Materials use in electricity generators in wind turbines – state-of-the-art and future specifications

Journal of Cleaner Production 87 (2015) 275e283

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Materials use in electricity generators in wind turbines e state-of-the-art and future specifications
ntegui* Roberto Lacal-Ara European Commission, DG Joint Research Centre, Institute for Energy and Transport, P.O. Box 2, 1755 ZG Petten, The Netherlands

a r t i c l e i n f o

a b s t r a c t

Article history: Received 24 April 2014 Received in revised form 22 August 2014 Accepted 14 September 2014 Available online 28 September 2014

The European Strategic Energy Technology Plan, adopted by the European Union in 2008, is a first step to establish an energy technology policy for Europe and to support the 2020 energy and climate change targets from the technology development point of view. One of its initiatives is to assess the characteristics of the materials that will be needed in order to achieve the 2020 targets, in terms both of amounts of materials and their technical specifications, along with the way to get there for the latter. The Materials Initiative was created to foster a roadmap which is based on a scientific assessment of the current situation. This paper presents the work of the author in the (wind turbine) electricity generator part of that assessment, it includes the aspects of technology and system state-of-the-art; material supply status; on-going research and players; materials specification targets for 2020/2030 and beyond. The assessment found that the performance of permanent magnets is the single item potentially to provide the most significant improvement in component specification, but that in order to achieve this perhaps new chemical components ebased on rare earths, as currently, or not-will be necessary in order to achieve these high-performance magnets. The search for these new materials is stimulated by the current dependency of the world in a nearly-monopolistic supplier of rare earth elements. The assessment also concluded that the improvement of materials specifications is challenging but achievable in most areas, and a crucial aspect for the necessary cost reductions in wind energy production. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Energy Wind power Materials Permanent magnet

1. Introduction This scientific assessment contributed to a roadmap for research and development of materials for wind power technology, the “Materials Roadmap Enabling Low Carbon Technologies”, a European Commission Staff Working Document published in December 2011 (EC, 2011). The Materials Roadmap aims at contributing to strategic decisions on materials research funding at European and Member State levels and is aligned with the priorities of the Strategic Energy Technology Plan (SET-Plan) (EC, 2007). Whereas the importance of materials in wind energy roadmapping has been highlighted (Daim et al., 2011), the Materials Roadmap (EC, 2011) is specifically intended to serve as a guide for developing specific research and development activities in the field of materials for energy applications over the next 10 years. This paper provides an in-depth analysis of the state-of-the-art and future challenges for the wind turbine electricity generatorrelated materials and suggests the targets for the research

* Tel.: þ31 224 565 390. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.jclepro.2014.09.047 0959-6526/© 2014 Elsevier Ltd. All rights reserved.

activities that will support the development of wind power technology both for the 2020 and the 2030 market horizons. It has been produced under the coordination the SET-Plan Information System (SETIS), which is managed by the Joint Research Centre (JRC) of the European Commission. The contents were presented and discussed at a dedicated hearing in which a wide pool of stakeholders participated, including representatives of the Wind Technology platform (TPWind), industry associations and the Joint Programmes of the European Energy Research Alliance. The main overall challenge for wind energy developments, and main objective of the technology evolution, is the reduction of the cost of wind energy. This needs to be obtained through reductions in capital investment and maintenance costs and through maximising the annual energy production.

1.1. Setting the scene: energy and wind power Energy is a vital input for social and economic development and a significant part of it energy is obtained from fossil fuels. However, countries without a native supply of fossil fuels suffer an external dependence which transfers welfare to the exporting countries.


ntegui / Journal of Cleaner Production 87 (2015) 275e283 R. Lacal-Ara

276

List of acronyms and abbreviations EC EU SET-Plan SETIS JRC TPWind HTS PE PM PMG DFIG MW DD

European Commission European Union Strategic Energy Technology Plan SET-Plan Information System Joint Research Centre European Wind Technology platform High-temperature superconductor Power electronics Permanent magnet Permanent magnet generator Doubly-fed induction generator Megawatt Direct-drive

One strategy to reduce the dependence on fossil fuel resources involves using renewable energy sources, of which wind energy is one of the most developed, a powerful and profitable way to satisfy the demand for sustainable development. Partly as a consequence, wind energy power generation has been growing over the last decade by an average 23.5% per year. The total worldwide installed capacity during 2013 was over 35 GW and at the end of 2013 the total installed capacity reached 318 GW (GWEC, 2014). In the EU the installed capacity at the end of 2013 was about 117 GW, whereas at country level China, the USA, Germany, Spain, and India are currently the world top five windpower-producing countries, and together they share 72% of global installed capacity (GWEC, 2014). It should be noted that in 2013, for the first time ever, wind power was the technology that contributed most to the annual electricity demand coverage in a country (Spain), with a share of 21.1% compared to 18.1% in 2012), reaching the same level as nuclear which contributed 21% (22.1% in 2012) (REE, 2014). The largest wind turbines installed are growing from 2 MW in 2000 to 8 MW today with rotor diameters up to 171 m,1 and manufacturers are working on designs up to 10 MW with rotor diameters of up to 180 m. Fig. 1 shows a schematic diagram of a wind turbine with gearbox. European wind installations have an average capacity factor of 22.3%.2 Offshore wind farms, with stronger, steadier winds, may reach capacity factors in the range of 45% (Lacal-Ar
antegui, 2014). Even considering the planning constraints relating to shipping lanes, fishing, bird migration, and the like, the world has abundant space for offshore projects. But conditions during offshore installation, operation and maintenance may be harsh, and so products must be highly reliable. Almost all commercial large turbines for both on and offshore locations are horizontal axis, three-bladed turbines with tubular steel towers.3 The main components of wind turbines include rotor blades, rotor hub, main shaft, gearbox, electricity generator and power converter, all generally hosted in a nacelle supported by a bedplate

1

The first prototype of Samsung's S7.0-171, with a 171-m rotor was commissioned at the Energy Park Fife, Scotland, UK, in March 2014, and the first prototype of the Vestas V164-8.0, with an 8 MW electricity generator, was installed at the Østerild Test Centre, Denmark, in January 2014. 2 10-year average 2003e2012. JRC calculation based on Eurostat electricity generation data and EWEA's annual installed capacity. 3 Several companies have 2-blade designs but only the Chinese Ming Yang seems able to achieve serial production of 2-blade turbines.

