A comparative analysis of carbon reduction potential for directly driven permanent magnet and doubly fed asynchronous wind turbines

Wind power generation does not emit greenhouse gases or pollutants, but there are some carbon emissions from the manufacturing, transportation, operation, and waste disposal of wind turbines. Directly driven permanent magnet and doubly fed asynchronous wind turbines currently have the largest market share in China, but few Chinese studies have compared their differences in carbon reduction potential. This paper uses life cycle assessment (LCA) to quantitatively analyze the full life cycle carbon emissions of the two wind turbines to determine which type of wind turbine has greater carbon reduction potential, obtaining the following results. (1) The full life cycle greenhouse gas emissions of 2.5 MW directly driven permanent magnet and doubly fed asynchronous wind turbines are 8.48 and 10.43 g CO2/kWh, respectively. The direct‐driven permanent magnet wind turbine is superior in terms of carbon reduction. (2) The stage with the greatest impact and the greatest difference between the two wind turbines in the full life cycle is the production stage, during which the carbon emissions of the directly driven permanent magnet and doubly fed asynchronous wind turbines are 1.045 × 106 and 1.210 × 106 kg, respectively. (3) According to sensitivity analysis, proper waste disposal and transportation can reduce carbon emissions from wind turbines. These research findings can be used to help achieve carbon peaking and neutrality goals, as well as the technological development of wind power enterprises.


| INTRODUCTION
The consumption of fossil fuels is increasing as industrialization progresses, resulting in more waste and pollutants in the natural environment, which significantly impacts the environment and human health. Wind power, which is green and low carbon, is an important way to address environmental and energy issues. 1 The cumulative installed capacity of wind power in China was 328 million kW at the end of 2021. The national wind power generation capacity was 652.6 billion kWh, a 40.5% increase year on year. Wind and photovoltaic power generation capacity in China represented approximately 11% of the total society's electricity consumption. When carbon peaks in 2030, wind power installed capacity is expected to reach 800 million kWh, accounting for 15% of total power generation. When carbon neutrality is expected in 2060, wind power installed capacity may exceed 2 billion kWh, and the wind power generation proportion will be greater than 30%. 2 Therefore, the growth of the wind power industry is critical to achieving the goals of carbon peaking and neutrality.
The most common wind turbine technologies in China are doubly fed asynchronous and directly driven permanent magnets. It can be seen from Figures 1 and 2 that the difference between the two types of wind turbines is that the wind wheel shaft of the doubly fed wind turbine is connected to the generator rotor after passing through the gearbox, rather than directly driven. The rotor shaft of a direct-drive wind turbine is directly connected to the generator rotor, and the gearbox is omitted. In 2020, the doubly fed asynchronous type accounted for 60.9% of newly installed offshore wind turbine capacity, while the direct-driven permanent magnet type accounted for 30.5%. 5 However, it is unknown which of the two types has a greater potential for carbon reduction and whether a greater carbon reduction effect can be achieved by increasing the proportion of a specific category. Therefore, analyzing and comparing the carbon reduction potentials of the two types of wind turbines is critical.
Although wind energy does not emit greenhouse gases, materials and energy are required throughout the life cycle of wind turbines, so some carbon emissions are unavoidable. 6 Its carbon emissions over its entire life cycle can be quantified using life cycle assessment (LCA). LCA is a valuable environmental management tool that can be used to assess carbon emissions and cumulative energy demands of products or services over their entire life cycle. 7,8 Most previous research on wind power using LCA focused on a single wind farm. For example, Al-Behadili and El-Osta 9 investigated the full life cycle and payback time of energy investment in Libyan wind farms; Ardente et al. 10 investigated the energy performance and conducted the LCA of an Italian wind farm with a 20-year time scale, and targeting a wind farm in Shenyang City, China, Gao et al. 11 calculated and compared the carbon emissions of wind turbine production, transportation, operation, and waste emission and treatment by LCA to those of coal power generation, concluding that wind power generation had significant energy-saving and environmental benefits. Other studies concentrated on wind farms that used both sea and land systems. Bonou et al. 12 conducted a life cycle analysis of wind farms on land and at sea. Xiang et al. used LCA to compare offshore and onshore wind power systems and analyzed the carbon emissions of wind farms equipped with various power turbines, finding that offshore wind farms with higher power provided greater carbon reduction benefits.
The difference in carbon reduction potential between directly driven and doubly fed wind turbines is rarely investigated in China. Most research is based on life cycle background databases from other countries, which do not accurately reflect Chinese reality.

