As a result of this measurement inaccuracy, the operating range must be restricted by a "guard-band" to ensure that the operating limits are not exceeded. In this example, the operating range must be restricted to the measured range of 22 percent to 78 percent instead of 20 percent to 80 percent. If the pack is expected to maintain the same range, a BMS with this accuracy will require additional battery capacity to compensate for the guard-band restrictions. Considering the 60 percent Usable SOC, the battery must be oversized by >7 percent (Oversize penalty equals 0.6/0.56) to compensate for cell measurement inaccuracy of ±10mV. For an HEV using a 5kWh pack costing $3000 ($600/kWh), this translates to an additional $214.
Guard-Band Requirements for ±10mV Cell Measurement Error
This argument can be extended to highlight the "guard-band penalty" for various cell measurement errors, and its dependence on the SOC range. As show below, a system with only 1mV of measurement error requires less than 1 percent oversizing, even when the pack is restricted to an SOC range of 25 percent to 75 percent (50 percent Usable SOC).
Dependence of Guard-Band on Cell Measurement Error
Although most lithium cells are generally well matched when first acquired, the SOCs of a long string of cells will diverge over time and charge cycling. This is due to small variations in cell characteristics and localized operating conditions, which can lead to small differences in self-discharge and load current. To avoid operating any cell beyond its SOC range, as the SOCs diverge, the total operating range will be slowly constricted by the most unbalanced cells. To address this, nearly all battery management systems include cell balancing.
With passive balancing, cells with higher SOC are discharged to normalize the SOC for all cells. This is a low cost, simple balancing method. However, it has significant limitations: Passive balancing only operates by removing charge. It wastes energy, as a function of the amount of imbalance, and it generates significant heat. This means that the balancing current must be kept relatively small, typically 5 percent or less of the cell capacity. As a result, passive balancing is primarily limited to operation offline and it requires significant time to complete. Passive balancing becomes less effective as the variations in SOC increase, and over time, SOC variation will increase due to diverging cell capacity.
Cells lose capacity as they age, a process that can differ from cell-to-cell due to a number of factors, such as gradients in pack temperature, and variations in cell manufacturing. With differences in capacity, cells will more readily become imbalanced. Allowing just one cell to operate beyond the SOC limit will simply exacerbate this problem by premature cell aging. Relying solely on passive balancing becomes increasingly difficult, as capacities diverge. To address the limitations of passive balancing, new battery management systems are implementing active balancing.
With active balancing, charge is moved between cells, rather than being dissipated with passive balancing. Active balancing can operate both during the charge and discharge cycles. When charging the pack, the active balancer can move charge from the weaker cell to the stronger cell. When discharging the pack, charge can be moved from the stronger cell to compensate the weaker cell. Instead of wasting energy, charge is transferred through a highly efficient circuit, such as a fly back converter. As a result, heat generation is limited, the balancing current is higher, and the balancing time is significantly reduced. This allows for active balancing while the pack is in use, where it can ensure extraction of the maximum capacity for each individual cell. New ICs, such as Linear Technology;s LTC3300 and LT8584, are enabling active balancing in automotive battery packs.
Ideally, active balancing should be enabled as the cells reach the ends of the SOC range.(For maximum efficiency, active balancing should be used when necessary. A system maintained well within the SOC limits, would require much less active balancing than one operating near the limits.) To illustrate, consider a pack containing many cells with uniform capacity, and one "weak" lower capacity cell.