FLORE

Repository istituzionale dell'Università degli Studi

di Firenze
Innovative device for mechanical treatment of End of Life
photovoltaic panels: Technical and environmental analysis
Questa è la Versione finale referata (Post print/Accepted manuscript) della seguente pubblicazione:

Original Citation:
Innovative device for mechanical treatment of End of Life photovoltaic panels: Technical and
environmental analysis / Del Pero F.; Delogu M.; Berzi L.; Escamilla M.. - In: WASTE MANAGEMENT. - ISSN
0956-053X. - ELETTRONICO. - 95:(2019), pp. 535-548. [10.1016/j.wasman.2019.06.037]

Availability:
This version is available at: 2158/1162161 since: 2021-03-30T20:35:08Z

Published version:
DOI: 10.1016/j.wasman.2019.06.037

Terms of use:
Open Access
La pubblicazione è resa disponibile sotto le norme e i termini della licenza di deposito, secondo quanto
stabilito dalla Policy per l'accesso aperto dell'Università degli Studi di Firenze
(https://www.sba.unifi.it/upload/policy-oa-2016-1.pdf)

Publisher copyright claim:

(Article begins on next page)

25 May 2023
This document is the accepted version of the article “Innovative device
for mechanical treatment of End of Life photovoltaic panels: technical
and environmental analysis”, to be used for sharing through Author
Institution repository.
Final published version is available at:
https://doi.org/10.1016/j.wasman.2019.06.037

Manuscript Draft

Title: Innovative device for mechanical treatment of End of Life

photovoltaic panels: technical and environmental analysis

Article Type: Full Length Article

Section/Category: Environmental Assessment

Keywords: End-of-Life; Photovoltaic panels; Mechanical treatment;

Recycling; Recovery; Environmental impact

Corresponding Author: Dr. Francesco Del Pero, Ph.D.

Corresponding Author's Institution: Department of Industrial Engineering

- University of Florence

First Author: Francesco Del Pero, Ph.D.

Order of Authors: Francesco Del Pero, Ph.D.; Massimo Delogu; Lorenzo

Berzi; Marta Escamilla

Abstract: The paper contributes at filling the lack of knowledge on

Photovoltaic (PV) panels recycling through the analysis of a mobile
mechanical treatment plant developed within the context of a European
project. The process, the machinery installed in the system and their
main functionalities are described. The data are used to perform a Life
Cycle Assessment (LCA) focused on the End-of-Life (EoL) process, assuming
as Functional Unit (FU) the treatment of a 20 kg PV panel. The system
boundaries include construction and operation of the device as well as
recycling and incineration of different material fractions performed
outside the plant. The inventory is mainly based on primary data coming
from a collection carried out directly on the recycling device. The
results show that impacts are concentrated on operation stage mainly due
to energy consumption involved in milling and separation activities. The
analysis of different operation steps reveals that pre-treatment gives
the highest contribution, followed by glass and silicon separation with
the lowest quota attributable to copper and polymeric fraction
separation. Considering also recycling and incineration processes of EoL
waste, the environmental credits due to the avoided production of virgin
raw materials counterbalance the burdens of construction and operation
for most of impact categories. The comparison of results with existing
LCAs of fixed recycling installations stresses that the use of a mobile
system involves considerable environmental benefits thanks to the
reduction of transports needed to move EoL PV waste to the recycling
facility site.
*Revised Manuscript (clean copy)
Click here to view linked References

1 Innovative device for mechanical treatment of End of Life photovoltaic panels: technical and
2 environmental analysis
3 Francesco Del Peroa, Massimo Delogua, Lorenzo Berzia, Marta Escamillab
a
4 Department of Industrial Engineering, University of Florence, Via di Santa Marta 3, 50139 Florence, Italy
b
5 Leitat Technological Centre, C/ de la Innovació, 2, 08225 Terrassa Barcelona, Spain

6 ABSTRACT

7 The paper contributes at filling the lack of knowledge on Photovoltaic (PV) panels recycling through the analysis of a
8 mobile mechanical treatment plant developed within the context of a European project. The process, the machinery
9 installed in the system and their main functionalities are described. The data are used to perform a Life Cycle
10 Assessment (LCA) focused on the End-of-Life (EoL) process, assuming as Functional Unit (FU) the treatment of a 20
11 kg PV panel. The system boundaries include construction and operation of the device as well as recycling and
12 incineration of different material fractions performed outside the plant. The inventory is mainly based on primary data
13 coming from a collection carried out directly on the recycling device. The results show that impacts are concentrated on
14 operation stage mainly due to energy consumption involved in milling and separation activities. The analysis of
15 different operation steps reveals that pre-treatment gives the highest contribution, followed by glass and silicon
16 separation with the lowest quota attributable to copper and polymeric fraction separation. Considering also recycling
17 and incineration processes of EoL waste, the environmental credits due to the avoided production of virgin raw
18 materials counterbalance the burdens of construction and operation for most of impact categories. The comparison of
19 results with existing LCAs of fixed recycling installations stresses that the use of a mobile system involves considerable
20 environmental benefits thanks to the reduction of transports needed to move EoL PV waste to the recycling facility site.

21 Keywords: End-of-Life, Photovoltaic panels, Life Cycle Assessment, Recycling, Recovery, Environmental impact.

22 Acronyms used in the text

23 CdTe: Cadmium/Tellurium based materials

24 Disposal: process as defined in Directive 2008/98/EU definition
25 EoL: End of Life
26 EVA: Ethylene-vinyl acetate polymer
27 FU: Functional unit
28 LCA: Life Cycle Assessment
29 LCI: Life Cycle Inventory
30 LCIA: Life Cycle Impact Assessment
31 LME: La Mia Energia (Company)
32 PV: Photovoltaic
33 PV–MOREDE: Photovoltaic Panels Mobile Recycling Device (Project)
34 Recovery: process as defined in Directive 2008/98/EU definition
35 Recycle: process as defined in Directive 2008/98/EU definition
36 WEEE: Waste from Electric and Electronic Equipment
37 WtE: Waste to Energy

38 1. Introduction

39 Our global society is strongly dependent on natural resources consumption involved in production and operation of
40 consumer goods (Hawkins at al., 2012; Delogu et al., 2016). At the same time the disposal of End-of-Life (EoL)
41 products represents one of the most challenging sustainability aspects (Tian and Chen, 2014; Pagnanelli et al., 2016). In
42 the last years, the need for environmentally friendly processes has been fast growing in several different areas, the most
43 influential being transportation (Banar and Özdemir, 2015; Del Pero et al., 2017) and energy production (Somorin et al.,
44 2017; Stenzel et al., 2017). Considering energy production systems, Photovoltaic (PV) panels technology is providing a
45 fundamental contribution to the shift from traditional fossil fuels to renewable energy sources. Solar panels have been
46 installed since the eighties but it is during the past decade that PV market has expanded exponentially with a world
47 cumulative installed capacity of 402.5 GW in 2017, of which about 28% is installed in Europe (EPIA, 2014; IEA,
48 2018). Further growth is expected, which could lead to 530 GW installed capacity at the end of 2019 according to
49 recent assessment (SPE, 2018). Today energy coming from PV panels is one of the most promising renewable sources
50 and forecasts for subsequent decades expect that it will provide up to 25 % of the global electricity demand by 2050
51 (Dias et al., 2017; Silva et al., 2014; Zuser and Rechberger, 2011).
52 Following such installation rate for PV systems, a parallel growth of e-waste coming from the sector is expected.
53 According to International Renewable Energy Agency data (IRENA 2018), the approximate life-span of solar panels is
1
 54 estimated in the range of 30 years; however, effective life can differ since early substitution are possible due to so called
55 “infant”, “mid-life” and “wear out” failures (Weckend et al., 2016). Therefore, the amount of PV waste on a global
56 scale is estimated to pass from the current 870 t/y to a level of at least 1.5 million t/y in 2030, with a worst case
57 prediction of 8 million t/y. Other authors confirm a growth determining the need to process more than 2 million t/y in
58 2038 (Jung et al., 2016). For 2050 scenarios, IRENA predictions provide an estimation of 60 to 78 million t/y of EoL
59 PV panels to be processed. Amongst various reasons for which landfilling of PV panels is not an option, a relevant
60 observation is that many products contain hazardous materials (such as lead, cadmium and bismuth) but also high-value
61 elements (such as silver, titanium, tellurium) (Marwede and Reller, 2012; Savvilotidou et al., 2017); the content
62 depends on the technology adopted by the panel. In this regard, the study by BioIS (BioIS, 2011) evidences the
63 environmental issues related to the improper disposal of solar panels, such as leaching of hazardous substances and
64 losses of both conventional and precious material resources. The introduction of efficient recovery systems for EoL
65 modules would involve two main beneficial effects. On one hand the use of recycled materials for the production of
66 new panels would strongly reduce the need for virgin resources; on the other hand, the development of specific EoL
67 treatments would avoid the dispersion in the environment of dangerous materials typical of PV modules. Regarding the
68 characteristics of panels installed up to now and the prevision for next years, it has to be considered that PV industry is
69 constantly being developed, thus leading to a possible modification of technologies and solutions adopted for modules.
70 Indeed, performance analyses of polycrystalline silicon PV panels in comparison with thin-film PV panels highlight that
71 the effective potential for energy production and profitability depends not only on maximum efficiency but also on the
72 overall condition of installation (Munshi et al., 2018). Therefore, according to recent efficiency growth estimations for
73 various technologies (IRENA, 2018) a definitive PV module technology strategy remains difficult to predict. From an
74 End-of-Life (EoL) point of view, this means that it is worth to investigate on treatment technologies for all the PV
75 families. Crystalline silicon-based cells, for example, currently represent a predominant share of the market (Lee and
76 Ebong, 2017) so that, given a certain time-delay from installation, a large amount of elements will need EoL treatment.
77 Regarding other notable technologies such as CdTe-based panels, the potential release of toxic substances on the
78 environment EoL stage has been demonstrated to be particularly worrying (Ramos-Ruiz et al., 2017). As a
79 consequence, even if a favorable overall environmental impact is assessed, EoL treatment different from landfilling is
80 advisable (Vellini et al., 2017; Tao and You, 2015). Other authors are stating that the risk of toxic contamination is not
81 critical even in case of landfilling, but their analysis still highlights the importance of recycling due to the favorable
82 environmental impact of secondary raw material production (Rocchetti and Beolchini, 2015). Another important point
83 is the European regulatory framework of PV panels EoL (Directive 2012/19/EU; Contreras-Lisperguer et al., 2017). For
84 solar modules, the regulation fixes collection rate up to 85% and recycling rate up to 80%. As European manufacturers
85 and distributors are legally obliged to guarantee take-back, recovery, and recycling of their products, the last years have
86 seen a great improvement in efficient collection programs and recovery processes for PV panels. Therefore PV
87 recycling is justified not only on the basis of mere economic feasibility or environmental assessment, but it is
88 mandatory, even if the implementation of the regulation is still affected by significant barriers (Besiou et Wassenhove,
89 2015). The approach is coherent with “waste hierarchy” criteria (Directive 2008/98/EU) and to other binding
90 regulations (such as those for End of Life Vehicles, see Berzi et al., 2013).
91 Considering the economic aspect, various framework for economic profitability assessment have been proposed but
92 they appear to be still affected by uncertainties of data which are difficult to assess with precision (Perez-Gallardo et al.,
93 2018; Duflou et al., 2018). The reduction of plant cost for WEEE treatment is a relevant condition for effective
94 economic operation (Cucchiella et al., 2016). Regarding the general category of WEEE, a few examples of portable
95 recycling plants are described in literature; such treatment plants are characterized by small capacity and treatment rate,
96 in the range of a few tons per hour or even less. Small plants have been proposed in order to achieve multiple targets,
97 such as reducing the investment needed for installation, enhancing the availability of recycling plants over territories,
98 reducing material transport needs, setting up highly specialized processes which can be suitable for relatively small
99 waste flows; a few examples of such systems will be described in the next paragraph.

