Science of the Total Environment 832 (2022) 155022

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Global environmental cost of using rare earth elements in green

energy technologies
Saeed Rahimpour Golroudbary a, , Iryna Makarava b,c,
⁎ ⁎⁎, Andrzej Kraslawski a,⁎⁎⁎, Eveliina Repo b
a
School of Engineering Science, Industrial Engineering and Management (IEM), LUT University, FI-53851 Lappeenranta, Finland
b
School of Engineering Science, Department of Separation Science, LUT University, FI-53850 Lappeenranta, Finland
c
Helmholtz-Zentrum Dresden-Rossendorf, Helmholtz Institute Freiberg for Resource Technology, D-09599 Freiberg, Germany

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Generation of green energy must be ana-

lyzed in a broad environmental context.
• Increase of green energy production by
1% causes 0.18% depletion of REE re-
serves.
• Growth of green energy production by 1%
leads to 0.90% growth of GHG emissions.
• Mining of known REEs deposits is not suf-
ficient to ensure sustainable development.

A R T I C L E I N F O A B S T R A C T

Editor: Daniel CW Tsang Decarbonization of economy is intended to reduce the consumption of non-renewable energy sources and emissions
from them. One of the major components of decarbonization are “green energy” technologies, e.g. wind turbines
Keywords: and electric vehicles. However, they themselves create new sustainability challenges, e.g. use of green energy contrib-
Green energy utes to the reduction of consumption of fossil fuels, on one hand, but at the same time it increases demand for perma-
Energy supply
nent magnets containing considerable amounts of rare earth elements (REEs). This article provides the first global
Critical materials
Decarbonization measures
analysis of environmental impact of using rare earth elements in green energy technologies. The analysis was per-
Environmental sustainability formed applying system dynamics modelling methodology integrated with life cycle assessment and geometallurgical
approach. We provide evidence that an increase by 1% of green energy production causes a depletion of REEs reserves
by 0.18% and increases GHG emissions in the exploitation phase by 0.90%. Our results demonstrate that between
2010 and 2020, the use of permanent magnets has resulted cumulatively in 32 billion tonnes CO2-equivalent of
GHG emissions globally. It shows that new approaches to decarbonization are still needed, in order to ensure sustain-
ability of the process. The finding highlights a need to design and implement various measures intended to increase
REEs reuse, recycling (currently below 1%), limit their dematerialization, increase substitution and develop new elim-
ination technologies. Such measures would support the development of appropriate strategies for decarbonization and
environmentally sustainable development of green energy technologies.

⁎ Corresponding author.
⁎⁎ Correspondence to: I. Makarava, School of Engineering Science, Department of Separation Science, LUT University, FI-53850 Lappeenranta, Finland.
⁎⁎⁎ Corresponding author.
E-mail addresses: Saeed.Rahimpour.Golroudbary@lut.fi (S.R. Golroudbary), Iryna.Makarava@lut.fi (I. Makarava), Andrzej.kraslawski@lut.fi (A. Kraslawski).

http://dx.doi.org/10.1016/j.scitotenv.2022.155022
Received 25 January 2022; Received in revised form 17 March 2022; Accepted 31 March 2022
Available online 4 April 2022

0048-9697/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

1. Introduction The novelty of this study consists in conducting a holistic assessment of

the quantitative impact of green energy technologies on other natural
Decarbonization has emerged as an important measure to keep global resources, i.e. rare earth elements and assessing their side effect and dy-
average temperature increase well below 2 °C above pre-industrial levels namic changes over time in the period 2010–2030. This work, unlike the
according to the goal agreed in the Paris Agreement (Waisman et al., previous publications, offers a detailed analysis of the subsequent stages
2019). Therefore, the transition from fossil fuels to renewable energy of rare elements life cycle. It studies material flow in the supply chain of
sources considerably accelerated in the last decade (increase by 16.4% fifteen rare earth elements. Moreover, this study considers 16 types of
per annum between 2007 and 2017 (Dudley et al., 2018)). However, the ores existing in the deposits (Bastnäsite, Monazite, Xenotime, Euxenite,
development of ‘green energy’ technologies has resulted in the growth of Apatite, Gadolinite, Loparite, Uraninite, Brannerite, Dolerite, Pyrochlore,
consumption of natural resources, e.g. growing use of ‘permanent magnets’, Perovskite, Zircon, Clay, Synchysite, and Parisite) while analyses offered
composed in 25–30% of critical materials, such as rare-earth elements by other publications are limited mainly to 1 or maximum 3 types of ores
(REEs). The ‘permanent magnets’, called Neodymium magnets (NdFeB), (usually Bastnäsite, Monazite, or ion-adsorption clays). Several studies
are used in wind turbines and direct drive motors in electric vehicles have been devoted to the growing global demand for REEs, e.g. (Dutta
(EVs) (Fishman and Graedel, 2019). et al., 2016). Some of them (Du and Graedel, 2013; Graedel et al., 2015)
As advocated by the Nature Reviews Materials presented in a recent se- suggest that supply risk can be expected for several elements such as Eu,
ries of articles (Herrington, 2021; Raw materials for a truly green future, Dy, and Er. This estimation results from the scarcity of resources. Hence,
2021), environmentally and socially sustainable ways of sourcing raw ma- the growing global demand for these elements stemming from their impor-
terials needed to meet emission reduction targets must be established. We tance in shifting towards a green economy, strongly encourages exploration
address those calls by showing that decarbonization of economy requires geochemists and technology developers to conduct new research to ensure
a much deeper analysis of environmental costs as it increases the consump- a sustainable supply of REEs. Therefore, there is a need for tracing the envi-
tion of critical materials. Our quantitative analysis provides an insight into ronmental cost of REEs consumption embodied in the global supply chains
the key question: to what extent the use of green energy is a sustainable from the primary sources to their use in green energy technologies. It
solution. should help in assessing the real environmental costs of transition towards
The existed literature has highlighted methodological challenges, high green economy. This study covers a comprehensive environmental assess-
uncertainty and lack of knowledge about the environmental impact of ment to address the discussed issues.
REEs production (Bailey et al., 2020; Schreiber et al., 2021). Several studies Using system dynamics modelling (Forrester, 1997), we quantify the
on the environmental impact of rare earth elements conducted using the key environmental concerns in mining and processing stages of fifteen
life cycle assessment (LCA) are summarized in Table 1. According to the lit- REEs including yttrium (Y) and lanthanoid group (except promethium
erature review, most studies have evaluated the carbon footprint of a spe- (Pm) that is radioactive, and does not have stable isotopes. In the current
cific stage of REEs production (mining and refining) or manufacturing of study, all REEs are considered in every stage of the supply chain. At the mo-
specific products (e.g. batteries or magnets production and its impact at ment, no technology exists that would allow to carry out a selective recov-
the local scale). Unfortunately, they do not allow for a holistic analysis of ery of selected REEs needed for specific applications. Therefore, we
environmental impacts of green energy value chain, e.g. energy and analyzed the life cycle of all rare earth elements in mining and processing
water consumption within all processes involving REEs. There are several stages, and focused on elements required for production of wind turbines
works addressing fragmentarily the problem of environmental impact of and EVs in the manufacturing stage. The issues that we intended to address
REEs supply chain. For example, previous studies on the environmental im- include: REEs deposits depletion, energy and water consumption, as well as
pact of REEs life cycle assess the recycling stage only, e.g. in the case of elec- related emissions. Energy sources used in mining and processing stages are:
tronic waste (Li et al., 2019; Liu and Keoleian, 2020), and electric vehicles crude oil, bitumen, shale oil, natural gas, coal, forest residues, pet coke, hy-
(Jin et al., 2018; Tintelecan et al., 2019). Considering the application of droelectric power, nuclear energy, geothermal sources, solar and wind
REEs in wind turbines, (Arshi et al., 2018) provide the environmental as- power. Water consumption considerations cover reservoir evaporation,
sessment of REEs limited to the regional scale of mining of monazite/ cooling, mining, and processing. We assess the following emissions CO2,
bastnasite deposits and ion adsorption clays from China and two CH4, SOx, NOx, CO, N2O, VOC, PM2.5, POC, BC, and PM10 at all stages of
manufacturing facilities located in Bayan Obo and South China. Gwenzi green energy supply chain.
et al. (2018) reviewed and highlighted the lack of quantitative analysis of In the production stage, we quantify the environmental impact (global
REEs flows from several perspectives e.g. environmental and human health warming, ecotoxicity, acidification, eutrophication, carcinogenic effects,
risks. Bailey et al. (Bailey et al., 2020) offer a limited analysis of the mining ozone depletion, and smog emissions) of the production of permanent mag-
stage by focusing only on 3 types of minerals: bastnäsite, ion adsorption nets.
clays and lateritic monazite. In contrast to the above-presented examples, We analyze trends observed in using EVs and installed wind power,
this study offers a complete supply chain analysis of all REEs elements on price, natural resource depletion and greenhouse gas (GHG) emissions.
the global scale for the years 2010–2030. Moreover, the significance of an The analysis covers years from 2010 to 2030. This period was selected to
integrated life cycle assessment and geometallurgical approach for reduc- explore the trends related to rare earth elements crisis caused by China ex-
ing the environmental impact of the extraction of raw materials essential port limitations and trends emerging in REEs consumption.
for the transition to a low-carbon economy has been demonstrated in the re-
cent review analysis by Pell et al. (2021). Most of the previous studies fo- 2. Materials and methods
cused only on China as the primary source of REEs ores. However,
nowadays there are various mines around the world extracting those ele- 2.1. Model of the global green energy supply chain
ments and their operation needs to be considered in a global sustainability
assessment. It is worth mentioning that several review articles analyzing Systematic environmental analysis of sustainability of the global supply
differences in mining and processing of REEs have been published recently chain of green energy technologies is still missing. Moreover, life cycle as-
(Pell et al., 2021; Schreiber et al., 2021). Nevertheless, there is a significant sessment of REEs is particularly sensitive to methodological differences,
gap in the literature as none of these studies discusses historical changes in e.g., variations in allocation procedures can significantly change the envi-
the mass flow of REEs through all the supply chain stages or estimates the ronmental impact per kg of product. Therefore, our analysis of REEs pro-
dynamics of trends in the near future. Therefore, a comprehensive study duction relies on both geology and processing route.
of the environmental impact of all supply chain stages for green energy This study proposes a dynamics model consisting of global REEs flows,
technologies and relevant decarbonization measures for REEs' mining and the use of respective resources, and emissions. The model consists of 1465
production at the global scale in a long-term perspective is still missing. variables and parameters. The Anylogic software (www.anylogic.com) is

