A Study on Active Balancing Methods by Battery Stack Power Recovery

Abstract

Recently, the demand for energy storage has increased, serial cell configurations are commonly used for high-capacity secondary batteries. Initially, when identical cells are configured in series, the battery cells operate without voltage differences. However, over time, voltage imbalances occur due to differences in the environmental and physical properties of the cells. In addition, the unique characteristics of cells change during charging and discharging, which greatly affects to the degradation of battery efficiency and performance. Various cell balancing techniques have been proposed to achieve high efficiency and performance in battery. This paper proposes a minimum voltage selective balancing topology based on a flyback converter for module balancing of batteries based on active cell balancing techniques. In the proposed topology, the output voltage of each battery is shared through a single transformer, the energy from the converter's primary side is transferred to the battery modules with the minimum voltage. The proposed module balancing circuit can be easily applied to the battery reuse industry. The proposed minimum voltage battery module selective balancing topology was verified through PSIM simulations and experiments.

1. Introduction

Recently, batteries are increasingly being used in automotive applications. In applications requiring high power and high voltage, such as electric vehicles (EVs), hybrid electric vehicles (HEVs), and energy storage systems (ESS), battery connections for cells connected in series and parallel are essential. Individual battery cell voltages vary from 1 V to 4.5 V, and cells are connected in series to increase the voltage level of the battery pack. However, study on battery balancing is being conducted to prevent voltage imbalance in serial cells due to manufacturing and thermal changes, internal impedance differences, and self-discharge rates [1]. The initial battery module has excellent charging/discharging characteristics because the cells have the same capacity. However, in the repeated process, differences in the battery's SoH (State of Health) temperature, other cell compound fluctuations, and differences in initial charge capacity are unavoidable [2]. As shown in Fig. 1, the capacity of the battery is limited in operation time during charging/discharging due to battery cells with reduced capacity, which affects lifespan. If charging continues with the same charging current in a battery voltage imbalance situation, aging batteries reduce efficiency and performance, and also cause battery aging and explosion. Accordingly, a battery management system (BMS) including battery charging and discharging, SoC (State of Charge), SoH charge/discharge cycle, and battery balancing is required to maximize the lifespan and performance of the battery [3-5].

Fig. 1 Decrease in batteries life time

Battery balancing methods can be divided into passive balancing methods and active balancing methods according to whether energy is consumed. The passive balancing method dissipates energy through passive components such as resistors based on a battery with a low voltage, the circuit configuration is relatively simple, and it has the advantage of being easy to control. However, since the energy dissipated from the resistor cannot be reused, so it has the disadvantage of low energy efficiency.

Active balancing has the advantage of being efficient and having excellent dynamic characteristics by moving the energy of a battery with a low voltage to a low battery to achieve an overall voltage balance without unnecessary balancing operations. However, compared to the passive balancing method, circuit configuration and control are complicated, and it has the disadvantage of being uneconomical in configuration [6].

2. Classification of Battery Cell Balancing

2.1 Cell Balancing Technology Trends

Fig. 2 shows the classification for battery balancing. Depending on the energy consumption method, it is broadly divided into passive and active balancing. The main operating principle of the passive method is to consume energy from a high voltage of an overcharged battery using a passive element such as a resistor. when the voltage between batteries is different, the voltage of the higher cell is balanced by consuming energy through a passive device based on the voltage of the lower cell to reduce voltage deviation [7-8].

Fig. 2 Battery balancing topology classification

Active balancing methods are divided into various methods using capacitors and methods using inductors, transformers, and power converters. Active balancing is a method of moving and redistributing energy towards insufficient battery cells without consuming overcharged energy, so that battery energy with a high voltage is moved to a battery with a lower voltage to achieve overall voltage balance [9].

Fig. 3 (a) shows a fixed parallel resistance method. The characteristic of a passive cell balancing topology is that a resistor is connected to each battery in parallel to consume energy in a high voltage battery. This method reduces voltage deviations between cells by consuming energy through resistors when the voltage of an overcharged cell is higher based on a low voltage battery. However, this method makes it difficult to fully adjust the voltage between batteries, and is not economical because a lot of energy is consumed through resistors during the balancing process. Also, the heat generated by the resistor can cause the system to overheat.

Fig. 3 Passive and active balancing of batteries

Fig. 3(b) is a switch capacitor method. when the number of cells is N, N battery cells are composed of switches and N-1 capacitors. The switch at this time is synchronized and controlled by a single control signal. The structure of the switch capacitor topology has a structure in which a single capacitor is continuously combined between two batteries, and the battery's capacitor is responsible for energy transfer [11-12].

