Comparison of the Environmental impact of 5 Electric Vehicle Battery
technologies using LCA
1
1
1
1
1
Julien Matheys , Jean-MarcTimmermans , W out Van Autenboer, Joeri Van Mierlo , Gaston Maggetto ,
2
2
2
3
Sandrine Meyer, Arnaud De Groof, W alterHecq , PeterVan den Bossche
1
Vrij
e Universiteit Brussel, Department ofElectrical Engineering and Energy Technology, Belgium
2
Université Libre de Bruxelles, Centre d’
Etudes Economiques et Sociales de l’
Environnement, Belgium
3
Erasmus Hogeschool Brussel, Departement Industriële W etenschappen en Technologie, Belgium
Abstract
The environmental assessment of various electric vehicle battery technologies (Lead-acid, NickelCadmium, Nickel-metal hydride, Sodium nickel-chloride, Lithium-ion)was performed in the context ofthe
European end-of-life vehicles directive (2000/53/EC).An environmental single-score based on a life-cycle
approach, was allocated to each ofthe studied battery technologies through the combined use ofthe
SimaPro® software and ofthe life cycle impact assessment (LCIA)method Eco-indicator99.The allocation
ofa single-score enables determining which battery technology is to be used preferably in electric vehicles
and to indicate how to further improve the overall environmental friendliness ofelectric vehicles in the
future.
Keywords
Electricvehicles, Battery technologies, environment, life-cycle analysis
(LCA) studies, the main difficulty encountered while
performing this study was the gathering of appropriate,
comparable and accurate data.
First ofall, a description ofthe LCA methodology used to
compare the different technologies is given.In a second
stage, a score is assigned to the different life phases of
the batteries.The final step is the compilation ofthese
results to obtain an overall environmental score foreach
battery type. The attribution of these scores is only
possible after normalisation and weighting of the
intermediate results. The overall scores of the different
batteries have been calculated, and the different battery
technologies can be ranked according to their
environmental performances.Finally, a sensitivity analysis
has been performed to demonstrate the robustness ofthe
results.
1 INTRODUCTION
The large-scale implementation of battery and hybrid
electric vehicles as substitutes for internal combustion
engine (ICE) vehicles can form part of the solution to
modern society’
s challenges such as urban airpollution,
fossil fuel depletion and global warming [1]. This
statement is valid independently of the electricity
production mix.Evidently when using renewable energy
sources, the beneficial effects of applying these
technologies are enhanced [2]. W hen it comes to the
environmental evaluation ofelectric vehicles, the battery
is often considered as the maj
orenvironmental concern.
W hatever the environmental impact ofthe battery might
be compared to the complete impact of the electric
vehicle, the environmental impact ofthe different battery
technologies should be assessed in order to determine
which technology should be preferred.This assessment
was performed through the SUBAT-proj
ect.The aim is to
evaluate the opportunity to keep nickel-cadmium traction
batteries for electric vehicles on the exemption list of
Directive 2000/53 on End-of-Life Vehicles. The proj
ect
delivers a complete assessment ofcommercially available
and forthcoming battery technologies for battery-electric
and hybrid electric vehicles (lead-acid, nickel-cadmium,
nickel-metal
hydride,
lithium-ion,
sodium-nickel
chloride...). The assessment is based on technical,
economical and environmental evaluations ofthe different
battery technologies.In this paper, only the environmental
analysis is presented. The impacts of the different
technologies should be analysed individually to allow
comparison and definition of the most environmentally
friendly battery technology forelectricvehicles.
A life cycle approach is a must when comparing the
environmental impact ofthe different battery technologies,
because the main environmental burden can be located in
different life stages fordifferent products.
The analysis started with a listing of the available
technologies for battery and hybrid electric vehicle
application.Afterwards, a model for the different battery
types has been developed and introduced in the
Simapro® software tool.This model allows an individual
comparison of the different phases of the life cycle of
traction batteries.As often during life cycle assessment
2 METHODOLOGY
LCA allows the practitioner to study the environmental
aspects and the potential impacts ofa product throughout
its life from raw material acquisition through production,
use and final disposal [3].
LCA is one of the most efficient tools to compare the
complete environmental burden ofdifferent products.This
is due to the fact that different products may present
environmental burdens in different parts oftheirlife cycle.