2
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

Table 1
Summary of investigation on the environmental impact of rare earth elements used in green energy technologies.
Reference Objective of study Methodology Supply chain scope Environmental Impact Geographical
scope
Mining Processing Manufacturing Recycling Energy Water Emissions

(Arshi et al., 2018) To assess the environmental LCA ✓ ✓ ✓ ✓ Local: China

footprint of rare earth products
(Lee and Wen, 2018) To analyze the environmental LCA and scenario building ✓ ✓ ✓ Local: China
impact of the supply of rare
earth elements in China
(Jin et al., 2018) To compare LCA ✓ ✓ ✓ Local: Case of
Neodymium-Iron-Boron manufacturing
magnet-to-magnet recycling for
electric vehicle motors
(Bailey et al., 2020) To assess life cycle of REEs to LCA ✓ ✓ ✓ Local:
provide information about data Australia
gap
(Zaimes et al., 2015) To evaluate the environmental LCA ✓ ✓ ✓ ✓ Local: China
impacts and resource intensity
of producing rare earth oxides
(Vahidi et al., 2016) Life cycle assessment on mixed LCA ✓ ✓ ✓ Local: China
rare earth oxides production
from ion adsorption clay
(Deng and Kendall, Life cycle assessment of heavy LCA ✓ ✓ ✓ Local: China
2019) rare earth oxides from
ion-adsorption clays
Current study Global environmental impact of System dynamics modelling ✓ ✓ ✓ ✓ ✓ ✓ Global
rare earth elements used in integrated with LCA and
green energy technologies geometallurgical approach

used for simulation. The main variables including flow, stock and auxiliary Survey (USGS) data sources (Gambogi, 2021). Therefore, the model covers
variables are presented in Table S1 and parameters together with all data two periods: the historical (1996–2020) and the future one (2020−2030).
required to run the model are available in Table S2 and Data S1 (#2, #4, Growth rate of the estimated primary production of REEs is based on future
#5 and #6). Fig. S4 provides details of the dynamic model and used data demand by 2030 (Hagelüken, 2014; Wang et al., 2020a).
sources. We divided the variables of the dynamic model into two groups in- The second stage of the model comprises all manufacturing streams
cluding endogenous and exogenous variables to specify the model bound- where REEs are applied. The focus of environmental analysis in this phase
ary. Endogenous variables affect and, at the same time, are affected by is on production of permanent magnets for wind turbines and EVs applica-
other system components and parameters, while exogenous variables are tions.
not directly affected by the system. Groups and types of all variables are
presented in Table S1. 2.2. Mathematical formulation
The model is composed of two main parts: (i) extractive stage which in-
cludes mining and processing of REEs and (ii) production stage of The model consists of two types of equations (state and rate) for quanti-
manufacturing permanent magnets applied in wind turbines and EVs. It ex- fying the stocks and flows in given systems. Results from the model are next
amines environmental impact of each process including energy consump- used as inputs to geographical distribution equations.
tion, water use, CO2, CH4, SOx, NOx, CO, VOC, PM2.5, N2O, POC, BC, and Stocks (state equations) in the mass flows of the model can be used to
PM10 emissions. analyze the environmental performance of countries involved in mining
In the first stage, REEs containing minerals are extracted, then sepa- and processing stages, as well as global green energy production over
rated, processed and transformed into chemicals. It should be noted that time. Flows (rate equations) correspond to the production of REEs,
the various types of deposits have particular characteristics of size and manufacturing of products, energy consumption, water use, and related
grade. For example, carbonatites tend to be medium to large tonnage and emissions.
high grade, whereas alkaline rock deposits are generally larger tonnage Eq. (1) corresponds to the global stock of REO (GRs(t)) over the period
but lower grade, tending to have higher proportions of heavy REE. Mineral “t0-t” where“t0” is the initial year and “t” is the final year. TMa−i (t) repre-
sands are low grade and REE minerals are by-products; ion adsorption de- sents the annual production rate of REO from mining by country “i”
posits are small and low grade but relatively rich in heavy REE (where i= 1,2, …,12) including Australia, Brazil, Myanmar, Burundi,
(Goodenough et al., 2016). Also, multiple stages are used for separation China, India, Madagascar, Malaysia, Russia, Thailand, US, and Vietnam;
of the individual REE, e.g., selective oxidation can be used for Ce, and REEa−ji(t) stands for annual production of element “j” (where j= 1,2,
also for Pr and Tb, which have a potential +4 oxidation state, whilst selec- …,15) including Cerium, Dysprosium, Erbium, Europium, Gadolinium,
tive reduction works for Eu, Sm and Yb because they have a potential +2 Holmium, Lanthanum, Lutetium, Neodymium, Praseodymium, Samarium,
valency (McNeice and Ghahreman, 2018; Strauss et al., 2019). In process- Terbium, Thulium, Ytterbium, and Yttrium, in country “i” (Eq. (2)).
ing, for example, the bastnäsite will be dissolved in acid and then subjected Z t 
to counter-current solvent extraction using immiscible organic and acid sol- GRs ðtÞ ¼ ∑12 15 12
i¼1 TM a−i ðt Þ−∑ j¼1 ∑i¼1 REE a−ji ðt Þ dt þ GRs ðt 0 Þ (1)
vents (Xie et al., 2014). The selective oxidation/reduction, fractional crys- t0
tallization, fractional precipitation, ion exchange or solvent extraction are
used for REEs purification (Kavun et al., 2021; Laatikainen et al., 2022; REEa−ji ðtÞ ¼ βji  TRs−i ðtÞ ð2Þ
Makarova et al., 2020). The exact steps of separation process depend on
the composition of the initial REE concentrate as well as the desired prod- where βji is a coefficient of element “j” processing by country “i” and TRs−i
ucts. The exact steps of separation process depend on the composition of (t) is the total REO stock in country “i” in year “t”.
the initial REE concentrate as well as the desired products. Historical data Eq. (3) describes the global stock of rare earth element “j” in country “i”
for the global mine production of rare earth oxide (REO) are available for (REEs−ji(t)) in the time period “t0-t” by a time integral of REEa−ji(t) as an-
the period between 1996 and 2020 from the United States Geological nual production of element “j” in country “i” minus REEa−jk(t) (Eq. (4))

