in Q1. The sector is on track to beat its number of installations from last year, provided the pace doesn't taper off significantly in the second half of the year. 

The Q2 storage market showed a gently creeping geographic diversification. California still clobbered everyone in the C&I space, with more than nine times the capacity deployed compared to runner-up New York. In the home storage arena, Hawaii fell just 52 kilowatts short of unseating California, thanks to its Customer Self-Supply program finally heating up.

And Arizona quietly took the lead for utility-scale deployments this quarter. Even without a state mandate or major incentives, Arizona's utilities have begun choosing storage based on the economics in notable cases.

One quarter doesn't make a trend, and this was an anomalous quarter at that. But it showed how lots of little projects can add up to more than a few big ones.

CAISO parties like it's 2014

Ok, 2014 doesn't sound like so long ago in most contexts, but that was a relatively primitive time for the storage industry compared with today.

Somehow data from that previous era formed the basis for a crucial California grid study about the necessity of a big new gas plant.

The California Energy Commission has to decide whether to allow NRG's Puente gas peaker to be built in Oxnard to replace gas plants retiring by the end of 2020. This process started years ago. These days, the state is making noise about a 100 percent renewable goal and pulled off 100 megawatts of storage in six months for grid reliability purposes, so the whole "this has to be a gas plant" premise has started to look a little shaky.

In walks the California Independent System Operator with a study of whether alternatives could meet the reliability needs of the local grid. It concludes that storage can do it at 2.7 times the cost. That doesn't look good for the folks advocating a storage solution instead of Puente.

I took a look into the footnotes of the study, though, and discovered those storage cost projections came from, yes, 2014

The upshot of this is that a massive storage procurement in 2020 will absolutely cost less than CAISO calculated based on the outdated numbers. That opens up the question of how cost-competitive other clean energy assets really would be. The clock is ticking to get a project in motion in time for the deadline, and regulators will need to figure out soon if they want to move ahead with the old plan or initiate a more robust investigation of alternatives.

This episode points to a larger structural challenge in storage procurement: how do regulators evaluate it when the cost projections keep on changing? CAISO needed to use publicly available data; it doesn't have access to proprietary cost projections, which storage companies jealously guard. The best publicly available data it could find wasn't much use in this fast-paced industry.

Another data point of just how fast things are moving, courtesy of the Energy Storage Monitor: GTM's cumulative five-year forecast has reached 7,463 megawatts, up 30 percent year over year. Even early 2016 cost projections have lost their relevance.

Hybridize it!

Another thing the CAISO study left out: solar+storage as a key solution. The nonprofit Clean Coalition, which would rather see clean resources than new fossil fuel plants, has since run its own study factoring that in.

They say the benefits of combining solar and storage for firmed local reliability, including the 30 percent federal Investment Tax Credit, makes the alternative solution cheaper than the gas plant -- $267 million compared to $299 million. 

This solution jibes better with the sentiments of Oxnard itself, which is sick of having its coastline gobbled up by looming power plants. It's very possible that more studies are on the way from other groups with their own take on the alternatives and their cost-effectiveness.

Microgrid advances

The startling force of Hurricane Harvey brought grid resilience to the forefront, as my colleague Jeff St. John reported. Among other things, the failure of the grid and local backup systems at the Arkema chemical facility in Crosby knocked out refrigeration for volatile chemicals, leading to them combusting.

Microgrids for local resilience are still an infant industry, mostly practiced at military bases and university campuses. But they're gaining wider traction, in large part stemming from Superstorm Sandy's battering of the New England coast, and may see renewed interest after Harvey. There's plenty of room for innovation in streamlining and improving this type of product.

Chicago-based S&C Electric Company pushed the functionality forward with a system they worked on at utility Ameren's testing and application center next to the University of Illinois.

"This particular installation is unique in that, when utility power comes back [after an outage], it doesn't have to turn off and reconnect," said David Chiesa, senior director for business development. "This system can sync with the utility signal and seamlessly transition back."

That sounds simple, but it's a significant development, verified in a 24-hour islanding test last month. Instead of momentarily shutting down to switch from grid-connected to island mode or vice versa, this system can move between them without a loss of power. If an outage is expected, the system could proactively switch to island mode without the loss of power an outage would cause.

Accomplishing this requires getting all the generation sources in the microgrid to harmonize with the frequency, voltage and phase angle of the broader grid. Without that, the rotational inertia of the grid could wreck a microgrid that connects while running at different specifications, Chiesa said.

That's one more step toward a world where microgrids fade into the background while toggling in and out as needed, without disruption.