The verdict is in: By unanimous ruling, the U.S. International Trade Commission found that increased imports are causing serious harm to U.S. solar cell and module manufacturing.
While domestic deployments of solar have grown nearly eightfold in the past five years, U.S. manufacturing has fallen behind. Previous trade cases were intended to stabilize pricing and result in new U.S. module capacity. But domestic production still hasn’t kept pace with deployments.
We estimate that 87 percent of U.S. solar installations in 2016 used foreign-produced panels.
Source: GTM Research Solar Supply Chain Service; GTM Research/SEIA U.S. Solar Market Insight Q3 2017
This is not an endorsement for Section 201-driven remedies. Far from it. We estimate that the remedies requested by the Section 201 petition would eliminate half of potential solar deployments over their term in exchange for limited new domestic module manufacturing.
But that doesn’t mean solutions for domestic upstream solar manufacturing should be abandoned.
Source: GTM Research U.S. Solar Outlook Under Section 201
Why is domestic solar manufacturing important?
Quite simply, manufacturing drives technology innovation. The National Science Foundation estimates that two-thirds of U.S. research and development dollars are spent by manufacturers (over 80 percent of which are from their own funds).
The 50+ percent reduction in solar costs over the past five years has primarily been brought about by innovations in the supply chain, from low-cost polysilicon production, to better throughput on materials, to improvements in efficiency.
If we believe that solar is a key part of the future of electricity, that the U.S. should be a leader in the clean electricity future, and that technology innovation is key driver toward that realization, then we must increase investment in domestic manufacturing.
It is possible to invest in R&D from the U.S. without domestic manufacturing; in fact, the U.S. boasts two global solar module leaders in SunPower and First Solar, who conduct much of their R&D in the U.S. while manufacturing most of their product abroad. However, the environment is changing, with more research-heavy competitors and less patient capital.
We also don’t dismiss the innovations and contribution from balance-of-systems technologies, which make up most of the remaining U.S. solar manufacturing landscape. But the module represents nearly one-third of PV system costs, and more than half of all hardware costs. No technology leadership strategy for solar can leave the module supply chain behind.
Nor are modules simply a commodity better suited for outsourced manufacturing. Even a simple survey of the Solar Power International exhibition floor in Vegas showed that foreign module providers are eager to compete on technologies beyond standard multicrystalline silicon: passivated emitter rear contact, half-cut cells, shingled cells, bifacial and even whispers of new challengers in n-type mono.
Upstream process improvements, from low-cost mono to diamond wire saw wafering, are a commercial reality. The bill of materials is cheaper than ever, and yet suppliers are comfortable offering better and longer warranties. These invisible innovations in the manufacturing process are driving lower costs for foreign manufacturers -- not just cheap labor.
The U.S. doesn’t risk losing the lead in solar technology. It needs to catch up.
Why Section 201 is unlikely to reinvigorate domestic manufacturing
Despite all this, levying tariffs or other protectionist measures is misguided. To quote one policymaker: “We should be in a race toward low-cost solar, not higher-cost solar.”
Tariffs that aren’t based on preventing illegal dumping (which, to be clear, is not the rationale for a Section 201 petition) fundamentally accept that the U.S. has been left behind. They aim to attack the symptoms and not the root cause, which is a systemic unwillingness to craft clear, long-term policies that support U.S. solar manufacturing.
Research further indicates that benefits to the domestic industry in safeguard cases are short-lived. For example, a 2013 Georgetown University Law Center study of three U.S. trade cases petitioned under Section 201 of the Trade Act of 1974 found that “none of the three industries achieved sustained competitiveness after safeguards terminated.”
For solar, the timeline for Section 201 creates many risks for manufacturers considering investing in U.S. capacity. To reach reasonably competitive costs, manufacturers must build at scale -- at least 500 megawatts, if not a true 1 GW facility.
And remember, the petition scope is on PV cell manufacturing at a minimum, a more difficult and expensive production stage than the final module assembly. This requires investments in the range of low triple-digit millions, all of which must be recovered within the period of the tariff. That will certainly prove tricky.
Section 201 remedies are set for a period of up to four years, with a potential extension up to eight. However, none of the six Section 201 safeguards initiated since the formation of the World Trade Organization have lasted for the full four-year term. Indeed, international pressure, whether through the WTO or retaliatory tariffs, could shorten the duration of any proposed safeguard.
