The Juxtaposition of Our Future Electrification Solutions: A View into the Unsustainable Life Cycle of the Permanent Magnet Electrical Machine
4.2. Ore Separation and Refinement
Numerous methods are available for separation and refinement of individual elements from ore. Depending on the desired material and methods available, the number of processes applied will vary [82]. The separation and refinement of the ore is a very complex process, requiring multiple methods and machines with many iterations in each in order to create a high-purity raw material. In almost all instances, the ore removed from the earth will typically be large rock forms, which require crushing into smaller parts before grinding into a fine powder for further processing [83]. Although there are many possible processes, some of the most common are covered in Table 5. Whilst the processes shown in this paper only outline a few of the available options, the effort and energy required to separate and refine individual elements from mined ore is apparent. An array of machinery, chemicals, power and heat sources are required, which can all pose issues.
Some of these methods have been used for many decades and are industry standard for the processing of the materials. However, there have been technological advances in ore refinement, which may recover greater quantities of elements from the ore being refined [86]. By increasing the yield per tonne of ore, less material will be considered waste, which already contributes significantly to waste streams globally [87]. Whilst some material may be kept to backfill mining operations when completed, others may be sold to secondary industries (such as rocks and soil to the construction sector). Increasingly, tailings and slag as being used in concrete or road production. However, inevitably, there will be hazardous waste produced that cannot be recycled [88]. By replacing virgin material with recycled content in products, there is less burden on finite resources to enable future electrification targets. Steel and copper, for example, are already widely recycled and utilised in new product manufacture. Using recycled steel rather than virgin materials can reduce energy consumption by up to 74% and is the equivalent of saving 1.5 tonnes of iron ore being mined, 1.28 tonnes of waste being produced, and air and water emissions of approximately 80% [89]. Whilst some material recycling is still in its infancy—that of magnets, for example—there are clear improvements to be made across the product manufacture by investing in recycling strategies and utilising existing recycled stock.
4.3. Product Manufacture
Magnets are typically manufactured in one of two ways: bonding or sintering. Bonding is a less common method than sintering; it combines different material blends with various non-magnetic binding agents, typically through one of four processing routes—calendering, extrusion, injection moulding and compression [90]. Sintering creates one solid product by fusing particles at temperature and/or under pressure [51] as part of a complex manufacturing process, outlined in Table 6. Bonded magnet processing can produce near-net shape products requiring little to no finishing operations, which is a key advantage over sintered magnets, as shown in Table 7. The process also allows for complexity in the magnet’s shape and offers more flexibility in the flux output for any given magnet. However, sintering can provide a higher flux output and operating temperature magnet than the bonded process.
The recycling of magnets has been a focus of research in recent years and, subsequently, there has been success in magnet recycling capability [104]. However, repurposing of magnets would be a more attractive solution than recycling within the circular economy hierarchy, as it is less energy intensive and does not require the material to be taken back to a base state. Alternative forms of manufacture may also be considered for improving the sustainability of these magnet. For example, additive manufacturing is being explored particularly as an alternative to some of the bonded magnet processes [105]. Redesign of components could also assist with the reuse of magnets in secondary applications. Högberg et al. presented a case study for the direct reuse of magnets from a 3 MW direct drive wind turbine generator, by making use of a segmented design [106]. Looking at cross-industrial markets could allow for repurposing of magnets into downstream applications, with the use of expertise to ensure quality of the product is maintained. An additional difficulty with magnets is the material composition, as well as the array of sizes, shapes, and coatings used. Whilst magnets are designed or selected based on their expected performance within a product, incorporating some standardisation would be beneficial for sustainable manufacturing, reuse in secondary applications, and efficient recycling streams. The current design of magnets is optimised for in-use performance, but EOL processing can be extremely difficult. For example, most magnets are glued in place- whether they are surface mounted or interior inserted. In addition, the size of magnets that will need to be removed for electrical machines will present major logistical issues. In some wind turbine applications, the magnets can be up to 1 m in length, with an aggregated weight of 2 tonnes [107]. Implementing design for disassembly procedures will be essential in the design of future products to ensure that a streamlined, economical end of life process will be feasible.
