Content uploaded by Parikhit Sinha
Author content
All content in this area was uploaded by Parikhit Sinha on Jun 25, 2015
Content may be subject to copyright.
1
Report IEA-PVPS T12-05:2015
Life Cycle Assessment of Future Photovoltaic
Electricity Production from Residential-scale
Systems Operated in Europe
INTERNATIONAL ENERGY AGENCY PHOTOVOLTAIC POWER SYSTEMS
PROGRAMME
Life Cycle Assessment of Future Photovoltaic
Electricity Production from Residential-scale Systems
Operated in Europe
IEA-PVPS Task 12, Subtask 2.0, LCA Report IEA-PVPS T12-05:2015
March 2015
ISBN 978-3-906042-30-5
Operating Agent:
Garvin Heath, National Renewable Energy Laboratory, Golden, CO, USA
Authors:
Rolf Frischknecht, René Itten, Franziska Wyss,
Contributors:
Isabelle Blanc, Garvin Heath, Marco Raugei, Parikhit Sinha, Andreas Wade
Citation: R. Frischknecht, R. Itten, F. Wyss, I. Blanc, G. Heath, M. Raugei, P. Sinha,
A. Wade, 2014, Life cycle assessment of future photovoltaic electricity
production from residential-scale systems operated in Europe, Subtask 2.0
"LCA", IEA-PVPS Task 12.
i
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
Abbreviations and Acronyms
ADEME Agency for Environment and Energy Management, France
APAC Asia & Pacific
BAU business-as-usual (scenario)
BWR boiling water reactor (nuclear power plant)
CCS carbon capture and storage
CdTe cadmium-telluride
CFC chlorofluorocarbon
CH Switzerland
CN China
CO2 carbon dioxide
CO2 eq carbon dioxide equivalents
CSP concentrating solar power (solar power production)
DE Germany
EAA European Aluminium Association
EH&S environmental, health and safety
ENTSO European Network of Transmission System Operators
EPIA European Photovoltaic Industry Association
FBR fluidized-bed reactor
FHI-ISE Fraunhofer Institute for Solar Energy Systems
GHG greenhouse gas
GLO global average
HFC hydrofluorocarbon
IEA International Energy Agency
IEA-PVPS International Energy Agency Photovoltaic Power Systems Programme
IIASA International Institute of Applied Systems Analysis
IPCC Intergovernmental Panel on Climate Change
kW kilowatt
kWh kilowatt-hour
kWp kilowatt-peak
LCA life cycle assessment
LCI life cycle inventory analysis
LCIA life cycle impact assessment
MG Metallurgical grade silicon
MJ megajoule
MJ oil-eq megajoule oil equivalents
Multi-Si multi-crystalline silicon based photovoltaics
ii
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
MW megawatt
NEEDS New Energy Externalities Development for Sustainability
NMVOC non-methane volatile organic compounds
NO Norway
NREPBT non-renewable energy payback time
OECD Organization for economic cooperation and development
OPT optimistic improvement (scenario)
PM10 particulate matter with a diameter of 10 µm and lower
PV photovoltaics
PVPS Photovoltaic Power Systems Programme
PWR pressure water reactor (nuclear power plant)
R & D Research and development
REAL Realistic improvement (scenario)
RER Europe
RLA Latin America and the Caribbean
single-Si single-crystalline
SO2 sulphur dioxide
tkm ton kilometre, unit for transport services
UCTE Union for the Coordination of the Transmission of Electricity
US United States (as used to define world regions = North America)
iii
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
Executive Summary
The photovoltaics (PV) industry is growing rapidly to meet the increasing demand of
green power. As the industry grows, the manufacturing processes and the material and
energy efficiencies of PV cells and panels are improving. To assess the impacts of this
trend, future scenarios of single-crystalline (single-Si, also known as mono-crystalline)
silicon and cadmium-telluride (CdTe) PV systems installed on European residences
were established. Assessment of the improvement potential of PV electricity-generating
technologies such as single-Si and CdTe could be considered in long-term energy
strategy decisions.
This study aims to provide scenario-based information about the environmental per-
formance of single-Si and CdTe PV modules produced and operated in the far future
(2030 to 2050). The deployment application assessed considers European residential
roofs. We made scenario-dependent projections of key parameters for single-Si and
CdTe PV panels manufactured in 2050. The parameters included cell efficiency, module
efficiency, wafer thickness, cutting losses, kerf losses, silver use, glass thickness and
operational lifetime (see Tab. S.1).
Tab. S.1 Key parameters of silicon-based single-crystalline and CdTe photovoltaic cells and modules
and values used in the three scenarios BAU, REAL and OPT.
Parameter
Single-Si
CdTe
TODAY
BAU
REAL
OPT
TODAY
BAU
REAL
OPT
Cell efficiency
16.5 %
25.0 %
27.0 %
29.0 %
15.6 %
22.8 %
24.4 %
26.0 %
Derate cell to module
efficiency
8.5 %
8.5 %
6.8 %
5.0 %
13.9 %
10.0 %
7.5 %
5.0 %
Module efficiency
15.1 %
22.9 %
25.2 %
27.6 %
13.4 %
20.5 %
22.6 %
24.7 %
Wafer thickness / layer
thickness
190 m
150 m
120 m
100 m
4.0 m
2.0 m
1.0 m
0.1 m
Electricity
demand in CdTe
laminate manufacture
-
-
-
-
100 %
86 %
81 %
74 %
Kerf loss
190 m
150 m
120 m
100 m
-
-
-
-
Silver per cell
9.6 g/m2
9.6 g/m2
5.0 g/m2
2.0 g/m2
-
-
-
-
Fluidized-bed reactor
(FBR) Share of Poly Si
Production
0 %
20 %
40 %
100 %
-
-
Glass thickness
4.0 mm
4.0 mm
3.0 mm
2.0 mm
3.5 mm
3.5 mm
3.0 mm
2.0 mm
Operational lifetime
30 years
30 years
35 years
40 years
30 years
30 years
35 years
40 years
iv
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
We combined developments for these parameters with projections of the environmental
performance of electricity mixes in the main manufacturing countries/regions (European
Union, China and the United States of America) and with projections of the environ-
mental performance of basic material production (aluminium, copper, magnesium,
nickel, pig iron, zinc, clinker and flat glass) in the far future. The three scenarios used in
the assessment of future PV electricity were categorized into three classes: “business as
usual” (BAU), “realistic improvement” (REAL) and “optimistic improvement” (OPT).
We estimate the current life cycle greenhouse gas emissions of single-Si PV electricity
produced on the roofs of European residences to be approximately 80 grams CO2-equi-
valent per kWh (g CO2-eq per kWh). Based on the projected changes to key parameters
and the background system, life cycle greenhouse gas (GHG) emissions could be
reduced to 65 % (scenario BAU), 31 % (scenario REAL) and 18 % (scenario OPT) of
that value in the far future. (The caption of Fig. S.1 specifies the module characteristics
evaluated.) Results for other life cycle assessment (LCA) metrics assessed here are also
shown in Fig. S.1: non-renewable cumulative energy demand, acidification potential,
human toxicity potential, photochemical ozone creation potential, particulate matter
formation potential, and land use.
v
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
Fig. S.1 Estimates of life cycle greenhouse gas emissions (using 100 year global warming potentials
from IPCC (2013)), non-renewable cumulative energy demand (following Frischknecht et al.
