Solar Forced Labor Issues, Part 1
More data points on the move from "just-in-time" to "just-in case"
For this post I want to get right to the point: the single biggest risk to the US solar industry in the next 6 to 12 months is that large quantities of modules are blocked from importation due to insufficient documentation regarding forced labor in their manufacture. The Wall Street Journal is reporting today of significant disruptions – WSJ reports that “solar-panel manufacturers Longi Green Energy Technology Co., Jinko Solar Co. and Trina Solar Co. are among those affected, people with knowledge of the events say. Longi has temporarily halted a panel factory in Vietnam that supplies the U.S. as a result, some of those people say.” I personally know of additional issues than reported in the article - more detentions, and some suppliers not shipping to the U.S. at all for the moment.
Failure of suppliers to be able to satisfy U.S. Customs (“CBP”) will result in significantly delayed equipment delivery (the product could be detained at the border for weeks or months, at high storage costs), or even equipment that cannot be imported into the US at all.
In this first post in a series on this topic, I discuss the forced labor issue generally, how it is manifesting in solar, and why this issue will become more of a risk in the second half of 2022. In later posts, we’ll get into the more advanced, detailed nitty-gritty of how this impacts contract negotiations and specific contract terms under negotiation right now.
Module contract negotiations are heating up, but forced labor concerns are a wrench in the near-term gears. I generally agree with Phil Shen from Roth Capital that forced labor law “continues to be a challenging situation . . . specifically, neither side wants the risk, and the parties are having a difficult time figuring out how to share the risk.” I would add to that: due to the leverage that some module vendors have – particularly the ones willing to take a shot at importing into the United States right now – I think inevitably buyers will have to share this risk, particularly as it relates to delays in delivery. More on the specifics in follow-up posts. First, we need to cover the basics.
Forced Labor and Importation
It’s been the law of the United States for a very long time – at least since 1930 – that you cannot import goods that were made with forced/involuntary labor. CBP is entitled to block importation of any product made with forced labor, and can also seize the goods. As a lawyer I’m tempted go into the intricacies of what “forced labor” means, but I think most people get the gist of it from the term itself.
The problem on the ground is proving the absence of forced labor in the manufacture of a product. CBP has a difficult job implementing the law in this regard: one cannot tell from physically examining any particular product at the border whether forced labor was used in that product’s manufacture. It’s an issue of documentation.
Given the complexities and challenges noted above, documentation to prove absence of forced labor does not become an issue unless CBP has some reason to suspect that a particular product, or any of its components, was made with forced labor. This suspicion could be based on government intelligence and/or reports from third-party organizations.
Once suspicion about a particular category of imports draws CBP scrutiny, the prospect of potential forced labor in a supply chain generally is like iodine in water – even if the actual amount is very small (and obviously any amount should rightfully be reduced to zero), it can be enough to impact the whole water supply, due to the documentation challenges that become required for the entire industry.
Because most products being imported into the US are not generally under suspicion of forced labor, intricate documentation about the working conditions of workers making any one specific product (or component of the product) are not generally kept in supply chains in a manner that links the specific labor utilized to make any one specific product. Even increasing ESG requirements only look at systems and working conditions generally, and not at the granular level of one exact worker tied to one exact product made.
For a solar module, there are many components and manufacturing steps: polysilicon, ingots/wafering, cell manufacture, aluminum frames, glass, back sheets, junction box, cable connectors, etc.
Some components of a module are interchangeable (I won’t say “commodity”). For example, aluminum frames enter into a factory and may look the same but could originate from different suppliers. The frames could be lumped together in the same pile of inventory on the factory floor before they are put into the final product – and this often has been general practice for certain components in a supply chain, because no one has ever had need to keep them separate, or keep documentation about how they are produced. Most critically: this is the same with different batches of polysilicon from different suppliers.
So, in order to address forced labor concerns, the nature and operation of whole supply chains will need to be documented, and often, processes changed.
The reason I go through all of this is to help highlight the magnitude of the shift required in supply chains to produce the level of documentation that CBP may require. It requires component procurement contracts to be amended with new requirements, manufacturing processes to be changed, individual items to be traced throughout the system in ways not previously required, and in some cases for suppliers to obtain entirely new sources of components and raw materials.
Solar Industry and Forced Labor Concerns
The issue began to arise in solar due to third-party reports of forced labor in polysilicon, particularly in the Xinjiang region of China. Roughly half of the world’s solar grade polysilicon is produced there.
In 2021 it became apparent that a few large module vendors – among them, Jinko Solar and Longi seemingly have had significant product detained until they could provide details proving the absence of forced labor. There could be others that have been detained (and/or have product presently detained), too – import matters with CBP are confidential and so there is no way to check public records to verify these things completely, in the absence of vendor self-disclosure.
Meanwhile, at the end of 2021, Congress passed the Uyghur Forced Labor Prevention Act (“UFLPA”), which was signed into law by President Biden. The UFLPA implementation just went into effect in June 2022, with CBP publishing guidance for importers.
Realistically, the UFLPA sets a higher bar for supply chain documentation sufficient to satisfy CBP that there is no forced labor in the manufacture of solar modules – and it seems importers were already having trouble with the level of the old bar. I think the UFLPA is basically a de facto ban on imports containing any content from Xinjiang, because the bar for bringing in goods with Xinjiang-content is now so high that realistically no importer will be able to do it (or will even try, given the costs and potential liability of having their goods blocked from importation or even seized by CBP).