3
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

gives the annual rate of element “j” used in global manufacturing output “k” which represent all sources of energy: fossil fuel, natural gas, petroleum,
(where k= 1,2, …,10) including magnets, battery alloys, metal alloys, cat- coal, non-fossil fuel, nuclear, renewables, and biomass.
alysts, petroleum, polishing powders, glass additives, phosphors, ceramics, The total cumulative and annual amount of water consumed in stage “n”
and other products. αjk is coefficient of element “j” used for production “k”. of REEs supply chain can be given by Eqs. (9) and (10).

Zt Zt
 
REEs−ji ðt Þ ¼ REEa−ij ðtÞ−REEa−jk ðt Þ dt þ REE s−ji ðt0 Þ (3) TW Sn−i ðtÞ ¼ TW an−i ðt Þdt þ TW Sn−i ðt0 Þ (9)
t0 t0

REEa−jk ðtÞ ¼ αjk  REEs−ji ðt Þ ð4Þ TW an−i ðt Þ ¼ REEa−ji ðtÞ  ∑4w¼1 υn,w (10)

Eq. (5) describes the global stock of product “l” (GPl(t)) in the time pe- where, the cumulative amount of water directly linked to the mining or pro-
riod “t0-t” by a time integral of REEa−jk(t) as the annual rate of element cessing of REEs (n=1,2) in country “i” in year “t” is shown by TWSn−i(t).
“j” used in the global manufacturing output “k” minus REEa−jl(t) (Eq. (6)) TWan−i (t) is annual water use for REEs mining in country “i” which is cal-
representing the annual rate of element “j” used in product “l” (where l= culated based on REEa−ji(t) and water sources used for procedure “w” (w=
1,2, …,20) including magnets for wind turbines and electric vehicles, 1,2,3,4 including water cooling, water mining, water process, and water
NiMH batteries, metallurgical alloys, catalysts for fluid cracking, automo- reservoir) in stage “n”.
bile catalytic converters, glass altering, glass colorizing or decolorizing, The life-cycles of emission rates from energy sources are adapted from
glass absorbing ultraviolet light, yttrium aluminum garnet laser, cathode the GREET model (Wang et al., 2020b) and previous studies (Jin et al.,
ray tube displays, fluorescent lamps, other color light applications, elec- 2018). GHG intensities are calculated by using IPCC AR5 100-year Global
tronic applications, ceramic glazes for color control, ferrites for oxygen sen- Warming Potential values (Stocker et al., 2013) of 1 (CO2), 36 (CH4), and
sors, chemicals containing REEs, military weapons, delivery systems, and 298 (N2O). We applied Eqs. (11) and (12) to estimate the total and annual
satellite systems. δl is coefficient of element “j” used in product “l” and Djl emissions of each life cycle stage of the green energy supply chain.
(t) is global demand of element “j” for using in product “l”. In Eq. (11), TGSn−i(t) is the cumulative emission “g” (g=1,2,3, …,11 in-
The growth of the rare earth metals market is driven by the growing use cluding CO2, CH4, SOx, NOx, CO, VOC, PM2.5, N2O, POC, BC, and PM10) in
of products intended to mitigate the climate change and global warming., stage “n” in year “t” and TGan−i (t) corresponds to the annual emission “g”
e.g. hybrid cars, electric car batteries, and wind turbines. Therefore, our in stage “n” in country “i” in year “t”. In Eq. (12), ϱg, n(t) is the rate of emis-
projection scenarios for REEs rely on the compound annual growth rate of sion generated in stage “n” in year “t”.
REEs demand until 2030.
Zt
Zt TGSn−i ðt Þ ¼ TGan−i ðtÞdt þ TGSn−i ðt0 Þ (11)
 
GPl ðtÞ ¼ REEa−jk ðtÞ−REEa−jl ðtÞ dt þ GPl ðt 0 Þ (5) t0
t0
 
TGan−i ðtÞ ¼ REEa−ji ðtÞ þ REEa−jk ðtÞ  ϱg,n ðtÞ (12)
REEa−jl ðt Þ ¼ δl  Djl ðtÞ ð6Þ