(2007c)), acidification potential, human toxicity potential, photochemical ozone creation
potential, particulate matter formation potential and land use (following Goedkoop et al.
(2009)) of electricity produced in the far future with single-crystalline silicon-based
photovoltaic laminates mounted on slanted roofs in Europe according to the three scenarios
(BAU, REAL and OPT). Results for “today” are defined to be 100%, with the three scenarios
as fractions thereof. Key assumptions are: module efficiency: 15.1 % (today), 22.9 % (BAU),
25.2 % (REAL), 27.6 % (OPT); annual yield (electricity generated per kWp of the PV power
plant and year): 975 kWh/kWp including degradation (10.5 % average for lifetime); solar
irradiation: 1 331 kWh/m2. Lifetime of the PV power plant: 30 years (today and BAU),
35 years (REAL), 40 years (OPT). The system includes mounting, cabling, inverter and
maintenance and considers production in different regions of the world (Europe, North
America, China and Asia & Pacific) using region-specific electricity mixes. This is a
prospective LCA for expected future development in the year 2050. The calculations are
performed using the software SimaPro with ecoinvent v2.2+ as background database.
We calculated the total energy payback time (EPBT) and the non-renewable energy
payback time (NREPBT) in the far future of single-crystalline silicon-based PV panels
operated in Europe by dividing the estimate of non-renewable cumulative energy
demand of PV electricity for the given scenario by the non-renewable cumulative
energy demand of the scenario-dependent national and regional non-renewable residual
electricity mixes.
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Greenhouse gas emissions
Cumulative energy demand, non-renewable
Acidification potential
Human toxicity potential
Photochemical ozone creation potential
Particulate matter formation potential
Land use
today
BAU
REAL
OPT
vi
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
We estimate the payback time could be reduced from 2.4 years today to 1.7, 1.2 and
1.2 years (scenarios BAU, REAL and OPT, respectively) in the far future, based on the
assumptions and projections in our analysis.
We estimate the current life cycle greenhouse gas emissions of CdTe PV electricity
produced on the roofs of European residences to be approximately 30 g CO2-eq per
kWh. Based on the projected changes to key parameters and the background system, life
cycle GHG emissions could be reduced to 70 % (scenario BAU), 44 % (scenario
REAL) and 32 % (scenario OPT) of that value in the far future (see Fig. S.2).
vii
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
Fig. S.2 Estimates of life cycle greenhouse gas emissions (using 100 year global warming potentials
from the most recent report of Working group 1 of the Intergovernmental Panel on Climate
Change IPCC (2013)), non-renewable cumulative energy demand (following Frischknecht et
al. (2007c)), acidification potential, human toxicity potential, photochemical ozone creation
potential, particulate matter formation potential and land use (following Goedkoop et al.
(2009)) of electricity produced in the far future with CdTe photovoltaic laminates mounted on
slanted roofs in Europe, according to the three scenarios (BAU, REAL and OPT). Results for
“today” are defined to be 100 %, with the three scenarios as fractions thereof. Key assumptions
are: module efficiency: 13.4 % (today), 20.5 % (BAU), 22.6 % (REAL), 24.7 % (OPT); annual
yield (electricity generated per kWp of the PV power plant and year): 975 kWh/kWp including
degradation (10.5 % average for lifetime); solar irradiation: 1 331 kWh/m2. Lifetime: 30 years
(today and BAU), 35 years (REAL), 40 years (OPT). The system includes mounting, cabling,
inverter and maintenance and considers production using region-specific electricity mixes. This
is a prospective LCA for expected future development in the year 2050; the calculations are
performed using the software SimaPro with ecoinvent v2.2+ as background database.
We estimate that the NREPBT of CdTe PV operated in Europe could be reduced from
1.1 years currently to 0.8, 0.7 and 0.7 years (scenarios BAU, REAL and OPT,
respectively) in the far future, based on the assumptions and projections in our analysis.
The NREPBT is lower in the scenario REAL than it is in the scenario OPT because of
the different residual electricity mixes that are replaced. The total energy payback times
(EPBT) are between 0 and 0.4 years higher than the NREPBT.
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Greenhouse gas emissions
Cumulative energy demand, non-renewable
Acidification potential
Human toxicity potential
Photochemical ozone creation potential
Particulate matter formation potential
Land use
today
BAU
REAL
OPT
viii
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
Tab. S.2 Future non-renewable residual electricity mixes for Europe (ENTSO-E) in the three scenarios
BAU, REAL and OPT
Power plant technology/fuel
BAU
REAL
OPT
Hard coal
34.1 %
0.0
8.7 %
Hard coal with carbon capture and storage (CCS)
0.0 %
8.1 %
6.3 %
Lignite
12.5 %
0.0 %
0.0 %
Fuel oil
0.8 %
0.3 %
0.0 %
Natural gas
3.9 %
0.1 %
24.2 %
Natural gas, gas combined cycle
20.2 %
4.3 %
59.4 %
Natural gas, gas combined cycle, CCS
0.0 %
53.2 %
1.4 %
Natural gas, fuel cell
0.0 %
0.2 %
0.0 %
Nuclear
28.5 %
33.9 %
0.0 %
Total
100.0 %
100.0 %
100.0 %
The study suggests that future developments in the PV industry, the electricity sector
and material supply could significantly reduce environmental impacts per kWh of PV-
generated electricity compared to those of today. The LCA results of this analysis could
help support long-term energy policy measures related to renewable energies. The
results are based on a set of assumptions and projections that use the best available
information and are specific to residential-scale rooftop systems operated in Europe.
They are subject to considerable uncertainty, especially when projecting more than
30 years for such fast-evolving technologies. Therefore, the results of this analysis are
best interpreted as indicating the currently expected direction and approximate relative
magnitude of change for the PV industry rather than as precise predictions of absolute
impacts in future years. While the absolute magnitude of results will change if different
locations or applications are considered, the direction and relative magnitude of
projected changes in impacts compared to the current situation is likely consistent with
those reported here, and therefore informative to energy decisions with long-term
consequences.
ix
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
Foreword
The International Energy Agency (IEA), founded in November 1974, is an autonomous
body within the framework of the Organization for Economic Cooperation and
Development (OECD) that carries out a comprehensive programme of energy co-
operation among its member countries. The European Commission also participates in
the work of the IEA.
The IEA PVPS is one of the collaborative R&D Agreements established within the IEA,
and was established in 1993. The overall programme is headed by an Executive
Committee composed of representatives from each participating country and/or
organisation, while the management of individual research projects (Tasks) is the
responsibility of Operating Agents. By early 2015, fifteen Tasks were established within
the PVPS programme, of which six are currently operational.
The IEA PVPS Implementing Agreement presently has 29 members and covers the
majority of countries active in photovoltaics, both in R&D, production and installation.
The programme deals with the relevant applications of photovoltaics, both for on-grid
and off-grid markets. It operates in a task-shared mode whereby member countries
and/or organisations contribute with their experts to the different Tasks. The co-
operation deals with both technical and non-technical issues relevant to a wide-spread
use of photovoltaics in these different market segments.