REE REO LV MV IEA EWEA CIS RoW t NdFeB CC YBCO AMSC MA

Rare earth elements Rare earth oxides Low voltage Medium voltage International Energy Agency European Wind Energy Association Confederation of Independent States Rest of the world tonne (International System) Neodymiumeironeboron Coated conductors Yttrium, barium, and copper alloys American Superconductor Million amperes

that is mounted on a tower and rotates thanks to a yaw bearing system see Fig. 1. Direct-drive turbines have a gearless drive train with a direct connection from the rotor hub to a large low-speed generator, see Fig. 2.

1.2. Materials used in a wind turbine The materials used in the main components are:  The blades are produced from polyester or epoxy reinforced with mainly glass fibres and to some extend carbon fibres in combination with polymer foam or balsa wood for the sandwich parts. The blades are mostly produced in two halves, the upper and lower part, and are joined using adhesive bonding. Both the aerodynamic shape as the structural parts of the blades are produced from fibre reinforced materials. The blades are bolted via a pitch bearing to the hub.  The rotor hub, the main shaft, the gearbox, the bedplate and the tower are basically produced from different types of low-alloy steels and cast irons. Again here the functional as well as the structural shapes are combined.  Electric generators are basically made of a rotor and a stator and materials used include magnetic steels and copper for wirings in electromagnet generators or steels, copper, boron, neodymium and dysprosium in permanent magnet generators. Hightemperature superconductor (HTS) generators, still in their development phase, use basically HTS wire and ceramics.  The power electronics (PE) converter, whether partial or full, ensures that the output electrical signal complies with the power quality of modern grid codes, and enables the variable speed operation of the wind turbine. In addition, PE devices are used in wind turbines as part of the control system to help achieving safe control and the maximum power output. The overall challenges for the materials used in wind energy that need to be addressed through research and development are:  Life cycle management, from ore processing until waste reuse and recycling. This needs to be done by means of environmentally-friendly production technologies. In many cases existing processes need to be adapted. New materials (nanomaterials, fibres and polymers for blades, lubricants, permanent magnets)  Resource management: Europe being a continent with very little raw material resources should assure its strategic access to


ntegui / Journal of Cleaner Production 87 (2015) 275e283 R. Lacal-Ara

277

Fig. 1. Turbine with a gearbox-based drive train. Source: Siemens.

these products and or develop alternatives for the critical materials,  New materials which make innovative solutions technically feasible,  Materials for extreme conditions of exploitation, such as offshore and cold climate conditions 1.3. Electricity generator Electric generators are basically made of a rotating part (rotor) and a fix part (stator). The rotating magnetic field induces electric energy in the windings in the stator. Their magnetic field can be provided by either electromagnets, which need electrical excitation and consumes reactive power, or permanent magnets (PM) which do not need electrical excitation (Hansen, 2005). Which generator is used in turbines depends on the whole drive-train design: gearbox-generator-power electronics. A drivetrain configuration with a 3-stage gearbox has evolved from using squirrel-cage induction generators into using doubly-fed induction generators (DFIG) with a partial power electronics converter. 64% of installed capacity in 2009, up from 34% in 2000 (Llorente-Iglesias et al., 2011), used this DFIG configuration which requires a highspeed generator. This configuration adapts well to most grid code requirements; it is proven and cost-effective and therefore preferred by investors and financing bodies for onshore wind farms. When there is no gearbox or there is but it has less than 3 stages the generator can be permanent magnet- or electromagnetbased, but it will always use a full PE converter (21% of the 2009 installed power according to Llorente-Iglesias et al. (2011)). A 2-

Fig. 2. Direct drive turbine (courtesy Siemens).

MW electromagnet generator is made of approximately 66% magnetic steel, 30% copper, and 3% silica (Martínez et al., 2009). An alternative to the DFIG with a 3-stage-gearbox is the multipole direct-drive (DD) generator which obviates the gearbox -a component whose failures cause significant downtime and repair costs. Electromagnet generators constituted more than 85% of DD generators installed up to 2009, but only 54% of those installed in 2009 worldwide (JRC, 2011). Electromagnet generators are less efficient below their rated power than PM generators, and turbines naturally generate below rated power for most of the time. New generators using permanent magnets (PMG) in the rotor achieve higher magnetic density: a 15-mm thick segment of permanent magnets can generate the same magnetic field as a 100e150-mm section of copper windings (BNEF, 2010). PM is an intrinsically more robust technology that has lower maintenance needs. PMGs are the choice for the majority of new offshore wind turbine designs and therefore they will be increasingly present in future offshore wind farms -as well as onshore installations through cascade effects. Main materials in PMGs include magnetic steel, boron, rare earth elements (REE), insulation and copper. The generator size is approximately one quarter less than its electromagnet counterpart (Siemens, 2011) of similar rating, but its specific mass depends on whether it is low-speed (which matches a DD configuration), medium-speed (coupled with a 1- or 2-stage gearbox), or high-speed (coupled with a traditional 3-stage gearbox). The magnet content of a 3-MW generator is around 650 kg/MW in the low-speed case, around 160 kg/MW for mediumspeed, and around 80 kg/MW for high-speed PMGs (Hill, 2010, 2011). The most popular permanent magnets are often called NdFeB magnets because they are manufactured from neodymium (Nd), iron (Fe) and boron (B), with a low but vital part of dysprosium (Dy). Praseodymium (Pr) may eventually replace up to 1 in 4 Nd, but with a certain loss in quality. Traces of terbium (Tb) can be used to prevent the PM from losing magnetic properties at high temperatures, although this is a function most commonly played by Dy. Bonded magnets are not used in the wind sector; it is the sintering process that produces the higher performance magnets required for larger wind turbine applications (Reyes, 2011). Nd, Pr, Dy, and Tb are REE subject to a difficult supply/demand balance (see Section 2). Manufacturers employ proprietary variations of elements within the magnets to produce the desired properties and proprietary process technologies for forming magnetic shapes. Fist-of-a-kind high-temperature superconductor (HTS) generators have been built in other sectors (e.g. ship propulsion) and are seen as the next technology for power generation in conventional plant and then for the wind sector, but at this stage they seem very far away from the market. In an HTS generator a superconducting rotor allows very high field strengths to be created in the air gap of