Windings will focus on copper as it is the primary material of choice. Copper is predominantly drawn into the wire that is used in windings. The wire is then covered with an insulating coating such as plastic or enamel, with resins often used to impregnate the stator assembly once the coils are wound in place, as outlined in Table 8.
Potential improvements in the sustainability of windings could be made on several fronts, including design, manufacturing, and assembly. The use of hollow coil conductors is used in larger applications to allow for direct cooling of the windings, and if implemented in a wider range of applications, could lead to an improved fill factor (and, hence, improved performance) and the removal of external cooling requirements [43]. Additive manufacture is also being investigated as a small-batch manufacturing option for complex, non-uniform windings [45]. The use of additive manufacturing for coil windings has several advantages including the design of compact end windings, the potential to achieve higher power densities, and the opportunities to include localised cooling of the coils. The current disadvantages include a lower IACS rating—approximately 88% [109]—and the scalability of the process for large-volume, economical production.
One of the largest environmental gains that could be achieved in the windings of electrical machines is the removal of resins. Encapsulation of the stator winding is included as a standard stage for most high-value machines produced currently, with the resins providing several functions, including thermal dispensation, electrical insulation, mechanical stiffness, and protection from external environmental influences. Although the coil resins provide many benefits during the in-use phase, they cause difficulty when disassembling the machine and add complexity to the recovery of material. If resins could be removed from the stator windings, or new resins are developed that are easier to remove at the EOL stage, then this would aid disassembly activities and thus enable easier implementation of circular economy strategies such as repair or remanufacture.
The stator core is most often built up of electrical steel sheets, which are cold rolled to thickness. These sheets are stamped into the required shape before being laminated together. This is discussed in Table 9. The rotor core is generally manufactured by hot forging carbon steel into the cylindrical blank of the size required. Holes for the shaft and slots for the magnets are then stamped, along with any other details required, as outlined in Table 10.
Two of the key issues associated with stator and rotor laminations are the yield losses incurred during profile stamping/cutting and the loss of material value in the recycling process. Several alternative options exist to reduce yield losses, including redesign of components in order to minimise the distance between components when cutting, research into new clamping techniques as a means to reduce wasted cutting space, and prioritising families of components across a company’s entire product range. Another sustainability issue associated with stator and rotor core production is the loss of high-value materials during the recycling process. Recycling of steel is an established industry in the EU, with approximately 100 million tonnes of scrap steel recycled every year. This equates to 56% of the steel produced in the EU steel coming from recycled sources [116]. Electrical steels are a high-value material and go through more processing stages than standard steel grades, which leads to higher costs and longer lead times. When the scrap electrical steel is collected after stator and rotor core profiling, this highly processed, expensive material is recycled in a general steel recycling stream. A value-added activity would be to consider a closed loop recycling system in order to maintain the higher value of this specific material.
The end caps are typically steel, cast into shape. The housing may also be cast steel, although if the profile is simple, these may be extruded. Aluminium is often used as a lightweight alternative to steel for these components. These processes are explored in Table 11. The shaft which passes through the rotor is likely to be an alloy steel in the applications of concern in this paper. As such, radial forging would be employed to allow a changeable profile, as discussed in Table 12. The containment sleeve is likely to be Inconel due to the high-performance requirements of electrification PMEMs and may be forged from a solid bar as outlined in Table 13.
Typically, the non-active components of an electrical machine account for 45–55% of the total weight, with these components not directly contributing to the power density of the machine. Their design and manufacture have also not been a priority, as research has focused on the improvement and optimisation of the active components, such as the windings and stator core. Several improvements could be made to the non-active components by considering alternative designs and manufacturing methods. One area of focus could be the use of near-net-shape (NNS) manufacturing preforms and processes where possible. Current manufacturing routes often include the use of bulk forgings or castings, which are then machined to the final geometry, resulting in high material wastage. The use of NNS processes can result in a reduction in cost and wastage, especially when high-value materials are required. Many NNS processes are also conducted at room temperature, which leads to grain refinement and therefore improved mechanical properties within the final component. An increase in mechanical properties can then facilitate a lightweight design, which can lead to energy savings during the in-use phase of the machine. As discussed in detail earlier, the increased use of recycled material was presented as a way to improve the sustainable production of electrical machines. Since the non-active components are mainly required for mechanical stability and external environmental protection, they are an ideal candidate for increasing the content of recycled materials. Steel, aluminium, and titanium are readily recycled within industry, and electrical machine manufacturers could benefit from this existing infrastructure.
