IN DEPTH: First Solar tech makes great leap forward

Throughout its 25-year history, First Solar, the largest US PV manufacturer, has always had plenty of things going for it. The efficiency of its thin-film modules, however, was never one of them. How times have changed.

The company’s cadmium-telluride (CdTe) modules were always less efficient at converting sunlight to electricity than crystalline silicon, the dominant PV technology, while at the same time being less expensive.

But fuelled by a massive R&D budget, First Solar’s panels have notched remarkable efficiency gains in recent years — catching up to, and even surpassing, many silicon-based rivals, and setting numerous cell-efficiency world records on the way.

Last summer, a First Solar module clocked an 18.6% efficiency in the lab, pushing CdTe for the first time past polycrystalline silicon — the technology it normally competes with in the utility-scale market — and within swinging distance of higher-efficiency monocrystalline.

Outside of the lab, its latest modules boast an average efficiency of 15.8%, higher than most rival polysilicon modules, including those made by SolarWorld and Suntech.

The Arizona-based OEM, which is also a formidable project developer, is now in an enviable position as it enters 2016. Its core market, the US, has just received a game-changing boost in the form of a five-year investment tax credit extension, and the company has carved out a leading position in many emerging solar hot spots, including India, the Middle East and Latin America.

Its balance sheet is the strongest in the industry, according to Goldman Sachs, and unlike many high-profile US-based PV companies, First Solar is profitable. After a 48% share-price surge in 2015, it is once again the world’s most valuable public solar company, having snatched back the title from SunEdison.

“No developer has more offensive firepower than First Solar heading into the next several years,” writes Brian Lee, senior clean-energy analyst at Goldman Sachs, in a recent research note.

Huge R&D investment

The efficiency gains of First Solar’s modules have not come cheaply.

Between 2010 and 2014, the company’s R&D team, based in Perrysburg, Ohio, spent nearly $650m — an unheard-of sum in the solar industry. “Compared to the competition, there’s no-one even close,” Nicholas Strevel, First Solar’s senior manager for module technology, tells Recharge.

In 2014 alone, First Solar spent $144m on R&D. In comparison, Trina Solar, the global market leader and record-holder for polycrystalline efficiency, spent $22.3m. Even SunPower, the second-largest US-based module maker, and a company prized for its highly efficient crystalline technology, spent just half of First Solar’s outlay.

In addition to its R&D investments, First Solar’s modules have benefited from the company’s absorption of GE’s CdTe intellectual property after acquiring GE-owned PrimeStar in 2013.

Against “number-one competitor” polycrystalline, “we’ve caught up and we’re pulling away”, Strevel says.

A decade ago, First Solar’s modules had an average efficiency of around 9.5%. Today its production lines turn out modules close to 16%, and its researchers believe that over time they can push that figure well above 20%, while continuing to keep production costs in check.

“To be blunt, many of our competitors are just manufacturers of existing technologies,” says Strevel.

“They buy their [manufacturing] equipment from third parties, they buy their feedstock from third parties, and they just put it all together and manufacture in the lowest-cost place they can, using the lowest-cost methods possible.

“That’s not what we’re doing. We’re a technology company.”

To the layman, an extra percentage point or two might not sound like much. But an increase in efficiency from 10% to 11% means the same panel can generate 10% more power, no small feat in the fiercely competitive world of electricity production. Over time, small efficiency gains make a big difference to the economics of solar energy.

More-efficient panels allow developers to build projects on less land using fewer mounting structures, while maintaining the same output. Higher efficiencies also reduce the amount of components that go into a given capacity of modules — such as glass, semiconductor materials, copper, wires, and the polymers that hold them together.

“Those same raw materials are going in whether you’re making a 12%-efficient module or a 20%-efficient one,” Strevel notes.

With CdTe now competing on efficiency alone, thin-film’s inherent advantages — long acknowledged by the industry — become more important.

Compared to the laborious, energy-intensive and multi-step process of turning a lump of silicon into a PV module, thin-film manufacturing is far simpler and quicker, involving the deposition of a very thin layer of semiconductor directly onto glass.

“If our [silicon] competitors could use glass as their substrate, they would,” Strevel says. “It gives us a dramatic cost advantage.”

Thin-film modules also have a more uniform and attractive appearance. “We’re not confined to a six-inch wafer that really you can only arrange in one or two different ways to make a useful module,” he says.