The mission of the IEA PVPS programme is: “To enhance the international
collaborative efforts which facilitate the role of photovoltaic solar energy as a
cornerstone in the transition to sustainable energy systems.” The underlying assumption
is that the market for PV systems is rapidly expanding to significant penetrations in
grid-connected markets in an increasing number of countries, connected to both the
distribution network and the central transmission network. At the same time, the market
is gradually shifting from a policy to a business driven approach.
Task 12 engages in fostering international collaboration in communicating and
assessing the environmental, health and safety aspects associated with the
environmental, health and safety (EH&S) aspects of PV technology over the life cycle
of the PV systems. Task 12 also disseminates reliable and accurate information on the
EH&S impacts of PV technology to policymakers, industry participants and the public
with the goal to improve consumer understanding and confidence, encourage industry
best practices and aid policymakers to make informed decisions in the course of the
energy transition. Furthermore, Task 12 brings its expertise in assessing methods and
standards for the evaluation of EH&S aspects of PV systems. The overall objectives of
Task 12 are to:
Quantify the environmental profile of PV electricity using a life cycle approach, in
order to contribute to the environmental sustainability of the supply chain and to
compare it with the environmental profile of electricity produced with other energy
technologies
x
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
Aim for a closed-loop supply chain by and help improve waste management of PV
by collective action on collection and recycling, including legislative developments
as well as development of technical standards
Distinguish and address actual and perceived issues touching the EH&S aspects of
PV technology that are important for market growth.
The first objective of this task is well served by life cycle assessments (LCAs) that
describe the energy-, material- and emission-flows in all the stages of the life cycle of
PV. The second objective will be addressed by assisting the collective action of PV
companies in defining material availability and product-recycling issues.
Within Task 12, a Subtask on “Life Cycle Assessment” includes three targets: to
quantify the environmental profile of electricity produced with PV systems (compared
to that from other sources); to evaluate trends in the environmental profile of PV; and,
to assess this profile with the help of "external" costs and other life cycle impact
assessment methods. In addition, Task 12 has produced and will continue to update
methodological guidelines for PV LCA. Further information on the activities and results
of the Task can be found at www.iea-pvps.org.
xi
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
Content
1 INTRODUCTION, MOTIVATION AND OVERVIEW 1
2 GOAL AND SCOPE 2
2.1 Goal of the Study 2
2.2 Functional Unit 2
2.3 System Boundary 2
2.4 Assumptions Related to the Operation of Photovoltaic Panels 2
2.5 Geographical, Temporal and Technical Validity 4
2.6 Scenarios 4
2.7 Data Sources and Modelling 5
2.8 Impact Assessment Methods 5
2.9 Non-renewable Energy Payback Time (NREPBT) 6
2.10 Total Energy Payback Time (EPBT) 6
3 TECHNOLOGIES ANALYSED AND KEY PARAMETERS VARIED 8
3.1 Overview 8
3.2 Photovoltaic Technologies Analysed 8
3.3 Key Parameters 8
3.3.1 Overview 8
3.3.2 Cell Efficiency 9
3.3.3 Derate from Cell to Module Efficiency 10
3.3.4 Module Efficiency 10
3.3.5 Wafer / Layer Thickness 12
3.3.6 Electricity Demand in CdTe Laminate Manufacture 13
3.3.7 Kerf Loss 14
3.3.8 Silver Contacts 14
3.3.9 Fluidised Bed Reactor (FBR) Share in Polysilicon Production 15
3.3.10 Glass Thickness 15
3.3.11 Operational Lifetime 16
xii
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2014
4 LIFE CYCLE INVENTORIES IN THE PV SUPPLY CHAINS 17
4.1 Overview 17
4.2 How to Read an EcoSpold Table 17
4.3 Solar-grade Silicon Production Using FBR 18
4.4 Metallization Paste 20
4.5 Photovoltaic Electricity Production Mixes 21
4.6 Non-renewable Residual Electricity Mixes for EPBT and NREPBT 23
5 LIFE CYCLE INVENTORIES IN THE BACKGROUND SYSTEM 24
5.1 Overview 24
5.2 Scenarios Applied 24
5.3 Electricity Mix ENTSO-E 25
5.4 Electricity Mix in China 29
5.5 Electricity Mix in the United States 32
5.6 Metals and Materials 37
5.7 Aluminium 58
5.7.1 Electricity Mix of the Aluminium Industry 59
5.7.2 Bauxite Mining 60
5.7.3 Aluminium Oxide Production 61
5.7.4 Anode Production 61
5.7.5 Liquid Aluminium Production 62
5.7.6 Primary Aluminium Production 65
6 CUMULATIVE RESULTS AND INTERPRETATION 67
6.1 Overview 67
6.2 Life Cycle GHG Emissions of Future 3-kWp Plants 67
6.2.1 Single-crystalline Silicon Photovoltaic Laminate 67
6.2.2 Cadmium-telluride Photovoltaic Laminate 68
6.3 Environmental Impacts of Future PV Electricity 69
6.3.1 Single-crystalline Silicon Photovoltaic Electricity 70
6.3.2 Cadmium-telluride Photovoltaic Electricity 73
6.4 Non-renewable and Total Energy Payback Time 76
6.4.1 Single-crystalline Silicon Photovoltaic Power Plant 76
1. Introduction, Motivation and Overview 1
1 Introduction, Motivation and Overview
Photovoltaics (PV) are considered a promising electricity producing technology that could
play an important role in replacing fossil and nuclear power plants and reducing the
environmental impacts of the electricity mixes of countries and regions. Long-term energy
planning and assessment relies on scenarios of the future development of the price of oil,
economic growth assumptions and the like. Similarly, the environmental assessment of future
electricity supply should rely on information about possible future developments with regard
to the energy and material efficiencies of electricity producing technologies as well as the
environmental efficiency in manufacturing these technologies. Furthermore, possible
developments in the supply chains of basic materials - such as aluminium (primary), copper,
cement - and in the electricity mixes of producing countries and regions need to be taken into
account.
Comprehensive and consistent assessments of the environmental impacts of power plants to
be operated between 2030 and 2050 were carried out in NEEDS, a research project
1
within
the 6th framework program of the European Union (Frischknecht et al. 2007a; Frischknecht &
Krewitt 2008; Frischknecht 2010). The results showed that the environmental impacts depend
on the scenario chosen and that it is important to adjust not only the foreground system (i.e.,
the PV supply chain) but also the background system (i.e., the material supply and the
electricity mix used in the PV manufacturing supply chain).
Because the PV industry has advanced since the first assessment in 2008, an update of the
environmental assessment of future PV electricity production would be helpful. It should be
emphasized that the specific results found in this study are based on a set of assumptions and
projections that use the best available information. As such, they are subject to considerable
uncertainty, especially when projecting 30 or more years for such a fast-evolving technology.
Therefore, the results of this assessment are best understood in the context of the information
available today and for the purpose of clarifying the currently expected direction and
approximate relative magnitude of change, rather than for their precise, absolute results. In
addition, this report only considers one PV application: small power plants installed on
European residential rooftops. While the absolute magnitude of results will change if different
locations or applications are considered, the direction and relative magnitude of projected
changes in impacts compared to the current situation is likely consistent with those reported
here.
This report describes the goal and scope of the key parameters (Section 2) used in the future
scenarios (Section 2.10) of the life cycle inventories (LCIs) of the systems analysed (Section
4), the LCIs used in the background system (Section 5) and the results of the future-oriented
LCAs (Section 6). The report is completed with conclusions in Section 7.
