LI-ION BATTERY packs are used in a variety of applications. Fig. 1 shows the varied power requirements for the numerous battery applications in Japan. The large consumer electronics market that has driven the industry's progress appears on the lower left side of Fig. 1. Consumer electronic devices such as cell phones, laptops, and PDAs require relatively small, low-power batteries.
While consumer electronics markets are still growing in Asia, they have begun to stabilize. The major growth areas for batteries in the future electric vehicles and bicycles. Also, there is a large movement to convert power tools and back-up batteries to Li-Ion. Equipment for these, as well as industrial, medical, and military applications requires bigger battery packs — 12-cell packs and larger — for applications such as electric vehicles.
Applications requiring high voltage and capacity are adopting Li-Ion technology because of advantages such as good cycle life and improving price points. But, most applications are looking for the high energy density, small size, and low weight this technology provides.
Fig. 2 compares the energy densities of the various competing rechargeable battery chemistries. Li-Ion cells usually provide an operating voltage of about 3.5 V, but very high voltages require many cells in series. High capacity is achieved with many cells in parallel while high rate capability is achieved with specially designed cells such as the manganese spinel or nano phosphate, or via higher-density battery packs to lower the effective C-rate.
Battery packs can challenge the designer because they are no longer a simple configuration of cells. They are carefully engineered products with many safety features. The battery packs in Fig. 3 are the backup power source for a portable medical device. As illustrated, the main components of a battery pack include:
- Cells, the primary energy source
- PCB provides system intelligence
- Fuel gauge calculates cell capacity
- LEDs indicate pack or cell status
- Serial data communications bus supplies pack status to the host equipment
- Plastic enclosure
- External contacts
Most importantly, the circuit board protects against unsafe operation. The safety circuit protects the pack from external stressors and is required for all Li-Ion battery packs. When overstressed, Li-Ion technology can be hazardous to a degree. Extra caution must be exercised during the design process to ensure that the cells are being used in a manner appropriate to the technology.
Essentially, the safety circuit provides the boundaries in which the cells and pack operate. Safe operation occurs within a window that protects against overcharging and overvoltage, and this window is reflected in the settings of the safety circuit. For example, the highest voltage usually allowed by the safety circuit is 4.3 V, while the charger's constant-voltage taper is at 4.2 V. There to handle over-discharge, the safety circuit is set just slightly lower than the device's normal operating cutoff, which is usually 3 V per cell. Short-circuit or overcurrent protection is also necessary, and the devices are typically rated to withstand extreme operating temperatures.
Safety circuits are available off the shelf. They are appropriate for most consumer applications that operate at room temperature and require relatively low currents; however, issues associated with cell imbalance and other problems can be readily apparent in larger battery arrays. Generally, if an application employs high current or more than 12 cells, it requires a custom solution for safety and cell balancing.
CHALLENGES FOR LARGE BATTERY ARRAYS
Fig. 4 is an example of a large battery array for a portable application that consists of 42 cylindrical Li-Ion cells. Each 18-mm-diameter, 65-mm-long cell yields 2.6 Ampere-hours (Ah) of capacity. This battery pack is an example of one that might replace a sealed lead-acid back-up battery.
The electric vehicle market is driving battery technology to even larger forms. The Tesla roadster's battery in Fig. 5 is one of the first electric vehicles on the market using Li-Ion technology. To power the car — which accelerates from 0 to 60 mph in less than 4 s and tops out at about 125 mph with a range of about 220 miles — Tesla had to design a custom battery solution. This microprocessor-controlled battery comprises almost 7,000 individual cells and weighs nearly 1,000 pounds. There are 11 modules, each with 9 cells connected in series and 69 cells connected in parallel to produce 375 V and 142 Ah. This modular approach is common for electric vehicles.
There are many challenges for large battery designs. The first to be considered are the practical issues of shipping and vendor support. Many cell vendors do not want their products used in multi-cell packs. This is especially true for prismatic cells, and most vendors limit the size of prismatic packs to three or four cells.
Shipping regulations are also an issue for large batteries in general. Three categories of batteries are defined in the U.S. DOT's rule based on their size, or equivalent lithium content (ELC). ELC is calculated in grams on a per-cell basis to be 0.3 times the rated capacity in Ah. Thus, the ELC for a battery pack is the rated capacity in Ah for a single cell multiplied by 0.3 and then multiplied by the number of cells in the battery pack. Small Li-Ion battery packs that have passed the UL testing requirements, including batteries packed with or installed in equipment, can be transported without restrictions. The battery packs do not have to be shipped as fully regulated Class 9 hazardous materials. Medium-size batteries can be shipped unregulated only by ground, but air shipment requires that they are classified as hazardous material. Larger Li-Ion packs must always be shipped as fully regulated Class 9 hazardous materials.
There are also design challenges for packs with many cells in parallel, i.e. high-capacity packs. High-current circuit design is non-standard and diodes are used between the cells. The issue of diode placement in the parallel string must be addressed and decided by the electronics designer. Fuel gauge limitations are also likely to be encountered. Off-the-shelf solutions often don't exist for high-current applications. The few solutions available use outdated coulomb counting technology. Some solutions exist or are in development for many of these problems.