Considering all mass flows over time, the environmental assessment 3. Results and discussion
(energy, water and emissions) is given in an identical form in the model.
Every stage in the REEs supply chain consumes energy obtained from differ- 3.1. Trends in REEs production
ent sources. There are three initial steps for mineral processing: mining,
beneficiation, and separation of an element. After separation, metals are Annual growth rate of REEs global consumption is expected to reach
produced by molten salt electrolysis and metallothermic reduction 4.4% between 2016 and 2026 (Wang et al., 2020a). Fig. S1 presents geo-
methods. During the mining of REEs, energy consumption is mainly associ- graphical distribution of REEs global reserves for top suppliers and by
ated with the use of mining machines and equipment. The industrial pro- type of deposits in 2020 – about 120 mt. More than 250 mineral species
cesses used are complex and require large amounts of electricity. are identified to contain REEs. However, few of those are considered eco-
Although all rare earth minerals are generally open-pit mined, energy re- nomically exploitable, e.g. silicates, fluorocarbonates, oxides and phos-
quirement for crushing and grinding varies depending on (i) their hardness phates. The highest proportion of the global REE reserves (95%) has been
and the gangue minerals associated and (ii) the chemical form that they ex- identified for Bastnäsite, Monazite, Xenotime or Ion-adsorption clay de-
tract (Peiró and Méndez, 2013). In the production stage, the manufacturing posits in China. We focus only on minerals with the highest concentration
of permanent magnets containing REEs is analyzed separately. Total cumu- of REEs (information on quantities and type of minerals for each country
lative and the annual amount of energy consumed in each stage are calcu- is presented in Data S1 (#1 and #2)). The results shown in Fig. S1 suggest
lated using Eqs. (7) and (8). that the US and Russia have the most diversified spectrum of REE-
containing minerals. US ranks 7th in reserves and 2nd in mining while
Zt
Russia occupies 4th place in reserves and 7th in mining.
TESn−i ðt Þ ¼ TEan−i ðtÞdt þ TESn−i ðt0 Þ (7) Fig. 1 presents geographical distribution of cumulative primary produc-
t0 tion in 2010–2020 of five the most significant REEs used in the production
of permanent magnets, i.e., Nd, Dy, Tb, Pr, and Gd. Fig. S2 shows geograph-
TEan−i ðtÞ ¼ REEa−ji ðt Þ  ∑8m¼1 φn,m (8) ical distribution of cumulative primary production of other REEs including
Yb, Ce, Y, Tm, Ho, Sm, Lu, La, Er, and Eu in the given period. The results
In Eq. (7), TESn−i(t) corresponds to the total cumulative amount of en- illustrate an increase in primary production of all elements between 2010
ergy consumption in country “i” in year “t” for the period 2010–2030, and 2020 – light rare-earth elements (Ce: 1.5 times, Er: 4.3 times, Ho: 3.4
n=1,2,3 represent mining, processing and production stages of the green times, La: 12.1 times, Nd: 1.6 times, Pr: 1.6 times, Tm: 9.3 times, and Y:
energy supply chain and TEan−i (t) is annual energy consumption in stage 4.1 times) and heavy rare-earth elements (Dy: 2.6 times, Eu: 4.4 times,
“n” in country “i” in year “t”. In Eq. (8), REEa−ji(t) is the amount of mass Gd: 2.3 times, Lu: 2.1 times, Tb: 1.4 times, Sm: 1.8 times, and Yb: 1.3
of element “j” in country “i” in year “t”, and φn, m is the energy required times). REEs never occur individually and they are accompanied by other
per 1 T of REEs flow in stage “n” from each energy source m= 1,2, …,8 elements in various minerals.

4
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

Fig. 1. Global cumulative primary production of rare earth elements by country in the period 2010–2020. a, Neodymium. b, Dysprosium. c, Terbium. d, Praseodymium. e,
Gadolinium.

Based on reports of 48 exporters and 75 importers in the United Nations fuel cell electric vehicles (FCEV) and plug-in hybrid electric vehicles
Commodity Trade Database (UNCOMTRADE), Fig. S3 shows the distribu- (PHEV) in 2020. The installation of public electric vehicle charging stations
tion of REEs suppliers and the position of countries in international trade. has been increasing exponentially from 30 thousands in 2011 to 1.2 million
Detailed information on quantity and value of exports and imports of in 2020. The number of BEV, FCEV, and PHEV globally is about to reach 8.2
REEs is presented in Table S3. The result shows that in 2020 exports and im- million, 35.2 thousands and 4.4 million, respectively. China dominates
ports of REEs reached around 131 kt (USD 1.22 billion) and 118 kt (USD with around 36% share of the market of EVs followed by Europe (20%),
1.53 billion), respectively. the United States (ca. 14%), and other countries (ca. 30%) in 2020. It is es-
timated that until 2027 EVs production will increase to 44% and 29% in
3.2. Use of REEs in green energy technologies China and Europe respectively. Detailed analysis of mass flows shows that
the annual growth rate of permanent magnets will grow to 6–9.5% between
Fig. 2 shows geographical distribution of wind turbines (Fig. 2a) and 2020 and 2030 (Keilhacker and Minner, 2017; Trench and Sykes, 2020).
different types of EVs (Fig. 2b) in 2020. BP Statistical Review of World En- The results indicate that the amount of primary REEs available for the
ergy revealed that 1590.19 terawatt-hours of global electricity was gener- global production of permanent magnets is about 24 thousand tonnes (kt)
ated by wind turbines in 2020 (Looney, 2020). The installed wind power in 2020 and will reach around 50 kt in 2030. However, the estimates ob-
capacity has been increasing by more than 50 GW each year and the cumu- tained from our model suggest that there is a need for 83 kt of those ele-
lative installed wind power capacity amounted to 651 GW at the end of ments based on demand for magnets production in 2030. So the primary
2019 (Jung and Schindler, 2020). production of REEs fails to meet the demand. The findings highlight that
REEs play an important role in the production of wind turbines, due to the contribution of other supply sources, such as extraction from waste
the increasing power per tower with turbines that have become taller, and recycling can be very important. However, recycling alone cannot
larger, and lighter, and the application of preferred more efficient configu- bridge the gap in the supply of critical materials needed for the develop-
rations of magnet-based technologies, such as gearbox or direct-drive per- ment of the above discussed technologies. Therefore, we should consider
manent magnet synchronous generators. Although, several technologies recycling as one of the core principles of circular economy (Rahimpour
are currently being developed that intend to reduce the use of REEs in per- Golroudbary et al., 2020). It will be essential to develop an effective
manent magnets, according to the IEA they will remain far from being com- recycling system, invest in the development of recycling technologies and
petitive with the existing wind technologies even until the next decade dedicated infrastructure for collecting, dismantling, and separating prod-
(Bobba et al., 2020). Therefore, we focus on magnet-based technologies ucts containing rare earth elements. Moreover, we need to consider other
in this study. complementary solutions such as substituting materials, introduction of
Wind power corresponds to 4.1% of electricity generation (coal 39%; new technologies, and more efficient use of materials.
natural gas 27%; nuclear 19%; hydropower 7%; biomass 1.5%; petroleum After 2030, it is expected that the amount of spent magnet-based prod-
and gaseous hydrocarbons 1% and geothermal, solar and other <1%) ucts for EVs will considerably increase and after 2040 the same phenomena
(Lampert et al., 2015). Fig. 2a shows that the biggest fraction of wind tur- will consider wind turbines. Therefore, there is a growing gap between
bines can be found in China (28.4% with cumulative installed capacity of available recycled magnets and steadily growing demand for REEs.
wind power 237 thousand megawatts) and the United States (21.2%, Hence, mining remains necessary to meet the demand for minerals needed
with 105 thousand megawatts). in decarbonization process.
Global sales of EVs rose by 43% in 2020 in comparison to 2019. It is es-
timated that in 20 years there will be around 500 million EVs. (Raw 3.3. Trends in environmental impacts of REEs mining
materials for a truly green future, 2021) To meet this demand, we will
need more critical raw materials such as REEs. Fig. 2b shows the distribu- In this study, the dynamic model is used to evaluate REEs global flows
tion of charging stations and the stock of battery electric vehicles (BEV), and their environmental impact in 2010–2030. The total annual energy

5
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

Fig. 2. Geographical distribution of green energy technology including wind turbines for energy supply and electric vehicles as consumers in 2020. a, Distribution of wind
turbines and leading countries based on wind energy capacity. b, Distribution of charging stations and stock of battery electric vehicles (BEV), fuel cell electric vehicles (FCEV)
and plug-in hybrid electric vehicles (PHEV).

