What is the equalization circuit of a lithium-ion battery, what is its function?
The EVERLASTING is a project funded by the European Union's Horizon 2020 Research and Innovation Program to improve the reliability, longevity and safety of lithium-ion batteries by developing more accurate and standardized battery monitoring and management systems. Predicting battery behavior over the life cycle enables positive and effective battery management.
The first article translated by Professor Bao introduces cell balance, especially how to evaluate the equilibrium strategy. Firstly, we discuss what kind of cell balancing is really needed. Secondly, we define different equalization strategies based on the application scenarios of energy storage. Finally, we propose a balancing strategy suitable for most scenarios.
1. The root cause of imbalance of batteries
Due to the decentralized energy infrastructure (fixed energy storage equipment) and the growing environmentally friendly travel tools (electric vehicles), reliable and economical energy storage solutions are needed. Lithium-ion batteries show the potential to meet these demands due to their high energy density. However, electric vehicles require not only high energy density but also high overall energy (large capacity battery pack). In order to meet these needs, the cells are composed of battery packs in series and in parallel, and there may be as many as several thousand cells in a battery pack (for example, 18650 or 21700 cells used by TESLA). The battery pack voltage is related to tolerable losses and power electronics design and usually determines the number of cells in series. On the other hand, the number of parallel cells depends on the energy and power requirements of the entire battery system.
The characteristics of the cells are slightly different due to manufacturing tolerances such as variations in electrode thickness and overall component connectivity. Due to the limited manufacturing precision, even the same batch of cells has different initial capacities and impedances. These parameter deviations are Gaussian (normal) distribution. Different cell capacities and impedances mean that in a series connection there is always one cell or multiple cell blocks (multiple cells in parallel) that first reach the end of charge or end of discharge voltage. Considering safety factors, these first reach the end of charging or the end of discharge voltage limit cannot be exceeded, so that the capacity of other cells cannot be fully utilized. In addition, different self-discharge and aging rates cause differences in cell voltage due to changes in internal parameters of the cell or the presence of temperature gradients. This difference in cell voltage results in a further premature limitation of the capacity of the battery pack. In order to avoid possible capacity limitations, equalization circuits are often required in battery systems.
In general, cell imbalance is related to cell quality. Cell quality includes changes in initial cell parameters and aging behavior under the same conditions, battery system quality, and especially thermal management system quality. If the exact same cell can be achieved and there is no temperature gradient inside the battery system, the equalization system will not be needed, and then this is an impossible task.
2, the goal of cell balance
The goal of cell balancing depends on the application scenario of the battery pack. The goal of electric vehicles is to achieve the maximum possible cruising range, while the goal of a fixed battery pack that participates in grid control is to provide the required power at any time.
In order to maximize the energy content of the battery pack for electric vehicles, it is necessary to make full use of the energy of each battery core. In the case of complete discharge, although the capacity and impedance of each battery core are different, it must be realized. The 100% SOC begins to discharge to 0% SOC. For fixed energy storage involved in grid control, the cell or cell block in the weakest block of the barrel effect needs to be at a fixed SOC to allow the positive and negative current pulses to be released and absorbed within the specified time (frequency modulation application) ). Since the application of electric vehicles is ahead of grid energy storage applications, maximizing the energy content of batteries is the primary goal of cell balancing.
3. Evaluation of battery balance algorithm
In the case of imbalance, not all cell capacity is fully utilized, and the remaining energy must be redistributed through the equalization circuit. The equalization circuit is typically capable of adjusting the energy level of a single cell or cell block. There are generally two equalization systems: a dissipative equalization system (passive equalization) and a non-dissipative equalization system (active equalization). Active equalization transfers energy from one cell or cell block to another without causing large losses. Almost all active systems require a large number of power electronics, such as coils, capacitors, and field effect transistors, as well as corresponding control schemes. This results in additional weight and cost, so there are fewer active equalization systems used in commercial applications. The passive equalization system is realized by a resistor and an electrical switch in parallel with the cell. It is favored for its simplicity and cost advantage. Passive equalization regulates the energy level of a single cell or cell by discharge. The actual evaluation of the commonly used passive equalization techniques will be made below.