The methods of assembly can be separated into two categories: reversible and non-reversible. Reversible processes include fixings, mechanical fasteners and interference fits, such as between the magnets and the rotor core when the magnets are integrated. Non-reversible processes include welding, resins and adhesives, often used to attach magnets to the rotor when surface mounted, for example. These must include design for maintenance, disassembly, and easier EOL processing.
4.4. In Use
Given that the focus of this research is with regard to future electrification solutions, two widely used PMEM products are electric vehicles and wind turbines. The average age of a vehicle on the road is 8–10 years, whilst the age of cars at scrappage is approximately 14 years [120,121]. Wind turbines have a life expectancy of 20 to 25 years, although current extension programmes are looking to keep turbines in service for 30 years or longer, depending on a number of factors such as their efficiency at site and maintenance requirements [122]. PMEMs will be designed for a specific application, with an assumed efficiency rating based on the variables assessed [123]. However, there are factors that can impact this working efficiency during use, such as working outside of the projected temperature range, insufficient air flow for cooling, overheating of the unit, excessive vibration from poor/degrading installation, electrical overload and contamination [124,125]. This can contribute to a degradation in performance of a product over its lifespan, with Staffell and Green [126] concluding that the output from an onshore wind farm decreases by an average of 1.6% per year. There is now a move to embed sensors within the turbine to provide real-time data, allowing for a more accurate assessment of the machine performance and subsequently allowing for pro-active maintenance [127]. Designing for the implementation of sensors and understanding what data need to be collected during the in-use phase will be crucial for the successful implementation of an optimised end of life strategy. In order for the full power of the circular economy to be realised, the correct intervention strategy must be implemented at the right time, and this will only be possible if the operating cycles of the machine can be traced and analysed. A comprehensive data collection strategy would also allow for predictive maintenance, which would assist with asset life extension. Ensuring a maintenance strategy is developed at the design stage, and stringently followed per manufacturer guidelines, will go a long way to ensure the longevity of high-value electrical machines and allow for preventative maintenance strategies to be employed. A comprehensive maintenance strategy will also provide opportunities for asset life extension, making use of embedded sensors to ensure product performance beyond the designed lifetime.
4.5. End of Life
Whilst some PMEMs may be disassembled by specialist recyclers, many will be disposed of as electrical waste since the technologies capable of recycling the complex components within PMEMs are not at a stage where they are commercially feasible or economically viable [12]. Concentrating on the UK, this lack of supply chain is not aided by the lack of appropriate waste streams available, as the treatment facilities across the countries vary in technology and capability [128]. The most common method of recycling PMEMs that is classified as electrical waste is shredding [129]. This process cuts the machine into smaller pieces and then sorts them into their base materials. However, this process does have a risk of contamination in the waste that results in lower-grade material outputs. Given the high value of some of the materials (for example, the silicon steel and cobalt iron used in the rotor and stator laminations), the lack of separation and salvaging of the individual materials is not only a quality or environmental loss but also a financial one. Larger PMEMs may be disassembled rather than shredded due to the fact they are simply too oversized for the shredding machinery. Disassembly would allow for a greater opportunity for reuse or recycling of the steels, for example, without as much risk of contamination as seen in the shredding method. However, due to the methods of assembly, it is possible that components can be damaged and therefore not suitable for reuse. This is particularly noticeable in copper windings, which are precoated with an insulating material, wound tightly onto the stator teeth and then commonly coated in resin, making it almost impossible to separate and unwind. Even so, the coating is subject to degradation, and, for this reason, it is common practice to recycle the windings rather than attempt any kind or reuse.
