Hybrid Energy Storage Systems in Electric Vehicle Applications

Abstract

This chapter presents hybrid energy storage systems for electric vehicles. It briefly reviews the different electrochemical energy storage technologies, highlighting their pros and cons. After that, the reason for hybridization appears: one device can be used for delivering high power and another one for having high energy density, thus large autonomy. Different energy storage devices should be interconnected in a way that guarantees the proper and safe operation of the vehicle and achieves some benefits in comparison with the single device storage system source. The chapter shows different topologies for interconnecting electrochemical technologies: passive, semi-active, and full-active, clarifying their benefits and drawbacks. The chapter concludes with a case study, an electric motorcycle, which is ridded using an urban profile. There, the hybridization was performed to extend its cycle life.

1. Introduction

Electrical vehicles require energy and power for achieving large autonomy and fast reaction. Currently, there are several types of electric cars in the market using different types of technologies such as Lithium-ion [1], NaS [2] and NiMH (particularly in hybrid vehicles such as Toyota Prius [3]). However, in case of full electric vehicle, Lithium-ion technology is used widely in automobiles, scooters, motorcycles, and busses [4]. It is known that the aging of a battery and its capacity is dependent on the type of use such as the current profile and the depth of discharge [5]. The deeper the battery is discharged and the higher the currents are, the smaller is the number of cycles for the battery. Thus, combining batteries with other energy sources, which can tolerate high currents, deep discharges, and high number of cycles, can reduce the use of the battery in “non-favorable” conditions. Thus, the combination of a supercapacitor (SC) with a battery can lead to longer cycle life of the battery [6].

However, to achieve this goal, a control system needs to select which source of energy should be used in real time. Thus, high complexity is added in the energy storage system (ESS). For that purpose, some DC/DC converter topologies are used with a controller that selects the energy source for each instant [4, 7]. To understand the advantages of hybrid energy storage systems (HESS), it is important to review the available energy sources.

2. Review of energy sources

There are many technologies for storing energy, most of them in an electrochemical way. Different technologies have different characteristics such as available power density, energy density, number of cycles, temperature range, calendar aging, and others. One of the most common diagrams for comparing energy storage technologies is the specific energy versus specific power plot, see Figure 1 [8]. There, different technologies are presented in a two-axis plot that clearly shows which technology is more useful for delivering power and which one is better for storing energy. Therefore, combining a high energy density technology with a high power density technology, an energy storage system that fits well in an electric vehicle can be achieved.

Figure 1.

Specific energy vs. specific power for different storage techniques [8].

In addition to those important indexes, the number of cycles is also important. In general, we can mention four main groups of storage technologies: electrochemical, electrostatic, electric double layer technology, and mechanical. There are also other types of storage technologies such as hydrogen fuel cells [9] and magnetic superconductors [10], but these are not covered in this chapter.

2.1 Electrostatic technologies

From Figure 1, notice that the devices that can deliver the highest specific power are the electrostatic devices; their limit is their internal resistances. Electrostatic devices include electrolytic capacitors, polyester, and polypropylene capacitors [11]. However the energy density is quite low. The number of cycles is very high and the temperature range as well. Commonly, electrolytic capacitors are used in power electronics for voltage sags in grid-connected power supplies; thus, the output power is not compromised during a temporary energy interruption in the milliseconds range. Polyester and polypropylene capacitors have much smaller internal resistance, so they can deal with higher powers, but the specific energy is extremely low; their use is in the micro to milliseconds range, usually inside power electronic converters [12].

2.2 Electrical double layer capacitors

The next solution is a hybrid between a “capacitor” and an electrochemical cell, the electrical double layer capacitor, supercapacitor, or ultracapacitor. This device has a notable bigger energy density and can deliver high power; two mechanisms can work inside the device, electrostatic and electrochemical processes. Depending on the manufacturer and the applications, one or the other process is dominant. Another advantage is that the number of cycles is much bigger than the traditional electrochemical cells. Manufacturers mention more than 1,000,000 cycles.