And ultimately, assuming the R&D budget is there, CdTe has more room to run, with significantly higher theoretical efficiency limits than standard silicon.

Business acumen

First Solar’s steady efficiency gains would not have been possible if the company had spent the past half-decade in desperate financial shape — like many of its rivals. But the company has cleverly navigated the market’s violent fluctuations. 

“Having watched them for years, what’s really impressed me most were their business decisions,” says Paula Mints, chief analyst at SPV Market Research. “In general they’ve made really good decisions that have kept them competitive. That’s given them a cushion so they could continue their R&D.”

In its early years, First Solar benefited from the financial backing of John Walton, heir to the Walmart fortune. The company went public in 2006 and for a period had a market capitalisation larger than that of General Motors.

During the financial crisis it profited enormously from the silicon shortage that sideswiped the global PV industry. With the price of polysilicon surging past $400/kg for a period (compared to less than $20/kg today), First Solar’s thin-film modules were able to gain share in the feed-in tariff-fuelled European market.

But the company’s strategic masterstroke was a series of acquisitions from 2007 to 2010 that gave it a multi-gigawatt project pipeline and a sophisticated in-house EPC team.

Whether born of foresight or simple necessity, the move has paid off tremendously. First Solar transitioned rapidly from a company focused on selling modules to one that instead built huge projects using its own equipment and then sold those projects to customers like NextEra Energy and GE Energy Financial Services. Nearly all of its major module rivals have since attempted to follow it downstream.

This perfectly timed push into the development business allowed it to act as its own largest customer during the module industry’s catastrophic downturn of 2011-13, when dozens of manufacturers went belly up due to a glutted module market.

As a result, even during the leanest years First Solar consistently maintained healthier finances than nearly all of its peers, while building some of the world’s largest solar projects — including the 550MW Desert Sunlight in California’s Mojave Desert.

Since going public a decade ago, First Solar has reported just two full-year losses, both relatively modest, in 2011 and 2012. It returned to profitability powerfully in 2013 and has not looked back since.

First Solar’s downstream prowess also feeds directly back into the company’s module technology.

If there’s a problem with its modules in the field, First Solar is usually the first to know, and it can quietly and quickly swap them out. Meanwhile, the lessons its EPC team learns by building solar farms around the world translate into better module designs.

That feedback loop will only intensify as First Solar retains long-term ownership of more of the plants it builds through its yieldco, 8point3 Energy Partners, launched with SunPower in 2015.

8point3 is unique in the yieldco universe in that the power-generation equipment used in its projects is made by its parent companies, taking vertical integration to the next level within the solar industry.

As Lou Trippel, director of First Solar’s line of module products, puts it: “We’re eating our own cooking.”


Ironically, for all of First Solar’s professed confidence in CdTe, the company has always had a wandering eye when it comes to competing technologies — and it makes no promises about the future.

“We’re thought of as a CdTe company, but fundamentally we’re technology-agnostic,” Trippel says.

In the past, First Solar has invested in CIGS (copper indium gallium selenide) research, the technology used by Japan’s Solar Frontier, its closest thin-film rival. And in 2013, it acquired TetraSun, a high-efficiency silicon start-up, ostensibly so it could break into the one booming market where it has virtually no presence today — rooftops. High-efficiency monocrystalline panels have a big advantage here over polycrystalline and thin-film, given the inherent space constraints.

In spite of its tendency to play the field, however, First Solar has never had more reason to be optimistic about CdTe, Trippel says.

TetraSun’s silicon technology will play a “complementary” role in First Solar’s future product line-up, Trippel insists. But the reality is that the acquisition has contributed little to the company’s bottom line. In First Solar’s most recent quarterly earnings presentation, TetraSun was not mentioned.

Meanwhile, in late 2014, First Solar became the largest investor in Clean Energy Collective, a Colorado-based developer of community solar projects. Many analysts believe community solar projects — typically ground-mounted arrays built with the backing of local utilities — will eventually pose a threat to the rooftop PV market.

Needless to say, such community projects are ideally suited for First Solar’s CdTe modules.

“There have been moments over the last decade where we’ve looked very hard at whether CdTe is the long-term answer for First Solar,” Trippel says. “We’ve had to prove to ourselves that it’s the platform that will allow us to continue to scale.

“If there were a better technology base out there for the coming decade, well, we have folks who look at that all the time.

“We retain a healthy paranoia,” he says. “But we don’t see any immediate threats.”