6
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

Fig. 3. Environmental performance of primary production of rare-earth elements in the mining stage between 2010 and 2030. a, Global energy consumption, in petajoules
(PJ). b, Global water consumption, in million cubic meters. c, Annual global carbon dioxide (CO2) emission by country, in thousands tonnes CO2. d, Global methane (CH4),
sulfur oxides (SOx), nitrogen oxides (NOx), carbon monoxide (CO), volatile organic compounds (VOC), particulate matter with sizes smaller than 2.5 μm (PM2.5), nitrous
oxide (N2O), particulate organic carbon (POC), black carbon (BC), and airborne particulate matter with sizes smaller than 10 μm (PM10) emissions through mining of
REEs in 2020, in tonnes.

and water consumption, as well as the trend in emissions related to the thousand t CO2 in 2030. Detailed calculations show that China, being the
REEs global primary production are given in Fig. 3. major polluter among the REEs mining countries, releases 138.9 t of CH4,
Results presented in Fig. 3a indicate an increase in global energy con- 132.6 t of SOx, 60.8 t of NOx, 33.2 t of CO, 9.8 t of VOC, 4.2 t of PM2.5,
sumption for the primary production of REEs from 2.55 petajoules (PJ) in 1.3 t of N2O, 0.9 t of POC, 0.4 t of BC, and 9.9 t of PM10.
2010 to 5.47 PJ in 2030, which corresponds approximately to a 2.14- The analysis shows that global GHG in carbon dioxide equivalent (CO2
times growth within 20 years. An identical trend can be seen in Fig. 3b eq) emitted by REEs mining has increased by 94% (152.97 kt CO2 eq) in
for water consumption in the primary production of REEs. It starts from 2020 compared to 79.01 kt CO2 eq in 2010 and it is estimated to reach
0.42 million cubic meters (m3) in 2010 to reach 0.91 million m3 in 2030. 169.49 kt CO2 eq in 2030. China emitted 88.06 kt CO2 eq, followed by
The results highlight the issues raised in the Carbon Disclosure Project the U.S. 23.90 kt CO2 eq, Myanmar 18.87 kt CO2 eq, Australia 10.69 kt
(CDP)(CDP, 2013) on criticality of sustainable water management. CO2 eq, and other countries 11.44 kt CO2 eq. China has the highest share
The results show that energy consumption is growing in all REEs pro- of GHG emissions in REEs supply chain due to its the largest share of
ducers. For example, in China it increased from about 2.4 gigajoules (GJ) primary production. To keep global average temperature increase well
in 2010 to 2.8 GJ in 2020 and is estimated to reach 3.2 GJ in 2030, and below 2 °C above pre-industrial levels, many regions in China are establish-
in the United States from 60.9 megajoules (MJ) in 2012 to 771.1 MJ in ing carbon neutrality targets, but this trend is the most noticeable in
2020 and is expected to reach 854.4 MJ in 2030. The same trend can be ob- the power and transport sectors or industries that heavily contribute
served for the use of water, e.g., China water consumption grew from about to GHG emissions e.g., magnesium production. However, it does not
0.40 million m3 in 2010 to 0.47 million m3 in 2020 and is estimated to concern the REEs industry itself yet (Liu et al., 2021; Rahimpour
reach 0.52 million m3 in 2030, and in the United States water consumption Golroudbary et al., 2022).
increased from 0.01 million m3 in 2012 to 0.12 million m3 in 2020 and is
expected to reach 0.14 million m3 in 2030. 3.4. Trends in environmental impact of REEs processing
Fig. 3c shows annual CO2 emissions generated by mining of REEs. In
Fig. 3d, there are presented the emissions (CH4, SOx, NOx, CO, VOC, We use a dynamic model to evaluate the global environmental impact of
PM2.5, N2O, POC, BC, and PM10) generated by the mining of REEs in processing of 15 REEs (Fig. 1 and Fig. S2) in 2010–2030. Energy and water
2020. In particular, it is estimated that CO2 emissions from REEs mining used in processing of each element, as well as emissions generated along the
will grow in China by the factor of 1.3, from 71.5 kt CO2 in 2010 to 92.5 process are presented in Fig. 4.

7
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

The result indicates that global energy and water consumption during emissions in the processing stage) – 2286.8 t of CH4, 2180.5 t of SOx,
the processing of REEs is 22 times higher than that in mining. The results 1001.6 t of NOx, 546.0 t of CO, 162.2 t of VOC, 72.6 t of PM2.5, 21.4 t
presented for energy consumption, Fig. 4a, and water analysis in Fig. 4b of N2O, 15.2 t of POC, 6.4 t of BC, and 0.2 t of PM10. We can also
show an increase (81%) of the consumed energy and water for processing learn that global GHG emissions from the processing stage of REEs in
of REEs in 2020 when compared to 2010. The estimates for 2030 show con- 2020 amounted to around 3.08 mt.
sumption of energy and water as 121.89 PJ and 20.24 million m3 respec- The results of simulation of environmental emissions from the mining
tively. According to Fig. 4b in 2010, the highest consumption of energy and processing of REEs for 5 years interval between 2010 and 2030 are
and water in the processing stage is attributed to Ce (30.81 PJ and 5.12 mil- available in Data S1 (#7 and #8).
lion m3) and Nd (12.22 PJ and 2.03 million m3). In 2020, annual trend It is important to note that there exist several routes of bio-accumulation
shows that in addition to Ce (46.8 PJ and 7.76 million m 3) and Nd of REE in plants, animals, and humans resulting from mining or processing
(19.40 PJ and 3.22 million m3), also La (8.31 PJ and 1.38 million m3) of REEs. Several activities in the primary production of REEs such as cut-
and Y (6.97 PJ and 1.16 million m3 ) are considered the highest con- ting, drilling, blasting, transportation, stockpiling, and processing can cre-
sumers of energy and water in the processing stage. When comparing ate dust containing REE, other toxic metals and chemicals. Therefore,
2030 and 2010, it is estimated that energy and water consumption environmental degradation (air, soil and water) may occur together
will increase for Ce by 1.8-times, for Er by 6.2-times, and for La by with negative impact on human health of hazards encountered in
15.5-times. waste disposal areas. Moreover, REE mining poses a serious challenge
Fig. 4c presents the annual trend of CO2 emissions from the processing due to the considerable amount of radioactive elements in minerals,
of REEs in years 2010–2030. The results show that CO2 emitted during the such as thorium and uranium (García et al., 2020). It shows that the
processing of Ce, Nd, Pr, and totally for other elements (La, Sm, Eu, Gd, Tb, mining and processing of rare earth elements, if not carefully controlled,
Dy, Y, Er, Ho, Tm, Yb, and Lu) increases from 0.91, 0.36, 0.09, 0.27 to can create serious environmental hazards. For example, during the pro-
1.64, 0.69, 0.18, 1.07 mt CO 2 in years 2010–2030, respectively. duction of one ton of rare earth oxides it is necessary to treat 1.4 tons of
Fig. 4d illustrates the global amount of emissions (CH4, SOx, NOx, CO, radioactive wastes in solid, liquid, or gaseous form (Talan and Huang,
VOC, PM2.5, N2 O, POC, BC, and PM10) generated by the processing 2022). The problem of the radioactive waste is one of the reasons why
of REEs in 2020. Detailed calculations show that the global Ce process- new policies addressing safety of mining activities of rare earth ele-
ing corresponds to the highest amount of emissions (47% of total ments are urgently needed.

Fig. 4. Environmental impact of processing rare-earth elements between 2010 and 2030. a, Global energy consumption, in petajoules (PJ). b, Global water consumption, in
million cubic meters (m3). c, Annual global carbon dioxide (CO2) emission by element, in million tonnes CO2. d, Global methane (CH4), sulfur oxides (SOx), nitrogen oxides
(NOx), carbon monoxide (CO), volatile organic compounds (VOC), particulate matter with sizes smaller than 2.5 μm (PM2.5), nitrous oxide (N2O), particulate organic carbon
(POC), black carbon (BC), and airborne particulate matter with sizes smaller than 10 μm (PM10) emissions through processing of REEs in 2020, in tonnes.