Supercapacitors can be understood as an electrochemical cell in which the redox reactions are not present. The cell has two electrodes, an electrolyte and a separator. There is not mass transfer between the electrodes and the electrolyte; thus, only a charge distribution appears. That charge distribution can be considered as a parallel plate capacitor in which one plate is the electrode and the other is the electrolyte. The “dielectric” is the distance between the electrolyte and the electrode. It is very small; this is why it creates a huge capacitance [13]. In addition, the porosity of the electrodes creates a huge specific area, which also maximizes the capacitance effect. The huge capacitance appears in the interface between one electrode and one electrolyte; thus, the full device has two huge capacitors in series. The separator allows the flow of ions and avoids the short circuit between the electrodes.

These devices have impressive capacitances, commercially available around 3500F, but as the distance between plates is very small, the maximum voltage of the device is small as well. For aqueous electrolyte, it is around 1 V, and for organic electrolytes, it can reach around 3.5 V. As the mechanism to store energy remains electrostatic, the number of cycles is still very high.

In addition to the electrostatic effect, there are other supercapacitors that also have faradic processes (redox reactions and other types of reactions), and due to that, they can achieve even a higher capacitance. However, by including this process, the number of cycles is reduced. This type of supercapacitors is sometimes called hybrid supercapacitors.

2.3 Electrochemical technologies

In this group, we can mention Lithium-ion, NiCd, Lead-acid, NiMH, NaS, and other electrochemical technologies that store their energy based on redox reactions. This group has high energy density, moderate number of cycles (a few thousands, depending on the specific design and use), and moderate temperature range. As an example, Table 1 shows some of the general characteristics of the most common technologies [19].

Technology	Voltage (V)	Energy density (Wh/dm^3)	Power density (kW/dm^3)
Lead-acid	2.0	50–80 [14, 15]	10–400
NiCd	1.2	60–150 [16]	150–300
NiMH	1.2	220–250 [17]	∼900
Li-ion	4.1	200–500 [18]	500–2000
NaS	2.0	150–250 [14, 18]	150–230

Table 1.

Electrochemical technologies for electric vehicles.

2.4 Other technologies

Regarding other ways of storing energy, mechanical storage devices were used in the past for massive storage. The most popular devices are: flywheels [20], which store energy in a kinetic fashion; compressed air energy storage (CAES); which storage energy by compressing and releasing the air; and water pumps, which store potential energy. Another technology is the redox flow batteries [21], which can be use of long periods of time, but its energy density is much lower than Lithium-ion batteries. All of these technologies are not suitable for vehicles, and they are mostly applicable in the electric grid.

Hydrogen fuel cells are also an interesting energy storage system that can fit in the electric vehicle technology and can be hybridized using an auxiliary energy storage such as lithium-ion or supercapacitors.

3. Hybrid energy storage systems (HESS)

There are several reasons for using a hybrid energy storage system instead of a single technology storage system (here, Battery Energy Storage System, BESS). All of them are related to the power sharing between a device that mainly stores energy and a device that mainly delivers power. There are several main benefits of power sharing:

• If the energy storage device (battery) delivers less instantaneous power (or current), the temperature is kept in safe operation area, which extends lifetime.

• If the energy storage device (battery) delivers less power (or current), it is expected that the extracted energy is larger. The less demanded power, the higher amount of energy that can be extracted from the battery. The device behaves more efficiently.

• Using an external device that can deliver more power, a new power limit can be achieved, higher than the one of the single energy storage device.

Therefore, with the aim of reducing the stress in the main battery, an auxiliary energy source is added, which creates an hybrid energy storage system (HESS). Thus, high currents can be shared, and the battery use is reduced, with the corresponding increase in life cycle.

3.1 Main topologies

Different topologies exist in order to connect two or more energy sources. They can be defined in terms of three main groups:

• passive HESS

• semi-active HESS

• full-active HESS.

