Resell, Recycle, or Dispose: How the End-of-Life Decision Gets Made for Commercial Solar Panels
The U.S. Department of Energy put the numbers in writing. In its Photovoltaics End-of-Life Action Plan, the Solar Energy Technologies Office estimated that sending a retired module to a recycler costs a waste generator roughly $15 to $45, while sending the same module to a landfill costs $1 to $5. That gap, somewhere between five and forty times, explains almost everything about how end-of-life solar has been handled in the United States so far.
It also explains why the decision is more interesting than it first appears. If the only question were cost per module, every array in the country would go to a landfill and the conversation would be over. It is not the only question. Regulation, residual generating value, material recovery, freight, and the paperwork your lender or your landowner will eventually ask for all push in different directions, and they push differently depending on the condition your modules are actually in.
What follows is a routing exercise, not a ranking. If you are responsible for a commercial rooftop array or a utility-scale site that is coming down, being repowered, or being partially replaced after damage, your practical work is establishing which condition each part of your fleet is in and then accepting what that condition costs. Those conditions are nowhere near equally expensive, and almost no site you will deal with holds only one of them.
The Three Facts That Decide Everything
Before you can route anything, you need three pieces of information about the modules. Everything downstream follows from these.
The first is generating output. A module that still produces a defensible percentage of nameplate has a market for you. A module that has degraded past the point where a buyer will underwrite it does not, regardless of how intact it looks.
The second is physical condition. Cracked glass, delaminated laminate, water ingress at the junction box, damaged frames, and burned bypass diodes all move a module out of the resale conversation and into the material recovery conversation.
The third is chemistry. Crystalline silicon and thin film behave differently under leaching tests, and older modules generally carry more lead in the solder and the ribbon than newer ones do. Chemistry is the fact that decides whether you are dealing with ordinary solid waste or with regulated hazardous waste.
Two definitions are worth fixing here, because the industry uses them loosely. Commercial volumes generally mean rooftop or ground-mount arrays in the hundreds to low thousands of modules, typically owned by a business, a developer, or a property owner. Utility-scale volumes mean the tens of thousands to millions of modules found on generating facilities that sell power onto the grid. Residential volumes, meaning a handful of modules off a single home, are a genuinely different logistics problem and are usually handled through installers or consumer marketplaces, not through project-scale disposition. This piece is about the two you are most likely dealing with.
Before Anything Leaves the Site, You Owe a Hazardous Waste Determination
This is the step that gets skipped, and it is the one with legal teeth. Under the federal Resource Conservation and Recovery Act, you are the party generating the waste, and the obligation to classify it sits with you. The Environmental Protection Agency states the position directly: it is the responsibility of the generator of the solar panel waste to determine if the panels are hazardous, either by performing the appropriate tests or by using generator knowledge.
The question test is the Toxicity Characteristic Leaching Procedure, usually shortened to TCLP. It is a laboratory method that simulates the acidic conditions inside a municipal landfill and measures what leaches out of a crushed sample. If lead comes back at or above 5.0 milligrams per liter, the waste exhibits the toxicity characteristic and is a hazardous waste under federal rules. Cadmium has its own threshold and matters most for certain thin film chemistries.
The result is not predictable from the outside. EPA’s own summary of end-of-life testing data says some panels exhibit the toxicity characteristic and some do not. Two arrays of the same age from different manufacturers can land on opposite sides of the line. That is why generator knowledge, meaning documented prior test results on the same make and model, is worth accumulating if you own a large fleet. It is cheaper than testing every lot.
There is one exemption that reorganizes your whole decision tree. EPA does not treat used modules as solid waste at all when they are sent to resellers for legitimate reuse or repair. A module going to a buyer is not a waste and does not require a determination. A module going to disposal is, and does. The moment you decide a module has no further generating life, you have created a regulated waste and inherited a set of obligations that did not exist a minute earlier. Deciding that too early, for a lot that could have been sold, is one of the more expensive unforced errors in this field.
