Solutions for recycling emerging wind turbine blade waste in China are not yet effective

The study addresses the emerging challenge of managing large volumes of composite wind turbine blade waste in China as wind capacity expands to meet carbon-neutrality targets. While metals in turbines are recyclable, blades made from glass-fibre reinforced polymers (GFRP) are bulky and costly to recycle, creating a potential waste burden. Conventional disposal (landfill, incineration) faces environmental drawbacks and regulatory limits. Existing global and national estimates of blade waste lack resolution, often ignore turbine size evolution, and underrepresent spatial variability and the dynamic energy context affecting treatment impacts. The research aims to quantify, at high spatial and temporal resolution, the generation of blade waste across manufacturing, operation and maintenance, and end-of-life stages, and to evaluate environmental and economic implications of current disposal and recycling options in China under evolving electricity mixes, thereby informing policies and strategies toward carbon neutrality.

Prior studies have provided coarse estimates of future blade waste at global and national scales but generally did not incorporate periodic increases in turbine capacity, lacked high-resolution inventory models for waste management strategies, and omitted location-specific dynamics of energy systems. Some works considered limited treatment options (e.g., landfill, cement co-processing) or static life-cycle impacts. Studies on composite recycling processes (mechanical, pyrolysis, fluidised bed, chemical) reported varying energy use and fibre quality outcomes, but translation to China’s provincial contexts and dynamic grid decarbonisation was underexplored. This study addresses these gaps by combining dynamic material flow analysis with provincial-level life cycle and cost assessments across multiple treatment routes.

The authors develop a multi-dimensional dynamic material flow analysis (DMFA) integrated with process-level life cycle assessment (LCA) and cost modelling for China (1989–2018 historical, projections to 2050). Key elements: (1) Wind power development scenarios: Base, minimum (IEA New Policies, 635 GW by 2050), and maximum (CNREC Active 2050 Advanced, 1600 GW by 2050), with base around 1000 GW; replacements maintain steady-state capacity. Offshore wind excluded. (2) Turbine size evolution: Based on historical market shares (2005–2018), average turbine size increases by ~0.5 MW every 6–8 years, dominant size rising from ~2 MW to 5.5 MW by 2050; 20-year lifetime assumed. (3) Blade inventory: High-resolution database with 104 models across 14 capacities (150–5500 kW); blade mass correlated with turbine size; assume 100% onshore blades are GFRP. (4) Waste generation: Quantifies manufacturing (MAN), operation & maintenance (O&M), and end-of-life (EOL) waste using size-specific ratios (from Supplementary data) and provincial distribution of 58 blade factories; estimates at national, regional, and provincial scales over time. (5) LCA of treatment routes: Functional unit = 1 kg of blade waste. Metrics: Primary energy demand (PED) and GHG emissions (IPCC 100-year CO2eq). Process models include handling, transport, processing, and credits from energy and material recovery. Seven options assessed: landfill; municipal incineration; cement kiln co-processing; mechanical recycling (with landfill or with incineration of residuals); pyrolysis; fluidised bed; chemical recycling. Recovered GF displaces virgin GF based on retained properties; displacement factor sensitivity assessed. (6) Dynamic background: Provincial and national electricity mixes evolve (2018–2050), affecting energy and GHG intensities of processes and credits. (7) Cost model: Net EOL cost includes dismantling, handling/shredding, transport, recycling/disposal, residuals disposal, and credits from recovered products; 2018 US$ with 2% inflation used; data from industry, literature, experiments, process models. (8) Integration: DMFA outputs feed scaled, time-varying LCA and cost assessments to 2050, with results presented at national, regional, and provincial levels.