100 The document contents following the introduction to the topic are organized as follows: section 1 provides a description
101 of the international context of PV panel recycling. Section 2 describes PV-MOREDE plant in detail, also summarizing
102 data used for impact assessment (LCA) analysis. Section 3 presents the results of the activity. Section 4 contains
103 concluding observation.

104 1.1. Recycling plants for EoL photovoltaic panels

105 The literature provides numerous studies regarding the EoL of solar panels (Jungbluth et al., 2005, 2012; Klugmann-
106 Radziemska et al., 2010b). Several papers (Dias et al., 2016; Gustafsson et al., 2014; Klugmann-Radziemska and
107 Ostrowski, 2010a) propose panels EoL routes based on two steps: a physical treatment (made up of shredding and
108 thermal process) and a thermo-chemical treatment (mainly based on ethylene-vinyl-acetate degradation and recovery
109 through pyrolysis). Further researches focus on alternative procedures for the extraction of resources from EoL solar
110 panels, such as organic solvent methods to recycle silicon cells used for conventional crystalline silicon PV modules
111 (Kanga et al., 2012; Doi et al., 2001). These studies show that recycling and recovery rates are comprised within 80 and
112 90 % depending on the specific material (silicon, copper, silver) with a value of more than 99% recoverability for the
2
113 polymeric ones. Potential values for glass recyclability if treated using proper physical processes are demonstrated to be
114 in the range of 80-85%, considering direct recycling as glass, while further 10% (corresponding to fraction finer than
115 0.08 mm) is estimated to be recoverable or recyclable through other processes (Granata et al., 2014). Therefore, an
116 highlight coming from literature is that the process of PV shredding and physical treatment is critical, since the
117 undesirable reduction of glass into small fines is potentially reducing the direct recyclability.

118 The studies hitherto presented deal with the development of PV panels recycling processes exclusively from a technical
119 point of view, without taking into account the eco-profile of the proposed methods. On the other hand, many papers
120 investigate the environmental impacts of the production and use of PV technologies as confirmed by some review
121 articles that deal with the topic (Bhat Varun et al., 2011; Evans at al., 2009; Hsu et al., 2012; Peng et al., 2013; Shervani
122 et al., 2010). Frisson et al. (2000) perform a comparison based on energy consumption of a standard PV panel and one
123 built using recycled wafers; the results show that the recycled wafer panel can lead to 40 % lower impacts. Frankl et al.
124 (2005) investigate the production of electricity by different PV technologies. The authors find that decommissioning
125 and disposal of a ground mounted PV plant represents roughly 4 % of total Life Cycle (LC) greenhouse gas emissions
126 while other impact categories present lower impacts. Shibasaki et al. (2006) investigate the production of 1 GJ of
127 electricity produced by thin film solar modules. The research takes into account the impacts related to the recycling
128 technology which includes module delamination, removal of the EVA layer, removal and recycling of metals. The study
129 reveals significant environmental advantages of thin film PV modules compared to conventional energy supplying
130 systems. Even if the relative environmental burdens are reduced, the impact expressed as absolute value is potentially
131 significant according to the expected growth of EoL panel arising in next decades. Additionally, the high impact of PV
132 panels EoL stage and the mandatory compliance with WEEE directive (Directive 2012/19/EU) highlight the need to
133 improve recyclability methods. Held and Ilg (2011) perform the Life Cycle Assessment (LCA) of the recycling of 1 m2
134 CdTe PV modules basing on the “First Solar” process as a sequence of mechanical and hydrometallurgical treatments
135 (First Solar, 2019). The results prove that solar power involves a notable environmental advantage with respect to the
136 country specific grid mixes. Wild-Scholten (2009) estimates some draft figures of energy consumption involved in PV
137 panels recycling. The author calculates that 250 MJ, 240 MJ and 150 MJ are needed for the taking back and recycling
138 respectively of mono-Si, multi-Si and CdTe PV devices. Rocchetti et al. (2013) and Zeng et al. (2015) deal with
139 portable e-waste recycling plants. Rocchetti et al. (2013) present a mobile plant installed in a container lorry that
140 processes e-waste residues through a hydrometallurgical process while Zeng et al. (2015) deal with an integrated mobile
141 e-waste recycling facility which combines dismantling, shredding and multi-level separation. These studies show that
142 valuable resources such as metals, plastic and glass can be fully separated for further recycling, thus leading to
143 remarkable benefits in terms of environmental protection and human health.

144 All the studies shown above do not consider the EoL stage since information regarding PV modules decommissioning
145 and recycling/disposal are lacking or the researches do not provide disaggregated information on the considered
146 recycling processes. A brief selection of papers that thoroughly examine the Eol of PV panels are Muller et al. (2005),
147 Fraunhofer Institut (2012) and, more recently, Carnevale et al. (2014), Corcelli et al. (2015) and Latunussa et al. (2016).
148 Muller et al., 2005 provide an environmental analysis concerning the EoL treatments of crystalline silicon PV modules
149 according to the “Deutsche Solar” recycling process. The study shows that even if the environment is damaged by the
150 inputs and outputs of the recycling process, the reuse of panels implies notable environmental advantages in terms of
151 CO2/SO2 emissions and resources depletion due to the avoided production of new cells. The research also compares the
152 impacts of the “Deutsche Solar” recycling process to the treatment in a municipal incineration plant and to the
153 shredding. The results show that the incineration and the shredding involve lower impacts, but also lower recyclability
154 rates. The authors justify these results with the different scale of incineration plant and recycling facility. The major
155 limitation of the study is that it does not explicitly declare material and energy input/output flows occurring in the
156 investigated processes. The research by Fraunhofer Institut (Fraunhofer Institut, 2012) performs a LCA analysis of a
157 recycling plant for EoL silicon modules. The first step of the treatment is the manual removal of aluminum frames and
158 junction boxes. Then the pre-treated modules are shredded and subsequently they enter the glass recycling line which
159 performs manual pre-sorting, laminates shredding, separation and materials extraction. Finally, the mix is separated
160 according to the different material fractions. The outcomes show that treatment activities have a much lower impact
161 than the potential credits achievable through the recycling of valuable materials. The study also stresses the significant
162 influence that transport activities have on eutrophication and photochemical ozone creation impact categories. The
163 authors evidence also that the recycling facility presents a low level of technological innovation and that it does not
164 allow a cost-effective recycling process for precious material fractions. Carnevale et al. (2014) perform a LCA
165 comparison between photovoltaic (both silicon based and thin film modules) and solar thermal systems including EoL
166 examination. The results highlight that the systems based on thermal solar collector are environmentally preferable
167 while thin film modules and solar thermal collector present the lowest values of energy and CO2eq payback time. On the
168 other hand, the emissions and energy saving associated to material recycling are relevant, especially for Si-based
169 modules. Corcelli et al., 2015 deal with an LCA applied to a laboratory-scale recycling process for silicon PV modules
170 based on thermal treatment. The authors build a scale-up scenario in order to assess the effective environmental
171 convenience of thermal treatment with respect to physical and chemical ones; different recycling scenarios according to

3
172 lower and higher recycling/recovery rates are taken into account. The LCA is performed by means of the SimaPro
173 software and the ReCiPe (H) midpoint is used for the impact assessment. The results are shown in aggregated and
174 normalized form, without the detail of the specific treatment steps. The study reveals negative environmental impacts
175 due to credits obtainable thanks to materials recycling. The authors conclude that recycling ensures the supply chain
176 sustainability in the long-term by enhancing potential credits for secondary materials production. Latunussa et al., 2016
177 deal with the LCA of an innovative process for the recycling of silicon PV panels. The process is composed of physical
178 (mechanical and thermal) and chemical (acid leaching and electrolysis) treatments. The focus of the paper is the
179 environmental impact due to the recycling treatment while credits coming from the potential production of secondary
180 raw materials are outside the system boundaries of the study. The research provides transparent and disaggregated data
181 for each stage of the recycling process. The authors stress that the impacts are mainly due to the incineration of the
182 panel encapsulation layers followed by the recovery of silicon metal, silver, copper and aluminum; the contribution of
183 transportation is also relevant for several impact categories.
184 The review of environmental studies of EoL PV modules recycling treatments highlights the following issues:

185 - Many relevant studies are type tailored, as they take into account EoL treatments for specific typologies of
186 solar panels, mainly crystalline silicon and Cd-Te modules (Kanga et al., 2012; Held and Ilg, 2011; Xu et al.,
187 2018). The existing processes developed specifically for PV panels recycling are based on fixed plants.
188 - The LCAs from literature do not always investigate in detail the inventories related to the considered treatment
189 systems. Data such as energy consumption, efficiency of recycling and recyclability/recoverability rates are
190 still under investigation, while many treatments are only described as laboratory process, still not consolidated
191 at industrial level (Bogacka et al., 2017; Pagnanelli et al., 2017).
192 - A direct comparison of EoL assessment data between different literature studies is often not possible due to the
193 adoption of different boundary conditions, such as FU, system definition and input materials considered;,
194 - Several LCA studies deal with the environmental impacts due to the production and use of PV technologies.
195 Even though EoL is recognized as a critical phase for the LCA of solar panels, the accurate analysis of this
196 phase is often excluded from the system boundaries or roughly estimated (Fraunhofer Institut, 2012; Corcelli et
197 al., 2015; Latunussa et al., 2016). Only a few papers examine the EoL of solar modules (Carnevale et al., 2014;
198 Corcelli et al., 2015). However, these studies focus only on potential reuse and they do not investigate in detail
199 the LCI as well as the environmental effects of materials recycling. Additionally, the information about the
200 efficiency of PV panels EoL processes and the achievable recycling/recovery rates are generally lacking or
201 quite incomplete.

202 This article performs the LCA of an innovative recycling process for EoL PV panels conceived and calibrated to be
203 used for mechanical recycling of crystalline-silicon based modules. The size of the plant is quite small and it is chosen
204 in order to make it tailored on the PV technology. The innovation is that the process is based on a mobile plant which
205 can operate in different locations over its life-time treating on-site and on demand modules characterized by different
206 composition/structure, thus representing a flexible and easily accessible alternative with respect to current fixed waste
207 systems. The system has been developed within the context of the European project “Photovoltaic panels Mobile
208 Recycling Device” (PV-MOREDE) and it has been constructed by the Italian company “La Mia Energia” (LME). The
209 study is built on a detailed LC inventory mostly based on primary data and it is aimed at filling the lack of knowledge
210 regarding the environmental benefits achievable by PV waste recycling.
211
212 2. Materials and method
213 The LCA methodology is applied to the PV-MOREDE pilot scale process for the recycling of solar panels. The LCA is
214 performed according to the ISO standards 14040, 14044 (ISO 14040/14044, 2006) and the ILCD handbook (Hiederer et
215 al., 2011). The study is described in detail in the following paragraphs.
216
217 2.1. Goal and scope
218 The scope of the study is assessing the potential environmental impacts of the PV-MOREDE recycling process and
219 identifying the main hotspots related to its operation. The Functional Unit (FU) is the recycling of a 20 kg EoL PV
220 panel (an average value observed on EoL panels) assuming a plant processing capacity of 0.8 t/h and 8 years as service
221 life-time. The processing capacity was assessed through preliminary operability test on the machine. Regarding plant
222 life, two main elements are used for the estimation:

223 - considering a yearly depreciation value of 12% (suitable for generic industrial machinery, according to Italian
224 system – see DM 31/12/1988) 8 years corresponds to zero residual value. In absence of other references, the
225 authors consider that this is a suitable assumption for the European context;
226 - since the whole PV-MOREDE project is motivated by the aim of promoting the machine on the market, an
227 investigation among potential buyers was performed. According to such confidential contacts, 8 years life is
228 indicated as a minimum requirement for the recycling system.
229

4
230 The study is based on a “cradle-to-gate” approach; the system boundaries includes construction of the PV-MOREDE
231 plant up to the operation stage, including recovery and disposal processes of EoL PV waste. Table 1 reports the mass
232 composition of the 20 kg panel taken into account for the study. Even if panels are coming from different
233 manufacturers, significant differences in terms of mass or dismantlability have not been found during the testing of the
234 system. However, a certain variability is expected depending on panel manufacturer, age and technology.
235
Component Quantity [kg] Percentage [%]

Glass 13.60 68.0

Frame (aluminum) 4.00 20.0
Polymer-based adhesive (EVA) encapsulation layer 1.10 5.5
Solar cell (silicon metal) 0.70 3.5
Back-sheet layer (Polyvinyl Fluoride) 0.20 1.0
Cables (copper and polymers) 0.20 1.0
Conductor (aluminum, copper) 0.16 0.8
Metals (silver, lead, tin) 0.04 0.2
Total 20.00 100.0
236
237 Table 1. Mass composition of the 20 kg PV panel as input to the PV-MOREDE recycling process. Data are adapted from literature (Latunussa et al.,
238 2016)

239 The construction takes into account the environmental impacts due to the manufacturing of the entire PV-MOREDE
240 system in terms of material/energy consumption, emissions to the environment and waste production. This stage
241 includes also transportation of the different plant components from the suppliers to the assembly facility site as well as
242 the transportation of PV-MOREDE plant to the operation site.
243 The operation stage assesses the impacts, which are originated by energy and materials consumption, as well as the
244 emissions to the environment due to the treatments performed within the plant. The operation of the PV-MOREDE
245 process is divided into four steps: pre-treatment, glass separation, silicon separation and copper & polymeric fraction
246 separation. The potential environmental impacts and credits related to further processing of materials (e.g. Aluminum to
247 foundry, Glass production from cullet, residuals to WtE) have been included since PV-Morede systems is enabling such
248 recycling and recovery processes. Therefore, the system boundaries of the EoL stage of PV waste include all the EoL
249 processes performed outside the PV-MOREDE plant:
250
251 - transportation of PV waste materials to recovery facilities;
252 - recycling of different material fractions separated within the PV-MOREDE system (mainly recycling of
253 aluminum, copper, glass, Silicon-rich fractions; segregation of polymeric and residual fractions);
254 - incineration with energy recovery of polymeric fraction (including polymers, elastomers and organic materials
255 in general). The amount of energy produced is considered a co-product of the recycling process and it is
256 modeled through an “incineration with energy recovery” process which implies both an environmental credit
257 (due to energy production) and an environmental impact (due to incineration emissions).
258
259 The study does not take into account the decommissioning of the recycling system.

260 2.2. Life Cycle Inventory (LCI)

261 2.2.1. Description of the recycling process

262 The PV-MOREDE system consists of a mobile device that guarantees a complete treatment of solar panels directly in
263 the installation site and it is able to treat any type of Mono and polycrystalline PV panels. The pilot scale plant has a
264 processing capacity of 0.8 t/h and 8 years service life-time. The system is aimed at reducing the panels into small parts
265 in order to enable recycling through proper segregation of different materials. As the plant is conceived and designed to
266 be transported within the limits of ordinary freight vehicles, all the machineries are arranged in three containers
267 disposed in a single line, thus allowing a continuous treatment process after being assembled.
268 The system is conceived for mono and polycrystalline panel type. The results of the study are expected not to vary
269 significantly between different PV panels manufacturers. A strong attention is paid on calibration of the separation
270 system of back-sheet polymers. In case that new technologies arise in the future (e.g. substitution of Tedlar-based back-
271 sheets with others), the expectation is that the system can be adapted with proper set-up and substitution of certain parts
272 (e.g. changing mesh size and vibration frequency in vibrating screen or modifying air flow speed for separator and
273 transport systems).
5
274
275 Figure 1 provides a schematic overview of the plant layout, including all sub-systems employed in the different
276 separation steps; the components are grouped according to the macro-phase they refer to. Pictures of the assembled
277 system, ready for operation, have been published on unrestricted documents, still available online (LME, 2015).

13 – Container n°1 14 – Container n°2 15 – Container n°3

PRE-TREATMENT GLASS SEPARATION SILICON COPPER & AUXILIARIES

SEPARATION POLYMERIC
Removal of Aluminum 2. Conveyor FRACTION 10. Dust aspiration
frame, electrical cables 3. Hammer mill 5. Conveyor SEPARATION system
and junction box (not 4. Vibrating screen 6. Hammer mill 11. Filtering system
represented in the 7. Vibrating screen 8. Conveyor 12. Compressed air
picture) 9. Three-stages circular generator
separator 13. Container n°1
PANEL SHEARING 14. Container n°2
15. Container n°3
1. Shearing machine

278 Figure 1. Layout and machineries of the PV-MOREDE system

279 Below each separation step is described in detail; numbers refer to subsystems numbering adopted in Figure 1.

280 Pre-treatment and panel shearing (separation of cables, connectors, aluminum frame, first size reduction and
281 introduction in the system). After a preliminarily weighting of the module, the pre-treatment is performed through the
282 separation of connection cables and junction boxes, usually by manual dismantling. This task is not labor-intense.
283 Handling of the panel is performed manually, but optionally a manipulator can be used to reduce effort of the operators.
284 After removal of the above mentioned electrical hardware, the panel is placed on a device for aluminum frame removal,
285 defined as “expander table”. Removal of the frame is done using a semi-automated hydraulic system, which applies
286 forces from the inside of the panel to the outside along all the perimeter of the panel. The expander machine has been
287 specifically developed for this application. The expander machine as well as a treatment plant which is a variant of the
288 here described PV-MOREDE system is visible on public website (Veolia, 2018). After preparation, the panel is reduced
289 to 100x100 mm pieces through the shearing machine (1). The shearing machine has also been developed for this
290 application and it is composed by an hydraulic shear-press system able to cut the panel into squares in order to make its
291 size suitable for the insertion in further size reduction machines (see also Figure 2) The machine acts on the panel “as-
292 is”, just cleaned from electrical connections and metal frame, shearing the layers all together. The aim is to reduce its
293 size for further processing. Each square still includes glass, cell materials, backsheet layer (see Figure 3).

