Environ. Sci. Technol.

2010, 44, 6550–6556

sodium-nickel-chloride (ZEBRA) batteries. New electric cars

Contribution of Li-Ion Batteries to typically use lithium ion (Li-ion) batteries. Major reasons
the Environmental Impact of Electric are the favorable material characteristics of lithium: it is the
lightest of all metals and offers the greatest electrochemical
Vehicles potential, which results in a high power and energy density
(2). Additionally, extensive experiences gained in the Infor-
mation and Communication Technology (ICT) industry have
DOMINIC A. NOTTER,* MARCEL GAUCH,
led to safe, long-lasting, and affordable products. The Li-ion
R O L F W I D M E R , P A T R I C K W Ä G E R ,
battery requires little maintenance, an advantage that most
ANNA STAMP, RAINER ZAH, AND
H A N S - J Ö R G A L T H A U S
other battery chemistries cannot claim. There is no memory
effect, little self-discharge, and no scheduled cycling is
Technology and Society Laboratory, Swiss Federal Laboratories
required to prolong the battery’s life. Li-ion battery chem-
for Materials Science and Technology (Empa),
istries and cell construction are rapidly developing and
Ueberlandstrasse 129, 8600 Duebendorf, Switzerland
changing: For instance, the commonly used, but expensive,
cobalt is being replaced by chemistries using iron phosphate
Received December 9, 2009. Revised manuscript received or manganese (3). Another development is the increase in
June 16, 2010. Accepted June 24, 2010. the content of active material by, for example, using bipolar
electrodes (4).
Commercial Li-ion cells are currently using various types
of cathode materials (5); one of them is lithium manganese
Battery-powered electric cars (BEVs) play a key role in oxide (LiMn2O4). Spinel type LiMn2O4 is attractive for BEVs
future mobility scenarios. However, little is known about the in many aspects, such as low cost, rather easy production
environmental impacts of the production, use and disposal of the process (3) and, last but not least, thermal safety (6). In
lithium ion (Li-ion) battery. This makes it difficult to compare addition, manganese is abundant in nature (7) and well
the environmental impacts of BEVs with those of internal established in the battery industry (8).
combustion engine cars (ICEVs). Consequently, a detailed The question remains as to whether the need for batteries
lifecycle inventory of a Li-ion battery and a rough LCA of BEV is causing an overcompensation of the potential benefits of
based mobility were compiled. The study shows that the the higher efficiency of BEVs compared to ICEVs. The study
environmental burdens of mobility are dominated by the operation presented in this paper aims at establishing a sound basis
for answering this question from an environmental point of
phase regardless of whether a gasoline-fueled ICEV or a
view.
European electricity fueled BEV is used. The share of the total
environmental impact of E-mobility caused by the battery
(measured in Ecoindicator 99 points) is 15%. The impact caused
Methods and Definitions
Life Cycle Assessment (LCA). LCA is an established but still
by the extraction of lithium for the components of the Li-ion evolving method primarily designed for accounting for and
battery is less than 2.3% (Ecoindicator 99 points). The major assessing the potential environmental impacts caused by
contributor to the environmental burden caused by the battery products, processes, or activities (9). LCA aspires to quantify
is the supply of copper and aluminum for the production of and evaluate the energy and material flows used in all stages
the anode and the cathode, plus the required cables or the of a product’s lifetime and the associated wastes and
battery management system. This study provides a sound basis emissions released to the environment. LCA is based on a
for more detailed environmental assessments of battery perspective called the functional unit. This is especially
based E-mobility. important when competing products are analyzed to allow
for comparative assessment.
This study analyzes the use of lithium-ion batteries in
Introduction electric vehicles as an environmentally viable option and
Society’s current individual mobility behavior is creating a evaluates whether the burdens related to the battery are likely
plethora of looming problems, such as fossil carbon intensity to offset the benefits related to the electric drivetrain. For
and the concomitant consequences regarding fossil resource this end it is necessary to model electric mobility (E-mobility)
supply or the emissions of pollutants such as nitrogen oxides including the vehicle’s production, use and disposal with a
(NOx), sulfur dioxide (SO2), and particulate matter. While strong focus on the Li-ion battery. The functional unit is
pollutant problems can partially be solved technically by chosen as one average kilometer driven by a vehicle with
catalytic converters and filters, expectations run high that electric drivetrain and Li-ion batteries on the European road
the greenhouse gas and resource problems can be addressed network. The corresponding reference flow is one vehicle-
by massively substituting internal combustion engine cars kilometer. The study encompasses a cradle-to-grave system
(ICEV) with battery powered electric cars (BEV). without predefined cutoff limits. Materials and processes
The widely used term “electric car” covers many types of are only neglected when their contribution to the potential
currently discussed and tested variations of electrical pro- environmental burdens is considered negligible based on a
pulsion systems for passenger cars, such as battery car, fuel combination of mass, energy demand, and expected burdens
cell car, and serial hybrid (range extender) car. Most of the per mass or energy unit. LCI data for the battery production
major car manufacturers recently announced that battery and for the production and use of an electric vehicle are
cars will soon become part of their product lines (1). compiled specifically for this study (see Figure 1), while LCI
In the past, the BEVs’ on-board energy supply has been data for the materials and processes in the background system
based on lead-acid, on nickel-metal hydride (NiMH), or on are taken from the ecoinvent database version 2.01.