2.2 Magnetic Flux Shared Battery Active Balancing

The magnetic flux shared battery balancing topology structure is a structure in which N batteries are connected in series to a single transformer having N windings. In the method of sharing magnetic flux, when the voltage of a battery is measured, the voltage can be estimated only by calculating the number of modules. Using a single voltage sensor can have a comparative advantage in circuit configuration and cost compared to previous methods where sensors are used for as many batteries.

Among isolated converters, the circuit structure is simple, and it has the advantage of being able to generate multiple outputs through a single transformer. Since voltage boosting and decompression are flexible by adjusting the winding ratio of the transformer, stable and efficient power conversion is possible while meeting various voltage requirements, and it can be manufactured at a relatively low cost, so it is widely used as an economical solution.

MODE 1 starts from the moment switch Q is turned on, and it is a section where energy is stored in the primary coil due to the input voltage.

Fig. 4 Flyback converter operation mode

At this time, an input voltage is applied from the primary side of the transformer, but there is no power transmission to the secondary side, and current is formed through the primary side leakage inductor and primary inductor.

The primary side current of the transformer is the same as equation (1).

\(\begin{align}i_{p}=\frac{V_{i n}}{L_{1}+L_{M}} t\end{align}\)       (1)

MODE 2 transfers energy stored in the primary transformer female inductor to the secondary side when the switch is off, and the voltage equations for the input and output voltages and the current on the primary and secondary sides of the transformer are as shown in equation (2).

\(\begin{align}\left[\begin{array}{c}V_{i n} \\ a V_{B}\end{array}\right]=\left[\begin{array}{cc}-\left(L_{1}+L_{M}\right) & -L_{M} \\ -L_{M} & -\left(a^{2} L_{2}+L_{M}\right)\end{array}\right]\left[\begin{array}{l}\dot{i_{p}} \\ \dot{i_{2}}\end{array}\right]\end{align}\)       (2)

It is a current type converter that controls the output voltage due to circulating current, and information on current damage is required for circuit analysis, and it is the same as equation (3), which takes the inverse matrix using equation (2) in the proposed converter.

\(\begin{align} \begin{array}{l} \left[\begin{array}{l} i_{p} \\ i_{s} \end{array}\right]=\frac{1}{A}\left[\begin{array}{cc} -\left(a^{2} L_{2}+L_{M}\right) & L_{M} \\ L_{M} & -\left(L_{1}+L_{M}\right) \end{array}\right]\left[\begin{array}{c} V_{i n} \\ \text { a Vbat_min } \end{array}\right]\\ \text {here}, A=a^{2} L_{2}\left(L_{1}+L_{M}\right)+L_{1} L_{M} \end{array} \end{align}\)       (3)

The current solution for the primary side and the secondary side through the equation of state is the same as equations (4) and (5).

\(\begin{align} \begin{array}{l} i_{p}=\frac{a L_{M} V_{\text {bat_min }}}{A}\left(t-t_{1}\right)-\frac{\left(a^{2} L_{2}+L_{M}\right) V_{i n}}{A}\left(t-t_{1}\right)+i_{p}\left(t_{1}\right)\\ \text {here}, A=a^{2} L_{2}\left(L_{1}+L_{M}\right)+L_{1} L_{M} \end{array} \end{align}\)       (4)

\(\begin{align}\begin{array}{l} i_{s}=\frac{a L_{M} V_{i n}}{A}\left(t-t_{1}\right)-\frac{\left(a^{2} L_{2}+L_{M}\right) V_{\text {bat_min }}}{A}\left(t-t_{1}\right)\\ \text {here}, A=a^{2} L_{2}\left(L_{1}+L_{M}\right)+L_{1} L_{M} \end{array}\end{align}\)       (5)

Fig. 5 Proposed flyback converter cell balancing topology

3. Proposed System Standby Power Reduction Circuit

3.1 System Idle Mode

Fig. 6 shows the structure of the Idle mode of the balancing system. A switch for input power input is added to reduce standby power, and the enable switch signal takes a structure that is generated at 2 locations, and when balancing is required in the upper BMS, the SMPS is operated with the pulse signal of the module balancing circuit. The DSP operates through the operation of the SMPS, and the DSP generates an operation signal of the SMPS at this time. It is a structure that measures the voltage of each battery, and when balancing is completed, the DSP generates an SMPS Disable signal and terminates the operation.

Fig. 6 System idle mode circuit​​​​​​​

In the operation of a minimum voltage selective battery balancing system with an SMPS Idle mode, DSP is operated by the SMPS power source when the SMPS Idle mode is disabled from the upper controller, and the balancing control mode can be selected between CC mode and CC-CV mode. Modbus communication was used to enable state transmission to the upper controller, and when the balancing operation was completed, the DSP generated an Idle mode signal to the SMPS to reduce standby power.

