Journal of Energy Storage 18 (2018) 175–184

Contents lists available at ScienceDirect

Journal of Energy Storage

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

Assessing the potential of an electric vehicle hybrid battery system T

comprising solid-state lithium metal polymer high energy and lithium-ion
high power batteries
⁎
Raphael Wegmanna, , Volker Dögea, Dirk Uwe Sauerb,c,d
a
Corporate Sector Research and Advance Engineering, Robert Bosch GmbH, 71272 Renningen, Germany
b
Electrochemical Energy Conversion and Storage Systems Group, Institute for Power Electronics and Electrical Drives (ISEA), RWTH Aachen University, 52066 Aachen,
Germany
c
Jülich Aachen Research Alliance, JARA-Energy, Germany
d
Institute for Power Generation and Storage Systems (PGS), E.ON ERC, RWTH Aachen University, 52074 Aachen, Germany

A R T I C LE I N FO A B S T R A C T

Keywords: In this work the characterization and application of a prototype solid-state lithium metal battery for an electric
Hybrid battery system vehicle application is addressed. This mid-temperature battery is operated at a temperature of 80 °C. In an
Solid-state lithium metal polymer battery innovative approach, this high energy technology is integrated into a hybrid battery system, where it is sup-
Electric vehicle ported by a highest power lithium-titanate technology. Three hybrid battery system configurations are sys-
tematically derived for a high class vehicle with different performance requirements. On the basis of an overall
vehicle simulation model, comprising a thermal model of the lithium metal polymer battery pack, the perfor-
mance of the system is evaluated under regular and extreme driving conditions. The results indicate that battery
hybridization makes a reasonable utilization of the lithium metal polymer battery in an electric vehicle possible.
The power capability of the overall battery system in discharge as well as charge direction is significantly
increased. Moreover, the system allows for a wider utilization of overall battery system's energy, resulting in an
increase of drivable distance. Furthermore, additional functionalities of the hybrid battery system are discussed,
such as warm-up procedure, low temperature performance enhancement and redundancy.

1. Introduction offers superior specific capacity in comparison to graphite [14–16].

The integration of battery systems into a vehicle's electrical drive-
train plays an important role for today's carmakers. But there remain
challenges in battery system operation and integration. Battery system
costs still outweigh those of other drivetrain components [1–5].
Moreover, the adequate dimensioning of battery systems with respect to
electrical power and energy for different vehicle types as well as the
achievement of a good lifetime performance [6–10] remain challen-
ging. In order to improve the cell's power and energy performance, cost
and safety, researchers work on the development of advanced battery
cell materials [11].
In today's electric vehicles the most prominent battery type is the
lithium-ion battery technology. Most vehicles’ battery cells comprise a
carbon based anode and a metal oxide material cathode [12,13]. In
order to achieve higher battery energy densities major effort is put on
the development of further battery technologies. One promising tech-
nology is the utilization of lithium metal anodes, since lithium metal
To avoid growth of lithium dendrites as well as to allow for an
electrochemically stable lithium metal anode operation the use of
polymer as well as inorganic solid-state electrolytes is discussed in lit-
erature [14–16].
The use of a solid-state polymer electrolyte may also lead to ad-
vantages in processing by supporting simpler and less production steps.
Moreover, due to the solid-state of the electrolyte versatile geometric
shapes are possible. Due to the comparably low ionic conductivity of
most polymer based solid electrolytes the battery operating tempera-
ture has to be increased [14].
This paper focuses on the integration of a lithium metal polymer
high energy battery in an electric vehicle. The investigated lithium
metal polymer battery is operated at a temperature of 80 °C. In order to
allow for advanced functionalities and an increased power capability
the lithium metal polymer battery is integrated into a hybrid battery
system topology.
The hybrid battery system combines two battery parts on system

⁎
Corresponding author.
E-mail address: Raphael.Wegmann@de.bosch.com (R. Wegmann).

https://doi.org/10.1016/j.est.2018.05.001
Received 10 March 2018; Received in revised form 30 April 2018; Accepted 1 May 2018
2352-152X/ © 2018 Elsevier Ltd. All rights reserved.
R. Wegmann et al. Journal of Energy Storage 18 (2018) 175–184

Nomenclature HBS hybrid battery system

LFP lithium iron phosphate
Symbols Li lithium
LTO lithium-titanate oxide
A surface NCA nickel cobalt aluminium oxide
Cth heat capacity PB high power battery part
E energy S semiconductor switch
F force SoC state of charge
I current SoE state of energy
L inductance act actual
N number of time steps amb ambient parameter
P power bat battery
Q, Q̇ heat, heat current batsys battery system
Rth thermal resistance ch charge
T temperature cond conduction
U voltage conv convection
cr, cw rolling friction and drag coefficient des desired
h heat transfer coefficient dch discharge
k time step dem demand
l length, thickness eb high energy battery part
s distance em electrical machine
t, τ time hbs hybrid battery system
λ thermal conductivity in inside the Li/LFP battery system
iso insulation
Acronyms nom nominal
pb high power battery part
BMS battery management system rms root mean square
D diode th thermal parameter
EB high energy battery part

level, where a battery part itself is an interconnection of multiple bat-
tery cells of one cell type. Several topologies of hybrid battery systems
are reviewed and suggested in [17,18]. Furthermore, the scaling of the
system is discussed e.g. in [19,20,4]. Here one battery part consists of
high energy density lithium polymer battery cells and the other battery
part consists of high power density lithium-ion cells. The cell types are
introduced in Section 2. The aim is to evaluate how much the energy
optimized mid-temperature battery part can be supported in specific
operating conditions.
Moreover, for further information about the dimensioning as well as
operation strategies of such a system the reader is referred to previous
work of the authors [10,19,21]. This paper consequently extends our
previous work, since focus has been given to battery hybridization
based on lithium-ion batteries in the previous work exclusively. In the
following, new results are presented, analysing battery hybridization
based on lithium metal polymer mid-temperature batteries.
The remainder of the paper is structured as follows. In Section 2 the
battery cells investigated here are introduced and characterized. Then
the integration of the lithium metal solid-state cell is discussed. In
Section 2.3 the hybrid battery system (HBS) and its functionalities on
system level are debated. The modelling environment as well as the
thermal model of the lithium metal polymer battery are presented in
Section 3. Details about how to ensure an operating temperature of
80 °C are given. Then, in Section 4 case studies are derived for a high
class vehicle in order to compare a single-cell reference battery system
and different hybrid battery system topologies. Subsequently results are
discussed in Section 5. Finally, a conclusion is given in Section 6.
2. Combining high energy lithium metal polymer and high power
lithium-ion batteries in electric vehicles
In Section 2.1 high energy density lithium secondary batteries are
compared to each other in terms of performance characteristics. Section
2.2 discusses two types of high power density energy storages. The
lithium-titanate battery used in this work is presented. Finally Section
2.3 introduces the innovative approach of battery hybridization taking
into account the combination of a lithium metal polymer high energy
battery and a lithium-titanate high power battery on system level.
2.1. High energy density secondary lithium batteries
The anode of today's lithium-ion batteries usually consists of gra-
phite. By using a lithium metal anode instead of such a graphite anode,
the battery's energy density can be increased significantly. This is due to
the considerably larger specific capacity and lower electric potential of
lithium metal in comparison to graphite [14,22]. Therefore, a lithium
metal polymer battery shall be examined in this work, with respect to
its usage as a high energy density battery in an electric vehicle.
The specific battery presented in this work is a prototype version of
a lithium metal polymer battery (Li/LFP). It comprises a lithium iron
phosphate (LFP) cathode and is operated at 80 °C. The basic char-
acteristics of this cell are shown in Table 1. The open circuit voltage,
derived from a measurement in the laboratory, is shown in Fig. 1 (black
curve). The flat potential profile of the LFP cathode in combination with
the lithium metal anode can be clearly seen [23]. Furthermore, the
open circuit voltage curve of a state of the art lithium-ion high energy
cell (Panasonic NCR 18650 BD) is shown in the figure for comparison.
For this reference cell, the basic characteristics are also given in
Table 1. It comprises a graphite anode and a nickel cobalt aluminium
oxide (NCA) cathode and has been characterized in a previous pub-
lication of the authors [10].
Moreover, the 10 s discharge and charge peak power capabilities
over state of charge (SoC) are shown in Fig. 2a and b respectively for
the two high energy lithium secondary battery cells. Actually the power
capability is normalized by the cell's nominal energy for better com-
parison. The peak power capability takes into account the respective
battery's SoC dependent open circuit voltage, the temperature and SoC
dependent inner resistance as well as the mostly temperature dependent