For example, one product may use less resources
compared to another product during the use phase, but
this may be at the cost of more resources used in its
production phase [4].
The life cycle assessment of a product will never be
completely exhaustive; as a consequence, the LCA
practitionercan choose to which degree ofdetail to model
the assessed life cycle. However, this choice clearly
influences the degree ofprecision and correctness ofa
study.
This study has been performed according to the fourISO
standards specifically designed forLCA applications (ISO
14040-14043).
2.1 Assumptions
To start with, an appropriate functional unit has to be
defined.As the different batteries have various life times
and consequently provide varying performances as
97
regards the number of charge-discharge cycles, the total
lifetime of the battery is not an appropriate functional unit.
Different functional units were evaluated, but in the end, it
was decided to choose a functional unit corresponding to
a battery enabling the car to cover a specific range, with
one charge. This “one-charge range” was set to 60 km
when driving up to 80% depth-of-discharge of the battery.
Furthermore, the environmental comparison will be based
on the impacts of a battery pack (or of the different battery
packs) enabling the vehicle to cover a 180.000 km lifetime
range, corresponding to 3.000 (60km) charge-discharge
cycles (80% depth-of-discharge). Depending on the
technology, the required number of batteries needed for
the functional unit has been determined. The considered
battery is applied to a car with a net weight of 888kg
(excluding the battery, including the 75kg driver). The
system boundaries had to be defined. The assessed time
period corresponds to the current state of the technology.
The related other life cycles (industrial buildings, trucks,
electric power plants, roads, etc.) have not been taken
into account. Self-discharge was not considered for any of
the assessed technologies because of the great
dependence of this parameter on the way the vehicle is
used. Neither was the maintenance of the batteries
because of the opinion this impact is comparatively small.
The electricity consumption has been considered using
the European (EU-25, 2003) electricity production mix as
a reference. It has been considered that the recycled
materials have the same quality as the original materials.
A collection rate of 100% was assumed and a recycling
rate of 95% was used regarding the recuperated materials
(except for the lead-acid recycling technology, which is
much older and which is very mature, where the lead
metal recycling rate is 98.3% ). It was assumed that the
electrolyte is neutralized before disposal (except for the
lead-acid technology where 90% is recuperated and 10%
is neutralized before disposal).
•
Extraction of raw materials,
•
Processing of materials and components,
•
Use phase of the battery,
•
Recycling of discarded batteries,
•
Final disposal or incineration
Figure 1 shows a schematized overview of the life cycle of
an electric vehicle battery.
2.2 Impact assessment
Information available in the literature, information obtained
by intensively interrogating the worldwide industry and
information obtained through commercially available
databases allowed performing the inventory analysis.
Starting from the data obtained from these sources, a
process tree of each stage of the life of the functional unit
was drawn and mass balances linked these subsystems
to each other. As an illustration, a process tree for the
lead-acid battery is presented in Figure 2.
The analysis of the use phase of the batteries can be
divided in 3 main parts. First of all, the use phase was
studied for an “ideal” battery (mass = 0 kg, energy
efficiency of the battery = 100% ). In other words, this is
the energy used to move the car (excluding the battery).
Secondly, the influences of the varying masses and
energy efficiencies of the different battery technologies
have been taken into account. The energy consumption of
the car varies slightly, depending on the mass of the
battery. These differences in energy consumption have
been simulated and calculated using the Vehicle
Simulation Program (VSP) [5].
The life cycle impact assessment methodologies LCApractitioners have at their disposal often differ and the
choice of the method to be used remains an important
decision [6].
In this study, the chosen LCIA method is Eco-indicator 99
(hierarchist perspective) [7,8,9].
The purpose of this study is to determine which type of
traction battery for electric vehicle applications is the most
environmentally friendly. This analysis is performed
considering the complete battery life cycle. Taking the
important number of calculations needed to perform an
LCA into account, the use of software is unavoidable.
Figure 1:The schematized life cycle of a battery.
