In order to get the most out of the battery and to ensure safe operation, current flow in and out of the cells that make up that battery must be carefully monitored and controlled. A battery management system (BMS) serves this function. Generally, we have no means of looking inside a cell or module after it is built. Instead, the BMS uses measurements of current, potential, and temperature to control charging and discharging, and to estimate the SOC and SOH. Additionally, the BMS must communicate with other systems that interact with the battery pack. It must also provide the electrical hardware and software to accomplish cell balancing. Our purpose in this chapter is not to examine the architecture of the BMS and describe how it works; these topics are beyond the scope of this text. Here, we are concerned with the electrochemical processes that drive the need for a battery management system.
The importance of and the sophistication required in the BMS depend on the cell chemistry, the size of the battery, and the application. With regard to cell chemistry, it is critical to understand the consequences of overcharging and overdischarging a cell. As you know, lithium-ion cells cannot be overcharged without incurring damage, whereas overcharging can help to prolong the life of lead–acid cells. An important function of the BMS is to manage overcharging and overdischarging of individual cells.
One of the most important issues in the management of a multicell battery is keeping the individual cells in balance. Cell balance is another way of stating that, to the extent possible, the SOC of all of the cells should be the same. Despite our best efforts, the cells in a string will not behave identically. There can be variations from the manufacturing process that introduce small differences in the cell resistance or in its capacity. Even if these nonuniformities are eliminated, cells assembled into a battery can experience different temperatures, depending on their location in the pack. Cells can age differently too. Although the same current passes through each cell in series, this does not ensure that the state-of-charge of a given cell remains synchronized with that of the other cells in the string. Cells will not have exactly the same coulombic efficiency, which will affect the usable capacity of the battery. For example, every time the pack is charged, cells with lower coulombic efficiencies will not charge to the same extent as other cells in the pack. In spite of their poorer charge performance, these weak cells are likely to be discharged to the same extent as the other cells. This means that their SOC relative to the other cells drops each cycle. Eventually, the SOC of the weak cells will drop to zero. Prior to that, the cell is likely to be damaged, perhaps irreversibly. Damage of other cells in the string is also possible. Illustration 8.3 demonstrates the impact of lower charge efficiency on the SOC.
ILLUSTRATION 8.3
A string of 1 A·h NiCd batteries are cycled between 30 and 80% SOC. If the nominal current efficiency for charging is 80%, but one weak cell has a coulombic efficiency of only 70%, what happens to the SOC of the weak cell with cycling?
The coulombs needed to restore the SOC of the nominal cells is
The weak cell receives the same number of coulombs, but a smaller fraction goes to restoring its SOC. Instead of 0.5, the change in SOC of the weak cell is
Thus, when discharged 0.5 A·h and subsequently charged, the new SOC of the weak cell is 0.3 + 0.44 = 0.74. After just one cycle, this cell only reaches 74% SOC. If the cycling continues discharging 0.5 A·h but only restoring with 0.44 A·h, this weak cell will quickly reach 0% SOC during discharge as illustrated in the figure. At this point, not only does the cell fail to contribute power, it actually consumes power and can be damaged.
A similar imbalance situation arises if cells have slightly different rates of self-discharge (see Problem 8.13). The approach to dealing with these imbalances depends on the cell chemistry and, in particular, the tolerance of the cells for overcharge. For chemistries that can tolerate overcharge, the individual cells can be brought to nearly the same SOC by overcharging the battery or pack. Although not energy efficient, this approach is the simplest one to implement and does not require continuous monitoring of individual cell potentials.
For cells, such as lithium-ion cells, that cannot tolerate any overcharge, strict monitoring of the potential of individual cells is required. Fortunately, rates of self-discharge are low in lithium-ion cells and current efficiencies are near unity, reducing the rate at which imbalances take place; nonetheless, over time imbalances will arise. For a string of cells, the useable capacity of the battery is controlled by the capability of the weakest cell. To avoid the condition shown in Illustration 8.3, the cells must be balanced. Since overcharging is not an option, special circuitry within the BMS is required to individually balance the charge in every cell. As we saw previously, during charge the potential is held at the cutoff value until the current decreases to a set (low) value (∼C/20). In a string of cells, the current is constant, but the voltage will vary slightly from cell to cell. In passive balancing, energy is dissipated in a shunt resistor for cells with excess SOC. In other words, to avoid having the potential of any cell exceed the prescribed voltage limit, some current bypasses the cell through a shunt resistor. An alternative to passive balancing is active balancing. Here, energy from a cell with a high SOC is moved to a cell with a lower SOC. Active balancing requires more complicated electrical circuitry, but has greater energy efficiency.
Knowledge of the SOC is an important aspect of cell balancing. There are two approaches for determining the SOC. The first means is to simply count the coulombs passed relative to a known condition. In doing so, current is taken as positive during charge and negative during discharge.
where SOC(t0) represents the known condition. Let’s consider some challenges with this approach. First, this method is only accurate if the current efficiency is close to one, although it can be corrected with use of an efficiency, if known. The SOC calculated by Equation 8.25 will also not be accurate if there are appreciable rates of self-discharge. Finally, since the capacity of the battery changes with time due to aging, the capacity used in Equation 8.25 should be the capacity at time t0. Temperature also influences capacity and must be considered in determining the SOC.
A second alternative to SOC determination is to measure the open-circuit potential of the cell or battery. With some chemistries, the potential change with SOC is minimal (Figure 7.4); with others it is substantial. When there is an appreciable change in the open-circuit potential with SOC, measuring the voltage of the cell is a quick way of estimating its SOC. Often, algorithms are used to combine these two approaches to estimate the SOC.
Finally, we close with a couple of comments about temperature. At either excessively high or low temperatures, the current in or out of the battery may need to be restricted by the BMS in order to avoid damage. Concerns include the possibility for thermal runaway at high temperatures, and the possibility of cell damage at temperatures that are too low for acceptable operation of the battery. Often individual lithium-ion cells have a positive temperature coefficient (PTC) current-limiting device that acts as a fuse to prevent thermal runaway. Thermal management is an important aspect of battery operation and will be considered in more depth in the next section.
Leave a Reply