When the Module Still Produces Power Somebody Will Buy
The markers you are looking for are a flash test within a range a buyer will accept, a sound laminate, a straight frame, and enough of the same make and model to make a truckload worth someone’s time. Repower projects are the classic source. If you swap ten-year-old modules for higher-output replacements, you are not generating waste at all, at least not for the portion that still works.
That window has narrowed considerably. The fifth annual PV Module Price Index compiled by EnergyBin and Buckstop, published in March 2026, found that used modules averaged $0.058 per watt in the fourth quarter of 2025, about 30 percent below where they sat in January 2024, while new TOPCon modules were listed around $0.090 per watt. When a new module with a full warranty costs less than four cents per watt more than a used one with no warranty and unknown handling history, the resale case gets thin fast.
If you have sat through a repower disposition plan built on optimistic resale assumptions, you already know how it ends. Model resale as an offset against disposition cost rather than as a revenue line. It reliably reduces what you spend, and it rarely pays for the project. Uniform lots of a single make and model, palletized properly and documented with flash test data, are the ones that will actually move for you.
When the Power Is Gone, but the Materials Are Not
The next condition covers your modules that will not generate any usable standard while the materials inside them remain intact and separable. Hail damage, transit breakage, cracked glass on otherwise sound laminates, and cell microcracks from racking failures all sit here.
You need to know what material recovery genuinely yields, because the marketing language around recycling tends to imply more than the mass balance supports. By weight, a crystalline silicon module is dominated by two materials. The Department of Energy’s figures put a silicon module at roughly 86 percent glass and 11 percent aluminum by mass. The aluminum frame is the most reliably valuable piece and comes off first, cleanly, with a good scrap price behind it. Copper from the junction box and wiring is the second dependable stream.
Glass is the awkward one. It is the bulk of the module and the least valuable per ton, and once it has been laminated to an encapsulant, it usually comes out as mixed cullet rather than as glass a float line will accept. Silicon and silver are the opposite problem: high value per gram, present in very small quantities, and chemically bonded inside the cell stack in a way that mechanical processing alone does not fully release. Silver in particular is worth more than the glass, aluminum, and silicon in a module combined despite being a rounding error by weight, which is why the more advanced separation processes exist at all.
The practical consequence is that recovery percentages quoted by weight and recovery percentages quoted by value describe two very different achievements. When you read a number, check which one it is.
When the Waste Rules Take Over the Decision
The third condition catches your modules that fail TCLP, modules from a fire or flood event, modules contaminated on site, and modules old enough and degraded enough that neither generating value nor clean material streams remain.
This is the most expensive outcome you can land in, and the cost is mostly regulatory rather than physical. As a hazardous waste generator, you face manifesting requirements, transporter and facility permitting constraints, accumulation time limits, and record retention obligations. The modules themselves are no harder to process than the merely damaged ones. The paperwork, the licensed transport, and the permitted destination are what move the price.
Two things soften it for you. First, RCRA contains exclusions that allow hazardous secondary materials to be recycled under conditions less onerous than full disposal, which is a large part of why recycling a failing module is often cheaper than disposing of it as hazardous waste. Second, several states have moved photovoltaic modules into universal waste, which is a reduced-burden subset of the hazardous waste rules designed for widely generated items. More on that below.
No Real Yard Sorts This Cleanly
The tidy version of this decision assumes your fleet arrives in one condition. In practice, your decommissioning yard will hold all three at once: a majority that still tests fine, a meaningful minority with broken glass from removal and handling, and a smaller group that was already failing before anyone touched it.
That mix creates a real operational problem. Sorting has to happen somewhere. Doing it on site costs you labor and lay-down space at the exact moment the site is busiest. Doing it at an intake facility means paying freight on modules that turn out to be worthless. Splitting the fleet across two vendors, one buying and one recycling, usually means two sets of logistics, two sets of documentation, and a reconciliation problem when your lender asks what happened to every serial number.
Your alternative is a single disposition path that carries both outcomes under one project file. So what should you expect commercial solar panel recycling to look like when resale and recovery sit in the same engagement? Where that path is available to you, as it is at Solar Recycling, which handles recycling and asset recovery for commercial and utility-scale fleets, your project moves through a fixed sequence instead of a series of unconnected transactions. Discovery establishes what you actually have, module by module and pallet by pallet. Estimation prices that inventory once it exists, which is the point at which your sort ratio stops being a guess. Logistics moves the fleet. Reprocessing takes whatever could not be resold. Certification closes the file, and your project ends with a Certificate of Completion recording what was resold, what was recycled, and where each material stream went.