There are generally three equalization algorithms: SOC-based equalization algorithms, model-based equalization algorithms, and voltage-based equalization algorithms.
SOC-based equalization algorithm
The SOC-based equalization is the most accurate because theoretically it utilizes all the charges by definition. However, the actual effect of the SOC-based equalization algorithm largely depends on the accuracy of the SOC. The SOC is a state that cannot be directly measured. The SOC usually uses estimation techniques (Kalman filtering, neural networks, etc.) and ampere-time integration. Even though these methods can provide accurate results at the beginning of the battery life cycle, the accuracy will drop rapidly during operation, and the estimation error will often exceed 2-3% (the estimated SOC estimation error for domestic BMS is 5%). The relative parameter variance of the most advanced lithium-ion batteries is far below 1%, which makes it difficult to use inaccurate SOC values as equalization input parameters.
Model-based equalization algorithm
An important feature of lithium-ion batteries is the nonlinear relationship between SOC and open circuit voltage (OCV), as shown in Figure 1. The OCV increases as the SOC increases. This relationship is determined by the potential of the positive and negative materials of the lithium ion battery. When there is a current load, the ohmic resistance, the transfer impedance, and the electric double layer effect cause an overpotential, so the measured terminal voltage (non-OCV) does not directly reflect the actual SOC. Model-based equalization Using the cell model to estimate the overpotential of the cell, it is possible to use the terminal voltage under current load to achieve SOC balance. However, model-based equalization algorithms have the same disadvantages as SOC-based balancing algorithms. The nonlinear nature of lithium-ion batteries makes it difficult to achieve robust and accurate models because all model parameters are in the SOC range and in different temperature ranges. It changes internally, especially during its service life.
Figure 1 OCV curve and equivalent circuit model of NMC/graphite cells
Voltage-based equalization algorithm
The most likely to apply the equalization algorithm is a voltage-based equalization algorithm because each battery system monitors the cell voltage. The cell voltage is usually adjusted by an equalization circuit during the charging process. However, the terminal voltage does not necessarily reflect the SOC, so the voltage balance under load may further deteriorate the imbalance of the battery pack. This imbalance depends on the actual charging current, the overcurrent of the cell, and the slope of the OCV. Reducing the current can alleviate this problem, so the constant voltage (CV) charging phase is suitable for cell balancing. However, since the end-of-charge voltage of the short-plate cells is usually the input of the charger current control, the voltage of the short-plate cells cannot be distorted, so voltage balancing in the CV phase is not recommended.
Due to the improved quality of advanced lithium-ion batteries, the internal parameters change little. In addition, the temperature gradient within the battery pack can be mitigated by specific design measures (thermal management systems). All of the above factors reduce the cell voltage difference, but the cell voltage difference still exists because it is not eliminated. After the end of charging, the overpotential of the cell is removed after a long period of time. At this time, the terminal voltage reflects the SOC of the cell. By discharging the voltage of all the cells to the minimum cell voltage in the battery pack, this process consumes only a small portion. The discharge energy is therefore considered acceptable as long as this voltage balance is not performed frequently.
Lithium-ion batteries exhibit differences in capacity, impedance, and self-discharge rate due to manufacturing tolerances, which results in voltage differences in the battery pack that limit the achievable discharge energy and may deteriorate further in the presence of temperature gradients. In order to maximize the energy content of the battery pack, a passive circuit with a bypass resistor is typically used.
It is recommended to apply voltage equalization during the rest of the battery pack after it is fully charged, during which all cell voltage discharges are adjusted to match the minimum cell voltage in the series battery pack.
If you want to know more information, you can click this link.