1
For more information about the New Energy Externalities Developments for Sustainability project, see
www.needs-project.org.
2. Goal and Scope 2
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
2 Goal and Scope
2.1 Goal of the Study
The goal of this study is to assess the life cycle environmental impacts of the future
production of electricity in Europe with two different types of PV laminates - silicon-based
single-crystalline silicon and cadmium-telluride - manufactured globally and operated
between 2030 and 2050 on Europe residential roofs.
PV laminates were selected for analysis because unframed and building-integrated laminates
cause fewer environmental impacts than framed and mounted panel laminates. Projections for
future material demand of the panel frames and the mounting structures are not part of this
study; therefore, the relative importance of the panel frames and the mounting devices relative
to the modules will increase over time (i.e., environmental impacts of the modules are
decreasing, whereas the impacts of the frames and the mounting systems remain constant).
Although the report often refers to the year 2050, the analyses carried out rather reflect a
situation between 2030 and 2050 and thus are long term future projections based on
assessments of changes in key parameters rather than projections for the specific year 2050.
2.2 Functional Unit
The functional unit used in this study is 1 kWh of electricity supplied to the grid in the long
term future.
2.3 System Boundary
The production system of the future PV electricity produced with crystalline silicon-based and
cadmium-telluride (CdTe) solar cells comprises:
- raw material extraction
- wafer, cell and module manufacture
- mounting structures manufacture
- inverters manufacture
- system installation
- operation (cleaning of the modules)
- end-of-life treatment.
2.4 Assumptions Related to the Operation of Photovoltaic Panels
The use phase is characterised by three main parameters: annual yield, degradation rate and
lifetime.
2. Goal and Scope 3
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
The annual yield of electricity depends on the location of installation, the mounting and
orientation of the modules (façade versus roof top, inclination and orientation) and the
degradation rate. Tab. 2.1 shows the cumulative installed PV power in Europe, according to
IEA-PVPS (2013) and the country-specific average yield at optimal angle in urban areas,
according to the report published by the European Photovoltaic Industry Association, EPIA
(2012). The annual average yield of optimally oriented modules in Europe, weighted accor-
ding to the cumulative installed PV power corresponds to 1 090 kWh/kWp (excluding degra-
dation effects) with an average solar irradiation of about 1 330 kWh/m2 and optimally
oriented panels.
Tab. 2.1 Cumulative installed photovoltaic power in Europe in 2012 according to IEA-PVPS (2013), country
specific average annual yield in kWh/kWp at optimal angle in urban areas according to EPIA (2012),
and average solar irradiation at optimal angle, based on data retrieved from PVGIS
2
. Degradation is
not included.
In line with the IEA-PVPS methodology guidelines (Fthenakis et al. 2011) and the Agency for
Environment and Energy Management (ADEME) methodology guidelines (Payet et al. 2013),
a degradation of 0.7 % per year is applied leading to a loss in yield of 21 % during the last
year of an operation time of 30 years. Hence, the weighted average yield of a PV module
installed in Europe and operated for 30 years is 10.5 % below the average yield shown in Tab.
2.1. The European PV modules are thus modelled with an annual yield of 975 kWh per kWp.
2
re.jrc.ec.europa.eu/pvgis/ (accessed on 29.04.2014)
Country
Cumulative
installed power
(MW)
Share
Average yield at
optimal angle in
urban areas
(kWh/kWp)
Average solar
irradiation at
optimal angle
(kWh/m2)
Austria 363 1% 1 027 1 314
Belgium 2 698 4% 930 1 100
Germany 32 462 51% 936 1 147
Denmark 332 1% 945 1 130
Spain 4 706 7% 1 471 1 812
France 4 033 6% 1 117 1 386
United Kingdom 1 901 3% 920 1 111
Italy 16 450 26% 1 326 1 611
Netherlands 345 1% 933 1 112
Portugal 210 0% 1 494 1 840
Sweden 24 0% 826 1 101
Europe (PVPS members) 63 524 100% 1 090 1 331
2. Goal and Scope 4
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
2.5 Geographical, Temporal and Technical Validity
The future PV supply chain covers four world regions and countries: Europe (RER), North
America (US), Asia & Pacific (APAC) and China (CN). In combination with information on
all levels of the PV supply chain, specific market mixes for the four regions are derived and
modelled. This includes both produced and installed PV capacities in the four regions. For the
purpose of this study, and without references providing alternative projections, the market
shares for the year 2012 is assumed to remain constant for all projection scenarios analysed
(see Section 2.6).
This project explores scenarios for the long term future, tied to projections for a set of key
parameters. The three scenarios represent pessimistic, realistic and optimistic projections of
technology development of producing polysilicon (also called metal grade silicon, the
feedstock material for semiconductor and PV industries), solar-grade silicon, of
manufacturing wafers, cells and panels (CdTe and single-crystalline silicon), of material and
energy efficiency of cells and panels and of supplying basic materials and electricity used in
the PV supply chain.
2.6 Scenarios
The future life cycle environmental impacts of two different major PV technologies are
analysed in this study: silicon-based single-crystalline (single-Si) PV modules and cadmium
telluride PV modules (CdTe). For each PV technology, three scenarios are evaluated: a
business-as-usual scenario (BAU), a realistic improvement scenario (REAL) and an optimistic
improvement scenario (OPT). Tab. 2.2 summarizes and describes the three scenarios. The
scenarios BAU, REAL and OPT correspond to the scenarios “pessimistic”, “realistic-
optimistic” and “very optimistic” according to the NEEDS terminology.
The assumptions and parameters being used in the three scenarios are described in detail in
Subchapter 3.3 and in Chapter 5.
Tab. 2.2 Overview and characterisation of the three scenarios
Scenario name
Abbreviation
Comment
Corresponding scenarion in NEEDS
Business-as-usual
BAU
Pessimistic scenario with limited
improvement
Scenario “pessimistic”: continuation
of established policies. No energy
goals are set.
Realistic
improvements
REAL
Realistic scenario between BAU und
OPT
Scenario “realistic optimistic”:
renewable energy sources as well as
energy efficient technologies are
pushed intensely. Key technologies
are advanced systematically and
energy politics have a high priority.
Optimistic
improvements
OPT
Optimistic scenario using the most
ambitious future projections for the
key parameters
Scenario “very optimistic”: highly
ambitiousenergy policies. Efficient
technologies are supported and
pushed
2. Goal and Scope 5
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
2.7 Data Sources and Modelling
The SimaPro (version 7.3.3) a commercial LCA software, is used to model the production
systems, to calculate the LCI and impact assessment results (PRé Consultants 2012). Most
background data are represented by ecoinvent data v2.2 (ecoinvent Centre 2010) and further
updates (LC-inventories 2012) unless otherwise noted. Key materials, electricity mixes of
selected countries as well as the PV supply chains are represented by data meant to represent
a longterm future. Data sets are documented and published in EcoSpold v1 format
3
.
2.8 Impact Assessment Methods
The environmental impact assessment is performed on the mid-point level, at some point
between the release of a substance and the potential damage it may cause. For instance, the
greenhouse gases are aggregated according to their radiative forcing potential. Thus we are
not quantifying environmental impacts in terms of damages but aggregating flows that
contribute to the same effect (e.g. climate change).