In the passive HESS topology, SCs and batteries are in parallel and connected directly to the load. It is a simple and low-cost topology, but the SCs’ contribution is poor [7]. The SC delivers energy only if its terminal voltage varies. Therefore, connecting the SC in parallel to the battery limits the voltage variations, thus the contribution is limited. This topology has been used for transient suppression under high current pulses [22]. The circuit topology is shown in Figure 2a.

Figure 2.

Topologies of hybrid energy storage systems: (a) passive, (b) B-HESS semi-active, (c) SC-HESS semi-active, (d) full-active using multiple DC/DC converters, (e) full-active using multiport converters.

Semi-active HESS topologies allow one of the energy storage devices, the battery or the SC-stack, to be controlled through a DC/DC converter, while the other storage device is directly connected to the load without any control [23].

On the one hand, if the SC-stack is the device under control (SC-HESS), the DC/DC converter should deliver high power and be fast enough to react to power pulses [24]. Otherwise, the battery would have to respond to the load, and no advantage would be obtained from the HESS. As a benefit, if the battery is connected directly to the motor drive, the voltage in the output port of the SC-HESS is almost constant because the battery voltage profile is quite flat. Lithium-ion batteries in particular have a very flat profile if the depth of discharge is lower than 70%. The circuit topology is shown in Figure 2b.

On the other hand, if the battery is the device under control (B-HESS), which is the topology shown in Figure 2c, the SCs have a direct connection to the motor drive, so they can react very quickly. The current from the battery can be controlled in order to keep the SCs charged while the battery has a smooth discharging profile, independent of the load profile [25]. Thus, the battery is protected. The main disadvantage is that the motor drive voltage (or SC-stack voltage) is not constant. This is because the SCs’ voltage is proportional to the stored electric charge and it changes as the current flows to the load. In order to allow the SCs to interact with the load and to deliver or absorb current, a voltage variation in the HESS output port must be allowed. The bigger the voltage variation, the bigger the current contribution from the SCs.

The full-active HESS solves the drawbacks of both semi-active topologies. Two main topologies exist: (a) parallel DC/DC converters topology [26] and (b) multiport DC/DC converters [27, 28] as shown in Figure 2d and e, where both the battery and the SCs are connected through a DC/DC converter to an output DC link that is connected to the load [25]. The SC-stack uses its entire operating voltage range; no high power pulses are demanded from the battery because its current is well controlled, and the DC link voltage can be correctly regulated through the DC/DC converters. However, the system is much more complex and efficiency is expected to be lower because the power is transferred always through DC/DC converters.

Semi-active and full-active systems are the most studied HESSs because they control the power that is delivered to the load and the SCs have a higher impact on the storage system compared to the passive topology. They are mainly used in photovoltaic, grid, and electric vehicle applications [24, 25, 26]. HESSs have been used for extending the power transfer capabilities of EV propulsion systems [25].

4. Case study: electro motorcycle

As an example of hybrid energy storage system for electric vehicle applications, a combination between supercapacitors and batteries is detailed in this section. The aim is to extend the battery lifetime by delivering high power using supercapacitors while the main battery is delivering the mean power.

HESSs have been used for extending the power transfer capabilities of EV propulsion systems [13, 19]. In [14] a control algorithm for a full-active HESS is proposed in which either the battery or the SC-stack is selected as energy source according to the frequency spectrum of the demanding current. A more detailed analysis was performed in [9], where the control algorithm also considers the extreme cases when the SCs are out of energy. As a result, these works presented an improvement in the power response capability of the system and a reduction in the current peaks demanded from the battery.