If the magnets are surrounded by a containment sleeve, this will have to be cut off due to the interference fit design and therefore cannot be directly reused. There are some commercially available recycling processes for carbon fibre, although this industry and an appropriate supply chain is still in its infancy and the majority of EoL carbon fibre will be disposed of rather than recycled [130]. The magnets themselves can also be damaged due to the design—for example, where the magnet is integrated to the rotor rather than surface mounted. Whilst some magnets may be reused, such as those from wind turbines due to the size [131], the majority will be waste or could potentially be recycled. Currently there is little recycling of magnets, although this is being carried out to some extent in China [132], and there are currently efforts to develop magnet recycling elsewhere [133,134]. With future targets requiring an increase in production volume, the use of the current materials already within the UK will need to be utilised to ease the strain on production of virgin stock and limit the unpredictability of the supply chain. High-value recycling streams could assist in retaining the value of key materials such as rare earth magnets and electrical steel. These streams will be complex to implement, and the challenge is bigger than a single sector; hence, a cross-sectoral approach will be required to make this economically and technologically feasible. For asset life extension, repair and remanufacturing activities to be fully utilised in PMEMs, design for disassembly will need to be a priority, and sustainability factors will need to be on par with performance and cost when designing future machines. Design for disassembly includes design choices such as the elimination of non-reversible assembly techniques in order to facilitate easy dismantling in any relevant waste stream.
In line with the zero-waste hierarchy, reuse of products is considered more beneficial than other practices; therefore, considering potential second use opportunities for the machine at the design stage could lead to significant reductions in the environmental footprint of these products. Potential second life opportunities could include reuse of the complete machine in the same primary application, or repurposing of the machine in secondary, downstream applications. These opportunities will need to be considered fully at the design stage, with a cross-sectoral approach in mind, as a lack of forward planning will likely make secondary use highly improbable.
Whilst the responsibility for this end-of-life management may lie with different parties (i.e., the manufacturer, consumer, decommissioner or otherwise), having the infrastructure available to manage these different products will be critical. Ideally, the end-of-life would be considered at the design stage and “built in” at the manufacturing stage, i.e., by manufacturing with disassembly potential in mind so that materials can be recirculated at end-of-life through appropriate channels, reducing resource depletion and embedded emission impact.
4.6. Transportation and Packaging
An aspect of the entire life cycle of the product not discussed above is transportation requirements. There are many stages over the life of the product where the materials and components will require transportation on a global scale for PMEMs, as shown in Figure 5. The supply chain of PMEMs is complex because of this, typically with multiple modes of transport required and a variety of (often single-use) packaging used, all of which has an environmental impact [135].
For example, NdFeB magnets are composed of more than ten individual materials, all of which must be mined. This is a global operation due to the availability of the ore, with neodymium primarily sourced in China, niobium mainly found in Brazil, iron mined in Australia, and so forth. The ore will be transported from where it is mined to where it is refined, and packaged, such as in powder or billet form. All these materials must then be transported to one location and mixed into the composition required, before finally being manufactured into the magnet. Magnets are likely sold business to business and will thus be transported again to the assembly site, before being sent to the distribution centre and then onto the customer. To overcome some of these challenges, creating localised markets would be a potential solution. Having closed loop recycling of materials in-house or within a local area to reduce virgin stock could reduce transportation costs and the associated environmental impact. Where virgin stock is required, localising the source location of this could also be advantageous from a transportation perspective. For example, whilst Australia is one of the highest producers of iron, it is mined in many other countries—selecting the one closest would reduce the impact of transportation whilst potentially also reducing costs and time. Contracting parts or manufacturing to external companies is not uncommon, but optimising the network of these could also have a positive effect. For example, rolls of sheet metal material may be manufactured in Asia or South America, shipped to Europe for component manufacture (i.e., stamping or laser cutting of laminations), and then shipped elsewhere for assembly before being sent to the company’s distribution location(s), whilst other component parts make similar global journeys for use in the final assembly. Re-considering each of the suppliers, their locations and making steps to reduce the number of stages or at least the distance over which parts are travelling could reduce impact and time. Utilising LCA to provide a baseline for current practices and having greater transparency of these operations could be beneficial to this exercise. In relation to the packing, utilising existing standardised products such as crates and pallets where possible, without single-use wrapping, would be beneficial. Minimising single-use packaging or replacing it with reusable, recyclable or compostable alternatives should be feasible due to the variety available today.
