By the time the Energy Department chose Solyndra for a half-billion-dollar loan in September 2009, a fateful market trend had already been building.
The price of polysilicon, the central ingredient in making the most prevalent solar panels, had skyrocketed for several years, into 2008. That enabled Solyndra's technology, which didn't use polysilicon, to compete with silicon-based solar modules, such as those made in China and Europe.
Solyndra had another cost advantage: Its panels were cheaper to install than other types of panels. But the price of polysilicon began a steep descent in February 2008, slumped again through the presidential election, and flattened out around May 2009 far below its peak, according to Bloomberg New Energy Finance. It fell from $475 per kilogram in February 2008 to $73 per kilogram in May 2009, the Bloomberg data show.
Polysilicon makes up about a quarter of the cost of a silicon-based solar module, according to Anthony Kim, a solar analyst at BNEF.
So the price drop meant that before Solyndra had received a single DOE loan guarantee dollar, a major part of its cost proposition had been undermined, if not erased.
The data add to questions about what the Department of Energy knew about Solyndra and when. Republicans on Capitol Hill, and some Democrats, are asking whether DOE performed adequate due diligence on the company, whether it rushed the loan application, and how it handled the loan once Solyndra began to struggle.
Another question is whether DOE was ready for the silicon price collapse, considering the data available at the time.
"When you're looking to support the manufacturing of an unprecedented technology and product in a volatile market, you need a pretty paranoid perspective: What are the ways the company could fail?" said Shyam Mehta, senior solar analyst at GTM Research.
In the middle of last decade, Mehta said, Germany and Spain began offering aggressive subsidies for solar power. The sudden demand led to groundbreakings for new polysilicon plants. Of course, these would take three to five years to build. So from about 2003 to 2008, he said, solar-grade silicon was in heavy undersupply -- and its price rocketed skyward.
The spike was artificial, Mehta said, because the new plants would eventually come online, bringing supply in line with demand and reducing the silicon price. But that prospect didn't stop new market entrants that could make a solar panel with little or no silicon. Solyndra was one of them.
Well-positioned for the 2009 market
Yesterday on Capitol Hill, a DOE official explained Solyndra's demise in terms of stiff Chinese competition and market conditions that neither DOE nor the private sector could have predicted.
Jonathan Silver, executive director of DOE's Loan Programs Office, said Solyndra's product seemed well-positioned in 2009. First, it could avoid polysilicon costs. Second, its panels had a unique tubular design -- rather than the conventional, flat panel -- that could be installed more cheaply.
"But polysilicon prices subsequently dropped significantly, taking Solyndra, and many industry analysts, by surprise," Silver said in testimony to a subcommittee of the House Energy and Commerce Committee.
In his written testimony, Silver cites data from Bloomberg New Energy Finance.
Silver also said DOE had commissioned a comprehensive review of Solyndra's technology, and the solar market, before approving the loan. The firm doing the review was R.W. Beck, a division of SAIC Inc.; a spokesman for SAIC said it is up to DOE whether to release the report.
Mehta said Solyndra's woes began in 2008, when Spain announced sharp reductions in its solar subsidies, slowing the world market. Then the financial crisis erupted, closing up the credit that was so crucial to building solar power.
By late 2009, when the Obama administration was finalizing its deal with Solyndra, the price of solar panels had dropped by half -- and the price of polysilicon dove along with it.
"That completely changed the outlook for Solyndra. The revenue model from the 2008 view of the world no longer applied," he said.
Losing a sprint against China
No one expected the polysilicon price to dive so quickly, Mehta admitted, but the crash itself "was always on the horizon. It was only a question of when, not if. You could make that claim."
When the crash came, Silver said, Solyndra began cutting costs. It redoubled its sales and marketing efforts, with some success: Its 2010 revenue was 40 percent higher than in 2009.
But the sprint to survive left it with little cash, and by summer 2010, bankruptcy loomed.
Solyndra asked DOE for a larger loan and was declined. Then Solyndra sought a $75 million emergency loan from its investors, and secured it in February 2011.
By this time, the Chinese solar giants were rolling.
Flush with $30 billion in loans from Beijing in 2010, these companies were producing solar panels by the gigawatt -- not the 250 or so megawatts Solyndra aspired to. They weren't dabbling in specialized technology but were making conventional, silicon-based solar panels.
"They're fighting over pennies in cost, tenths of a percentage in efficiency, these guys," said Matthew Feinstein, a Shanghai-based analyst with Lux Research. "They're the ones that are making it tough for high-cost American and European players to survive in this kind of a market."
In 2010, Germany and Italy had announced cuts or limits in their solar subsidies. Now, new Chinese capacity had joined the market, causing oversupply. The economy wasn't picking up.
All these factors caused the price of polysilicon to crash again in 2011.
"That was the straw that broke Solyndra's back," Mehta said. "When the second wave of price collapse hit the market, they realized there was no way to lower their costs in such a way they could sell their systems at a profit."