A battery balancing circuit that reduces system standby power can greatly reduce the risk of battery discharge due to self-discharge when battery energy is not used for a long period of time. Also, the proposed SMPS Idle method can be easily configured when combined with a commercially available SMPS dedicated chip.

As a result of using a method combined with the LM2576, a commonly used SMPS-dedicated chip, the standby power was reduced to 30 [uA].

Fig. 7 Proposed battery balancing circuit​​​​​​​

4. Performance Simulation and Review

4.1 Simulation

Fig. 8 is a simulation circuit diagram for verifying the validity of the minimum voltage battery module selective balancing topology using the proposed flyback converter. The simulation circuit was configured for 4 batteries, an RCD snubber circuit was configured to prevent overvoltage, and the duty ratio was set to 0.35 and 20 [kHz].

Fig. 8 Battery balancing PSIM simulation circuit​​​​​​​

Fig. 9 shows a battery voltage and current simulation waveform. The voltage for each module was set to Vbat1 = 6 [V], Vbat2 = 10 [V], Vbat3 = 14 [V], and Vbat4 = 18 [V].

Fig. 9 Analysis of battery voltage and current simulation​​​​​​​

As a result of the simulation, the balancing operation is completed at about 0.8 [sec] of the module voltage, and the operation sequence is as follows. Since the voltage of module 1 is minimal during initial startup, the output of the flyback converter forms current only on the diode D1 to raise the voltage of module 1. As a result, the remaining modules 2, 3, and 4 transfer power for balancing to the flyback converter, and the voltage drops. In about 0.1 seconds, the voltages of module 1 and module 2 become the same, and the flyback converter supplies power through D1 and D2 to increase the voltage of the two modules equally. At approximately 0.35 [sec], the voltages of module 1, module 2, and module 3 become the same, and the flyback converter supplies power through each diode D1, D2, and D3 to increase the voltage of the three modules equally. At approximately 0.8 [sec], the voltages of all modules 1, 2, 3, and 4 are balanced, and the balancing operation ends.

4.2 Experiments

Fig. 10 establishes an experimental environment to determine the dynamic characteristics and validity of the prototype performance of the proposed paper. This system consists of 5 super capacitors that replaced batteries, and TMS320F280025 was used as the DSP for PWM control. IC LM2576 for SMPS was used to implement the Idle mode of the proposed system.

Table 1. Battery balancing system parameter​​​​​​​

Fig. 10 Experimental environment configuration​​​​​​​

Fig. 11 shows that the test was conducted from 0.1 to 0.3 with the test result waveform according to Duty of the proposed minimum voltage module selective balancing, and it was confirmed that when the duty ratio was set to 0.1, 0.2, and 0.3, respectively, it converged at 180 [sec], 120 [sec], and 40 [sec].

Fig. 11 Balancing analysis waveform​​​​​​​

Fig. 12 (a) shows the voltage of the battery during the battery balancing operation and the output current waveform of the flyback converter connected to each battery. In the case of a battery with a low voltage, energy is supplied through a flyback converter and the voltage rises. In a battery with a high voltage, the output current of the flyback converter is not supplied as 0, but since the battery is used as an input power source for balancing, the voltage is lowered. Finally, it can be confirmed that balancing is performed by repeating this operation until the voltage of all batteries is the same.

Fig. 12 Input current waveform due to module voltage​​​​​​​

Fig. 12 (b) shows the voltage fluctuation characteristics and the primary side current waveform of the flyback converter. It can be confirmed by the battery side current and the primary side current of the flyback converter that the balancing operation ends as the battery voltage converges to the target value through the waveform.

At this time, the battery balancing efficiency was found to be approximately 78.6%.

5. Conclusion

In this paper, a balancing topology was proposed using a magnetic flux type flyback converter to eliminate each battery voltage imbalance. The proposed method verified dynamic characteristics and validity through PSIM simulations to verify the validity of modular battery balancing, and the following conclusions were drawn by designing a prototype of the proposed paper. A power cycle method that uses the output of a serial module as a power source for balancing has superior characteristics compared to a method of supplying power from a specific module by driving a balancing circuit using the same current of each module. Also, since the voltage of the battery module is not measured individually, the design is relatively simple and economical due to the relatively simple circuit configuration and simple control algorithm. Also, when applying the Idle mode of the system, the balancing operation ends, and after a certain period of time, it enters standby power mode to reduce power and improve efficiency. It is believed that it can be easily applied to systems that require large capacity such as ESS or in the reuse industry of aging batteries.