Meanwhile, new equipment (and moving equipment from outside the U.S.) would likely take 12-18 months to reach completion -- and longer to ramp up to full capacity. By the time most new manufacturers are shipping modules in bulk, the tariffs could be nearing sunset.
Any investment in manufacturing would have to be predicated on a belief in long-term competitiveness at scale without tariffs. With potential exclusions for Free Trade Agreement countries in the ITC’s injury determination, the U.S. would also compete with other geographies that may offer better support or cheaper costs.
But U.S. upstream solar manufacturing isn't impossible (and given the interest, one or two suppliers might pull the trigger regardless). Paper exercises by manufacturers indicate that pricing (with good margins) for domestically produced multicrystalline modules could be between $0.42 per watt to $0.50 per watt at the gigawatt-scale by 2020. That’s compared to projected average volume pricing of $0.25 per watt to $0.34 per watt for Chinese Tier 1 suppliers. And while the gulf seems large, a number of creative alternatives to tariffs could help bridge the gap.
Six alternatives to tariffs
So what could the U.S. do to support real investment in domestic solar manufacturing without sacrificing a strategic market?
1) Support domestic products with a tiered investment tax credit
The federal Investment Tax Credit has been a clear driver for historical solar growth and will soon step down as solar begins to compete economically with traditional power generation. Instead of stepping the ITC down from 30 percent to 10 percent, keep it at 30 percent for domestic manufacturing.
With turnkey utility PV system engineering, procurement and construction (EPC) prices nearing $0.85 per watt by 2020, a 20 percent difference in the ITC would level the playing field between foreign and domestic manufacturing. Arguably, it could also drive up the federal government’s returns on the ITC.
2) Expand federal targets for renewable procurement
Government and military renewable purchases can already give preferential treatment to U.S. products through the Buy American Act and Trade Agreements Act-compliant procurement. By expanding federal renewable (specifically solar) procurement targets, the federal government can reduce costs, increase energy independence and support domestic suppliers at the same time.
3) Direct collected duties toward supporting domestic manufacturing
If the Trump administration truly believes in the art of a (good) deal, it should take the duties collected from existing tariffs on Chinese and Taiwanese solar products and equitably redirect them toward new manufacturing investment. In other words, let’s build a wall of solar manufacturing and get China to pay for it.
4) Provide loan support or guarantees for U.S. suppliers
The much-maligned DOE loan guarantee program could be another effective tool in two ways. First, the program can offer loan guarantees directly to domestic manufacturing facilities (hopefully conducting heavy diligence to avoid another Solyndra). Second, the program can offer preferential treatment to loan guarantees for power plants that utilize innovative, domestically manufactured technology.
5) Subsidize the solar supply chain
A dated (but still informative) research study by NREL from 2013 indicated that China’s advantage in cost of module production primarily came from purchasing power at scale, as well as low-cost regional equipment and material supply.
Indeed, part of the Department of Commerce’s rationale to levy existing anti-dumping/countervailing duty tariffs on Chinese solar modules (which have been in place since 2012) was predicated on its belief that China’s government involvement distorted pricing on key materials like polysilicon and aluminum extrusions.
To beat China at its own game, the U.S. needs to invest in the full bill of materials, not just the primary pieces of the module value chain. Copying from China’s textbook, low-cost loans, technology development assistance, cheap land and other forms of non-monetary encouragement should go beyond silicon, wafer, cells and modules to include encapsulants, coatings, solar glass and other solar materials.
6) Provide assistance for workforce and technology development
One common reason for not bringing manufacturing to the U.S. is the relative scarcity of experienced solar manufacturing engineers. Like the proposal to invest in technology, the U.S. also needs to invest in ideas and people.
While a talent pool of solar production engineers may not seem like much, we always point to knowledge-sharing as a key driver toward pushing crystalline-silicon solar costs down. Indeed, a parallel example for CdTe came just a year after First Solar acquired the IP from GE’s failed CdTe effort (PrimeStar) -- efficiencies skyrocketed to parity with standard multicrystalline silicon in the span of a few years.
Proliferation and free movement of solar manufacturing expertise increases the chances of continued innovation that can further drive down solar costs -- and R&D with day-to-day access of operating factories at scale smooths the path of technology from lab to roof.
We can’t say for certain that these investments will yield a thriving U.S. solar manufacturing sector for years to come. But if the U.S. hopes to lead solar innovation, an investment that seeks to lower the cost of domestic solar is the better path.
MJ Shiao is the head of Americas Research at GTM. Shayle Kann is the senior vice president at GTM and head of GTM Research.