If you like our site, mark us as a preferred source on Google — so you’ll see our articles more often in search!
★Mark us as a preferred sourceOne rarely stated consequence of the solar boom is that everything installed today will have to come down again in twenty-five to thirty years. A study published in Nature in August 2026 has now modelled what this means at global scale: solar panel waste recycling will remain loss-making for the next decade, break-even is expected between 2035 and 2040, and once the sector finally becomes viable, it will distribute the benefits very unevenly across world regions.
The international team led by researchers at Shandong University built its calculations on 32 world regions, 1,708 recycling practice scenarios and four raw materials (copper, aluminium, silver and silicon). The model links an integrated assessment framework (GCAM) with dynamic material flow analysis, life-cycle assessment and life-cycle costing. The analysis covers crystalline silicon modules only, since these have accounted for more than 90 per cent of global installations since 2012.
A 150- to 200-Fold Increase in Twenty Years
The starting point is deployment dynamics. Annual global solar installations have exceeded 100 GW since 2020, and installed capacity rises from below 800 GW in 2020 to between 3,781 and 5,037 GW by 2030, and to between 8,703 and 24,544 GW by 2060, depending on the scenario.
Decommissioning follows the same curve, delayed by roughly a quarter of a century:
- 2020: 0.02 million tonnes of retired modules per year
- 2030: 0.5 million tonnes per year
- 2060: more than 19 million tonnes per year
Cumulative volumes increase roughly 150- to 200-fold between 2030 and 2060, reaching 297 to 402 million tonnes by 2060. The lower and upper bounds reflect different socioeconomic and climate policy pathways — and, counterintuitively, the stricter climate target produces more waste, because it also means more panels installed.
For comparison, earlier IRENA-based estimates put the 2050 figure at 200 million tonnes. The newly published model therefore does not contradict the earlier picture; it pushes it out in time and revises it upwards.
Europe Leads Today, China Will Lead in 2060
The geography of the waste burden shifts as the period progresses. Between 2020 and 2040, high-income regions still account for more than half of global volumes (12.2–12.5 million tonnes), with the EU-15 as the largest single contributor at 5.8–5.9 million tonnes, or 25.2–26.0 per cent of the world total.
By 2060, upper-middle-income regions take over:
- China: 112.8–160.5 million tonnes (36.2–39.9 per cent of the global total)
- EU-15: 30.9–39.7 million tonnes
- United States: 25.7–39.0 million tonnes
- India: 26.8–35.2 million tonnes (8.6–9.2 per cent)
The share of lower-middle-income regions almost doubles between 2040 and 2060, from 7.7–8.1 per cent to 15.9–16.9 per cent. The lowest-income countries stay below 2 per cent throughout — a figure that, as we shall see, does not mean the issue will spare them.
Why Delay Is Not an Option: Lead, Cadmium and Silver in One Panel
The research starts from the premise that unregulated disposal creates losses in two directions at once. First, an environmental risk: heavy metals such as lead and cadmium can leach from modules into soil and groundwater. Second, a materials management loss: the panels contain silicon, silver, copper and tellurium — precisely the raw materials needed to manufacture the next generation of solar modules.
Recycling is therefore not only an environmental task but a supply security question. Regulation follows the same logic, though so far in only a few places: through the WEEE Directive, the EU mandated an 80 per cent recycling rate for decommissioned modules by 2018; China is targeting the recycling of 250,000 tonnes of retired modules by 2027; and the Australian state of Victoria has introduced a landfill ban on solar panel waste.
The Turning Point: Between 2035 and 2040
The study’s most newsworthy finding concerns the timing of break-even. Today, revenue per tonne of recycling stands at 131.8 to 304.3 US dollars, while costs run from 342 to 2,550 dollars — meaning the process is deeply underwater in economic terms.
Two effects reverse the picture by 2040. Prices for recovered materials, primarily silver, aluminium and silicon, rise, lifting unit revenue to as much as 457.1 dollars. At the same time, learning by doing and larger plant capacity bring costs down to between 55.9 and 402.7 dollars. The result: recycling becomes economically viable between 2035 and 2040 in every scenario examined.
