Your daily dose of photovoltaic technology developments and solar news

• Most Popular Articles

  • UK feed-in tariff is announced: payback begins April 1
  • Intel outside: Chipmaker moves to install more PV power systems, equipped with First Solar modules
  • Suntech predicts excess solar supply will move to U.S.
  • New Product: SunPower ‘T20 Tracker Evolution’ delivers maximum energy with 128-cell, 400-watt mo
  • Project Focus: GA-Solar to build 300MW solar PV power plant in New Mexico

• Popular Technical Papers

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

• Subscribe to our weekly newsletter

  Name
  Job title
  Company
  Email

One of the key tenets in First Solar’s cost-cutting and continuous improvement mantras is the ever-increasing conversion efficiency of its cadmium telluride thin-film panels.  "A tremendous amount of our ability to drive down cost has to do with improving the efficiency, one of the fundamental metrics that drives our technology," said company president Bruce Sohn during last week’s analysts meet. Given how progress has gone so far, 11% module efficiency should be right around the corner, with management exuding confidence that the magic number will pop to 12.5% within a few years.

Market watchers like Barclays Capital’s Vish Shah share this level of conviction, stating in his FSLR analyst day takeaway report that "we believe 12.5% efficiency targets are more near term (perhaps 2012-2011 timeframe). Our checks suggest that efficiency enhancement programs in Ohio are largely complete," he continues. "With the Malaysia ramp now over, Ohio efficiency programs can be rolled over to the Malaysian factory."

Although it remains unclear just what those best-of-breed conversion efficiencies may be on First’s top-performing "mothership" process development line at its Perrysburg, OH, site (and don’t expect an answer from the exec team on that one), it can be reasonably assumed they’ve already crossed the 11% bar.

But what of the longer term? When David Eaglesham, the company’s seldom-seen VP of technology, proclaimed during his presentation that there’s plenty of room for CdTe conversion efficiencies to push beyond the limits of the well-established 16.5% record set by NREL with its CdTe "hero cell," his comments elicited a collective eyebrow raise for those more technologically minded listeners.

Eaglesham shared data showing that since 2001, CdTe has closed the efficiency gap with poly- and monocrystalline silicon PV by about 3.3X, improving some 70% over the decade so far—a much brisker clip than the improvements shown by the cSi siblings. Part of that advantage results from what he called CdTe’s "inherently faster cycles of learning" compared to the bulk silicon boys and girls.

When you’re dealing with cycle times where glass goes in one end of the line and 2-1/2 hours later a finished module comes out, technologists can get test results pretty fast.

He spoke of how First Solar does a lot of heavy analytical lifting to determine which efficiency components are higher risk and how to find multiple pathways to possible solutions to help mitigate that risk. Considering the 11% efficiency goal "looked like a big mountain to climb" when they were dwelling in the low 9s, the approach seems to have worked. 

He also said he "felt lucky as a module technologist to have a direct line of sight to the installation and EPC businesses," which offers an "opportunity to really understand" what’s going on in the field at a detailed level. And of course, the company’s production scale and financial vigor don’t hurt the efficiency improvement process either.

But for all its chest-thumping about the benefits of economies of scale, size also can have a detrimental effect. You could even call it growing pains.

Eaglesham admitted that "as the scale of the company gets larger, it does become more challenging to improve efficiencies," noting how "implementing improvements on a very large base" becomes more difficult. He noted the choice between retrofitting existing facilities with improved processes and doing so on freshly minted production lines, and admitted that "in many cases, you can implement on new capacity more readily."

Sohn spoke to this issue during the meeting’s question-and-answer session as well. "Historically, we’ve always said that our learning rate was about ½% in efficiency per year and about 3% points in line run rate per year," he said. "But we’ve been beyond that rate since 2006 until today it’s been about 25% CAGR."

"If we look out in the future, while we anticipate continued growth, it’s actually slower than what we had over the last several years," he continued. "The ability to understand the technology becomes a real challenge. One of the good things about executing to a roadmap is of course we become very predictable in terms of our operational performance. But it also gets us to a point where we start to sometimes forget that we’re dealing with a high technology product and our ability to understand the improvements comes down to our understanding of the physics and the fundamental science and the chemistry and the mechanics and so forth, in terms of the way the factory operates.

"That’s an ongoing challenge that the engineers and improvements teams inside the factories and the R&D center are constantly working with," concluded Sohn.

