Retired wind turbine blades are among the large industrial wastes that are hardest to recycle. Many blades are made of glass fiber held together by a thermosetting resin. The material is lightweight and strong, but the resin forms a rigid cross-linked structure, so it does not melt when heated and cannot be reshaped like ordinary plastic. Blades have traditionally been landfilled, or processed by pyrolysis or chemical treatment to recover the fibers. Those routes face problems with the energy they require and the degraded quality of the recovered material.
A research team from Tsinghua University, Peking University, and the Beijing Graphene Institute in China has developed a way to turn scrap blades into a different high-performance material. In the study by Bi et al., "Upcycling of waste wind turbine blades by electrothermal reduction toward circular resin composite applications", published in the journal Chem Circularity, crushed blade waste is heated almost instantly to very high temperatures by "flash Joule heating." The silica in the glass fiber and the carbon derived from the resin are converted into a composite powder of silicon carbide and graphene. Rather than regenerating the original glass fiber, the method converts the waste into a new material that can reinforce resin.
Thermosetting resin cannot be melted, and conventional recycling has limits
Wind turbine blades are designed to withstand wind pressure and repeated loading over many years. Their main material, glass-fiber-reinforced plastic (GFRP), is a composite of high-strength glass fiber bound with a thermosetting polymer such as epoxy resin. A three-dimensional cross-linked structure forms inside the resin. It gives the material excellent mechanical strength and durability, but also makes it hard to break down or reprocess after use, because, unlike thermoplastics, it cannot be melted and reshaped.
Conventional recycling technologies include pyrolysis and chemical dissolution. In pyrolysis, the material is heated to roughly 400–600°C in an oxygen-free or low-oxygen environment, breaking the resin down into gas and oil-like components so the glass fiber can be recovered. However, prolonged external heating consumes a great deal of energy, and the heat treatment damages the fiber surface and lowers its tensile strength. As a result, the recovered fiber is hard to use in structural materials equivalent to the original blade and is often used as relatively low-cost short-fiber reinforcement or filler.
Chemical dissolution uses strong acids, hydrogen peroxide, or high-temperature, high-pressure organic solvents to cut the chemical bonds of the resin. Depending on conditions, fiber damage can be limited, but solvent recovery and reuse and waste liquid treatment bring cost and environmental burdens. Both approaches aim to separate the tightly bonded composite back into its original components, which requires substantial energy and processing equipment.
Bi et al. depart from the idea of extracting the original fiber and resin materials as they were. The blade waste is crushed, and the silicon, oxygen, and carbon it contains are reacted at high temperature to convert them into different materials: silicon carbide and graphene. The basic idea is not simply to recover waste but to remake it into a high-performance material for different uses.
About 2,000°C from electrical heating produces silicon carbide and graphene
At the heart of the technology is flash Joule heating, which uses Joule heat from an electric current to heat a sample rapidly. Because glass fiber conducts electricity poorly, the team mixed conductive carbon black into the crushed turbine blade waste. The powder is packed into a reaction cell made of quartz tubing or similar, and pulsed direct current is passed through it via electrodes, causing the sample itself to generate heat.
When current is applied, the sample reaches about 2,000°C in roughly 0.5 seconds. At this temperature, carbothermal reduction, in which silicon dioxide (SiO₂), the main component of glass fiber, reacts with carbon, produces silicon carbide (SiC). At the same time, carbon from the thermal decomposition of the epoxy resin and the mixed-in carbon material rearranges into thin-layer graphene. The result is a composite powder containing silicon carbide and graphene.
The process is a dry one that needs no solvents or precious-metal catalysts, and the heating time is extremely short. Instead of heating an entire reaction vessel for a long time as in a conventional furnace, current is passed through the conductive sample itself, generating heat directly inside it. This delivers the required high temperature in a short time. The rapid heating and cooling also play a role in forming the structure of the carbon material and ceramic particles, enabling material synthesis different from long-duration heating.
The composition and recovery rate of the powder deserve caution, however. According to a related technical report, the solid recovery rate under representative experimental conditions was about 55.6% of the input mass. Much of the remaining mass is thought to have been lost as volatile components such as carbon monoxide and low-molecular-weight hydrocarbons produced by thermal decomposition of the resin. The recovered solid was reported to be about 33.6% silicon carbide and about 66.4% flash graphene by mass.
Importantly, these are not separately purified, high-purity silicon carbide and graphene. The product is a composite powder in which silicon carbide particles are dispersed within a graphene network. The team indicated that the composition may be controllable by adjusting current conditions and raw material mixing ratios, and also examined reusing the recovered graphene-containing powder as a conductive additive in the next processing run to reduce the carbon black added from outside. These results, however, concern production and reuse of the mixed powder; they do not show that high-purity silicon carbide and graphene can be separately recovered.