4.1 Aging, thermal, and electric model for the battery

For understanding how the battery lifetime can be extended, aging models have been introduced in the past. Among aging models for batteries, [29] presents a degradation model that considers the depth of discharge in cycles and temperature. The model is based on crack propagation theory [30]. It predicts the capacity reduction and proposes an equivalent electric model that is, modified according to the age of the battery, so it combines the prediction in capacity reduction (Ah) with electric equivalent circuit. The model, without considering the calendar aging, is:

$$∆L=k_{\mathit{co}}.N_e.e^{\frac{{\sigma }_{\mathit{SoC}}-1}{k_{\mathit{ex}}}.\frac{T_N}{T_a}}.e^{k_{\mathit{soc}}\frac{\overline{\mathit{SoC}}-0.5}{0.25}}.e^{k_T\frac{T-T_a}{\frac{T_a}{T_N}}}.\left(1-L_0\right)$$E1

where L is the capacity reduction starting from 0 to 0.2 C when the battery capacity (C) is reduced 20%, which is normally considered as the end of the life time, and L₀ is the initial capacity degradation (L₀ = 1-C₀) during the time interval where ΔL is been computed. K_co, k_ex, k_soc, and k_T are the parameters of the model and need to be fitted using data. $\overline{\mathit{SoC}}$ and${\sigma }_{\mathit{SoC}}$ are the average and normalized standard deviation of the state of charge in the interval. For example, ${\sigma }_{\mathit{SoC}}=1$ and $\overline{\mathit{SoC}}=0.5$ are the values for a time interval that includes a full charge-discharge cycle. N_e is the number of equivalent full charge-discharge cycles during the time interval. T, T_a, and T_N are the battery temperature, ambient temperature, and nominal temperature (298 K) in K. Using (1), the battery capacity degradation, L, can be calculated at any moment: L = L₀ + ΔL.

This model expresses that the aging depends on the number of equal cycles, the previous aging (1-L₀), depth of discharge within a cycle, and temperature. However, for the hybrid energy storage system, the main control parameter is the battery current, particularly, how smooth the discharge current is. This means that what the ratio between RMS and average value is during the cycle. The smaller is the ratio, the smoother is the current.

Unfortunately, this ratio is not explicitly shown in the model. This ratio affects the battery temperature, T, and then, the temperature has an impact in the aging. In order to understand the impact, a relation between the internal temperature of the cell and the RMS/average ratio is needed. A simple model is used as shown in Figure 3a and b shows the electric circuit. The thermal model is as simple as possible: a thermal capacitance, related to the materials and battery size, and a thermal resistance, which allows the battery to extract the heat. The electric circuit is also simplified as the main goal is not to accurately predict the output, but to estimate the losses.

Figure 3.

(a) Thermal and (b) electrical equivalent circuits.

Figure 4 shows two types of discharge-charge cycle and the temperature effect of it. The constant discharge-discharge cycle has a notable less impact in the temperature although both of them discharge at 1C rate. The variable discharge-charge cycle has a higher temperature impact, which is reflected in the aging. The aging plot at the end of the figure shows the effect of cycling both constant and variable cycles, considering and neglecting the temperature effect. In both cases, the temperature effect notably increases the degradation of the battery, and the variable discharge process achieves the maximum degradation. In addition, the figure also presents the effect of the degradation only because of the depth of discharge and the current. These parameters also play an important role in the degradation of the battery.

Figure 4.

Example of impact of the temperature in battery capacity degradation.

4.2 Motorcycle description, battery current profile with and without HESS

Figure 5 shows the main diagram of the electric motorcycle traction system. In the example, the main characteristics are: in-wheel 5 kW brushless motor, a total weight of 200 kg, a 70 V 50 Ah battery, and a motor drive that can work in the 60 to 100 V range. The battery consists of 22 Lithium-ion 50 Ah cells (LiFePO4) in series.

Figure 5.

Motorcycle traction system.

In the BESS case (only batteries), the battery is connected directly to the motor drive and delivers the current demanded by the motor. For this case study, we propose the use of B-HESS case (Figure 2b); the battery is connected through a DC/DC converter to a SC-stack, and both then connect to the motor drive. The semi-active B-HESS was selected because:

1. Passive topologies are robust but cannot take full advantage of the SCs because the power sharing is controlled based on their own internal impedances, so no real-time control.

