So much attention gets focused on thin film around these here parts that it's easy to forget that crystalline silicon (c-Si) PV still makes up more than three-fourths of the module market (for more details on this, you can find our free research note on 2009 cell and module production here). To some extent, this is understandable: the obvious exception of CdTe-producer-that-shall-not-be-named aside, thin film is still very much a work in progress as regards commercialization, with much scope for future technological and economic improvement; c-Si, on the other hand, is considered a mature and well-understood (read: boring) technology, with only incremental improvements in cost and efficiency expected over the coming years. And of course, Davids are far more interesting than Goliaths.
It is generally true that efficiencies for most c-Si firms have registered only incremental gains over the past few years; however, amongst this large mass of relatively undifferentiated firms, a small handful of players are attempting to drive step-function improvements in cell efficiency. While some of these firms have been ahead of the rest of the pack for years, the initiatives of others are still in early-stage commercialization. Besides the bragging rights and distinctiveness they confer, efficiency improvements also drive reductions in cost, both on the module manufacturing and the BOS fronts, although the R&D spend required to maintain improvements is not insignificant, and higher efficiency cell configurations can be more expensive to manufacture. Below, we profile the firms that stand at the forefront of crystalline silicon efficiency, both now and looking to the future, and put their efficiency initiatives under the microscope.
Technology: All-back contact monocrystalline
High-Efficiency Product Status: Volume production (2009: 398 MW)
Commercialized Cell Efficiency: 22%
SunPower has been the heavyweight champion of the world when it comes to commercialized cell and module efficiencies for the last half-decade, and by some measure. The company's back-contact cell design, in commercial production since 2005, moves the metal contacts to the back of the wafer, maximizes the working cell area, and eliminates redundant wires (for details, see this). Impressively, SunPower has been able to achieve consistent improvements in efficiency with each successive generation of commercialized cells, and this has translated to gains in the module arena as well. Its Gen 2 cells, currently in high-volume production, have an efficiency of 22%. Further improvements are on the way: Gen 3 cells, which reportedly have already started shipping, have efficiencies in excess of 23%.
The Verdict: As Gen 3 rolls out and exceeds efficiencies of 24% (something the company has already achieved in low volume), SunPower is likely to be the efficiency leader when it comes to high-volume PV cells and modules for the foreseeable future. The problem, as this article by Michael Kanellos points out, is that 24% is awfully close to the realistic ceiling, meaning there may not be much further to go from there. As the other firms on this list start to narrow the difference, the company's price premium will erode, and its high cost structure will come under increased scrutiny. SunPower has already recognized this, and has aimed at what seems to be a realistic target of $1/W by 2014. Whether this will be enough to survive in a commoditized world of low-cost Chinese manufacturing remains to be seen. Fortunately for the firm, though, its downstream business does afford it some measure of insulation.
Technology: Heterojunction with Intrinsic Thin Film (HIT)
High-Efficiency Product Status: Volume production (2009: 255 MW)
Commercialized Cell Efficiency: 19.8%
Ahead of the rest, but a distant second behind SunPower, Sanyo's high-efficiency product has been in volume production for quite some time -- since way back in 1997, to be exact. Its proprietary HIT cell is a hybrid of monocrystalline silicon surrounded by ultra-thin amorphous silicon layers (see here for details). The amorphous silicon layer enables superior temperature characteristics and low light performance compared to standard crystalline silicon technology. Continuous improvements have led to best commercialized cell efficiencies of 19.8% (launched this year), compared to 18.4% six years ago.
The Verdict: Sanyo has the same basic problem as SunPower: HIT costs considerably more to manufacture than standard c-Si. At the same time, its cells are about two percent less efficient than SunPower's, which means the cost pressure is significantly more. Sanyo should continue to hold the number-two spot as regards commercial efficiency over the next three years, but unless it can start driving step-function improvements in either cost or efficiency, this will matter less and less in the commoditized global market. The company will, however, enjoy a competitive advantage in its home country of Japan, where residential systems dominate and space constraints mean that there will always be a preference for higher efficiency products. Additionally, the company is banking on the success of specialty products (e.g., BIPV modules, combined module-battery packs) in less price-sensitive markets going forward to ensure demand.