Battery. We chose to model a LiMn2O4 battery since it
* Corresponding author e-mail: dominic.notter@empa.ch; phone: seems reasonable to assume that manganese will in the near
+41 44 823 47 60. future be substituted for the nickel and cobalt commonly
6550 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 17, 2010 10.1021/es903729a  2010 American Chemical Society
Published on Web 08/09/2010
used in many of today’s batteries because of the much lower
price, and better availability of manganese. Calculations were
done on different cathode materials containing nickel, cobalt
or iron-phosphate in order to check the sensitivity of the
results. Details on the production of the battery and its
components are given in the paragraph below headed
Description of Unit Processes.
Electric Vehicle. The vehicle we studied was comparable
to a Volkswagen Golf in size and power, had a range of around
200 km per charge (battery weight, 300 kg; battery capacity,
0.114 kWh/kg battery (10)) and an assumed lifetime of 150 000
km. These assumptions were tested with a sensitivity analysis
for an extension of vehicle life to 240 000 km. In this case the
BEV would require a battery replacement. The energy
consumption of the electric vehicle’s operation was estimated
based on existing vehicles and theoretical considerations (for
details see Supporting Information); 14.1 kWh of electric
energy is needed per 100 km to propel a Golf-class vehicle
with an overall efficiency of 80% (including charging losses
and recuperation gains) in a standard driving cycle (New
European Driving Cycle, NEDC). Heating, cooling, and
electronic devices consume 2.9 kWh/100 km (for details see
Supporting Information). The BEV thus required a total of
17 kWh/100 km. The influence of energy consumption on
the environmental burden was tested by varying energy
demand by (20%.
The average electricity production mix (UCTE) in Europe
(11) was chosen for the operation of the BEVs in agreement
with the criteria used in the rest of the study and in the
ecoinvent database. The environmental burden for the
operation of BEV depended mainly on the choice of electricity
production. This was tested by replacing the UCTE-mix with
electricity from hard coal and hydropower. It is important
to state that neither the vehicle nor the battery was meant FIGURE 1. Model structure of electric mobility. The present study
to represent specific products but rather a technically sensible considers all relevant processes to deliver a journey of 1 km but
option. focuses on the materials and components in the gray boxes.
Reference Vehicle. A new efficient gasoline car (Euro 5
standard (12)) was chosen as a basis for comparison. This demand (CED) of which only the nonrenewable (fossil fuel
ICEV consumes 5.2 L of gasoline per 100 km in the NEDC, and nuclear) are disclosed (17) and the Ecoindicator 99 using
resulting in a direct emission of 0.12 kg CO2 per km (13). The the hierarchic perspective and an average weighting (EI99
car chosen was representative of neither the European fleet H/A) (18). Resource depletion is indicated as abiotic depletion
nor the fleet of new cars sold in Europe in 2009: 51.4% of the potential (ADP), one of the impact categories in the CML
latter consists of diesel fuel cars with an average fuel method (19, 20). All impact assessment methods are used as
consumption of 5.7 L/100 km diesel or 6.6 L/100 km gasoline implemented in the ecoinvent database version 2.01. The
(calculated from CO2 emissions reported in ref 14). However, first two indicators were chosen for their broad acceptance
the ICEV was chosen to represent a technological level similar and relevance in decision making. The EI99 H/A was chosen
to that of the BEV. since the other indicators are almost exclusively energy driven
Allocation and System Boundary. Allocation of the inputs and exclude (toxic) effects on human health and ecosystems,
and output flows to the various products is a critical issue as well as damage to resource quality. ADP was evaluated to
in LCA studies because the choice of allocation principles include the use of resources, especially metals. Apart from
can predetermine the outcome of the LCA. Thus, it is very the impact assessment results, we also present cumulative
important that allocation procedures are in line with the particulate matter (PM10), SO2, and NOx emissions, since
goal of the study (15). Since a present study aimed to these indicators are widely used in discussions on environ-
determine the potential contribution of batteries to the overall mental issues of mobility.
burdens of mobility, allocation in the foreground system was Description of Unit Processes. Figure 1 depicts the
chosen in such a way as to result in the highest possible production steps required for the Li-ion battery ranging from
environmental burdens for the battery. Thus, all expenditures the extraction of lithium and the electrode production to the
for the exploitation of the lithium salts were allocated to the battery pack, the components of the electric vehicle, and the
lithium salts, even though the saline brine yields other mobility with the electric vehicle. The dashed line refers to
byproduct as well. the functional unit chosen for this study. For all productions
In line with this principle, end-of-life (EOL) products that steps, the required thermal and electrical energy to produce a
are being recycled are not allocated any expenditure from 1 kg Li-ion battery is quoted. The mass used for the calculation
their production. Thus, all the burdens from material are based on a Kokam battery (21) and the cathode material
production are allocated to the first life of a product even is assumed to be LiMn2O4. Detailed input-output tables for all
though the product might even be reused (e.g., after 5 years gray boxes and the assumptions for transport distances,
in a vehicle the batteries might be used in stationary infrastructure, and electricity mixes are provided in the Sup-
applications). porting Information.
