1. Introduction
Mitigating climate is undoubtedly the major challenge humanity is facing today. With the evidence of fossil fuels being a major contributor to human carbon dioxide emissions over the last 200 years [
1], Germany is a pioneer in decarbonizing its energy sector. The transformation aims to substitute fossil energy sources (e.g., coal, oil, gas) with renewable energies such as wind turbines or photovoltaics [
2]. Until 2030, 80% of the electricity is aimed to come from renewable energies, mainly wind power and solar power, and ultimately achieving a climate-neutral status in 2045 [
3]. Wind turbines, as mature and widely applied technology, are already the key contributors to Germany’s renewable electricity sector [
4]. However, this transformation comes with a demand for new materials for these new energy technologies.
Modern wind turbines oftentimes use permanent magnet synchronous generators (PMSG), which can easily contain up to three tons of high-performing permanent magnets [
5,
6]. Since their market introduction in 1984, neodymium–iron–boron (NdFeB) permanent magnets are considered to be the best available magnets due to their superior energy product [
7]. Depending on their application, the magnet’s properties are enhanced by adding other rare-earth elements (REEs) like praseodymium, dysprosium or terbium [
8]. Generally, there are three types of magnets: sintered, bonded, and hot pressed or hot deformed magnets. Sintered magnets are superior in terms of their magnetic performance [
9], and more than 90% of all magnets produced are sintered magnets [
5,
9,
10]. REE originated from mining activities and is predominantly located in China. Geopolitical conflicts and the COVID-19 pandemic are massively influencing the market [
11], and the increasing demand of regional markets is causing bottlenecks in supply [
12,
13]; these elements are becoming increasingly valuable. Furthermore, the recent developments of tariffs in global trade, driven by the United States and retaliatory measures by China, increase the supply risk of these REEs for the EU and Germany [
14].
The European Commission considers these elements as critical raw materials (CRM), meaning their high economic importance mismatches their high supply risk [
15]. Furthermore, extracting and refining these elements causes severe environmentally harmful impacts [
12], and the environmental benefit of recycling NdFeB magnets has been investigated by comparative Life Cycle Assessments of primary produced magnets and recycled magnets [
16,
17,
18,
19,
20]. These studies found that the environmental impacts, in general, were below those of the primary production. To address the major issues, establishing a circular economy for REE is a major goal in the German raw material strategy [
21]. Additionally, for CRM, the European Commission aims to ensure its own domestic critical raw materials supply chain within the ‘Critical Raw Material Act’ (Regulation (EU) 2024/1252) by targeting to extract at least 10% of the critical raw materials it uses from within Europe by 2030, processing at least 40% within the EU and recycling 15% of the EU’s annual consumption of raw materials by 2050.
Various circular economy measures for magnets have been developed and can be categorized into the reuse of magnets, magnet-to-magnet recycling and the elemental recycling of REE [
5,
7,
22,
23,
24,
25,
26,
27,
28]. There are European and German startups with ambitions to recycle magnets in the future [
29,
30]. However, industrial-scale magnet recycling is in the beginning of being created in Germany and Europe with recycling rates of ~1% [
26]. Nowadays, the availability of low-cost primary magnets from China dominates the market. Nevertheless, to reduce supply risk associated with RE magnets as well as to strengthen domestic production within the EU, a constant supply of End-of-Life (EoL) magnets to ensure reproducible properties of recycled magnets is the aim of the EU’s critical raw material act [
26]. Several studies address the technical feasibility of recovering magnets and REE from EoL products [
31], mainly waste electric and electronic equipment (WEEE) [
32,
33,
34,
35]; the recovery of magnets and REE from wind turbines is not addressed often [
36]. Despite being used for 20 years, more wind turbine generators will form a reliant input for recycling NdFeB in the future [
5,
37]. Several studies have investigated the potentially available EoL magnets for Recycling today and in the future, mainly focusing on global neodymium or NdFeB markets [
8,
37,
38,
39]. However, the potential contribution of End-of-Life magnets from wind turbines in Germany today and a future circular economy of NdFeB magnets needs to be addressed. Therefore, this work gives an overview of the decommissioning of wind turbines in Germany and the potential accessibility of NdFeB magnets for recycling or recovery of the contained rare-earth elements for the time frames of 2024, 2025–2030, 2031–2044 and 2045–2050 as well as an outlook of the potentially available secondary material.
