New Hampshire, USA — Back in mid-August, Vine Fresh Produce in Ontario unveiled a 2.3-MW solar rooftop array on its greenhouse, the largest commercial rooftop project under the province’s feed-in tariff (FIT). This system notably incorporates a technology that’s been more familiar in the U.S. residential solar market: microinverters. (The devices, made in Enphase’s Ontario plant, helped the project qualify for that FIT.) Weeks ago Enphase followed that up with another large-sized project using microinverters, 3.1-MW of distributed solar across 125 buildings for the San Diego Unified School District.
Vine Fresh Produce’s 2.3-MW (2-MW AC) solar project in Ontario, Canada. Credit: Enphase.
Those announcements were meant as stakes in the ground. “We’ve proven [microinverter technology] in residential, we’re proving ourselves in small commercial… but our ambitions are much bigger than that,” said Raghu Belur, Enphase co-founder and VP of products and strategic initiatives. “We’re seeing people deploy [microinverters] in significantly larger systems.”
The technology is rapidly gaining traction, according to Cormac Gilligan, IHS senior PV market analyst. Microinverter shipments will reach 580 MW this year, with sales topping $283 million, and average global prices sinking 16 percent to $0.49/Watt, he projects. By 2017 he sees shipments soaring to 2.1 GW with revenues of about $700 million, and expansion beyond the U.S. into several regional markets, especially those in early stages of development that might be more open to newer technologies: Australia, France, the U.K., Switzerland, and even Hawaii. Japan’s big residential solar market is especially attractive, but poses certification challenges and strong domestic competition.
But as those two Enphase projects illustrate, there’s another growth area for microinverters that’s emerging alongside regional expansion — up into commercial-sized rooftop solar installations. The same reasons residential customers like microinverters apply to small-scale commercial projects as well: offset partial shading, more precise monitoring at the individual module level, provide a more holistic readout of what the system is producing, and improve safety because they typically use a lot lower voltage. Just nine percent of microinverter shipments in 2012 were to commercial-scale use, noted Gilligan — but he sees those surging to nearly a third of shipments by 2017.
Who’s Making Microinverters
The microinverter space is getting crowded (see table below), if not yet a model of parity. Enphase continues to dominate with more than half of the sector’s revenues in 2012, four million units cumulatively shipped and four product generations. “We are a high-tech company that happens to be in the solar sector,” Belur explained. Compared with what he called the “big iron, big copper guys” who are now broadening their inverter portfolios with microinverters, “we’re all about semiconductors, communications, and software.” The company designs its own chips for its microinverters, and outsources manufacturing to Flextronics.
SMA got its entry into the game with the 2009 acquisition of Dutch firm OKE. “In the residential market it became clear to us that customers were interested in the microinverter architecture,” said Bates Marshall, VP of SMA America’s medium-power solutions group. SMA also sells the string inverters that have gained favor over big centralized inverters, so SMA’s simply broadening its portfolio. With the emergence of the U.S. solar end-market, SMA is more willing to push some R&D and product development over here; “we get to drive the bus to a greater extent,” he said. SMA recently started shipping microinverters to the U.S. from its German inventories, but a production line is now being qualified at the company’s Denver facility.
Similarly to SMA, Power-One (recently bought by ABB) aims to supply whatever type of power conversion capability customers need, noted Chavonne Yee, Power-One’s director of product management for North America. So far demand for microinverters has come in the U.S. residential market, offering high granularity and maximum power point tracking (MPPT), but she sees most of the commercial-scale demand switching from traditional central inverters to three-phase string inverters, not microinverters.
Module supplier ReneSola sells a standalone microinverter, touting the typical features with some higher (208-240) voltage options for small light commercial, but at a 15-20 percent lower price point, explained Brian Armentrout, marketing director for ReneSola America. “We are seeing some demand” in small light commercial applications ranging from 50-kW up to 500-kW at which points there’s “the breaking point where string inverters make more sense.” Down the road the company wants to take the end-around route of integrating microinverters directly onto panels; its gen-2 microinverter should be available in the spring of 2014. Armentrout projects ReneSola will be “in the top three” next year for microinverter sales, while simultaneously aiming high for the top spot in module shipments.