176
R. Wegmann et al. Journal of Energy Storage 18 (2018) 175–184

Table 1
Characteristics of the reference lithium-ion high energy cell (Panasonic), the prototype lithium metal polymer battery (Li/LFP) high energy cell and the high power
battery cell used in this work. The performance measures are derived for the respective cell's nominal operating temperature, i.e. 80 °C for the Li/LFP cell and 23 °C
for the other cells. Current limits of the lithium metal polymer battery cell are related to a prototype cell version.
Secondary lithium battery Panasonic NCR 18650 BD High energy Li/LFP prototype cell High power LTO cell

Nominal capacity Qnom 3.18 Ah 10.6 Ah (C/3) 2.9 Ah

Nominal voltage Unom 3.6 V 3.42 V 2.4 V
Operating voltage range 2.5 V … 4.2 V 2.5 V … 3.6 V 1.8 V … 2.8 V
Rdc,dch (50%, 5 s) 38.11 mΩ (1 C, 23 °C) 10.8 mΩ (1 C, 80 °C) 3.5 mΩ (5 C, 23 °C)
Specific discharge power 454 W kg−1 396 W kg−1 2400 W kg−1
Specific energy 244 Wh kg−1 212 Wh kg−1 46 Wh kg−1
Discharging peak current 2C 2C 70 C
Charging peak current 1C 1C 70 C
Cont. discharging current 1C 1C 10 C
Cont. charging current C/2 C/6 10 C
Cell weight 0.047 kg 0.171 kg 0.155 kg
Cell volume 0.0177 l 0.13 l 0.08 l

voltage and current limits. For a deeper insight into the calculation of
the peak power values the reader is referred to a previous work of the
authors, where the calculation is explained in more detail [21]. Espe-
cially at low or high SoC and/or at low temperatures the accurate
calculation of the peak power capability is necessary in order to meet
the battery's operational boundaries. This is also pronounced in [24].
The authors furthermore show that small deviations from the peak
power value can occur, depending on the calculation method. Together
with an adequate battery state estimation, e.g. based on machine
learning [25], a robust method for peak power calculation helps opti-
Fig. 1. Open circuit voltage curves of the solid-state lithium metal polymer LFP
mizing the battery system operation.
and the lithium-ion NCA battery. The lithium-ion NCA cell has been char-
The power capability graph for the lithium metal polymer LFP cell is
acterized in a previous publication of the authors [10].
shown for the operating temperature 80 °C. The power capability graph
of the high energy lithium-ion NCA battery cell is shown for the oper-
ating temperature 23 °C. The comparison shows that there remains in- passivating layers and there is no risk of lithium metal plating, since the
creased discharge power capability for the lithium metal LFP cell in low potential of LTO vs. Li/Li+ is substantially higher [32]. Moreover, a
state of charge region, as can be seen in Fig. 2a. This is due to the LTO battery can provide high power capability even at low and sub-
relatively high overpotential that can be drawn from the battery over a zero temperatures, allowing for recuperation even at sub-zero tem-
wide state of charge range. The full cell voltage only decreases from its peratures.
medium voltage of around 3.42 V at very low and high states of charge The LTO cell used in this paper has been introduced in a previous
as can be seen in Fig. 1. In addition to this there remains increased publication of the authors [10]. It's nominal specific discharge power is
charge power capability1 for the lithium metal LFP cell in high states of 2400 W kg−1. The essential characteristics of the high power LTO cell
charge. Thus, power capability is sustained over a wider state of charge are shown in Table 1. The table also shows the current rate limitations
range than it is for the lithium-ion battery, whose open circuit voltage that are implemented in the energy management of the vehicle simu-
decreases more continuously with decreasing state of charge. lation environment, see Section 3.
A Ragone plot, comparing the discussed battery cells, is shown in
Fig. 3. The plot proves that specific power and specific energy of the
2.2. High power lithium-ion battery and electrochemical capacitor overall battery system can be scaled appropriately by battery hy-
bridization. With dimensioning of the two hybrid battery parts the
The combination of secondary batteries that are designed towards a adequate scaling of the vehicle's power and energy requirements can be
high energy density with electrochemical double-layer capacitors can achieved. The latter will be discussed in Sections 2.3 and 4.2.
be found in many literature references [26,27,20,28,29]. These double-
layer capacitors provide very good sub-zero temperature power per-
2.3. Hybrid battery system topologies
formance [30], a high efficiency and a high cycle life as well as an
excellent cost-per-power ratio [31–33]. However, the gravimetric and
As discussed, hybrid battery systems are defined as a combination of
volumetric energy density is much lower than that of lithium-ion bat-
a high energy cell based battery part and a high power cell based
teries and the cost-per-energy is considerably higher than that of li-
battery part. The battery parts themselves consist of multiple inter-
thium-ion batteries [30,32].
connected high energy and high power battery cells respectively.
Here a lithium-ion battery with a lithium-titanate anode (LTO) is
When combining the two battery parts several concepts are possible.
chosen, which is designed towards a high power density [34]. It is an
In the end different criteria have to be weighed, e.g.:
off-the-shelf battery cell. In particular the high power lithium-ion LTO
battery cell is presented as a counterpart to the high energy battery.
Multiple LTO battery cells are interconnected and utilized in the high • Trade-off between cost and component lifetime, weight and volume.
power battery part. • Electrical efficiency.
An advantage of the LTO anode over a graphite anode is that the • Functionality: Power-, range-, thermal- and drivability-require-
ments.
LTO anode does not show such strong ageing effects caused by
• Safety.
1
Note the sign convention of {P, I} < 0 for charging power or current. By means of these main criteria an evaluation has been conducted.