The interaction of the functional unit with nature is
assessed considering the following life stages of the
battery:
98
P ROCEEDINGS OF LCE2006
1p
LC Lead-Acid F.U
Range (M1/R1+ R2)
1.08E3
6p
Assembly Lead-acid
F.U. Range (M1)
1.09E3
81.4
65.6 p
Assembly monblock
lead-acid
(opgesplitst)
1.09E3
42 p
Separator (/kg)
139 p
Case (/kg)
16.5
1.21E4 MJ
European mix
(2002)
81.4
1.32E3 p
Electrodes (/kg)
79.3
1.21E4 MJ
Electricity (due to
mass)
834
2.08E4 MJ
Electricity (due to
battery ef f iciency )
140
581
2.08E4 MJ
European mix
(2002)
2.07E3 kg
Disposal Scenario
-809
8.61E4 MJ
European mix
(2002)
140
563 p
Electroly te (/kg)
4.93
8.61E4 MJ
Electricity (f or ideal
battery )
581
2.07E3 kg
Recy cling scenarios
(R1 and R2)
-809
2.07E3 p
Energy Mix
Prodcution
156
1.03E3 kg
Recy cling scenario
(R2)
1.03E3 kg
Recy cling scenario
(R1)
-406
-403
Figure 2: Illustration of a lead-acid process tree with SimaPro®.
On the basis of technical and commercial interest, a
number of battery technologies have been selected and
have been analysed quantitatively: lead-acid, nickelcadmium, nickel-metal hydride, sodium-nickel chloride
and lithium-ion. Some of the data, required to compare
the battery technologies, were extracted from the
technical part of the Subat report (www.batteryelectric.com) and are presented in Table 1.
Specific
Energy
(Wh/kg)
Pb-acid
NiCd
NiMH
Li-ion
NaNiCl
40
60
70
125
125
Number of
cycles of 1
battery
pack
500
1350
1350
1000
1000
Energy
efficiency
82.5%
72.5%
70.0%
90.0%
92.5%
Losses
due to
heating
When analysing the other part of the environmental
impact of the battery (excluding the use phase
completely), it appears that the lead-acid battery has got
the highest impact, followed by nickel-cadmium, lithiumion, nickel-metal hydride and sodium-nickel chloride.
Additionally, the recycling phase allows compensation of a
significant part of the environmental impacts of the
production phase.
500.00
140
243
400.00
271
7.2%
81.4
300.00
59.7
Table 1: Battery properties
The results obtained by using SimaPro® and ecoindicator 99, are expressed in eco-indicator points. One
eco-indicator point being equivalent to one thousandth of
the yearly environmental impact of one average European
inhabitant. Next to this, to allow an easy comparison of
the environmental rating of the different battery
technologies, the results were all compared to the
environmental impact of the lead-acid battery, which was
taken as a reference.
3
Additional Energy consumption
due to battery efficiency
Additional Energy due to battery
mass
Assembly + Recycling
Eco-indicator Points
RESULTS
3.1 Environmental impact assessment
When considering the life cycle of the batteries, a
significant impact on the environment is induced by the
energy losses in the battery and the energy losses due to
the additional mass of the battery (Table 2 and Figure 3).
However, this impact is strongly dependent on the way
electricity is produced. In the present calculations the
European electricity production mix has been used, but
this impact would be strongly decreased if renewable
energy sources were used more intensively.
66.9
200.00
52.4
21.7
168.16
189.40
281.97
22.8
240.82
100.00
99.5
111.83
0.00
Pb-Ac
NiCd
NiMH
Li-ion
NaNiCl
Figure 3: Environmental impact (in Eco-indicator points) of
the assessed technologies, including the losses due to
battery masses and to battery efficiencies during use.
When including the effects of the losses due to the battery
(battery efficiency and battery mass), three technologies
have a somewhat higher environmental impact compared
to the other two. Inclusion of battery efficiency (table 1)
results in a higher environmental impact for nickelcadmium and nickel-metal hydride batteries and in a lower
one for lithium-ion batteries as compared to the others.
13th CIRP I NTERNATIONAL C ONFERENCE ON L IFE C YCLE E NGINEERING
99
Production
Pb-acid
1091
Additional
Energy
221
Recycling
Total
-809
503
NiCd
861
303
-620
544
NiMH
945
323
-777
491
Li-ion
361
89
-172
278
NaNiCl
368
122
-256
234
Table 2: Environmental scores (Eco-indicator points) of
the life stages of the assessed battery technologies.
3.2
Sensitivity analysis
A sensitivity analysis has been performed as the results
need to be reliable. This analysis consisted in the
modification of the assumptions made during the
development of the model. The consequences of these
modifications on the results were analysed.