The following indicators are used in this study:
Greenhouse gas emissions (kg CO2-eq), assessed assuming 100 year global warming
potentials based on the latest IPCC 2013 (Tab. 8.A.1): This indicator aggregates
greenhouse gases emitted according to their radiative forcing capacity relative to the
reference substance CO2.
Cumulative energy demand, non-renewable (MJ oil-eq) (Frischknecht et al. 2007b): This
indicator aggregates fossil and nuclear energy resources on the basis of their upper heating
value (fossil energy resources) and the energy extractable from 1 kg fissible uranium in a
nuclear light water reactor.
Acidification potential (kg SO2-eq) evaluated using the ReCiPe midpoint method
(hierarchist perspective) (Goedkoop et al. 2009): This indicator aggregates pollutants
potentially contributing to the acidification of water bodies and soils based on the capacity
of binding H+ ions relative to the reference substance SO2.
Human toxicity potential
4
(kg 1,4-DB eq), evaluated using the ReCiPe midpoint method
(hierarchist perspective) (Goedkoop et al. 2009): This indicator aggregates substances
potentially toxic to humans relative to the reference substance 1,4-dichlorobenzene.
Photochemical ozone creation potential (kg NMVOC), evaluated using the ReCiPe
midpoint method (hierarchist perspective) (Goedkoop et al. 2009): This indicator
aggregates substances potentially contributing to summer smog situations (via ozone
formation), expressed relative to the ozone creating potential of an average NMVOC.
3
The files in the EcoSpold v1 format are available at www.treeze.ch/projects/case-studies/energy/photovoltaic
4
More recent impact category indicators are available (USETox) which may lead to substantially different
outcomes, in particular with regard to the impact reduction potential in the longterm future.
2. Goal and Scope 6
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Particulate matter formation potential (kg PM10-eq), evaluated using the ReCiPe midpoint
method (hierarchist perspective) (Goedkoop et al. 2009): This indicator aggregates
particulate matter directly emitted and substances transformed in the atmosphere to
particulate matter (secondary particulates). It covers PM, NOX, SO2 and NH3 emitted to
air with PM being the reference substance.
Urban land occupation (m2a), evaluated using the ReCiPe midpoint method (hierarchist
perspective) (Goedkoop et al. 2009). This indicator quantifies the area multiplied by time
of land being occupied by buildings, power plants, factories and other infrastructures
(roads, railway tracks, dams).
2.9 Non-renewable Energy Payback Time (NREPBT)
The non-renewable energy payback time (Fthenakis et al. 2011, Frischknecht et al. 2007c) is
defined as the period required for a renewable energy system to generate the same amount of
energy (in terms of non-renewable primary energy equivalent) that was used to produce the
system itself. It considers non-renewable energy sources such as hard coal, lignite, crude oil,
natural gas and uranium. The NREPBT is calculated using the following formula:
Emat: Non renewable primary energy demand to produce materials comprising PV system
Emanuf: Non renewable primary energy demand to manufacture PV system
Etrans: Non renewable primary energy demand to transport materials used during the life cycle
Einst: Non renewable primary energy demand to install the system
EEOL: Non renewable primary energy demand for end-of-life management
Eagen: Annual electricity generation
EO&M: Annual Non renewable primary energy demand for operation and maintenance
G: Grid efficiency, average non renewable primary energy to electricity conversion efficiency at the demand
side
2.10 Total Energy Payback Time (EPBT)
The total energy payback time (Fthenakis et al. 2011) is defined as the period required for a
renewable energy system to generate the same amount of energy (in terms of total primary
energy equivalent) that was used to produce the system itself. It considers all non renewable
and renewable energy sources, except for the direct solar radiation input during the operation
phase, which is not accounted for as part of EO&M. The EPBT is calculated using the follow-
ing formula:
2. Goal and Scope 7
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Emat: Total primary energy demand to produce materials comprising PV system
Emanuf: Total primary energy demand to manufacture PV system
Etrans: Total primary energy demand to transport materials used during the life cycle
Einst: Total primary energy demand to install the system
EEOL: Total primary energy demand for end-of-life management
Eagen: Annual electricity generation
EO&M: Annual total primary energy demand for operation and maintenance, excluding direct solar radiation input
G: Grid efficiency, average total primary energy to electricity conversion efficiency at the demand side
3. Technologies Analysed and Key Parameters Varied 8
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
3 Technologies Analysed and Key Parameters Varied
3.1 Overview
The PV technologies analysed in this project are described in Section 3.2. In Section 3.3, the
key parameters varied over time are introduced, and scenario-dependent values are presented
and substantiated.
3.2 Photovoltaic Technologies Analysed
This future analysis focuses on two PV technologies: unframed single-crystalline silicon-
based PV laminate and unframed CdTe PV laminate. The use of framed PV panels mounted
on roofs is not part of the analysis; only installation as unframed laminate integrated in the
building is analysed for both technologies.
Single-crystalline (single-Si) silicon-based PV modules cover more than 40 % of the annual
global production of PV power plants (kWp) in 2012, based on FHI-ISE (2013)
5
, and they are
selected for analysis here because of their high share in the PV electricity production mix.
Cadmium-telluride PV modules are selected as a comparably inexpensive and emerging
technology in the PV market with a considerable annual global production of PV power plants
(kWp) in 2012 (6.3 %, according to FHI-ISE (2013)5).
3.3 Key Parameters
3.3.1 Overview
Several key parameters determine the environmental performance of electricity produced with
PV modules. Future projections of these key parameters of the different PV technologies are
made for each of the three scenarios. The key parameters, which are part of the future
scenario analysis in this study, are listed in Tab. 3.1. The definition of each key parameter is
described in detail in Sections 3.3.2 to 3.3.9.
Balance of system components such as inverters or mounting structures were not subject to
future forecasts. Hence, the material and environmental efficiency of these components is
assumed to remain constant and reflects the current situation in all three scenarios.
5
FHI-ISE (2013) cites the data originally published by Navigant Consulting (2012).
3. Technologies Analysed and Key Parameters Varied 9
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Tab. 3.1 Key parameters of silicon-based single-crystalline and CdTe photovoltaic cells and modules.
Parameter
Single-
Si
CdTe
Comment
Section
Cell efficiency
yes
yes
Describes the efficiency of the solar cells
3.3.2
Derate cell to
module
efficiency
yes
yes
Describes the efficiency loss between cells and modules
3.3.3
Module
efficiency
yes
yes
Describes the efficiency of the PV module
3.3.4
Wafer thickness /
layer thickness
yes
yes
Describes the thickness of the silicon wafer and the CdTe
layer, respectively
3.3.5
Electricity
demand
no
yes
Describes the electricity demand during the manufacture of
CdTe PV laminate
3.3.6
Kerf loss
yes
no
Describes the kerfing losses during single-Si wafer sawing
3.3.7
Silver per cell
yes
no
Describes the amount of silver used for electric contacts in
a single-Si cell
3.3.8
Fluidized-bed
reactor (FBR)
Share of Poly Si
Production
yes
no
Describes the share of the most efficient production
technology for polysilicon (silicon feedstock): fluidized-
bed reactor
3.3.9
Glass thickness
yes
yes
Describes the thickness of the solar glass used on the back
and the front side of the solar cell
3.3.10
Operational
lifetime
yes
yes
Describes the lifetime of the PV modules
3.3.11
The LCI data of all the modelled technologies and processes are available at www.treeze.ch.