That sentence can be read two ways. Optimistically: break-even is not a fiction but a calendar question. Pessimistically: for at least another decade, an industry that does not generate a profit on its own will have to be kept alive through regulation and subsidy — precisely during the period when waste volumes grow exponentially.
Mechanical, Thermal, Chemical: The Technology Mix Matters
The model puts three processes in competition and examines four technology pathways: business as usual (BAU), economic priority, carbon priority and technology diffusion.
The finding is unambiguous. Thermal processes deliver the highest benefit per unit, while mechanical processing — the least capital-intensive option, and therefore dominant in developing regions — generates the least value, because its metal recovery efficiency is lower.
On the carbon-priority pathway, assuming local treatment only, the model yields 670.8 billion dollars in cumulative economic benefits and 2.44 billion tonnes of CO₂-equivalent savings by 2060. That is more than 60 per cent better than the BAU pathway, mainly because the share of thermal processing rises from 10.1 per cent to 44.0–51.6 per cent. The technology diffusion pathway is more modest, delivering a 37.0–44.3 per cent improvement, because mechanical processing stays above 50 per cent within it.
The authors also flag a technological bottleneck: solar-grade silicon requires purity above 6N, increasingly 8N–9N, and removing carbon and metallic impurities at industrial scale remains economically difficult. Recovered silicon therefore typically ends up in downgraded applications such as battery anodes. Investment in high-purity refining technologies, such as directional solidification, is urgently needed.
Efficiency and Equity Cancel Each Other Out
This is where the research reaches its most interesting section. When the model allows international waste shipment, global benefits rise — but asymmetrically.
The best global outcome comes from carbon-priority trade organised around producer responsibility: 529.1 to 935.5 billion dollars in cumulative net benefits and 2.2 to 3.32 billion tonnes of CO₂ equivalent in emission reductions by 2060. Achieving this, however, requires waste to flow towards regions with advanced technology and low processing costs. If it does, the share of benefits accruing to upper-middle-income regions rises from roughly two-thirds under local treatment to around three-quarters.
The numbers are just as telling in the other direction. High-income regions generate 34.0–36.3 per cent of decommissioned modules but receive only 28.7 per cent of cumulative net economic benefits, because their processing costs are higher. The lowest-income regions, meanwhile, capture only 1.3–2.1 per cent of the benefits (4.2–10.6 billion dollars) and 1.2–1.8 per cent of avoided emissions even on the technology diffusion pathway.
More balanced trade regimes improve distribution but worsen overall performance. Broader global trade reduces global net economic benefits by 5.8–8.6 per cent, and regional trade by 2.1–5.4 per cent. The share going to lower-middle-income regions, however, rises from 3.6–4.3 per cent to 6.7–8.2 per cent.
The worst combination is the BAU pathway paired with producer-responsibility trade. Here, with mechanical processing at close to 60 per cent, benefits fall to 29.3–55.9 per cent of the theoretical maximum, while inequality is the highest of all scenarios. In other words, a poor technology mix and free waste flows together are the most damaging: they generate little, and concentrate even that in few hands.
The Declining Subsidy Is the Best Deal
The authors test five subsidy schemes, and the result is unusually clear-cut. The phased-out, declining subsidy delivers the largest improvement in inter-regional equality (5.8–8.1 per cent), and does so at just 2.9–11.9 per cent of the cost of continuous support — 0.4 to 1.9 billion dollars by 2060, against 16 billion for the continuous scheme. That is a 9.5-fold cost difference for a marginal performance advantage.
The mechanism is simple: support is withdrawn as soon as processing becomes self-sustaining. According to the model, middle-income regions turn profitable around 2032 and can phase out subsidies by 2044; high-income regions become viable by 2040 and complete phase-out by 2058.
Carbon-price-linked support, by contrast, backfires. More than 85 per cent of high-carbon-price schemes increase inequality, because high-income regions receive up to 200 dollars per tonne — nearly four times what developing economies get, many of which have no established carbon market at all. The inter-regional gap widens to 1,198–1,434 dollars per tonne, 1.4–4.3 per cent worse than the no-subsidy case.
This finding is remarkable in itself: a climate policy instrument intended to finance waste management can, under certain conditions, achieve the exact opposite of what was expected of it.