294 Glass separation (grinding and mechanical separation of fragments). A conveyor (2) moves the squares to the impact
295 swinging hammers mill (3) where further fragmentation occurs (see also Figure 4). The fragments are sifted by the first
296 vibrating screen (4), which is in the form of rectilinear sieve. The size of the mesh network hole is 6 mm. This section
297 provides three main flows:

298 - the over-screen fraction, which mainly comprehends glass fragments and heavy fraction which is sent to the
299 refinement;
300 - the under-screen fraction, which comprehends a flow with low content of glass and high content of silicon
301 parts and copper;
302 - dust (mainly silicon fines), which is recovered by the dust collection system.

6
303 The refinement of the scrap glass is performed through an optical micro-sorting machine. A scrap in compliance with
304 the EU Regulation no. 1179/2012 (Commission regulation, 2012) is obtained, so that recycling of glass is effectively
305 possible.

306
307 Figure 2. Shear-press machine used for the panel size reduction into 100x100 mm squares (first fragmentation stage). Component as described in
308 EP3089825B1 (Reggi, 2017)

309
310 Figure 3. Panel square portion, output of the shearing machine.

311
312 Figure 4. Swinging hammers mill (second fragmentation stage). Component as described in EP3089825B1 (Reggi, 2017)

313 Silicon separation (further grinding/shredding of the small fraction and mechanical treatments for material separation).
314 The under-screen fraction coming from glass separation are transported by a conveyor (5) which carries the material to
315 a second impact mill (Figure 5) with a fixed hammer (6) whose outputs are:

316 - plastic fraction, considered useful for WtE processes, separated through a mesh screen of 2 mm;

7
317 - a mixed compound of plastic materials, copper and silicon;
318 - dust, recovered by the collection system.
319

320
321 Figure 5. Fixed hammers mill (third fragmentation stage). Component as described in EP3089825B1 (Reggi, 2017)

322 Copper & polymeric fraction separation (further sieving of the plastic-copper compound). A conveyor (8) moves the
323 compound plastic/copper to the third vibrating screen (9) (Figure 6), a circular one, which performs the separation of
324 silicon and copper fractions:

325 - mixed compound of plastic and copper (and, eventually, small fractions of other metals) with 0.5-2 mm
326 grain size. This fraction is sent to eddy current separator which performs the final segregation of non-
327 ferrous metals and the separation of polymers;
328 - fraction with 0.315-0.500 mm grain size, low silicon content and classified as recyclable (see also Table
329 4);
330 - fraction with grain size less than 0.315 mm, high silicon content and classified as recyclable (see also
331 Table 4), but also potentially valuable for further refinement.
332

333
334 Figure 6. Three stages screen for fines separation. Component as described in EP3089825B1 (Reggi, 2017)

335 Where possible, transportation of materials and fines is performed through pneumatic systems. In conclusion, the entire
336 recycling process can be divided into three main size reduction and separation phases:

337 1) Shearing from panel to squares

338
339 2) Shredding from squares to particles with 6 mm grain size
340
341 a) Over-screen: glass sent to refinement
342 b) Under-screen: material to further shredding
343
8
344 3) Shredding from 6 mm to 2 mm grain size particles
345
346 a) Over-screen: mainly plastics (expected destination: recovery through WtE processes)
347 b) Under-screen: to three-stages sieving
348
349 i) 0.5-2 mm grain size particles: to eddy current separation for metal recovery
350 ii) 0.500-0.315 mm grain size particles: low-silicon product
351 iii) <0.315 mm grain size particles: high-silicon product.

352 The containers, as well as the supporting structures, are designed and built specifically for this application (Grassi et al.,
353 2018). In comparison with standard freight containers, the structures offer increased stiffness and resistance in order to
354 hold up the basement of the machineries. The three containers present different lengths in order to adapt to systems
355 disposition and dimensions. Each container has doors and openings for operation and maintenance activities. The main
356 machines of the plant are covered by patent EP3089825B1 (Reggi, 2017) and they are listed in Table 2.

357 2.2.2. LCI data collection

358 The inventory is mainly based on primary data coming from a detailed gathering performed on the pilot scale plant. As
359 secondary data the Ecoinvent database v3 (Ecoinvent database, 2017) and GaBi Thinkstep database 8.6.0.20 are used.
360 Below the LCI data collection is described in detail for each one of the LC stages of the PV-MOREDE recycling
361 process.
362
363 Plant construction. Data collection regarding plant manufacturing is the typology and quantity of materials that
364 constitute the entire PV-MOREDE system. Table 2 reports material composition of each machine/component of the
365 device as well as the LCI datasets adopted for the environmental modelling of the manufacturing stage.

9
366
367
PLANT CONSTRUCTION (manufacturing)

Equipment Component Sub-component Material Mass [kg] Ecoinvent DB process

Gear motor Fixed dye and counter die High Speed Steel (HSS) 15 Steel, chromium steel 18/8 - Hot rolled {GLO}
Shearing machine
Oleo dynamic & hydraulic equip. Material extraction tape Rubber 5 Synthetic rubber {GLO}
Conveyor tape Rubber 2 Synthetic rubber {GLO}
Transfer and control unit Steel S 235 JR 8 Steel, unalloyed {GLO}
Chain conveyor n°1 Gear motor
Rollers 10
Inox steel AISI 304 Steel, chromium steel 18/8 {GLO}
Tensioners 5
Four-pole three-phase asynchronous
Hammer mill Swinging hammers Steel (high content of Mn and Cr) 85 Steel, low-alloyed {GLO}
electric motor
Frame Inox steel AISI 304 4 Steel, chromium steel 18/8 {GLO}
Vibrating screen n°1 N° 2 gear motors
N° 4 springs Steel Si-Cr-Ni 2 Sinter, iron {GLO}
Straps
Steel S 235 JR 7 Steel, unalloyed {GLO}
Pulleys
Chain conveyor n°2 Three-phase electric motor
Conveyor chains
Inox steel AISI 304 5 Steel, chromium steel 18/8 {GLO}
Pipes
Four-pole three-phase asynchronous Impact mill (structure) Cast Iron 50 Cast iron {GLO}
Hammer mill
electric motor Fixed hammer Steel (high content of Mn and Cr) 35 Steel, low-alloyed {GLO}
Frame Inox steel AISI 304 4 Steel, chromium steel 18/8 {GLO}
Rectangular vibrating sifter N° 2 gear motors
N° 4 spring Steel Si-Cr-Ni 2 Sinter, iron {GLO}
Straps
Steel S 235 JR 5 Steel, unalloyed {GLO}
Pulleys
Chain conveyor n°3 Three-phase electric motor
Conveyor chains
Inox steel AISI 304 7 Steel, chromium steel 18/8 {GLO}
Pipes
Lower band with exhaust pipe
Intermediate band with cone for
Three- stage circular
Three-phase electric motor recycling and discharge Inox steel AISI 304 40 Steel, chromium steel 18/8 {GLO}
separator
Lower band with exhaust pipe
N° 2 rings for grid holding
Separator of non-ferrous Full-magnetic drum housing
Electric motor Steel S 235 JR 45 Steel, unalloyed {GLO}
materials vibrating feeder and control panel
Collecting container Metal container internal bag Steel S 235 JR Steel S 235 JR - Steel, unalloyed {GLO}
Electrum vacuum cleaner N° 7 receiving hoods / pipes/
Dust aspiration system Inox steel AISI 304 90 Steel, chromium steel 18/8 {GLO}
Three-phase electric motor fittings
Filtering system
Compressor (600 l) N° 1 over-pressure shutter Aluminum 50 Aluminum, cast alloy {GLO}
Container n°1 Steel S 235 JR 1800
Metallic structure box Container n°2 Steel S 235 JR 2885 Steel, unalloyed {GLO}
Container n°3 Steel S 235 JR 3200
Total: 8361

Table 2. Components of PV-MOREDE system LCI data collection for plant construction

10
368 For the transportation of

369 - plant components from suppliers to the assembly site

370 - assembled plant to the operation site and subsequent transportation events during its life

371 data collection involves the determination of transport typology and travelled distance. Table 3 reports the LCI data
372 collection for transportation, including LCI datasets used for the environmental modelling.

PLANT CONSTRUCTION (transportation)

Travelled distance
Component Transport typology (Supplier – Assembly site) Ecoinvent DB process
[km]

Freight transport Transport freight, lorry 3.5-7.5 metric ton,

Hammer mill 676 EURO5 {GLO}
EURO 5

Freight transport Transport freight, lorry 3.5-7.5 metric ton,

Vibrating screen 704 EURO5 {GLO}
EURO 5
Transportation plant components

Freight transport Transport freight, lorry 3.5-7.5 metric ton,

Chain conveyor n°2 713 EURO5 {GLO}
EURO 5

Freight transport Transport freight, lorry 3.5-7.5 metric ton,

Hammer mill 676 EURO5 {GLO}
EURO 5

Rectangular vibrating Freight transport Transport freight, lorry 3.5-7.5 metric ton,
704 EURO5 {GLO}
sifter EURO 5

Freight transport Transport freight, lorry 3.5-7.5 metric ton,

Chain conveyor n°3 713 EURO5 {GLO}
EURO 5

Three- stage circular Freight transport Transport freight, lorry 3.5-7.5 metric ton,
704 EURO5 {GLO}
separator EURO 5

Separator of non- Freight transport Transport freight, lorry 3.5-7.5 metric ton,
2230 EURO5 {GLO}
ferrous materials EURO 5

Freight transport Transport freight, lorry 3.5-7.5 metric ton,

Collecting container 539 EURO5 {GLO}
EURO 5

Dust aspiration system Freight transport Transport freight, lorry 3.5-7.5 metric ton,
Filtering system
133 EURO5 {GLO}
EURO 5

Travelled distance
Transport typology Ecoinvent DB process
Transportation PV

(Operation site) [km]

MOREDE plant

Freight transport Transport freight, lorry 11

3000* metric ton, EURO5 {GLO}
EURO 5

*
Travelled distance to reach operation sites during plant LC

373 Table 3. LCI data collection for transportation during plant construction (plant components and PV-MOREDE plant)

374 Plant operation. For the operation stage the inventory is built with the following data:

375 - the amount of material processed within each separation step;

376 - the electricity consumption for each treatment step;
377 - the LC inventory datasets used for the environmental modelling of electricity production.