The LCI data can be downloaded and adjusted to model alternative future projections of
certain parameters such as efficiency and lifetime.
3.3.2 Cell Efficiency
Tab. 3.2 shows the future projections of the single-Si and CdTe cell efficiencies used in the
three scenarios. Each scenario shows a considerable increase in cell efficiency compared to
the current average efficiency of single-Si PV cells (16.5 %, de Wild-Scholten 2013) and
CdTe PV cells (15.6 %, module efficiency according to First Solar 2014) and own
calculations using derate cell to module efficiency according to Garabedian (2013).
The calculated maximum efficiency for silicon-based crystalline PV cells is 33 %, according
to the Shockley-Queisser limit (Shockley & Queisser 1961). This efficiency is 4 % higher
than the efficiency of 29 % used in the scenario OPT. According to Swanson (2005) 29 % is
the maximum efficiency, which can be achieved, due to practical reasons.
The calculated maximum efficiency for CdTe PV cells according to Garabedian (2013) is
30 %. The maximum practical efficiency is derived analogously as for silicon-based single-
crystalline PV cells (maximum theoretical efficiency minus 4 %) and corresponds to 26 %.
3. Technologies Analysed and Key Parameters Varied 10
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Tab. 3.2 Scenario-dependent cell efficiency of single-Si and CdTe PV cells.
Cell Efficiency
Single-Si
Source
CdTe
Source
Unit
Percent
Percent
Current LCI
16.5%
de Wild-Scholten 2013,
Photon International 2013
15.6%
calculated based on module
efficiency and cell to module
derate
BAU
25.0%
Goodrich et al. 2013
22.8%
Garabedian 2014, mid term
target
REAL
27.0%
interpolated
24.4%
interpolated
OPT
29.0%
Swanson 2005
26.0%
Garabedian 2013, max
efficiency minus 4%
Theoretical maximum
33.0%
Shockley-Queisser
30.0%
Garabedian 2013
3.3.3 Derate from Cell to Module Efficiency
Tab. 3.3 shows the derate of the efficiency from cell to module of single-Si and CdTe PV
modules used in the three scenarios. The derate of the current CdTe PV modules (13.9 %) is
higher than the derate of the single-Si PV modules (8.5 %). Due to technological
improvements, derates comparable to the single-Si PV modules are expected for the CdTe PV
modules. Thus, the future derate varies between 5 % and 10 %, as reported by Goodrich et al.
(2013) for single-Si. A derate of 5 % is assumed in the scenario OPT. Derates of 8.5 % and
10 % are used in the BAU scenario for single-Si and CdTe modules, respectively. For the
scenario REAL, the average between the derates of the OPT and BAU scenarios is used.
Tab. 3.3 Scenario-dependent derate of the efficiency from cell to module of single-Si and CdTe PV modules.
Derate Cell to
Module Efficiency
Single-Si
Source
CdTe
Source
Unit
Percent
Percent
Current LCI
8.5%
calculated based on de Wild-
Scholten 2013, Photon
International 2013
13.9%
calculated based on Garabedian
2014
BAU
8.5%
based on current LCI
10.0%
Goodrich et al. 2013
REAL
6.8%
interpolated
7.5%
interpolated
OPT
5.0%
Goodrich et al. 2013
5.0%
Goodrich et al. 2013
3.3.4 Module Efficiency
Tab. 3.4 shows the scenario-dependent module efficiency of single-Si and CdTe PV modules.
The future module efficiencies are calculated based on the future cell efficiencies (see Tab.
3.2) and the future cell-to-module derates (see Tab. 3.3).
3. Technologies Analysed and Key Parameters Varied 11
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Each scenario shows a considerable increase in the module efficiency compared to the current
average efficiency of single-Si (15.1 %) and CdTe (13.4 %) PV modules.
Tab. 3.4 Scenario-dependent module efficiency of single-Si and CdTe PV modules.
Module Efficiency
Single-Si
Source
CdTe
Source
Unit
Percent
Percent
Current LCI
15.1%
de Wild-Scholten 2013,
Photon International 2013
13.4%
First Solar 2014
BAU
22.9%
calculated based on cell
efficiency, derate 8.5%
20.5%
calculated based on cell
efficiency, derate 10%
REAL
25.2%
calculated based on cell
efficiency, derate 6.8%
22.6%
calculated based on cell
efficiency, derate 7.5%
OPT
27.6%
calculated based on cell
efficiency, derate 5%
24.7%
calculated based on cell
efficiency, derate 5%
Shockley-Queisser
31.4%
calculated based on cell
efficiency, derate 5%
Fig. 3.1 shows the development of the module efficiency of silicon-based single-crystalline
cells between 2000 and 2012, according to de Wild-Scholten (2013) and Photon International
(2013), with linear extrapolation until 2050. Both the scenario-dependent module efficiencies
of the single-Si technology and the Shockley-Queisser limit (33 % cell efficiency) at a derate
of 5 % (dashed line) are also shown. Fig. 3.1 also shows that the expected module efficiency
in the scenario OPT would require a slightly higher annual increase in efficiency as compared
to the past 12 years, whereas the expected efficiency in the scenarios BAU and REAL reflect
a decrease in annual efficiency improvements. The most optimistic module efficiency is about
4 % below the technical maximum given by the Shockley-Queisser limit.
3. Technologies Analysed and Key Parameters Varied 12
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Fig. 3.1 Development of the module efficiency of silicon-based single-crystalline cells between 2000 and 2012
according to de Wild-Scholten (2013) and Photon International (2013) with linear extrapolation to the
year 2050 and selected module efficiencies of the three scenarios BAU, REAL & OPT and the
Shockley-Queisser limit (33 % cell efficiency) at a derate of 5 %.
3.3.5 Wafer / Layer Thickness
Tab. 3.5 shows the silicon wafer thickness of single-Si modules and CdTe layer thickness of
CdTe PV modules. The wafer thickness and the thickness of the CdTe layer influence the
silicon demand of the single-Si modules and the CdTe demand of CdTe modules.
The reduction in the wafer thickness of single-Si modules is based on future projections by
the International Technology Roadmap for Photovoltaic (ITRPV 2013). The roadmap projects
wafter thicknesses between 100 and 120 µm for the PV cells; these values are used in the
scenarios REAL (120 µm) and OPT (100 µm). In the BAU scenario, a reduction of the wafer
thickness to 150 µm is assumed, based on author expert judgment.
The reduction in the layer thickness of CdTe modules are based on future projections shown
in Marwede & Reller (2012) and Woodhouse et al. (2013). The future projections of Marwede
& Reller (2012) and Woodhouse et al. (2013) have different time references (2040 and 2030,
respectively). However, the future projections are considered comparable because they are
both long-term projections (i.e., more than 15 years).
The utilisation rate of CdTe per µm of the cell layer is assumed as constant. The amount of
CdTe used in the production of the PV laminate is of only minor importance (compared to the
energy use).
0
5
10
15
20
25
30
35
1990 2000 2010 2020 2030 2040 2050 2060
Module efficiency in percen t
Year
Single-Si Single-Si BAU
Single-Si REAL Single-Si OPT
Shockley-Queisser, derate 5% Linear (Single-Si)
3. Technologies Analysed and Key Parameters Varied 13
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Tab. 3.5 Scenario-dependent silicon wafer and CdTe layer thickness of single-Si and CdTe PV modules.