177
R. Wegmann et al.
Fig. 2. Comparison of the high energy solid-state lithium metal polymer pro-
totype battery cell (Li/LFP) with a state of the art high energy lithium-ion NCA
battery cell. The lithium-ion NCA cell has been characterized in a previous
publication of the authors [10]. (a and b) 10 s discharge as well as charge peak
power-to-nominal energy ratios of two high energy lithium secondary battery
cells. Shown for the prototype lithium metal polymer LFP cell at operating
temperature 80 °C and at 23 °C (2 C discharging, 1 C charging) for the lithium-
ion NCA cell.
Fig. 3. Ragone plot showing discharge characteristics of the high power LTO
cell, the high energy NCA cell and a representative operation point of the solid-
state lithium metal polymer cell. The LTO's and the NCA's Ragone plot are
shown for an initial cell temperature of 23 °C and the solid-state battery's
Ragone plot is shown for the initial cell temperature 80 °C.
The hybrid battery system topology resulting from the evaluation is
shown in Fig. 4.
The high energy battery part comprises the lithium metal polymer
solid-state battery cells. Thus, this battery part is operated at 80 °C. The
thermal model of this battery part is introduced in Section 3.1. The high
Fig. 4. Drivetrain with hybrid battery system. High energy battery part contains
solid-state lithium metal polymer battery cells and high power battery part
contains lithium-ion LTO battery cells.
178
Journal of Energy Storage 18 (2018) 175–184
power battery part comprises the high power LTO cells. The overall
hybrid battery system concept allows for a substantial increase in power
capability and improved functionalities as will be discussed in Section
5.
Here, the high power battery part is directly linked to the inverter
dc-link. Thus no separate dc-to-dc converter that connects the high
power battery part and the dc-to-ac inverter in Fig. 4 is necessary. This
reduces cost, weight and volume. Furthermore, electrical losses that
would occur in an additional dc-to-dc converter can be avoided. The
linkage of the high power battery part to the dc-link is also preferable,
since this battery part is operated in limited medium state of charge
range for most of the operating time in order to provide power cap-
ability both in charge as well as discharge direction. Therefore the high
power battery voltage level is designed towards a preferable inverter
input voltage of around 400 V.
When connecting the two battery parts of the hybrid battery system
no galvanic isolation of the dc-to-dc converter is necessary, since only
moderate voltage ratios between the two battery parts occur for the
investigated systems. This can be seen from the battery system setups,
whose dimensioning is discussed in Section 4.2. Taking into account the
number of serially connected cells in Table 3 and the respective cell's
voltage limits, the maximum voltage ratio between the high power and
the high energy battery part is about 1.6. With the chosen dc-to-dc
converter topology, the power flow can be well converted between
those two voltage levels. Moreover, the ratio 1.6 corresponds to an
extreme operation point from the lowest voltage of one battery part to
the highest voltage of the other battery part. In real hybrid battery
system operation, the voltage ratio will be lower than 1.6 most of the
time, since both the Li/LFP as well as the LTO battery cell have a flat
open circuit voltage curve over SoC. A bi-directional power flow shall
be enabled in order to allow for an energy and power efficient operation
of the overall battery system. A 4-switch buck-boost converter topology
is used here. It is shown in Fig. 5, comprising semiconductor switches
(S), diodes (D) and the inductance L. Both battery parts operate in a
similar voltage range and recuperation can be performed to both bat-
tery parts. Moreover, with the presented system topology the high
power battery part can be charged while the high energy battery part is
discharged and vice versa. A discussion on the explicit dimensioning of
the battery parts is carried out in Section 4.
3. Vehicle modelling environment and operation strategy
To evaluate the electric vehicle performance in different driving
situations a vehicle simulation model is developed. An outline of the
model comprising the different drivetrain components and the overall
energy management is given in Fig. 6.
Next to detailed models of the two battery parts as well as the other
drivetrain components (dc-to-dc converter, chassis, dc-to-ac inverter,
transmission) an energy management controller as well as a battery
management unit (BMS) are implemented. The driver is modelled by
means of a PI-controller, which controls the vehicle velocity according
to a predefined target velocity (driving cycle). This approach offers the
advantage to immediately feed back and react on limitations of one of
the drivetrain components. E.g. when an electrical limit is reached the
vehicle controller will restrict vehicle acceleration and velocity in order
Fig. 5. Bi-directional 4-switch buck-boost dc-to-dc converter topology.
R. Wegmann et al.
Fig. 6. Electric vehicle simulation model with drivetrain components including
the hybrid battery system (block battery system).
to operate all drivetrain components within their operating boundaries.
Hence, a driving mission and different driving situations can be mod-
elled with the comprehensive vehicle model. Furthermore, the perfor-
mance of the hybrid battery system and the related energy management
strategy can be tested.
In extreme driving situations the battery operating limits have to be
fulfilled too. In order to meet requirements in terms of peak and con-
tinuous current limits, moving average currents are introduced.
Comparable approaches of battery current root mean square calculation
in order to allow for a specific battery operation can be found in
[28,35,36]. In discharge direction of the respective battery the dis-
dch
charge root mean square definition Irms is given in Eq. (1).
1 t
dch
Irms =
Δtdch
∫τ=t−Δt dch
Ibat 2 (τ )dτ
(1)
dch
In the given formula the root mean square battery current is
Irms
calculated over an elapsed time period Δtdch. The period is in the range
of 60–200 s. Likewise an operation-related mean charging current cal-
culation is implemented in the modelling environment.
During battery operation the charge and discharge current limits of
the respective battery part are related to the respective root mean
square current by the energy management system. When a certain root
mean square limit is reached, the current limits are reduced so that the
root mean square currents remain within certain specified ranges.
3.1. Thermal model of the lithium metal polymer battery
In order to model the temperature behaviour of the lithium metal
polymer battery system a simple temperature model is developed. The
nominal operating temperature of this battery is 80 °C. Hence, a pack
insulation is needed. The structure of the thermal pack and the related
thermal parameters are outlined in Fig. 7a.
The heat capacity of all battery cells is combined in Cth. Heat current
resulting from ohmic losses and polarization in the batteries is given by
Q̇source . In order to keep the battery at an adequate temperature a
heating current term Q̇ heating is considered. This term allows for an ad-
ditional heat source when the battery temperature drops below 80 °C.
The related power that is necessary for the heating is taken from the
Fig. 7. Approach of the thermal lithium metal polymer battery pack model. The battery is operated at 80 °C, therefore, a thermal insulation is needed. (a) Sketch of
the battery thermal pack. (b) Thermal network model.
179
Journal of Energy Storage 18 (2018) 175–184
battery system. It is found that during the investigated driving scenarios
no additional cooling but heating of the batteries is needed. Hence, no
additional cooling system or strategy is investigated here. The battery
temperature is given by Tin. The batteries are isolated from the en-
vironment with an insulation material of thickness liso. In literature one
can find high performance insulating materials that achieve a thermal
conductivity as low as 0.0038 W m−1 K−1 [37]. Inevitably the inside of
the battery pack has to be connected to the outside of the battery pack.
For a high voltage power cable connection from the battery cells to the
electric drive this feedthrough is needed. Also for the monitoring cable
and thermal system periphery a connection has to be established.
Hence, the thermal insulation gets worse due to a good thermal con-
ductivity of electric cables consisting of metal. An overall thermal
battery pack conductivity of λiso = 0.007 W m−1 K−1 and an insulation
thickness liso = 1 cm are assumed here.