Others are looking to integrate microinverters directly into the modules. SolarBridge has worked closely with SunPower and BenQ to design its microinverters to eliminate several components that typically fail, notably the electrolytic capacitors and opto-isolators, explained Craig Lawrence, VP of marketing. They also minimize other typical costs such as cabling, grounding wires and even tailoring the microinverter for a specific module type to optimize the microinverter’s firmware, he explained. He sees the trend to bring microinverters into the commercial-scale environment, particularly with SolarBridge’s more recent second-generation microinverters in the past year or so.
Microinverters vs. String Inverters
In general, installers are making a choice between microinverters and string inverters, comparing functionalities and costs. Both sides make a case for reliability: microinverters use fewer components and represent lower cost when something does fail; string inverter vendors point out microinverters have only been on the market for a few years and can’t make substantial claims about reliability. IHS’s Gilligan noted the sheer number of microinverter devices in the field potentially requiring repair/replacement could be daunting.
Solar panels on a building for the San Diego Unified School District. Credit: Enphase.
SolarBridge’s Lawrence argues in favor of microinverters on an operations & maintenance basis. Central inverters account for half of an operations & maintenance budget and it’s the single highest failure component in a solar PV system; that’s why there’s been a shift from those to string inverters on commercial-scale solar. “All the reasons you’d do that, are the exact same reasons to go from string inverters to microinverters,” he said. “You want as much redundancy and granularity as you can possibly get, to maximize your rooftop utilization and simplify your O&M.” Factoring in replacement costs, labor savings in not having work with high-voltage DC, “for most of our customers that alone is enough to justify the additional [price] premium.” With a microinverter you’ll know when (and which) one panel is underperforming, and it might be tolerable to just leave it alone; on a string inverter you might not know where the problem is while you lose power over the entire string, he pointed out.
Scott Wiater, president of installer Standard Solar, acknowledges that microinverter technologies and reliability have improved over the past couple of years, but he’s not convinced this is an argument in their favor vs. string inverters. “I have concerns over the long term,” he said. “If you truly believe you’re going to get 25 years out of a microinverter with no maintenance, that might hold true, but we haven’t had that experience.” In fact he advises that any residential or commercial system should plan to replace whatever inverter it uses at least once over a 20-year lifetime.
Commercial-Scale Adoption: Yes or No?
Talking with both inverter vendors and solar installers, the choice of microinverters vs. string inverters for commercial solar settings is making some initial inroads into light commercial applications, but might not be quite ready to move up in scale at that commercial level.
“For projects under 50kW, we have found that microinverters can be positive for the project LCOE on an ‘all-in’ basis,” explained Jeremy Jones, CTO of SoCore Energy, an early adopter of microinverters, including commercial solar projects into the hundreds of kilowatts in size. In general the technology’s “high granularity of real time data is very useful in the ongoing asset management,” and SoCore’s projects with microinverters “have consistently outperformed our other string inverter and central inverter sites.” The technology stacks up favorably to central and string inverters (especially for three-phase 208-volt systems) in terms of added costs, he said: warranty extensions, third-party monitoring, and other balance-of-systems costs. Microinverters’ performance and low-cost warranties also benefit longer-term finance deals, he added.
However, above 50kW “we have had a harder time making microinverters ‘pencil’ on typical projects,” Jones added. Until costs come down, those larger-sized projects where microinverters can make sense tend to be unique cases where there’s a higher value per kilowatt-hour (higher electric rates or SREC values), or sites that can maximize kWh per kW due to high balance-of-systems costs, such as parking canopies, he explained.
SMA’s Marshall is “bullish on the commercial market, that’s where the volume will be” for inverters in general, but he doesn’t see it as a big boon for microinverters because of what he calculates as a 25-30 cents/Watt cost delta from residential string inverters. In the residential space there are ways to knock prices down to mitigate that difference, but in the commercial space that gap is too big for the average buyer, he said. “As a mainstream option? We don’t see it today.” Microinverters may have a play for “some unique projects” such as campuses or municipalities spanning multiple buildings, but the big growth in commercial solar will be in large retailers, “big flat open roofs, and big flat structures like carports,” he said, and there a three-phase inverter “blows the door off in terms of raw economics.”
SolarBridge’s Lawrence is “seeing a lot of activity” in smaller commercial settings (100-kw or less), tallying to 15-20 percent of the company’s product installations. But while the company is bidding into projects ranging up to 1-MW, it’s “harder to make the case above 250-kW,” he acknowledged; “those don’t pencil out for us right now.”
