Online information source for semiconductor professionals

Although the Solar World 1 photovoltaic-powered car driven by the University of Bochum (Germany) team finished a very respectable third in the just-completed Dallas-to-Calgary North American Solar Challenge race, the company itself has the number-one position in terms of vertically integrated crystalline silicon-based solar manufacturing on the continent.

At last week's Intersolar show in San Francisco, company chair/CEO Frank Ausbeck said the new 100-MW capacity, fully automated monocrystalline module-production line in Camarillo, CA, has been completed. Expansion and modernization work at the former Shell Solar site began in 2007 (before that, Siemens and Arco owned it, btw), and SolarWorld has said it spent at least $26 million on the project last year, boosting output at the factory from its previous manual-/semiautomated line maximum of 35 MW.

German engineering excellence finds automotive and PV synergy
in latest solar challenge race. (Photo courtesy: SolarWorld) 

Gordon Brinser, Solar World USA's VP of operations, told me after his keynote on the future of crystalline silicon at the North American PV Advanced Manufacturing Technology Conference that the company's new integrated wafer and cell fab in Hillsboro, OR, "will have 100 MW of capacity functional by year end... with production output reaching 100 MW by Q109." The facility (bought from Komatsu, reportedly on the cheap, in early 2007) includes 480,000 square feet of manufacturing floorspace, which will be ramped up to 250 MW next year and possibly 500 MW by 2010, making it the largest c-Si production plant in North America.

The first wafers were cut from silicon ingots "the last week of June," he confirmed in a follow-up email, and the "tools are being moved in, and we're working with the normal delivery and startup issues with specific vendors." Some equipment deliveries have been pulled in, others pushed out, but most of the systems are now on the factory floor. "We work together with the tool vendors to really drive the overall efficiency in their 'tool space.' Whether this is through higher productivity or small tools or lower-cost processes, it all gets us to our target--grid parity--much faster."

Brinser focused on "supply partners challenges" during the latter part of his conference presentation, citing cost improvement, speed, and innovation as the keys to success. He broke down six "areas of opportunity" for equipment and consumables suppliers. While the tool companies need to work on their productivity per cubic foot and consumables houses must get more consistent quality and eliminate "the variation in their products," both supplier categories could improve their efforts in standardization, sustainability, and innovation, and both need to be wary of what he called "creeping elegance."

Noting the "capacity limitations" of the supply chain, Brinser said there are real opportunities for "new and existing suppliers that can capture the real essence of PV need," arguing that companies must remember that solar "is different from semiconductor" in terms of the requirements for bigger cost reductions and more speed, as well as the much higher volume involved.

In terms of cost, he spoke of the need to "lean the tools," to reduce the total equipment footprint, with minimal weight, auxiliary, and structural add-ons, as well as relatively uncomplicated operational features. He called on "productivity per cubic meter" to become an "embedded metric," and for time-conscious cost reductions to be driven across the areas of installation, startup, and maintenance.

He bemoaned the lack of both hardware and software standardization, noting that common hardware standards would greatly speed up "initial connection and startup," reduce the "cost of maintenance and troubleshooting," and provide more "flexibility to accommodate innovation." Interface and other software components would also benefit from this push for standardization.

As for the avoidance of "creeping elegance," Brinser said that "increases in specifications must be driven by data and clear needs," since "every additional 'feature' adds complexity and cost." The question that he thinks should be asked is, "Does the new, more costly approach add value to the module in the field?"

The industry "must avoid the previous models that choked innovation." He cited improved automated wafer handling as a critical need as well as an acceleration in the development of "crystallization and wafering technologies," which he sees as "lagging." A "mentality of sustainability" must be "built into all phases of tool design, the manufacturing of tools, consumables, and components, packaging, and end of life" of the products, and that this, Brinser believes, will "eventually differentiate companies."

He concluded his talk with a goal/challenge for the solar PV production community: "The target must be to double the watt-peak throughput of a square foot of manufacturing space every 24 months." Although he admitted during the Q&A that no baseline numbers yet exist on which to quantify this metric, he urged the industry to come up with them as soon as possible as part of an overall drive for more standardization.

Taking Brinser up on his call to arms, Curt Ward of ABRE, who spoke about SEMI's standards activities (which include getting a PV global roadmap in place by the second half of next year), buttonholed the SolarWorld exec during the break and encouraged his participation.

For standards to really take hold and proliferate in the solar manufacturing community, there must be significant input from Brinser and other factory operations experts who see the cost-reduction and productivity improvement benefits of getting the industry on the same page.