In order to model the discussed thermal battery pack behaviour, an
equivalent circuit based thermal model is established. It is given in
Fig. 7b. The above mentioned thermal quantities can be found there. In
addition the thermal heat flow from the inside to the outside of the
battery pack is considered by thermal resistances. The battery pack
insulation is provided by Rth,iso. It is calculated with the insulation
material thickness liso, the outer insulation housing surface Aiso and the
mentioned overall thermal conductivity λiso with the relationship
Rth,iso = liso · λiso−1 · Aiso−1. Heat transfer from the outer insulation
housing to the environment is considered by convection. It is assumed
that the case surface temperature is close to the environment tem-
perature Tamb, and thus radiation is neglected. A convective heat
transfer coefficient hconv = 13.5W m−2 K−1 is assumed. This value re-
flects the heat transfer at a free convection situation according to
measurements in [38]. The related thermal resistance is then calculated
with Rth,conv = hconv−1 · Aiso−1.
The overall thermal behaviour can then be modelled by the heat
balance equation given in Eq. (2). The convective heat current Q̇conv can
be calculated according to Eq. (3).
dTin
Cth· = Q˙ source + Q˙ heating − Q˙ conv
dt (2)
Tin − Tamb
Q˙ conv =
Rth,in + Rth,conv (3)
3.2. Thermal behaviour of the LTO battery
A thermal system of the LTO high power battery part is not ex-
plicitly investigated in this work. Nevertheless, the effects of the load
characteristics on the thermal behaviour of this battery part shall be
discussed briefly. About 80% of energy throughput at the LTO high
power battery part occurs at charge and discharge current rates below
10 C and about 45% energy throughput at rates below 5 C. Hence,
significantly less energy throughput occurs above 10 C and mostly these
are short-duration peak loads. The latter has also been seen for all li-
thium-ion hybrid battery systems for BEV in another publication of the
authors [10].
R. Wegmann et al.
Laboratory measurements have revealed uncritical thermal beha-
viour of the LTO battery. Even for the much more stressing situation of
non-stop symmetric 10 C cycling (compared to the driving cycle sce-
narios described above) at an ambient temperature of 23 °C, the cell's
surface temperature converges to a stationary value of around 34 °C
(cell standing in the climate chamber with low convective turbulences).
Of course, due to stacking of multiple cells and differing ambient con-
ditions the thermal behaviour in a battery system will change.
Nevertheless, based on the explanations and the robust cell's thermal
behaviour, it is assumed, that the monitoring and managing of the LTO
high power cell's temperature can be well handled in a real vehicle
system.
4. Operation under different driving conditions
In this section it is evaluated how a lithium metal polymer battery
can be used as a part of a hybrid battery system. We derive the di-
mensioning of the hybrid battery system for different performance re-
quirements. Moreover, we discuss the system behaviour in different
driving scenarios.
4.1. Electric vehicle performance
Depending on the demanded vehicle performance and driving
range, power and energy requirements arise, which have to be fulfilled
by the battery system and the electric drive. Here, different battery
system configurations for a high class vehicle are discussed.
The high class vehicle comprises an overall mass of 2041 kg, an
aerodynamic surface of 2.375 m2, a rolling friction coefficient of
cr = 0.0095 and an aerodynamic drag coefficient cw = 0.24. These
characteristics have been derived based on the Tesla Motors Model S
high class vehicle, which is available on the market [39,40]. Fig. 8
shows the resulting peak discharge power requirements of a high class
vehicle for different acceleration requirements. The resulting power
requirements for an acceleration procedure starting from 0 km/h to
100 km/h as well as from 80 km/h to 120 km/h in target time Δt are
shown in this graph. The calculation of the peak power is based on the
kinetic energy difference between the vehicle's initial and target velo-
city ΔEkin. According to [41], the peak power can then be calculated
with Ppeak ≈ 2 · ΔEkin · Δt−1. It can be seen that the peak power demand
increases exponentially with decreasing target acceleration time Δt. The
markings highlight the acceleration requirements for different perfor-
mance classes that are introduced here.
Other vehicle requirements, such as the demanded vehicle's range,
the maximum velocity as well as the performance on a road with slope,
contribute to the battery system and electric drive requirements too.
Table 2 shows different vehicle performance classes. The classes
Fig. 8. Peak discharge power requirements of the high class vehicle for three
performance classes “Dynamic”, “Performance” and “Sport”. The graphs show
the peak power demand that is necessary to accelerate the vehicle from the
given initial to final vehicle velocity within time step Δt.
180
Journal of Energy Storage 18 (2018) 175–184
Table 2
Characteristics of the vehicle performance classes “Comfort”, “Dynamic”,
“Performance” and “Sport”.
Requirement Unit Comfort Dynamic Performance Sport
Δt, 0 … 100 km/h s 12 9.2 7.2 4.5
Δt, 80 … 120 km/h s 9.7 6.8 4.5 2.4
Max. velocity km/h 140 160 180 180
Up-hill velocity km/h 50 50 50 50
Up-hill slope % 12 12 12 12
“Dynamic”, “Performance” and “Sport” are utilized in the following.
The requirement of maximum velocity on a plane road as well as the
up-hill velocity requirement mainly contributes to the continuous
electrical power capability that has to be provided by the vehicle's
battery system. The power demand depends on vehicle velocity, road
slope and the discussed vehicle parameters. It can be calculated ac-
cording to the theory of longitudinal dynamics, which has been fre-
quently discussed in literature, e.g. in [41,42]. A hybrid battery system
is designed for each of those classes in the next section.
4.2. Battery system dimensioning
Previously it was explained how different performance require-
ments together with the overall vehicle body data contribute to the
electrical power requirements of the battery system. In addition to this,
the requested vehicle range mainly affects the nominal energy of the
battery system. To determine this energy, the route-specific energy
demand in “Wh/km” is calculated for the high class vehicle for different
driving cycles. Based on the maximum route-specific energy demand
and the vehicle range the requested battery system nominal energy is
then derived.
In Table 3 the resulting battery system requirements for the high
class vehicle with three different performance classes and a required
vehicle range of 400 km are shown. Moreover, hybrid battery system
configurations are given that meet the respective power and energy
requirements.
In order to compare the hybrid battery system topology to a con-
ventional battery system, as it can be seen in today's electric vehicles, a
single-cell reference pack is displayed too. This battery system com-
prises only one cell type, i.e. the lithium metal polymer high energy
cell. The single-cell type pack is designed with sufficient energy to meet
the range requirement.
In literature weight and volume packaging factors can be found,
which specify the share of the cumulated cell mass or volume in the
overall battery system weight or volume respectively. Depending on the
individual form factors and specific vehicle integration requirements,
these packaging factors can diverge. Thus, the two factors are given by
“cumulated cell mass/pack mass” and “cumulated cell volume/pack
volume”. For BEV battery systems weight packaging factors around 0.7
or 0.75 can be found [4,43]. Moreover, a volume packaging factor
around 0.5 or 0.6 is given in [43]. Compared to the BEV battery system,
additive weight and volume can arise in the Li/LFP hybrid battery
system due to the dc-to-dc converter, the thermal system, the battery
pack's thermal isolation and a comparably low energy density and
specific energy of the high power cells. Hence, for the hybrid battery
systems here, a weight packaging factor of 0.7 and a volume packaging
factor of 0.55 is assumed. For the single-cell reference pack the litera-
ture values 0.75 and 0.6 are assumed respectively. Finally, the calcu-
lated overall battery system weight and volume is given in Table 3.
4.3. Driving scenarios
Here extreme driving conditions are investigated in order to eval-
uate the performance of the hybrid battery system. Hence, two driving
scenarios on roads that comprise high proportions of slope are chosen.
R. Wegmann et al. Journal of Energy Storage 18 (2018) 175–184