278


ntegui / Journal of Cleaner Production 87 (2015) 275e283 R. Lacal-Ara

the machine, making power density extremely high. Introducing the HTS technology removes an additional 50% of the mass of the generator, making even greater capital cost reductions possible (Hill, 2010). It is claimed that an HTS generator offers another advantage over the PMG, namely that the peak power consumed by the cryogenics of a DD-HTS generator is a fraction of the normal losses in a conventional rotor, and significantly lower than the heating effects on magnets (Hill, 2010). This would give the HTS rotor the highest possible efficiency and additional earnings over the lifetime of the machine. Superconducting wire costs, a key component of HTS, are reducing at a rate of around 10% per annum, and the rate of saving is accelerating as more wire becomes available. The first HTS generator in the renewable energy sector was expected to be delivered for a run-of-river scheme at Hirschaid in Germany (PowerGen, 2011). The supplier stated that a reduced size DD-HTSG design for the wind sector was to be demonstrated and trialled in 2011 and 2012, followed by a fully rated project from 2013 e but this has been delayed. The first use of this technology in volume serial production is likely to be determined by the relative availability and cost of permanent magnets and HTS wire (Hill, 2010). The market evolution from DFIG towards DD-PMG will bring about an increase in the use of copper, due to the larger size of the generator: a 3-MW DFIG has approximately 800 kg of copper per MW (Ancona and Mc Veigh, 2001; Martínez et al., 2009; Vestas, 2009) which can be compared with 2.1 t/MW (Frost and Sullivan, 2010) in the case of a 3.3-MW DD-PMG (weight 70 t of which 10% is copper). Therefore the move towards DD-PMG wind turbines will result in a more than doubling of the copper demand for the corresponding moved market share (Hill, 2011; JRC, 2011). Significant as it is in relative terms, this change will not make copper a critical material for wind turbine generators although the environmental aspects of ever-lower-grade copper production could become critical in the future. Nowadays the large majority of generators supply low voltage (LV, less than 1000 V) whereas new wind turbines in the multi-MW range are increasingly being designed for medium voltage (MV) generation of 3.3 kV or above.

2. Material supply status and challenges The analysis of materials supply and challenges was based on the consumption in 2010 as well as on scenarios for future consumption.

2.2. Materials for the electricity generator and power converter Iron, copper and all the other materials used for manufacturing magnetic steels and copper wirings for electromagnet generators are abundant in the world, come from a wide geographical distribution and have a world market which ensure transparent prices. This is not the case for the rare earth elements (REE) necessary for PMG. Copper is not considered a critical material. However, researchers alert that in addition to market and socio-political factors other factors including mining energy and water consumption and ground usage can turn copper into a critical element (MII, 2010). Leading copper producing countries in 2009 were Chile (34% of global production), Peru (8%), U.S. (8%), China (6%), Indonesia (6%), Australia (6%), Russia (5%), Zambia (4%), Canada (3%), Poland (3%), Kazakhstan (3%), and Mexico (2%); Mongolia and the Democratic Republic of Congo have major, recently discovered reserves (APS, 2011). REE occur in low concentrations in metal ores, they are often coproduced with other metals and among themselves in concentrations which vary widely from ore to ore. The revenue stream from REE in those mines used to be a minor part of the revenue from the major product, and thus mine managers' decisions did not depend on the REE price signals from the market but on those of the main ore. However, recent scarcity (see below) and price increases are stimulating the development of REE-only mines or the re-opening of those in Western countries which closed in the 1990s because of the competition of low-price Chinese REE (DoE, 2010). In effect, the cost of REE used to represent a low percentage of the total cost of the generator which contributed to its demand being largely inelastic with respect to price. Therefore, even large changes in the prices of rare earths, as shown in Fig. 3, had a relatively low impact on producer costs, but this has recently changed. Notwithstanding the recent price reductions, the rare earth industry expects prices to increase above 2013 levels (Kingsnorth, 2013).

Why the REE price increases? The REE price crisis was triggered by a drastic reduction of Chinese export quotas (see Table 1) and a nervous market realising that at 97% of world production, China had a strong market power. However, as in any crisis there were many other factors contributing, e.g. the technology fundamentals that in many fields make REE products to have superior specifications (e.g. in permanent magnets) and which reduce their price elasticity.

2.1. Deployment rate From the scenarios used in the SET-Plan, the EWEA (2010) and IEA (2008) scenarios for Europe and the world were evaluated for the period 2010 to 2020/2030 and 2030 to 2050. These scenarios lead to a maximum deployment rate of about 50 GW per year over the periods considered. This maximum comes from the IEA-BLUE scenario in the period 2030 to 2050.4 However the deployment rate in 2010 was over 39 GW already and BTM-Consult (2011) predicted much higher annual deployment rates, up to 140 GW by 2020. Therefore it was decided to use as a reference for this analysis an annual deployment rate of 100 GW, of which about 10% is assumed to be installed offshore.