378 Regarding the treatment processes performed within the PV-MOREDE system, the inventory is based on primary data
379 coming from preliminary machine testing activities and measured on the pilot plant by the LME company experts.
380 Where data are not available or not accurate enough, assumptions on the basis of machine nominal capabilities are used.
381 Table 4 reports the LCI data collection for the operation stage in terms of material and energy inputs.

382 EoL PV waste. Concerning the EoL PV waste stage, the inventory consists of the following data:
11
383 - transport typology and travelled distance for transporting PV waste materials to recovery facilities;
384 - transportation, material recycling and incineration with energy recovery processes performed within plants
385 external to the PV-MOREDE facility;
386 - typology and amount of materials processed within the recycling and incineration with energy recovery
387 processes;
388 - LCI datasets used for the environmental modelling of transportation, recycling and incineration with energy
389 recovery processes.

390 The LCI data are from the GaBi Thinkstep database and they refer to average processes in the market. Table 4 shows
391 the LCI data collection for the EoL PV waste stage in terms of material flows to transportation, material recycling and
392 incineration with energy recovery processes. For each flow the table reports separation step, recovery process and LC
393 inventory dataset used for the environmental modelling. The assumptions for material treatment are summarized as
394 follows.

395 The recycling of aluminum (2.0 kg), copper (0.2 kg), glass (13.1 kg) and junction boxes (0.05 kg Polyamide and 0.05
396 kg Polyethylene) scrap is modeled as open loop recycling through the implementation of a specific substitution rate.
397 The substitution rate provides the impact credits due to the substitution of primary with secondary material (avoided
398 production of primary material) and it is calculated net of

399 - lower quality of secondary material with respect to primary material

400 - impacts caused by energy consumption and emissions associated with recycling processes (i.e. removal of
401 impurities/washing, re-melting and refining).
402
403 Concerning aluminum, literature generally estimates the substitution rate in a range between 10 % and 94 % (LCI
404 8.6.0.20 version GaBi dataset; (Koffler and Florin, 2013; Schrijvers et al., 2016). Lower values are suitable for low-
405 quality aluminum scrap (low-grade irony Alu), higher values are suitable for scrap aluminum extrusions. The case of
406 PV-MOREDE frames is comparable to high quality aluminum extrusions scraps, so that even a substitution factor of
407 about 90-95 % would be suitable. Considering that in large treating plants it is possible the mixing of aluminum scraps
408 coming from different sources, a substitution rate of 0.4 is conservatively assumed.

409 The substitution rate for copper scrap is assumed 0.65, coherently with recycling potential estimation provided by LCI
410 8.6.0.20 version GaBi dataset.

411 The substitution rate for glass scrap is assumed 0.8, basing on LCI 8.6.0.20 version GaBi dataset and literature
412 references which suggest high values (Ferreira et al., 2017; Rigamonti et al., 2009); even if known data are mainly
413 referred to glass cullet originated by packaging, we assume that the glass cullet provided by the PV-MOREDE has
414 similar substitution rate due to its high quality (obtained through screening with optical sorting).

415 For junction boxes, on the basis of the material examined during plant operativity it has been assumed that 50 % (0.1
416 kg) is sent to incineration with energy recovery (Latunussa et al., 2016) due to degradation, damage and/or impossibility
417 to recognize the constituting material - most time being PVC, PA or PE as usual for WEEE insulating components. The
418 remaining 50 % is recycled as material. For this fraction, the substitution rate (sample considered: 0.05 kg Polyamide
419 and 0.05 kg Polyethylene) is assumed 0.5, a conservative factor on the basis of literature references for WEEE-derived
420 plastics (Wäger and Hischier, 2015). Reuse, even if cited in literature (Park and Park, 2014), has not been considered.

421 For the remaining quota of material scrap

422 - 0.2 kg cables

423 - 2.0 kg polymeric fraction
424 - 1.5 kg low/high-content silicon material
425 - 0.8 kg residual
426
427 the assumed waste EoL scenario is incineration with energy recovery. It is assumed, therefore, that also fractions
428 comprehending mostly inert materials (e.g. silicon) are sent to incineration with energy recovery as mixed residuals; in
429 this case, a model specific for inert fraction in incinerators is used.

430 A specific model has been adopted for each fraction, as described in Table 4. The environmental modeling of
431 incineration with energy recovery accounts for both impacts due to energy consumption/emissions and credits
432 associated with energy production. However, the environmental burdens due to energy consumption/emissions are
433 much higher with respect to credits coming from energy production. As a confirmation, the EoL PV waste impact
434 reported in Figure 9 is positive for all materials/parts forwarded to incineration process (cables, junction boxes,
435 polymeric fraction, silicon materials and residual).

12
 PV MOREDE
Item Quantity Ecoinvent DB process
process
20.00 kg Pre-treatment
PLANT OPERATION

17.60 kg Glass separation

PV waste panel 3.30 kg Silicon separation -
Copper & Polymeric
1.00 kg
Input to the fraction separation
PV MOREDE plant 0.79 kWh Pre-treatment
0.51 kWh Glass separation
Electricity 0.46 kWh Silicon separation Electricity, medium voltage {RoW}
Copper & Polymeric
0.16 kWh
fraction separation

Origin Destination
Item Quantity (PV MOREDE process) (recovery facilities)
GaBi Thinkstep DB process

- GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*

- EU-28: Diesel mix at refinery ts (Transportation)
Aluminum 2.00 kg Retrieving foundry
- Aluminum - Scrap credit open loop (User) (Recycling)
- Aluminum ingot mix ts (Recycling)
- GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts(Transportation)*
- EU-28: Diesel mix at refinery ts (Transportation)
Junction boxes (Polyamide) 0.05 kg Pre-treatment
- Polyamide 6.6 - Scrap credit open loop (User) (Recycling)
Retrieving in - DE: Polyamide 6.6 Granulate (PA 6.6) Mix ts (Recycling)
conventional plants - GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*
Recycling route - EU-28: Diesel mix at refinery ts (Transportation)
EoL PV WASTE

Junction boxes (Polyethylene) 0.05 kg

(including transportation) - Polyethylene High Density (HDPE/PE-HD) - Scrap credit open loop (User) (Recycling)
- DE: Polyethylene High Density Granulate (HDPE/PE-HD) Mix ts (Recycling)
- GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*
Retrieving in - EU-28: Diesel mix at refinery ts (Transportation)
Glass 13.10 kg Glass separation
glassware - Glass - Scrap credit open loop (User) (Recycling)
- EU-28: Float flat glass ts (Recycling)
- GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*
Copper & Polymeric fraction - EU-28: Diesel mix at refinery ts (Transportation)
Copper 0.20 kg separation Retrieving foundry
- Copper - Scrap credit open loop (User) (Recycling)
- EU28: Copper sheet
- GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*
Cables 0.20 kg - EU-28: Diesel mix at refinery ts (Transportation)
Incineration route - EU-28: Cable waste in waste incineration plant ts (Incineration)
Pre-treatment Incineration of EVA
(including transportation) - GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*
Junction boxes
0.10 kg - EU-28: Diesel mix at refinery ts (Transportation)
(Polyamide/Polyethylene) - EU-28: Waste incineration of plastics (unspecified) ts (Incineration)

13
 - GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*
Polymeric fraction 1.20 kg Glass separation - EU-28: Diesel mix at refinery ts (Transportation)
- EU-28: Waste incineration of plastics (unspecified) ts (Incineration)

Low-content silicon materials 0.60 kg - GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*
Silicon separation - EU-28: Diesel mix at refinery ts (Transportation)
High-content silicon materials 0.90 kg - EU-28: Waste incineration of glass/inert material ELCD/CEWEP ts (Incineration)

- GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*

Polymeric fraction 0.80 kg - EU-28: Diesel mix at refinery ts (Transportation)
Copper & Polymeric fraction - EU-28: Waste incineration of plastics (unspecified) ts (Incineration)
separation
- GLO: Truck, Euro 5, 7.5-12t gross weight / 5t payload capacity ts (Transportation)*
Residual 0.80 kg - EU-28: Diesel mix at refinery ts (Transportation)
- EU-28: Waste incineration of glass/inert material ELCD/CEWEP ts (Incineration)

* Assumed distance to recycling / recovery treatment facilities = 150 km

436 Table 4. LCI data collection for plant operation and EoL of PV waste (all data refer to 1 FU - 20 kg PV crystallin-silicon based panel)

14
437 2.3. Life Cycle Impact Assessment (LCIA)
438 The base methodology chosen for the impact assessment of the considered system is the ILCD midpoint method
439 (Hiederer, 2011), recommended by the European Commission and scientifically accepted. This LCIA method includes
440 16 midpoint impact categories. Considering the goal of the study, the following impact categories are selected:

441 - Global Warming Potential

442 - Ozone Depletion Potential
443 - Photochemical Ozone Formation Potential
444 - Acidification Potential
445 - Terrestrial Eutrophication Potential
446 - Freshwater Eutrophication Potential
447 - Mineral, fossil & renewable resource depletion Potential

448 3. Interpretation of results and discussion

449 The LCIA results of the recycling of 20 kg PV panel through the PV-MOREDE process are illustrated in Table 5.

Mineral, fossil & renewable resource depletion Potential

Photochemical Ozone Formation Potential

Freshwater Eutrophication Potential

Terrestrial Eutrophication Potential
Global Warming Potential

Ozone Depletion Potential

Acidification Potential
[kg NMVOC eq.]

[Mole of H+ eq.]
[kg CFC11-eq.]