| Research object and method
Until 2021, the wind turbine with the most installed capacity among onshore wind farms in China was the model with a single unit capacity of 2.5 MW, accounting for 40% of the onshore wind farm installed capacity. Therefore, the 2.5 MW directly driven permanent magnet wind turbine and 2.5 MW doubly fed asynchronous wind turbine are chosen as research objects in this paper. The full life cycle carbon footprints and cumulative energy demands of two types of wind turbines are compared using eFootprint, China's first LCA software with independent intellectual property rights, and the domestic life cycle background database to determine the type with greater carbon reduction potential.

| System boundary
A wind turbine's system boundary includes four stages: production, transportation, operation, and waste disposal. For more information, see Figure 3. The main purpose of this study is to analyze the differences in the carbon footprint of two types of wind turbines. During installation and commissioning, the carbon footprint of the two types of wind turbines is almost the same. According to the calculation and review of relevant studies, although this process will also emit a lot of greenhouse gases, its carbon footprint accounts for less than 1% of the total carbon footprint, so this process will not be separately listed in the 2.4 Inventory analysis.

| Model assumptions
Three assumptions are made in this paper. (1) The tower is 80 m tall, and the ground is level. (2) Wind turbines have a service life of 20 years. 13,14 (3) The wind turbine is located in a wind power plant in the Chinese province of Guangdong, and its annual power generation time is 2630 h. 15

| Inventory analysis
The wind power foundation, tower, blade, hub, nacelle cover, nacelle chassis, transmission mechanism, generator, anemometry system, and electronic control system are the main components of the doubly fed asynchronous wind turbine. In contrast to the doubly fed type, the low-speed wind wheel of a directly driven wind turbine is directly connected to the generator, removing the need for a complex transmission mechanism. The anemometry and electronic control systems are essentially electronic equipment, with volumes and masses that are less than 5% of the total unit. They have many parts, and their data are difficult to obtain, so they are not discussed in this paper. 16,17 The materials and energy consumption of the components in the 2.5 MW doubly fed asynchronous and directly driven permanent magnet wind turbines are listed in Table 1. The majority of the data comes from wind turbine manufacturers' product manuals.
T A B L E 1 Material consumption of wind turbine components at the production stage.

