*2.4. New Trends in Battery Technology and the Implication for EVs*

Lithium-ion (Li-ion) storage technology prices have decreased, while the manufacturing volumes have increased. Experience in manufacturing batteries for consumer electronics has driven cost reductions to the benefit of EV packs as well as stationary storage.

Figure 4 illustrates the cost reductions comparative to the cumulative manufactured capacity across Li-ion storage technologies used in various applications. It also shows that Li-ion batteries have proved significant cost reductions since their market introduction in the 1990s.

**Figure 4.** Evolution of Lithium-ion storage technology prices vs. installed capacity adapted from [27].

The early development of batteries for consumer electronics (e.g., smartphones, laptops, etc.) provided invaluable experience in the production of Li-ion cells, underpinning the attainment of the cumulative production capacity of 100 gigawatt-hours (GWh) by 2010 [27], enabling the achievement of very significant cost reductions and performance improvements over the past decade. These same developments made the development of Li-ion battery packs for EVs increasingly viable.

At the state-of-art, most of battery packs used in EVs exploit the lithium-ion technology. This technology is reaching a maturity level that is enabling the design of EVs with a performance comparable to ICE vehicles. Current battery packs for light-duty applications have gravimetric energy densities of 200 Watt-hours per kilogram (Wh/kg) [28], and volumetric pack energy densities of 200–300 Watt-hours per liter (Wh/L) [29]. The lifetime of the battery is another important parameter. For EV batteries, a good proxy is the expected mileage associated with a battery's lifetime, as well as its ability to retain a good share of its initial capacity (usually 80%).

According to the recent literature, modern Li-ion chemistry for EV batteries is able to withstand 1000-cycle degradation [30]. To compare this value to the lifetime of an ICE-based vehicle, let us assume a vehicle with a battery capacity of 35 kWh and an average consumption of 0.2 kWh/km. Withstanding 1000-cycle degradation suggests that the cycle life threshold would not be reached over the first 175,000 km of driving. Considering a car used mainly in an urban scenario for about 12,500 km/year, then 175,000 km and 1000-cycle recharging degradation means a lifetime of 14 years

for the battery pack. Therefore, the lifetime of the battery is compatible with the expected lifetime for a car.

Notwithstanding the complexity of battery design and manufacturing, four key cost and performance drivers have been identified for Li-ion batteries—chemistry, capacity, manufacturing capacity, and charging speeds. The cost per kWh of the currently available battery chemistries varies because of the different energy densities and material needs.

The size of the battery packs used at the state-of-art in EVs vary considerably. For BEVs, the size of the battery packs ranges from about 20 kWh to about 100 kWh. In China, which is, as discussed before, the worldwide market leader for EVs, the three bestselling EVs have battery sizes ranging between 18.3 kWh and 23 kWh, mainly because the market in China is focused on small vehicles, and their design is focused on affordability. Instead, in Europe and North America, for mid-sized cars, the capacity of battery packs ranges between 23 kWh and 60 kWh, whereas larger cars and SUVs have battery capacities ranging from 75 kWh and 100 kWh.

As far as the current charging speed is concerned, at the state-of-art, fast chargers enable 80% recharging in less than 1 h. Such a charging speed does not constitute a challenge for current battery design. Further increasing the maximum speed of charging to ultra-fast charging (which implies working at power levels as high as 300 kW or 400 kW) is a desirable feature that would decrease the performance gap of EVs compared to ICE vehicles. However, designing batteries for ultra-fast charging has the negative effects of increasing the complexity of their design and of shortening their lifetime. Accommodating fast charging requires specific battery design considerations, such as decreasing the thickness of the electrodes. These added design constraints tend to increase the cost of the battery and to decrease its energy density. With an appropriate design and appropriately sized thermal management system, the increases in fast charging are not expected to affect the battery's lifetime. On the other hand, an analysis conducted for the United States Department of Energy suggests that the change in battery design to accommodate 400 kW charging would nearly double the cell costs [31]. Indications from the recent assessments of battery technologies suggest that lithium-ion is expected to remain the technology of choice for the next decade (see Figure 5). The main developments in cell technology that are likely to be deployed in the next few years include the following:


The next generation of Li-ion batteries entering the mass production market around 2025 is expected to have a low cobalt content, high energy density, and NMC 811 cathodes. To increase the energy density up to 50% silicon can be added in small quantities to the graphite anode [28]. To contribute, better performance electrolyte salts that are able to withstand higher voltages, can be used. As reported in Figure 5, lithium-ion is assumed to remain the technology of choice for the next decade, when it is expected to take advantage of a number of improvements to increase the battery performance. According to Figure 5, other technology options are expected to become available after 2030.

**Figure 5.** Expected battery technology commercialization timeline.

#### **3. Hybrid-Electric 48 V Vehicles**

Hybrid vehicles may ensure a rapid and smooth transition from the current generation of petrol/diesel-based propulsion to new electrified mobility. HEVs may be realized by combining a downsized ICE with an electric motor, usually within 10 kW and with power DC buses up to 48 V nominal. In such cases, the on-board energy storage to provide extra energy for torque assist, or to store recovered energy when braking, can be limited to less than 1 kWh. The immediate benefit vs. current vehicle generations is fuel saving and CO2 reduction. According to this approach, an electrical machine able to work in both motor and generator modes, with power levels to 10 or 15 kW [12,17,21], can implement several "green" functions, aimed at fuel saving and/or a reduction of pollutant emissions. Among these "green" functions, it is worth mentioning the following: start-and-stop to cut ICE emissions in urban scenarios, torque assistance to improve efficiency at low-rpm regimes where an ICE is not working in optimal conditions, and regenerative braking to avoid just dissipating energy when a deceleration is needed. Besides assisting with an electric machine, the ICE for the propulsion, a saving of energy and of pollution emissions can be achieved through the electrifications of comfort and/or chassis control functions. The latter include, as an example, the electrically controlled air-conditioning compressors or the electric steering.

In this context, a belt-driven starter generator (BSG), or an integrated starter generator (ISG), replaces the conventional alternator, thus creating a parallel hybrid architecture. Figure 6 shows the layout of a BSG. The main idea is that the starter-generator electrical machines replaces the conventional alternator. In this way, there is a low impact on the layout of the engine compartment, and just the redesign of the belt tensioner is needed. Differently from the BSG solution, in the ISG solution, the starter-generator electric machine should be inserted between the ICE and the gearbox. Hence, the ISG approach requires a complete revision of the engine compartment. This is why the authors focused their work, as reported in this survey paper, on the design of a BSG electrical machine.

**Figure 6.** 48 V belt-driven starter generator (BSG) interfaced to the electronics (**left**) and to the mechanical-transmission (MT) and internal combustion engine (ICE) (**right**).

For hybrid vehicles using BSG or ISG machines, the classic 12 V automotive DC bus reaches its limit as follows: at 12 V, the supplied current would be above 600 A in the case of cranking (e.g., peak current at cars start). This will require cables with a high cross section, with a too high increase in wiring cost and size. To solve this issue, there is a trend toward the adoption of a 48 V DC bus technology (52 V in generator mode) for hybrid vehicles. With respect to a classic 12 V solution, the drawn current at equivalent power is reduced by a factor of four. As DC voltages below 60 V do not require electrical shock protection, adopting a 48 V DC bus does not increase the cost of the on-board electrical implant by too much. In the short term, to reduce the cost of migrating from ICE-based vehicle generation using 12 V electrical systems to a new 48 V HEV generation, most of the 12 V automotive components will be reused. As a consequence, new 48 V HEVs require the development of energy-efficient and compact DC/DC converters [20,21]. These DC/DC converters should be able to interconnect the 48 V DC domain to the low voltage supply domains required by low-power components (such as ECUs and memories and sensors). In the case of a fault and/or malfunction, the 48 V domain should be isolated from the low-voltage domain. In emerging 48 V HEVs, another issue is the design of an integrated H-bridge for rotor excitation. The integration is important for reducing the use of discrete devices, thus reducing the size, weight, and cost. In summary, the main issues of hybrid vs. ICE vehicles are the introduction of (i) BSG/ISG machines and related power drives; (ii) 48 V architecture; and (iii) DC/DC converters among the 48 V and 12 V buses to reuse most of the 12 V components.

Figure 6 shows the vehicle electrical architecture for a hybrid vehicle, including a BSG or ISG unit. The BSG or ISG unit is a mechatronic sub-system; it receives the electric power from the battery pack (at 48 V in Figure 6) plus proper command/configuration signals from the vehicle control unit (VCU). The VCU is interfaced also to local sensors, including the ignition signal, and through the main car bus it is connected to all of the other ECUs. At the state-of-art, the controller area network (CAN) is the de facto standard to connect vehicle ECUs with smart actuator and sensing units. As it will be discussed in Section 5, the fact that the BSG or ISG machine (providing torque assistance, start and stop, and electric energy generation functionalities) is connected to the CAN poses several cybersecurity issues to be addressed, as they can lead to safety issues. The core of the BSG proposed in our research work is a six-phase wounded rotor synchronous machine, showed with its dedicated electronics in Figure 7. As we detail in the literature [21], the architecture of the proposed BSG electrical machine includes a wounded rotor and a double three-phase stator. The prototype of the BSG electrical machine is shown in Figure 7. The power level of the proposed BSG electrical machine is up to 8 kW in operating conditions, with a peak of 15 kW. The adoption of two stators instead of a classic solution with two stators allows for the use of lower current levels in each stator and of smaller copper winding. The two stators in the BSG of Figure 7 are electrically shifted by 30 degrees. This solution allows for reducing the ripple on the output rectification. The new BSG electrical machine of Figure 7, within the EU project ATHENIS-3D with the Valeo company, has been integrated and tested on a commercial car—a Peugeot308 hybrid [21]. The proposed drive reaches a max speed of 20,000 rpm (no load condition),

and a max torque at a low-rpm of 70 Nm, and is able to start a 1.5 L diesel engine in less than 400 ms, going from 0 to more than 1000 rpm.

**Figure 7.** Prototype of the 48 V BSG plus control and sensing electronics.

The electronics of the BSG electrical machine in Figure 6 include the following two main power sections:


The control electronics of the BSG electrical machine include the following two power domains:


Another innovation discussed in this work is the use of direct copper bonding (DCB) technology for the 48 V H-bridge. The aim of the DCB usage is the reduction of the ON-resistance ("drain-source on

resistance", is the total resistance between the drain and source in a Metal Oxide Field Effect Transistor, or MOSFET when the MOSFET is "on") of the H-bridge switches. Indeed, low ON-resistance values are needed to minimize power dissipation when the current is flowing through the H-bridge switch. The ON-resistance is due not only to the RDSON of the power MOSFETs (that can be reduced by increasing the silicon area), but also to the resistance of the connections between the power MOSFETs and the electronic board on which the devices are assembled. To achieve low ohmic transistor connections in the 48 V H-bridge, DCB technology can be adopted when designing the electronic board and assembling the devices. In the DCB approach, an interconnect plate with Al2O3 ceramic substrate and electroplated copper is used. The thickness of the electronic board is 0.38 mm. The thickness of the copper is about 0.3 mm, with a surface plating for die stacking with a 5 μm Ni and 0.03 μm Au layer. Each transistor chip is connected with four lines (two source lines and two drain lines) to the electronic board. A stress analysis has been carried out within the ATHENIS-3D EU project [21] with an AMS partner on a 48 V H-bridge, using transistors implemented in High Voltage Power MOSFET (HV-MOS) technology, with a DCB connection to the board. This stress analysis has demonstrated that the DCB technology can allow for operations at high temperatures (up to 175 ◦C).

The supply voltage of the H-bridge, used in Figure 7, is nominal 48 V (in generator mode the voltage is 52 V). The rotor currents are up to 12 A in nominal conditions and up to 17 A in transient conditions. The four transistors of the H-bridge are realized in AMS HV-MOS technology, with DCB on ceramic substrate, as two P-channel MOSFETs for the high side devices and two N-channel MOSFETs for the low side devices. Using P-channel MOSFETs as high side devices in the H-bridge allows for avoiding the use of complex charge-pump devices, although at the price of a higher ON-resistance vs. N-channel MOSFETs. The minimum ON-resistance is 8 mΩ for N-channel MOSFETs and 11 mΩ for P-channel MOSFETs when usinga5V gate-source supply for the transistors. Table 1 shows a comparison of the performance parameters of the integrated H-bridge vs. the state-of-art of the integrated devices for rotor coil driving. With reference to the H-bridge in the literature [33], implemented using the same HV-MOS technology, the solution proposed in this work sustains a current 2.125 times higher and a voltage 4 times higher, whereas the ON-resistance is reduced by 6 times. The improved performance of the H-bridge proposed in this work vs. that in the literature [33] (×2.125 higher current, ×4 higher voltage, and ×6 lower ON-resistance) is justified by both the increase of the transistor area and by the adoption of the DCB approach, missing in the literature [33]. Thanks to the DCB technology, the transistors are assembled on top of the ceramic substrate. The DCB technology also allows for a reduction of the printed circuit board (PCB) size with respect to the layouts of electronic boards, where large conductive plates are placed around the chip to achieve low-ohmic contacts, as in the literature [33]. Since in this work, the electronics is coupled with the 48 V BSG machine, and the adopted technology is a low-cost one, and the DCB approach allows for a reduction of the PCB size, then the increased area of the transistors is not an issue.


**Table 1.** 48 V H-bridge vs. state-of-art, rotor current control.

#### **4. Full Electric Vehicles**

In recent years, EVs or BEVs have been gaining popularity, and one of the most important reason behind this is their contribution to reducing greenhouse gas (GHG) emissions.

EVs, including the different types of BEVs, HEVs, PHEVs, and fuel-cell EVs (FCEVs), have been becoming more commonplace in the transportation sector in the last five years. A block diagram with the main EV's components is depicted in Figure 8. Comparing this figure for EV with that of Figure 6 (right), it is clear how full EV needs a major change in current car architectures with lot of new electronic and electrical sub-systems. This is why many leading OEM's and suppliers are developing 48 Volt architecture, as discussed in Section 3, on the basis that it can achieve large efficiency gains at lower costs in the medium term rather than full electrification, as discussed in this section.

**Figure 8.** Car's block diagram with the main EV's components.

The main energy source of BEV is the high voltage battery, while the most important components of the EV are the electrical motors and the high voltage battery, which, together with the transmission system, builds the vehicle drivetrain architecture. These main components are assisted by a number of auxiliary subsystems, such as an ECU, a battery management system (BMS), and power electronics converters. An ECU, which can be called "e-motor control", is liable for the electric motors operation mode. BMS, also called the battery monitoring and control unit (BMCU), supervises the storage operations, including charging and discharging states. Most EVs contain on-board single/phase or three/phase unidirectional or bidirectional chargers; hence, an AC/DC power electronic converter is required. Power electronic converters, for the charging operation mode are able to transform energy from AC to DC, while for the discharging operation mode, they should be able to transform energy from DC to AC. At the same time, the EV system operation requires other energy conversions in order to supply other subsystems. Therefore, several additional converters are needed, such as the following:


EVs can produce significant impacts on the environment, on the network for power system distribution, and on other related sectors. The actual power system could face huge instabilities with enough EV penetration, but with a suitable management and coordination system, EVs can be turned into a major contributor to the successful implementation of the smart grid concept.