4 Habib and Wenzel (2014) carried out a thorough literature review of scenarios for future deployment of wind turbines differentiating the direct-drive machines from the rest.

There are first signs that REE for permanent magnets, and very specially dysprosium, are starting to be seriously scarce with current supply levels. This, in turn, suggests that client firms may have to invest in the extensive (and expensive) research and development (R&D) needed to identify substitutes and lowering the material intensity of REE in magnets. End-user strategies, i.e. by wind turbine manufacturers, are also affected and for example a technological option of a PMG in combination with a 1- or 2-stage gearbox (the hybrid model) is gaining more attention because of the reduced rare-earth content of the corresponding medium/highspeed PMG. On the positive side projections of new mines suggest that REE scarcity will ease for Nd by 2015 and for Dy by 2017 (Hatch, 2011). Once mined or coproduced, REE ore can be separated into a concentrate, processed into a mixed rare earth solution and elementally separated to oxides by solvent extraction. Rare earth


ntegui / Journal of Cleaner Production 87 (2015) 275e283 R. Lacal-Ara

279

Fig. 3. Price development of REOs used in magnets as shown by the quarterly FOB China prices. Note: Dy and Tb on left axis and Nd and Pr on right axis. Source: Lynas Corporation quarterly reports.

oxides (REO) are ultimately used to produce rare earth metals, alloys and powders. REO production was estimated at 150 000 t in 2009, and recycling constitutes between 4500 and 15 500 t (DoE, 2010) (Lograsso, 2010). The vast majority of REE mining currently occurs in China (89%), the rest in the US, India, Russia, and Australia (USGS, 2014,5). Chinese REO exports are restricted as shown in Table 1. REO world reserves are estimated around 140 million t, of which 40% in China, 16% in Brazil, 9% in USA, 2.2% in India and 1.5% in Australia (USGS, 2014). The demand for REO in 2008 was split between magnets (21%), catalysts (20%), metal alloys (18%), polishing (12%), glass (10%), phosphors (7%), ceramics (6%) and others (6%) (Lograsso, 2010). Estimates suggest that by 2014 permanent magnets will be the main user of REO with 30% of the demand (Lynas, 2011a). Most magnets are manufactured in China (75e80 %) with Japan € (17e25 %) and Europe (3e5 %) well behind (Oko-Institut, 2011). Master patents on NdFeB magnets are controlled by two firms: Hitachi Metals (formerly Sumitomo) in Japan and Magnequench, a former U.S. firm that was sold to a Chinese-backed consortium in 1995. Magnequench merged with Canadian based AMR Technologies in 2005 to form Neo Materials Technologies. It now operates as a Canadian-based company with Chinese operations (DoE, 2010). Superconductor generators are based on coated conductors (CC) which is mostly made of yttrium, barium and copper (YBCO). The challenges include the development of low-cost industrial manufacturing for CC, an improvement in the performance of the cryogenic systems from 20% today to 30%, and a dramatic reduction of the cryogenic cost by a factor of 10 to reach 18 V/W at 70 K (Tixador, 2010). The magnetisation method of the bulk superconductor and low AC loss superconducting wires or tapes for a superconducting stator are still under development. 3. On-going research and players in the field, and challenges World wind energy associations and other key players (Fichaux, 2009; IEA, 2010; TPWind, 2009) suggest that basic and applied research in the area of materials technology should focus on the development of structural and functional materials. Main wind turbine components made from structural materials include the rotor, nacelle machinery and tower, shall be characterised by low mass-to-strength ratio, and at the same time the cost of production

5 It has to be noted that the 2014 issue of this report included a significant correction on the previous year issue, namely the reduction from 7000 to 800 t of the US 2012 estimated production of rare earths.

shall be reduced as well. To fulfil these goals main basic and applied research challenges shall be focused on the tower top weight: rotor (hub þ blades), drive-train, gearbox, generator, pitch system, nacelle bedplate, yaw drive and the tower. For offshore turbines the support structure and transition piece have to be added to this list. The EU sixth Framework Programme (FP6) UpWind project explored from this perspective the design of wind turbines of eight to ten MW rated capacity. Part of the research was targeted at possible barriers that might be encountered when designing a 20 MW turbine with a rotor diameter of 256 m. The FP7 project RELIAWIND focussed in general on the reliability of wind turbines and the development of related testing procedures. Electromagnet is a very established generator technology with few possible breakthrough research proposals. Exploratory wire research aims at increasing conductivity through nanotechnology e.g. the Ultraconductus project (LANL, 2011). The general lines for permanent magnet research focus on performance, cost and recycling. Performance research include higher energy products (discovery of new materials); better high temperature performance (substitution/use of additives); higher electrical resistivity (e.g. reducing eddy current losses/use of additives); and enhanced mechanical properties (additives/processing). Cost-related research focuses on raw materials costs; decrease or eliminate RE constituents (new materials/nanoprocessing); and processing costs (direct conversion, selective separations). Finally, recycling research approaches include manufacturing scrap e swarf (processing/selective separations); and post-consumer (processing) (Lograsso, 2010).

Table 1 China's REE export quotas and demand from the rest of the world (RoW). Sources: DoE (2010); TMR (2011); USGS (2013, 2014). Year

Export quotas (t REO)

Change from previous year

RoW demand (t)

RoW supply (t)

2005 2006 2007 2008 2009 2010 2011 2012 2013

65 61 59 56 50 30 30 30 30

e 6% 4% 5% 12% 40% 7%a þ2% 0%

46 50 50 50 25 48

3850 3850 3730 3730 3730 6700 7000 9700 11 700

609 821 643 939 145 258 246 996 999

000 000 000 000 000 000

a The 2011 figure cannot be directly compared to the 2010 one because one additional category was included in the 2011 quota, ferroalloys. The figure of 7% results from a deeper analysis by Lynas Corp. (Lynas, 2011b).