Wafer thickness / layer
thickness
Single-Si
Source
CdTe
Source
Unit
micrometer
micrometer
Current LCI
190
Jungbluth et al. 2012
4.0
Jungbluth et al. 2012
BAU
150
Expert judgment
2.0
Marwede & Reller 2012
Woodhouse et al. 2013
REAL
120
ITRPV 2013 (upper range)
1.0
Marwede & Reller 2012
Woodhouse et al. 2013
OPT
100
ITRPV 2013 (lower range)
0.1
Marwede & Reller 2012
Woodhouse et al. 2013
3.3.6 Electricity Demand in CdTe Laminate Manufacture
Tab. 3.6 shows the electricity demand during the manufacturing of CdTe laminate relative to
current demand. The reduction of the electricity demand was derived using the following
assumptions
6
(the three effects are additive):
10 % overall reduction of the electricity demand (manufacturing efficiency gains)
4 % reduced electricity demand per 0.7 mm of reduced glass thickness
2 % reduced electricity demand per 1 µm of reduced CdTe layer
The electricity demand during the manufacturing of single-Si laminate remains unchanged,
but the electricity demand of the global production of the raw materials (solar-grade silicon,
fluidized-bed-reactor (FBR)) is reduced.
The electricity demand of the single-crystalline-silicon global production remains unchanged
because no future projections of the energy demand have been available.
6
Personal communication: Parikhit Sinha, First Solar, 08.01.2014
3. Technologies Analysed and Key Parameters Varied 14
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Tab. 3.6 Scenario-dependent electricity demand during the manufacturing of CdTe laminate, relative to today’s
electricity demand.
Electricity Demand
CdTe
Source
Unit
%
Current LCI
100%
Jungbluth et al. 2012
BAU
86%
pers. Communication Parikhit
Sinha, First Solar, 08.01.20146
REAL
81%
pers. Communication Parikhit
Sinha, First Solar, 08.01.20146
OPT
74%
pers. Communication Parikhit
Sinha, First Solar, 08.01.20146
3.3.7 Kerf Loss
Tab. 3.7 shows the kerf loss (i.e., the loss of silicon due to slicing of multi- or single-
crystalline silicon) of single-Si modules used in the three scenarios. The kerf loss is assumed
to correspond to the wafer thickness shown in Tab. 3.5; hence, the thinner the wafer, the less
the kerf loss. For a detailed description of the selected values, see Section 3.3.5.
Tab. 3.7 Scenario-dependent kerf loss of single-Si modules.
Kerf Loss
Single-Si
Source
Unit
micrometer
Current LCI
190
Jungbluth et al. 2012
BAU
150
Expert judgment
REAL
120
ITRPV 2013 (upper range)
OPT
100
ITRPV 2013 (lower range)
3.3.8 Silver Contacts
Tab. 3.8 shows the scenario-dependent amount of silver used per cell of single-Si module.
The silver is used for electrical contacts on the cells. Silver can be replaced with less-
expensive copper. This replacement is taken into account in the LCIs by increasing the
amount of copper used in the projection scenarios.
3. Technologies Analysed and Key Parameters Varied 15
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Tab. 3.8 Amount of silver per cell of single-Si module used in the three scenarios.
Silver per Cell
Single-Si
Source
Unit
g / m2
Current LCI
9.6
Jungbluth et al. 2012
BAU
9.6
unchanged
REAL
5.0
interpolated
OPT
2.0
ITRPV 2013
3.3.9 Fluidised Bed Reactor (FBR) Share in Polysilicon Production
Tab. 3.9 shows the share of FBR polysilicon global production in the silicon supply of single-
Si modules used in the three scenarios. FBR polysilicon production is more energy efficient
than traditional processes and therefore helps reduce environmental impacts in the silicon
supply chain. A share of 100 % FBR is assumed in the scenario OPT.
Tab. 3.9 Scenario-dependent share of FBR polysilicon global production in the silicon supply of single-Si
modules.
FBR Share of Poly Si
Production
Single-Si
Source
Unit
percent
Current LCI
100% Siemens
Jungbluth et al. 2012
BAU
20% FBR
/ 80% Siemens
expert judgement
REAL
40% FBR
/ 60% Siemens
ITRPV 2013
OPT
100% FBR
expert judgement
3.3.10 Glass Thickness
Tab. 3.10 shows the scenario-dependent thickness of the solar glass of single-Si and CdTe
modules. The glass used in CdTe modules is slightly thinner than the one used in single-Si
modules. For the future, a thickness of 2 mm is assumed for both technologies, as there are no
major differences regarding the module construction that would justify different glass
thicknesses.
3. Technologies Analysed and Key Parameters Varied 16
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Tab. 3.10 Scenario-dependent thickness of the solar glass of single-Si and CdTe modules used in the three
scenarios. (The values in brackets indicate the thickness of the additional glass layer which is needed
in case of frameless modules with glass-glass structure
7
. Such a scenario could be evaluated in future
assessments of single-Si PV; it represents current technology of CdTe PV.)
Glass thickness
Single-Si
Source
CdTe
Source
Unit
mm
mm
Current LCI
4.0 (+4.0)
de Wild-Scholten & Alsema
2007, Jungbluth et al. 2012
3.5 (+3.5)
Jungbluth et al. 2012, First
Solar 2011
BAU
4.0 (+4.0)
unchanged
3.5 (+3.5)
unchanged
REAL
3.0 (+3.0)
interpolated
3.0 (+3.0)
interpolated
OPT
2.0 (+2.0)
ITRPV 2013
2.0(+2.0)
ITRPV 2013
3.3.11 Operational Lifetime
Tab. 3.11 shows the scenario-dependent operational lifetime of single-Si and CdTe modules.
According to IEA (2010), an increase from 30 years to 40 years in the operational lifetime of
the PV modules can be expected (scenario OPT). However, in the scenario BAU, an
unchanged lifetime of 30 years is assumed. The lifetime used in the scenario REAL is
interpolated between these two values.
Tab. 3.11 Operational lifetime of single-Si and CdTe modules used in the three scenarios.
Operational lifetime
Single-Si
Source
CdTe
Source
Unit
years
years
Current LCI
30
Jungbluth et al. 2012
30
Jungbluth et al. 2012
BAU
30
current value
30
current value
REAL
35
interpolated
35
interpolated
OPT
40
IEA 2010
40
IEA 2010
7
Personal communication: Andreas Wade, First Solar, 12.12.2013
4. Life Cycle Inventories in the PV Supply Chains 17
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
4 Life Cycle Inventories in the PV Supply Chains
4.1 Overview
The data used in this analysis are based on the inventories of the PV supply chain described in
Itten et al. (2014, Part I). The key parameters are adjusted in the scenarios according to the
description in Section2.10. For sake of readability, only unit process raw data of the adjusted
processes in the PV supply chain are described in this report; data sets of processes that
remained unchanged can be found in Itten et al. (2014).
In Section 4.3, the unit process data of the newly introduced production technology for solar-
grade silicon (fluidized bed reactor, FBR) are shown. In Section 4.4, the adjusted unit process
data of the metallization paste (reduced use of silver) are shown. Section 4.5 documents the
adjusted PV electricity production mixes, and Section 4.6 shows the non-renewable residual
electricity mixes, which are used to calculate the NREPBT.