Table 3
Battery system setups for the different vehicle performance classes for the high class vehicle comprising a range of 400 km. Furthermore, a single-cell reference
system is derived providing a nominal energy that equals the respective hybrid battery system configurations (HBS).
Single-cell reference pack Hybrid battery system configurations

HBS “Dynamic” HBS “Performance” HBS “Sport”

Battery system peak discharge power requirement – 185.28 kW 279.97 kW 524.95 kW

Battery system continuous power requirement – 47.34 kW 64.30 kW 64.3 kW
Battery system nominal energy requirement – 70.32 kWh 70.32 kWh 70.32 kWh

Battery system weight (including dc-to-dc converter) 510 kg 564 kg 609 kg 728 kg
Battery system volume (including dc-to-dc converter) 484 l 538 l 564 l 633 l

High energy battery nominal energy (res. Single-Cell Pack) 70.4 kWh 69.6 kWh 67.7 kWh 62.8 kWh
High energy battery cell configuration (86s|26p) (85s|26p) (86s|25p) (95s|21p)
High power battery nominal energy – 0.75 kWh 2.63 kWh 7.57 kWh
High power battery cell configuration – (107s|1p) (126s|3p) (136s|8p)

Fig. 11. Accumulated recuperated energy (absolute value) depending on re-

cuperation phase duration and mean recuperation power. The right y-axis
shows the relative energy throughput by recuperation for a battery part of
1 kWh nominal energy.

Fig. 9. Altitude profiles of the driving cycles “Kasseler Berge” and “Schwäbisch
Hall”.
Fig. 9a shows the altitude profiles of a highway section near Kassel
in Germany.
It is the section of Autobahn A7 between Kirchheim (postal code
36275) and Göttingen (postal code 37079). This section is well known
for its high road slope. A ride with a constant velocity of 100 km/h is
utilized as a benchmark highway drive here. Furthermore, the altitude
profile of a road section in the geographic area “Schwäbisch Hall” in
Germany (postal code 74523) is shown in the graph. This section serves
as a second extreme driving scenario. Note that the altitude in the be-
ginning of the respective driving cycle is higher than the altitude at the
end of the drive. Thus, the vehicle starts with an increased potential
energy. Fig. 10 shows an excerpt from the downhill driving phase of the
“Schwäbisch Hall” driving cycle. A high mean recuperation traction
power Ptraction arises. When the inverter losses are taken into account
the power Pbatsys can be re-charged to the battery system (negative
power accounts for charging direction).
Depending on battery system energy, the applied technology and
operational requirements, conventional battery systems may fail to
sustain power demands in extreme driving phases and for a longer time
period.
Taking into account the high peak and continuous power capability
of high power type batteries, as provided by the LTO cell investigated in
this work, a simple example is given in Fig. 11 of how the addition of a
high power battery part can help to sustain power capability and in-
crease the ratio of recuperated energy. The figure displays the linear
relationship of recuperated energy, which is made available to the
battery system over a time of up to 100 s and for two constant re-
cuperation power graphs. On the right side the second y-axis shows the
resulting relative energy increase when the whole recuperation energy
is stored to a battery with the size of 1 kWh. It can be seen that even for
a period of about 1 min of recuperation with up to −20 kW, only a
fraction of the nominal energy (< 60%) is recuperated. Hence, even an
energetically small high power battery part can provide essential power
capability and functionality, taking into account the comparably low
peak and continuous power capability of the high energy battery part.
Highest power density storage technologies like the lithium-titanate
battery used in this work show peak power-to-energy ratios that allow
for high charging/discharging power even if the storage part's nominal
energy is comparably low or high respectively.

4.4. Evaluation criteria

Fig. 10. Extract of a downhill driving cycle section in the geographic area
“Schwäbisch Hall”, Germany. A long recuperation phase with high average
battery system charging power results during the drive (negative power for
charging).
When the batteries – either the battery parts of the hybrid battery
system or the single-cell reference battery pack – reach low states of
charge, the power capability decreases significantly. Consequently not
all propulsion power demands can be provided by the battery system
and the vehicle decelerates so that not all torque required by the driver

181
R. Wegmann et al. Journal of Energy Storage 18 (2018) 175–184

Table 4
Resulting performance measures for the high class vehicle, two driving cycles, the single-cell battery system and the different hybrid battery systems (HBS). Measures
1–3 are calculated for the whole driving distance. Measures 4 and 5 are calculated beginning at a respective battery system state of energy of 5% until stop of
simulation.
No. Measure Unit “Kasseler Berge” “Schwäbisch Hall”

Single-cell ref. HBS HBS HBS Single-cell ref. HBS HBS HBS
– Dynamic Performance Sport – Dynamic Performance Sport

