First Solar CEO Mark Widmar has warned that the US is at risk of “adding Ohio, Texas, Arizona and other states to the locations that host China’s downstream solar manufacturing.” We agree. Widmar also advocated that 45X advanced manufacturing tax credits be restricted so as to not apply to Chinese manufacturers.
Widmar stated that China has a strategy of maintaining upstream supply chain control while also seeking to control downstream manufacturing outside its borders. This may be a deliberate strategy, but it is also an organic and predictable outcome, since reshoring efforts were always going to find wafer-polysilicon hardest, as this is where China’s energy cost advantage is most salient. It’s also not surprising that Indian and US policies incentivizing manufacturing expansion should attract the interest of Chinese companies since those are the overwhelming majority of the global industry.
India’s latest manufacturing expansion has mostly used Mono PERC, which is slightly obsolete compared to the N-type expansion underway in China – meaning that Chinese companies have a technical edge they can use to expand within India itself (as First Solar has done), going beyond merely supplying production line equipment to Indian companies as they have done so far. In the US, the latest related announcement is Jinchen opening a subsidiary in Texas last week, with bigger developments already announced a year ago – such as 5 GW module production capacity from LONGi in Texas working with Invenergy, Jinko already manufacturing modules in Florida since 2018, a 5 GW cell factory announced by Canadian Solar in Indiana in October (Canadian Solar’s manufacturing arm is Chinese), a $200 million investment from Trina Solar in Texas announced in September, and so on and on.
Our recent coverage at Rethink Energy has touched on the continued strength of China’s grip on the solar wafer segment (and thus also polysilicon). Last week we came to the conclusion that Oman’s under-construction polysilicon factory is likely Chinese-owned and will initially be making only metallurgical-grade with the polysilicon dependent on future market conditions, and the previous month saw CubicPV cancelling its 10 GW wafer factory in the US, while Nexwafe has yet to announce one and Indian progress remains small.
Our conclusion from those recent events is that US (and perhaps also Indian) reshoring efforts have failed to penetrate upstream to wafer-polysilicon, effectively making it only a cell-module reshoring agenda. The EU has never gotten on board with a thorough reshoring strategy of its solar manufacturing efforts, preferring to wallow in a gigantic stockpile of Chinese imports. If near-future US political developments include a cut to IRA subsidies, this outcome will become conclusive, requiring something entirely new like perovskites to change it (and perhaps not even then, since tandems include silicon PV).
Wafer and polysilicon production outside of China is continuing to die off, with China actually strengthening its grip since the Inflation Reduction Act was brought in (CubicPV’s planned factory would’ve been reliant on many Chinese employees, because they are the people with the up-to-date expertise). At least cell and module manufacturing outside of China has increased dramatically in the past year – up almost twofold, as can be determined from Chinese wafer and cell export figures – and will continue to do so. Additionally, Widmar also mentioned a 30 GW to 40 GW module oversupply in the US in 2023.