[Mole of N eq.]
[kg CO2 eq.]

[kg Sb eq.]
Plant Manufacturing 3.38E-02 1.34E-09 1.36E-04 1.73E-04 3.44E-04 [kg P eq.]
1.28E-05 1.82E-06

construction Transportation 5.00E-03 1.33E-16 1.44E-05 1.65E-05 8.34E-05 2.48E-08 1.83E-09

Plant operation 1.39E+00 4.03E-08 4.00E-03 1.03E-02 1.38E-02 6.53E-04 3.87E-06

Transportation 6.47E-01 1.74E-14 1.87E-03 2.15E-03 1.08E-02 3.24E-06 2.39E-07

EoL PV waste
Recovery processes -1.26E+01 5.87E-09 -6.92E-02 -1.66E-01 -3.58E-01 -9.54E-06 -1.40E-03

Total -1.06E+01 4.75E-08 -6.32E-02 -1.54E-01 -3.33E-01 6.60E-04 -1.39E-03

450
451 Table 5. Potential LC impacts for the recycling of 20 kg PV panel through the PV-MOREDE process
452
453 Figure 6 reports the contribution analysis of impacts due to plant construction and operation stages. The LCIA results
454 show that operation is definitely more influential than the construction, with a contribution that exceeds 95% for 6 of 7
455 impact categories. The reason for this is the considerable energy consumption required by the operation of the PV-
456 MOREDE process. On the other hand for the impact category Mineral, fossil & renewable resource depletion Potential
457 the contribution of construction stage is considerable (32 %) and it is due to the large amount of metals required to build
458 up the system.

15
459
460 Figure 6. Contribution analysis of impacts due to plant construction and operation

461 Figure 7 shows the contribution of each separation step to the overall operation stage impact. The pre-treatment and
462 panel shearing phases (considered together) present the highest contribution (about 40 %) for all the impact categories
463 mainly due to the amount of energy required for separating the aluminum frame and reducing the size of panel. This
464 step is also crucial in order to allow a high efficiency of the recycling process in terms of quantity and quality of the
465 different recyclable fractions. The contribution of both glass and silicon separation is about 25 % for all the impact
466 categories with the lowest contribution of copper & polymeric fraction separation (approximately 10 % for all the
467 categories). An interesting point emerges from the analysis of the operation stage. Unlike literature LCAs referring to
468 fixed recycling plants (Fraunhofer Institute, 2012; Latunussa e al., 2016), no contribution due to the transportation of
469 waste modules to the recycling/recovery facilities is added up, since the PV-MOREDE system is placed relatively close
470 to the installation site of EoL panels.

471
472 Figure 7. Contribution analysis by separation step of impacts due to plant operation. “Pre-treatment” includes also panel shearing
473 into squares.

474 Figure 8 reports the contribution analysis by LC stage of impacts including also recovery processes performed outside
475 the PV-MOREDE system (EoL PV waste stage). Recycling of aluminum, glass and copper for new PV panel
476 manufacturing or for other industrial activities yields significant environmental benefits due to avoided impacts for
477 production of virgin materials. Similarly, WtE recovery processes allow avoiding impacts due to the generation of
478 energy. Results show that credits from recovery of EoL waste counterbalances by far the environmental load related to
479 plant construction and operation, thus leading to a considerable reduction (negative values) of total LC impacts. This is
480 true for all the LCIA categories with the exception of impact categories Ozone Depletion Potential and Freshwater
481 Eutrophication Potential. For the Ozone Depletion Potential all LC stages provide positive impacts, the operation being
482 the most influential one; the burdens associated with EoL PV waste stage are primarily caused by air emissions of
483 incineration processes (incineration of polymeric fraction). Considering the Freshwater Eutrophication Potential, the
484 credits from recovery processes do not counterbalance the impacts of the other LC stages, especially plant operation.

16
485
486 Figure 8. Contribution analysis by LC stage of the overall LC impacts (plant construction, plant operation and EoL of PV waste)

487 Figure 9 shows the contribution analysis by material category of the impacts due to EoL PV waste stage. For the
488 environmental credits, the major quota is attributable to recycling of aluminum, glass and copper. On the other hand, the
489 incineration processes of polymeric and residual fractions provide a positive impact for all impact categories, especially
490 for the Ozone Depletion Potential for which credits do not counterbalance the impacts.

491
492
493 Figure 9. Contribution analysis by material category of impacts due to EoL PV waste

494 Considering literature regarding LCA of PV panels recycling processes, a direct comparison of results is not feasible
495 since existing studies often make use of different LCIA methods/impact categories, present results in aggregated form
496 or do not explicitly declare boundary conditions of the study (i.e. FU, system boundaries, assumptions on data source
497 and cut-off criteria). In particular, the main difference between the data here presented and other literature studies is that
498 the focus is not on the PV panel itself, but on the activity of recycling a panel using a certain plant (PV-MOREDE).
499 That said, it can be argued that the PV-MOREDE recycling system allows the achieving of notable environmental
500 benefits with respect to fixed installations thanks to the reduction of transportation processes for moving EoL panels
501 and resulting materials to the recycling site. As a confirmation, existing LCAs of fixed plants (Fraunhofer Institute,
502 2012; Latunussa et al., 2016) reveal that waste supply covers a significant amount of operation impact.

503 The main limitations of the work concern primarily two aspects. The first one is that LCI primary data regarding system
504 operation (technical/operational features of the plant, energy consumption for operation, amount of input/output
505 materials) refer to the pilot case plant built up by LME; as a consequence, process parameters could vary in the
506 transition to the industrial scale production. The second limitation is related to the modelling of processes performed
507 outside the PV-MOREDE facility (plant construction and recovery of EoL PV waste). These are production and
508 construction of different plant machines, transportation of system components to the LME assembly site, materials and
509 energy recovery processes. For these steps primary data are not available, and the LCI data collection is based on
510 secondary data coming from commercial LCI databases. Since recycling and incineration processes have a strong
511 influence on LC impact for all the considered impact categories, the representativeness of the used LCI databases is
512 crucial.
513 The added value of the work is that it deals with a mobile recycling process able to treat on-site and on demand modules
514 characterized by different composition/structure. The study provides transparent and disaggregated LCIA impacts with
515 a LCI mainly based on primary data. Considering that a substantial increase in the amount of EoL solar panels is
516 expected in the next decades, the need for recycling this type of waste will grow rapidly and the outcomes of the paper
17
517 can be relevant for different professional figures such as recyclers, LCA practitioners and policy makers. The work
518 could be also helpful in order to define strategies for the design of future PV panels characterized by higher
519 recoverability rate and lower environmental impacts.
520
521 4. Conclusions

522 This article performs the LCA of an innovative recycling process for EoL solar panels, experimented at a pilot scale
523 within the PV-MOREDE project. The investigation on plant construction and operation reveals that energy
524 consumption during operation involves much higher impacts than the consumption of material and energy resources
525 used for the construction. This is true for all the considered impact categories with the exception of Mineral, fossil &
526 renewable resource depletion Potential for which the large amount of metals required to build up the system makes the
527 contribution of construction comparable to the one of operation. Considering the operation stage, the pre-treatment step
528 gives the highest contribution (about 40 %) for all the impact categories due to the high energy consumption of the
529 shredding machine. Glass and silicon separation present similar quota (about 25 %) with the lowest contribution of
530 copper and polymeric fraction separation (approximately 10 % for all the categories).

531 The investigation of all LC stages including also processes performed outside the PV MOREDE system points out that
532 recycling from different material fractions (aluminum, glass, copper, silicon and plastics) allows achieving great
533 environmental benefits due to avoided production of virgin materials and energy. The LCIA results show that credits
534 from recovery counterbalance by far the burdens due to plant construction and operation for most of impact categories
535 and this is mainly due to recovery of metals and glass. The only exceptions are represented by impact categories Ozone
536 depletion Potential and Freshwater Eutrophication Potential for which the impact of plant operation is notably higher
537 with respect to the other LC stages. The investigation of literature shows that the environmental benefits achievable
538 through a mobile system are relevant with respect to fixed recycling installations not only due to the tailored process
539 which can be obtained on small scale systems, but also thanks to the overall reduction of transportation processes
540 needed to move EoL panels, materials and residuals to and from the recycling facility.

541 This LCA is the only work that deals with a mobile recycling process developed specifically for EoL PV panels. The
542 contribution of the study to the sustainability assessment of PV waste recycling lies in providing transparent and
543 disaggregated LCIA performed through a LC inventory mainly based on primary data. The findings of the research
544 could be relevant for different professionals such as recyclers, LCA practitioners and policy makers committed to
545 satisfying the ever-growing demand for environmentally friendly processes and products.
546
547 Acknowledgements
548 The present work is a part of the project PV–MOREDE (Photovoltaic Panels Mobile Recycling Device), co–funded by
549 the Eco–Innovation initiative of the European Union, contract number ECO/12/333078.
550 https://ec.europa.eu/environment/eco–innovation/projects/en/projects/pv–morede http://www.pvmorede.it/
551 The Publication as provided reflects only the authors’ view.
552 The authors would like to thank “La Mia Energia” Company (Venafro, Isernia – Italy) for their courtesy.
553

18
554 References
555
556 Banar, M., Özdemir, A., 2015. An evaluation of railway passenger transport in Turkey using life cycle assessment and
557 life cycle cost methods, In Transportation Research Part D: Transport and Environment, Volume 41, 2015, Pages
558 88-105, ISSN 1361-9209, https://doi.org/10.1016/j.trd.2015.09.017.

559 Berzi, L., Delogu, M., Giorgetti, A., Pierini, M., 2013. On-field investigation and process modelling of End-of-Life
560 Vehicles treatment in the context of Italian craft-type Authorized Treatment Facilities. Waste Management 33, 892–
561 906. https://doi.org/10.1016/j.wasman.2012.12.004

562 Besiou, M., Wassenhove, L.N.V., 2016. Closed-Loop Supply Chains for Photovoltaic Panels: A Case-Based Approach.
563 Journal of Industrial Ecology 20, 929–937. https://doi.org/10.1111/jiec.12297

564 Bhat Varun, I.K., Prakash, R., 2011. LCA of renewable energy for electricity generation systems - a review. Renew.
565 Sustain. Energy. Rev.15 (2011) 1625-1636.