8
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

3.5. Trends in environmental impacts of magnets production a result of the global production of permanent magnets between 2010
and 2030.
In this section, a detailed analysis is presented of emissions related to The results of simulation for 5 years interval between 2010 and 2030
the production of permanent magnets in years 2010–2030. Production are available in Data S1 (#9).
process comprises strip casting, hydrogen decrepitation, jet milling,
and electroplating. As shown in Fig. 5a, pollution related to global 3.6. Rebound effect of green energy
warming and ecotoxicity is increasing rapidly by taking into account
the production of magnets for wind turbines and EVs. The environmen- A bi-directional positive reinforcing causal relationship between renew-
tal analysis of manufacturing magnets shows an increase of GHG emis- able energy consumption and economic growth has been demonstrated in
sions from 2.4 billion CO2 eq in 2010 to 5.4 billion CO2 eq in 2030 for previous research studies (Pirlogea and Cicea, 2012; Zafar et al., 2020).
the transportation sector and from 499.6 million CO 2 eq in 2010 to We discovered that the rapid development of green energy technologies
1.1 billion CO 2 eq in 2030 for energy generation. The ecotoxicity of causes a rebound effect as shown in Fig. 6. On the one hand, the number
magnets' manufacturing is estimated to grow from 16.9 billion Compar- of produced EVs and installed wind power capacity have increased between
ative Toxic Unit equivalent (CTUe) in 2010 to 37.4 billion CTUe in 2030 2010 and 2020 (Fig. 6a) triggering technological progress and production
in EVs manufacturing and from 3.5 billion CTUe in 2010 to 7.7 billion diversification. As a result, REEs prices (except Lu) have decreased sharply
CTUe in 2030 in wind turbines production. Subsequently, Fig. 5b in 2020 compared to 2011 (e.g. by 53% for Yb, 76% for Tb, 78% for Pr, and
shows the impact of magnet production on water acidification and eu- 82% for Dy), reducing the price of wind power by 71% in 2010 (Fig. 6b and
trophication. The results indicate that permanent magnets production Data S1 (#3)). On the other hand, lower prices of REEs elements led to the
in EVs manufacturing emitted 18.7 mt SOx in 2010 and is estimated to increase of demand for them for EVs (from 8 kt in 2010 to 18.5 kt in 2020)
reach 41.6 mt SO x in 2030 while in wind turbine manufacturing the and wind turbine production (from 1 kt in 2010 to 7 kt in 2020), as shown
amount of sulfur oxides emitted grows from 3.8 to 8.5 mt SOx in 2010 in Fig. 6c. The results given in Fig. 6d show annual GHG emissions from
and 2030, respectively. The same trend is observed for eutrophication, mining and processing stages of REEs, as well as the production of magnets
in the unit of nitrogen equivalent (N eq), which increased from 26.3 to applied in wind turbines and EVs. The calculations indicate that every 1%
58.2 million N eq in EVs manufacturing and 5.4 to 11.9 million N eq increase of green energy production causes roughly a 0.18% depletion of
in wind turbine manufacturing. Fig. 5c represents the projected rare-earth elements reserves and 0.90% increase of CO2 eq. It means, cumu-
emissions of carcinogenic and non-carcinogenic components in the latively 32 billion t CO2 eq of greenhouse gas emissions in the period
unit of human toxicity impacts (CTUh) and CH4 and NOx emissions as 2010–2020, globally.

Fig. 5. Environmental effects of the production of permanent magnets applied in electric vehicles and wind turbines between 2010 and 2030. a, Global warming and
ecotoxicity b, Accidification and eutrification c, Carcinogenic and non-carcinogenic component emissions, ozone depletion and smog pollution.

9
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

Fig. 6. Rebound effect of the transition to green energy based on the use of critical materials and their environmental impact between 2010 and 2020. a, Annual green energy
production. b, Annual price of rare-earth elements and wind power (the price for thulium is not available). c, Consumption of rare-earth elements applied in green energy
technology. d, Global greenhouse gas emissions from mining and processing of rare-earth elements and from the production of magnets.

3.7. Discussion manufacturing of permanent magnets remain substantial. It is worth noting

that the energy required in mining and processing of minerals usually
One of key elements in the development of green energy technologies is comes from fossil fuels and not from wind or solar power stations.
utilization of wind power generators and electric vehicles. The use of REEs Therefore, mining and processing are one of the global top GHG emitting
in those applications is hardly substitutable due to their physical and chem- industries.
ical properties. According to the IEA, achieving net-zero emissions globally Energy production sector accounts for 72% and transportation corre-
by 2050 would require 6 times higher mineral input (3–7 times more REEs) sponds to 15% of global GHG emissions including CO2, CH4 (16%), N2O
in 2040 than today. Building wind farms and manufacturing EVs generally (6%), and fluorinated gases (2%) (Ritchie and Roser, 2020). Also, statistics
require more minerals (6–9 times) than their fossil fuel-based counterparts. shows that global GHG emissions from the use of energy in the manufactur-
The results of our simulations show that critical materials could be a bottle- ing sector corresponds to 24.2% of the total GHG emissions (49.4 billion t
neck in the propagation of green energy technologies. Our findings reveal a CO2 eq in 2016) (Ritchie and Roser, 2019) and our results show the share
rebound effect of the current decarbonization activities aimed at ensuring of REEs mining, processing and manufacturing of permanent magnets for
sustainable development. wind turbines and EVs to be 0.05%. Considering total GHG emissions
It is generally accepted that wind energy generators emit no pollution to from the energy sector of the REEs producing countries (Australia, Brazil,
air or water, require no mining or drilling for fuel, use no water in electric- China, India, Russia, United States and others), the obtained results show
ity generation, and create no hazardous or radioactive wastes that would re- that mining and processing account for 0.12% and 0.19% of CO2 eq emitted
quire a permanent storage. Also, it is estimated that wind power has the by this sector in 2010 and 2020, respectively. The detailed calculation
potential to cumulatively avoid one-third of global annual carbon emissions shows that an increase by 1% of green energy production represents
until 2050 — a 2 MW wind turbine containing around 350 kg REEs can re- roughly a 0.90% increase of GHG emissions in exploitation phase including
duce up to 42 kt of CO2 eq over its life time (Change, 2014; Nugent and mining, processing, and production stages. Moreover, the results show that
Sovacool, 2014). However, those benefits are limited only to the system ex- 1% of green energy production leads to 0.18% REEs resource depletion.
ploitation phase. They are considerably limited in the raw material extrac- Moreover, our analysis shows that China, the US and Australia are the
tion and equipment production phases. Therefore, a complete life cycle top suppliers of raw materials demanded by the green energy sector and
analysis of the system should be considered. For EVs, it is estimated that their production has the highest environmental impact. Focusing on the
total GHG emissions associated with EV-related handling operations case of China, the eco-cost analysis of 19 development pathways for REEs
(e.g., charging, and driving) are typically lower than those of a conven- mining industry has provided us with the net cost ranging between 14.8
tional gasoline vehicle, particularly when electricity is generated from re- and 16 US$ billion reported for 2015 and 2025, respectively (Lee and
newable energy sources, e.g. wind. However, as demonstrated in this Wen, 2018). In the longer perspective, diversification of supplies to meet
paper, energy and water consumption and emissions generated in the the actual demand for REEs from the existing technologies, leaving aside

10
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

the sustainability issues, shifts environmental problems from one place to —review and editing. Eveliina Repo: Writing—review and editing. All au-
another and moves them forward into the future while no attempts are thors have read and agreed to the published version of the manuscript.
made to mitigate the most urgent issues. This finding highlights the need
to design and put in place various measures intended to increase REEs
reuse, recycling (the current rate is less than 1%), limit dematerialization, Data and materials availability
increase substitution and develop new elimination technologies (e.g. an en-
gine without permanent magnets). Such actions would support an environ- The authors declare that all data supporting the findings of this study
mentally sustainable development of green energy technologies. can be found in the article and/or its Supplementary Information files.

