Recent industry events have provided a forum for analysts to opine on the future of PV technology, both thin-film and crystalline.
At a recent Applied Materials/IEEE PV Technology Symposium, Paula Mints, principal analyst, PV Services Program, and associate director of the energy practice at Navigant Consulting, made it clear that the industry needs a diversified product portfolio and all technologies have a place.
“All technologies are working toward higher efficiency,” Mints told SST. While one of the attributes of thin-film technologies is more kWhrs out of the system, she pointed out that thin-film systems are more expensive to install, as the technology is lower in efficiency and requires more balance of systems and area, and may have a higher price. “Thin-film offers more kWhrs out of the system (performance) because of better conversion efficiency at low light conditions and conditions of diffused light, better performance at higher temperatures and improved shadow tolerance,” she said. And given the silicon raw material that will soon be coming online, Mints thinks there will be a greater amount of higher-efficiency crystalline available at a lower price. “Since crystalline silicon is available at a lower price, it will be more difficult for thin-film to compete. Thin-film PV will need to be priced lower than crystalline in order to compete.”
Mints observed that thin-film technologies share gains in 2006-2007 were primarily because of the silicon shortage, which gave them “an opportunity to overcome market wariness of lower efficiency and perceived lower reliability.” She expects the raw material supply to begin easing in late 2009, however, and “most of the silicon coming online is already sold under long-term contract.”
Jeff Watt, IEEE co-chair of the AMAT/IEEE event, observed that lack of enthusiasm about thin-film technology growth on the part of some is most likely driven by the view that its recent rapid growth (primarily attributed to First Solar) is a result of the shortage of silicon, but as the shortage of silicon abates, the growth of thin-film cells (especially CdTe) will slow. “At the same time, we saw that Applied Materials sees significant growth and cost reduction in thin-film silicon cells based on its SunFab production lines,” Watt told SST. “Both crystalline and thin-film silicon cells continue to achieve incremental efficiency improvements.” Referring to presentations at the event, Watt also noted that while multijunction cells have achieved the largest rate of efficiency improvement, they remain a niche technology in concentrator applications and it is not clear that the technology will ever achieve a significant market share.
IEEE co-chair Philippe Jansen (also of IMEC) observed that the first spot in shipments is held by crystalline PV, and due to its high efficiency and current investment in new equipment for less breakage and improved yield it should grow its lead over thin-film PV. “Shipments of thin-film PV today are small and comprise only a handful of fabs,” he pointed out. “Moreover, the lower efficiency of thin-films currently makes them not interesting for markets such as the residential market, despite their aesthetic advantages.” He also noted that while thin-film PV solutions address the high substrate cost disadvantage of crystalline PV, their cost remains high as larger R&D and capital investments need to be made compared to crystalline PV to improve breakage, yield, and throughput.
Sheila Bailey, senior physicist, NASA Glenn Research Center, told the AMAT/IEEE event attendees that there is a kind of stalemate in getting a thin-film solar cell from the lab to a commercially viable product, and further explained to SST that the efficiency losses (e.g., optical losses, recombination losses, fill factor losses) are challenging. “The trouble with the thin-films [is] there are much more complicated materials issues,” she said. “And amorphous Si is amorphous, it’s not crystalline, so there are a lot of material concerns that go into this difference in efficiency between the lab cells and the commercial modules.” These problems are being addressed in the research arena, and over time the efficiencies have risen, but she noted that progress is not instantaneous; it’s more of an evolutionary process, not revolutionary. And there are no shortcuts, with a variety of material problems that will need to be investigated and understood. “I think there are even fundamental physical modeling things that are necessary to understand how the defects and impurities interact in a cell, and this is particularly true of the advent of nanocrystalline materials in a cell,” she said. “It’s getting there, it just takes time.”
Alfonso Velosa, research director at Gartner, told attendees of the research firm’s Semiconductor Industry Briefing (Oct. 9 in San Jose, CA) that thin-film PV holds the potential for the lowest cost/kWhr, but “the biggest issue for the PV industry is that it is so expensive.” He noted, though, that “there are strong efforts to reduce costs in crystalline PV, and thin-film PV may reach grid parity between 2012 and 2015.”
While Gartner forecasts a 73% CAGR for thin-film PV between 2008 and 2012, Velosa noted that crystalline silicon PV, with a forecast CAGR of 42% from 2008-2012, remains the market core. “This is due both to the large installed base of manufacturers as well as to its higher efficiency, which makes it attractive for users with limited footprints and a requirement for the maximum possible electricity,” he explained. Crystalline silicon-based PV is expected to reach 13GW sold by 2012; thin-film-based PV has the potential to reach >4.5GW sold by 2012, according to his projections.
Thin-film will gain market share, observed Velosa, especially for solar farms and BIPV (building-integrated PV). Like Mintz, he expects that thin-film PV and crystalline-silicon PV will coexist; however, “the central issue for players in the industry will be how to scale up production in order to achieve economies of scale, while building effective teams and management processes.”
Solar cell production cost comparison. (Source: Displaybank)
Moreover, Brad Jung, solar market senior analyst/GM at Displaybank, thinks that thin-film solar cells have an advantage over silicon wafer-based solar cells in terms of production costs (see table). “Bulk-type silicon solar cell cost is ~US$3/Watt, whereas the thin-film solar cell cost is <$2/Watt,” he told SST following the firm’s recent San Jose Conference. “In terms of silicon usage, an a-Si solar cell utilizes ~1/100 of silicon compared to bulk type solar cells. CIS/CIGS and CdTe types use compound semiconductors so they have relatively lower pressure on material cost as the silicon supply does not affect those two types.
Marketshare forecast by solar cell. (Source: Displaybank)
Despite such advantages, the a-Si solar cell market entry (see figure) will be determined by developing proper manufacturing processing technology for issues such as a product’s long-term security, lifetime, and conversion efficiency, Jung pointed out. “The current a-Si type has 6%-8% efficiency, CdTe has ~9%-10%, and CIS/CIGS has 10%-12%,” he said. And while triple junction and tandem-type a-Si solar cells are being researched, which create multiple thin-film layers to further promote efficiency, “it appears difficult to replace crystalline solar cells in a short period of time due to the low efficiency and the high installation investment cost compared to crystalline solar cells.” He does note however, that the potential in terms of material cost reduction is higher than for the crystalline solar cell.