566 Bio Intelligence Service (BioIS), 2011. Study on photovoltaic panels supplementing the impact assessment for a recast
567 of the WEEE directive - Final report, 2011.

568 Bogacka, M., Pikoń, K., Landrat, M., 2017. Environmental impact of PV cell waste scenario. Waste Management 70,
569 198–203. https://doi.org/10.1016/j.wasman.2017.09.007

570 Carnevale, E., Lombardi, L., Zanchi, L., 2014. Life Cycle Assessment of solar energy systems: Comparison of
571 photovoltaic and water thermal heater at domestic scale. Energy 77, 434–446.
572 https://doi.org/10.1016/j.energy.2014.09.028

573 Commission Regulation (EU) No 1179/2012 of 10 December 2012 establishing criteria determining when glass cullet
574 ceases to be waste under Directive 2008/98/EC of the European Parliament and of the Council.

575 Corcelli, F., Ripa, M., Leccisi, E., Cigolotti, V., Fiandra, V., Tammaro, M., Sannino, L., Graditi, G., Ulgiati, S., 2015.
576 Material recovery and energy savings from c-Si PV panels end-of-life. Life Cycle Assessment of PV panels thermal
577 treatment. Proceedings of the Global Cleaner Production and Sustainable consumption Conference, Sitges, Spain, 1-
578 4 November, 2015.

579 Contreras-Lisperguer, R., Muñoz-Cerón, E., Aguilera, J., Casa, J. de la, 2017. Cradle-to-cradle approach in the life
580 cycle of silicon solar photovoltaic panels. Journal of Cleaner Production 168, 51–59.
581 https://doi.org/10.1016/j.jclepro.2017.08.206

582 Cucchiella, F., D’Adamo, I., Lenny Koh, S.C., Rosa, P., 2016. A profitability assessment of European recycling
583 processes treating printed circuit boards from waste electrical and electronic equipments. Renew. Sustain. Energy
584 Rev. 64, 749–760. doi:10.1016/j.rser.2016.06.057

585 Del Pero, F., Delogu, M., Pierini, M., 2017. The effect of lightweighting in automotive LCA perspective: Estimation of
586 mass-induced fuel consumption reduction for gasoline turbocharged vehicles. Journal of Cleaner Production,
587 Volume 154, 2017, Pages 566-577, ISSN 0959-6526, https://doi.org/10.1016/j.jclepro.2017.04.013.

588 Delogu, M., Del Pero, F., Pierini, M., 2016. Lightweight Design Solutions in the Automotive Field: Environmental
589 Modelling Based on Fuel Reduction Value Applied to Diesel Turbocharged Vehicles. Sustainability 2016, 8(11),
590 1167; doi:10.3390/su8111167.

591 Dias, P., Javimczik, S., Benevit, M., Veit, H., 2016. Recycling WEEE: polymer characterization and pyrolysis study for
592 waste of crystalline silicon photovoltaic modules. Waste Manage. http://dx.doi.org/10.1016/j.wasman.2016.08.036.

19
593 Dias, P., Javimczik, S., Benevit, M., Veit, H., 2017. Recycling WEEE: Polymer characterization and pyrolysis study for
594 waste of crystalline silicon photovoltaic modules. Waste management, 60:716-722. doi: 10.1016/j.wasman. 2016.08.
595 036. Epub 2016 Sep 2.

596 DM 31/12/1988, 1988. Coefficients of depreciation of the cost of material assets used in for commercial activities, arts
597 and professions.

598 Directive 2012/19/EU of the European Parliament and of the Council of 4 July 2012 on waste electrical and electronic
599 equipment (WEEE) (recast), 2012.

600 Directive 2008/98/EC of the European Parliament and of the Council of 19 November 2008 on waste and repealing
601 certain Directives (Text with EEA relevance), 2008. , 312.

602 Doi, T., Tsuda, I., Unagida, H., Murata, A., Sakuta, K., Kurokawa, K., 2001. Experimental study on PV module
603 recycling with organic solvent method. Sol. Energy Mater. Sol. Cells 67, 397-403.

604 Duflou, J.R., Peeters, J.R., Altamirano, D., Bracquene, E., Dewulf, W., 2018. Demanufacturing photovoltaic panels:
605 Comparison of end-of-life treatment strategies for improved resource recovery. CIRP Annals 67, 29–32.
606 https://doi.org/10.1016/j.cirp.2018.04.053

607 Ecoinvent database, 2017. Available at: www.ecoinvent.org/database/database.html (accessed 12.12.17).

608 EPIA, European Photovoltaic Industry Association, 2014. Global Market Outlook for Photovoltaics 2014-2018.

609 Evans, A., Strezov, V., Evans, T.J., 2009. Assessment of sustainability indicators for renewable energy technologies.
610 Renew. Sustain. Energy. Rev. 13 (2009) 1082-1088.

611 Ferreira, S., Cabral, M., De Jaeger, S., Da Cruz, N.F., Simões, P., Marques, R.C., 2017. Life cycle assessment and
612 valuation of the packaging waste recycling system in Belgium. J Mater Cycles Waste Manag 19, 144–154.
613 https://doi.org/10.1007/s10163-015-0383-x

614 First Solar, 2019. First Solar’s state-of-the-art recycling facilities [WWW Document]. URL
615 http://www.firstsolar.com/Modules/Recycling (accessed 2.21.19).

616 Frankl, P., Menichetti, E., Raugei, M., Lombardelli, S., Prennushi, G., 2005. Final report on technical data, costs and
617 life cycle inventories of PV applications - Deliverable no. 11.2-RSIa of the NEEDS (New Energy Externalities
618 Developments for Sustainability) project, Rome, Italy, 2005.

619 Fraunhofer Institut für Bauphysik, 2012. Life Cycle Assessment (LCA) screening of the Maltha recycling process for
620 Si-PV modules – Executive Summary, available at: http://www.pvcycle.org/wp-content/uploads/Exec-Sum mary-
621 LCA-Screening-of-a-Recycling-process-of-silicon-based-PV-modules- 2012-07.pdf, (accessed 01.12.17).

622 Frisson, L., Lieten, K., Bruton, T., Declercq, K., Szlufcik, J., DeMoor, H., Goris, M., Benali, A., Aceves, O., 2000.
623 Recent improvements in industrial PV module recycling, in: Proceedings of the 16th European Photovoltaic Solar
624 Energy Conference, Glasgow, 1–5 May, 2000.

625 Granata, G., Pagnanelli, F., Moscardini, E., Havlik, T., Toro, L., 2014. Recycling of photovoltaic panels by physical
626 operations. Solar Energy Materials and Solar Cells 123, 239–248. https://doi.org/10.1016/j.solmat.2014.01.012

627 Grassi, A., Delogu, M., Baldanzini, N., Berzi, L., Pierini, M., 2018. Structural analysis of a mobile device for the End–
628 of–Life treatment of photovoltaic panels. Procedia Structural Integrity. Volume 8, Pages 594-603, ISSN 2452-3216.

629 Gustafsson, A.M.K., Foreman, M.R.S.J., Ekberg, C., 2014. Recycling of high purity selenium from CIGS solar cell
630 waste materials. Waste Manage. 34 (10), 1775-1782.

631 Hawkins, T.R., Singh, B., Majeau-Bettez, Anders Hammer Strømman, A.H., 2012. Comparative Environmental Life
632 Cycle Assessment of Conventional and Electric Vehicles. Journal of Industrial Ecology. DOI: 10.1111/j.1530-
633 9290.2012.00532 (2012).

634 Held, M., Ilg, R., 2011. Update of environmental indicators and energy paybacktime of CdTe PV systems in Europe,
635 Prog. Photovolt.: Res. Appl. 19 (2011) 614-626.

20
636 Hiederer, R., 2011. International reference life cycle data system (ILCD) handbook: general guide for life cycle
637 assessment: provisions and action steps. European Commission, Joint Research Centre - Institute for Environment
638 and Sustainability.

639 Hsu, D.D., O’Donoughue, P., Fthenakis, V.S., Heath, G.A., Kim, H.C., Sawyer, P., Choi, J.K., Turney, D.E., 2012. Life
640 cycle greenhouse gas emissions of crystalline silicon photovoltaic electricity generation: systematic review and
641 harmonization. J. Ind. Ecol.16 (S1)(2012) 122-135.

642 IEA, 2018. Snapshot of global photovoltaic market 2018 - Report PVPS T1-33:2018 [WWW Document]. URL
643 http://www.iea-pvps.org/?id=266 (accessed 10.21.18).

644 IRENA, 2018. Renewable power generation costs in 2017. International Renewable Energy Agency, Abu Dhabi.

645 ISO 14040/14044, 2006. Environmental management - Life Cycle Assessment - Principals and framework/
646 Requirements and guidelines. Genew, Switzerland.

647 Kanga, S., Yoo, S., Lee, J., Boo, B., Ryu, H., 2012. Experimental investigations for recycling of silicon and glass from
648 waste photovoltaic modules. Renew. Energy 47, 152-159.

649 Klugmann-Radziemska, E., Ostrowski, P., Drabczyk, K., Panek, P., Szkodo, M., 2010a. Experimental validation of
650 crystalline silicon solar cells recycling by thermal and chemical methods. Sol. Energy Mater. Sol. Cells 94, 2275-
651 2282.

652 Klugmann-Radziemska, E., Ostrowski, P., 2010b. Chemical treatment of crystalline silicon solar cells as a method of
653 recovering pure silicon from photovoltaic modules. Renew. Energy 35 (8), 1751-1759.