4. Conclusions
Declaration of competing interest

This paper contributes to the enhanced understanding of environmental

The authors declare that they have no known competing financial inter-
impact of green energy technologies and supports energy policy makers in
ests or personal relationships that could have appeared to influence the
developing appropriate strategies for decarbonization. Contrary to the pre-
work reported in this paper.
viously published works on REEs focused on local scale analysis, this publi-
cation, for the first time, presents a global scale analysis of environmental
Acknowledgments
cost of using REEs in green energy technologies. It has been performed
using system dynamics modelling integrated with life cycle assessment
The authors would like to acknowledge Viipuri Management Research
and geometallurgical approach. The obtained results confirm the existence
Lab of LUT University for providing access to AnyLogic software (University
of a rebound effect - rapid development of existing green energy technolo-
8.5.0).
gies leads to unsustainable consumption of rare earth elements. In the lon-
ger perspective, it could trigger shifting environmental problems from one
Appendix A. Supplementary data
location to another and pushing them forward into the future if any rele-
vant attempts would be made to mitigate the core issues. Therefore, it is
Supplementary data to this article can be found online at https://doi.
necessary to propose strategies aimed to develop new technologies required
org/10.1016/j.scitotenv.2022.155022.
to substitute rare earth elements used for green energy generation. Another
approach would be to develop efficient recycling policies to increase the
circularity of rare earth elements. For example, large scale implementation References
of closed loop supply chain of permanent magnets used in wind turbines
and electric vehicles. Arshi, P.S., Vahidi, E., Zhao, F., 2018. Behind the scenes of clean energy: the environmental
It is important to switch to more sustainable energy sources. However, footprint of rare earth products. ACS Sustain. Chem. Eng. 6, 3311–3320.
the considerable environmental cost of implementing green energy tech- Bailey, G., Joyce, P.J., Schrijvers, D., Schulze, R., Sylvestre, A.M., Sprecher, B., Vahidi, E.,
Dewulf, W., Van Acker, K., 2020. Review and new life cycle assessment for rare earth pro-
nologies and its rebound effect at a global scale cannot be overlooked. duction from bastnäsite, ion adsorption clays and lateritic monazite. Resour. Conserv.
The insight offered by this study should facilitate the analysis of real envi- Recycl. 155, 104675.
ronmental costs of transition to green energy technologies. Our findings Bobba, S., Carrara, S., Huisman, J., Mathieux, F., Pavel, C., 2020. Critical Raw Materials for
Strategic Technologies and Sectors in the EU—A Foresight Study. Eur. Comm, Brussels,
will hopefully help in developing adequate policies aimed at the implemen- Belgium.
tation of sustainable decarbonization measures. These polices should care- CDP, 2013. Metals & Mining: a Sector Under Water Pressure Analysis for Institutional Inves-
fully consider ways to increase the share of non-fossil fuels, the tors of Critical Issues Facing the Industry. Carbon Disclosure Project (CDP), United
Kingdom (2013), p. 20p.
implementation of negative GHG emission technologies, or creating a Change, I.C., 2014. In: Edenhofer, O., et al. (Eds.), Mitigation of Climate Change, pp. 1–33.
basis for economically viable processes for capturing/storing CO2 or its Deng, H., Kendall, A., 2019. Life cycle assessment with primary data on heavy rare earth ox-
use as a raw material. ides from ion-adsorption clays. Int. J. Life Cycle Assess. 24, 1643–1652.
Du, X., Graedel, T.E., 2013. Uncovering the end uses of the rare earth elements. Sci. Total En-
The most critical parameters that influence the results are: volatility of
viron. 461–462, 781–784. https://doi.org/10.1016/j.scitotenv.2013.02.099.
rare earth markets caused by the political situation and substitution of Dudley, B., et al., 2018. BP Statistical Review of World Energy. BP Stat. Rev, London, UK
NdFeB technology by another one requiring smaller amount of REEs. How- accessed Aug 6, 116.
Dutta, T., Kim, K.H., Uchimiya, M., Kwon, E.E., Jeon, B.H., Deep, A., Yun, S.T., 2016. Global
ever, data on the social impacts of using rare earth elements for green en-
demand for rare earth resources and strategies for green mining. Environ. Res. 150,
ergy generation are still missing. Therefore, a comprehensive assessment 182–190. https://doi.org/10.1016/j.envres.2016.05.052.
of sustainability of green energy issues still needs to be considered. The ac- Fishman, T., Graedel, T.E., 2019. Impact of the establishment of US offshore wind power on
tivities aimed at the reduction of rare earth elements consumption in green neodymium flows. Nat. Sustain. 2, 332–338.
Forrester, J.W., 1997. Industrial dynamics. J. Oper. Res. Soc. 48, 1037–1041.
energy generation will be an important factor contributing to the enhance- Gambogi, J., 2021. Rare Earths Statistics and Information. AvailableUS Geol. Surv. (USGS).
ment of environmental sustainability of renewable energy technologies. https//www.usgs.gov/centers/nmic/rare-earths-statistics-and-information.
The closed-loop supply chain of green technology, which differs signif- García, A.C., Latifi, M., Amini, A., Chaouki, J., 2020. Separation of radioactive elements from
rare earth element-bearing minerals. Metals (Basel) 10, 1–22. https://doi.org/10.3390/
icantly from the forward supply chain in many aspects, is not considered in met10111524.
the analysis. Therefore, the limitation of this study is the exclusion of the Goodenough, K.M., Schilling, J., Jonsson, E., Kalvig, P., Charles, N., Tuduri, J., Deady, E.A.,
potential circularity of REEs originating from end-of-life wind turbines Sadeghi, M., Schiellerup, H., Müller, A., Bertrand, G., Arvanitidis, N., Eliopoulos, D.G.,
Shaw, R.A., Thrane, K., Keulen, N., 2016. Europe’s rare earth element resource potential:
and electric vehicles. Another limitation are missing data on life cycle as- an overview of REE metallogenetic provinces and their geodynamic setting. Ore Geol.
sessment of all stages of the REEs supply chain. Rev. 72, 838–856. https://doi.org/10.1016/j.oregeorev.2015.09.019.
Graedel, T.E., Harper, E.M., Nassar, N.T., Nuss, P., Reck, B.K., 2015. Criticality of metals and
metalloids. Proc. Natl. Acad. Sci. 112, 4257–4262.
CRediT authorship contribution statement Gwenzi, W., Mangori, L., Danha, C., Chaukura, N., Dunjana, N., Sanganyado, E., 2018.
Sources, behaviour, and environmental and human health risks of high-technology rare
Saeed Rahimpour Golroudbary: Conceptualization, Data curation, earth elements as emerging contaminants. Sci. Total Environ. 636, 299–313.
Hagelüken, C., 2014. Recycling of (critical) metals. Crit. Met. Handb, pp. 41–69.
Methodology, Verification and validation, Formal analysis, Investigation,
Herrington, R., 2021. Mining our green future. Nat. Rev. Mater. 6, 456–458.
Visualization, Writing—original draft preparation, Writing—review and Jin, H., Afiuny, P., Dove, S., Furlan, G., Zakotnik, M., Yih, Y., Sutherland, J.W., 2018. Life
editing. Iryna Makarava: Data curation, Methodology, Formal analysis, In- cycle assessment of neodymium-iron-boron magnet-to-magnet recycling for electric vehi-
vestigation, Visualization, Writing—original draft preparation, Writing— cle motors. Environ. Sci. Technol. 52, 3796–3802.
Jung, C., Schindler, D., 2020. Introducing a new approach for wind energy potential assess-
review and editing. Andrzej Kraslawski: Conceptualization, Verification ment under climate change at the wind turbine scale. Energy Convers. Manag. 225,
and validation, supervision, Writing—original draft preparation, Writing 113425.

