In search of a niche


“It’s five minutes to midnight,” warned Peter Fath, CTO of Centrotherm, at last year’s Fab Managers Forum. In light of increasing global competition and the dominance of Asian manufacturers, the clock has continued to tick and the grace period for European cell and module manufacturers has become even shorter. This became clear at the two-day professional forum of the SEMI PV Group.
Virtually all of the ten cell and module manufacturers with the largest sales are Asian, in particular Chinese, companies. They produce with an annual capacity of at least one gigawatt; a size where most European and German manufacturers are no longer able to compete. Analysts such as Henning Wicht from iSuppli expect to see developments in global capacity of more than 40 gigawatts in terms of module and cell capacities, while the worldwide demand for photovoltaic systems is only half as great, at presumably twenty-one gigawatts.
A view shared by most of the participants at the forum was that such overcapacities will result in further pressure on prices as well as price collapses. There are two sides to the coin: “In order to put real momentum into the market the module price must drop from approximately – 1.60/watt currently to about 1.00/watt – and solar electricity will then move a step further in the direction grid parity,” pointed out Wicht on the one hand. And on the other hand, he explained: “The margins of the module and cell manufacturers are already very small in any case, they hardly have any latitude to lower prices any further at short notice.” The consequence of this is relentless competition, in which additional smaller manufacturers in Germany and Europe that already manufacture more expensively than Asian competitors could come up with the short end of the stick. Whether or not it is also possible to advance quality assurance with this cost pressure remained one of the big unanswered questions at the well-attended mc-Si Forum.

Increasing cooperation

With intensified cooperation, seventeen European cell and module manufacturers and wafer suppliers organized as the SEMI PV Group aim to strike a balance between cost reduction and quality assurance and to counter the competition out of Asia. To this end they presented an International Technology Roadmap for Photovoltaics (ITRPV) in Berlin. Its forerunner was the PV Roadmap for Crystalline Silicon, was presented at last year’s Fab Managers Forum (see pv magazine 04/2010).
More economical use of materials such as silver and polysilicon are regarded as a key instrument in order to save on costs.
While a strongly reduced share of silver for cell metallization was unanimously welcomed at the meeting as being both feasible and worthwhile, there was spirited debate about the suggested use of thinner wafers in the ITRPV Roadmap: processing wafers with a thickness of less than 180 or even 150 micrometers (µm) is more challenging because the rate of breakage normally increases and efficiency declines. This requires new handling and cell concepts.

