Playing it safe

Middle-of-the-road is a phrase often used disparagingly, from its association with safe, central positions in politics to the more dire output of 1970s pop music. But it probably best sums up the global photovoltaics market right now in terms of the multicrystalline silicon (multi c-Si) technology that dominates and the type of modules that have proliferated, which typically exhibit decent performance at low cost.
PV manufacturers aggressively pursuing high efficiency cell technology road maps are the exception, not the norm, and are certainly not the leaders by market share. Within the global PV market still dominated by crystalline silicon (c-Si), the mainstream “standard” module technology of choice is multi c-Si. Monocrystalline silicon (mono c-Si) cells, which are made from the highest-grade silicon and associated with step-changes in increased efficiency, are more costly than cells made from polysilicon wafers.
The market share of modules based on multi c-Si cell technology has grown consistently in recent years. In 2008, the market was roughly split between mono c-Si and multi c-Si. Today, multi c-Si dominates, according to NPD Solarbuzz. The trend is likely to continue over the next few years.

No ambition for high efficiency

Several years ago many of the leading PV module producers communicated, to varying degrees, proprietary technology, through R&D partnerships and collaborations, for making high efficiency solar cells, typically requiring complex new cell architectures and designs using mono c-Si technology, which, in turn, would entail investments in new production processes and toolsets.
Instead, manufacturers have relied on making gradual improvements to multi c-Si wafer and cell processes. The result has been the slow upward creep of conversion rates across much of the industry year-on-year.
GTM Research suggests that the average increase in efficiency of conventional panels, based on multi c-Si technology, is likely to improve by roughly 2% over seven years, or, on average, a rate of about 0.3% annually. This is confirmed by IHS Research, which began measuring module efficiencies in 2009.
“We see this pattern continuing over the next three years at least. All the top players are progressing on their mainstream products at the expense of any serious effort to introduce high efficiency module technologies based on advanced cell architectures,” says IHS Research Senior Director Henning Wicht.
Only two of the largest PV manufacturers have made any real effort to bring into volume production high efficiency cell technologies.
They are Suntech, with its Pluto cell production process, in a licensing agreement and R&D partnership with the University of New South Wales, and more recently Yingli, which has worked with the Energy Research Centre of the Netherlands (ECN) and Amtech Systems, using n-doped mono c-Si technology instead of the industry’s standard p-type multi c-Si technology. But Yingli’s annual capacity for making its high efficiency Panda modules is about 600 MW and it is not the main focus of module shipments. The company declined to be interviewed for this article to discuss demand trends for high efficiency modules versus standard modules.
Yingli, which has replaced Suntech in the number one spot, achieved a module production figure of 1.95 GW in 2012 and probably counts as one of the biggest spenders on production equipment in recent years.
But until its lines are run at optimum utilization rates, Yingli is likely to delay upgrading more of its production capacity to produce the Panda modules, according to NPD Solarbuzz Vice President Finlay Colville.
Suntech also had its own challenges with implementing the Pluto cell production process reliably but now that the company is going through restructuring proceedings it is questionable if any new business to emerge from a buyout will be emphasizing the Pluto technology.
Aside from upgrades, for example acquisitions of new tools to replace older equipment or to achieve incremental efficiency improvements, genuine efforts by the biggest manufacturers to introduce high efficiency modules into mass production have remained isolated and have not caught on across the industry.

Road maps

Going back several years, there was a common belief that once the PV industry reached high volumes, then technology road maps would kick in. In other words, the leading PV manufacturers would all buy into a collective effort of making continued improvements and investments in manufacturing, resulting in modules with higher power ratings, based on new solar cell technologies that could lead to efficiency increases of 1% or more, as opposed to incremental efficiency improvements.
In other high-tech industries, including semiconductor and LCD/flat panel displays (FPDs), every manufacturer has to acknowledge road maps, otherwise they are surpassed by the competition. Road maps create the pull for technological innovation, translating into demand for new processes, tools and production equipment, allowing equipment suppliers to align the development of their own offerings with the road maps of their manufacturing customers.
###MARGINALIE_BEGIN###

###MARGINALIE_END###
Conventional solar photovoltaic cells using p-type multi c-Si, as well as mono c-Si wafers, are already nudging the limits of power conversion efficiencies, “which cannot be exceeded by cheap improvements in production: a technology shift is therefore necessary,” equipment supplier Meyer Burger stated in a paper it published recently concerning its high efficiency heterojunction cell technology. Technology shifts tend to entail a rush of new equipment purchases.