11
S.R. Golroudbary et al. Science of the Total Environment 832 (2022) 155022

Kavun, V., van der Veen, M.A., Repo, E., 2021. Selective recovery and separation of rare earth Rahimpour Golroudbary, S., Makarava, I., Repo, E., Kraslawski, A., Luukka, P., 2022. Magne-
elements by organophosphorus modified MIL-101(Cr). Microporous Mesoporous Mater. sium life cycle in automotive industry. Procedia CIRP 589–594.
312, 110747. https://doi.org/10.1016/j.micromeso.2020.110747. Ritchie, H., Roser, M., 2019. Our World in Data: CO2 and Greenhouse Gas Emissions.
Keilhacker, M.L., Minner, S., 2017. Supply chain risk management for critical commodities: a Ritchie, H., Roser, M., 2020. CO₂ and Greenhouse Gas Emissions. Our World Data.
system dynamics model for the case of the rare earth elements. Resour. Conserv. Recycl. Schreiber, A., Marx, J., Zapp, P., 2021. Life Cycle Assessment studies of rare earths
125, 349–362. production-findings from a systematic review. Sci. Total Environ. 148257.
Laatikainen, M., Makarova, I., Sainio, T., Repo, E., 2022. Selective acid leaching of rare earth Stocker, T.F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y.,
elements from roasted NdFeB magnets. Sep. Purif. Technol. 278, 119571. https://doi. Bex, V., Midgley, P.M., et al., 2013. Climate change 2013: the physical science basis.
org/10.1016/j.seppur.2021.119571. Contrib. Work. Gr. I to Fifth Assess. Rep. Intergov. Panel Clim. Chang. 1535.
Lampert, D.J., Cai, H., Wang, Z., Keisman, J., Wu, M., Han, J., Dunn, J., Sullivan, J.L., Strauss, M.L., Mishra, B., Martins, G.P., 2019. Selective reduction and separation of europium
Elgowainy, A., Wang, M., et al., 2015. Development of a Life Cycle Inventory of Water from mixed rare-earth oxides recovered from waste fluorescent lamp phosphors. Miner.
Consumption Associated With the Production of Transportation Fuels. Met. Mater. Ser. 305–309. https://doi.org/10.1007/978-3-030-10386-6_35.
Lee, J.C.K., Wen, Z., 2018. Pathways for greening the supply of rare earth elements in China. Talan, D., Huang, Q., 2022. A review of environmental aspect of rare earth element extraction
Nat. Sustain. 1, 598–605. processes and solution purification techniques. Miner. Eng. 179, 107430. https://doi.
Li, Z., Diaz, L.A., Yang, Z., Jin, H., Lister, T.E., Vahidi, E., Zhao, F., 2019. Challenges and op- org/10.1016/j.mineng.2022.107430.
portunities for carbon neutrality in China. Res. Conserv. Recyc. 149, 20–30. https://doi. Tintelecan, A.C., Dobra, A., Marţiş, C., 2014. LCA Indicators in Electric Vehicles
org/10.1016/j.resconrec.2019.05.025. EnvironmentalImpact Assessment. Electric Vehicles International Conference (EV).
Liu, L., Keoleian, G.A., 2020. LCA of rare earth and critical metal recovery and replacement https://doi.org/10.1109/EV.2019.8892893.
decisions for commercial lighting waste management. Res. Conserv. Recyc. 159, Trench, A., Sykes, J.P., 2020. Rare earth permanent magnets and their place in the future
104846. https://doi.org/10.1016/j.resconrec.2020.104846. economy. Engineering 6, 115–118.
Liu, Z., Deng, Z., He, G., Wang, H., Zhang, X., Lin, J., Qi, Y., Liang, X., 2021. Challenges and Vahidi, E., Navarro, J., Zhao, F., 2016. An initial life cycle assessment of rare earth oxides pro-
opportunities for carbon neutrality in China. Nat. Rev. Earth Environ. https://doi.org/10. duction from ion-adsorption clays. Resour. Conserv. Recycl. 113, 1–11.
1038/s43017-021-00244-x. Waisman, H., de Coninck, H.C., Rogelj, J., 2019. Key Technological Enablers for Ambitious
Looney, B., 2020. BP Statistical Review of World Energy. Climate Goals Insights From the IPCC Special Report on Global Warming of 1.5° C.
Makarova, I., Ryl, J., Sun, Z., Kurilo, I., Górnicka, K., Laatikainen, M., Repo, E., 2020. One-step Wang, J., Guo, M., Liu, M., Wei, X., 2020. Long-term outlook for global rare earth production.
recovery of REE oxalates in electro-leaching of spent NdFeB magnets. Sep. Purif. Technol. Resour. Policy 65, 101569.
251, 117362. https://doi.org/10.1016/j.seppur.2020.117362. Wang, M., Elgowainy, A., Lee, U., Bafana, A., Benavides, P.T., Burnham, A., Cai, H., Dai, Q.,
McNeice, J., Ghahreman, A., 2018. Selective oxidation of cerium in rare earth solutions, a Gracida-Alvarez, U.R., Hawkins, T.R., et al., 2020. Summary of Expansions and Updates
comparison of four oxidants. Miner. Met. Mater. Ser. (PartF5), 15–25 https://doi.org/ in GREET®2020.
10.1007/978-3-319-72350-1_2. Xie, F., Zhang, T.A., Dreisinger, D., Doyle, F., 2014. A critical review on solvent extraction of
Nugent, D., Sovacool, B.K., 2014. Assessing the lifecycle greenhouse gas emissions from solar rare earths from aqueous solutions. Miner. Eng. 56, 10–28. https://doi.org/10.1016/j.
PV and wind energy: a critical meta-survey. Energy Policy 65, 229–244. mineng.2013.10.021.
Peiró, L.T., Méndez, G.V., 2013. Material and energy requirement for rare earth production. Zafar, M.W., Shahbaz, M., Sinha, A., Sengupta, T., Qin, Q., 2020. How renewable energy con-
JOM 65, 1327–1340. sumption contribute to environmental quality? The role of education in OECD countries.
Pell, R., Tijsseling, L., Goodenough, K., Wall, F., Dehaine, Q., Grant, A., Deak, D., Yan, X., J. Clean. Prod. 268, 122149.
Whattoff, P., 2021. Towards sustainable extraction of technology materials through inte- Zaimes, G.G., Hubler, B.J., Wang, S., Khanna, V., 2015. Environmental life cycle perspective
grated approaches. Nat. Rev. Earth Environ. 1–15. on rare earth oxide production. ACS Sustain. Chem. Eng. 3, 237–244.
Pirlogea, C., Cicea, C., 2012. Econometric perspective of the energy consumption and eco- Raw materials for a truly green future. Nat. Rev. Mater. 6, 455.
nomic growth relation in European Union. Renew. Sust. Energ. Rev. 16, 5718–5726.
Rahimpour Golroudbary, S., El Wali, M., Kraslawski, A., 2020. Rationality of using phospho-
rus primary and secondary sources in circular economy: game-theory-based analysis. En-
viron. Sci. Policy 106, 166–176. https://doi.org/10.1016/j.envsci.2020.02.004.

12