Environmental Impact Assessment. The environmental The production of concentrated lithium brine includes
burdens are expressed as global warming potential (GWP) inspissations of lithium containing brine by solar energy in
applying a time frame of 100 years (16), the cumulative energy the desert of Atacama. Diesel fuel is required for pumping

VOL. 44, NO. 17, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6551
FIGURE 2. Shares correlating with the components of an internal combustion engine car (ICEV, value in % of the BEV) and an
electric battery powered car (BEV, the BEV is set as 100%) assessed with four impact assessment methods: abiotic depletion
potential (ADP), nonrenewable cumulated energy demand (CED), global warming potential (GWP), and Ecoindicator 99 H/A (EI99 H/A).
Road includes construction, maintenance, and end of life treatment (EOL). The absolute values of the components are provided in the
Supporting Information.

the brine (22) between different basins. The concentrated Cathode, separator, and anode are calendared, slit to size,
lithium brine is further treated with additives for the removal winded, and packed to a single cell in a polyethylene envelope.
of boron, followed by a purification step. Finally, the addition In an inert atmosphere, the electrolyte (LiPF6 dissolved in
of soda for carbonation results in the precipitation of lithium C3H4O3) is added to the electrode (26). Finally, single cells,
carbonate (Li2CO3). The salt is filtered, washed and dried the battery management system and cables are assembled
which results in a purity of 99% or higher (23). in a steel box.
Manganese oxide (Mn2O3) is produced by a two stage The electric car represented in this LCI was derived from
roasting process whereby manganese carbonate is roasted the existing Golf LCI (34). The glider (chassis, car body parts,
in an atmosphere low in oxygen content, followed by roasting wheels, interiors, safety devices, acclimatization devices)
in an atmosphere high in oxygen content (24). Subsequently, remained unaltered, but the drivetrain was replaced by an
lithium manganese oxide (LiMn2O4) is made from Mn2O3 electric drivetrain (composed of the electric power control,
and Li2CO3 by means of several roasting stages in a rotary an electric motor and the transmission) and by a Li-ion battery
kiln (3). During the different stages, the atmosphere in the (see Scheme S1 for Supporting Information). The use of the
rotary kiln changes from an inert (addition of N2) to an car takes into account electricity consumption and all
oxidizing (addition of O2) condition. The powder is then infrastructures needed (vehicle, road and electricity network)
suspended with water followed by spray drying (evaporation including EOL treatment. The data set for a new efficient
of the water). gasoline passenger car with reduced fuel consumption (Euro
Base materials for the electrolyte are an organic solvent, 5 standard) based on the ecoinvent Database was used as a
typically ethylene carbonate (C3H4O3) (25), and the electrolytic reference.
salt, typically lithium hexafluorophosphate (LiPF6) (26). For Emissions and Impacts. The Li-ion battery plays a minor
the production of the LiPF6, lithium fluoride (LiF) is role regarding the environmental burdens of E-mobility
manufactured with a reaction of Li2CO3 and hydrogen fluoride irrespective of the impact assessment method used. Transport
at room temperature. The filtrate is titrated with ammonia services with an ICEV cause higher environmental burdens
(pH 7.5), washed with water, and dried (27). Phosphorus than with a BEV (ADP, + 37.47% or 261 kg antimony
pentachloride (PCl5) (28) and LiF are then combined in an equivalents; GWP, + 55.3% or 37,700 kg CO2 equivalents;
autoclave and cooled down to -78 °C. Thereafter, hydrogen CED, +23.5% or 593,000 MJ-equivalents; EI99 H/A, +61.6%
fluoride is added in excess for complete chlorine-fluorine or 2530 points; Figure 2). The share of the total environmental
exchange in PCl5 (29). The reaction in the autoclave occurs impact of E-mobility caused by the battery is between 7 (CED)
in an inert nitrogen atmosphere. and 15% (EI99 H/A). Analysis with EI99 H/A showed a relative
The production of the cathode and anode requires the share of E-mobility caused by the battery that is twice as
mixture of a few components (binder and solvent, black high as analysis with the other impact assessment methods,
carbon, LiMn2O4 and graphite respectively) in a ball mill to and this is mainly at the expense of the operation phase.