1 batsys
P peak,dch kW 151 192 296 553 151 193 297 560

2 batsys
P peak,ch kW −84 −111 −189 −464 −84 −110 −182 −438

3 Δs+ % 0 2 2.1 5.7 0 9.7 9.7 12.8

4 batsys
P peak,dch kW 56 88 185 361 46 84 174 302

(SoE ≤ 5%)
5 batsys
P peak,ch kW −84 −113 −199 −494 −83 −110 −182 −462

(SoE ≤ 5%)

is provided. Driving cycles are repeated until a point is reached where
the battery system for the respective vehicle is not able to further
provide necessary power to follow the driving cycle anymore. Thus, the
following stopping criteria is defined: During the deceleration proce-
dure the driven distance sact deviates from the desired reference dis-
tance of the given driving cycle sdes, the deviation is described by
d2
Δs = sdes − sact. Then the second derivation (dt )2 Δs is evaluated, it gives
the deceleration of the distance deviation in the physical unit m/s2. A
distance deceleration of 3 m/s2 is chosen here as a stopping criteria for
the vehicle simulations. This measure is comparable to the discomfort
index introduced in [44].
The power capability of the battery system depends on the states of
energy2 (SoE) or respectively on the states of charge of the battery
parts. In order to compare the power capabilities the arithmetic mean
power values for different situations are calculated. In Eq. (4) the cal-
culation of the hybrid battery system mean discharge power capability
over N discrete time steps is shown.
N −1 eb pb
batsys
∑k = 0 Pdch,peak (k ) + Pdch,peak (k )
P peak,dch =
N (4)
The arithmetic mean discharge power capability for the single-cell
battery system is calculated accordingly, with the difference that no
power battery part has to be considered. In the same manner the ar-
ithmetic mean charge power capability (subscript ch.) is calculated for
the battery configurations.
Moreover, the mean power capabilities within the last 5% of the
respective overall battery system's energy are investigated.
A further measure that is calculated for the whole driving trip is
given in Eq. (5). It defines the additional distance that can be provided
by the respective hybrid battery system (hbs) in comparison to the
hbs
single-cell reference system. stotal defines the total driving distance of
single − cell
the respective hybrid battery system and stotal defines the total
driving distance of the single-cell reference system in the respective
driving cycle.
hbs
⎛ stotal ⎞
Δs+ = 100%·⎜ single-cell − 1⎟
⎝ stotal ⎠ (5)
5. Results and discussion
As discussed in the previous section, the single-cell reference battery
system and hybrid battery systems for three different performance
classes are compared to each other on two driving routes. A drive in the
2
In the following the battery “state of energy” (SoE) is used. It indicates the relative
share of the respective battery system energy in percent.
“Schwäbisch Hall” as well as the “Kasseler Berge” driving cycle is si-
mulated for all battery systems with the presented vehicle simulation
model. In order to use a large amount of the battery system's energy,
scenarios are preferred which make the vehicle drive a long distance.
All scenarios are started with 100% states of charge in the battery parts.
The cycles are then repeated until the stopping criteria is reached,
which was discussed in Section 4.4.
In Section 5.1 the performance of the battery systems is evaluated
for the whole driving missions in terms of power capability. In Section
5.2 special focus is given to the performance at low states of energy.
Then, in Section 5.3 several functionalities that can be achieved with
battery hybridization are discussed.
5.1. Performance for the whole driving mission
An evaluation of all driving scenarios is carried out and the results
are shown in Table 4. Measures 1–3 evaluate the performance of the
battery systems over the respective whole driving mission.
When comparing the arithmetic mean discharge and charge power
of the respective battery system, given by measure 1 and 2, the gradual
increase in power capability with an increase in the high power battery
part's nominal energy can be seen.
A comparison of the single-cell battery system and the hybrid bat-
tery systems for the “sport” performance class shows that the hybrid
battery systems allow for an up to 3.7 times higher arithmetic mean
batsys
discharge power capability P peack,dch than the single-cell battery system
in the “Schwäbisch Hall” and the “Kasseler Berge” driving cycle re-
spectively. The respective arithmetic mean charge power capability
batsys
P peack,ch is up to 5.5 times smaller (i.e. the charge power value for hy-
brid battery system setups is smaller since charge power is defined to be
negative.). Hence, the possibility of a battery hybridization to provide
substantial power capability is emphasized. This could be interesting
for example in sports car settings as well as in aged battery states with
power deficits.
Measure 3 in Table 4 reflects the additional distance that can be
provided by the respective hybrid battery system (hbs) in comparison to
the single-cell reference system, as discussed in Section 4.4. It is worth
noting that an increase in distance of up to 5.7% for the “Kasseler
Berge” and 12.8% for the “Schwäbisch Hall” driving scenarios can be
achieved with the hybrid battery systems. This is due to the higher
discharge power capability of the hybrid battery systems within low
states of energy, as will be discussed in the following section. The hy-
brid battery systems manage to provide comparably high discharge
power capability in the final phase of the driving scenario. The power
capability per energy of the high power lithium-titanate battery, in-
troduced in Section 2, is some orders of magnitude higher than the
power-to-energy ratio of the investigated high energy cell. As a con-
sequence higher power capability is provided by the battery system.