| Stage of transportation
Because wind energy resources are often found in remote areas, the carbon footprint in the transportation stage is primarily due to the consumption of fossil fuels by transportation vehicles. This paper assumes that transportation is provided by 1.000 (10 4 kg) gasoline-powered trucks. The trucking life cycle history is derived from the database of the LCA software eFootprint. It is assumed that the truck transportation distance to deliver the wind turbine components to the wind farm is 1000 km, and the concrete required by the wind power foundation is 50 km. 18 manufacturing stage, compensating for some materials' carbon emissions, the total carbon emission in this stage (−1.742 × 10 5 kg) is negative, accounting for -15.6% of the total. The carbon emission of the doubly fed asynchronous wind turbine is 1.371 × 10 6 kg and the carbon emission in the manufacturing stage is also the highest (1.298 × 10 6 kg), accounting for 94.7% of the total. The carbon emissions from the operation, transportation, and waste disposal stages are 2.334 × 10 5 , 7.290 × 10 4 , and -2.336 × 10 5 kg, respectively, accounting for 17.0%, 5.3%, and -17.0%. The carbon footprint of a directly driven permanent magnet wind turbine is 81.4% times that of a doubly fed asynchronous wind turbine.
Therefore, the carbon emissions of direct-driven permanent magnet and doubly fed asynchronous wind turbines are 8.48 and 10.43 g CO 2 /kWh, respectively, according to Equation (1 where b is the carbon emission per kWh (g CO 2 /kWh); B is the total carbon emission in the entire life cycle (g CO 2 ), and Q is the annual average power generation capacity (kWh/year). The production stage is when the carbon emissions of the two wind turbines are highest, and their difference is greatest. Figure 5 depicts the carbon emission ratio of each component. The main structural difference between the two is in the generator and transmission mechanism. Because of its low revolving speed, the directly driven permanent magnet generator requires more magnetic poles (typically above level 90), resulting in a larger volume and weight than the doubly fed asynchronous generator. The generator mass in the 2.5 MW directly driven permanent magnet wind turbine is approximately 6.500 × 10 4 kg, whereas the doubly fed asynchronous generator mass is only approximately 1.200 × 10 4 kg. 28 The transmission mechanism of a directly driven wind turbine is simplified because its wind axle is directly connected to the generator rotor, eliminating the speed-up gearbox, and greatly reducing transmission mechanism mass.
The nacelle mass of doubly fed asynchronous wind turbines is generally greater than that of directly driven permanent magnet wind turbines. The nacelle mass of the 2.5-MW doubly fed asynchronous wind turbine (including the impeller and generator) is approximately 1.530 × 10 5 kg, whereas it is approximately 1.320 × 10 5 kg in the directly driven permanent magnet type. A larger nacelle mass necessitates increasing the tower and wind power foundation mass (assuming a flat terrain). Figures 5 and 6 show that the tower is responsible for most of the carbon footprint at this point. Furthermore, the carbon footprint of the tower in Figures 5 and 6 stems solely from its manufacturing process and does not include any other carbon footprint components.
To summarize, the materials and energy consumption of the doubly fed asynchronous wind turbine are greater than those of the directly driven permanent magnet type during the manufacturing stage, and the carbon emission of the former is 124.3% of that of the latter.
The mode of transportation determines the amount of carbon emitted during transportation: the distance traveled and the mass of the goods. A truck with a load of 1.000 × 10 4 kg transporting 1.000 × 10 3 kg of goods for 1 km emits 0.140 kg of CO 2 . Because the two wind turbines' transportation modes and distance are assumed to be the same, carbon emissions are solely determined by the mass of goods. At this point, the carbon emissions  of the doubly fed wind turbine are 7.290 × 10 4 kg, which is 1.390 × 10 4 kg higher than that of the directly driven wind turbine.
Carbon emissions in the operating stage are caused by daily maintenance and component replacement. Routine inspection and replacement of lubricating oil are part of daily maintenance. Because the doubly fed type has a gearbox, its carbon emissions are slightly higher in this process than the directly driven type. The failure and replacement rates of each component of the wind turbine are related to component replacement. The speed-up gearbox is eliminated in the directly driven wind turbine, lowering the overall failure rate, and the overall mass is lower than in the doubly fed type. Therefore, the carbon emission from directly driven wind turbine component replacement is only 79.7% of that of the doubly fed asynchronous type.
According to the assumptions of this paper, metals are recycled in the waste disposal stage, while other materials are treated as municipal solid waste to be discharged. Therefore, the doubly fed asynchronous wind turbine consumes more metals and can offset 17% of its carbon emissions, whereas the directly driven type can only offset 15.6%.

| Payback period for energy
The energy payback time is the number of operating years required for a wind power system to recover its primary energy consumption over its life cycle. The sum of energy required in the wind turbine production, transportation, operation, and waste disposal stages is typically referred to as consumption. The relationship between total energy consumption and annual system power generation can intuitively reflect the return on investment of unit energy. 29 The annual operation time of the wind turbines studied in this paper is 2630 h, resulting in an annual average power generation capacity of 6.575 × 10 6 kWh, equivalent to 23,670 GJ. Throughout the life cycle, the cumulative energy demands of 2.5-MW doubly fed asynchronous and directly driven permanent magnet wind turbines are 1.890 × 10 7 and 1.560 × 10 7 MJ, respectively. Therefore, the energy payback times for the 2.5 MW doubly fed asynchronous and directly driven permanent magnet wind turbines are 0.80 and 0.66 years, respectively, according to Equation (2) where CED is the cumulative energy demand (MJ); Q is the annual average power generation capacity (MJ/years); EPT is the energy payback time (years).