280


ntegui / Journal of Cleaner Production 87 (2015) 275e283 R. Lacal-Ara

Fig. 4. Turbine sizes at time of market introduction (Beurskens, 2010).

Ongoing research aims at reducing the REE content of magnets e.g. by substituting REEs e.g. at the grain boundaries (Kobe, 2010); increasing their efficiency, magnetic density and temperature range. This is partly done by searching alternative materials. Basic science research focuses on the synthesis of highest quality polycrystals and single crystals, advanced characterization methods, especially neutron and magnetic X-ray scattering and first principles modelling (Ames Laboratory, US); anisotropic magnets based on the hightemperature, rare-earth-based alloy previously designed for isotropic bonded magnets (Ames); bulk nano-structured magnetic materials (General Electric-GE-in the US). Applied research includes higher-flux density PM generators (QM Power in the US), magnets based on Fe-, Co-, or Mn-rich materials (Univ. of Delaware, US); 640 kJ/ m3 and 80% less rare-earth mineral content (GE in the US); gas-phasemodified TM materials and FeNi phases with the L10 structure (both at Tohoku University, JP) (Kobe, 2010); high-performance anisotropic nanocomposite PM with low RE content (Hitachi in JP). Japan's Rare Metal Substitute Materials Development Program targets a 30% reduction in the Dy and Nd content of PM through grain refinement and nanostructure techniques, and REE-less/free alloys or other elements. Korea has developed since 2007 a comprehensive research programme supported by an annual budget of one billion USD. The main research institutions in Europe are the Materials Innovation Institute (M2i, NL); Leibniz Institute for Solid State and Ma
el, Grenoble (FR); Trinity terials Research, Dresden (DE) Institut Ne €lten University of Applied Sciences (AT); College, Dublin (IE); St. Po Vienna University of Technology (AT); Jo zef Stefan Institute, Ljubljana, (SI); Vacuumschmeltze, GmbH, (DE); Siemens GmbH, (DE). Applied research and demonstration include direct current (DC) generators which reduce the mass of the active part of a DD-PMG by up to 25% (Converteam, 2010). Incidentally, a DC generator in a DC wind farm grid has the potential to increase output by reducing losses from AC/DC/AC power conversion. Superconductor research includes reducing the cost of wires: more efficient processes to deposit the layers of YBCO (YBa2Cu3O7), the superconducting cuprates that form coated conductors (Risø). Specific HTS research projects include Risø's Superwind project (DK); Converteam/Zenergy research on an 8-MW HTS generator; AMSC/ TECO Westinghouse on components for a 10-MW HTS generator. 4. Target materials specification for 2020/2030 and 2050 Realising the international targets for the implementation of wind energy means that both on- and offshore wind energy must

become economically competitive with fossil fuel and nuclear power generation. Therefore the current trends, the scaling up of the wind turbines, see Fig. 4, and of wind farms, needs to be continued in the considered time frame. To achieve the required reduction in investment and maintenance costs, breakthroughs are required in the technology development of wind turbines. Especially for offshore wind the need to reduce investment, operation and maintenance costs puts high demands on reliability, accessibility and operation and maintenance strategies. Achieving the required high level of reliability and availability of almost 100% requires condition monitoring of all critical components. Condition monitoring facilitates that preventive, well-planned maintenance is carried out in periods of average low wind speeds when power generation losses are relatively low. To lower the initial investment costs all components must be optimised with respect to functionality and material consumption. New, more demanding functional and structural designs will put higher demands on reliability of the design process and the materials used. Since wind energy is a renewable energy source, environmental considerations have played a key role in its support by society. For consistency reasons, the recyclability of all its components and materials has to be strived for. For most of the metallic components this has already been achieved nowadays. However for a number of other parts like electronic components and especially the blades this is not yet the case. For the development of recycling processes for electronic components wind energy can probably rely on the developments in the large world market of this type of components, since these are used in a lot of other applications too. However for the blades this has to be specifically a wind energy development. Since no other market sector uses fibre reinforced components with the size, quantity and wall thickness as wind turbine blades do. 4.1. Target specification for electricity generators in 2020/2030 Literature research, experts' opinion and stakeholders' consultation suggest that the possibilities for higher performance electricity generators fall mainly in two kinds of generators, permanent-magnet (PMG) and high-temperature superconductor (HTS) generators. The following list determines technical indicators and the values they should ideally achieve for commercial operation by 2020/ 2030:


ntegui / Journal of Cleaner Production 87 (2015) 275e283 R. Lacal-Ara

➢ Direct-drive generators:

281

of 5.5114 by 2020 and 5.4889 by 2030, and of the coefficient B of 11.045 by 2020 and 12.702 by 2030.

B Weight: a 12-MW direct-drive generator should weigh less than 170 t by 2020, and less than 130 t by 2030. Current estimates put a 10-MW PMG at 300 t, initial data by AMSC (2009) suggested 120 t would be possible for similarlysized HTS technology. B There should be a performance indicator “specific cost”; unfortunately price information is highly sensible, which would make the value of this indicator difficult to find. ➢ Permanent magnets: specifications for commercial magnets (military magnets have higher grades)

B Rare earth content in mass at equal magnetic density should aim at a reduction for neodymium from the current 30e32 % to around 28% by 2015, 25% by 2020 and 20% by 2030; for dysprosium today there is a large margin of 3e6 %, should aim at 2% by 2015, 1.8% by 2020, and 1% by 2030.

