Search sponsor
Search sponsor
The post-Intersolar North America debriefing continues, with analyses, observations, and exclusive information garnered at the event from across the photovoltaic technology value chain. Topics range from PV’s niche multiplicity to the growing market momentum and emerging applications space of the CIGS thin-film sector to the problem of ridiculously divergent market forecasts to a young company with a low-cost single-crystal silicon mousetrap.
Niche multiplicity. I had a nice chat with the upbeat Alain Harrus, a partner at VC house Crosslink Capital and former semiconductor industry exec, who sits on the boards of solar early-stagers Alta Devices and SoloPower. He believes there is “a multiplicity of market niches for all kinds of PV” that fit into the “efficiency, performance, and cost envelope.” But small companies are finding it increasingly difficult to get enough capital to build out production, he noted, giving the "more mature" companies a better shot at success. He also pointed out a dilemma for the Chinese solar manufacturers—since they count manufacturing costs in dollars, but sell in Euro (at least in Europe, of course), they could face "currency-driven margin compressions."
Finally, Harrus had a word for those who might engage in obsessive competitive eyeballing: “Knowing what the other guy is doing or not doing is a waste of time—it’s all about execution, taking care of one’s own shop.”
CI(G)S spells success? Execution is finally starting to take place in one of the industry sectors that Harrus knows well. The growing market momentum of copper-indium-gallium-(di)selenide (CIGS) thin-film PV and its CIS cousins emerged as a trend worth noting at the show. Sulfurcell’s U.S. sales rep Boris von Bormann told me the German CIS (the “S” stands for sulfur in this case) company started shipping sample modules to select American rooftop-oriented customers in October 2009. He expects the domestic UL certifications to come soon for the framed 1.3 x 0.66m panels.
The modules are produced at the firm’s 75MW capacity plant near Berlin, which von Bormann said has a 2010 output goal of about 20MW, with 35-40MW projected for 2011—all the while with continuous improvements in efficiency and wattage ratings.
Clever improvements on the balance-of-systems and installation side have also led to the company’s systems achieving 98% of the power output of a crystalline-silicon-based system on the same amount of roof space.
The modules may bear a “Made in the USA” label one day, since Sulfurcell (which counts Intel Capital among its investors) is considering building a fully integrated production facility in the U.S., according to von Bormann.
A CIGS company already draped in red, white, and blue--flexible module maker Ascent Solar--has ramped the first 6-8MW of its 30MW production plant in Colorado. Head honcho Farhad Moghadam told me that they have most of the tools on the factory floor for 20MW, with a few incremental/additional systems coming soon. In terms of laser scribe and lamination gear, there’s already more than 30MW ready to go.
He also noted that modules have been submitted for IEC certification, with the UL process “right behind it.” The IEC tests should be completed by “August or so,” which once completed, would allow the company to start shipping product to the BIPV market in earnest.
An intriguing emerging sector that we discussed was transportation, where flex modules like Ascent’s could be used on automobiles to help power ventilation systems to keep the cars cool in hot climes. Another, more esoteric application would involve putting the conformal PV panels on tops of big-rig trucks or even trains.
Moghadam said that with larger trains, one could install as much as 5-6KW of PV per rooftop, which when multiplied by many cars is the equivalent of a commercial/industrial-sized power system. He claimed it would also be a very lucrative specialty business, with 100MW of product sold into the subsector generating the revenue equivalent of 300MW of modules sold into more mainstream applications.
On the opposite side of the globe, Showa Shell subsidiary Solar Frontier (SF for short, though this ain’t no sci-fi outfit) is a great example of a non-Chinese Asian company trying to penetrate/dominate the U.S./Americas markets with a non-crystalline silicon product line. Greg Ashley, the newly hired COO/VP for the region (who’s previously worked for Canadian Solar and SunEdison), enthused profusely about how SF was “a strong, bankable company” with “the best technology he’s encountered in the last nine years.”
“No other company that’s 110 years old is as committed to solar as Showa Shell/Solar Frontier,” he remarked. “This is not a boutique business…coming out of Japan gives it the credibility of global reach.”