Technology: ARTisun monocrystalline
High-Efficiency Product Status: Volume production (2009: 16 MW)
Commercialized Cell Efficiency: 18.3%
The brainchild of PV pioneer Dr. Ajit Rohatgi, a Georgia Tech scientist, Suniva began commercial production of its monocrystalline cells in late 2008. Unlike many struggling PV startups that entered the market around that time, the company has gone from strength to strength over the last 18 months. It has exhibited one of the quickest production ramps of any Western PV company, going from an initial 32 MW to 96 MW to a current 170 MW of cell capacity, and is sold out for 2010. By its own admission, Suniva's technology does not represent a radical innovation; rather, the company has its own paste and texture recipes, is able to customize and optimize every layer of the cell design to its own specifications, and has leveraged its considerable R&D experience to optimize each processing step to a high degree.
The Verdict: While Suniva is clearly not going to overtake SunPower or Sanyo any time soon, reports suggest that the company has a much better cost structure compared to these two players, one that is more in line with low-cost manufacturers. That, combined with its current efficiency advantage over other firms, makes it competitively positioned for right now. A 19% efficiency cell is in the works and should maintain competitiveness in the near future as well. The key question is whether the company can maintain this advantage going forward, given that major Chinese players are hell-bent on playing catch-up (see below). Moreover, the company does not really have a differentiated technology that can guarantee this.
4. Suntech Power
Technology: Pluto monocrystalline
High-efficiency Product Status: Low volume (2010 run rate of 4 MW per month)
Commercialized Cell Efficiency: 19%
The Chinese cell/module behemoth threw its hat into the next-gen c-Si ring in spring 2009, when it announced the development of its proprietary "Pluto" technology, which can be used to retrofit existing cell lines. The Pluto design is based on the PERL (passivated emitter with rear locally diffused) technology developed at Australia's University of New South Wales, where efficiencies of 25 percent have been achieved in the laboratory. Unique texturing technology with lower reflectivity ensures more sunlight can be absorbed throughout the day even without directsolarradiation, and thinner metal lines on the top surface reduce shading loss. Average cell efficiencies in low-volume production were 19%, with plans to hit 20% in two years. The company aimed to reach 450 MW of Pluto capacity by mid-2010, and envisioned that Pluto would eventually become its core product over time.
The Verdict: At 19%, Pluto would place Suntech behind only SunPower and Sanyo in the efficiency stakes. Importantly, Pluto's offers higher efficiency with the potential to simultaneously lower costs: as this GTM article outlined, the cells are made with copper, rather than more expensive silver paste contacts. Pluto thus holds the key to global domination for Suntech. Unfortunately, the company has had trouble ramping production beyond its current levels of 4 MW per month, which it describes as "glitches" with the process flow (see this article for a detailed explanation). Although it is too early to be certain, one is inclined to think that the snags will eventually be resolved; the question is more 'when' than 'if'. Too long, and Suntech runs the risk of lagging behind its Chinese brethren (Yingli and Trina, see below) on both cost (which it already does) as well as efficiency, and facing heated competition from less differentiated Chinese manufacturers (Eging PV, Ningbo, Neo Solar).
5. Trina Solar
Technology: Quad Max square monocrystalline
High-efficiency Product Status: Development (first shipments expected Q3 2010)
Commercialized Cell Efficiency: 18.1% (pilot)
Trina's new cell tries to avoid cutting corners, quite literally -- Quad Max's square shape allows it to harvest more sunlight by avoiding surface area loss typical with traditional monocrystalline cells, which are octagonal-shaped (also known as "pseudo-square"). In a 72-cell module, the additional active surface area translates into a power output advantage of eight percent. The company has developed a new process for the technology, which involves two high-temperature thermal processes, an additional printing and cleaning step, and usage of special paste for the cell surface. Initial shipments are expected in the third quarter of 2010, but don't expect meaningful megawatts until 2011.