182
R. Wegmann et al.
5.2. Performance at low states of energy
For an electric vehicle it is desirable to sustain high discharge power
capability even at low states of energy in order to satisfy all desired
driver torque requests over a wide energy range of the battery system.
Therefore, the arithmetic mean power capability for the discharge as
well as the charging direction is compared for the battery system
topologies within the last 5% of battery system state of energy for the
different driving scenarios. See measures 4 and 5 in Table 4. When
comparing the single-cell and the hybrid battery systems for the “sport”
performance class it can be seen that hybrid battery systems allow for
an up to 6.6 times higher arithmetic mean discharge power capability
than the single-cell battery system in the “Schwäbisch Hall” and the
“Kasseler Berge” driving cycle respectively. The respective arithmetic
mean charge power capability within the last 5% of battery system's
energy is up to 5.9 times smaller. For all hybrid battery system setups
the arithmetic mean discharge power is higher than the related single-
cell battery system power. Also the negative arithmetic mean charge
power is smaller for all hybrid battery systems, what in fact implies
higher charging power capabilities of the hybrid battery systems. The
additional driving distance that can be provided by the hybrid battery
systems, which was discussed in Section 5.1, is due to this improvement
in power performance, especially in low states of energy.
In Fig. 12 an excerpt of two vehicle simulations is shown. Starting
from the 5% leftover of the overall battery system's nominal energy the
graphs show the peak power discharge capability for the hybrid battery
system as well as for the single-cell reference battery pack. For better
comparison the graphs of the single-cell reference as well as for the
hybrid battery system start at time 0. Furthermore, the battery part's
overall energy trajectories are shown in the lower subfigure. It can be
seen that the discharge power capability within the last 5% of the
battery system energy is higher for the hybrid battery system config-
urations in both driving scenarios. The hybrid battery system for the
“sport” performance class comprises a comparably large high power
battery part in terms of energy, i.e. 7.57 kWh. This configuration allows
for a very high discharge power capability, even until a battery system's
state of energy of as low as 0.3% is reached in the “Kasseler Berge”
scenario, as can be seen in Fig. 12b. The single-cell reference system's
state of energy can only be utilized to 2.8% until the stopping criteria is
reached and the vehicle has to be decelerated quickly in order to op-
erate the drivetrain components within their operating boundaries.
The power performance behaviour is shown for the “Schwäbisch
Hall” driving route in Fig. 12a. Within the presented section the hybrid
battery system's high energy battery part SoE is below 1%. Never-
theless, substantial discharge power capability is provided by the high
183
Journal of Energy Storage 18 (2018) 175–184
Fig. 12. Trajectories of peak power
capability and state of energy of the
different battery parts for two driving
cycles. Graphs are shown for the high
energy battery part (EB) and the high
power battery part (PB) of the hybrid
battery system, as well as the sum of
both (EB + PB). The single-cell re-
ference is compared to the “sport” hy-
brid battery system. The graphs start at
the same time 0 s that is related to the
5% battery system nominal energy. (a)
Results for driving cycle “Schwäbisch
Hall”. (b) Results for driving cycle
“Kasseler Berge”.
power battery part of this battery system.
Altogether, it is shown that the hybrid battery system allows for a
wider utilization of overall battery system's energy while providing
adequate power capability. This is due to the relatively large share of
high power battery part energy in the overall battery system energy.
5.3. Providing battery system functionality via battery hybridization
Another advantage of the hybrid battery system as presented, is the
possibility to use the high power battery part as an energy source for
heating processes in situations where the lithium metal polymer battery
needs to be warmed-up in order to be in adequate operating tempera-
ture range. When taking into account the thermal equation
Qwarm−up = Cth · ΔT and assuming that the high power battery energy
can be completely converted to the heat Qwarm−up, the temperature
increment for the hybrid battery system configuration of the “sport”
performance class is ΔT = 69.5 K.
Moreover, the hybridization of the lithium-metal polymer battery
system with the LTO battery allows for further functionalities. A short
summary is given in the following:
• Provide energy for a warm-up process of the mid-temperature li-
thium-metal polymer battery after times of vehicle rest period.
• Shift power demand to the high power battery part in phases where
the mid-temperature lithium-metal polymer battery did not reach its
target operating temperature 80 °C.
• Redundancy: Provide electrical energy for safety relevant features
and actuators in case of failure of one battery part. Guarantee a
minimum drivable distance, e.g. to the roadside, in case of failure.
• Adjust battery system power-to-energy ratios according to different
vehicle classes and requirements.
• Provide recuperation functionality even at sub-zero temperature.
• Reduce current amplitudes and charge throughput per driving dis-
tance on the high energy battery part.
6. Conclusion
In this work the integration of a high energy lithium metal polymer
battery (Li/LFP) into electric vehicles (BEV) with varying power per-
formance was investigated. This battery technology is operated at 80 °C
and currently beyond state of the art in battery systems. Moreover, the
Li/LFP battery is combined with a high power lithium-ion battery via
dc-to-dc converter, forming a hybrid battery system. The high power
battery comprises a lithium titanate anode. Three hybrid battery sys-
tems, addressing the varying BEV's power performances, are compared
R. Wegmann et al.
in two extreme driving scenarios with high shares of road slope. A
single-cell battery system serves as a reference.
It is shown that the hybrid battery system's arithmetic mean dis-
charge power capability over a wide state of charge range during a
driving mission is up to 3.7 times higher than the power capability of
the respective single-cell reference system. Moreover, a significant in-
crease of charge power capability can be achieved by battery hy-
bridization. Then, a wider utilization of overall battery system's energy
while providing adequate power capability can be achieved by hy-
bridization, which in fact allows for additional driving distance of up to
12.8% when comparing the most performant hybrid battery system and
the single-cell reference system. It is shown that by BEV's battery system
hybridization, substantial power capability is provided even when the
battery system's energy is below 5% of the rated energy.
Due to its excellent low temperature performance the LTO battery
can be used for recuperation functionality at sub-zero temperatures or
for warm-up scenarios of the lithium metal polymer battery. For the
most performant hybrid battery system, a warm-up temperature in-
crement of up to 69.5 K is provided.
Moreover, hybrid battery systems feature additional functionalities:
Storage redundancy, e.g. for autonomous driving solutions, support of
aged battery states, reduction of damaging loads on the high energy
battery and an adequate adjustment of battery system power-to-energy
ratios.
Nevertheless, when designing an automotive battery system, the
trade-off between cost, weight, volume, lifetime and functionalities has
to be found.
In the future it would be interesting to investigate whether the
presented hybrid battery system configurations, especially for the in-
vestigated lithium-metal solid-state cells, allow for a reduction of bat-
