Your daily dose of photovoltaic technology developments and solar news

• Most Popular Articles

  • Q-Cells CEO exits amid investor anxiety
  • Moser Baer achieves 7.3% module efficiency through its proprietary process
  • Solar photovoltaic market grew 6% in 2009, says Solarbuzz
  • Yingli Green to add 300MW of capacity; tops $1B in annual revenues, achieves record module shipments
  • Project Focus: SunPower scores 200MW rooftop PV contract with Southern California Edison

• Popular Technical Papers

  • Photovoltaic Fab Managers’ Forum Review
  • Photovoltaic industry 2009: a journey into uncertainty
  • Overview of automation in the photovoltaic industry
  • Thin-film CIGS starts to come of age
  • CIGS, CdTe thin-film PV equipment sector emerges, standardization remains elusive

• Subscribe to our weekly newsletter

  Name
  Job title
  Company
  Email

While Q-Cells weathers the price erosions and market uncertainties of the crystalline-silicon sector, holding its cSi cellmaking production capacity steady for now, its top-performing thin-film unit and flagship of the new Q-Cells Modules brand—Solibro—is about to surpass the century nameplate mark as its second manufacturing plant comes online. The German copper-indium-gallium-(di)selenide company with Swedish roots will have 90MW of nominal capacity to go with its existing 45MW, reaching a total of 135MW, according to Q-Cells boss, Anton Milner. But based on what Solibro COO Johannes Segner told me at EU PVSEC in September, that new line in Thalheim may actually reach the 100MW mark, once it’s fully optimized and running at volume. And that’s just a taste of what we discussed.

The COO said that the new fully automated plant will have all tools installed and qualified by the end of this year, only a year or so after groundbreaking took place. The first line has seen total area conversion efficiencies for the glass-glass CIGS modules go from around 9% in August 2008 (the same month the company shipped its first products to customers) to stable production runs of about 11% +/- 0.5%, with very tight deviations, he said.

Champion SL1 CIGS modules on Line One have reached at least 12.3% efficiencies, and Segner believes the median production number should exceed 12% in 2010, with further tailoring of the CIGS coevaporation process’s concentration profiles and other enhancements. Eliminating shunts through contamination control and defect avoidance approaches will help boost the efficiencies to the high 12s, he added.

Solibro has hit efficiencies as high as 16.7% on small-area devices at its R&D facility in Uppsala, Sweden, and Segner sees gains from a worst case of 13% to a best case of 15% possible within a couple of years. But to get to 15% and beyond will require changes in certain parts of the process (such as one possible scheme to leave zinc oxide out and put the conductive elements directly on the CIGS itself), and any modifications will have an as-yet-to-be-determined impact on the other steps in the process. Still, the operations chief is optimistic that his team will be able to hit the big one-five, given the quality of the R&D results he’s seen.

Another part of Segner’s optimism has to do with the quality and quantity of his 60 or so engineers working on the two production lines (and another 30 at the R&D site). “We take engineering very seriously,” he said, citing the company’s work on building new laser and mechanical scribe tools, modifying the CVD system, changing the ZnO approach completely, and generally improving tool uptimes as well as product quality and reliability.

Inline metrology has been upgraded, with 100% of the products checked after every process step, and every piece of glass identified and tracked. This info is linked to a database that facilitates data mining and root cause analyses that often reveal lots of effects--what Segner calls “scatter”--so that when the process is changed, the impact can really be seen, helping to speed up the production ramp. 

The COO (pictured above) had strong words for the vendor-supplier community. “There’s no unstable process, only unstable equipment,” he quipped.  Although he believes the tool suppliers are in transition and “moving in the right direction,” he would like to see them become closer and “more intimate” with their customers (such as his company).

He explained that every Solibro supplier gets a weekly uptime report with lots of details, but some vendors don’t take advantage of the opportunity to do some root-cause analysis of their own, instead reacting with “is it our fault?,” instead of “do we have a problem?” and trying to solve it in cooperation with the customer-partner.

Noting that he has to “force” certain toolmakers to come into Solibro’s fabs, he sees a shortage of applications engineers in the vendors’ ranks and thinks adding more talent in that area would be a “way for them to differentiate themselves” from the competition.  He cited another example, recounting  when he wrote down the technical specs he wanted from the tool suppliers that were meant to reduce the number of spare parts in the equipment. The vendors were at first reluctant, but three of them thanked him later for forcing them into making the changes and helping them to differentiate.

As a result of efforts like these, “we now have leaner operations,” he said, noting that the number of vendors of production-related tools has shrunk from 45 when they started the first line to 15 now. He also revealed that the price for the complete automation system for the new fab came in at 6 million Euro, about half the amount quoted by some of what he called the “industry leaders” (none of which ended up getting the order).     

The new factory will be producing Solibro’s latest module, the sleek black SL2, which is slated for market launch in Q2 2010. This frameless panel features 150 CIGS cells compared to the SL1’s 118, and it’s 25% wider than the existing models—79cm vs. 63cm—and will have a watt-peak range of 95-115 compared to 70-85Wpk for the current devices.

A preliminary SL2 data sheet distributed at PVSEC showed nominal standard test condition efficiencies up to 12.2%, with improved low light intensity and temperature coefficient performance. Given the enhanced energy yields—Solibro loves to point out its “kilowatt-hour per kilowatt-peak installed” advantages versus CdTe and crystalline (SL1s have supposedly shown an average of 8% higher yields in all orientations in side-by-side field tests with SolarWorld’s 225W models)—the new, more powerful modules will also extend the company’s targeted market sectors from residential and commercial-industrial rooftop/BIPV applications to free-field utility-scale installations.

During a PVSEC presentation by Solibro CTO Lars Stolt, he offered a cost-per-watt-peak production goal of €0.78, slightly more ambitious than the €0.80 figure cited by Q-Cells’ Milner. The company’s chief techie noted that when annual production capacities go above 100MW—which Solibro will do soon—that materials costs for making CIGS account for >50% of the COGS, with front glass, contact layers, the CIGS itself, and substrates making up the largest slices of the pie chart.  He also said that reducing equipment costs by 50% (like automation perhaps?) has the same impact as a 1% improvement in conversion efficiency.

Will Solibro hit its efficiency and cost targets as its scales into triple-digit megawattage, and what kind of headroom does it have for future improvements? Although getting well below a Euro per watt is laudable and puts the company ahead of most CIGS manufacturers, it’s still a large slice off from the €0.57 ($0.85) achieved by the CadTel crew at First Solar in its most recent quarterly results.

PHOTOS BY TOM CHEYNEY