Betting on the future: A look at Solar Frontier's Kunitomi CIS thin film PV manufacturing facility

by Christian Rolseund
August 24th, 2011

Mangoes and CIS production in Southern Japan

It is a strange place for one of the largest and most advanced solar photovoltaic (PV) manufacturing facilities in the world. A few kilometers outside a sleepy seaside town in Southern Japan, past rice fields and small farms, as the road rises into the hills, the massive white forms of a factory emerge between cedar-lined hills.

Aerial view of Solar Frontier's Kunitomi manufacturing facility. Image source: Solar Frontier K.K.
Aerial view of Solar Frontier's Kunitomi manufacturing facility. Image source: Solar Frontier K.K.

Likewise, the tourist brochures in the hotels of Miyazaki speak of mangoes and rocky cliffs, not high tech manufacturing. Yet inside those walls, every five seconds a PV module is produced, using the most sophisticated equipment and processes from three continents, and is then stacked in the cavernous spaces of an automated warehouse, ready for shipment.

The sheer scale of the interior space of the Kunitomi PV factory is breathtaking, with line upon line of machines stretching for tens of meters, which provide an annual production capacity of 900 MW.

The facility represents a big wager on the future of PV, and particularly of copper indium gallium diselenide (CIGS or CIS) technology. Only 426 MW of CIGS modules where sold in 2010, and at full capacity, the Kunitomi factory can produce double that quantity.

In speaking with plant officials, it is clear that Solar Frontier feels remarkably secure in taking this risk. 

 

Thin-films and CIS: A niche, or the future?

Despite being touted as the "next generation" of PV, thin-films - thin coatings of semiconductor material on a various substrates - have not yet entirely emerged from their status as a niche technology.

Thin films represented only 13% of the global PV market share in 2010, down from 19% in 2009, largely as a result of Chinese crystalline silicon PV producers out-pacing thin-film leader First Solar Inc. (Tempe, Arizona, U.S.) in 2010.

Crystalline silicon (c-Si) supporters will point their higher conversion efficiencies - from 16% to over 20% for the highest efficiency c-Si modules, versus 12% to over 13% for the most efficient mass-produced thin-film panels. However thin films offer a number of other advantages, including lighter weight and better performance in hot climates and low-light conditions.

Most significantly, as First Solar has shown, it is possible to produce thin films at a much lower cost per watt than crystalline silicon. Thin film production processes utilize fewer processes than crystalline silicon, and less semiconductor material. Solar Frontier also notes that its process uses less than half of the energy of crystalline silicon production.

 

CIS promises conversion efficiencies similar to crystalline silicon

CIS currently has the smallest market share of any thin-film, representing only 2% of the global market in 2010. However, CIS may also hold the biggest promise of any thin-film. The U.S. Department of Energy's National Renewable Energy Laboratories has achieved conversion efficiencies over 20% with CIS, higher than the world records for either cadmium telluride (CdTe) or amorphous silicon thin films.

And while it is a long way from the laboratory to the shop floor, a number of analysts, including PHOTON Consulting, have identified CIS as the technology with the greatest potential to achieve a technological breakthrough in the next few years, which could lead to a dramatic change in CIS' market share.

CIS has also outpaced CdTe efficiencies in mass production. On March 29th, 2011, Q-Cells subsidiary Solibro achieved a world record of 13.4% efficiency with mass-produced CIS modules. Solar Frontier is currently mass-producing 155 watt modules at 12.6%, roughly equivalent to First Solar CdTe modules, and has developed a 13.4% efficient "champion" module.

Another advantage of the technology is the "light-soaking effect". Unlike the light-induced degradation which afflicts amorphous silicon, CIS modules gain a 5-10% performance boost after initial exposure to sunlight.

Solar Frontier has begun mass production of its 150 watt SF 150 PV modules at the Kunitomi plant. Image source: Solar Frontier K.K.

A note on indium

Critics of CIS allege that the use of indium in the semiconductor layer inherently puts CIS production in a position of precarity. Indium is a rare earth element, and there is a limited supply of the material that naturally occurs in the earth's crust.

However, Solar Frontier officials are quick to dispel any notions that their production is at risk of being imminently shut down due to global political crises or other plausible interruptions of supply.

"There is more indium currently in the earth than silver," notes Solar Frontier Global Product Manager James Plastow. "China controls a lot, but it is spread around the earth as well."

 The company also notes that the greatest use of indium is in the LED industry, not the solar industry. While it would not disclose specific details, Solar Frontier notes that much of its indium comes from recycling operations.

The amount of indium used per module in Solar Frontier's 1.2 - 1.5 micrometer thick CIS layers is also very small, and the company notes that even with nearly 1 GW of total annual production, its current total usage represents a small portion of the total global supply.

 

Solar Frontier's parent company has been working with CIS since 1993

Solar Frontier is far from a newcomer to CIS technology. Parent company Show Shell Sekiyu has been investigating PV technologies since 1978. After a period of exploring amorphous silicon in the 1980's, the company began working with CIS in 1993, through research funded by Japan's New Energy and Industrial Technology Development Organization (NEDO).

Show Shell Sekiyu committed to CIS manufacturing in 2005, and the Kunitomi manufacturing plant is Solar Frontier's third CIS factory, with a capacity more than ten times greater than its 20 MW Miyazaki Plant 1 and 60 MW Miyazaki Plant 2 factories combined.

 

Size matters: Automation and scale

The Kunitomi Plant represents a massive investment, which Solar Frontier estimates at USD 1 billion. Solar Frontier finished the factory on an accelerated time line, and began commercially producing modules in February 2011, only 16 months after groundbreaking in September 2009. The plant officially opened in April 2011.

The plant is highly remarkable in both its scale and its very high degree of automation.

The majority of the plant's 700 employees work in product inspection and testing. Image source: Solar Server
The majority of the plant's 700 employees work in product inspection and testing. Image source: Solar Server

At 400,000 square meters, the plant is composed of a series of enormous, cavernous spaces for module production and storage. And yet the production floor is eerily quiet, except for the hum of machines.

All in all, the factory employs less than one worker for every MW of annual production capacity, mostly in product inspection and testing.

"Our facility requires 7-800 people," observes Solar Frontier Executive VP of Communications and Operations Brooks Herring. "A facility like this in China would require 7,000 - 8,000."

The company cites multiple advantages to this high degree of automation, including both lower production costs and a greater ability to control production quality.

Though Solar Frontier is careful to note that thin film production is not as dependent upon large scale to achieve low costs as crystalline silicon, the very large scale of the plant offers advantages to Solar Frontier's bottom line.

"There are a lot of costs, were if you can spread it over a GW, it's a lot better than spreading it over a MW," notes Solar Frontier Global Product Manager James Plastow.

 

Solar Frontier's CIS process

There are other companies visible that could raise the money to build similarly enormous factories, and choose not to. Solar Frontier's massive investment in the Kunitomi plant represents not only the deep pockets of parent company Show Shell Sekiyu, but also the tremendous faith that Solar Frontier has in its CIS process.

Solar Frontier uses a slightly different process than other manufacturers. The basic cell and module components are the same: glass, a molybdenum electrode, a CIS p-type absorber, a buffer layer, a transparent front contact and cover glass. However the details are different.

Solar Frontier uses proprietary selenization and sulfurization steps, based on a process developed by Siemens, to create its CIS layer, as opposed to the co-evaporation process used by some other CIS manufacturers. More unique is the company's Zn(S,OH)x buffer layer, as most companies use cadmium sulfide.

Scientists at Solar Frontier's Atsugi Research Center have achieved a 17.2% conversion efficiency with a 30 cm x 30 cm sub-module; Image source: Solar Frontier K.K.
Scientists at Solar Frontier's Atsugi Research Center have achieved a 17.2% conversion efficiency with a 30 cm x 30 cm sub-module; Image source: Solar Frontier K.K.

Using its proprietary processes, Solar Frontier scientists at its Atsugi Research Center in Tokyo achieved an efficiency of 17.2% on a 30 cm x 30 cm submodule in 2011.

The company has set a roadmap to transfer the technical achievements that achieved that efficiency to its production lines.

Solar Frontier is careful not to let on too much about how its processes work, noting that these two are the core of what differentiates its CIS product.

"We have ten steps," states Solar Frontier VP Brooks Herring. "Six of these are really easy. The other three, three and a half are really hard and nobody else can do them."

 

Details of selenization and sulfurization

Solar Frontier's selenization and sulfurization processes begins with the application of a copper, indium and gallium precursor via sputtering, followed by annealing processes using H2Se and H2S gas in two separate steps. Solar Frontier describes the second sulfurization step as "a kind of advanced selenization".

"We make another, wider bandgap layer on top of copper indium diselenide layer, intentionally making a hetero-junction type banding," explains Solar Frontier Chief Technology Officer Satoru Kuriyagawa "That is very important for stable performance, that is the reason that we are using sulfurization after selenization process."

Solar Frontier says that this process allows for uniform, high speed-production. The company states that its selenization process allows it to use 1/40th the selenium material used in co-evaporation. Another advantage is the ability to recycle excess copper, gallium and indium used, due to the lack of contamination in the sputtering process.

While technical papers show possible reaction pathways for the two processes starting at 200 degrees and reaching up to 500 degrees centigrade, Solar Frontier does not reveal the temperatures at which it applies H2Se and H2S. However, Solar Frontier CTO Satoru Kuriyagawa states that high-speed production is necessary for low cost processes, and the higher the temperature, the longer the cooling time before the next step of the process.

 

Zinc hydroxide buffer layer

A second innovation of Solar Frontier is its use of a Zn(S,OH)x buffer layer, as opposed to the cadmium sulfide layers used by other CIS manufacturers, including Würth Solar, Soltecture, Global Solar, and AVANCIS. This layer is applied using a MOCVD process, which the company says provides better crystalline growth than sputtering.

Solar Frontier states that the zinc hydroxide layer allows for more absorption of shorter wavelength light, which means greater energy production in the afternoon and evening. The company also notes that this allows it to avoid the use of cadmium, which is moderately toxic.

 

CIS-markets: A good product, but at a bad time?

However, having superior technology and processes does not alone guarantee market success.. While Solar Frontier has undergone a lengthy process to arrive at full commercial operation in the Kunitomi Plant, the timing of launching such large production can be seen as less than optimal.

A significant portion of the company's sales have traditionally been in the Japanese market, particularly its sales for residential installations which totaled 23.6 MW in November 2010. The Fukushima Disaster has strongly affected the nation's view of nuclear power, and Japan is on the verge of passing a new feed-in tariff. However, the nation still struggles with very high balance of system costs for PV, and its political establishment remains strongly tied to the country's nuclear power industry.

Solar Frontier PV modules on the roof of a traditional Japanese house. The majority of the company's residential PV sales have been in the Japanese market.Image source: Solar Frontier K.K.
Solar Frontier PV modules on the roof of a traditional Japanese house. The majority of the company's residential PV sales have been in the Japanese market.Image source: Solar Frontier K.K.

To sell the product that Solar Frontier can produce at the Kunitomi plant will require that the company achieves dramatic sales successes in multiple major PV markets. The company has a significant installed capacity in the utility and commercial European PV markets, at 55.5 MW as of November 2010, but as of yet a smaller presence in North American markets.

The global solar industry has seen a very difficult first two quarters of 2011. Many companies have reported falling revenues and nearly every PV manufacturer has reported dramatically reduced margins as a result of collapsed PV prices in both the first and second quarters of the year.

Additionally, a number of industry analysts are predicting a reduced global PV market in 2012 and 2013, as other national markets mature to replace anticipated reductions in the German and Italian markets.

In this environment, Solar Frontier will have to work twice as hard to sell the capacities that it can produce, which would involve more than doubling the current market share of CIS.

 

Partnerships with EPC giants; support by Showa Shell Sekiyu

However, Solar Frontier has a long range view of the global PV industry, and just might be able to sell out its product, even under difficult circumstances. Besides its strong value arguments for its PV modules, the company has secured enviable partnerships with EPC giants Belectric and General Electric.

In fact, Solar Frontier President Shigeaki Kameda states that unlike most PV manufacturers, the company currently is not in a position of oversupply, and that its biggest challenge is to increase capacity quickly to meet demand.

Furthermore, the support of Show Shell Sekiyu mean that the company could weather a few bad quarters without collapsing or restructuring, as some companies have done in recent months.

The next few quarters will bear out if Solar Frontier has chosen wisely with its enormous investment in the Kunitomi plant, but the company seems to feel that the risk of inaction is greater.

"It took us 5-10 years to go from a small scale into commercialization," notes Kameda. "If we are too slow, we will be left behind in the race to develop technologies."

Solar Frontier President Shigeaki Kameda in the Kunitomi manufacturing facility. Image Source: Solar Frontier K.K.
Solar Frontier President Shigeaki Kameda in the Kunitomi manufacturing facility. Image Source: Solar Frontier K.K.