For the near term, Eaglesham said that First Solar doesn’t just have "high confidence" it will reach the 12.5% module efficiency metric, they have "very high confidence," with all the concepts necessary to achieve that goal "locked in." The pathway will be through improving "light transmission into the existing device" and won’t require any major changes to the basic design.

As he shifted into dance mode yet again, the VP discussed the "clear avenues" available to achieve this. Reducing the thickness of the cadmium-sulfide layer, a "significant loss mechanism," can help, but the challenge is do it in a manufacturable, controlled manner (since "thinner" doesn’t necessarily mean "easy to manufacture"). The transparent conductive oxide (TCO) films offer another area for improvement, and evidently First is working on some proprietary chemistries to "drive to best in class" capabilities. Glass transmission is another piece of the puzzle where continuous enhancements lead to more light getting to the existing device, he said. And then there’s First’s ubiquitous scale and how it "brings advantages" across all of the areas.

But what of future conversion efficiency improvements, which might reach and even pass the storied 16.5% hero cell created by NREL?

Eaglesham ran through some slides on the theoretical limits of PV (30% for a perfect single-junction cell), the "top-down efficiency potential of CdTe" (21% is a reasonably established number for a theoretical device, then after factoring in module losses and production averages, a 16-17% production module seems plausible) and then "bottom-up efficiency potential" (largely the same conclusions as the top-down chart, but the production panels pop up to the 17-18% range). At the heart of the matter is "voltage," which he called "the biggest lever in closing the gap between 16.5% and 30%."

That’s right, First Solar posits the average conversion efficiency of its CdTe panels—not champion cells, but production modules—nesting in the 16-18% range on the company’s extended roadmap.

In discussing a slide titled "1000 pathways to >16% and beyond," Eaglesham pointed to a quintet of material science and device physics  engineering avenues to get to the high teens: optical, contact, grain-boundary, band, and dopant.

These categories appear on every thin-film PV efficiency roadmap, he said, and taken together, they offer a "laundry list" of technical and engineering tricks that underscore one of his main takeaways: "CdTe still has lots of headroom for improvement."   

The optical pathway, which offers solutions that "improve the given amount of light hitting the front side [of the cell] that winds up in that device," includes tweaking or directing the absorption of the light, enhancing the coupling, and the using better antireflective films, he said.

Like all thin-film PV purveyors, First looks hard at the contacts, since both CdS and back metal "are  significant in driving performance" and can be improved with more highly engineered, even combinatorial materials.

Along the grain boundary, a place near and sometimes not so dear to the hearts of CIGS and CdTe techies, figuring out how to "get minority carriers across the boundary is inherently difficult," he pointed out. To engineer that grain structure, researchers might add impurities to change the electric fields or alter process temperatures to get better behaving structures. 

Band engineering, a familiar tool in the semiconductor process kit, involves playing with the composition of materials to produce different kinds of fields within the device and using energy bands to manipulate those carriers, he said.

He didn’t mention any specifics about how dopant engineering came into play with CdTe, categorizing that as more of a concern for those working with micromorph silicon thin films.

"You can’t drive to 16% devices by focusing on any one of them," Eaglesham opined, "you need to focus on all of them." Again, he reiterated the familiar First mantra of scale, how the company’s resources allow it the ability to "invest across a wide portfolio" of potential approaches and to maintain its sustainable differentiation and its lead in PV’s technological competition in the years ahead.

While Eaglesham’s tantalizing propeller-head dance routine offered a peak into the First Solar technology vault and underscored how they "keep raising their own bar," it provided little true granularity of just what the heck they’re doing in their labs to keep the conversion efficiency needle moving. (Single-walled carbon nanotube networks as next-gen transparent back contacts, anyone? An NREL team is working on that.)

Still, when the horse-race handicapper concluded by saying "we believe that 16-18% [efficiency] as a practical production potential is very achievable," it was clear that he and the rest of the First team remain quite confident that they know how to get from here to there, and to continue moving forward with agility, despite their company’s increasing girth and complexity.

To go to the first installment of the two-part blog on First Solar's analyst meeting, click here.

First Solar's Bruce Sohn will deliver the keynote address at the "Challenges in PV Thin Film Manufacturing" conference on Thursday morning, July 16, at Intersolar North America in San Francisco.