a slurry (26, 30), followed by coating the collector foil There are no differences between ICEV and BEV with
(aluminum and copper respectively) with the slurry. The respect to the environmental burden related to road use
binder (modified styrene butadiene copolymer (31)) is water- (infrastructure, maintenance, and disposal) and the glider.
soluble and has the advantage that no organic solvent is Small differences are related to the drivetrain, maintenance,
needed. For the production of the separator, a porous and disposal of the car. The main difference is reflected in
polyethylene film is coated with a slurry consisting of a the operation phase, which rises far above the impact of the
copolymer, dibutyl phthalate and silica dissolved in acetone battery. Operation obviously dominates the LCA of both
(32). Thermal heat energy for anode, cathode and separator E-mobility and mobility with an ICEV, while it is distinctly
is used to heat up the slurry to 130 °C, to evaporate the solvent higher for mobility with an ICEV.
and to completely dehumidify the components of the PM10-, NOx-, and SO2-emissions caused by E-mobility
electrode in a dry channel (H2O content <20 ppm) (33). (PM10 100%, 16.2 kg; NOx 100%, 49.5 kg; SO2 100%, 83.7 kg)

6552 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 17, 2010
FIGURE 3. Environmental burden of the main components of the Li-ion battery and the electrodes expressed with Ecoindicator 99 H/
A (EI99 H/A), cumulated energy demand (CED), global warming potential (GWP), and abiotic depletion potential (ADP). Components of
the anode (C. Ano), components of the cathode (C. Cat). Absolute values are provided in the Supporting Information.

are higher compared to mobility with an ICEV (PM10 79.0%, The sensitivity of the LiMn2O4 as an active material was
12.8 kg; NOx 87.9%, 43.5 kg; SO2 74.7%, 62.5 kg; Supporting tested (EI99 H/A) by comparing the environmental burden
Information Figure S1 and Table S20). All these emissions of LiMn2O4 compared to the also widely used active
result mainly from operation independently of the vehicle material Li(Mn(1/3)Ni(1/3)Co(1/3))O2 and LiFeO4. While the
type. The production of the battery, the glider, and the scenario with the active material containing nickel and
drivetrain also emits considerable amounts of PM10, NOx, cobalt results in an increase in environmental burden of
and SO2. 12.8% for the battery, LiFeO4 as cathode material decreases
The production of the Li-ion battery is dominated by the the impact for 1.9%. The difference on the level of transport
production of the anode, the cathode and the battery pack service is much smaller (Li(Mn(1/3)Ni(1/3)Co(1/3))O2, +2.0%;
(Figure 3). Single cell, separator, lithium salt, and solvent LiFeO4, -0.30%).
play a minor role. In addition to its cells, the battery pack The results of the sensitivity calculated for a vehicle life
contains a steel box, cables, and the printed wiring board. of 240,000 km (the BEV needs 2 sets of Li-ion battery) shows
These components cause a relatively high share of more than a decrease of the total environmental burden per vehicle-
20% throughout all impact assessment methods. kilometre (EI99 H/A) of 7.5% (BEV) and 7.1% (ICEV).
Concerning EI99 H/A, the production of the anode A variation of electricity consumption of (20% (mean,
generates the highest impact, while CED, GWP, and ADP 0.17 kWh/km; -20%, 0.14 kWh/km; +20%, 0.20 kWh/km)
show the highest impact for the production of the cathode. results in a modification of (8.2% of the environmental
Copper in the anode is needed as collector foil, which has burden (EI99 H/A) for E-mobility.
a share of 43% (EI99 H/A) of the environmental burden of The impact (EI99 H/A) of the transportation of a BEV
the Li-ion battery. Copper used in other components (e.g., using electricity produced with hard coal (UCTE) increases
cables) comes in addition. Graphite and all other components for 13.4%, while it decreases for 40.2% when applying
of the anode only have a small impact. The results for the electricity from hydropower plants. Thus, using hydropower
anode look different when assessed with the ADP, CED,or electricity as fuel for the BEV reduces the share of operation
GWP. The anode has a much smaller share on the total impact on total environmental burden of transport service sub-
of the battery. Within the anode, graphite shows a higher stantially to 9.6%.
environmental burden than copper, at least when assessed
with ADP and CED.
The cathode causes a higher GWP, CED, or ADP than the
Discussion
anode. The collector of the cathode, made of aluminum foil, The main finding of this study is that the impact of a Li-ion
has a higher share of the environmental burdens than the battery used in BEVs for transport service is relatively small.
active material throughout all impact assessment methods. In contrast, it is the operation phase that remains the
All other components (binder, carbon black, energy use, etc.) dominant contributor to the environmental burden caused
cause a very small environmental burden for the production by transport service as long as the electricity for the BEV is
of the cathode. not produced by renewable hydropower. This finding is in
Environmental burdens caused by the two lithium good accordance with other studies showing that the impact
containing components LiMn2O4 and LiPF6 are between 10 of operation dominates in transport service (35, 36). In these
(EI99 H/A) and 20% (GWP), whereas the share of LiMn2O4 studies, infrastructure, maintenance, and service have minor
(EI99 H/A 5.60%; GWP 13.8%) is higher than the share of shares of the environmental impact imposed by transport
LiPF6 (EI99 H/A 3.79%; GWP 6.47%). services. We found the same pattern for the environmental
The printed wiring board, process heat, and nitrogen are burden of the different components to transport service
other important contributors to the total impact of a Li-ion (Figure 2).