B Power density range increasing from 270e422 kJ/m3 in the most performing magnets of today to 360e500 by 2020 and 460e535 kJ/m3 by 2030.

After this proposal was put forward, discussions with the industry and academic experts suggested that these specifications are highly ambitious, and that the current NdFeB compositions of magnets are unlikely to ever reach the proposed specifications by 2030. Therefore, new magnet compositions will be necessary that could or could not be based as currently in rare earth elements, which could be substituted by new chemical components.

B Remanence (HcJ) versus coercivity (Br).

➢ High-temperature superconductors

Remanence and coercivity are two physical characteristics that reflect very well the trade-offs needed in the design of permanent magnets for use in generators and motors, e.g. magnetic power versus resistance to demagnetisation from higher temperatures. For this reason an indicator that includes both characteristics is necessary as a way to better monitor overall performance improvements. The current highest-performing magnets from several manufacturers were plotted in a HcJ/Br curve which we call HcJ(2010) (see Fig. 5) The function that relates them was found to be linear (y ¼ A x þ B) and therefore it conveniently reflects the trade-offs among these two key parameters of a permanent magnet. The values of this function are:

HJcJð2010Þ ¼ 5:5581Br þ 9:2671 In this equation and Fig. 5, HcJ is expressed in MA/m and Br in Tesla. Future specifications of this relationship are reflected in that figure for 2020 and 2030. The specifications proposed for 2020 involve a 20% improvement over the current specifications, and a further 15% is proposed for 2030. This proposals result in values of the coefficient A

B Superconducting wire cost should aim to decrease from around 300 V/kA-m today to 30e40 V/kA-m by 2020 and below 15 V/kA-m by 2030 B The corresponding superconductor operational temperature should aim to increase from 70 K today to 85 K by 2015 at laboratory level and by 2020 in the market. The reduction of component mass has the knock-on effect of enabling the optimisation and consequent materials and mass reduction of supporting structures (e.g. the bedplates, nacelle, tower, foundations), which is particularly important for offshore substructures. Work by Converteam (Hill, 2011) suggests that nacelle mass savings in an 8 MW, 12 rpm generator design are multiplied by up to four times in the tower and offshore foundation structures. When consulting with different companies working in HTS, there seemed to be general support for the claim that costreduction of superconducting wire depends exclusively on economies of scale of wire manufacture, in what seems to be a catch-22 situation: because there is no use at such high prices there is no investment in large-scale production facilities that would reduce the cost, and vice versa.

Fig. 5. Proposed specifications for key permanent magnet parameters by 2020 and 2030.

282


ntegui / Journal of Cleaner Production 87 (2015) 275e283 R. Lacal-Ara

In addition to the target levels described for individual components, the integration of components, e.g. power electronics integrated in the generator or a gearbox-generator, offers further possibilities for using less materials. 4.2. Target specification for 2050 Any attempt to describe what the 2050 specification should be is merely speculative; the main element of uncertainty is possibly the introduction of scientific breakthroughs. Within these limitations it is suggested that the weight of a 20MW direct-drive generator should be less than 200 t by 2050, which should possible at least with HTS technology. Another objective for 2050 is the development of low-cost manufacturing processes for coated HTS. 4.3. Synergies with other technologies The achievement of those materials specifications will be supported by the use of the same materials in other industrial sectors. Wind turbine generators, whether electromagnets, permanent magnets (PM) or high-temperature semiconductors, present synergies with many sectors using generators or motors including hydropower, marine currents, cogeneration, vessels (generation and propulsion); train traction; and industrial sectors (pumps, fans, crushers, mills, rolling motion, shredders, etc.). Of upmost importance are synergies with the electric vehicle sector because of its projected development: according to DoE projections electric vehicles demand of neodymium (Nd) will outweigh wind turbines by 4 times by 2025, and of dysprosium (Dy) by 3 times (DoE, 2010). Beyond this, all industrial uses of motors, other power generation subsectors, and propulsion machines (e.g. for ships) present strong potential synergies with the wind energy sectors. Other sectors competing for rare earths include catalysts, NiMH batteries electronics, glass, ceramics, metal alloys and laser technologies (Smith, 2010). The manufacturing of PM from rare earths presents few synergies because of its specific manufacturing process. Those synergies would include other metal-oriented manufacturing processes. 5. Conclusions: needs and recommendation of activities to achieve the specifications In a context in which the support for renewable energy aims to achieve a secure and affordable supply of clean energy, the improvements of the specifications in the materials used in wind power electricity generators has naturally to contribute to a lower cost of wind electricity. This paper explores, discusses and proposes the optimum specifications of materials that can give electricity generators built by 2020 the improved characteristics needed to reach the policy objective of a high penetration of wind energy in the European electricity system. This paper proposes targets specifications for materials density by 2020 and beyond, e.g. a 12-MW direct-drive generator of less than 170 t by 2020, and less than 130 t by 2030. An increase in the power density range of permanent magnets is proposed as well, from 270e422 kJ/m3 in today's most performing magnets to 360e500 by 2020 and 460e535 kJ/m3 by 2030; and a significant increase in the remanence versus coercivity ratio of these magnets. In order to monitor the latter improvement, this paper proposes an empirical formulae based on the currently best performance magnets. In addition, for high-temperature superconductor generators research should aim at bringing down the price of superconducting