• National waste quantities: In the base case, cumulative blade waste reaches 12.9 Mt (2018–2050), comprising 3.9 Mt MAN, 7.8 Mt EOL, and 1.1 Mt O&M. Scenario range: 7.7–23.1 Mt by 2050. 2018 waste was ~507 kt; a ~20-fold increase by 2050 (base). - Temporal dynamics: Before 2025, MAN waste dominates (50–60 kt/y). From 2025–2040, total waste rises from 88 to 557 kt/y, peaking at ~618 kt/y in 2035, with EOL growing from 5% (2025) to 67% (2040). From 2040–2050, MAN stabilises while EOL continues to grow. - Spatial patterns: Manufacturing facilities concentrate in Jiangsu (11), Inner Mongolia (8), Hebei (7), shaping MAN waste hotspots. 2018 top MAN waste provinces: Jiangsu (10.6 kt), Inner Mongolia (9.2 kt), Hebei (7.0 kt), Shanghai (4.7 kt), Gansu (4.7 kt). By 2040, top annual waste provinces: Hebei (53.6 kt), Jiangsu (51.5 kt), Inner Mongolia (50.4 kt), Shandong (29.1 kt), Xinjiang (26.1 kt). In 2050, top five (Inner Mongolia, Hebei, Xinjiang, Jiangsu, Gansu) generate ~50% of national waste; EOL dominates most provinces. - Regional growth modes: Seven regions exhibit exponential, logarithmic, Kuznets, or slow growth patterns in waste arisings due to differences in resource, deployment, and policy. - Environmental impacts per unit: Recycling generally reduces GHG vs landfill/incineration, except pyrolysis (highest per-unit GHG). Mechanical recycling offers the largest GHG reduction; fluidised bed yields negative net emissions (credits exceed burdens). Cement kiln co-processing appears near net-zero GHG under assumptions (displacing coal), but still emits fossil carbon unless CCS/bio-based polymers are used. - Scaled and cumulative impacts: Annual scaled PED and GHG increase sharply after 2025 with growing waste. Pyrolysis PED reaches ~14.68 PJ/year by 2050; cement kiln co-processing achieves net PED reduction of ~−9.35 PJ/year by 2050. Fluidised bed annual GHG remains negative and becomes more negative over time despite grid decarbonisation impacts (e.g., from about −0.04 to −0.39 MtCO2eq/year). From 2035 to 2045, grid decarbonisation reduces unit intensities (pyrolysis −13%, chemical −188% noted for intensity changes) but total pyrolysis scaled GHG rises ~6.3× (0.20 to 1.28 MtCO2eq/yr). Cumulative GHG to 2050 ranges from +25.3 MtCO2eq (worst) to −8.7 MtCO2eq (best). - Provincial variation in impacts: Top 10 provinces (Jiangsu, Inner Mongolia, Hebei, Gansu, Shanghai, Tianjin, Liaoning, Jilin, Xinjiang, Guangdong) account for 79–93% of GHG impacts in 2018 across options, declining to ~58–59% in 2050 as waste distribution evens. Provincial electricity grid intensities alter rankings between waste quantities and environmental impacts. - Costs: Landfill is low per-unit cost but increasingly restricted (ban on composite landfill in China). Mechanical recycling has the lowest cost among advanced options and can generate net revenue; estimated cumulative income US$1.86–1.91 billion by 2050. Fluidised bed annual cost ~US$176 million/year by 2050; cumulative ~US$2.19 billion. Municipal incineration annual cost grows from US$8.5 million/year (2018) to US$226.3 million/year (2050). Pyrolysis is most expensive: annual cost ~US$320.6 million/year by 2050; cumulative ~US$4.0 billion. Cement kiln co-processing cumulative cost ~US$4.1 billion. Landfill cumulative cost ~US$3.2 billion (but restricted). - Displacement factors: Substituting recycled GF for virgin GF reduces impacts and costs. For PED, net-zero achievable at ~40% displacement (fluidised bed) and ~65% (chemical). Chemical can achieve net GHG reduction at high displacement factors (>70%). Mechanical+landfill shows net PED and GHG reductions at very low displacement (~8%), indicating strong potential when high-quality fibre recovery is achieved. - GF sector relevance: China’s GF production emits ~3.04 tCO2eq/t GF. Using recycled GF from cumulative blade waste (~60% of 12.9 Mt) could save ~23.5 MtCO2eq by 2050.

The findings show China will face rapidly growing volumes of GFRP blade waste, with EOL streams dominating after 2030, concentrated in specific provinces. Recycling pathways can significantly reduce energy use and GHG emissions compared to conventional disposal, but benefits depend on process energy intensity, product quality, and the evolving electricity mix. Mechanical recycling and fluidised bed processes offer the most attractive environmental profiles; cement kiln co-processing can approach net zero GHG only under specific assumptions, while pyrolysis is environmentally and economically unfavourable under current conditions. Provincial electricity carbon intensities alter the relative benefits and should inform region-specific strategies. Policies such as extended producer responsibility, targeted R&D and pilots in top-impact provinces, industrial symbiosis in manufacturing hubs, and efforts to expand markets for recycled GF are key to scaling recycling. The current landfill ban on composite residuals limits the viability of certain low-impact routes (mechanical+landfill), indicating policy review may be needed to enable best overall outcomes while ensuring stringent emissions control. Overall, integrating circular strategies across the value chain—from design and manufacturing waste reduction to reuse/repurposing and advanced recycling—will be necessary to manage blade waste sustainably.

China is projected to generate 7.7–23.1 Mt of wind turbine blade waste by 2050 (12.9 Mt base case), with EOL waste becoming dominant. High-resolution modelling identifies temporal and provincial hotspots, enabling strategic planning for recycling capacity. Life cycle and cost analyses indicate no single universally superior solution, but mechanical recycling and fluidised bed offer the most promising combinations of environmental performance and cost; cement co-processing can reduce PED and approach net-zero GHG under assumptions, while pyrolysis remains both impact- and cost-intensive. Policy support, producer responsibility, and market development for recycled GF are crucial. The modelling framework provides a transferable basis for other countries. Future research should refine cost models with region-specific data, evaluate uncertainty in lifetimes and offshore blade waste, improve recycling process efficiency and product quality, and design tailored provincial strategies aligned with grid decarbonisation trajectories.

• Offshore wind turbines and their waste streams are excluded; offshore lifetimes and logistics may differ. - A fixed 20-year turbine lifetime is assumed; lifetime uncertainty and extensions are not modelled. - Onshore blades are assumed 100% GFRP; carbon fibre content is neglected. - Some cost data use generic or best-available estimates; detailed regional cost factors (e.g., wages, logistics, facility costs) are simplified or excluded in summary tables. - The analysis assumes dynamic but scenario-based electricity mix decarbonisation; deviations would alter results. - Pyrolysis modelling assumes no condensate recovery of chemical vapours, likely overestimating GHG impacts for future optimised systems. - Landfill policy constraints in China restrict some otherwise low-impact routes (mechanical+landfill), affecting feasibility. - Environmental credits rely on displacement assumptions for recycled GF and resin; real-world displacement factors and product performance may vary.