| Sensitivity analysis
The wind turbine's waste disposal stage is critical in determining its carbon footprint throughout its life cycle. Therefore, the treatment of waste materials directly impacts the environmental effects produced at this stage. Sensitivity analysis is performed for the wind turbine transport process, waste disposal treatment methods, and metal recovery rate. 30

| Waste disposal method
Wind turbine waste materials are treated in three ways: recycling, landfill, and incineration. 31,32 Metals (steel, copper, iron) are recycled, while nonmetallics (glass fiber, epoxy resin, polyester resin, acetone) are landfilled or incinerated. Keeping all other variables constant and assuming that each method's utilization rate is 100%. The carbon emissions generated by the full landfill of nonmetallic materials for the doubly fed asynchronous wind turbine will be 4.430 × 10 3 kg and the carbon emissions generated by the full incineration will be 7.415 × 10 4 kg. The carbon emissions generated by the full landfill of nonmetallic materials for the directly driven permanent magnet wind turbine will be 4.230 × 10 3 kg and the carbon emissions generated by the full incineration will be 7.056 × 10 4 kg, as shown in Figure 7. Landfills are currently used primarily in China to dispose of nonmetallic materials. Currently, landfills are the primary method of disposing of nonmetallic materials in China.

| Recovery rate of metals
Metal recovery can significantly reduce wind turbine carbon emissions over their entire life cycle. When all other variables remain constant, carbon emissions from doubly fed asynchronous and directly driven permanent magnet wind turbines can be reduced by 3.950 × 10 4 and 2.790 × 10 4 kg, respectively, for every 10% increase in recovery rate, as shown in Figure 8.

| Transportation
Highway transportation is classified as either gasoline or diesel. When all other factors are held constant, the carbon emissions of doubly fed asynchronous and directly driven permanent magnet wind turbines in heavy-duty gasoline transportation are 7.290 × 10 4 and 5.900 × 10 4 kg, respectively. Carbon emissions from doubly fed asynchronous and directly driven permanent magnet wind turbines are 9.214 × 10 4 and 7.457 × 10 4 kg, respectively, when compared to heavyduty diesel transportation, as shown in Figure 9. For doubly fed asynchronous and directly driven permanent magnet wind turbines, the carbon emissions reductions in transportation using gasoline over diesel are -1.924 × 10 4 and -1.557 × 10 4 kg, respectively.
F I G U R E 7 Sensitivity analysis on disposal methods of nonmetallic materials.

| Conclusion
(1) The carbon emission of the 2.5-MW doubly fed asynchronous wind turbine is 10.43 g/kWh. In comparison, the carbon emission of the directly driven permanent magnet wind turbine with the same power is 8.48 g/kWh, a difference of only 81.4%. Their total energy demands are 1.890 × 10 7 and 1.560 × 10 7 MJ, respectively, and their energy payback times are 0.80 and 0.67 years, as shown in Figure 10. In terms of carbon reduction, the directly driven permanent magnet wind turbine is more promising, but its emissions are far lower than the 1050 g/kWh of traditional thermal power generation. 33 (2) The 2.5-MW asynchronous wind turbine production process contributes 94.7% of the total carbon emissions in the wind turbine life cycle, while the 2.5-MW directly driven permanent magnet wind turbine production process contributes 93.7%. 2.5-MW doubly fed asynchronous and directly driven permanent magnet wind turbines contribute 5.30% and 5.65% of carbon emissions during the transportation stage. The two wind turbines contribute 17% and 16.7% in the operation stage, respectively.
Wind turbine metal recycling significantly impacts the full life cycle results in the waste disposal process. Carbon emissions from 2.5-MW doubly fed asynchronous and directly driven permanent magnet wind turbines are -17.0% and -15.6%, respectively.
The carbon emissions from the two wind turbines in their entire life cycle are primarily caused by the manufacturing stage, followed by operation and waste disposal.
(3) According to the findings of the sensitivity analysis, the method of waste disposal of nonmetallic materials affects the carbon footprint of wind turbines throughout their life cycle. In the case of the 2.5 MW doubly fed asynchronous wind turbine, landfill increases carbon emissions by 0.32% over its entire life cycle, whereas incineration increases carbon emissions by 5.41%. For the 2.5-MW directly driven permanent magnet wind turbine, landfill increases carbon emissions by 0.38% over its entire life cycle, whereas incineration increases carbon emissions by 6.32%. Therefore, the carbon footprint of landfills is significantly lower than that of incineration. (4) Metal recovery has a significant impact on the carbon footprint of wind turbines over their lifetime. Carbon emissions are reduced by 2.88% for every 10% increase in recovery rate for the 2.5-MW doubly fed asynchronous and directly driven permanent magnet wind turbines, respectively. Therefore, increasing metal recovery rates can effectively reduce carbon emissions. (5) The carbon footprint of heavy-duty gasoline transportation of the two wind turbines is 20.9% lower than that of heavy-duty diesel transportation. Clearly, heavy-duty gasoline transportation emits less carbon dioxide than heavy-duty diesel transportation.