wire from around 300 V/kA-m today to 30e40 V/kA-m by 2020 and below 15 V/kA-m by 2030, and to increase the temperature of commercial operation to 85 K. The former seems to depend heavily on economies of scale of wire manufacture. Because the basic and applied research on both higher-efficiency use, and substitution, of rare earth elements in permanent magnets presents high risk and has a European scope (and beyond), the European Union common research mechanisms should allow European firms and research centres to share and optimise the cost of this research. Government funding of pre-competitive research on enabling material alternatives combined with international coordination; early (financial) involvement of industry to steer and apply material research. The practical objectives should be reducing rare earth content in permanent magnets, use of alternative materials for permanent magnet manufacture, and design for recycling. The kind of research needed to obtain breakthroughs is plagued with risks. However, most research programmes are adverse to risks, but no risks means no or a very low possibility for breakthroughs. Some programmes, in particular in basic science, should be re-designed to allow for higher risks. Acknowledgements The scientific assessment of materials specifications for wind turbines was commissioned by the European Commission's DirectorateGeneral for Research and Innovation to a group of experts led by. L. G. J. Janssen, and including the author of this paper (in charge of the electricity generator aspects), Paul Brøndsted, P. Gimondo, A. Klimpel, B. B. Johansen and P. Thibaux. I would like to show my high appreciation to these co-authors for their guide and support. The draft scientific assessment was presented and discussed in a stakeholders meeting organized by de European Commission on the 4th of April of 2011 which included the presence of Albert Kriener (E.ON); Jan Schamp (Hansen Transmissions); Pep Prats (Alstom Power-TPWind); Jos Beurskens (Energy research Centre of the Netherlands-TPWind); Athanasia Arapogianni and Filippo Gagliardi (EWEA-TPWind); Jaap de Boer (Suzlon); Bertrand de Lamberterie (Steel Research, ESTEP); John F. Hill (Converteam); Jose Maria Albuquerque (ISQ-EUMAT) and Amaia Igartua (TEKNIKEREUMAT); Brian Norton (Indestructible Paint-BCF) Martijn Geertzen (NEN-CENELEC); Larissa Gorbatikh (K. U. Leuven); and George Kariniotakis (MINES ParisTech). The comments made in this meeting and the subsequent exchanged were used to update this assessment. In addition to the former, we would like to acknowledge the very valuable contribution from the following persons: Alain Guy (Converteam), H. Polinder (T. U. Delft), J. F. Reyes (Atlas Magnetic ~ o (European Group), J van Rossum (Dredging International), F. Nun Copper Institute), N. Athanassopoulou (Cambridge University), L. Vandebbossche (Arcelormittal). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jclepro.2014.09.047. References AMSC, 2009. AMSC and U.S. Department of Energy Agree to Collaborate on 10 Megawatt-class Superconductor Wind Turbines. Press release 10/02/2009, available at: http://ir.amsc.com/releasedetail.cfm?releaseid¼642764 (accessed 20.08.14.). Ancona, D., Mc Veigh, J., 2001. Wind Turbine e Materials and Manufacturing Fact Sheet. US Department of Energy, pp. C-1eC-8 (accessed 23.04.14.) at. http://


ntegui / Journal of Cleaner Production 87 (2015) 275e283 R. Lacal-Ara www.perihq.com/documents/windturbine-materialsandmanufacturing_ factsheet.pdf. APS, February 2011. American Physical Society Panel on Public Affairs/Materials Research Society, US: Energy Critical Elements: Securing Materials for Emerging Technologies. Available at: http://www.aps.org/policy/reports/popa-reports/ loader.cfm?csModule¼security/getfile&PageID¼236337. Beurskens, J., 2010. Developing Offshore Wind Energy now and in the future. In: Third We@Sea Conference, IJmuiden, November 25, 2010. Available at: http:// www.we-at-sea.org/leden/docs/conference2010/1.pdf (accessed 27.02.11.). BNEF, 2nd June 2010. Wind Energy Research Note: Driving to 5 MW and beyond. Bloomberg New Energy Finance. BTM Consult, 2011. World Market Update 2010. BTM Consult AsP, Copenhagen. Converteam, 2010. Press Release 6th December 2010: “Converteam and Scottish and Southern Energy Join Forces on Offshore Wind Energy” to Demonstrate a DC Generator that Can Be Integrated in a DC Array and Connected to the Shore through a HVDC Cable. Daim, T.U., Amer, M., Brenden, R., 2011. Technology Roadmapping for wind energy: case of the Pacific Northwest. J. Clean. Prod. 20, 27e37. http://dx.doi.org/ 10.1016/j.jclepro.2011.07.025. DoE, 2010. Critical Materials Strategy. United States Department of Energy. December 2010. EC, 2007. Communication from the Commission to the Council, the European Parliament, the European Economic and Social Committee and the Committee of the Regions - a European Strategic Energy Technology Plan (SET-plan)-towards a Low Carbon Future. COM/2007/723. European Commission. EC, 2011. Commission Staff Working Paper SEC(2011) 1609 Final: Materials Roadmap Enabling Low Carbon Energy Technologies. Brussels, Belgium. Available at. : http:// setis.ec.europa.eu/newsroom-items-folder/materials-roadmap-enabling-lowcarbon-energy-technologies-published. EWEA, November 2010. Powering Europe: Wind Energy and the Electricity Grid. European Wind Energy Association, Brussels. Fichaux, N., 2009. Delivering Today the Energy of Tomorrow. EWI e The European Wind Initiative Eurpean Wind Energy Technology Platform. July 2009. http:// www.windplatform.eu/fileadmin/ewetp_docs/Events/European_Wind_ Initiative_-_JP.pdf. Frost, Sullivan, 2010. Copper Content Assessment of Wind Turbines, De-brief Meeting. Presented to the European Copper Institute, 12th July 2010. GWEC, 2014. Global Wind Statistics 2013. Global Wind Energy Council, Brussels. February 2014. Habib, K., Wenzel, H., 2014. Exploring rare earths supply constraints for the emerging clean energy technologies and the role of recycling. J. Clean. Prod. 84, 348e359. http://dx.doi.org/10.1016/j.jclepro.2014.04.035. Hansen, A.D., 2005. Generators and power electronics in wind turbines. In: Ackermann, T. (Ed.), Wind Power in Power Systems. John Wiley & Sons, Ltd, Stockholm, p. 745. Hatch, G.P., 2011. Critical Rare Earths e Global Supply & Demand Projections and the Leading Contenders for New Sources of Supply. Technology Metals Research. August 2011. Hill, J.F., 2010. Wind Turbines e the Era of the Permanent Magnet Generator. Presented at the seminar on “Permanent Magnets e The Rare Earth Crisis” of the UK Magnetics Society, 2nd December 2010. Hill, J.F., 2011. Marketing Manager, Converteam Renewables. Email Communications, 7th and 24th February, and 6th April 2011. IEA, 2008. Energy Technology Perspectives 2008, Scenarios and Strategies to 2050. International Energy Agency, Paris. IEA, 2010. Global Gaps in Clean Energy RD&D. Update and Recommendations for International Collaboration. IEA Report for the Clean Energy Ministerial. JRC, 2011. Internal Database of Wind Turbine Models and Installations. Joint Research Centre of the European Commission. Kobe, S., 2010. Presentation on Rare Earth Magnets in Europe at the Trans-Atlantic Workshop on Rare Earth Elements and Other Critical Materials for a Clean Energy Future. Jozef Stefan Institut, Slovenia, 3rd December 2010.