tery ageing over the driven distance, as energy throughput can be sig-
nificantly reduced. Moreover, a detailed analysis of overall system cost
and a benchmark to existing technologies should be performed.
References
[1] C. Pillot, The rechargeable battery market and main trends 2014–2025, 32nd
International Battery Seminar and Exhibit, Fort Lauderdale, USA, 2015.
[2] V. Hennige, The Tesla model S battery: a battery pack analysis study, Advanced
Automotive Battery Conference (2015).
[3] H. Budde-Meiwes, J. Drillkens, B. Lunz, J. Muennix, S. Rothgang, J. Kowal,
D.U. Sauer, A review of current automotive battery technology and future pro-
spects, Proc. Inst. Mech. Eng. Part D: J. Automob. Eng. 227 (5) (2013) 761–776.
[4] S. Rothgang, T. Baumhöfer, H. van Hoek, T. Lange, De Doncker, W. Rik, D.U. Sauer,
Modular battery design for reliable, flexible and multi-technology energy storage
systems, Appl. Energy 137 (2015) 931–937.
[5] N. Hartmann, E.D. Özdemir, Impact of different utilization scenarios of electric
vehicles on the German grid in 2030, J. Power Sources 196 (4) (2011) 2311–2318.
[6] M. Broussely, P. Biensan, F. Bonhomme, P. Blanchard, S. Herreyre, K. Nechev,
R.J. Staniewicz, Main aging mechanisms in Li ion batteries, J. Power Sources 146
(1-2) (2005) 90–96.
[7] J. Vetter, P. Novak, M.R. Wagner, C. Veit, K.-C. Möller, J.O. Besenhard, M. Winter,
M. Wohlfahrt-Mehrens, C. Vogler, A. Hammouche, Ageing mechanisms in lithium-
ion batteries, J. Power Sources 147 (1-2) (2005) 269–281.
[8] H.-p. Lin, D. Chua, M. Salomon, H.-C. Shiao, M. Hendrickson, E. Plichta, S. Slane,
Low-temperature behavior of Li-ion cells, Electrochem. Solid-State Lett. 4 (6)
(2001) A71.
[9] Y. Ji, C.Y. Wang, Heating strategies for Li-ion batteries operated from subzero
temperatures, Electrochim. Acta 107 (2013) 664–674.
[10] R. Wegmann, V. Döge, J. Becker, D.U. Sauer, Optimized operation of hybrid battery
systems for electric vehicles using deterministic and stochastic dynamic program-
ming, J. Energy Storage 14 (Part 1) (2017) 22–38.
[11] X. Hu, C. Zou, C. Zhang, Y. Li, Technological developments in batteries: a survey of
principal roles, types, and management needs, IEEE Power Energy Mag. 15 (5)
(2017) 20–31.
[12] J.B. Goodenough, Y. Kim, Challenges for rechargeable Li batteries, Chem. Mater. 22
(3) (2010) 587–603.
[13] M. Winter, J.O. Besenhard, M.E. Spahr, P. Novák, Insertion electrode materials for
rechargeable lithium batteries, Adv. Mater. 10 (10) (1998) 725–763.
[14] A. Varzi, R. Raccichini, S. Passerini, B. Scrosati, Challenges and prospects of the role
of solid electrolytes in the revitalization of lithium metal batteries, J. Mater. Chem.
184
Journal of Energy Storage 18 (2018) 175–184
A 4 (44) (2016) 17251–17259.
[15] B. Chen, Q. Xu, Z. Huang, Y. Zhao, S. Chen, X. Xu, One-pot preparation of new
copolymer electrolytes with tunable network structure for all-solid-state lithium
battery, J. Power Sources 331 (2016) 322–331.
[16] C. Xu, B. Sun, T. Gustafsson, K. Edström, D. Brandell, M. Hahlin, Interface layer
formation in solid polymer electrolyte lithium batteries: an XPS study, J. Mater.
Chem. A 2 (20) (2014) 7256–7264.
[17] T. Zimmermann, P. Keil, M. Hofmann, M.F. Horsche, S. Pichlmaier, A. Jossen,
Review of system topologies for hybrid electrical energy storage systems, J. Energy
Storage 8 (2016) 78–90.
[18] M.A. Hannan, M.M. Hoque, A. Mohamed, A. Ayob, Review of energy storage sys-
tems for electric vehicle applications: issues and challenges, Renew. Sustain. Energy
Rev. 69 (2017) 771–789.
[19] V. Döge, C. Kurtulus, V. Hennige, R. Wegmann, Scalable energy storage systems for
effective electrified mobility concepts, Transport Research Arena (TRA), vol. 14
(2016) 3621–3630.
[20] Y. Kim, J. Koh, Q. Xie, Y. Wang, N. Chang, M. Pedram, A scalable and flexible
hybrid energy storage system design and implementation, J. Power Sources 255
(2014) 410–422.
[21] R. Wegmann, V. Döge, D.U. Sauer, Energy optimal operation of hybrid battery
systems and comparison to a single-cell reference system for electric vehicles in-
cluding the aged state of high energy cells, 29th International Electric Vehicle
Symposium (EVS29), Electric Drive Transportation Association, Montreal, Canada,
2016, pp. 8–19.
[22] J. Groß, H. Eitouni, Drivers and technologies for Li-metal solid-state batteries,
Advanced Automotive Battery Conference (2016).
[23] V. Srinivasan, J. Newman, Discharge model for the lithium iron-phosphate elec-
trode, J. Electrochem. Soc. 151 (10) (2004) A1517.
[24] X. Hu, R. Xiong, B. Egardt, Model-based dynamic power assessment of lithium-ion
batteries considering different operating conditions, IEEE Trans. Ind. Inform. 10 (3)
(2014) 1948–1959.
[25] X. Hu, S.E. Li, Y. Yang, Advanced machine learning approach for lithium-ion battery
state estimation in electric vehicles, IEEE Trans. Transp. Electrif. 2 (2) (2016)
140–149.
[26] J. Ruan, P.D. Walker, N. Zhang, J. Wu, An investigation of hybrid energy storage
system in multi-speed electric vehicle, Energy 140 (2017) 291–306.
[27] Z. Cabrane, M. Ouassaid, M. Maaroufi, Analysis and evaluation of battery-super-
capacitor hybrid energy storage system for photovoltaic installation, Int. J.
Hydrogen Energy 41 (45) (2016) 20897–20907.
[28] A. Castaings, W. Lhomme, R. Trigui, A. Bouscayrol, Comparison of energy man-
agement strategies of a battery/supercapacitors system for electric vehicle under
real-time constraints, Appl. Energy 163 (2016) 190–200.
[29] V. Herrera, A. Milo, H. Gaztañaga, I. Etxeberria-Otadui, I. Villarreal, H. Camblong,
Adaptive energy management strategy and optimal sizing applied on a battery-su-
percapacitor based tramway, Appl. Energy 169 (2016) 831–845.
[30] J. Libich, J. Máca, J. Vondrák, O. Čech, M. Sedlaříková, Supercapacitors: properties
and applications, J. Energy Storage 17 (2018) 224–227.
[31] L. Zhang, X. Hu, Z. Wang, F. Sun, D.G. Dorrell, A review of supercapacitor mod-
eling, estimation, and applications: a control management perspective, Renew.
Sustain. Energy Rev. 81 (2018) 1868–1878.
[32] T.B. Reddy, D. Linden, D.H.o.b. Linden, Linden's Handbook of Batteries, 4th edition,
McGraw-Hill, New York and London, 2011.
[33] D. Cericola, R. Kötz, Hybridization of rechargeable batteries and electrochemical
capacitors: principles and limits, Electrochim. Acta 72 (2012) 1–17.
[34] B. Zhao, R. Ran, M. Liu, Z. Shao, A comprehensive review of Li4Ti5O12-based
electrodes for lithium-ion batteries: the latest advancements and future perspec-
tives, Mater. Sci. Eng. R: Rep. 98 (2015) 1–71.
[35] R.A. Dougal, S. Liu, R.E. White, Power and life extension of battery-ultracapacitor
hybrids, IEEE Trans. Compon. Packag. Technol. 25 (2002) 120–131.
[36] A. Lahyani, A. Sari, I. Lahbib, P. Venet, Optimal hybridization and amortized cost
study of battery/supercapacitors system under pulsed loads, J. Energy Storage
(2016) 222–231.
[37] Y. Chen, Three-dimensional thermal modeling of lithium-polymer batteries under
galvanostatic discharge and dynamic power profile, J. Electrochem. Soc. 141 (11)
(1994) 2947–2955.
[38] Y. Wang, J. Jiang, J.R. Dahn, The reactivity of delithiated Li(Ni1/3Co1/3Mn1/3)O2, Li
(Ni0.8Co0.15Al0.05)O2 or LiCoO2 with non-aqueous electrolyte, Electrochem.
Commun. 9 (10) (2007) 2534–2540.
[39] D. Sherman, Drag Queens: five slippery cars enter a wind tunnel; one slinks out a
winner, Car and Driver (2014) 86–92.
[40] www.nie-mehr-benzin.de, Tesla Model S Article (access date: 20th October 2017).
[41] L. Guzzella, A. Sciarretta, Vehicle Propulsion Systems, Springer, Berlin, Heidelberg,
2013.
[42] M. Ehsani, Y. Gao, A. Emadi, Modern Electric, Hybrid Electric, and Fuel Cell
Vehicles: Fundamentals, Theory, and Design. Power Electronics and Applications
Series, 2nd edition, CRC Press, Boca Raton, 2010.
[43] D.A. Crolla, D.E. Foster, T. Kobayashi, N. Vaughan, Encyclopedia of Automotive
Engineering, Wiley, Chichester, 2015.
[44] N. Mutoh, Y. Hayano, H. Yahagi, K. Takita, Electric braking control methods for
electric vehicles with independently driven front and rear wheels, IEEE Trans. Ind.
Electron. 54 (2007) 1168–1176.