The sensitivity analysis assessed the effects of the
assumptions (concerning average battery composition,
energy consumption, etc.) and of possible variations in the
collected data on the results. This analysis was performed
by varying the assumed parameters. The implemented
variations included calculations using different relative
sizes of the components of the battery (10% more weight
of one component, compensated by an equivalent
decrease of another component). This resulted in the
alteration of the proportional masses of the electrodes,
electrolytes and cases. Also, the recycling efficiencies and
rates were modified as well as the required energy to
produce and recycle the different types of batteries. The
calculations of the environmental impacts were performed
over again for each of the alterations included in the
functional unit data.
Figure 4 summarizes the relative environmental scores as
well as the results of the sensitivity analysis. Figure 4 only
includes the results originating from production, recycling
and the energy losses due to the battery mass and to the
battery efficiency, but not the energy use in the hypothesis
of an ideal battery, as this parameter doesn’t vary from
one battery technology to another. Moreover; this energy
use is related to the use of the vehicle and not to the
battery itself. The bars in Figure 4 represent the relative
environmental impacts of every battery type, considering
the lead-acid as a reference. The overall environmental
score of the lead-acid battery has been set to 100. The
error bars represent the intervals containing all the results
obtained during the sensitivity analysis.
Figure 4 shows that the assumptions didn’t have any
significant impact on the results as the conclusions remain
100
the same within the error bars. This reveals the results of
this study are reliable and demonstrates the robustness of
the model. It should be noted that data concerning the
environmental impacts of the electrolytes of the lithium-ion
batteries were very difficult to obtain, as these electrolytes
are very specific and new. As a consequence, these
impacts have not been taken into account in this study.
Also, no realistic data were obtained concerning the
energy consumption during the manufacturing of the
sodium-nickel chloride batteries. As a consequence, the
environmental impacts of both the lithium-ion and the
sodium-nickel chloride batteries could be slightly higher
than the results shown in Figure 4.
Environmental
impact
115
95
75
55
100.0
108.0
97.7
35
55.2
46.5
15
-5
Lead-acid
Nickelcadmium
Nickel-met
al
hydride
Lit
hium-ion Sodium-nickel
chloride
Figure 4: Graphical overview of the relative environmental
scores (including the sensitivity analysis). The lead-acid
technology has been taken as a comparison basis and set
to 100.
Finally, other “one-charge ranges” have been assumed
(50 or 70 km instead of 60 km). The results of the
changes in the “one-charge range” are discussed
separately from the other results of the sensitivity
analysis, as they imply the establishment of new and
different functional units, which are consequently not to be
directly compared.
The environmental impacts of the batteries presenting 50
and 70 km “one-charge ranges”, are shown in figure 5.
These results are based on the same reference as figure
4 (lead-acid with a 60 km range = 100). It appears that the
absolute environmental impacts are different from the
ones obtained using the 60 km range. But the main
trends, and thus the conclusions, stay the same within
each of the assessed “same-range batteries”.
P ROCEEDINGS OF LCE2006
Environmental Impact
50km
120
60km
100
70km
80
130.4
123.6
122.8
60
108
100
40
97.7
86.7
79.1
81.5
65.5
55.8
55.2
44.9
20
46.5
37.8
0
Lead-acid
Nickel-cadmium
Nickel-metal hydride
Lithium-ion
Sodium-nickel
chloride
Figure 5: Graphical comparison of the environmental impacts of the different battery types when assuming three
different ranges.The 60km range lead-acid battery has been kept as a comparison basis and set to 100.
4 CONCLUSIONS
An essential conclusion is that the impacts of the
assembly and production phases are compensated to a
important extent when collection and recycling of the
batteries is efficient and performed on a large scale.
Excluding the energy losses occurring during the use
phase (due to the battery efficiencies and the additional
masses of the batteries), results in the following
environmental
ranking
(decreasing
environmental
impact): lead-acid, nickel-cadmium, lithium-ion, nickelmetal hydride, sodium-nickel chloride.
Looking at the global results, the following environmental
ranking is obtained (decreasing environmental impact):
nickel-cadmium, lead-acid, nickel-metal hydride, lithiumion and sodium-nickel chloride. Globally three battery
technologies (lead-acid, nickel-cadmium and nickel-metal
hydride) have very comparable environmental impacts. It
can consequently be stated that, taking the sensitivity
analysis into account, these technologies have a higher
environmental impact than the lithium-ion and the sodiumnickel chloride batteries.