battery, besides the copper- and aluminum collector foils Another explanation for the small impact of the battery
and the active materials graphite and LiMn2O4. on the overall assessment of transport service is the tiny
A closer look at the damage categories of the EI99 H/A share of the lithium components on the environmental
indicates that the production of a Li-ion battery predomi- burden for the Li-ion battery. This finding can be explained
nantly causes damage to human health (44%) and resource first of all by the fact that the lithium content accounts for
quality (39%), whereas the quality of ecosystems is affected only 0.007 kg per kg Li-ion battery. Thereby, the lithium
less (17%). Inorganic emissions affecting the respiratory content of the active material (LiMn2O4) and the lithium in
system, such as PM10, SO2, NOx, etc., cause the highest impact, the electrolyte is included. In addition, the processes used
followed by the use of fossil fuels and minerals (for detailed to extract lithium from brines are very simple and have a low
information see Supporting Information Figure S2). energy demand. Although lithium occurs in average con-

VOL. 44, NO. 17, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6553
centrations lower than 0.01% in the Earth’s crust and hence Li-ion battery, are located in industrial areas where the
can be considered to be a geochemically scarce metal (37), population density is rather small. The releases of emis-
assessment with ADP does not result in a high impact for the sions from operation are prevalent in urban areas with a
lithium components. Li2CO3, the base material for the cathode high population density. The NOx-emissions from an ICEV
active material and the lithium salt have an impact of only that originate prevalent from operation, consequently have
1.9%. Compared to other components, for example, Mn2O3 a high damage potential to human health.
(4.4%), copper (5.3%) or aluminum (15.1%), the abiotic The relationship between operation versus battery pro-
depletion of lithium resources does not seem to be critical. duction is different when assessed with EI99 H/A compared
However, these results are valid only as long as Li2CO3 is to the other impact assessment methods, even though the
produced from brines. If the lithium components were based comparisons for transports with BEV or ICEV look very similar
on spodumene, a silicate of lithium and aluminum, the regardless of the method used. EI99 H/A indicates copper as
extraction of the lithium would require a considerable amount being a large contributor to the environmental burden,
of process energy (38). whereas it has a rather small share when assessed with the
The major contributors to the environmental burden for other methods. On the contrary, aluminum and LiMn2O4
the production of the battery, regardless of the impact which contribute considerably when assessed with CED, GWP
assessment method used, are metal supply (Figure 3) and or ADP, only account for a small share when assessed with
process energy. Metals appear above all in the production EI99 H/A. This is suggested by the different information that
of the anode (copper collector foil), the cathode (aluminum can be inferred when using the EI99 H/A method. GWP,
collector foil), and the battery pack. The battery pack requires CED and ADP are driven exclusively by the use of minerals
cables (copper), steel for the box of the battery and a battery and energy, while EI99 H/A also appraises toxicity to humans
management system, which contains different metals, for and ecosystems. The human health damage category within
example, copper, gold, tin. A high energy demand occurs in EI99 H/A accounts for 43% of the complete environmental
the production of aluminum, the production of wafers burden caused by the production of a Li-ion battery (see
for the battery management system, the production of Figure S2 in Supporting Information). When analyzing only
graphite, the roasting processes of manganese carbonate to copper (43% of overall impact of the entire battery produc-
Mn2O3 or Li2CO3 and Mn2O3 to LiMn2O4 or the use of heat tion) with EI99 H/A, the damage to human health (40%) and
to dry the electrodes. ecosystem quality (27%) inflicts a greater environmental
Graphite has a distinctly higher impact regarding CED burden than the extraction of the mineral (30%) including
compared to GWP. Hard coal coke is the base material which energy consumption (3%). This evidence explains the dif-
is transformed into graphite. The material itself contains a ferent pictures produced by EI99 and the other assessment
lot of energy and contributes to the CED, but not to the GWP methods.
as the carbon remains in the product and only a low level Uncertainty and Sensitivity. The inventory data presented
of CO2 emissions are generated in the process. for the Li-ion battery and both the BEV and the ICEV do not
Another remarkable contributor to the environmental rely on performance data representing specific products,
impact of the Li-ion battery is LiMn2O4 which reaches its hence, uncertainties adhere to the LCI. Also the choice of
highest values when assessed with GWP. The high score is allocation procedures and other modeling choices elicit
explained with the energy input for the roasting process of variances that might affect the outcome of the study. The
Mn2O3 and LiMn2O4 and the concomitant high use of the most critical points are therefore discussed in the following
resource. The extraordinarily high value in terms of GWP section.