| Outlook
This paper makes four recommendations based on a comparison of the carbon emissions of directly driven and doubly fed wind turbines over their entire life cycle, as well as investigations of energy and power enterprises.
(1) The 2.5 MW directly driven permanent magnet wind turbine emits 1.116 × 10 6 kg of CO 2 over its entire life F I G U R E 10 Comparison of carbon emission and energy payback time of the two types of wind turbine. cycle, with a cumulative energy demand of 1.560 × 10 7 MJ. The 2.5 MW doubly fed asynchronous wind turbine has two values: 1.371 × 10 6 kg and 1.890 × 10 7 MJ. The carbon emissions and cumulative energy demand of the 2.5-MW directly driven wind turbine are less than those of the doubly fed wind turbine, with the former being 81.4% and the latter being 82.5%. Therefore, the direct-drive permanent magnet wind turbine is preferred in terms of energy conservation and emission reduction; increasing its market share can help save energy and reduce emissions. (2) Wind turbine production, operation, and transportation all contribute to carbon emissions and energy consumption over their entire life cycle. In contrast, recycling wind turbine metals during the waste disposal stage reduces carbon emissions and energy requirements throughout the entire life cycle. Therefore, the waste treatment methods chosen significantly impact carbon emissions and energy demands throughout the life cycle of wind turbines. According to the sensitivity analysis, the carbon emissions produced by landfill are much lower than those produced by incineration, and no other harmful gases are produced. Wind turbine carbon emissions can be reduced by increasing the metal recovery rate. Therefore, it is suggested that metal materials with higher recycling values be used in the manufacturing process, that the recovery rate of wind turbine metals be improved during the waste disposal stage, and that waste materials be treated in landfills.
(3) Manufacturers can optimize the wind turbine's design and manufacturing process. The total carbon emissions of directly driven permanent magnet and doubly fed asynchronous wind turbines are primarily from the manufacturing stage, accounting for 93.7% and 94.7% of total carbon emissions. Therefore, increasing energy efficiency and material utilization during the manufacturing stage while designing and manufacturing more lightweight and ecological wind turbines is effective in reducing carbon emissions and wind turbine cumulative energy demands. (4) Figure 5 shows that the carbon emission of the tower represents the largest proportion in the production stage, and the tower's main composition is steel. Therefore, by adopting cleaner steel production methods, accelerating the transformation, energy conservation, and efficiency increase of the steel industry, and reducing carbon emissions and energy consumption in steel production, the wind power industry's energy conservation and emission reduction can be indirectly promoted. Furthermore, wind power manufacturers should actively seek out more environmentally friendly materials that can replace steel, as material selection plays a significant role in reducing carbon emissions from wind turbines.

AUTHOR CONTRIBUTIONS
All authors contributed to the study's conception and design. Material preparation, data collection, and analysis were performed by Zhi-Yu Zhuo, Meng-Jie Chen, and Xiu-Yu Li. The first draft of the manuscript was written by Zhi-Yu Zhuo and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.