Ashley said that small commercial shipments have been made to the U.S. already, with some “fairly large projects” to be announced soon to provide the uptake for what he called “a major commitment/allocation” of panels to the U.S. and the rest of the region.
The source of those CIS (here, the “S” stands for selenide) modules is the company’s fast-growing manufacturing complex in Miyazaki, where its third production facility is ramping. Deputy GM Ichiro Sugiyama told me that production equipment is coming into the new 900MW-rated megafactory this month, and the first modules are scheduled for shipment in Q1 2011.
Sugiyama-san described the basics of SF’s own CIS process. The back-contact is standard-issue sputtered molybdenum. The absorber layer also involves sputtering, with a precursor deposition step followed by selenization. He admitted that the company does use a little bit of gallium, making the stack in effect a “CIgS” semiconductor.
Because of Japan’s ban on cadmium in manufacturing, SF came up with an alternative to the usual CdS buffer layer, choosing a chemical bath deposition of zinc-sulfide hydroxide that produces the 200-300nm-thick film. The top contact/TCO relies on a more standard chemical cocktail—zinc oxide.
The technologist, who is based at the company’s Atsugi Research Center, told me that no antireflective coatings are used in the films or on the module glass. The company’s approach to scribing employs a combination of mechanical and laser methods, noting that they could go “100% laser eventually.” He admitted this would be “interesting” because of lasering’s reduced scribe losses and ability to produce lines of narrower widths.
The production of a bigger, 1.255 × 0.977m module (to be released to the market early next year) forced the company into optimization mode, given the challenges of adjusting uniformities to the larger surface area—although Sugiyama-san would not divulge any specific uniformity or yield numbers. He did say that inline process inspection and control are prevalent throughout the production line, including for incoming glass.
The company’s smaller, 1.235 × 0.641m modules are already certified for the U.S. market, and the new megaplant itself will be UL-certified by next year. Median conversion efficiencies off the manufacturing line range from 11% to 11.5%, according to the Japanese exec, and the roadmap calls for those numbers to reach 14% (as measured from “frame edge to frame edge”) by 2014.
Ridiculous uncertainty. A look ahead to 2014 also came from SunPower’s über-spokeswoman Julie Blunden, who stepped into the breach to represent the company at the PV Fab Managers Forum on the topic of U.S. manufacturing challenges. One of those challenges actually has to do with how the uncertainty of market predictions makes corporate and project planning rather difficult.
She showed a slide that included demand forecasts for the global PV market in 2014 drawn from a variety of “reputable” data-crunching research-house sources. The numbers ranged from a low of 14GW to a high of 45MW (!). Given the admittedly ridiculous 31MW spread in predictions, “one can say that it will be a strong growth market in 2014,” the SunPower EVP smiled.
Single-crystal persuasion. One new company that might be around when the PV market grows to 14 or 24 or 41GW is Confluence Solar. In a classic case of trade-show kismet, I ran into company CEO Tom Cadwell and got an exclusive news update on the spot.
I hadn’t seen the industry veteran since he was running silicon components maker Integrated Materials Inc. earlier in the decade. He once headed a division of MEMC, and also helmed CMP equipment manufacturer Strasbaugh before that stint.
His involvement with Confluence should come as no surprise, since the company makes what it calls “a premium quality, single-crystal silicon substrate,” which can supposedly be produced and delivered at a lower price point than the usual multicrystalline materials.
The technology cavorts in the realm of high-efficiency N-type applications. They’re able to make silicon with consistent/constant resistivity and controlled lifetime characteristics “all the way down the ingot,” according to Cadwell.
He told me that the development work is done, and that Confluence’s secret-sauce silicon is being run through multiple customer qualifications. “The next three months are critical… we’re far along with quals with some of the big guys.”
An initial 150MW facility is being built out in Hazelwood, MO, but longer-term plans call for the construction of a $200 million, 200,000 sq ft 1GW factory in Tennessee, with ground-breaking possible by the first quarter of next year—if the funding materializes.
News on the capital-raising front should be coming soon. Cadwell said that the investors are involved in their Series B diligence—a process that he hopes will close in a month or so.
BEES PHOTO BY TOM CHEYNEY; REST OF PHOTOS COURTESY OF RESPECTIVE COMPANIES