When the calculations are performed with batteries
having different energy storage capacities (batteries
allowing the coverage of different ranges with a single
charge), the main conclusions remain the same. In other
words, three of the assessed technologies (lead-acid,
nickel-cadmium and nickel-metal hydride) have a similar
environmental burden, which is higher than the burdens
of the other two technologies, lithium-ion and sodiumnickel chloride. However, these results could be mitigated
due to the great rareness of environmental data
concerning some aspects of the lithium-ion and the
sodium-nickel chloride batteries (for example concerning
the electrolyte).
When analyzing the results of this study, it should be kept
in mind that the environmental impacts of the batteries of
electric vehicles are small compared to the impact of the
rest of the vehicle as well as compared to its use when
using the current electricity production mix (whatever the
used battery technology might be). Additionally it should
be remembered that the latter impacts are also small
when comparing them to the environmental burden
caused by vehicles equipped with internal combustion
engines [1]. Therefore the results of this study should be
seen as an indication on how to even enhance the
environmental friendliness of electric vehicles. When
compiling these conclusions with the technical and
economic conclusions provided by the Subat project the
following conclusion is drawn.
The will to improve the environmental friendliness of
transportation
(by
improving
the
environmental
friendliness of batteries for electric vehicles) should not
discourage the electric vehicle manufacturers. Some time
should be provided to the vehicle manufacturers to adapt
their production modes and to integrate some more
environmentally sound battery technologies in their
vehicles. During the discussions the consortium had with
various stakeholders, it appeared this cannot be
performed within 4 years from now (2006).
ACKNOW LEDGMENTS
This research has been made possible thanks to the
support and funding of the European Commission through
the 6th Framework Program (Action 8.1.B.1.6; EU-project
n°502490).
REFERENCES
[1] Maggetto, G. & Van Mierlo (2001). Electric Vehicles,
Hybrid Electric Vehicles and Fuel Cell Electric
Vehicles: State of the Art and Perspectives. Annales
de Chimie - Science des matériaux, 26 (4), 9-26.
[2] Van Mierlo, J., Timmermans, J.-M., Maggetto, G.,
Van den Bossche, P., Meyer S., Hecq, W.,
Govaerts, L. & Verlaak, J. (2004). Environmental
rating of vehicles with different alternative fuels and
13th CIRP I NTERNATIONAL C ONFERENCE ON L IFE C YCLE E NGINEERING
101
[3]
[4]
[5]
[6]
102
drive trains: a comparison of two approaches.
Transportation Research Part D 9, 387-399.
ISO 14040 (1997). Environmental management –
Life cycle assessment – Principles and framework.
International Standard ISO 14040.
Finnveden G. (2000). On the limitations of Life Cycle
Assessment and Environmental Systems Analysis
Tools in General. International Journal of LCA 5,
229-238.
Van Mierlo, J. & Maggetto, G. (2004). Innovative
Iteration Algorithm for a Vehicle
Simulation
Program.
IEEE
Transactions
on
vehicular
technology 53 (2), 401-412.
Dreyer L.C., Niemann A.L. & Hauschild M.Z. (2003).
Comparison of Three Different LCIA Methods:
EDIP97,
CML2001
and
Eco-indicator
99.
International Journal of LCA 8 (4), 191-200.
[7]
[8]
[9]
VROM (Ministerie van Volkshuisvesting, Ruimtelijke
Ordening en Milieubeheer) (1999). The Ecoindicator 99: A damage oriented method for Life
Cycle Impact Assessment. Methodology Report,
Publicatiereeks productenbeleid, nr. 1999/36A. The
Netherlands.
SimaPro (2004). Introduction to LCA with SimaPro,
PRé Consult, Amersfoort, The Netherlands, pp. 71.
SimaPro (2005). Online 12th January 2005: [
http://www.pre.nl/simapro/default.htm ]
CONTACT
Julien Matheys
Vrije Universiteit Brussel, Department of Electrotechnical
Engineering and Energy Technology, Pleinlaan 2, B-1050
Brussels, Belgium, jmatheys@ vub.ac.be
P ROCEEDINGS OF LCE2006