originates from the fact that the conversion of manganese The chemistry chosen for the Li-ion batteries investigated
carbonate to Mn2O3 and further the reaction of Li2CO3 and in this study was based on manganese. Today, numerous
Mn2O3 to LiMn2O4 releases considerable amounts of CO2. other materials are serious contenders for automotive
The LCA result of BEVmobility mainly depends on the batteries, for example, nickel, cobalt, or iron. The sensitivity
environmental profile of the electricity mixes considered, analysis of different lithium-based cathode materials showed
as the vehicle tailpipe emissions are shifted to the power only small changes in the environmental burden. Hence, for
generation units (36). E-mobility causes the highest a generic assessment it seems reasonable to neglect the
possible EI99 H/A, CED, GWP ADP results whenever an diversity of many different active materials to reduce the
electricity mix is used that contains a high share of fossil complexity of battery chemistry.
fuel such as the UCTE electricity mix (share of fossil fuel The sensitivity analysis on electricity consumption for
>50% (11)). Nevertheless, the operation of an ICEV alone the BEV or the sensitivity analysis for a modified lifespan
causes impacts that are roughly just as high (CED, 92%; showed rather small variances concerning environmental
GWP, 125%; Figure 2) as the total environmental impacts burdens for both, mobility with an ICEV or BEV. However,
of E-mobility (100%). A break even analysis shows that an the choice of the electricity generation led to considerable
ICEV would need to consume less than 3.9 L/100km to variations in the results. Propelling a BEV with electricity
cause lower CED than a BEV or less than 2.6 L/100km to from an average hard coal power plant increases the
cause a lower EI99 H/A score. Consumptions in this range environmental burden by 13.4%. On the other hand, using
are achieved by some small and very efficient diesel ICEVs, electricity from an average hydropower plant decreases
for example, from Ford and Volkswagen (13, 39). environmental burden by 40.2%. This results in a decrease
Transport service affects the environment largely by for the operation from 41.8% (UCTE mix) to 9.6% when
contributing to global air pollution. PM10, SO2, and NOx charging the battery with electricity from hydropower
traffic emissions contribute significantly to environmental plants.
problems such as acidification and eutrophication (SO2 The modeling applied to EOL treatment for the vehicles
and NOx), photochemical air pollution (NOx) or have including the Li-ion battery resulted in a worst case
adverse effects on human health, for example, cell toxicity, scenario, as no benefits were derived from the potentially
damage to genetic material by means of oxidative stress useful materials in the battery. Batteries are recycled at a
or by triggering allergies (PM10, SO2, and NOx). With respect very high rate, since recycling and recovery rates prescribed
to the LCI results for the pollutants PM10, SO2, and NOx, by the EU legislation are 85% for 2006 and 95% for 2015
transport with a BEV leads to higher environmental burden (40). For a conventional car, the EI99 H/A scores would
than transport with an ICEV. However, the emissions be reduced to 88.8% if the modeling approach included
caused by the production of the vehicle, in particular the the benefits of recycled material being substituted for other

6554 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 17, 2010
(virgin) materials (41). This reduction is expected to be (14) EU Measures for monitoring the average specific emissions of
even higher for electric vehicles since the exergy analysis carbon dioxide from new passenger cars. http://ec.europa.eu/
by Dewulf et al. (42) shows that switching from virgin environment/air/transport/co2/co2_monitoring.htm (Accessed
June 14, 2010).
resource supply to recycling for Li-ion battery cathode (15) Werner, F.; Althaus, H. J.; Richter, K.; Scholz, R. W. Post-consumer
material results in a 51% natural resource savings. Thus, waste wood in attributive product LCAsContext specific evalu-
the EOL modelling approach applied in this study un- ation of allocation procedures in a functionalistic conception
derlines the ecological advantage of E-mobility over of LCA. Int. J. Life Cycle Assess. 2007, 12 (3), 160–172.
mobility with an ICEV. (16) IPCC Climate Change 2007: Synthesis Report; IPCC, Intergov-
ernmental Panel on Climate Change, Cambridge University
All the facts taken together, the results of the LCA, the
Press, The Edinburgh Building Shaftesbury Road: Cambridge,
various sensitivity analyses, the modeling applied for EOL, U.K., 2007.
the assumption for the used electricity mix, etc., suggest (17) Hischier, R.; Weidema, B.; Althaus, H.-J.; Bauer, C.; Doka, G.;
that E-mobility is environmentally beneficial compared Dones, R.; Frischknecht, R.; Hellweg, S.; Humbert, S.; Jung-
to conventional mobility. The Li-ion battery plays a minor bluth, N.; Köllner, T.; Loerincik, Y.; Margni, M.; Nemecek, T.
role in the assessment of the environmental burden of Implementation of Life Cycle Impact Assessment Methods;
Swiss Centre for Life Cycle Inventories: Dübendorf, Switzerland,
E-Mobility. Thus, a Li-ion battery in an BEV does not lead
2009.
to an overcompensation of the potential benefits of the (18) Goedkoop, M.; Spriensma, R., The Eco-indicator 99, A Damage
higher efficiency of BEV compared to an ICEV. Oriented Method for Life Cycle Impact Assessment; Methodol-
ogy Report. 2nd ed.; Pré Consultants, B. V.: Amersfoort,
Netherlands, 2000.