2. Materials and Methods
A graphical display of the methodology is displayed in
Figure 1. The methodology is further explained below.
As a first step of the potential analysis, the material flows of NdFeB of wind turbines for each time frame into the End-of-Life phase need to be determined. For this, the use of magnets in wind turbines is of interest, as is the number and capacity of decommissioned wind turbines. Especially for future time frames, assumptions are made regarding the development of installed capacity, the usage of permanent magnets onshore and offshore, and the average lifetime. Currently, generators are the only component of wind turbines containing magnets [
40]. In Germany, five types of generators are used [
41]. They can be differentiated in terms of the generator technology (synchronous or asynchronous), the type of excitation (electrically or permanent), and the type of gear unit (direct drive, high speed or middle speed). The only drivetrain using permanent magnets is the permanent magnet synchronous generator (PMSG). They can be operated at all wind speeds and do not need excitation current like electrical excited synchronous generators (EESG) [
40,
42]. Moreover, they can be used in onshore and offshore applications [
43] by direct drive or with a gearbox [
42,
44].
For the time frame of 2024, the number of decommissioned wind turbines can be taken from national statistics, claiming a decommissioning of 712 MW (557 turbines) [
45] onshore and no decommissioning offshore [
46]. For future decommissioning, the average lifetime, development of installed capacity and number of turbines, as well as their year of construction, needs to be determined.
Data for the average lifetime of wind turbines vary from 16 years over 20 or 25 years up to 30 years [
5,
47,
48]. Limiting factors are, e.g., the profitability and the duration of subvention [
5,
49]. The average lifetime for this work is determined to be 20 years onshore and offshore based on the end of the subvention of wind turbines in Germany after 20 years [
5].
Today’s (31 December 2024) installed capacity of wind power in Germany is 63.551 MW onshore (28.6717 wind turbines) [
45] and 9.222 MW offshore (1.639 wind turbines) [
46]. By considering an average lifetime of 20 years, the installed capacity of 2030 determines the amount of material reaching the End-of-Life stage in 2050. The German government plans to expand wind power to an installed capacity of 115 GW onshore and 30 GW offshore, respectively, until 2030 [
50]. Additionally, an estimated decommissioning of onshore turbines of 17 MW until 2030 has to be covered by new installations [
48]. Furthermore, onshore decommissioning is determined because of the age of wind turbines given in [
45], assuming all turbines built until 2010 are powered down and a linear capacity distribution between the years 2009 and 2023. Regarding offshore decommissioning, until 2030, only the first German wind park, ‘alpha-ventus’, is expected to be decommissioned. Consequently,
Table 1 summarizes the capacity of wind turbines reaching the End-of-Life phase in Germany, their application and year.
The capacity of turbines containing rare-earth magnets is determined by identifying the share of PMSG for each time step. The graphical data by Berkhout et al. [
51] are taken to cover the shares of decommissioned PMSG for 2024 to 2030. Zimmermann et al. [
40] project a market penetration of direct drive PMSG in 2030 of around 15% onshore. Since high-power turbines (>5 MW) mainly use PMSG [
5]. The market share of geared PMSG in 2030 is assumed based on the trend of overall increase in PMSG usage [
5] combined with being the major application of PMSG onshore so far [
51] to 30%.
Regarding offshore applications, the literature values are inconsistent. Viebahn et al. [
41] claim that until 2012, all German offshore wind projects were based on asynchronous generators, not using RE-magnets. Glöser-Chahoud et al. [
5] assume 50% for offshore application (direct drive only), and Zimmermann et al. [
40] project an installation of 15% direct drive PMSG in 2030 offshore. To gather reliable data on the PMSG usage in German offshore wind turbines, a screening of installed and planned offshore wind parks until 2030 covering the amount, installed capacity and usage of PMSG is performed [
52]. A detailed list of installed wind parks in Germany is displayed in
Supplementary Materials—Table S1, as well as a list of wind parks planned and under construction until 2030.
