In Toledo Solar the US has a Cadmium Telluride (CdTe) player to begin the push into rooftop installations, which could be important in the current market situation – with another 18 months of polysilicon shortages to go, and then likely more tariffs and sanctions holding back the Chinese solar industry
As we look at the global landscape for CdTe the first name on everyone’s lips is First Solar. Its business has grown steadily and now it has a factory planned in India, the first outside of the US and Indochina. But one thing has remained constant – its solar panels are exclusively made for and sold to utility-scale projects, and it dominates the US large-scale CdTe sector.
Talking to Aaron Bates, the CEO of Toledo Solar he thinks of his firm as the “little brother” of First Solar, a CdTe manufacturer founded in 2016 whose product is instead made purely for rooftop projects. There are other Cadmium-Telluride manufacturers in both the West and in China, but they are all very small or still getting themselves set up – Toledo therefore claims it has the number 2 spot, with a production capacity of 100 MW, which it expects to double every few years, reaching 800 MW in 2028 or so. It activated its first factory in 2020.
Toledo’s operation is located in Perrysburg, Ohio, less than 10 minutes from First Solar’s first and biggest factory. Perrysburg has the most solar manufacturing in the world outside of China, and it should be noted that Cadmium-Telluride is very much an American technology and even a product specifically of Perrysburg R&D. Despite closeness and similarity the two companies don’t compete. “We work well with First Solar – for 20+ years, First Solar has only sold its panels to utility-scale projects”, Bates explains. “Their business model has been very focused, in a Ford Motor way – one car, one color.”
Bates asserts that that there’s only two ways to cost-effectively manufacture CdTe semiconductors and that’s thin film, with First Solar and Toledo Solar both using vapor transport deposition with processing times as low as 15 seconds: the two companies have a technological overlap, but also each have their own IP as well, “The only two patents which allow for large-scale manufacturing of CdTe are theirs and ours, if you want to do it larger than in a laboratory.”
And that’s why Toledo Solar was founded – it had the technology, and an entire market segment, rooftop, was being ignored by First Solar, which is doing just fine for itself in the utility sector. Most of Toledo’s engineers are from First Solar and the two participate in consortia together. But Aaron Bates himself has a past career in Mergers & Acquisitions, not in solar. He met the owner of a shut-down solar factory in the Obama era, and saw the potential to revive a company on his own behalf, for a change.
It almost seems strange that Toledo isn’t simply an arm of First Solar – but there are some other things the “little brother” has going for itself. It is developing a transparent CdTe product, naturally very promising for BIPV, and it has its own manufacturing process – basically the same as First Solar’s but “horizontal instead of vertical.” But perhaps the single biggest distinction is module size – First Solar’s Series 6 product is 600mm by 1200mm, so 2 feet by 4 feet.
Since 2016 First Solar has rebuilt its manufacturing facilities, especially glass, to operate with larger, more cost-efficient units – which is great for utility projects, but to serve rooftop they would need dedicated supply lines producing smaller units. Since CdTe is thin-film, the problem isn’t weight, and it isn’t purely physical size either – in fact it’s the high voltage of these panels which violates fire codes. So First Solar could go into rooftop if it really wanted to – but why bother when it is still far away from saturating the utility-scale segment? Its next product, Series 7, will be larger yet.
Bates made one thing very clear – CdTe is underrated when judged by its nameplate efficiency ratings, it has a far broader bandgap, or usable spectrum, than silicon photovoltaics, which leads to it performing less well in test conditions, but better in suboptimal real-world conditions; and it handles diffuse light better. That weaker nameplate efficiency claim is one major reason why CdTe has been limited to utility-scale projects – developers have the expertise to know better and judge by kWh not kW, homeowners do not, though their knowledge should improve as the market grows.
Bates points out that in the “Wild West” days when solar first took off in the US, the disparity between claims and real performance lead to legal actions against several manufacturers – to the benefit of First Solar. “Silicon PV has a lot of limitations – even leaving aside the silicon price, China’s subsidies supporting solar manufacturing, and the Uyghur issue, there’s this 900 nm to 1100 nm bandgap – whereas CdTe aligns almost perfectly with our sun.”
And then there’s yet another backend advantage for CdTe – it doesn’t get as hot as silicon PV and it doesn’t lose as much efficiency for each degree of temperature increase. In fact, Bates points out that CdTe even benefits from “warming up” – not in terms of temperature, but due to some little-understood process related to atomic physics which increases its electrical output. Ongoing research into the phenomenon indicates that this is due to CdTe being a natural diode (unlike silicon), and that its diode property is reinforced by the passage of electrons.
At present, Toledo Solar’s average salary is six figures – this is not a manpower-intensive manufacturing process. So total wage costs are very low – and the single biggest manufacturing cost is actually glass, it’s just two sheets of glass with a thin-film semiconductor between. Asked about the Capex of a CdTe factory, Bates spitballs a figure of “maybe $200 million to $250 million for a 1 GW facility – cheaper than silicon PV. This is relevant as is but will enable a big acceleration if the Biden Administration manages to get solar manufacturing subsidies past Senator Manchin before the 2022 midterms.