That closing document is the one your lender, your landowner, and any regulator with a question will eventually ask to see. It is far easier to produce when a single file covers the whole fleet, and considerably harder when two vendors each hold half of the story.
Why Module Recycling Economics Do Not Behave Like E-Waste
People coming from IT asset disposition often expect solar to work the same way, and it does not. The differences are worth naming, because they explain the pricing you will be quoted.
Value density is the first. A retired server concentrates gold, palladium, and copper into a chassis that weighs a few kilograms. A photovoltaic module weighs roughly twenty kilograms and is mostly glass. The recoverable value per kilogram is an order of magnitude apart.
Freight is the second, and it is usually the largest single line item. Modules are bulky, fragile, and awkward to palletize, and the material inside them does not appreciate with distance. Shipping a container of modules across three states can cost more than processing them. This is why facility location and the density of a provider’s intake network affect your price more than the gate fee does.
Design is the third. A module is laminated on purpose, engineered to survive thirty years of weather without separating. Every property that makes it durable in the field makes it expensive to take apart. That is a permanent structural cost, not an inefficiency someone will optimize away next year.
The Department of Energy’s own target is to bring the cost of recycling a module below $3 by 2030, roughly a tenfold reduction from today’s range. That target exists precisely because the current economics do not close on their own.
Where the Project Sits Changes the Answer
Federal rules set the floor. State rules frequently set something higher, and they are the reason your fleet can be routed two different ways in two different states.
At the federal level, solar panels are not currently classified as universal waste. EPA initiated a rulemaking in October 2023 proposing to add them, and that rulemaking has not been finalized. Until it is, the federal position remains a case-by-case hazardous waste determination by the generator.
Two states have already moved. EPA’s own reference material lists California and Hawaii as regulating photovoltaic modules as state universal waste. Universal waste status is generally good news for you as the asset owner: it keeps modules inside the hazardous waste system while relaxing manifesting, accumulation, and transport requirements enough to make compliant handling practical at volume. Washington took a different route years earlier with a manufacturer responsibility program.
Separately, several states and individual landfill operators restrict or refuse photovoltaic modules outright, either through disposal bans or through facility acceptance policies. Do not assume the landfill option is available just because it is cheap on paper. Confirm acceptance before you build a budget around it, because discovering the restriction after the modules are already palletized is how a disposal plan turns into a storage problem.
What Each Outcome Costs or Returns
Rough shape of the numbers, so the conditions carry weights.
Resale returns money to you instead of costing it, but at secondary market prices that have fallen sharply, and only for uniform, tested, documented lots.
Material recovery on damaged modules is a net cost, with third-party figures from the Department of Energy putting recycling at $15 to $45 per module against $1 to $5 for landfill, and with freight often exceeding the processing charge on long hauls.
Hazardous disposition is the highest cost in the set, driven by manifesting, licensed transport, and permitted destination facilities rather than by processing difficulty.
A mixed fleet lands somewhere between the three, weighted by your actual sort ratio, which is the number you should establish before you request a single quote.
The Variable You Control
You do not control the price of silver, the freight market, or the pace of EPA rulemaking. What you control is the scope you write before any of it starts.
Modules presented with output data, a module inventory, and a TCLP result or documented generator knowledge get routed accurately and priced accordingly. Modules presented as pallets in a yard, mixed in condition and stripped of records, get quoted as a single undifferentiated waste stream, which means your whole lot gets priced at the cost of its worst units. The difference between those two outcomes is usually larger than the difference between any two vendors you might be comparing.
So the leverage sits earlier than most owners expect. By the time competing quotes are on the table, the range those quotes can occupy has already been fixed by your removal scope: what it required to be tested, what it required to be recorded, and how it told the crew to stack a pallet. Write that document carelessly and you will spend the rest of the project choosing between numbers that are all worse than they needed to be.