283

Kingsnorth, D., (26.07.2013). Kingsnorth Believes Rare Earth Element Demand and Prices to Increase e See more at: http://investorintel.com/rare-earth-intel/ dudley-kingsnorth-believes-that-ree-demand-and-prices-will-double-by2015/#sthash.CdpRoYgW.dpuf. (Retrieved 06.05.14.), from Investor Intel: http:// investorintel.com/rare-earth-intel/dudley-kingsnorth-believes-that-reedemand-and-prices-will-double-by-2015/ LANL, 2011. Los Alamos National Laboratory: Ultraconductus, 2010 R&D 100 Awards Entry Summaries consulted on 25/02/11 at. http://www.lanl.gov/source/orgs/tt/ awards/rd100/summaries/10_execsum.shtml. Lacal-Ar
antegui, R., 2014. 2013 JRC Wind Status Report. Publications Office of the European Union, Luxembourg. http://dx.doi.org/10.2790/97044.
ntegui, R., Aguado-Alonso, M., 2011. Power electronics Llorente-Iglesias, R., Lacal-Ara evolution in wind turbines e a market-based analysis. Renew. Sustain. Energy Rev. 15, 4982e4993. http://dx.doi.org/10.1016/j.rser.2011.07.056. Lograsso, T., 2010. In: Presentation at the Trans-Atlantic Forum, Boston (US) 3rd December 2010, Citing IMCOA as the Source. T. Lograsso from the Ames Laboratory, US DoE. Lynas, 2011a. In: Lynas Corporation Ltd., Investors Presentation May 2011, where it Estimated that by 2014 Magnets Would Consume 50 kt of Rare Earths or 28th of that Year's Total REE Demand. Lynas, 2011b. In: Lynas Corporation Ltd. Press Release 15th July 2011: Chinese Rare Earths Export Quota Released for Second Half 2011. Available on line at. http:// www.asx.com.au/asxpdf/20110715/pdf/41zshtqzvgbdx9.pdf. consulted 3rd January 2012.
nez, E., Blanco, J., 2009. Life cycle assessment Martínez, E., Sanz, F., Pellegrini, S., Jime of a multi-megawatt wind turbine. Renew. Energy 34, 667e673. MII, 2010. In: Presentation of the Materials Scarcity Report and Industrial Perspectives at the Trans-Atlantic Forum. Materials Innovation Institute, Boston (US) 3rd December 2010. € Oko-Institut, January 2011. Study on Rare Earths and Their Recycling. Final report for The Greens/EFA Group in the European Parliament. Darmstadt. PowerGen, 1st February 2011. Superconducting Generator: One Small Step? PowerGen World Wide (accessed on 15.02.11.) at. http://www.powergenworldwide. com/index/display/articledisplay/8435051715/articles/power-engineeringinternational/volume-18/issue-2/features/superconducting-generator.html.
ctrica de REE, July 2014. The Spanish Electricity System Annual Report 2013. Red Ele ~ a. Available at: http://www.ree.es/en/publications/spanish-electricalEspan system/spanish-electricity-system-2013 (accessed 20.08.14.). Reyes, J.F., 2011. Atlas Magnetics Europe B.V., Personal Conversation. Siemens, 2011. Wind Power Technical Solutions. Available at: http://www. automation.siemens.com/mcms/large-drives/en/ld/solutions-industry/ windpower/Pages/windpower.aspx. consulted 25/02/11. Smith, M.A., 2010. In: Testimony of Mark A. Smith, CEO, Molycorp Minerals LLC, to the Subcommittee on Investigation and Oversight, House of Science and Technology Committee, US Congress 16.03.10. Tixador, P., 2010. Development of superconducting power devices in Europe. Phys. C. 470, 971e979. TMR, 2011. Different Entries in the Blog Technology Metals Research. Available at: http://www.techmetalsresearch.com/. TPWind, 2009. The Way Forward. European Wind Technology Platform, Brussels. March 2009. USGS, 2013. Survey Mineral Commodity Summaries 2013. US Geological Survey. Available at: http://minerals.usgs.gov/minerals/pubs/mcs/2013/mcs2013.pdf (accessed January 04.06.14.). USGS, 2014. Survey Mineral Commodity Summaries 2014. US Geological Survey. Available at: http://minerals.usgs.gov/minerals/pubs/mcs/2014/mcs2014.pdf (accessed January 04.06.14.). Vestas, 2009. Lifecycle Assessment of a V90-3.0 MW Offshore Wind Turbine. Vestas Wind Systems A/S, Copenhagen.