Supplementary Materials—Table S2 summarizes the shares of PMSG reaching the End-of-Life stage, respectively, for each time frame and onshore or offshore application. For this, all offshore wind parks in Germany are determined [
52], then each type of wind turbine is determined [
53], and finally, their drive type [
54] results in the overall share of direct-drive or geared PMSG usage.
Considering the shares and decommissioned capacity of
Table 2, the decommissioned capacity of PMSG for each time frame is displayed in
Table 3.
Sintered magnets form the major market share and are superior in their magnetic properties [
5,
9], leading to the assumption that wind turbines use exclusively sintered magnets. The literature data on the magnet content vary from 80 kg/MW to 650 kg/MW depending on their drivetrain [
5,
41,
55]. Viebahn et al. [
41] even assume a reduction in the material intensity of magnets until 2025. This work uses the data of van Nielen et al. [
6], 625 kg/MW for direct drive (low speed) and 134 kg/MW geared (high speed), with both values valid for onshore and offshore applications.
Table 4 displays the number of RE magnets reaching the End-of-Life stage, respectively, their application in direct drive or geared generators of onshore and offshore applications in Germany, as well as the expected time frame of decommissioning.
NdFeB magnets are mainly based on the Nd
2Fe
14B phase, giving a theoretical Nd content of 27%. Depending on the performance of sintered magnets, other REEs such as Ps, Dy and Tb are added. Different alloy configurations are possible. Since Terbium and Dysprosium are added for the same purpose of increasing coercivity and Tb is more expensive and rarer than Dy, Tb is not considered [
5]. For sintered magnets, the typical overall content of REEs is 31 wt% [
6]. For the Dy-content in wind turbines, Binnemans et al. [
22] claim a share of 4.1 wt%. For the ratio of Nd to Ps, the literature names 3:1 [
56] or 4:1 [
39]. By choosing a ratio of 4:1, the REE composition is consequently at Nd 21.52 wt%, Ps 5.38 wt%, Dy 4.1 wt%.
Table 5 summarizes the amount of each REE for each time frame.
3. Results
The availability of data for today’s End-of-Life treatment of RE-magnets from wind turbines is limited. Guidelines for decommissioning [
49,
57,
58,
59] mention the separation and special treatment of RE magnets in the End-of-Life phase, which is not further explained. Despite several pathways to recycle RE-magnets and recover REEs having been developed, no commercial-scale recycling has been established in Germany or the EU. Today, recycling plants for RE magnets only exist in the laboratory and pilot stages, leaving the REE recycling rate below 1% [
28]. This is mainly because no primary production of REE is carried out in Europe [
39].
Today, there is no commercial-scale recycling of magnets in Germany or Europe. Technically, impeccable plants can be reused in a new location. Turbines are frequently exported and rebuilt in developing and emerging economies [
5]. In fact, there is a market for used wind turbines. A screening of the website “
wind-turbine-models.com” showed around 200 advertisements in Germany [
60]. A number of 557 turbines have been decommissioned in 2024 [
45]. Moreover, according to the German Federal Environment Agency [
39], 70 Mg of post-consumer NdFeB-magnets were collected in 2015, mainly coming from wind turbines or electric vehicles. Additionally, 100 Mg of magnet production scrap is exported to China or Japan to be used in RE-metal production [
39]. Reimer et al. [
8] claim a total of 150 Mg, including production scrap, to be exported. Furthermore, NdFeB magnets can end up in the scrap yard and be melted together with steel scrap to produce recycled steel [
61]. This mainly happens for smaller magnets used in electronic applications [
39]. Due to their Fe-content of 67–73 wt.% [
62], NdFeB-magnets ending in the scrap yard can be melted together with steel scrap [
5]. Steel recycling is an established process, and so-called tramp elements, which are not intentionally present in steel, are kept in the steel cycle once they enter [
63]. However, REEs are then “lost” in the steel cycle.