A small addition raises resin stiffness by 27% and thermal diffusivity by 41%
The team investigated whether the composite powder could function as a reinforcing agent in new resin materials. They prepared test specimens by adding small amounts of the recovered powder to epoxy resin and evaluated tensile and flexural properties against pure epoxy without additives.
Specimens with 0.3 wt% (0.3% by mass) of recovered powder showed a Young's modulus (tensile modulus) about 27% higher than pure epoxy. Young's modulus indicates how resistant a material is to deformation under force; a higher value means greater stiffness. The improvement in tensile strength, the stress a material can withstand before breaking, was about 11%. Flexural tests also showed gains, with flexural modulus up about 9% and flexural strength up about 6%.
These results suggest that even an addition of just 0.3 wt% may improve the mechanical properties of the resin. But higher stiffness does not translate into an equivalent gain in breaking strength: in this test, Young's modulus rose 27% while tensile strength rose only 11%. Also, although the available published materials give improvement rates, they do not list absolute values, such as MPa or GPa, for the Young's modulus or tensile strength of the pure epoxy and the additive specimens.
For comparison, the team also prepared resin specimens containing the same 0.3 wt% of waste blade powder that had not been electrically heated. Resin with the untreated powder had lower mechanical strength than pure epoxy. Merely crushed glass fiber and resin fragments probably did not adhere well to the surrounding epoxy and caused stress concentration under load. By contrast, the powder converted into silicon carbide and graphene by electrical heating worked effectively as a reinforcement.
データを表で見る
| Change vs. pure epoxy (%) | |
|---|---|
| Young's modulus (0.3 wt%) | 27 |
| Tensile strength (0.3 wt%) | 11 |
| Flexural modulus (0.3 wt%) | 9 |
| Flexural strength (0.3 wt%) | 6 |
| Thermal diffusivity (0.1 wt%) | 41 |
| Thermal conductivity (0.1 wt%) | 1.4 |
The evaluation of thermal properties gave an interesting result that differed from the mechanical one. Specimens with 0.1 wt% of recovered powder showed thermal diffusivity about 41% higher than pure epoxy, while thermal conductivity rose only about 1.4%. In other words, thermal diffusivity changed greatly, but the ability to transmit heat itself did not improve at the same rate. For these thermal properties, too, the available published materials show only rates of change, not absolute values of thermal conductivity (W/m·K) or thermal diffusivity (mm²/s).
Understanding this difference requires looking at the relationship between thermal diffusivity and thermal conductivity. Thermal diffusivity () is defined as thermal conductivity () divided by the product of density () and specific heat capacity ().
Thermal conductivity expresses how readily heat travels through a material, while thermal diffusivity expresses how quickly temperature changes propagate through it. So even if thermal conductivity barely changes, thermal diffusivity rises if the product of density and specific heat capacity falls. The team interprets the 41% rise mainly as the additive causing an apparent decrease in the composite's specific heat capacity.
This result cannot be read as a 41% improvement in the heat dissipation of wind turbine blades. Thermal diffusivity and thermal conductivity are different properties, and heat transfer in real structures also depends on shape, thickness, and surroundings. The evaluation also used small resin specimens made in the lab, not actual turbine blades. Performance that matters in practice, such as resistance to long-term cyclic loading, surface erosion from wind and rain, moisture ingress, and degradation from temperature changes, will need to be verified separately.
The main results and their scope are summarized below.
| Item | Condition / comparison | Reported result | Basis | Main limitation |
|---|---|---|---|---|
| Tensile stiffness | 0.3 wt% added to epoxy | Young's modulus up about 27% | Measured on small specimens | Comparison with pure epoxy; absolute values not found in available published materials |
| Tensile strength | 0.3 wt% added to epoxy | Up about 11% | Measured on small specimens | Smaller gain than stiffness; performance in real structures untested |
| Flexural properties | 0.3 wt% added to epoxy | Flexural modulus up about 9%, flexural strength up about 6% | Measured on small specimens | Dispersion and performance at larger additions untested |
| Thermal diffusivity | 0.1 wt% added to epoxy | Up about 41% | Measured on small specimens | Interpreted as mainly due to specific heat change, not improved thermal conductivity |
| Thermal conductivity | 0.1 wt% added to epoxy | Up about 1.4% | Measured on small specimens | Limited gain compared with thermal diffusivity |
| Untreated powder addition | 0.3 wt% added to epoxy | Mechanical strength decreased | Control experiment | Merely crushed waste did not act as an effective reinforcement |
| Greenhouse gas emissions | Compared with pyrolysis / chemical dissolution | About 83% / 92% reduction | LCA model estimate | Not measured values from a commercial plant |
| Processing profitability | Per ton of waste blades | About $80 profit | Techno-economic assessment model | Depends on assumptions such as the sale price of the recovered powder |
From 0.3 g to 200 g: challenges remain in scaling up
There is still a large gap between synthesizing materials in the lab and a commercial facility that processes industrial waste in bulk. Most tests in this study used small batches of about 0.3 g of raw material. The team also ran tests at the 1.2 g and 10 g scales, but processing actual turbine blades in volume would require far greater scale-up.