Bigger cells and modules are in development and will offer the advantage of providing an off-the-shelf built-in solution to some problems, like thermal management. However, they are typically built around a specific application and may not adequately meet everyone's needs. For example, a module from Electrovaya consisting of pouched prismatic cells delivers 1.5 kWh, and is equivalent in capacity to 166 standard cobalt 18650s or 356 nano iron phosphate 18650s. The scale-up to a larger system, such as those required for an electric vehicle, is easier when you start with these larger modules. Heat sinks and active cooling can be used during charging and discharging for thermal management, and large ICs can be used to accommodate the current.
To deliver a given wattage, high series cell counts or high voltage is more effective than high parallel cell count. Therefore, high series design challenges are more prevalent. While thermal management issues are similar, balancing paralleled cells to ensure they all supply the same voltage and current is more challenging. Pack reliability and cycle life can be compromised by cells going out of balance. The pack will perform to the lowest-common-denominator cell, and cell imbalance will grow over multiple charge and discharge cycles. Self discharge, especially due to uneven heating, will exacerbate the problem.
High-voltage chemistries are sometimes explored as an alternative. Commercially available cells deliver 3.3 to 3.7 V. The potential difference between lithium-based anode material and the cathode's oxide insertion material determines the cell's voltage. Unfortunately, the list of potential candidates is limited to delivering maximum charge voltages of about 5 V, far from the hundreds of volts necessary for an electric vehicle or even the 14 V for a motor. This is because the electrolyte has a fundamental operating window. If the voltage difference is greater than this window, the electrolyte will decompose.
To conquer the challenge of designing a high-voltage pack the problem of cell imbalance must be understood. Poor cell-capacity matching can cripple a battery pack right off the manufacturing floor. First, cells must be matched so that the pack does not start out of balance. A quality battery-pack manufacturer will test all incoming lots and should reject lots of cells with high variance. Impedance and chemical efficiency variations in the cells can have a similar effect to poor capacity matching.
Packs should never be assembled using more than one manufacturer or even more than one lot. Regardless, heat can cause the pack to unbalance over time even if cells are well matched from the start. Non-uniform thermal stress is a common problem often due to poor design of the host device's batteries. Because self-discharge doubles for each 10°C rise in temperature, even heat from a microprocessor can cause radical differences in self-discharge across a multi-cell battery pack. Non-uniform electrical loading of the pack causes the same uneven discharge and high discharge rates can exacerbate all of these issues.
TEMPERATURE AND CELL BALANCING
Temperature becomes a great factor in very large battery arrays simply because the gradients are larger, and also the physical cell arrangement can influence the temperature gradients and pack effects. Active cooling may not evenly affect the cells, so careful pack design is important. Fig. 6 is an experimental setup featuring a battery pack with four cells in series and six in parallel. Fig. 7 is a real thermal image of the battery pack in operation.
Clearly, the cells are heating unevenly; some cells remain at 40° or 45°C while the hottest cells are about 60° to 65°C. While these are safe operating temperatures for all the cells, the 20° difference can quadruple the self-discharge rate, making an active balancing system necessary. Another feature to notice in this thermal image is the connector to the right of the pack. It is white-hot and well over 70°C. Obviously, connector design and placement is an important consideration when a pack is delivering significant current as it can compound the problems.
Pack design should minimize the gradients to which cells are exposed, but this may not be sufficient in very large packs. The solution, then, is to employ cell balancing. There are a couple of strategies to implement cell balancing, but their ultimate purpose is the same: deliver as much energy during discharge as possible and extend the cycle life of the battery pack by minimizing the difference in energy stored in each cell. The techniques fall into two basic categories: bypass or active redistribution.
Also known as bleed-balancing, the bypass technique provides an alternative current path to a cell that is out of balance with other cells in series. This is the traditional, simple technique and is the least expensive for low current.
Bleeding off excess energy represents a fundamental tradeoff between energy conservation in the long run and energy delivered. It is recommended to balance during charge cycle, and there are duty-cycle limitations; the amount of energy moved is limited to by time, temperature, and current. There is a trade-off in the cost of high-current resistors and low-resistance FETs, and the technique is thermally challenging at high-temperature portions of pack life.
A newer strategy is active balancing, or charge redistribution, which actually moves charge from higher-charged cells to lower-charged cells in series. It is basically a method for energy transfer between adjacent cells and the circuitry moves energy where and when it's needed to minimize global imbalance. The current path is outside of the charge and discharge path. Unlike the bypass strategy, active balancing can be implemented during charge, idle, and discharge periods.
There are a couple of topology choices: capacitive and inductive. For capacitive, there is a switch capacitor, from higher cells to lower cells, that requires a simple higher-voltage to lower-voltage measurement and shuttle of energy. Unfortunately, this technique only works during times of peak voltage, at the end of the cycle. There is a maximum 50% efficiency.
The inductive method stores energy from the higher cell before delivering it to the lower cell. A FET capacitor and inductor create a small dc-dc boost-converter and bi-directional energy transfers efficiently between adjacent cells.
Redistribution is allowed anywhere in pack. The system moves energy where and when it is needed to minimize global imbalance and is not as efficiency challenged at mid-capacity levels. The downside of the inductive method is that it has a higher part count and cost.
Large battery arrays represent unique challenges for the pack designer. Yet, they enable new markets to employ lighter, smaller, and more efficient Li-Ion technology. Fortunately, the huge potential in this market has prompted the development of new, innovative solutions. The most recent and meaningful attacks the issue of cell imbalance due to thermal and electric gradients.