What the Model Handles Only in a Footnote: Reuse
The authors themselves note that refurbishment and reuse can in principle deliver substantially greater environmental and resource-efficiency benefits than material recovery, and can delay the need to build expensive processing infrastructure in regions that currently lack capacity.
Reuse also decouples part of the circular benefit from raw material prices: if no new module has to be manufactured, rising silver or silicon prices do not hit price-sensitive developing economies. Large-scale reuse is nonetheless constrained by variation in module condition, latent reliability risks and limited market confidence in second-hand products. The study therefore positions reuse for now as a transitional, complementary strategy rather than an alternative to processing.
What Does This Mean for European Practice?
The study is not a Hungarian or EU regulatory proposal, but its conclusions are directly relevant to European practice.
The first concerns the structure of targets. The authors warn that weight-based recycling targets — as in the WEEE Directive — naturally reward the recovery of glass and aluminium frames, because these make up most of a module’s mass. The substantive materials management benefit, however, lies in the silver and silicon present in low concentration but high value. Regions with strong institutional capacity therefore need more targeted, material-specific goals to move the market towards higher-value processing.
The second is trade transparency. Against a background of tightening global waste trade rules — the Basel Convention, the EU battery regulation, China’s import bans — and growing geopolitical uncertainty, the authors argue for transparent, standardised cross-regional cooperation mechanisms. Not to block waste flows, but to keep them out of informal processing and to clarify where responsibility sits.
The third is the logic of subsidy policy. The declining subsidy principle is not new in waste management, but it is rare to see a calculation demonstrate so clearly that support concentrated in the early phase and withdrawn on schedule is not the thriftier compromise against continuous support — it is simply the better solution, distributional effects included.
Frequently Asked Questions About Solar Panel Waste Recycling
How much solar panel waste will be generated worldwide by 2060?
According to the model published in Nature, 297 to 402 million tonnes of solar panel waste will accumulate by 2060. Annual volumes rise from 0.02 million tonnes in 2020 to 0.5 million tonnes by 2030, and to more than 19 million tonnes per year by 2060. Cumulative volumes grow roughly 150- to 200-fold between 2030 and 2060.
When will solar panel recycling become profitable?
Between 2035 and 2040 in every scenario examined. Revenue per tonne currently stands at 131.8 to 304.3 dollars, against costs of 342 to 2,550 dollars. Rising prices for recovered silver, aluminium and silicon lift unit revenue to 457.1 dollars, while learning effects and larger plant capacity cut costs to 55.9–402.7 dollars.
Which country will generate the most solar panel waste?
Until 2040, high-income regions still account for more than half of global volumes, with the EU-15 as the largest single contributor. By 2060, China takes the lead with 112.8 to 160.5 million tonnes, or 36.2–39.9 per cent of the global total. The EU-15 then stands at 30.9–39.7 and the United States at 25.7–39.0 million tonnes.
Which recycling technology is the most effective?
Thermal processes deliver the highest economic and climate benefit per unit, while mechanical processing delivers the least, because its metal recovery efficiency is lower. On the carbon-priority pathway, the share of thermal processing rises from 10.1 per cent to 44.0–51.6 per cent by 2060, producing more than a 60 per cent gain over the BAU pathway.
Why does carbon-price-linked subsidy not work?
Because more than 85 per cent of high-carbon-price schemes increase inter-regional inequality. High-income regions receive up to 200 dollars per tonne, nearly four times what developing economies get, many of which have no established carbon market. The gap widens to 1,198–1,434 dollars per tonne, worse than the no-subsidy case.
Why can reuse not replace recycling?
The authors note that refurbishment can in principle deliver greater environmental and resource-efficiency benefits, and decouples the circular benefit from raw material prices. Large-scale reuse is nonetheless constrained by variation in module condition, latent reliability risks and limited market confidence in second-hand products, so it remains a transitional, complementary strategy.
Source:
Wang, C., Zuo, J., Chen, X. et al. Towards an equitable future of global photovoltaic waste recycling. Nature (2026). Published 12 August 2026. DOI: 10.1038/s41586-026-10905-w. Open access publication.
