How active and passive cell balancing works

In a Power Systems Design article, “Active and passive balancing for battery management systems (North America > December 2015 > Page 21),” Stefano Zanella described how a multicell system becomes unbalanced. In this post, I would like to explore how batteries become unusable if they’re not balanced and expand a bit on the effects of battery-capacity mismatch. I will focus on automotive lithium-ion (Li-ion) batteries, but in general these principles apply to all batteries.

Multicell batteries are often built as an array of cells, in series and or in parallel. A higher number of cells in series will lead to higher battery-pack voltages, while more cells in parallel will lead to higher overall battery capacity (expressed as ampere-hour rating, or Ahrs). The battery capacity will then dictate the number of cells in parallel, by the capacity being equal to the number of cells in parallel times the cell capacity required for the system to run. Automotive vehicles tend to use 96 Li-ion cells in series and 24 cells in parallel, depending on cell type. For instance, an electric vehicle with a range of 100 miles will need a battery from 20-30kWh, depending on the weight of the vehicle, the anticipated usage profile and the efficiency of the various systems in the car. Several aspects of the system will dictate the battery-pack voltage, including the overall size and type of the electric motor, cable size, and isolation requirements.

Multicell batteries charge by supplying current to the positive terminal of the cell on top of the stack. (Assume that the battery comprises n cells in series.) In other words, the cells of the battery do not charge individually. If you read Stefano’s article, you already know that at the end of a charge, the amount of charge left in each cell is different; and as you repeatedly charge and discharge the battery (in the absence of balancing), this difference increases. This animation below shows this process.

Figure 1: How active and passive cell balancing works

If you imagine the two cells in Figure 1 as identical containers of charge, driving an electric vehicle will result in extracting energy from the battery, which will deplete those containers. Charging an electric vehicle injects charge into the battery, thereby filling those containers. Not all cells are identical to one another, nor do they age evenly; therefore, the weaker cells will charge and discharge at a slightly different rate. The voltage level of each cell will slowly rise and fall as the cells charge and discharge, respectively.

Let’s start from a full battery. All of the energy (usable energy) contained in the cells can power a car. To not overdischarge the cells (because overdischarging reduces cell lifetimes and can impact safety), when the first cell reaches the undervoltage threshold (plus a safety margin that often depends on the protector), discharging must stop. To not overcharge Li-ion cells, when the first cell reaches the overvoltage threshold, charging must stop. The cells that lag, however, are not fully charged yet, leaving some energy in the battery that cannot be used for driving because, again, when the first cell is full, charging must stop.

In other words, after the first charge/discharge cycle, some energy is stranded in the pack. It can never be used to power the car.

As the battery charges and discharges over and over, the stranded energy increases, thus decreasing the usable energy. Plus, the loss of usable energy is twice the stranded energy because the stranded energy isn’t usable and an equivalent amount of charge cannot be injected into the other cell.

After enough charge and discharge cycles, the usable energy starts to approach zero. How do you avoid this problem? Balancing! You can achieve cell balancing by dissipating the stranded energy onto a resistor, regaining the ability to top off the cells and reach full charge.

As long as all cells have the same capacity, complete balancing is not necessary at the end of each and every charge cycle – because the effects of charge imbalance are fully reversible. I have seen a case during battery electronics development where the passive-balancing portion of a battery was not implemented until after a number of charge/discharge cycles. When the balancing system was ready, the usable energy had diminished by more than 25%. However, after balancing all of the cells, the pack could be fully charged with only a minimal loss of usable energy.

You should choose the amount of balancing current based on the application and thermal considerations. For example, in a 24kWh system (96 cells in series), assuming that the cells have less than a 1% charge-time difference at their end of life (differences in charge times increase over time), a 66Ah system will need to compensate for 660mAh. With a balancing current of 200mA, you could balance this system in 3.3 hours, but it would take twice as much time to balance with a 100mA current.

Table 1: Application specific monitoring and protection devices 

If you’re looking to start balancing your battery’s cells, Table 1 matches the perfect monitoring and protection device with your application. If your application is not included in this table or you have questions about your current design, connect with a TI battery expert on the TI E2E battery management forum.  

Additional resources:

• Get more information on all of TI’s battery management products at www.ti.com/battery.
• Read the blog: From millivolts to miles: How the performance of battery-management ICs affects your car’s performance