The International technology roadmap for photovoltaics

The International Technology Roadmap for Photovoltaics (ITRPV) strives to reduce module manufacturing costs on a per piece basis by eight to nine percent, or by nine to ten percent on a per watt-peak basis annually by the year 2020.
The following measures and assumptions form part of the ITRPV Roadmap.
Materials crystallization and wafering
The price of silicon and consumables for crystallization and wafering needs to be reduced by fifty percent by the 2020.
Materials cell processing
In order to reduce the cost of materials the thickness of current wafers with an average of 180 micrometers (µm) is on the one hand to be reduced to 140 µm in 2015 and to 100 µm in the year 2020.
On the other hand consumption of expensive silver pastes in the metallization process will be significantly reduced. The remaining portion of silver per cell is to be reduced from the present 0.30 g/cell to clearly less than 0.05 g/cell by the year 2020. As of 2015 silver should be replaced by Cu on a large scale basis.
Materials module
It is assumed that the reflectivity at the front side of the glass-air interface will be reduced from typically four percent to less than two percent by introducing anti-reflection glass into the mainstream as of 2013 onwards. Moreover, it is necessary to halve the absorbance in glass and the encapsulant from the current one percent by the year 2020, according to the Roadmap.
Processes manufacturing
It is possible to increase the throughput of the crystallization process by changing the common formats of the ingots, according to the Roadmap. In order to do so the ingot mass of multicrystalline silicon (mc-Si) must be increased from the current 400 kilograms to 1,100 kilograms, and from 150 kilograms to 300 kilograms for monocrystalline silicon (mono-Si).
Correspondingly, the throughput per tool for ingot growth is to be increased by twenty percent (mono-Si) and/or forty-five percent (mc-Si) by the year 2020. In the areas of wafer sawing and cleaning a fifty-percent increase is anticipated.
It is assumed that by the year 2020 mechanical yield losses during cell processing is expected to decrease from two percent to below one percent.
Another important factor consists in improvement of the tool uptime of the chemical processes from 96 to 97 percent (2020), as well as the thermal processes from 95 to 96.5 percent and the metallization and classification processes from 91.5 to 96 percent.
The throughput of production tools is to be at least doubled in order to reduce the unit costs per cell: for front end processes from 3,600 to 7,200 wafers per hour and from 3,000 to 7,200 wafers per hour in the case of single line back end processes.
Correspondingly the aim will be to reduce the relative number of operators by forty percent by the year 2020.
As a result of the improvements in manufacturing technologies and equipment it is expected that the investment costs for new cell factories per megawatt-peak will be reduced by 45 by the year 2020.
The expected yield trend in module manufacturing (ratio of good cells in good modules out to good cells in) is rising from 98.75 percent to 99.5 percent by the year 2020 despite ongoing reductions in wafer thickness.
In order to achieve the required cost reductions, module manufacturing equipment should occupy less floor space and have a higher throughput, according to the Roadmap. Thus it expected that floor space per megawatt-peak will be reduced by more than forty percent and throughput virtually doubled by 2020.
Process technology
In addition to production parameters, the efficiency is also expected to improve. Thus the Total Thickness Variation (TTV) is to be more than halved from the current average of 27.5 µm by the year 2020. The kerf loss is expected to be reduced from 155 to 100 µm.
In order to minimize recombination losses the Roadmap requires that recombination currents have to be greatly reduced in the next ten years. Correspondingly, the emitter sheet resistance is to be virtually doubled from approximately 65 ohm/square to 120 ohm/square (for n-type emitters).
The finger width needs to be reduced from 100 µm to 45 µm without a significant reduction in conductivity, says the Roadmap.
Furthermore, there should be an improvement in alignment precision. Approximately 30 to 50 µm can be achieved with current screen printing technology. Alignment precision needs to increase to ten µm by the year 2015, according to the Roadmap.
In order to get as much power out of the assembled cell as possible, the module-to-cell power ratio has to be improved according to the Roadmap, i.e. the mono-Si power ratio from 96 percent to 98.5 percent, and the mc-Si power ration from 97.5 to 100 percent by 2020.
The share of n-type mono wafers is expected to increase from the current eight percent up to fifty percent by 2020, and the share of n-type multi-wafers to just under ten percent.
The stabilized efficiency for mc-Si solar cells in mass production will increase from 16.5 percent to 18.5 percent, for mono-Si cells from 18 to 20 percent.
Correspondingly it is expected that the output power of mono-Si modules will increase from 250 to almost 290 watt-peak (Wp), and the mc-Si module power from 235 to more than 270 Wp.
Finally it is assumed that the share of rear contact cells as fraction of worldwide production will increase from five percent today to 40 percent by 2020.

Moreover, the most important global module manufacturers, particularly in China, are already fully integrated in general. For them the wafer is not the expensive cost factor as in the case of other cell and module manufacturers who have to purchase them.

The global perspective

“The fully integrated top players are hardly interested in reducing the wafer thickness with all of the associated risks and disadvantages if it does not pay off for them at all,” explained Centrotherm management Board Member Fath. Speaking with pv magazine he also criticized the fact that the ITPVR Roadmap, which was largely developed by German cell and module manufacturers, does not take enough account of international developments, particularly in Asia.
“Since our main customers are in Asia, in particular China, a road map that is oriented toward German requirements and perspectives does not help us very much,” stressed Fath who at the same time is a member of the governing board of the German Engineering Federation (VDMA).
For example, according to Fath, the submitted road map does not take into account the fact that leading Chinese suppliers increasingly offer mixed mc-Si and mono-Si ingots that can be used for manufacturing wafers for highly efficient and standard cells.
In addition Fath makes reference to several contradictions in the ITPVR Roadmap between the proposed increase in efficiency and the required cost reduction. For example, with intensified introduction of anti-reflection glass and reduced adsorption of glass, the increased throughput per tool for wafer sawing and wafer cleaning with simultaneous reduction of the wafer thicknesses or the expected yield trend in module manufacturing with the use of thinner wafers.

Too conservative

On the other hand, the ITPVR is too conservative in several points in Fath’s opinion; thus, for example, the state of the art in the case of the tool uptime of thermal processes when it comes to cell production is already above the initial roadmap value of ninety-five percent and the relative number of operators in a cell production line and the floor space per megawatt-peak in the case of module production could be reduced even more than is specified in the roadmap.
Altogether Fath regards the planned cost reduction targets of the roadmap from eight to nine percent annually for module production on a per piece basis and from nine to ten percent annually on a per watt-peak basis as insufficient. “This may be fine for Asian manufacturers, but for European and German manufacturers it is not aggressive enough,” stressed the Centrotherm CTO.
It will be interesting to see how the roadmap continues to develop and what further strategies and measures will be taken by German and European module and cell manufacturers in order to be able to keep pace in the global playoff.

This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: