First Solar

Daniel Riley
December 6, 2010

Submitted as coursework for Physics 240, Stanford University, Fall 2010

First Solar, Inc. is the world's largest photovoltaic module manufacturer in the world, with a production of 1.1 gigawatts worth of solar panels (from now on I will refer to amount of solar panels simply in terms of watts as is customary in the literature) in 2009 and 4,700 employees as of February 12, 2010. Their mission is to produce non-subsidized electricity from photovoltaics at equal to the cost of commonly used non-renewable sources such as coal or natural gas. They specialize in producing cadmium telluride (CdTe) thin-film photovoltaic modules and constructing large-scale commercial installations. [1] CdTe modules, while not as efficient as crystalline and polycrystalline silicon solar cells, are the cheapest modules to produce. [2] Their corporate headquarters are based in Tempe, Arizona while their production plants are built in Perrysburg, Ohio; Frankfurt, Germany; and Kulim, Malaysia. [1,3] First Solar has experienced significant growth in the past decade as demonstrated by the fact that their first production facility, which was built in Perrysburg, Ohio, became fully operation in only 2004 and produced 100 MW/yr, and they are now producing 1.1 gigawatts annually. [1,3] I will discuss the history of First Solar and the factors that have enabled them to attain such monumental growth. I will also explain the advantages and disadvantages of CdTe technology. Finally I will explore what obstacles First Solar faces in the future and where they appear to be heading.

First Solar's story begins in 1992 in "The Glass City," Toledo, Ohio, which was filled with people who had expertise in the glass industry, which translated well to solar panel manufacturing. They credit John Walton, son of Walmart founder Sam Walton, who invested $150 million in the company, with being one of the primary reasons the company was able to take off. With such a patient investor the company was able to incubate long enough to develop the necessary manufacturing process. Silicon Valley venture capitalists generally expect a return-on-investment within 5 years, whereas John Walton stuck with them for 12 years before the first manufacturing line even became operational. In 1999 True North Partners, LLC purchased a controlling share of the company and reformed the company into what it is today, First Solar. By 2004 they had built their first fully operational 100 MW production line. [3]

2004 was the year that First Solar began its explosive growth, however this wasn't a result of the U.S. market or legislation. Mike Ahearn, CEO of First Solar says "When I was discussing this story with a friend, he remarked, 'Only in America.' And it's true that our story at first glance does seem to have many of the hallmarks of the classic American dream. But in fact First Solar is to a large extent a German success story." First Solar was looking for a large market so they could produce high volumes to drive down the module price, but Washington wasn't interested in helping them. In 2004 Germany had begun the first solar feed-in program, where utilities were required to purchase electricity from customers producing solar power at a rate set by national law for 20 years. This created a large and stable market that First Solar decided to enter. Ahearn commented that it is difficult to enter the American market, because of unstable government incentive programs that start and stop every few years. Many other European countries adopted similar programs in Europe which drove tremendous growth and economies of scale, and as a result 86% of First Solar's net sales in 2009 were generated from customers headquartered in the European Union. [1,3] Concurrently the price of silicon, which was being used in 92% of panels, tripled from 2003 to 2005. [1] Despite the fact that First Solar, a company which had begun in Toledo, Ohio, was progressing towards becoming the largest PV manufacturer in the world, Sen. George Voinovich of Ohio voted against a national renewable energy portfolio standard and against a 30% investment tax credit for building solar energy. [3] There are still hopeful signs of pro-renewable legislation in the U.S. in states such as California, which has created a significant renewable energy portfolio standard.

With First Solar's support thin-film CdTe is one of the most widespread photovoltaic technologies. As of 2009 they are able to produce a CdTe module at a cost of $0.85/watt, whereas the cost of thin film silicon is about $2-3/watt. [4] While CdTe cells are generally less efficient than CIGS or Si solar cells, they are much cheaper to produce. They use 1/100 of the thickness and 1/10 of the manufacturing time required for crystalline Si cells. They are much more efficient than a-Si cells. CIGS cells provide comparable thickness films and slightly higher efficiencies, but manufacturing cells with quaternary compounds is difficult and expensive. Some immediate drawbacks to CdTe cells include careful handling of the material components due to toxicity issues (although CdTe as a compound is quite stable), and lower demonstrated efficiencies than Si or CIGS. [5] Due to the reduced efficiency, CdTe cells require more land area. The U.S. Energy Information Administration predicts that the world will consume 31.6 trillion kWh annually by 2030. [6] If we assume an average solar irradiation of 2000 kWh/m2/yr, a CdTe module efficiency of 10%, and a panel to grid efficiency of 80% then it would take 76,000 miles2 of CdTe panels to power the world. This is roughly the size of Nebraska and would require huge amounts of cadmium and tellurium. The biggest long term question for CdTe technology is the availability of the extremely rare element tellurium.

There is a significant debate occurring right now over whether tellurium supplies will keep up with demand as First Solar expands its operations. Estimates vary, but world production of tellurium, which is primarily obtained from electrorefining copper, is approximately 500 MT/yr. [2] First Solar is currently producing panels at a rate of 1.1 GW/yr which consumes 105 MT/yr, a significant fraction of the global tellurium supply. [1,2] In 2000 Björn Andersson argued CdTe production would be limited to 20 GW/yr as of 2020, however in a 2010 Science paper Ken Zweibel argued that previous estimates had not taken consideration just how thin or efficient CdTe cells may become or how much the tellurium supply may increase once prices increase. [2,7] Zweibel outlined scenarios in which CdTe cells were produced at a rate of ~65 GW/yr in 2020 assuming 1% growth in copper mining rates, module efficiencies that had increased to 15%, and film thickness reduced to 0.3 μm. [2] He didn't include the possibility of increased refining of already mined copper to produce more tellurium, even though Ojebuoboh has described how the tellurium supply could triple by electrorefining more copper if the price of Te increased. [8] Zweibel has also shown that CdTe solar cells could account for 5-20% of the world's power by 2030 based on differing scenarios of film thickness and overall photovoltaic market growth. [2] If we consider my estimate that powering the world in 2030 would take 76,000 miles2 of CdTe panels, and the fact that a square mile of panels requires 16.8 MT of Te, then 1.29 MMT will be required to fully power the world. [7] At current film thicknesses and recovery rates it would take over 2000 years to recover this much tellurium, which motivates the necessity for higher recovery rates and thinner films. While the future of tellurium in the production of solar panels is still very unclear it's definitely an issue that many people, including those at First Solar, are thinking about.

First Solar's future is all about continued aggressive growth by expanding into new markets and continuing to reduce costs. They have recently signed a memorandum of understanding to construct a 2 GW power plant in Orods City, China. They also expect to increase manufacturing capacity to 1.8 GW/year by 2012. While they've received the most media attention for being able to produce modules so cheaply they're also working on reducing costs of hardware installation and labor, which account for over half the cost of setting up a system. [1] Finally, the most interesting possibility for First Silicon's future is that they may be expanding their technological portfolio into other thin-film technologies. In April 2010 Reuters reported that First Solar's secretive Silicon Valley lab was looking into CIGS technology. [9] While these reports are unconfirmed by First Solar, it is intriguing nonetheless to think what advances they could make if they applied their manufacturing expertise to CIGS technology.

© Daniel Riley. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.


[1] 2009 Annual Report, First Solar, Inc. (2009).

[2] K. Zweibel, "The Impact of Tellurium Supply on Cadmium Telluride Photovoltaics," Science 328, 699 (2010).

[3] T. Friedman, Hot, Flat, and Crowded: Why We Need a Green Revolution and How it can Renew America (Farrar, Strauss, and Giroux, 2008), pp. 386-391.

[4] K. Zweibel, "Multi-Gigawatt Thin Film PV," Proc. 34th IEEE Photovoltaic Specialists Conference (PVSC), 17 Feb 09.

[5] A. R. Jha, Solar Cell Technology and Applications (Auerbach Publications, 2010), pp. 146-149, 174-178.

[6] "International Energy Outlook 2010," U.S. Energy Information Administration, DOE/EIA-0484(2010), July 2010.

[7] B. A. Andersson, "Materials Availability for Large-Scale Thin-Film Photovoltaics," Progress in Photovoltaics Research and Applications 8, 61 (2000).

[8] F. Ojebuoboh, "Selenium and Tellurium from Copper Refinery Slimes and Their Changing Applications," World of Metallurgy - ERZMETALL 61, 33 (2008).

[9] P. Gupta, "Exclusive: First Solar Exploring New Panels in Silicon Valley," Reuters, 8 Apr 10.