Risk averse

In contrast to the prevailing belief within the other industries, the solar PV industry has not seen the central requirement for a road map, so most PV manufacturers have not implemented one and those few that have tried have often failed.
This is due to several factors, including overcapacity in the industry and the decline in the profitability of all PV manufacturers, according to Colville. But the oversupply of polysilicon, the raw material required for making multi c-Si wafers which are processed into solar cells, has probably had the biggest impact. “Big players have not had to implement any high risk approaches to achieve higher solar cell efficiencies based on new cell architectures, requiring advanced production processes and new tools and equipment, as the falling cost of polysilicon has reduced manufacturing costs by so much,” he says.
This will only change when the market leaders see a reason to produce high efficiency solar panels as part of a strategy to achieve new market share. Now there is no need for them to take the risk involved to do this.
Transitioning to higher efficiency cell production requires extensive R&D, testing and qualification, requiring capital, and this is something solar companies are in short supply of right now. It is a state of affairs that is unlikely to change for the next two years, Colville and Wicht agree.
“The industry is still going through a shakeout, so why would companies that see competitors going out of business see a need to implement major changes in manufacturing,” adds Colville. The overarching desire is to return to profitability slowly.

End user market needs

End use markets have a role to play in manufacturers’ decisions. Supported by attractive solar PV incentive rates and a strong rooftop market because of the limited availability of space for ground-mounted installations, demand for premium high efficiency mono c-Si based modules is more acute in Japan than in most other markets.
Traditionally the solar rooftop market was expected to sustain demand for modules with higher power ratings, as space constraints favor smaller panels. However, the creeping efficiency improvements for standard panels mean that the gap is closing between c-Si panels and more expensive, premium mono c-Si products.
Wicht says: “Most conventional multi c-Si based modules have 250 W power ratings, compared with 210 W five years ago, while good mono c-Si based modules are between 265 and 270 W. Furthermore, it’s questionable whether the mainstream rooftop market is likely to have power requirements that exceed a module with a 270 W output.” What about the ground-mounted market?
As well as having high power ratings, higher power conversion efficiencies, mono c-Si modules can achieve the longest lifetimes, tend to perform better than similarly rated polycrystalline solar panels in low-light conditions and suffer less efficiency losses in very high temperatures compared with multi c-Si modules. PV equipment manufacturers, including Meyer Burger, argue that future solar markets, particularly those in the sunbelt regions, will benefit from high efficiency solar cell technologies.
“For ground-mounted solar projects, typically utility market-driven, it is even more important for the overall cost of the project to be kept down, which means using standard modules that are lower cost,” says Wicht.
Or, as Colville puts it, if the industry’s downstream channels can fill sales pipelines and meet return on investment requirements with p-type multi c-Si modules, then this technology will be sold and will create the demand that drives the utilization rates of different production lines.