Table 4 and
Table 5 display the theoretical potential or stock of NdFeB magnets and REEs used in wind turbines that are reaching the End-of-Life phase for each time frame in Germany. These numbers may be affected by an actual collection rate, assumed by Schulze et al. [
37] at 90%. Nevertheless, a potential amount of secondary material of NdFeB can be assessed using only the material efficiency of different recycling processes. These processes can be categorized into magnet-to-magnet (direct recycling) and elemental/recycling of REE (indirect recycling) [
22,
23]. Furthermore, corrosion or other magnet degradation is neglected in order to address the theoretical potential first. Moreover, metallurgical recycling routes are able to process oxidized material [
22].
Table 6 gives an overview of the different recycling pathways, their estimated TRL and their material efficiency in terms of recovery.
Given the numbers of
Table 6, magnet-to-magnet recycling routes, if they were commercially available today, could have provided 3555–4825 kg of recycled NdFeB magnets of End-of-Life magnets collected in 2024. The values of recycled magnets for the time frames of 2025–2030, 2031–2044 and 2045–2050 range from 205,929 to 279,475 kg, 4,870,786 to 6,610,353 kg, and 8,509,306 to 11,648,343 kg, respectively.
Table 7 summarizes the potential amount of recycled magnets, the time frame and the recycling pathway.
However, this does not consider the type, application and magnetic properties of the recycled magnet. The actual recycling process chain depends on the desired magnetic product. Hydrogenated NdFeB powder from hydrogen decrepitation or HDDR, e.g., can be further processed to sintered magnets. Depending on their application, magnets have varying process chains, which may include the addition of (heavy) REE [
9,
20,
22,
26]. Ribbons produced by Melt Spinning can be further processed to bonded or hot pressed and/or hot deformed magnets [
9,
26]. However, they also vary in magnetic properties and application. Given these challenges of recycling magnets, a substitution potential is not determined in this work.
Elemental recycling routes, if they were commercially available today, would be able to recover 601–1082 kg of neodymium, 150–270 kg praseodymium and 114–206 kg dysprosium from End-of-Life magnets collected in 2024. The values for the time frames from 2025–2030, 2031–2044 and 2045–2050 are summarized in
Table 8.
Recovered REE can then be reintroduced in already-existing production chains for magnets or different goods containing REE [
9,
22]. Neglected in these numbers is that today, there is a market for used wind turbines to be exported [
5,
54]. How potential magnet recycling affects this market is out of scope.
The potential analysis shows that the expected increase in installed wind power capacity, especially offshore, is a significant input for a future circular economy of NdFeB magnets. However, even potential future recycling heavily mismatches the current demand of German industries for NdFeB magnets of 13,000 t [
66]. With an increased demand in the future, there is a need to diversify the source of EoL magnets for recycling. To put the calculated numbers of this work into the context of previous studies, it can be said that Schulze et al. [
37] determined global values of NdFeB coming from wind turbines in 2030, ranging from 1480 t to 2390 t. Assuming that Germany has a share of 6.1% of global wind power installations [
67,
68], it would result in 90–146 t of EoL magnets from wind turbines, which is double the value of this work. This can also be seen in the comparison of REE. However, this may result from a different scope and approach to determining amounts. Berkhout et al. [
51] claim a substitution potential of NdFeB for Germany of 1700 t and 4800 t for 2030 and 2040, respectively. This is exceeded by this work, considering wind turbines are responsible for 10–18% of German NdFeB demand. For the values of 2050, no comparison of the literature data is possible yet. Moreover, it needs to be pointed out that all of these studies [
8,
37,
38,
39] differ, in parts heavily, in their results because of different scopes, assumptions and methodologies.
The data presented in this paper come with uncertainty due to the use of secondary data. Nevertheless, this work quantitively displays the theoretical potential of NdFeB magnets as input for recycling to a circular economy and provides an outlook on a potential amount of secondary magnets. However, despite providing a range of material efficiency, the recycling processes are still in development and, therefore, give a very broad range of expected secondary material. Furthermore, since magnets are functional materials specially designed in shape and properties for their application [
9,
26], recycled magnets need to address the specifications of desired products as well. Technical developments in recycling processes for magnets may improve their performance; however, they cannot be predicted in this work. Future studies need to consider the performance of recycled magnets and their application in order to fully assess a substitution potential for rare-earth magnets in terms of a circular economy.