PV module recycling and deep processing: a working market study
This section is written for someone deciding whether to build a plant, not for someone writing a brochure. It sets out the installed base and its timi
| Component | Share of module mass | Recovery route | What limits recovery |
|---|---|---|---|
| Glass | 67.45% | Cullet for glass processing or secondary applications | Encapsulant residue and metal fragments; both introduced upstream |
| Aluminium frame | 12.73% | Direct remelt after decoating and screening | Silicone sealant and glass shards bonded to the sections |
| Encapsulant (EVA) | 6.69% | Energy recovery or thermal processing | Bonded by design; this is the material that makes dismantling hard |
| Backsheet and other polymers | ≈ 6.08% | Energy recovery or specialist processing | Mixed polymer types are difficult to sort |
| Silicon in cells | 2.73% | Cell concentrate to a refiner or a metal route | Needs concentration before any chemical route can run economically |
| Junction box, silicone, cables | ≈ 2.80% | Copper-bearing and polymer streams | Only recoverable if removed before glass handling |
| Interconnect (Cu, Sn, Pb) | 1.48% | Recovered with the cell fraction | Locked into the laminate; released only after delamination |
| Silver-bearing metallisation | ≈ 0.04% | Hydrometallurgical route on the cell concentrate | Concentration, reagents, effluent and assay |
Shares from Fraunhofer ISE module composition data (2021), quoted as published. Component grouping differs slightly from the original table so that each row maps to a recovery route.
The following is a sensitivity calculation, not a market price. It takes the composition above, expresses it per tonne of end-of-life modules, and applies stated commodity prices to show how the value is distributed. It assumes complete recovery of each material, which no real process achieves.
| Stream | Mass per tonne of modules | Illustrative price applied | Illustrative value per tonne of modules |
|---|---|---|---|
| Aluminium frame | ≈ 127 kg | USD 1,800–2,200 per tonne of clean scrap | ≈ USD 230–280 |
| Glass | ≈ 675 kg | USD 0–30 per tonne, or a disposal cost | ≈ USD 0–20, or negative |
| Silicon in cells | ≈ 27 kg | Not a traded commodity in this form | Value sits in the concentrate, not the raw silicon |
| Silver (at USD 25/ozt) | ≈ 0.34 kg | USD 0.80 per gram | ≈ USD 275 |
| Silver (at USD 35/ozt) | ≈ 0.34 kg | USD 1.13 per gram | ≈ USD 385 |
| Silver (at USD 45/ozt) | ≈ 0.34 kg | USD 1.45 per gram | ≈ USD 495 |
| Copper and other metals | ≈ 15 kg | Market dependent | Small in value, useful in stream cleanliness |
Illustrative only. The silver figure assumes the silver-bearing fractions in the published composition data are predominantly silver, and assumes 100% recovery from the full tonne of modules — which requires the cell fraction to be separated first. Real recovery is lower, and reagent, energy, labour and effluent costs are not deducted.
Two trends push the value of an old module down. Modules are getting more efficient, so the same wattage needs less material — silicon usage has fallen from roughly 16 g/Wp in 2004 to about 2.0 g/Wp in 2024 according to Fraunhofer ISE. And the industry is actively reducing silver, through multi-busbar and busbar-less cell designs and through substituting silver with copper or aluminium.
For a plant being planned now that has a favourable implication and an unfavourable one. The modules reaching recyclers today were made when silver loadings were higher, so the near-term silver case is better than the long-term one. But a plant that depends on silver for its returns is depending on a feedstock characteristic that is deliberately being designed out.
That is an argument for treating silver recovery as a staged investment rather than the foundation of the business, and for making sure the plant's core economics work on aluminium and glass alone.
Send a description of the modules you receive — type, vintage and condition. We will tell you which streams the composition logic suggests and which ones your line cannot reach yet.
By material content, an illustrative calculation puts aluminium at a few hundred dollars per tonne and the silver at a few hundred more at recent metal prices, with glass worth little or nothing — and possibly a disposal cost. Real net value is far lower, because recovery is incomplete and processing costs are not deducted. Your own gate fee, outlets and costs decide the actual number.
Per tonne it is, because it is the most expensive element present. In total value per module, the aluminium frame is usually more significant, because there is vastly more of it. That is why we treat silver recovery as a conditional add-on rather than the foundation of a plant's economics.
It is a bulk material with a low price per tonne and a high transport cost relative to its value, and it has to compete with locally available virgin glass. Whether it is an income stream or a cost depends almost entirely on whether a processor within economic distance will accept your grade.
The trend is downward per tonne, not upward, because modules are using less material per watt and less silver in particular. Value per tonne also depends on commodity prices, which move independently. A robust plant design does not rely on material value rising.
Tell us what arrives at your gate and what has to leave. The engineering team replies with the station configuration, the scope boundary and the information still missing before drawings start.
Prefer to talk first? WhatsApp +86 182 0399 3035 or email [email protected].
Three details are enough to start: what modules you receive, how fast you need to process them, and what output you have to produce.