The Verdict: "True" square mono has been a talking point in the industry for a number of years without anything to show for it. Trina's move is therefore a much-needed step in the right direction. At 18.1% efficiency, though, it places Trina at the bottom of the pack as far as high-efficiency initiatives are concerned. This will matter less as long as Quad Max does not represent a meaningful increase in manufacturing costs, since Trina is currently the second cheapest manufacturer of c-Si PV in the world, and Quad would drive a 0.6% increase in module efficiency, which would boost product margins. It is still early days for the technology, though: as Suntech's example shows, there is potential for problems galore when going from low- to high-volume production.
6. Yingli Solar
Technology: PANDA N-type monocrystalline
High-efficiency Product Status: Pilot (commercial launch in Q3 2010)
Best Commercialized Cell Efficiency: 18.5% (pilot)
Yingli's foray into the world of high-efficiency cell technology has come courtesy of a three-way research collaboration with the Energy Research Center of the Netherlands (ECN) and process tool maker Amtech Systems, announced in June 2009. PANDA uses ECN's design, the solar diffusion technology and dry phosphosilicate glass (PSG) removal expertise of Amtech's Tempress Systems subsidiary, and Yingli's process technology. The PANDA cell is N-type (for more on that, see here), which means it has greater impurity tolerance and does not suffer from the light-induced degradation that conventional P-type cells do. Yingli claims the corresponding module will also have better performance under high-temperature and low-light conditions. Plans for PANDA are aggressive: in March 2010, the company announced it would construct 300 MW of ingots, wafers, cells and module capacity by the end of the year, and first shipments are expected by the end of October.
The Verdict: As with Trina, Yingli has a ways to go as far as commercial ramp-up of PANDA is concerned, but average cell efficiencies of 18.5% in pilot production are comfortably above Quad Max's 18.1%, although comfortably behind Pluto's 19%. Given N-type's higher impurity tolerance, PANDA also gives Yingli the option of using lower quality (and thus cheaper) polysilicon for its cells, which confers a direct cost advantage. This would further cement the firm's position as the lowest-cost c-Si manufacturer in the world and make life very difficult for its competitors indeed. And with a $5.3 billion loan in hand, the company has some cash to burn before it gets the recipe right.
7. JA Solar
Technology: SECIUM nanoparticle ink
High-efficiency Product Status: Pilot (production began May 2010; commercial production expected H2 2010)
Best Commercialized Cell Efficiency: 18.9% (pilot)
The secret sauce in JA Solar's high-efficiency cell comes by way of California startup Innovalight, which manufactures a proprietary nanotechnology-based silicon ink and licenses a process which allows a simple upgrade to cell lines to boost efficiency - currently by a full percentage point. Importantly, the modification to the production line is relatively simple, requiring only one additional step: the ink is applied using the screen-printing technology typically used by semiconductor lines during back-end metallization. Pilot production is already underway and first commercial shipments are expected any time soon.
The Verdict: Unlike the other firms discussed here, most of which only sell modules, the bulk of JA Solar's business comes from cell sales, which means it is not a direct competitor to them. Success with SECIUM would place JA head and shoulders above other pure-play cell firms in terms of efficiency; only Suniva would come close. And there is potential for further upside -- as discussed in this GTM article, the ink-aided efficiency bump could double to two percent in 2011. Since JA is already the cost leader in cell manufacturing, SECIUM (if ramped up to volume successfully) could provide it pole position on both cost and efficiency fronts. There are two caveats: one, the incremental cost better not outweigh the efficiency gains, and two, nothing really stops other firms from jumping onto the silicon ink bandwagon at a later point -- indeed, as of January 2010, Innovalight claimed it had lined up five other customers*.
That each of these seven firms is employing a different approach to commercializing high-efficiency products should dispel notions that we have reached the end of the road as concerns technological progress in crystalline silicon manufacturing. As with the larger question of PV absorber materials, there is a long way to go before the dust will truly settle on which variant(s) of c-Si will emerge as the dominant leader in the space, if any. There is still much room -- and reward -- for innovation.
*As this article went to press, news broke of Yingli's jumping on to the Innovalight bandwagon as well. For details, see here.