Closing down the efficiency gap

This does not mean that manufacturers are averse to promoting a product range that offers standard and more premium choices.
Trina’s Honey series, pitched at the rooftop market, includes multi and mono products. The 60 cell multi c-Si modules have a power rating of 250 to 260 W, based on a maximum module efficiency of 15.9%, while the company’s mono c-Si 60 cell Honey M modules have a power rating of 260 to 270 W based on a maximum module efficiency of 16.5%.
In Yingli’s Panda series, its 60 cell modules have a maximum power rating of 270 W based on a maximum module efficiency of 16.5%. In the company’s multi c-Si range its 60 cell modules achieve a power rating of 250 W based on 15.3% module efficiencies. Though JA Solar has made some recent announcements with regards to solar cell efficiencies – 18.3% for multi c-Si cells and over 20% for mono c-Si cells – these have yet to be integrated into the company’s manufacturing lines.
In any case, the point is that multicrystalline silicon modules are closing the efficiency gap.
When comparing this to some of the most powerful modules available on the market, from SunPower – the company’s new X Series panel stands at over 21% – then the biggest manufacturers’ mainstream panels and their high efficiency versions are all bunching around the 15 to 16.5% mark.
Several factors are contributing to the gradual performance improvements that are being teased out of multi c-Si modules, for example improvements to yield distribution, achieving better quality wafers and turning up fab utilization rates. These developments all bring about some tangible benefits.
Much of this is achieved by further advances in multi c-Si casting, implementation of better monitoring and qualification of production processes and inspection of incoming wafers and equipment upgrades for process steps to achieve incremental efficiency increases.
But this low-hanging fruit concerning efficiency increase has mostly been harvested by now, according to ECN. There is still great potential for climbing, as opposed to crawling, up the efficiency ladder.
But this is only truly possible if advanced processes and device designs are introduced, states ECN.
In July of this year the trade press reported that new PV manufacturing capacity expansions in China would only be authorized if these met minimum efficiency thresholds of 20% conversion efficiencies for mono c-Si cells and 18% for polysilicon cells.
Might such a plan give the industry the kick that it needs to implement a technology road map?
Unfortunately, sources in the PV industry cannot confirm the announcement as coming from the government and until the development is qualified further it should probably be treated with caution.
Some sources speculate that if the Chinese government is seriously considering such a plan then it would be to force a shakeout of the tier-2 and tier-3 producers as tier-1 manufacturers all have high efficiency technologies in-house.

The glass-glass question

With such acute cost pressures applied to PV, the focus on innovation is switching to other areas.
More and more PV producers are offering glass-glass modules. Interesting claims are being made about these modules, where PV cells are encapsulated between two very thin sheets of glass, typically 2 mm or less in thickness, rather than the standard module format which consists of PV cells sandwiched between a front panel of glass and a polymer backsheet.
Dual glass modules do not require framing, reducing the overall bill of materials and are durable for decades. German company Solarwatt made headlines earlier this year when it announced it would be converting all of its modules to glass-glass and Austria-headquartered PV Products also supplies a dual glass module.
By the end of this year Talesun, a Chinese PV company focusing on high quality as opposed to low cost PV modules, will also make its glass-glass module available once the product receives certifications.
These companies all offer glass-glass modules as part of a strategy to differentiate their product line. But now Trina, one of the largest PV manufacturers, is also promoting the availability of its own glass-glass offering.
Wicht says: “It’s a bit too early to say what sort of impact these glass-glass modules will have on the market. The key driver for demand for glass-glass is whether it is cheaper to use these frameless modules compared with conventional ones.” Trina Solar’s PDG5 60 cell frameless glass-glass module, which needs no grounding, comprises two layers of 2.5 mm heat-strengthened glass. The company calculates that for a 2 MW commercial rooftop in New Jersey, compared with a standard 60 cell 245 W module with frame and commercial racking, the PDG5 module, with equivalent output and using Trina’s own racking system, achieves a balance of system (BOS) cost that is US$0.10/W lower, so a 5% reduction of the payback period can be achieved.
The building-integrated and building facade market, where glass-glass modules are a more aesthetic option, is forecast to remain niche.
The real test of whether glass-glass modules are going to present a threat to the dominance of the standard framed silicon PV module is if they prove more cost-effective in the ground-mounted market.
Despite their advantages such as durability, glass-glass modules are more fragile to handle, because of the thinner, albeit strength-treated, glass. But if the use of robotics and related technologies continues to rise in ground-mounted project installations, at the expense of manual labor, the automation of solar farm array installations could be compatible with glass-glass modules.
However, it is still too early to predict the degree of impact that glass-glass modules will have, according to Wicht.
NPD Solarbuzz is in the middle of compiling research into the glass-glass module market, so it cannot yet share findings. For now the jury is out.

Conclusion

In an industry going through a shakeout, with more exits inevitable, no one wins prizes for investing in new, riskier high efficiency module technologies. For the foreseeable future polysilicon will continue to take the lion’s share of the global PV market. It’s what the industry wants right now.