“Anything below around 1 megawatt, we are shifting from a central to more of a string inverter, but we’re certainly not going to the microinverter level yet — nor do we think we will anytime soon,” said Standard Solar’s Wiater. “The economics behind the projects and having it pencil out, microinverters just can’t compete with string or central inverters on a larger scale.” While microinverters can help on some rooftop applications where shading might be an issue (close to elevator shafts, vents, HVAC units), a more tightly-designed system with an efficient string inverter “can have a much better return for the customer,” he said.
Jeff Jankiewicz, project/logistics manager at Renewable Energy Corporation in Maryland, “definitely considers” microinverters as part of a system design; “we like the performance and efficiency they provide.” But for his company it’s really only for residential and small commercial projects; the largest they’ve done is a 20-kW system out in Maryland’s horse country. Any bigger than that and it’s a case-by-case comparison, specifically looking at shading and energy conversion.
Microinverters and the Grid: The Solar Industry’s Next Battle
Everyone we talked with about microinverters agreed on one thing, however: there’s a trend coming that will incorporate more advanced grid management capabilities, such as reactive power and low-voltage ride-throughs, to give utilities more control and the ability to reach in and curtail availability to support grid reliability. California’s Rule 21 proceedings [http://www.cpuc.ca.gov/PUC/energy/Procurement/LTPP/rule21.htm] is the first such example, seeking to mandate control functions in distributed generators. Those grid-management capabilities are already coming and “very, very soon,” Lawrence urged, pointing to new requirements being codified in Australia and the U.S. probably following within a year or so.
SMA is becoming very vocal about this topic. Its microinverter architecture incorporates a multigate feature with wired Ethernet that allows for a single point of interface into the array, which he emphasized is important for modern grid codes and providing grid management services, Marshall emphasized. Power-One’s Yee, ReneSola’s Armentrout, and SolarBridge’s Lawrence echoed the concern over regulations and requirements coming down the road that will necessitate microinverters becoming more grid-friendly. They also questioned whether all microinverter architectures are suited for such site-level controls — specifically market-leading Enphase, which they said is limited in its architecture and topology.
Enphase’s Belur responds strongly to this debate. “We 100 percent support the need for advanced grid functions, and we are absolutely capable of providing those,” he replied, calling those criticisms an “oversimplification of the problem.” Enphase, he said, is “the most proactive company” pushing for those grid-management requirements — but is seeking to do it judiciously through standards bodies and with proper certification and testing bodies, “and you cannot ignore the policy on top of that,” he said. “It needs to be done; let’s do it properly,” he said.
Integration of energy storage, which also recently got a California state mandate, is another looming question as it relates to inverters. Standard Solar’s Wiater thinks that’s a bigger challenge for inverter functionality than grid-friendly controls, to more directly address the issue of buffering solar energy’s intermittency. Some inverters are being designed to interact with energy storage, he noted, but he questions how that would work for a microinverter because it “defeats the purpose” to switch from DC to AC on a roof, then convert back to DC again. Power-One’s Yee, meanwhile, sees more distributed solar combined with battery storage as a tipping point in favor of multi-port string inverters being a more cost-effective approach.
Wiater agrees that grid management features are coming, and that the bigger inverter technologies have been out in front of some of these requirements, e.g. to curtail output. On the installer side, SoCore’s Jones isn’t seeing customers or utilities push strongly for such capabilities yet, but “spec’ing these features in now will allow us to future proof our designs and open up possible future revenue streams.”
This issue might have bigger ramifications than just competitiveness between inverter suppliers. Once distributed solar generation gets enough penetration into the grid, utilities will say they can’t support it without stronger control capabilities, Lawrence warned. That’s likely going to be hashed out as a negotiation between the solar industry and utilities and implemented via codes and standards applicable to everyone, and the industry needs to get out in front of that resolution, he pointed out. “The solar industry is going to have to participate, or utilities will have a good case why they can limit the penetration of solar PV,” he said. He cited discussions with a large U.S. solar developer who listed these smart-grid control capabilities as one of their top-four priorities for the coming year: “They believe it’s coming,” he confirmed. Getting the solar industry working together to help these speed these capabilities along “will help head off utility objections to more and more solar.”