Section 232 Polysilicon Tariffs: What EPCs and Developers Should Be Doing in the Next 120 Days

On August 7, the White House signed Section 232 tariffs covering polysilicon and the products made from it. The structure is unusual, and the timing matters more than the headline number.

Here is what changed, and what it means for anyone with modules to procure between now and the middle of next year.

What the order actually does

Two mechanisms, applied together:

A 15% ad valorem tariff on imports of polysilicon and derivative products across the solar supply chain.

Minimum import prices — a floor beneath which imported product cannot enter, regardless of what a supplier quotes:

Product Minimum import price Polysilicon $21 / kg Ingots and wafers $100 / kg Solar cells $0.22 / W Solar modules $0.38 / W

Both take effect at 12:01 a.m. Eastern on December 4, 2026.

The price floor is the part worth reading twice. A tariff raises landed cost proportionally. A floor sets an absolute bottom on what imported product can cost, which changes the shape of the market rather than just its level. It removes the low end entirely.

The administration's stated basis is a Commerce finding that current import volumes and conditions pose a national security risk, paired with the reality that the United States imports essentially all of its ingots, wafers, and cells today. The order's goal is to make domestic polysilicon production — both solar-grade and semiconductor-grade — financially viable enough to sustain.

Two provisions leave room for movement. The U.S. Trade Representative can negotiate country-specific arrangements that adjust how the tariffs and floors apply. And Commerce is authorized to build an incentive program for companies onshoring polysilicon, ingot, wafer, or cell production, with a construction-start deadline of January 20, 2029. Neither changes anything before December 4, but both mean the landscape in 2027 may not look exactly like the one described this month.

What it does to module cost

Roth Capital Partners published estimates shortly after the announcement. These are their projections, not settled prices — but they are the clearest available read on direction and magnitude:

• U.S. assembly using imported cells: roughly $0.40/W, an increase of about $0.11/W

• U.S. assembly importing both wafers and cells: roughly $0.48/W, an increase of about $0.05/W

• Directly imported finished modules: roughly $0.38/W, up from approximately $0.24/W

Averaged across configurations, Roth estimates about a $0.10/W increase in module cost. Carried through to project economics, they project PPA rates would need to rise somewhere in the range of $4.00 to $5.00/MWh to fully absorb it.

Worth noting what the middle row implies: the configuration with the smallest increase is also the most expensive in absolute terms. Domestic wafer-and-cell assembly was already carrying cost; the tariff just compressed the gap between it and everything else.

The constraint most people are underweighting

The cost math will get the attention. The supply constraint is the harder problem.

Anza's read, which we think is the most operationally useful commentary published so far, is that developers pushed toward domestic product will find there is not enough domestic TOPCon available to meet demand — which means a meaningful share of projects get pushed back toward domestic PERC. That is a technology step backward for projects that were specced around TOPCon performance assumptions, and it has consequences for energy modeling, land use, and racking layout that do not show up in a per-watt price comparison.

Anza also flags wafers as the real story. Domestic wafer capacity became dramatically more valuable overnight, and they expect significant new investment there within about six months. That investment does not help anyone procuring in the next two quarters.

So the practical picture for Q4 and Q1: higher cost, thinner domestic TOPCon availability, and a possible spec change on projects that were already engineered.

What to do in the next 120 days

Identify which projects cross the December 4 line. Anything landing before that date prices under current rules. Anything after does not. If you have flexibility on order timing, this is the single highest-leverage decision in front of you.

Re-run economics on anything marginal. A $0.10/W swing does not break a strong project. It does break the ones that were already thin. Better to know in August than in January.

Pressure-test your domestic TOPCon assumptions. If a project is specced around domestic TOPCon and you do not have supply committed, treat that as an open risk rather than a line item. Understand what the PERC fallback does to your production model before you are forced into it.

Keep your domestic content and FEOC documentation current. These classifications were already consequential. Under a price-floor regime with country-specific carve-outs still to be negotiated, precision about what qualifies — and on what basis — becomes considerably more valuable than a general claim.

Widen the supplier conversation now. Concentration risk gets expensive fast when a policy change removes options from the low end of the market. This is the moment to know what your alternatives are, not the moment to start looking.

Where we fit

Exel Solar US is a sourcing and logistics partner. We work across a broad supplier network rather than a single line card, and we move product through hubs in California, Texas, and Ohio.

That model is built for exactly this kind of disruption. When availability tightens and the qualifying-product question gets more complicated, the useful thing is not a bigger warehouse — it is knowing where product actually is, what it qualifies for, and how fast it can reach a job site.

If you are working through what December 4 means for a specific project pipeline, we are glad to talk it through. Schedule a conversation.

Figures in this article are drawn from the White House fact sheet on Section 232 tariffs for polysilicon and its derivatives, published August 7, 2026. Price projections are attributed to Roth Capital Partners and represent that firm's estimates. This article is provided for general information and does not constitute pricing, availability, or procurement advice for any specific project.