Acknowledgments (19) Guinée, J. B., (final editor); Gorrée, M.; Heijungs, R.; Huppes,
We thank Dr. Margarita Osses for helpful translations and G.; Kleijn, R.; de Koning, A.; van Oers, L.; Wegener Sleeswijk, A.;
Thomas Ruddy for proofreading. We appreciate the Suh, S.; Udo de Haes, H. A.; de Bruijn, H.; van Duin, R.;
financial support for this research, which was provided by Huijbregts, M. A. J.; Lindeijer, E.; Roorda, A. A. H.; Weidema,
Axpo Holding AG. We thank Daniela Desormeaux and B. P. Life Cycle Assessment: An Operational Guide to the ISO
Daniel Pizarro from SQM Chile for providing Supporting Standards, Part 3; Scientific Background; Ministry of Housing,
Spatial Planning and Environment (VROM) and Centre of
Information about the process of Li2CO3 production. Environmental Science (CML): Den Haag and Leiden, The
Netherlands, 2001.
Supporting Information Available (20) Guinée, J. B., (final editor); Gorrée, M.; Heijungs, R.; Huppes,
Detailed life cycle inventory data for the drive train and G.; Kleijn, R.; de Koning, A.; van Oers, L.; Wegener Sleeswijk, A.;
Suh, S.; Udo de Haes, H. A.; de Bruijn, H.; van Duin, R.;
the glider of the ICEV and the BEV, input-output tables Huijbregts, M. A. J.; Lindeijer, E.; Roorda, A. A. H.; Weidema,
for the Li-ion battery, and absolute values to environmental B. P. Life Cycle Assessment: An Operational Guide to the ISO
burden for E-mobility and conventional mobility. This Standards, Parts 1 and 2; Ministry of Housing, Spatial Planning
information is available free of charge via the Internet at and Environment (VROM) and Centre of Environmental Science
http://pubs.acs.org/. (CML): Den Haag and Leiden, The Netherlands, 2001.
(21) Kokam Co., L. Cell specification data. http://www.superlatticeinc.
com/downloads/SLPB100216216H.pdf (Accessed Sept 23, 2009).
Literature Cited (22) SEIA-CONAMA EIA - Cambios y Mejoras de la Operación Minera
(1) Taylor, E. Start-Ups Race to Produce ‘Green’ Cars. The Wall en el Salar de Atacama; San Pedro de Atacama, 2006.
Street Journal, May 6, 2008, p B1;http://online.wsj.com/article/ (23) SEIA-CONAMA Ampliación Planta Carbonato de Litio a 48.000
SB121002128666768637.html. ton/año; Antofagasta, 2007.
(2) Armand, M.; Tarascon, J. M. Building Better Batteries. Nature (24) Kajiya, Y.; Tasaki, H. Manganese oxide producing methods.
2008, 451 (7179), 652–657. European Patent Application No. EP1493715, published on 1/05/
(3) Heil, G.; Kormann, C.; Adel, J. Lithium Oxide Containing Lithium 2005.
Intercalation Compounds. European Patent Application No. (25) Birnbach, S.; Dockner, T.; Mohr, J.; Benfer, R.; Bieg, W.; Peters,
EP1204601, published on 2/19/2003. J.; Ruge, B.; Weinle, W.; Zehner, P. Method for the continuous
(4) Marsh, R. A.; Russell, P. G.; Reddy, T. B. Bipolar Lithium-Ion production of 1,3-dioxolan-ones. European Patent Application
Battery Development. J. Power Sources , 65 (1-2), 133–141. No. EP1076654, published on 6/25/2003.
(5) Dinger, A.; Ripley, M.; Mosquet, X.; Rabl, M.; Rizoulis, D.; Russo, (26) Brodd, R. J.; Tagawa, K. Advances in Lithium-Ion Batteries, 1st
M.; Sticher, G. Batteries for Electric Cars: Challenges, Op- ed.; Kluwer Academic/Plenum Pub.: New York, 2002.
portunities, and the Outlook to 2020; The Boston Consulting (27) Friedrich, H.; Pfeffinger, J.; Leutner, B. Verfahren zur Herstellung
Group: Boston, 2010. von hochreinen Lithiumsalzen. Patent DE19809420A1, 1999.