2. Full-active topologies are attractive, but size, cost, and complexity are above a scooter design and price.

3. Semi-active topologies offer two variants: with SC-HESS, the converter should tolerate SC currents, which can be much larger than the battery currents, so it will be more expensive, bulkier, and heavier. On the contrary, B-HESS is a good trade-off between control, complexity, and cost and allows the full control of the battery current, which is important for keeping the temperature within the safe operational area while still delivering high power.

Therefore, B-HESS was selected.

In this configuration, the SC-stack voltage is allowed to fluctuate between 60 and 100 V, in order to transfer energy to the motor drive. The motor drive is a traditional voltage source inverter, which can tolerate that voltage variation. However, if a wider input voltage is needed, other inverters must be used, such as a Z-source converter [31].

The current that the motor drive demands from the battery is shown in Figure 6. Notice that the currents peaks achieve 3C for this particular battery 150A and there is a notable variation in the current from zero to the maximum value. However, by using the B-HESS (Figure 2b) and using the converter as a low pass filter, the high frequency currents can be delivered by the supercapacitors (SCs) and the average current by the Lithium-ion battery. Figure 7 shows the profile of the current in the battery using B-HESS. B-HESS consisted of the same battery, a bidirectional half-bridge converter [7], and a set of 33 SC in series as a SC-stack 3 V-3000F. Notice that the high frequency currents are delivered by the SCs, while the main current is delivered by the battery. Also, notice that for long pulses, the supercapacitors cannot deliver the power, so in that case, the battery is the one that is delivering the power. Therefore, if the SCs are exhausted, the battery still provides the energy. This allows a robust performance.

Figure 6.

Battery discharging profiles for an urban ride in a scooter using only BESS.

Figure 7.

Battery and SC discharging profiles for an urban ride in a scooter using B-HESS.

4.3 Aging estimation

Finally, both the BESS discharging current profile from Figure 6 and the B-HESS discharging current profiles from Figure 7 were cycled using a 1C-constant charging profile for more than 300 cycles. The results are shown in Figure 8 for the only battery case (BESS) and in Figure 9 for the B-HESS case. The rise in the temperature is slightly higher in the BESS case, but the main effect is the current variation, which is much softer in the B-HESS. For both cases, the presented aging model were used with the same parameters. It predicts that the degradation achieves 10% of capacity loss for 900 cycles in the BESS mode; however, for the B-HESS, the capacity loss is 8%. This means around 20% of life cycle extension. In addition, the real battery capacity was measured in each cycle, and the results of the first 300 cycles are in concordance with the proposal.

Figure 8.

Degradation test in only battery case, BESS.

Figure 9.

Degradation test in B-HESS case.

As a future work, a complete analysis reaching the end of the life should be performed for several cells and several profiles. Unfortunately, these tests are very time demanding, so the validations are limited but express are big potential in life cycle increase using HESS technologies.

5. Conclusion

This chapter briefly describes the technologies for storage energy and from that extracts the idea of combining two different technologies in order to have high power available and high energy density. It suggests that a good combination for this is an HESS with supercapacitors and batteries. For that, several topologies are mentioned: passive, semi-active, and full-active, with their own advantages and disadvantages. The most attractive topologies are semi-active and full-active, due to the flexibility that they offer.

Using HESS, combining SC and batteries, current stress and temperature of the battery can be reduced. These two aspects affect the life cycle of the battery, so the chapter also presents an aging model that helps to quantify their effect on the battery life cycle.

Finally, the chapter shows a practical case study, in which the energy storage system of an urban scooter is replaced by an HESS. The study clearly shows how the current in the battery becomes smoother and how this impacts the temperature and the capacity degradation. This is shown using the aging model and experimental validation.

Acknowledgments

This chapter was supported by Skoltech NGP Program.