4.2 How to Read an EcoSpold Table
The Ecospold tables are the tables presented in the following Subchapters and in Chapter 5
describing the life cycle inventory datasets developed within this project.
How to read the tables.
The light green fields describe the name of the product/process, its region (e.g. RER stands
for Europe) and the unit data it refers to. It is the output product (the reference output) of the
process and always equal to '1'. The yellow fields show the inputs and outputs of the
respective processes. The grey fields specify whether it is an input from or an output to nature
or technosphere and the compartment to which a pollutant is emitted. For each product,
additional descriptive information is given in separate tables.
The location codes (an extended ISO alpha-2 code-set) have the following meaning:
Regions: Countries:
APAC Asia Pacific AU Australia JP Japan
ENTSO European electricity network CH Switzerland KR South Korea
GLO Global CN China NO Norway
OCE Oceanic DE Germany NZ New Zealand
RER Europe ES Spain US United States of
America
4. Life Cycle Inventories in the PV Supply Chains 18
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
4.3 Solar-grade Silicon Production Using FBR
Tab. 4.1 shows the unit process data of solar-grade silicon production using fluidised bed
reactors (FBR) in Europe (RER), China (CN), North America (represented by US) and Asia
& Pacific (APAC). The inventory is based on the solar-grade silicon production as described
by Jungbluth et al. (2012).
The electricity consumption of the FBR deposition process is significantly lower than that of
the Siemens process. de Wild-Scholten & Alsema (2005) estimate that the electricity
consumption is about 30 kWh/kg, but no information is provided regarding possible other
energy sources or working materials.
To approximate the new FBR technology, the electricity demand reported in the unit process
data set of the Siemens process is reduced to 30 kWh/kg. All other material uses and
emissions remain unchanged, as no detailed LCI data are available. This LCA could be
updated after a complete LCI for the FBR process becomes available.
Tab. 4.2 shows the silicon production mixes for PV in the four different regions distinguished
in this analysis. The share of the new FBR technology in supplying solar-grade silicon is
varied between the scenarios as shown in Tab. 3.9. Tab. 4.2 shows the silicon production
mixes for the scenario OPT with a share of 100 % FBR technology.
4. Life Cycle Inventories in the PV Supply Chains 19
LCA of future photovoltaics electricity production IEA-PVPS T12-05:2015
Tab. 4.1 Unit process data of solar-grade silicon production using fluidised bed reactors (FBR) in Europe
(RER), China (CN), North America (US) and Asia & Pacific (APAC).
Name
Location
InfrastructureProcess
Unit
silico n, solar
grade, fluidis ed
bed reactor (FBR),
at plant
silico n, solar
grade, fluidis ed
bed reactor (FBR),
at plant
silico n, solar
grade, fluidis ed
bed reactor (FBR),
at plant
silico n, solar
grade, fluidis ed
bed reactor (FBR),
at plant
UncertaintyType
StandardDeviation95%
GeneralCom ment
Location RER CN US APAC
InfrastructureProces s 0 0 0 0
Unit kg kg kg kg
product
silico n, solar grade, fluidis ed bed reactor (FBR),
at plant
RER 0 kg 1 0 0 0
silico n, solar grade, fluidis ed bed reactor (FBR),
at plant
CN 0 kg 0 1 0 0
silico n, solar grade, fluidis ed bed reactor (FBR),
at plant
US 0 kg 0 0 1 0
silico n, solar grade, fluidis ed bed reactor (FBR),
at plant
APAC 0 kg 0 0 0 1
technosp here MG-silicon, at plant NO 0 kg 1.13E+0 0 0 0 1 1.10 (2,3,1,2,1,3); Litera ture
MG-silicon, at plant CN 0 kg 0 1.13 E+0 0 0 1 1.10 (2,3,1,2,1,3); Literature
MG-silicon, at plant US 0 kg 0 0 1.13E+0 0 1 1.10 (2,3,1,2,1,3); Litera ture
MG-silicon, at plant APAC 0 kg 0 0 0 1.13E+0 1 1.10 (2,3,1,2,1,3); Literature
hydrochloric acid, 30% in H2O, at plant RER 0 kg 1 .60E+0 1.60E+0 1.60E+0 1.60E+0 1 1.14
(3,3,1,2,1,3); de Wild 2007, sha re of NaOH,
HCl and H2 es timated with EG-Si data
hydrogen, liquid , at plant RER 0 kg 5.01E-2 5.01 E-2 5.01E-2 5.01E-2 1 1.14
(3,3,1,2,1,3); de Wild 2007, sha re of NaOH,
HCl and H2 es timated with EG-Si data
sodiu m hydroxide, 50% in H2O, production mi x,
at plant
RER 0 kg 3.48E-1 3.48E-1 3.48E-1 3.48E-1 1 1.14
(3,3,1,2,1,3); de Wild 2007, sha re of NaOH,
HCl and H2 es timated with EG-Si data
transport, lorry >16t, fleet average RER 0 tkm 2.66E+0 2.66E+0 2.66E+0 2.66E+0 1 2.09
(4,5,na,na,na,na); Dis tance 2000km plu s
100 km for chem icals
transport, freight, rail RER 0 tkm 2.40E+0 2.40E+0 2.40E+0 2.40E+0 1 2.09
(4,5,na,na,na,na); 600km for chemicals
includin g solvent
transport, trans oceanic freight ship OCE 0 tkm 5.30E+0 0 0 0 1 2.06
(2,3,2,2,3,2); Transport of REC sil icon from
US to European m arket
electricity, at cogen 1MWe lean burn, allocation
exergy
RER 0 kWh 9.78E+0 0 0 0 1 1.10
(2,3,1,2,1,3); on-site plan t of Wacker in
Germany
electricity, hydropower, at run-of-river power plant R ER 0 kWh 1.68E+1 0 0 0 1 1.10
(2,3,1,2,1,3); production of REC and of
Wacker's hydropower pl ant
electricity, medium voltage, at grid NO 0 kWh 3.40E+0 0 0 0 1 1.10
(2,3,1,2,1,3); de Wild-Scholten & Alsem a
2005, Environme ntal Life Cycle Inventory of
Crystalline Sil icon Photovoltaic Module
Production
electricity, medium voltage, at grid CN 0 kWh 0 3.00E+1 0 0 1 1.10
(2,3,1,2,1,3); de Wild-Scholten & Alsem a
2005, Environme ntal Life Cycle Inventory of
Crystalline Sil icon Photovoltaic Module
Production
electricity, medium voltage, at grid US 0 kWh 0 0 3.00E+1 0 1 1.10
(2,3,1,2,1,3); de Wild-Scholten & Alsem a
2005, Environme ntal Life Cycle Inventory of
Crystalline Sil icon Photovoltaic Module
Production
electricity, medium voltage, at grid KR 0 kWh 0 0 0 3.00E+1 1 1.10
(2,3,1,2,1,3); de Wild-Scholten & Alsem a
2005, Environme ntal Life Cycle Inventory of
Crystalline Sil icon Photovoltaic Module
Production
heat, at cogen 1MWe lean burn, alloca tion exergy RER 0 MJ 1.85E+2 1.85E+2 1.85 E+2 1.85E+2 1 1.10 (2,3,1,2,1,3); literature , for proces s heat
silico ne plant RER 1 unit 1.00E-11 1.00E-11 1.00E-11