(6) Kitoh, K.; Nemoto, H. 100 Wh Large size Li-ion batteries and (28) Munster, G.; Riess, G.; Russow, J. Process for the manufacture
safety tests. J. Power Sources 1999, 81-82, 887–890. of phosphorus pentachloride. United States Patent Application
(7) Wellbeloved, D. B.; Craven, P. M.; Waudby, J. W., Manganese No. 06/167660, published on 5/5/1981.
and Manganese Alloys, 5th ed.; Wiley & Sons: London, 1997; (29) Belt, H.-J.; Rudolph, W.; Seffer, D. Process for preparing
2nd release 2005, 7th ed. lithium hexafluorophosphate. European Patent Application
(8) Thackeray, M. M.; Johnson, C. S.; Vaughey, J. T.; Li, N.; Hackney, No. EP0816289, published on 1/26/2000.
S. A. Advances in Manganese-Oxide “Composite” Electrodes (30) Na, S.-h.; Kim, H.-s.; Moon, S.-i.; Doh, C.-h. Method for
for Lithium-Ion Batteries. J. Mater. Chem. 2005, 15 (23), 2257– manufacturing high power electrode for lithium secondary
2267. battery. United States Patent Application No. 11/132185,
(9) ISO 14040: Environmental Management-Life Cycle Assessment- published on 12/8/2005.
Requirements and Guidelines; 14044; International Standard (31) LICO Technology Corporation Aqueous binder LHB-108P for
Organisation: Geneva, 2006; p 54. lithium ion electrodes. http://www.predmaterials.com/en_batt/
(10) Protoscar Lampo (pure battery EV): Specifications. http://www. AqueousBinder.html (Accessed June 5, 2009).
protoscar.com/pdf/LAMPO2/LAMPO2_Technical_Specifications. (32) Noh, H.-g. Method of manufacturing lithium secondary cell.
pdf (accessed March 13, 2010). United States Patent Application No. 6610109, published on
(11) Frischknecht, R. T. M., Faist Emmenegger, M., Bauer, C. Dones, 8/26/2003.
R Strommix und Stromnetz. In Sachbilanzen von Energiesys- (33) Naarmann, H.; Krueger, F. J.; Theuerkauf, S. Battery electrodes
temen: Grundlagen für den ökologischen Vergleich von Ener- with enlarged surfaces and mehtod for production thereof.
giesystemen und den Einbezug von Energiesystemen in Ökobi- International Patent Application No. PCT/EP2003/009119,
lanzen für die Schweiz; Swiss Centre for Life Cycle Inventories: published on 4/3/2004.
Dübendorf, Switzerland, 2007. (34) Schweimer, G. W.; Levin, M. Life Cycle Inventory for the Golf A4;
(12) Jungbluth, N. Life Cycle Inventories of Bioenergy; Swiss Centre Research, Environment and Transport Volkswagen AG, Wolfs-
for Life Cycle Inventories: Dübendorf, Switzerland, 2007. burg and Center of Environmental Systems Research, University
(13) Volkswagen Der Golf mit BlueMotion Technology. http:// of Kassel: Wolfsburg, Kassel, Germany, 2000.
www.volkswagen.ch/ch/de/Volkswagen/Innovation/Bluemotion/ (35) MacLean, H. L.; Lave, L. B. Life Cycle Assessment of Automobile/
bmt_modelle/golf.html (Accessed March 9, 2010). Fuel Options. Environ. Sci. Technol. 2003, 37 (23), 5445–5452.

VOL. 44, NO. 17, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6555
(36) Samaras, C.; Meisterling, K. Life Cycle Assessment of Green- (40) EU End of life vehicles. http://ec.europa.eu/environment/waste/
house Gas Emissions from Plug-In Hybrid Vehicles: Implica- elv_index.htm (Accessed June 14, 2010).
tions for Policy. Environ. Sci. Technol. 2008, 42 (9), 3170– (41) Spielmann, M.; Althaus, H.-J. Can a Prolonged Use of a Passenger
3176. Car Reduce Environmental Burdens? Life Cycle Analysis of Swiss
(37) Skinner, B. J. 2nd Iron-Age Ahead. Am. Sci. 1976, 64 (3), 258– Passenger Cars. J. Cleaner Prod. 2007, 15 (11-12), 1122–1134.
269. (42) Dewulf, J.; Van der Vorst, G.; Denturck, K.; Van Langenhove,
(38) Ebensperger, A.; Maxwell, P.; Moscoso, C. The Lithium Industry: H.; Ghyoot, W.; Tytgat, J.; Vandeputte, K. Recycling recharge-
Its Recent Evolution and Future Prospects. Resour. Policy 2005, able lithium ion batteries: Critical analysis of natural resource
30 (3), 218–231. savings. Resour., Conserv. Recycl. 2010, 54 (4), 229-234, DOI:
(39) Ford Ford Fiesta ECOnetic. http://www.ford.de/Pkw-Modelle/ 10.1016/j.resconrec.2009.08.004.
AlternativeAntriebe/FordECOnetic/tabid)tab1 (Accessed March
11, 2010). ES903729A

6556 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 17, 2010