What exact level of documentation will be required from now on for solar modules? While CBP has released detailed general UFLPA guidance, there is no way for CBP to bless, in a pre-determined manner, all of the different supply chain combinations and formatting of record-keeping processes, for all different suspect products. In addition, it is not 100% certain that all CBP agents will exercise the same judgment calls on the sufficiency of documents – think baseball umpires that have similar, but not quite the same, calls of balls and strikes.
It’s becoming clearer that not only will an importer of crystalline silicon solar modules need to trace the sources of its polysilicon (and ensure no mixing or matching that makes the poly traceability indeterminate), but the polysilicon manufacturer will need to trace the quartzite (more on this in later posts).
For any Tier 1 module vendor, many do not make their own polysilicon (or even their own wafers). In turn, polysilicon manufacturers may procure quartzite from another supplier. So, for a module vendor, there could be anywhere from one to half a dozen or so companies in between the quartzite and the final, fully-assembled module. Until recently, module vendors did not have to request documents like this up the supply chain (or even anywhere in the supply chain), and it’s unlikely all key component procurement contracts had contractually required this documentation to accompany the physical components sold to the module vendor.
Module vendors will likely be requiring this documentation from their sub-suppliers now. But the question is, how long will it take for the supply chains to respond to these documentation needs?
One could argue that module vendors should have known that this was coming, and have been prepared already, when detentions of products by CBP began last year. The Solar Energy Industries Association published guidance regarding traceability requirements over a year ago, in April 2021.
I have empathy for that position – it has some truth. Though to be fair, the exact granular requirements for importation documentation are arguably not yet known even now. But regardless of anything that could have been done in the past, the reality of the present is that many vendors just don’t know what they don’t know. That is fundamentally why this risk is so big. Most vendors have never been through something like this before. Many smaller manufacturers have probably never hired a U.S. trade lawyer to assist them with something as complex as this. They have historically viewed CBP more as the DMV accepting and stamping their import forms, and collecting tariffs when applicable, rather than how they should really look at it: CBP as the FBI – a federal agency with knowledgeable professionals who can and will investigate and act as strict gatekeepers of federal law and guardians of the policy objective to eliminate use of forced labor.
Also, the uncertainty around the Auxin petition wrenched the industry and diverted attention from this issue, since most vendors stopped shipping to the U.S. for a time anyway. There is an overlooked interplay here – why go through entire supply chain reform for traceability documentation up and down the entire supply chain – when you don’t even know if you’ll be serving the U.S. market anyway, due to the Auxin tariff? In that regard, the havoc that tiny little Auxin Solar wreaked on the industry will still be felt for a time to come, even with a tariff waiver executive order. It ripples into this issue, too.
So that’s where we are now: newer, stricter, maybe-not-yet-100%-clear documentation requirements for importation, requiring traceability of components in a manner than most companies never have had to do, and many of whom don’t know what they don’t know.
As a lawyer, I often get requests to “draft language” to protect a client on issues like these, whether seller or buyer – and of course, that’s part of my job.
But clever ink in a contract alone cannot mitigate this risk, due to all of the complexities involved. Module manufacturers need to reform their supply chain documentation if they want to serve the U.S. market. CBP, as the gatekeeper, will eventually know who has the proper documents and who doesn’t.
Buyers need to do their due diligence on a vendor’s supply chain and recognize these complexities.
Final Thought: Just-in-Time to Just-in-Case
Besides diligence, and doing your best with the contract ink (more on that coming up), what’s the best way for a module buyer to mitigate this risk?
Time.
Allow for a potential delay of months in delivery. This is extremely difficult and costly to do in many instances – I get it, this is easy to say but hard to do. It impacts EPC schedules, financing, company metrics reported to investors, etc. But the alternative could be worse.
This risk is just one of many data points painting a bigger picture: the renewables industry is in the midst of a significant shift that many other global supply chains are experiencing – a shift from “just-in-time” to “just-in-case” operation. Management teams and their lawyers who recognize the full extent of this shift – as it will manifest particularly in solar and storage in the years ahead – can avoid headaches and extra costs. Management teams who not only recognize this but who have investors and stakeholders who also understand this will do even better – over the long-term, even if the shorter-term metrics don’t initially look as good.
Also, as I noted at the end of my first post – modules are not really commodities. A module that gets to your project two months late beats a module that gets locked in a bonded warehouse at the border for 10 months, beats a module that never gets there at all. That is true product differentiation, though it won’t show up as a line on the module spec sheet.
As related examples, we’ve seen persistent global transportation delays, and exogenous events like COVID and war. It is now more difficult to rely on the concept of your product showing up from a factory Asia to your U.S. project site just when you want it – not too soon (because you’d have to store it and/or pay for it earlier than you would like and/or have product degradation) but also not too late (where delays increase your costs or even threatens the project), and all at rock bottom cost.
Even when the industry gets over this particular forced labor documentation hurdle – which it will – “just-in-case” will likely be the new normal for years to come. Maybe U.S. factories spurred by incentives can eventually fix this risk – but even there, as a developer I wouldn’t have my whole project portfolio turn on yet-to-be-constructed U.S. manufacturing plants being timely built just-in-time to deliver modules just-in-time for my extremely tight project timeline. Have your lawyer like me load up the ink on the Delivery LD amounts against the vendor, to protect yourself, if you can win those. But ultimately, the plan may just need some slack, too.
Len Conapinski
Len Conapinski is a partner at DuFour Conapinski Ha LLP. Please note that all views here are my own personal thoughts, are not intended to be taken as legal advice, and are not reflective of the views of my law firm colleagues, or our firm as an organization, or any client of our firm.