654 Koffler, C., Florin, J., 2013. Tackling the Downcycling Issue—A Revised Approach to Value-Corrected Substitution in
655 Life Cycle Assessment of Aluminum (VCS 2.0). Sustainability 5, 4546–4560. https://doi.org/10.3390/su5114546

656 Jung, B., Park, J., Seo, D., Park, N., 2016. Sustainable system for raw-metal recovery from crystalline silicon solar
657 panels: From noble-metal extraction to lead removal. ACS Sustainable Chem. Eng. 4, 4079-4083.

658 Jungbluth,N., 2005. Life cycle assessment of crystalline photovoltaics in the Swiss ecoinvent database, Prog. Photovolt.
659 13 (5) (2005) 429–446.

660 Jungbluth, N., Stucki, M., Flury, K., Frischknecht, R., Büsser, S., 2012. Life cycle inventories for photovoltaics:
661 Version 2012 – Report by the ESU – services Ltd on behalf of the Swiss Federal Office of Energy (SFOE), 2012.

662 Latunussa, C.E.L., Ardente, F., G.A. Blengini, Mancini, L., 2016. Life Cycle Assessment of an innovative recycling
663 process for crystalline silicon photovoltaic panels. Solar EnergyMaterials&SolarCells156(2016)101-111.

664 Lee, T.D., Ebong, A.U., 2017. A review of thin film solar cell technologies and challenges. Renewable and Sustainable
665 Energy Reviews 70, 1286–1297. https://doi.org/10.1016/j.rser.2016.12.028

666 LME, 2015. PV-MOREDE DELIVERABLE D 3.1 First PV-Morede manufactured [WWW Document]. URL
667 http://www.pvmorede.it/public/D%203.1%201st%20PVMOREDE%20MANUFACTURED.PDF

668 Marwede, M., Reller, A., 2012. Future recycling flows of tellurium from cadmium telluride photovoltaic waste.
669 Resources, Conservation and Recycling 69, 35–49. https://doi.org/10.1016/j.resconrec.2012.09.003

670 Müller, A., Wambach, K., Alsema, E., 2005. Life cycle analysis of solar module recycling process. Proceedings of the
671 20th European Photovoltaic Solar Energy Conference, Barcelona, Spain, 6–10 June, 2005.

672 Munshi, A.H., Sasidharan, N., Pinkayan, S., Barth, K.L., Sampath, W.S., Ongsakul, W., 2018. Thin-film CdTe
673 photovoltaics – The technology for utility scale sustainable energy generation. Solar Energy 173, 511–516.
674 https://doi.org/10.1016/j.solener.2018.07.090

675 Pagnanelli, F., Moscardini, E., Altimari, P., Abo Atia, T., Toro, L., 2016. Cobalt products from real waste fractions of
676 end of life lithium ion batteries. Waste Management, Volume 51, 2016, Pages 214-221, ISSN 0956-053X,
677 https://doi.org/10.1016/j.wasman.2015.11.003.

678 Pagnanelli, F., Moscardini, E., Granata, G., Abo Atia, T., Altimari, P., Havlik, T., Toro, L., 2017. Physical and chemical
679 treatment of end of life panels: An integrated automatic approach viable for different photovoltaic technologies.

21
680 Waste Management, Volume 59, 2017, Pages 422-431, ISSN 0956-053X,
681 http://dx.doi.org/10.1016/j.wasman.2016.11.011.

682 Park, J., Park, N., 2014. Wet etching processes for recycling crystalline silicon solar cells from end-of-life photovoltaic
683 modules. RSC Adv. 4, 34823–34829. https://doi.org/10.1039/C4RA03895A

684 Peng, J., Lu, L., Yang, H., 2013. Review on life cycle assessment of energy paybackand greenhouse gas emission of
685 solar photovoltaic systems, Renew. Sustain. Energy Rev.19 (2013) 255–274.

686 Perez-Gallardo, J.R., Azzaro-Pantel, C., Astier, S., 2018. A Multi-objective Framework for Assessment of Recycling
687 Strategies for Photovoltaic Modules based on Life Cycle Assessment. Waste Biomass Valor 9, 147–159.
688 https://doi.org/10.1007/s12649-017-9878-0.
689 Ramos-Ruiz, A., Wilkening, J.V., Field, J.A., Sierra-Alvarez, R., 2017. Leaching of cadmium and tellurium from
690 cadmium telluride (CdTe) thin-film solar panels under simulated landfill conditions. Journal of Hazardous Materials
691 336, 57–64. https://doi.org/10.1016/j.jhazmat.2017.04.052

692 Reggi, R., 2017. De-assembling system for a photovoltaic panel enabling salvage of original materials. EP3089825B1.

693 Rigamonti, L., Grosso, M., Giugliano, M., 2009. Life cycle assessment for optimising the level of separated collection
694 in integrated MSW management systems. Waste Management 29, 934–944.
695 https://doi.org/10.1016/j.wasman.2008.06.005

696 Rocchetti, L., Veglio, F., Kopacek, B., Beolchini, F., 2013. Environmental impact assessment of hydrometallurgical
697 processes for metal recovery from WEEE residues using a portable prototype plant. Environ. Sci. Technol. 47,
698 1581-1588.

699 Rocchetti, L., Beolchini, F., 2015. Recovery of valuable materials from end-of-life thin-film photovoltaic panels:
700 environmental impact assessment of different management options. Journal of Cleaner Production 89, 59–64.
701 https://doi.org/10.1016/j.jclepro.2014.11.009

702 Savvilotidou, V., Antoniou, A., Gidarakos, E., 2017. Toxicity assessment and feasible recycling process for amorphous
703 silicon and CIS waste photovoltaic panels. Waste Management 59, 394–402.
704 https://doi.org/10.1016/j.wasman.2016.10.003

705 Schrijvers, D.L., Loubet, P., Sonnemann, G., 2016. Developing a systematic framework for consistent allocation in
706 LCA. Int J Life Cycle Assess 21, 976–993. https://doi.org/10.1007/s11367-016-1063-3

707 Sherwani, A.F., Usmani, J.A., Varun, 2010. Life cycle assessment of solar PV based electricity generation systems: a
708 review. Renew. Sustain. Energ. Rev.14 (2010) 540–544.

709 Shibasaki, M., Warburg, N., Springer, J., Lombardelli, S., 2006. Recycling of thin-film solar modules life cycle
710 assessment case study, in: Proceedings of the 21st European Photovoltaic Solar Energy Conference, Dresden,
711 Germany, 4–8 September, 2006.

712 Silva, M.D. et al., 2014. Integration of technologies and alternative sources of water and energy to promote the
713 sustainability of urban landscapes. Resour. Conserv. Recycl. 91, 71-81.

714 Somorin, T.O., Di Lorenzo, G., Kolios, A.J., 2017. Life-cycle assessment of self-generated electricity in Nigeria and
715 Jatropha biodiesel as an alternative power fuel, In Renewable Energy, Volume 113, 2017, Pages 966-979, ISSN
716 0960-1481, https://doi.org/10.1016/j.renene.2017.06.073.

717 SPE, 2018. Global Market Outlook 2018-2022 - Solar Power Europe. http://www.solarpowereurope.org/wp-
718 content/uploads/2018/09/Global-Market-Outlook-2018-2022.pdf (last visualization: January 2019).
719 Stenzel, P., Schreiber, A., Marx, J., Wulf, C., Schreieder, M., Stephan, L., 2017. Renewable energies for Graciosa
720 Island, Azores – Life Cycle Assessment of electricity generation. Energy Procedia, Volume 135, 2017, Pages
721 62-74, ISSN 1876-6102, https://doi.org/10.1016/j.egypro.2017.09.487.
722 Tao, J., Yu, S., 2015. Review on feasible recycling pathways and technologies of solar photovoltaic modules. Solar
723 Energy Materials and Solar Cells 141, 108–124. https://doi.org/10.1016/j.solmat.2015.05.005

22
724 Tian, J., Chen, M., 2014. Sustainable design for automotive products: Dismantling and recycling of end-of-life vehicles.
725 Waste Management, Volume 34, Issue 2, 2014, Pages 458-467, ISSN 0956-053X,
726 https://doi.org/10.1016/j.wasman.2013.11.005.

727 Vellini, M., Gambini, M., Prattella, V., 2017. Environmental impacts of PV technology throughout the life cycle:
728 Importance of the end-of-life management for Si-panels and CdTe-panels. Energy 138, 1099–1111.
729 https://doi.org/10.1016/j.energy.2017.07.031

730 Veolia, 2018. Veolia opens the first European plant entirely dedicated to recycling photovoltaic panels [WWW
731 Document]. Veolia. URL https://www.veolia.com/en/newsroom/news/recycling-photovoltaic-panels-circular-
732 economy-france (accessed 3.1.19).

733 Wäger, P.A., Hischier, R., 2015. Life cycle assessment of post-consumer plastics production from waste electrical and
734 electronic equipment (WEEE) treatment residues in a Central European plastics recycling plant. Science of The
735 Total Environment 529, 158–167. https://doi.org/10.1016/j.scitotenv.2015.05.043

736 Weckend, S., Wade, A., Heat, G., 2016. End-Of-Life Management - Solar Photovoltaic panels - IEA-PVPS Report
737 Number: T12-06:2016.

738 Wild-Scholten, M.J., 2009. Energy pay-back times of PV modules and systems. Proceedings of the Workshop
739 Photovoltaik-Modultechnik. Koln, 26-27 November, 2009.

740 Xu, Y., Li, J., Tan, Q., Peters, A.L., Yang, C., 2018. Global status of recycling waste solar panels: A review. Waste
741 Management 75, 450–458. https://doi.org/10.1016/j.wasman.2018.01.036.

742 Zeng, X., Song, Q., Li, J., Yuan, W., Duan, H., Liu, L., 2015. Solving e-waste problem using an integrated mobile
743 recycling plant. Journal of Cleaner Production, 90 (2015) 55-59.

744 Zuser, A., Rechberger, H., 2011. Considerations of resource availability in technology development strategies: the case
745 study of photovoltaics. Resour. Conserv. Recycl. 56 (1), 56-65.

23