Blades on large wind turbines can be tens of meters long and weigh from several tons to tens of tons. Batch processing in small reactors alone cannot handle waste generated at tens of thousands of tons a year. Equipment that can heat large quantities of powder at once, or feed and process raw material continuously, will be essential.
The team also reports having designed and built a 200 g-scale electrothermal reactor for scale-up. But building a reactor is a separate matter from sustaining large-volume processing at stable quality. A particularly important question is whether current can flow uniformly through the entire powder sample and heat it to the required temperature.
With small samples, the distance between electrodes is short, and the electrical resistance and packing state of the whole powder are relatively easy to manage. In a larger reactor, however, differences in powder density and contact can create regions where current flows easily and others where it does not. That can produce hot spots with locally high temperatures and cold spots that are not heated enough. In cooler regions, silica reduction is insufficient and unreacted glass remains, while overheated regions may see material evaporation or unintended reactions.
Such temperature variation can change the ratio of silicon carbide to graphene and the particle structure in the recovered powder, ultimately affecting resin reinforcement performance. It is therefore important to verify heating uniformity and product quality at the 200 g scale, and to see whether the same performance can be reproduced at even larger throughput. Electrode durability, powder packing methods, and exhaust gas treatment must also be considered in designing commercial equipment.
For practical use as a material, long-term fatigue tests and environmental durability tests are needed beyond static tensile and flexural tests on resin specimens. Turbine blades are subject to fluctuating wind pressure, vibration, temperature changes, and humidity. Whether the gains in stiffness and strength shown here persist under such conditions is not yet known.
The team suggests the technology could also be applied to composite waste from aerospace, automotive, electronics, and construction materials as well as wind blades. But different resins and fibers contain different elements and have different thermal decomposition behavior and required reaction conditions. The conversion efficiency and product characteristics confirmed here cannot simply be applied to other composites.
Up to 92% lower greenhouse gas emissions, but does the profitability estimate hold?
In addition to the synthesis experiments, the team carried out a life cycle assessment (LCA) and techno-economic assessment (TEA) assuming future industrial use. LCA evaluates environmental burden from raw material procurement through processing, while TEA estimates a technology's economics by considering equipment and operating costs and revenue from products.
The LCA model estimated that the electrothermal process could cut greenhouse gas emissions by about 83% compared with conventional pyrolysis and about 92% compared with chemical dissolution. Behind this is the ability of electrical heating to reach the required high temperature in an extremely short time. Conventional pyrolysis requires heating a furnace for long periods, and chemical dissolution consumes energy maintaining high temperature and pressure and recovering solvents. By contrast, this method can reach about 2,000°C with roughly 0.5 seconds of current, so the model rates its processing energy demand as lower.
However, the 83% and 92% reductions are not emission cuts measured at a commercial facility. They are model estimates based on assumed comparison processes and power supply conditions, and they change with those assumptions. The available published materials do not list absolute greenhouse gas emissions for each process in CO₂-equivalent mass. A real assessment of environmental burden would need to include not only power consumption during heating but also collection and transport of waste blades, crushing, production of the conductive additive, exhaust gas treatment, and equipment operation.
On economics, the team also indicates the method may be more advantageous than conventional technologies. In the TEA model, processing one ton of waste blades incurs a net cost of about $460 with pyrolysis and about $1,800 with chemical dissolution, whereas the electrothermal method yields a profit of about $80. Unlike conventional approaches that bear the cost of disposing of waste, the idea is that if the composite powder can be sold as a high-value resin additive, the processing itself can become a source of revenue.
That profitability, though, depends heavily on the selling price of the recovered material. The product is a mixed powder of about 33.6 wt% silicon carbide and about 66.4 wt% graphene, which differs in quality and application from high-purity silicon carbide for semiconductors or separately purified graphene. Market prices for typical high-performance materials therefore cannot necessarily be applied as they are. How much actual resin manufacturers would pay for this mixed powder will be a key factor in commercialization.
Costs for crushing and pretreating the raw material, procuring carbon black, replacing electrodes, and treating exhaust gas are also not negligible. At commercial scale, equipment for controlling current distribution and temperature could be more complex than in the lab. The short heating time alone does not justify concluding that overall energy consumption and operating costs will be low. The LCA and TEA results should be regarded as an early assessment of the conditions under which the technology could work.
What this study shows is the possibility of converting waste from hard-to-recycle wind turbine blades into a new functional material without separating it into its original constituents. Improvements in stiffness and strength were also confirmed in small resin specimens, pointing to an upcycling path different from conventional fiber recovery. Challenges remain for practical use, however: heating uniformity in large-volume processing, stability of product quality, long-term durability, and market selling price. The focus going forward will be whether the 200 g reactor can demonstrate stable processing performance, and whether the environmental and economic advantages hold up when scaled to commercial size.